nutritional and environmental factors regulating
TRANSCRIPT
The Pennsylvania State University
The Graduate School
Department of Animal Science
NUTRITIONAL AND ENVIRONMENTAL FACTORS REGULATING
BIOLOGICAL RHYTHMS OF MILK SYNTHESIS IN DAIRY CATTLE
A Dissertation in
Animal Science
by
Isaac J. Salfer
2019 Isaac J. Salfer
Submitted in Partial Fulfillment
of the Requirements
for the Degree of
Doctor of Philosophy
August 2019
ii
The dissertation of Isaac J. Salfer was reviewed and approved* by the following
Kevin J. Harvatine
Associate Professor of Nutritional Physiology
Dissertation Advisor
Chair of Committee
Paul A. Bartell
Associate Professor of Avian Biology
W. Burt Staniar
Associate Professor of Equine Science
Connie J. Rogers
Associate Professor of Nutritional Immunology
Terry D. Etherton
Distinguished Professor of Animal Nutrition
Graduate Program Head – Animal Science Department
*Signatures are on file in the Graduate School
iii
ABSTRACT
Dairy cows exhibit well-characterized daily patterns of feed intake and milk synthesis.
These patterns may represent circadian rhythms, endogenous repeating cycles of approximately
24 h that govern most physiological processes. Circadian rhythms are controlled by a set of
‘clock’ transcription factors that oscillate over 24 h and govern gene expression across the day.
There is an increasing body of evidence demonstrating a role of nutrient intake in modifying
circadian rhythms in peripheral tissues. However, this effect has not been well-characterized in
dairy cattle. In addition to a daily rhythms of milk synthesis, dairy cows display seasonal changes
of milk production that are not well understood. The objective of this dissertation was to examine
the factors affecting daily and annual rhythms of milk synthesis. Specifically, we wanted to
examine the effects of nutrient intake on the daily rhythms of milk synthesis, and the relationships
between cow-level and environmental factors and the annual rhythms of milk production.
To accomplish these objectives, 6 experiments were conducted, with the first 4
examining the role of nutrient intake on the daily rhythms of milk synthesis, and the final 2
characterizing factors influencing annual rhythms of milk production. First, the effect of time-
restricted feeding on the daily patterns of milk synthesis and plasma hormones and metabolites
was examined in an experiment where feed was temporally restricted to cows for 16 h/d either
during the day (0700 to 2300 h) or at night (1900 to 1100 h). This experiment demonstrated that
night restricted feeding shifted the peak of milk yield from morning to evening, while shifting the
peak of milk fat and protein concentration from evening to morning. Moreover, the daily rhythms
of plasma glucose, nonesterified fatty acids, plasma urea nitrogen, and insulin were shifted about
8 h by night-restricted feeding. The objective of the second experiment was to investigate if the
changes in daily patterns of milk synthesis during time-restricted feeding were caused by changes
in the molecular circadian clock of the mammary gland. This experiment compared a control
iv
group with feed available continuously to cows under night-restricted feeding (feed available for
16 h/d from 2000 h to 1200 h). Results demonstrated that night-restricted feeding altered the daily
rhythms of expression of clock genes circadian locomoter output cycles kaput, cryptochrome 1
and REV-ERBα, suggesting that nutrients entrain the molecular circadian clock of the mammary
gland. The final two experiments examined the role of specific nutrients on the daily rhythms of
milk synthesis. In one experiment, a high C18:1 oil was abomasally infused either 24 h/d or for 8
h during the day (0900 to 1700 h) or the night (from 2100 to 0500). Day-infusion increased the
amplitudes of milk and milk fat and protein yield while reducing the amplitudes of milk fat and
protein concentration. However, treatment had little influence on the phase of the rhythms,
suggesting that the timing of fatty acid infusion did not entrain the mammary clock. In the other
experiment, sodium caseinate solution was abomasally infused either 24 h/d or for 8 h during the
day (0900 to 1700 h) or the night (from 2100 to 0500). Night-infusion induced a daily rhythm of
milk yield, but no rhythm was present during continuous or daytime infusion. Infusion of sodium
caseinate during the day reduced the amplitude of fat concentration and increased the amplitude
of protein concentration, while night-infusion decreased the amplitude of protein concentration.
Treatment had little effect on the time at peak of the rhythm, suggesting that the mammary
circadian clock was not entrained to the time of protein infusion. Furthermore, night-infusion
shifted the peak of the nonesterified fatty acid rhythm from morning to evening.
The final 2 experiments characterized factors influencing the annual rhythms of milk
synthesis. First, annual rhythms of milk fat and protein concentration were compared among
regions of the United States using monthly averages of Federal Milk Marketing Orders. The
annual rhythms had greater amplitudes in northern regions compared to southern regions
suggesting that they are related to photoperiod. In the same report, the effect of cow-level factors
on annual rhythms of milk, fat, and protein yield, and milk fat and protein concentration were
examined using data from 11 dairy herds in Pennsylvania. Rhythms were consistent regardless of
v
herd, lactation number, and genotype, suggesting that they are highly conserved among individual
cows. Finally, annual rhythms of milk, fat and protein yield and milk fat and protein
concentration were studied using dataset obtained from Dairy Records Management Systems and,
relationships between annual rhythms of production and daylength, change in daylength, and
environmental temperature were determined. Annual rhythms differed between northern and
southern states, with northern states having greater amplitudes of milk fat and protein
concentration and southern states having greater amplitudes of milk yield. Moreover, the rhythms
of milk fat and protein concentration were highly correlated with absolute daylength, while the
rhythm of milk yield was highly correlated with the change in daylength, suggesting that they are
controlled through two separate oscillatory mechanisms.
In conclusion, the daily rhythms of milk synthesis are modified by the time of feeding
and this effect appears mediated by the molecular circadian clock. The timing of both fatty acids
and protein affect the robustness of milk synthesis rhythms, without large effects on their time of
peak. Finally, yearly patterns of milk production appear to constitute an annual rhythm, with the
rhythms of fat and protein concentration being entrained by daylength and the rhythm of milk
yield being entrained by the change in daylength.
Keywords: Biological rhythm, milk synthesis, dairy cattle, food entrainment
vi
TABLE OF CONTENTS
List of Figures .................................................................................................................... xi
List of Tables ................................................................................................................... xiii
List of Supplemental Figures ........................................................................................... xiv
List of Supplemental Tables ............................................................................................. xv
List of Abbreviations ....................................................................................................... xvi
Acknowledgements .......................................................................................................... xix
Chapter 1 Introduction ........................................................................................................ 1
Chapter 2 Literature Review ............................................................................................... 5
Biological Rhythms ......................................................................................................... 5
Molecular Structure of Biological Clocks ....................................................................... 7
Transcriptional-Translational Feedback Loops ........................................................... 7
Post-Transcriptional Modifications ............................................................................. 9
Post-Translational Modifications .............................................................................. 10
Epigenetics and the Clock ......................................................................................... 11
Entrainment of Biological Clocks ................................................................................. 11
Light Entrainment ...................................................................................................... 12
Food Entrainment ...................................................................................................... 13
Temperature Entrainment .......................................................................................... 15
Circadian Rhythms and Nutrient Metabolism ............................................................... 16
Cellular Energy and Redox Status and the Circadian Clock ..................................... 16
Glucose Metabolism and the Clock ........................................................................... 18
Lipid Metabolism and the Clock ............................................................................... 20
Circadian Rhythms in Microorganisms ..................................................................... 22
Annual Rhythms ............................................................................................................ 25
Annual Rhythms and Reproduction .......................................................................... 26
Seasonal Changes in Feeding Behavior and Metabolism .......................................... 28
Seasonal Changes in Immunity ................................................................................. 31
Epigenetic Control of Seasonality ............................................................................. 34
vii
Conclusions ................................................................................................................... 35
Figures ........................................................................................................................... 36
Chapter 3 The effects of day- versus night-restricted feeding of dairy cows on the daily
rhythms of feed intake, milk synthesis and plasma metabolites ....................................... 39
ABSTRACT .................................................................................................................. 40
INTRODUCTION ......................................................................................................... 41
MATERIALS & METHODS........................................................................................ 42
Animals and Treatments ............................................................................................ 42
Milk Sampling and Analysis ..................................................................................... 43
Feeding Behavior Observation and Analysis ............................................................ 44
Plasma Sampling and Analysis.................................................................................. 44
Body Temperature Recording ................................................................................... 45
Statistical Analysis .................................................................................................... 45
RESULTS...................................................................................................................... 46
Eating Behavior ......................................................................................................... 46
Rhythm of Milk and Milk Components .................................................................... 47
Milk FA Yield and Profile ......................................................................................... 48
Plasma Hormones and Metabolites ........................................................................... 49
Body Temperature ..................................................................................................... 50
DISCUSSION ............................................................................................................... 50
FIGURES ...................................................................................................................... 55
TABLES ........................................................................................................................ 60
SUPPLEMENTAL FIGURES ...................................................................................... 61
SUPPLEMENTAL TABLES ........................................................................................ 62
Chapter 4 The effects of night-restricted feeding on the molecular clock of the mammary
gland in lactating dairy cows ............................................................................................ 65
ABSTRACT .................................................................................................................. 66
INTRODUCTION ......................................................................................................... 68
MATERIALS AND METHODS .................................................................................. 70
Animals and Treatments ............................................................................................ 70
Milk Sampling and Analysis ..................................................................................... 71
Mammary Tissue Collection ..................................................................................... 71
viii
Plasma Sampling and Analysis.................................................................................. 72
Gene Expression Analysis ......................................................................................... 73
Statistical Analysis .................................................................................................... 73
RESULTS AND DISCUSSION ................................................................................... 74
Dry Matter Intake ...................................................................................................... 74
Milk Yield, Milk Composition, and Milk Fatty Acid Yields .................................... 75
Molecular Clock Gene Expression ............................................................................ 76
Plasma Hormones and Metabolites ........................................................................... 78
CONCLUSIONS ........................................................................................................... 80
FIGURES ...................................................................................................................... 82
TABLES ........................................................................................................................ 87
SUPPLEMENTAL TABLES ........................................................................................ 88
Chapter 5 The effects of time of fatty acid infusion on the daily rhythms of milk synthesis
and plasma hormones and metabolites in dairy cows ....................................................... 91
ABSTRACT .................................................................................................................. 92
INTRODUCTION ......................................................................................................... 94
MATERIALS & METHODS........................................................................................ 95
Animals and Treatments ............................................................................................ 96
Milk Sampling and Analysis ..................................................................................... 96
Plasma Sampling and Analysis.................................................................................. 97
Body Temperature Analysis ...................................................................................... 98
Statistical Analysis .................................................................................................... 98
RESULTS AND DISCUSSION ................................................................................... 99
Milk and Milk Components Synthesis ...................................................................... 99
Daily Yields and Daily Rhythms of Milk Fatty Acids ............................................ 101
Daily Rhythms of Plasma Metabolites .................................................................... 103
Daily Rhythm of Body Temperature ....................................................................... 105
CONCLUSIONS ......................................................................................................... 106
FIGURES .................................................................................................................... 108
SUPPLEMENTAL FIGURES .................................................................................... 116
SUPPLEMENTAL TABLES ...................................................................................... 117
ix
Chapter 6 The effect of timing of protein infusion on the daily rhythms of milk
production and plasma hormones and metabolites ......................................................... 119
ABSTRACT ................................................................................................................ 120
INTRODUCTION ....................................................................................................... 122
MATERIALS & METHODS...................................................................................... 124
Animals and Treatments .......................................................................................... 124
Milk Sampling and Analysis ................................................................................... 125
Plasma Sampling and Analysis................................................................................ 125
Statistical Analysis .................................................................................................. 126
RESULTS AND DISCUSSION ................................................................................. 127
Daily Milk Yield and Milk Components ................................................................. 127
Daily Rhythms of Milk Yield and Milk Components ............................................. 128
Daily Yields and Daily Rhythms of Milk Fatty Acids ............................................ 130
Daily Rhythms of Plasma Metabolites .................................................................... 131
CONCLUSIONS ......................................................................................................... 135
FIGURES .................................................................................................................... 136
SUPPLEMENTAL FIGURES .................................................................................... 142
SUPPLEMENTAL TABLES ...................................................................................... 143
Chapter 7 Annual rhythms of milk and milk fat and protein production in dairy cattle in
the United States ............................................................................................................. 145
INTRODUCTION ....................................................................................................... 148
MATERIALS & METHODS...................................................................................... 149
USDA Milk Market Data ........................................................................................ 149
Cow-Level Data ....................................................................................................... 150
RESULTS.................................................................................................................... 151
USDA Milk Market Data ........................................................................................ 151
Cow-Level Data ....................................................................................................... 153
DISCUSSION ............................................................................................................. 156
CONCLUSIONS ......................................................................................................... 163
FIGURES .................................................................................................................... 164
TABLES ...................................................................................................................... 169
SUPPLEMENTAL FIGURES .................................................................................... 170
x
Chapter 8 Annual rhythms of milk synthesis in dairy herds in four regions of the United
States and their relationships to environmental predictors. ............................................ 175
ABSTRACT ................................................................................................................ 176
INTRODUCTION ....................................................................................................... 178
MATERIALS & METHODS...................................................................................... 179
Data Collection ........................................................................................................ 179
Cosinor rhythmometry ............................................................................................. 180
Effect of breed on annual rhythms of production .................................................... 180
Comparison of annual rhythm with temperature ..................................................... 181
Relationships among environmental variables and milk production responses ...... 181
RESULTS.................................................................................................................... 182
Annual Rhythms within Selected States .................................................................. 182
Annual Rhythms among Breeds .............................................................................. 183
Comparison of Annual Rhythm and Temperature Models ..................................... 185
Relationships among Environmental Variables and Milk Production .................... 186
Regression Equations to Adjust for Seasonal Changes in Production .................... 187
DISCUSSION ............................................................................................................. 187
CONCLUSIONS ......................................................................................................... 193
FIGURES .................................................................................................................... 195
TABLES ...................................................................................................................... 198
SUPPLEMENTAL TABLES ...................................................................................... 201
Chapter 9 Integrative Discussion .................................................................................... 205
References ....................................................................................................................... 212
xi
LIST OF FIGURES
Figure 2-1. Parameters used in the quantification of biological rhythms. ....................... 36
Figure 2-2. Structure of the mammalian transcriptional circadian clock. ........................ 37
Figure 2-3. Physiological mechanisms governing seasonal reproduction. ...................... 38
Figure 3-1. Effect of day- versus night-restricted feeding on the daily pattern of feed
intake. ................................................................................................................................ 55
Figure 3-2. Effect of day- versus night-restricted feeding on daily rhythms of milk yield
and milk components. ....................................................................................................... 56
Figure 3-3. Effect of day- versus night-restricted feeding on the daily production and
daily pattern of milk fatty acids. ....................................................................................... 57
Figure 3-4. Effect of day- versus night-restricted feeding on daily rhythms of plasma
hormones and metabolites................................................................................................. 58
Figure 3-5. Effect of day- versus night- restricted feeding on the circadian rhythm of
body temperature in dairy cows. ....................................................................................... 59
Figure 4-1. Schedule of feeding, lighting and sampling during the experiment. ............. 82
Figure 4-2. Effect of night-restricted feeding on average daily feed intake and milk
synthesis. ........................................................................................................................... 83
Figure 4-3. Effect of night-restricted feeding on milk fatty acid yields by origin. .......... 84
Figure 4-4. Effect of night-restricted feeding on the daily patterns of expression of
molecular clock genes in the mammary gland. ................................................................. 85
Figure 4-5. Effect of night-restricted feeding on the daily patterns of plasma metabolites.
........................................................................................................................................... 86
Figure 5-1. Schedule of feeding, lighting and sampling during the experiment. ........... 108
Figure 5-2. Effect of time of fatty acid infusion on dry matter intake. .......................... 109
Figure 5-3. Effect of time of fatty acid infusion on daily milk, fat and protein yield and
fat and protein concentration. ......................................................................................... 110
Figure 5-4. The effects of time of fatty acid infusion on the daily rhythms of milk
synthesis. ......................................................................................................................... 111
Figure 5-5. Effect of time of fatty acid infusion on the daily yields of fatty acids. ....... 112
Figure 5-6. Effect of time of fatty acid infusion on daily rhythms of fatty acids by source.
......................................................................................................................................... 113
Figure 5-7. The effects of time of fatty acid infusion on daily rhythms of plasma
metabolites. ..................................................................................................................... 114
Figure 5-8. The effect of the time of fatty acid infusion on the daily rhythms of core body
temperature. .................................................................................................................... 115
Figure 6-1. Schedule of feeding, lighting and sampling during the experiment. ........... 136
Figure 6-2. Effect of time of protein infusion on dry matter intake............................... 137
Figure 6-3. Effect of time of protein infusion on daily milk, fat and protein yield and fat
and protein concentration. ............................................................................................... 138
xii
Figure 6-4. The effects of time of protein infusion on daily rhythms of milk synthesis.
......................................................................................................................................... 139
Figure 6-5. Effect of time of protein infusion on daily yields and daily rhythms of milk
FA by source. .................................................................................................................. 140
Figure 6-6. The effects of time of protein infusion on daily rhythms of plasma
metabolites. ..................................................................................................................... 141
Figure 7-1. Map of United States Federal Milk Marketing Orders. .............................. 164
Figure 7-2. Annual rhythms of fat concentration in selected Federal Milk Market orders.
......................................................................................................................................... 165
Figure 7-3. Annual rhythms of protein concentration in selected Federal Milk Market
orders............................................................................................................................... 166
Figure 7-4. The effect of parity on annual rhythms of milk yield and composition. ..... 167
Figure 7-5. The effect of DGAT1 K232A polymorphism on annual rhythms of milk and
milk fat concentration and yield. .................................................................................... 168
Figure 8-1. Annual rhythms of milk, fat and protein yield and fat and protein
concentration in Holstein dairy herds in Florida (FL), Minnesota (MN), Pennsylvania
(PA), and Texas (TX). .................................................................................................... 195
Figure 8-2. Annual rhythms of milk and milk fat and protein yield and milk fat and
protein concentration by breed........................................................................................ 196
Figure 8-3. Relationship among environmental predictors and annual rhythms of milk
and milk fat and protein yield and milk fat and protein concentration by breed. ........... 197
xiii
LIST OF TABLES
Table 3-1. Effect of day versus night-restricted feeding on total daily DMI, milk yield,
and milk composition. ....................................................................................................... 60
Table 4-1. Bovine primers used to quantify mammary gene expression with RT-qPCR. 87
Table 7-1. The acrophase and amplitude of a cosine function with a 12-month period fit
to the regional monthly averages of milk fat and protein concentration. ....................... 169
Table 8-1. Comparison between models testing the fit of a cosine function versus
maximum temperature. ................................................................................................... 198
Table 8-2. Equations for cosine regression equations for milk and milk fat and protein
yield and milk fat and protein concentration for Florida (FL), Minnesota (MN),
Pennsylvania (PA), and Texas (TX). .............................................................................. 199
Table 8-3. A summary of milk yield and milk fat and protein concentration responses in
experiments testing the effects of heat stress in environmental chambers. .................... 200
xiv
LIST OF SUPPLEMENTAL FIGURES
Supplemental Figure S3-1. Effect of the day versus night-restricted feeding on feeding
behavior............................................................................................................................. 61
Supplemental Figure S5-1. Effect of time of fatty acid infusion on milk urea nitrogen
concentration. .................................................................................................................. 116
Supplemental Figure S 6-1. Effect of time of protein infusion on lactose yield. ......... 142
Supplemental Figure S7-1. Annual rhythms of fat concentration in Federal Milk Market
orders............................................................................................................................... 170
Supplemental Figure S7-2. Annual rhythms of protein concentration in Federal Milk
Market orders. ................................................................................................................. 171
Supplemental Figure S7-3. Variability in acrophase and amplitude of daily rhythms of
milk, fat and protein yield among 11 herds in Pennsylvania. ......................................... 172
Supplemental Figure S7-4. Variability in acrophase and amplitude of daily rhythms of
milk fat and protein concentration among 11 herds in Pennsylvania. ............................ 173
Supplemental Figure S7-5. Repeatability of annual rhythms across years. ................. 174
xv
LIST OF SUPPLEMENTAL TABLES
Supplemental Table S3-1. Diet and nutrient composition of the experimental diet. ...... 62
Supplemental Table S3-2. Effect of day- versus night-restricted feeding on total daily
and daily rhythms of individual milk fatty acids. ............................................................. 63
Supplemental Table S4-1. Diet and nutrient composition of experimental diet............. 88
Supplemental Table S4-2. Effect of night-restricted feeding on daily yields of individual
milk fatty acids. ................................................................................................................. 89
Supplemental Table S5-1. Diet and nutrient composition of the experimental diet. .... 117
Supplemental Table S5-2. Fatty acid profile of oil used for abomasal infusion of fatty
acids. ............................................................................................................................... 118
Supplemental Table S6-1. Diet and nutrient composition of the experimental diet. .... 143
Supplemental Table S6-2. Amino acid profile of sodium caseinate. ........................... 144
Supplemental Table S8-1. Adjustment factors to correct for annual rhythms of milk
production in Florida....................................................................................................... 201
Supplemental Table S8-2. Adjustment factors to correct for annual rhythms of milk
production in Minnesota. ................................................................................................ 202
Supplemental Table S8-3. Adjustment factors to correct for annual rhythms of milk
production in Pennsylvania. ............................................................................................ 203
Supplemental Table S8-4. Adjustment factors to correct for annual rhythms of milk
production in Texas......................................................................................................... 204
xvi
LIST OF ABBREVIATIONS
AA-NAT Arylalkylamine N-acetyl transferase
ACC Acetyl-CoA carboxylase
ACTB Beta-actin
ADF Acid detergent fiber
AICAR AMPK-inhibitors 5-aminoimidazole-4-carboxyamide ribonucleoside
AMPK (AMP)-activated protein kinase
AMS Agricultural Marketing Service
B2M Beta-2 microglobulin
BCFA Branched-chain fatty acid
BH Biohydrogenation
BHBA β-hydroxybutyrate
BMAL1 Brain and muscle ARNT-like 1
BVD Bovine viral diarrhea
CCG Clock controlled gene
CCK Cholecystokinin
CLOCK Circadian locomotor output cycles kaput
CIRP Cold-inducible RNA-binding proteins
CK1ε Casein kinase 1ε
CP Crude protein
CPT Carnitine palmitoyltranserase
CRY Cryptochrome
CYP7A1 7-alpha-hydroxylase
DBP D-box binding protein
DGAT1 Diacylglycerol O-acyltransferase 1
DHIA Dairy Herd Information Association
DIO Deiodinase
DM Dry matter
DMH Dorsomedial hypothalamus
DMI Dry matter intake
DRMS Dairy Records Management Systems
E4BP4 E4- promoter binding protein
EE Ether extract
FA Fatty acid
FAA Food anticipatory activity
FASN Fatty acid synthase
FBXL3 F-box/LRR-repeat protein 3
FEO Food-entrainable oscillator
FFA Free Fatty Acid
FMMO Federal milk marketing order
xvii
FOXO1 Forkhead box protein O1
GIP Gastric inhibitory peptide
GLP1 Glucagon-like peptide 1
GRE Glucocorticoid-response element
HAT Histone acetyltransferase
HDAC Histone deacetylase
HLF Hepatic leukemia factor
HMG-CoA 3-hydroxy-3-methyl-glutaryl-Coenzyme A
HSF1 Heat-shock factor 1
IGF-1 Insulin-like growth factor
ipRGC Intrinsically photosensitive retinal ganglion cell
KLF15 Krüppel-like factor 15
LSM Least square means
MEL Melatonin
miRNA MicroRNA
mTOR Mechanistic target of rapamycin
MTP Microsomal triglyceride transfer protein
MUN Milk urea nitrogen
NAMPT Nicotinamide phosphoribosyltransferase
NDF Neutral detergent fiber
NE Norepinephrine
NEFA Nonesterified fatty acids
NF-κB Nuclear factor-κB
NK Natural killer
NPAS4 Neuronal PAS domain protein 4
OBCFA Odd- and branched-chain fatty acids
OCFA Odd-chain fatty acid
OTU Operational taxonomic unit
PBMC Polymorphonuclear blood monocyte
PER Period
PGC-1α Proliferator-activated receptor gamma cofactor 1α
PPAR Peroxisome proliferator-activated receptor
PPRE Peroxisome proliferator-activated receptors response element
PUN Plasma urea nitrogen
psbAI Photosystem II protein
PT Pars tuberalis
PTB Polypyrimidine tract-binding protein
PVN Paraventricular nucleus
qPCR Quantitative polymerase chain reaction
RAPTOR Regulatory-associated protein of mTOR
ROR Retinoic acid-related orphan receptor
RORE Retinoic acid-related orphan receptor binding element
RPS9 Ribosomal protein 9
xviii
S14 Spot 14
SCC Somatic cell count
SCD Stearoyl-CoA desaturase
SCN Suprachiasmatic nucleus
SDPP Short day photoperiod
SEM Standard error of the mean
SFA Saturated fatty acid
SHP Small heterodimer partner
siRNA Small interfering RNA
SIRT1 Sirtulin 1
SREBP Sterol regulatory element-binding protein
T3 Triiodothyronine
T4 Thyroxine
t10 trans-10 C18:1
t11 trans-11 C18:1
TEF Thyrotrophic embryonic factor
TMR Total mixed ration
TSH Thyroid stimulating hormone
TTFL Transcription translation feedback loop
USDA United States Department of Agriculture
UV Ultraviolet
VDR Vitamin D receptor
VFA Volatile fatty acid
VMH Ventromedial hypothalamus
xix
ACKNOWLEDGEMENTS
I have frequently joked that all the friends and family members that have provided love
and support throughout my Ph.D. program and should be listed as co-authors on all of my
manuscripts because the research truly could not have been performed without them. In the
sincerest sense, this is true. I have been extremely blessed to have an incredible support system
around me over the past 4 years, and there is no conceivable way I could have finished my Ph.D.
without them.
First, I would like to thank my advisor Dr. Kevin Harvatine. While his incredible
intellect, critical thinking skills and creativity are impressive, what I am most inspired by is his
quiet humility and the deep kindness he shows to everyone. I am truly fortunate to have had the
opportunity to work with him for my Ph.D.
I would also like to thank the other members of my graduate committee. Dr. Paul Bartell
is responsible for helping me develop a true understanding of circadian biology, and has been
instrumental in the design and analysis of my experiments. I learn something new and exciting
from him nearly every time we interact. Dr. Burt Staniar has been a great mentor and friend, and
always asks important questions I would never have considered. I have also learned many “non-
scientific” things from him about teaching, mentoring, and collaborating with other faculty
members, and always enjoy our discussions of opportunities to improve the academic system. Dr.
Connie Rogers has been a great resource about basic nutrient metabolism, and has sparked my
interest in nutritional immunology, which is a field I hope to pursue in the future.
There are several other members of the Penn State Animal Science Department that I
must thank. Dale Olver recruited me to help coach the dairy judging team on my first day at Penn
State, and since then he has continued to provide opportunities to be involved with the Dairy
Club, and the Pennsylvania dairy industry. I would like to thank Dr. “Doc” Dan Kniffen for
xx
always coming to visit our lab to tell jokes and stories, as well as for providing me with the
opportunity to help T.A. the beef management course. I would like to thank Dr. Judd Heinrichs
for our discussions about everything from cheese to dairy nutrition, to historical changes in the
dairy industry and for inviting me to celebrate Thanksgiving with his family and fellow graduate
students for the past two years. I would like to thank Dr. Chad Dechow for assistance with
statistical analysis, for providing data four our seasonal rhythm experiments, and for always being
kind and supportive. Thank you to Dr. Tara Felix for always bringing up excellent questions in
journal club and providing advice on how to be a young faculty member. Thank you to Dr. Larry
Specht for the conversations about the evolution the dairy industry, and for always making sure to
remind me when Penn State beat the University of Minnesota in sports. Thank you to Dr. Mike
O’Connor for the invitation to the Bracketology 101 class. Finally, I must thank Drs. Craig
Baumrucker, Dr. Joy Pate, Alan Johnson, Troy Ott, and Alex Hristov for generously allowing me
to use of their laboratory equipment.
The completion of my degree would truly not have happened if it were not for the
amazing technicians we have had in the lab. Jackie Ying was incredibly instrumental in teaching
me the laboratory techniques I needed to complete my PhD, and helped organize and execute my
early experiments. When Jackie left the lab, I thought she was irreplaceable, but we were
fortunate to catch lightning in a bottle twice when Dr. Harvatine hired Beckie Bomberger and
Elaine Barnoff. Both of them have been very helpful keeping me organized, helping with
circadian experiments so I could actually get some sleep, and performing lab analyses. They both
will be missed very much.
I must thank my other fellow labmates that I have had the pleasure to work with over the
past 4 years. Natalie Urrutia and Michel Baldin were very welcoming when I started in the lab
and will remain lifelong friends. I am also thankful to Richie Shepardson, Cesar Matamoros, Elle
Andreen, and Reilly Pierce, Yifan Fan, Jocely Souza, Li Junrong, Chengmin Li and Ahmed
xxi
Elzenary for their immense amount of help with sample collection. A special shout out goes to
“Harvateam” members Richie (the “Crystal Crusader”), Cesar (“Captain Ceviche”) and Elle (the
“RuminatoR”) for putting up with me through good times and bad, and for all the fun times we
had together. I am excited to see where all of us end up in the future.
I am incredibly grateful for the friendship I have shared with many other graduate
students in the department. I’d like to give special thanks to Kahina Ghanem for always providing
a spark of positivity and fun whenever I talk to her, for providing a ton of support when I was
finishing my dissertation, and for cooking unique and delicious food for myself and the rest of the
department. Special thanks also go out to Camilla Hughes for the discussions about physiology,
the academic system and politics, and to Isaac Haagen for inviting me to their family holidays on
several occasions, especially when I was new at Penn State. I also want to thank Pedro Carvalho,
Chen Lu, G. Ali Chisti, Susanna Ræsanen, Siga Laspinkas, Martyna Łupicka, Jasmine Dillon,
Lucas Mitchell, Alberto López, Mike Harper, Evelyn Weaver, Neha Oli, Sreelakshmi Vasudevan,
Manasi Kamat, Sonia Arnold, Lydia Hardie and Patricia Ochonski for the fun and friendship.
Lastly, I must thank my wonderful family. More than anyone else, they have made
completion of my PhD possible, If not for my dad’s love of dairy science and my mom’s love of
education; I would not have even thought to pursue a graduate degree in animal science. I cannot
express in words the gratitude I have for their unwavering support and love. I know me living
halfway across the country for 4 years was not easy for them, but they never discouraged me from
pursuing my degree and always made tremendous efforts to visit me here. I would be remiss if I
didn’t also thank my brother Lucas for always providing opportunities for intellectual
discussions, and my sister Hannah for constantly finding new ways to make me laugh.
1
Chapter 1
Introduction
Milk production from dairy cows provides an important source of nutrients for human
diets. As ruminants, dairy cows convert fibrous plant material into high-quality protein, making
them integral to sustainable food production. Moreover, the dairy industry in an important piece
of the United States agricultural economy and accounts for approximately 10% of total
agricultural sales according to the USDA Economic Research Service. Globally, the per capita
demand for dairy products has increased a total of 9% since 1970, and is expected to rise another
20% from now until 2030 as economic development continues throughout the world (FAO,
2012). The dairy industry, along with all of animal agriculture, has been receiving increased
scrutiny regarding potential negative environmental impacts. There has been public pressure to
reduce greenhouse gas emissions from livestock species in order to slow the effects of global
climate change. Additionally, over the past 4 years, the U.S. dairy industry has experienced a
period of unprecedentedly low milk prices, making margins for dairy producers low to
nonexistent. Together, these demands have created imminent desire for creative solutions to
improve the efficiency of milk production.
One potential approach to improve the efficiency of dairy cattle is by developing a better
understanding of the biological rhythms governing milk synthesis. Biological rhythms are
endogenous cycles that govern the behavior and physiology of nearly all organisms and allow
them to predict changes in the external environment before they occur. These rhythms occur
across multiple time scales ranging from a year to less than a day. Of these, the most commonly
described are circadian rhythms, which are approximately 24 h in length, and circannual rhythms,
2
which are approximately 1 year. Both circadian and circannual rhythms are controlled at a
cellular level through a host of proteins comprising a ‘molecular clock’. In recent years, there has
been an increasing body of literature demonstrating an intimate relationship between these
biological rhythms and metabolism in laboratory models and humans. Diseases such as obesity
and type II diabetes have been associated with disruption of circadian rhythms and nutritional
approaches such as time-restricted feeding appear to have potential benefits for metabolic health
and reduced aging (Kessler and Pivovarova-Ramich, 2019). However, there is still a dearth of
research examining the regulation of these biological rhythms in livestock species.
While still poorly understood, there is compelling evidence to suggest that biological
rhythms are important for milk synthesis in dairy cows. Dairy cows exhibit differences in milk
yield and composition across the day, with milk yield being greater in the morning than evening
and milk fat and protein concentration being greater in the evening (Gilbert et al., 1972). Initial
research has suggested that these patterns represent a daily rhythm and that this rhythm is
responsive to changes in feeding time (Rottman et al., 2014). Additionally, several hormones and
metabolites follow a diurnal pattern in dairy cows, suggesting that the metabolism of other tissues
is also controlled by circadian rhythms (Giannetto and Piccione, 2009). Milk yield also follows a
yearly pattern, with increases in milk and milk components during the winter and decreases in the
summer (Wood, 1976). While these rhythms appear to be important, little is known about their
regulation. Developing a better understanding of these biological rhythms may allow for
development of nutritional and management strategies to improve the efficiency of dairy
production.
The overall objective of this dissertation was to examine the relationship factors
regulating the biological rhythms of milk synthesis. Specifically, we investigated the effects of
nutrient intake on the daily rhythms of milk synthesis, as well as the relationships among cow-
and environmental factors and the annual rhythms of milk production. We expect that total food
3
intake and individual macronutrients will entrain the daily rhythms of milk synthesis by
modifying the circadian clock of the mammary gland. Moreover, we anticipate that an annual
rhythm governs yearly changes in milk production, and that this rhythm is consistent among
years, herds, and individual cows, but is responsive to changes in photoperiod.
First, a review of the literature providing a background of the basic molecular
mechanisms governing circadian and circannual rhythms, as well as their relationships to food
intake and metabolism is presented (Chapter 2). The aim of the first experiment (Chapter 3) was
to examine the effects of the time of feed intake on the daily rhythms of milk synthesis by
restricting the time of feed availability to either the day or night. We expected that the daily
rhythms of milk synthesis would be shifted by the time of feed restriction. To investigate if
changes in the rhythms of milk synthesis due to the time of feed intake were mediated by the
molecular circadian clock, a follow-up experiment examining the role of time-restricted feeding
on the daily patterns of clock gene expression in the mammary gland was conducted (Chapter 4).
We expected that expression of clock genes would be rhythmic and that the rhythms would be
shifted by night-restricted feeding. The next two experimented focused on the importance of
individual nutrients for entrainment of the daily rhythms of milk synthesis. Chapter 5 examined
the role of the timing of fatty acid infusion and Chapter 6 examined the role of the timing of
protein infusion over the day on the daily rhythm of milk synthesis. We expected that the time of
fat infusion would alter the daily rhythms of milk and milk fat synthesis, while the time of protein
infusion would alter the daily rhythms of milk and milk protein synthesis.
Chapters 7 and 8 focus on the factors contributing to annual rhythms of milk synthesis.
The analysis reported in Chapter 7, which has been published in the Journal of Dairy Science,
examines the effects of geographic region on annual rhythm of fat and protein concentration
using data from U.S. milk markets, as well as the factors such as lactation number and genotype.
We expected that lactation number and genotype will not affect annual rhythms, but that they will
4
be affected by geographic region. This experiment was followed by another analysis using a large
dataset from Dairy Records Management Systems to determine better elucidate the effects of
region of the U.S. on the annual rhythms of milk, fat and protein yield and fat and protein
concentration, as well as examine potential environmental signals entraining these rhythms
(Chapter 8). Based on the results of the previous experiment, we expected that geographic region
would affect the annual rhythms, and these rhythms would be highly correlated with the change
in photoperiod. Finally, an integrative discussion describing the potential implications, limitations
of these experiments, and future directions is provided in Chapter 9.
5
Chapter 2
Literature Review
Biological Rhythms
The ability of organisms to perceive time and coordinate behavior and metabolism
bestows a tremendous evolutionary advantage, allowing them to maximize energetic resources
and avoid harmful circumstances. Reliably predicting food availability and predator activity, even
by mere minutes, can be the difference between survival and death for many species.
Furthermore, the ability for an organism to monitor time of day allows them to temporally
segregate incompatible physiological processes and maximize energetic efficiency by preventing
futile cycles of metabolism. To facilitate the coordination between the temporal changes in the
external environment, and their own intrinsic metabolism, animals have developed inherent
biological rhythms. These rhythms persist in the absence of external cues and are responsive to
entrainment, or resetting environmental signals.
The most well-studied biological rhythms are circadian rhythms, which are repeating
biological cycles of approximately 24 h. The term circadian was coined in 1959 by Franz
Halberg and is derived from the Latin terms circa, meaning ‘about’; and diēm, meaning ‘day’
(Halberg et al., 1959). However, descriptions of daily rhythms had been made much earlier.
Androsthenes, a scribe for Alexander the Great discovered daily patterns of tamarind
(Tamarindus indica) leaf movements in the 4th century B.C. (Moore-Ede, 1982). These first
recorded observations indicating that daily rhythms may be endogenous were made Jean-Jacques
d’Ortous de Mairan in 1729. He noted that the flowering plant Mimosa pudica, which goes
6
through cycles of open leaves during the day and closed leaves overnight, continued to follow this
pattern even when plants were placed in constant darkness (de Mairan, 1729). This discovery was
revisited nearly 100 years later by Alphonse de Candolle who observed that the rhythm of plant
movement also persisted under constant light, and that the period of leaf movement was slightly
shorter than 24 h (De Candolle, 1832).
While these initial discoveries were important to developing the idea of endogenous daily
rhythms, the field of chronobiology – the study of biological rhythms - was not established until
the middle of the 20th century. Three scientists, Erwin Bünning, Jürgen Aschoff and Colin
Pittendrigh are credited for developing the foundational principles of biological rhythms.
Together they established that biological rhythms are endogenous, meaning that they are self-
sustained through physiological mechanisms in the animal, that they are free-running, meaning
that they can persist without external stimuli, and that they are entrainable, meaning that they are
adaptable to changes in the environment (Daan, 2000). At the same time, biological rhythms on
time scales greater or less than 24 h were proposed. Aschoff (1955) first suggested that seasonal
changes in behavior such as hibernation or migration may be due to an endogenous ‘circannual’
rhythm. This proposition was later confirmed in Willow Warblers (Phylloscopus trochilus),
which exhibited seasonal migratory behavior when placed in constant conditions (Gwinner,
1968). Moreover, great steps in the quantification of biological rhythms were made during this
time. Biological rhythms are characterized by their period length, which is the amount of time to
complete one cycle (e.g. 24 h, 12 m), acrophase, which is the time of the rhythm’s peak, and
amplitude, or the distance from mean to peak (Figure 2-1).
A dogmatic shift in the understanding of biological rhythms was initiated by the now
classic experiment performed by Konopka and Benzer (1971), who used mutagenesis to
demonstrate that the length of a free-running circadian rhythm was controlled at the level of the
gene. Later, an entire host of genes were discovered to make up what is now known as the
7
‘biological clock’, a network of transcription factors that differentially regulate gene expression
across the 24 h day (Huang, 2018). These same transcription factors were later discovered to be
responsible for the generation of circannual rhythms (Lincoln, 2019).
Molecular Structure of Biological Clocks
Circadian rhythms are generated in mammals through a hierarchical network of
independent cellular clocks that communicate through neural and humoral signals. Every
individual cell possesses their own molecular clock that is regulated by intracellular signaling
cascades. A region of ~20,000 neurons in the hypothalamus known as the suprachiasmatic
nucleus (SCN) is commonly acknowledged as the master pacemaker that is principally
responsible for entrainment of rhythms in peripheral tissues (Partch et al., 2014). In addition to
receiving inputs from the SCN, peripheral tissues relay rhythmic information to each other. For
example, melatonin produced by the pineal gland and glucocorticoids produced by the adrenal
cortex are important entraining signals for the clock of the pancreas, liver, and mammary gland
(Mohawk et al., 2012). Peripheral tissues can also be entrained by other cues, primarily feed
intake (Stokkan et al., 2001). Uncoupling of central SCN rhythms from rhythms of feed intake is
proposed as a mechanism for the development of the metabolic disturbances caused by circadian
maladies such as shift-work disorder chronic jet lag (Vetter and Scheer, 2017).
Transcriptional-Translational Feedback Loops
At its most basic level, the molecular circadian clock operates through interconnected
transcriptional-translational feedback loops employing an array of ‘clock’ transcription factors
(Figure 2-2). In mammals, the core negative feedback loop includes brain and muscle ARNT-like
1 (BMAL1), circadian locomotor output cycles kaput (CLOCK), period (PER) 1, 2 and 3, and
cryptochrome (CRY) 1 and 2. BMAL1 and CLOCK form heterodimers that bind to specific DNA
8
sites called E-boxes [canonically CANNTG] in the promoter region of clock-controlled genes.
Additionally, the BMAL1-CLOCK complex enhances the expression of PER and CRY proteins,
which themselves dimerize and feedback on the BMAL1 and CLOCK genes to repress their
expression (Takahashi, 2017). This cycle of activation and repression generates a rhythm of clock
gene expression that is completed in approximately 24 h, which leads to 24 h rhythms of clock-
controlled genes.
An ancillary feedback loop consisting of the orphan nuclear receptors REV-ERBα and β
and RAR-related orphan receptor (ROR) α, β, and γ adds addition level of regulation to stabilize
the core clock and provide secondary outputs for regulation of clock-controlled genes. Both
receptors bind to retinoic acid-related orphan receptor binding elements (ROREs)
[(G/A)GGTCA], with the ROR enhancing transcription and REV-ERB repressing transcription
(Takahashi, 2017). Both BMAL1 and CLOCK contain ROREs and competitive binding by REV-
ERB and ROR leads to either repression or activation of these transcription factors (Guillaumond
et al., 2005). A third level of clock regulation utilizes the transcription factors D-box binding
protein (DBP), thyrotrophic embryonic factor (TEF), hepatic leukemia factor (HLF) and E4-
promoter binding protein (E4BP4, also known as NFIL3). The expression of DBP, TEF, and HLF
is enhanced by the binding of the BMAL1-CLOCK complex to E-Box sequences, and themselves
act as transcription enhancers by binding to D-Box promoter elements. Alternatively, E4BP4
expression is activated by REV-ERBα, and represses expression of D-Box element-containing
genes through interference with DBP, TEF, and HLF binding (Papazyan et al., 2016). The
transcription of clock-controlled genes is therefore coordinated through regulation by a wide host
of transcription factors that act on either E-Box, RORE, or D-Box elements to influence their
expression.
9
Post-Transcriptional Modifications
Beyond the transcriptional-translational feedback loops, a number of post-transcriptional
modifications add an additional layer of regulation to affect both the rhythms of clock gene
expression and their outputs. The importance of these modifications is becoming increasingly
clear, with Reddy et al. (2006) demonstrating that nearly half of the rhythmic proteins present in
the liver lacked corresponding rhythms in mRNA expression. Transcriptional termination is an
important point of regulation for mRNA expression because it is required before transcriptional
initiation can proceed. In addition to its role in E-box mediated transcriptional repression, PER
represses transcriptional initiation through rhythmically binding to several RNA helicases, which
make up the transcriptional termination complex for certain genes (Lim and Allada, 2013).
Recent evidence suggests that alternative splicing of pre-mRNA is also regulated in a circadian
manner (McGlincy et al., 2012). Microarray data demonstrated that 62 of 227 measured splicing
factors displayed rhythmic gene expression (Grosso et al., 2008). Furthermore, Preußner et al.
(2017) illustrated that circadian rhythms of body temperature entrain cellular rhythms of alternate
splicing in the cerebellum and liver with the effect via heat-sensitive SR proteins.
The stability of mRNA is also influenced by circadian rhythms through silencing by
RNA-binding proteins and microRNA (miRNA). Period 2 and CRY1 stability is decreased by
RNA-binding proteins polypyrimidine tract-binding protein (PTB) and AU-rich element RNA-
binding protein 1. Deletions of these proteins result in increased amplitudes of the circadian
mRNA expression (Green, 2018). MicroRNA are small (~22 nt) non-coding RNA molecules that
interact with the 3’ untranslated region of mRNA to repress its translation. The light inducible
miRNA miR-132 augments the rhythms of PER1 expression in murine SCN neurons (Cheng et
al., 2007). In the same study, miR-219-1 was identified as clock-controlled miRNA whose
rhythm is ablated in CRY1/CRY2 double knockout mice. Later research implicated miR-219 as a
link between shift-work disorder and increased breast cancer. Night shift workers have greater
10
methylation of the miR-219 promoter, which downregulates its expression and attenuates its
anticarcinogenic activity (Shi et al., 2013). The miRNA 192/194 cluster was discovered to inhibit
translation of all three PER genes, and overexpression and knockdown of these miRNAs leads to
shortening and lengthening of the circadian period, respectively (Nagel et al., 2009). Furthermore,
miR-122 is regulated in a circadian manner in the liver and causes rhythmic degradation of
several genes involved in cholesterol and lipid metabolism (Mehta and Cheng, 2013).
Post-Translational Modifications
Post-translational modifications are also vital to the generation of 24-h rhythms within
cells. Modifications to clock proteins introduce a delay between the activating and repressing
arms of the transcriptional-translational feedback loop, ensuring that the cycle is maintained at a
period of ~ 24 h. Without this delay, the cycle of activation and repression would last just a few
hours (Gallego and Virshup, 2007). The most well-established example of post-translational
regulation of circadian period length in mammals occurs through casein kinase 1ε–mediated
phosphorylation. Casein kinase 1ε (CK1ε) decreases the stability of PER through
phosphorylation, leading to its degradation via the ubiquitin-mediated proteosomal pathway
(Vielhaber et al., 2000).
The importance of phosphorylation in the establishment of cell-autonomous rhythms is
perpetuated by presence of circadian rhythms in cells that are completely devoid of rhythmic gene
expression. Prokaryotic cyanobacteria can maintain 24-h cycles of growth, nitrogen fixation, and
photosynthesis with a molecular clock based only on rhythmic phosphorylation of KaiA, KaiB,
and KaiC proteins (Takai et al., 2006). Moreover, circadian oscillations of peroxiredoxin
oxidation occur in differentiated red blood cells which contain no nucleus or DNA, suggesting
that these oscillations are maintained through exclusively post-translational mechanisms (O’Neill
and Reddy, 2011). Phosphoproteomic analysis is further uncovering the indispensable role of
11
phosphorylation in generation of circadian rhythms. Robles et al. (2017) determined using mass
spectrometry that 25% (> 20,000 phosphosites) of the murine hepatic phosphoproteome are
follow a 24 h rhythmic pattern, with phosphorylation cycles having markedly greater amplitudes
than the rhythms of transcript or protein abundance.
Epigenetics and the Clock
Epigenetic modifications, such as DNA methylation and histone acetylation/deacetylation
have a bidirectional link with the circadian clock. In the liver of mice, acetylation of the H3 histone
follows a circadian rhythm, causing chromatin remodeling near the promoter regions of Per1, Per2,
and Cry1, permitting their expression (Etchegaray et al., 2003). The circadian transcription factor
CLOCK is itself a histone acetyltransferase (HAT), and its function is potentiated by its
heterodimer BMAL (Doi et al., 2006). REV-ERBα, a transcription factor of the ancillary molecular
clock, helps recruit the histone deacetylase 3 (HDAC3) to specific sites on histone 3, inhibiting the
transcription of target genes (Yin et al., 2007). Azzi et al. (2014) demonstrated that DNA
methylation patterns in the SCN of mice were associated with changes in the light-dark cycle.
Entrainment of Biological Clocks
Entrainment is the process of synchronizing endogenous biological clocks with external
environmental rhythms. Environmental cues, called Zeitgebers (German for “time giver”), confer
information about the environmental time to circadian oscillators to synchronize an organism’s
internal clock (Aschoff, 1963). All organisms have endogenous rhythms that differ slightly from
24 h which allow them to be responsive to entrainment signals from the environment. When
zeitgebers are absent, the rhythm is said to be freerunning, meaning that it expresses its
endogenous period length. The length of the freerunning period, also called tau (τ), differs
between species and even individuals. Humans, for example, have an average freerunning period
12
of 24.2 h (Czeisler et al., 1999). Entrainment of circadian clocks can occur through several cues,
including the light/dark cycle, nutrient intake, and body temperature.
Light Entrainment
Light is the most prominent zeitgeber known to entrain the central clock of the SCN.
Light is detected by the eye and transmitted to the SCN through two separate pathways. Rod and
cone cells are the primary visual photoreceptors in the retina. After sensing light, they signal the
SCN via retinal ganglion cells, setting its circadian clock (Drouyer et al., 2007). However, light
entrainment can occur completely independent of rod and cone cells. Foster et al. (1991)
demonstrated that mice possessing the rd/rd mutation, which causes degeneration of rod and cone
cells and complete blindness, still entrained to the light-dark cycle, suggesting that an additional
photoreceptor was involved in circadian entrainment. It was later discovered that the non-visual
photopigment melanopsin can entrain the master clock through excitement of intrinsically
photosensitive retinal ganglion cells (ipRGCs) in the eye (Panda et al., 2002). Exposure of
ipRGCs to light causes photoisomerization of the melanopsin protein, leading to a G-protein
signaling cascade which causes cell depolarization and increases the nerve firing rate (Hattar et
al., 2002). The ipRGCs directly innervate the SCN via the retinohypothalamic tract, conveying
light signals directly to the master clock. Panda et al. (2003) later confirmed the independence of
circadian oscillation and rod and cone cells using rd/rd mice, as well as mice lacking melanopsin
(Opn4-/-) and mice with both mutations. They found that photoentrainment occurred in mouse
strains with single deletions, but did not occur in the absence of both receptor systems. In addition
to establishing two separate pathways of circadian entrainment this study submitted that unlike
birds, reptiles, fish and invertebrates which can detect light extra-ocularly in the pineal gland or
deep-brain photoreceptors, mammals require ocular photoreceptors for photoentrainment.
13
Food Entrainment
Circadian rhythms of mammals can be entrained by nutrient intake. Early observations of
rat behavior revealed an increase in locomotor activity in the 2 to 4 h prior to feeding (Richter,
1922). These patterns of food anticipatory activity (FAA) were later shown to persist for up to 3
to 4 days after complete food removal (Stephan, 2002). While the light dark cycle is the
predominant zeitgeber for entraining behavioral rhythms, circadian rhythms of FAA can overtake
them during conditions of restricted food availability. Stokkan et al. (2001) detected a phase shift
of Per1 expression corresponding to feeding time in the liver and lung of mice, but not the SCN
when food was restricted to the inactive period of the day. The amount of entrainment is
proportional to the energy content of the feed, feeding a highly palatable, high-energy feed at a
specific time each day is sufficient to entrain FAA, even if food availability is not restricted
(Mistlberger, 1994). Additionally, food entrainment can maintain circadian behavioral rhythms
when signaling from the central clock is absent. Krieger et al. (1977) observed that rats with
lesions of their SCN still exhibited entrainment of daily locomotor activity, core body
temperature, and blood glucocorticoid concentration rhythms to once per day feeding. These
discoveries suggests that food entrainment is controlled by an independent oscillator separate
from the SCN.
While there has been evidence substantiating the presence of food-entrainable circadian
rhythms, the search for the so-called food-entrainable oscillator (FEO) has proven onerous.
Initially, the adrenal gland was targeted as the source of food-entrainable oscillations because of
the role of glucocorticoids in food anticipation, but adrenalectomy failed to affect the circadian
rhythms of FAA (Stephan, 2002). Several regions of the central nervous system have been
explored for their potential role as an FEO. Ablation of the ventromedial hypothalamus (VMH), a
region involved in satiety signaling and thermoregulation, eliminated the entrainment of body
temperature and glucocorticoids to meal time in some early short-term experiments, but this
14
effect did not occur when recovery of greater than 14 weeks was allowed following surgery
(Davidson, 2009). More recent experiments have failed to show correlations between VMH
lesions and FAA (Gooley et al., 2006). The paraventricular nucleus (PVN) is a region of the brain
involved in appetite regulation and stress. Lesioning of the PVN failed to reduce food-
anticipatory food-bin approach behavior, although general locomotion was reduced (Mistlberger
and Rusak, 1988).
The dorsomedial hypothalamus (DMH) has been highly implicated as the location of an
FEO, but results have been inconsistent. It was initially explored as a candidate because of its
previously discovered roles in feeding behavior regulation, stress response, and involvement in
the circadian rhythms of cortisol release (Chou et al., 2003). Gooley et al. (2006) observed that
food restriction synchronizes daily rhythms of DMH neuronal firing to the time of feed delivery.
Additionally, rhythmic expression of Per1 and Per2 is induced in the DMH during time-restricted
feeding (Mieda et al., 2006). However, Landry et al. (2006) failed to impair daily rhythms of
FAA upon partial or complete lesioning of the DMH. These results were later corroborated by
Landry et al. (2007) and Moriya et al. (2009). More recent evidence suggests that the effect of
DMH lesions on FAA is time-of-day dependent. Landry et al. (2011) demonstrated that DMH-
lesioned rats had less robust, but still present, rhythms of FAA when fed during the day, but that
FAA was unaffected when they were fed at night. The inability to isolate the FEO to a specific
region of the brain suggests that food-entrained rhythms may actually be generated by a network
of separate oscillators that communicate with each other to maintain food-anticipatory rhythms of
behavior, body temperature, and glucocorticoid signaling.
Communication from the viscera to the brain has been investigated as a mechanism for
entrainment by nutrient intake. However, ablation of both vagal and non-vagal afferents nerves
emanating from the viscera did not alter FAA (Davidson, 2009). Ghrelin is a peptide hormone
produced by the oxyntic cells of the stomach that is directly responsible for hunger signaling. The
15
expression of Per1 and Per2 in oxyntic cells is in antiphase with ghrelin secretion, and the
rhythms of their expression are altered by feeding time (LeSauter et al., 2009). Knockout of the
ghrelin receptor greatly reduces food-anticipatory locomotor activity in mice, suggesting that
ghrelin is important for this behavior (Blum et al., 2009). Research performed in goldfish
(Carassius auratus) demonstrated that ghrelin treatment elevated hepatic and hypothalamic Per1,
2 and 3 expression by over 2-fold and that treatment with a ghrelin antagonist completely
abolished daily rhythms of FAA (Nisembaum et al., 2014).
Temperature Entrainment
Environmental temperature is a common entraining cue for poikilotherms, who
experience dramatic changes in physiology when external temperature changes. Additionally,
Some plants and algae can entrain to just 1 to 2°C differences in temperature across the day
(Rensing and Ruoff, 2002). In mammals, the circadian rhythms of body temperature can entrain
biological clocks in peripheral tissues. Brown et al. (2002) ascertained that rhythms of Per2
expression could be entrained in rat fibroblast cells by temperature cycles that oscillated by 4°C.
Saini et al. (2012) later revealed similar results in human and mice fibroblast, but revealed that as
little as a 1°C change in temperature is sufficient for entrainment. These authors also revealed
that temperature entrainment was dependent on the protein heat-shock factor 1 (HSF1). In
addition to HSF1, cold-inducible RNA-binding proteins (CIRPs) link body temperature and the
circadian clock. Cycles of body temperature affect the regulation of polyadenylation sites by
CIRP, thus affecting the stability of target RNAs (Ki et al., 2015). While peripheral tissues appear
responsive to temperature, the SCN possesses mechanisms to resist feedback of body temperature
on the central clock. Buhr et al. (2010) discovered that communication between the dosoromedial
and ventrolateral regions of the SCN via L-type calcium channels prevents temperature from
resetting the SCN clock.
16
Circadian Rhythms and Nutrient Metabolism
As mentioned above, feeding behavior is a strong clue for circadian entrainment. In
addition to entrainment of whole-body rhythms by meal delivery or high-calorie feed
supplementation, specific nutrients can directly affect the molecular clock within individual cells.
Cellular energy status, redox status, and direct action by metabolically important hormones can
directly impact the circadian clock (Peek et al., 2012). Furthermore, the molecular clock exerts a
reciprocal relationship on cellular metabolism. By temporally segregating biochemically
incompatible processes, the circadian clock prevents futile cycling and maximized energetic
resources within the cell. Moreover, the efficiency of nutrient utilization is optimized by the
circadian clock because it coordinates the proper metabolic machinery with the time of nutrient
availability.
Cellular Energy and Redox Status and the Circadian Clock
Cellular energy status can entrain circadian rhythms through adenosine monophosphate
(AMP)-activated protein kinase (AMPK). During states of prolonged fasting and ATP depletion,
AMPK is activated. Activated AMPK can then affect the period length of the cellular circadian
clock by two separate pathways leading to CRY or PER degradation. First, AMPK
phosphorylates CK1ε at serine 389, which causes an increase in its activity leading to increased
degradation of PER2 and shortening of the circadian period (Jordan and Lamia, 2013).
Furthermore, AMPK can directly phosphorylate CRY1 and CRY2, which then interact with F-
box/LRR-repeat protein 3 (FBXL3), leading to their polyubiquitination and proteosomal
degradation (Lamia et al., 2009). Treatment with the AMPK antagonists, AMPK-inhibitors 5-
aminoimidazole-4-carboxyamide ribonucleoside (AICAR) or metformin, has been shown to
cause a phase shift in the circadian clock of the liver, suggesting that AMPK may play a role in
17
entrainment (Jordan and Lamia, 2013). In addition to its effects on circadian clock proteins,
AMPK appears to be responsible for circadian regulation of its target proteins. Both acetyl-coA
carboxylase (ACC), the rate-limiting enzyme for fatty acid biosynthesis, and regulatory-
associated protein of mTOR (RAPTOR), a protein involved in signaling cascade of insulin to
mechanistic target of rapamycin (mTOR), display 24 h rhythms of AMPK-mediated
phosphorylation (Davies et al., 1992; Lamia et al., 2009).
Cellular redox state is dependent on the ratio of the oxidized and reduced forms of the
cofactor nicotinamide adenine dinucleotide (NAD+/NADH). The molecular clock directly
impacts redox status through regulation of nicotinamide phosphoribosyltransferase (NAMPT), the
rate-limiting enzyme in the NAD+ salvage pathway (Peek et al., 2012). The CLOCK:BMAL1
complex binds to E-boxes in the NAMPT promoter, upregulating its expression and causing
NAMPT and NAD+ to oscillate in a diurnal manner (Ramsey et al., 2009). The diurnal pattern of
NAD+ also likely affects the circadian cycling of other metabolically important proteins though
its effect on sirtulin 1 (SIRT1). Sirtulin 1 is an NAD+-dependent deacetylase that regulates the
expression of a wide array of metabolically important transcription factors, including peroxisome
proliferator-activated receptor gamma cofactor 1α (PGC-1α, involved in mitochondrial biogenesis
and respiration), sterol regulatory element-binding protein 1c (SREBP-1c; involved in lipid
synthesis and glucose metabolism), and forkhead box protein O1 (FOXO1; involved in insulin
signaling) (Peek et al., 2012). Nakahata et al. (2009) confirmed that SIRT1 activity follows a
circadian pattern and that selective inhibition of NAMPT can abolish this rhythm. Furthermore,
SIRT1 can also feedback to inhibit the BMAL1-CLOCK complex through deacetylation of
BMAL1, preventing activation of clock-controlled genes (Tong et al., 2015).
The cellular redox state or peroxiredoxin proteins has also been shown to itself be a
circadian oscillator (Peek et al., 2012). Peroxiredoxins are proteins that both serve as antioxidants
which scavenge hydrogen peroxide (H2O2) and whose regulation is a component of H2O2-
18
mediated signal transduction (Rhee et al., 2005). Therefore, the oxidation status of peroxiredoxins
can affect the transcription and regulation of other target genes and proteins. O’Neill et al. (2011)
discovered that peroxiredoxins can maintain self-sustained circadian oscillations, independent of
transcription. They later discovered that this mechanism allows red blood cells, which lack a
nucleus, to express circadian rhythms (O’Neill and Reddy, 2011). The oxidation rhythms of
peroxiredoxins is out of phase with the rhythms of Bmal1 mRNA, and this phase relationship
differs among tissues, suggesting the peroxiredoxins may represent an entirely different oscillator
than the TTFL (Edgar et al., 2012).
Glucose Metabolism and the Clock
Glucose metabolism in mammals is highly influenced by the circadian system. Plasma
glucose and insulin concentrations exhibit 24 h rhythms in a variety of species including humans
(Van Cauter et al., 1991), mice (Sadacca et al., 2011), rats (Bellinger et al., 1975), sheep (McMillen
et al., 1987), and cattle (Rottman et al., 2014). Furthermore, liver glycogen concentration follows
an entrainable circadian rhythm (Halberg et al., 1960). Both Bmal1 and Clock are essential for
maintenance of these rhythms, and global knockout of either gene abolishes the daily oscillations
of plasma glucose concentration and hepatic gluconeogenesis (Rudic et al., 2004). In addition to
absolute glucose and insulin concentrations, insulin-stimulated glucose uptake is controlled by
circadian rhythms. In the medical field, knowledge that glucose tolerance in humans varies across
the day has existed since the late 1960’s (Bowen and Reeves, 1967). Van Cauter et al. (1997)
determined that 2 h post-infusion blood glucose concentrations are typically 30 to 50 mg/dL greater
when glucose infusion is performed the afternoon compared to in the morning. The insulin-
secreting islet β cells of the pancreas express 24 h rhythms of molecular clock genes in mice, and
when these rhythms are abolished by tissue-specific Bmal1 knockout, subjects exhibited severe
glucose tolerance and insufficient glucose-stimulated insulin production (Sadacca et al., 2011).
19
Besides its effects on glucose metabolism, the circadian clock appears to affect other aspects of
carbohydrate metabolism. Thaiss et al. (2016) determined using metabolomic analysis that the
conversion of sucrose to lactate was rhythmic.
Circadian rhythms of glucose metabolism may also be controlled through rhythms of
incretin release. Incretins, which include glucagon-like peptide-1 (GLP-1) and gastric inhibitory
peptide (GIP) are released from the enteroendocrine cells of the gut in response to feed intake and
enhance insulin release from the pancreas. Like insulin, the responsiveness of incretins to feed
intake shows robust circadian rhythms in humans and model organisms (Reimann and Reddy,
2014). L cells, responsible for the production of GLP-1, show 24 h oscillations of Bmal1 and
Per2 in vitro (Reimann and Reddy, 2014). Gil-Lozano et al. (2014) illustrated that GLP-1
secretion in response to various secretagogues followed a diurnal pattern. A less well-
characterized incretin, oxyntomodulin, appears to play an important role in food entrainment.
Landgraf et al. (2015) detected that oxyntomodulin entrains the rhythms of Per1 and Per2
expression via binding to the glucocorticoid receptor and that knockout of this protein blocks
food entrainment of the liver. Therefore, it is a promising candidate for the link between food
intake and the molecular clock.
Another link between circadian clocks and glucose metabolism is glucocorticoid
signaling. Glucocorticoids are a class of steroid hormones produced by the adrenal cortex that
have a somewhat antagonistic role to insulin, causing greater hepatic gluconeogenesis, reduced
glucose uptake, and increased lipid mobilization (Kuo et al., 2015). They are increased during
fasting, but are also greatly elevated in response to stress. In addition to their role in stress and
metabolism, glucocorticoids are a major output of the core circadian clock. Glucocorticoids
follows a consistent daily rhythm that peaks at the start of the active phase in both nocturnal and
diurnal animals (Dickmeis, 2009). This rhythm is modulated directly by the SCN, which
innervates the adrenal gland and affects its release in a circadian manner (Mohawk et al., 2012).
20
Glucocorticoids directly entrain circadian rhythms in peripheral tissues. This mechanism is
frequently exploited when studying circadian rhythms in vitro, when the synthetic glucocorticoid
dexamethasone is used to synchronize cultured cells (Balsalobre et al., 2000). The promoter
regions of Bmal1, Cry1, Per1, and Per2 contain glucocorticoid-response elements (GREs) that
regulate the circadian clock in peripheral tissues. (Mohawk et al., 2012). Furthermore, treatment
with glucocorticoids decrease the rate of entrainment of peripheral clocks to the time of food
intake (Minh et al., 2001). So et al. (2009) discovered that deletion of the GRE in Per2 protects
individuals from insulin resistance, demonstrating the intimate link between the glucocorticoid
rhythm and glucose metabolism.
Lipid Metabolism and the Clock
Several aspects of lipid metabolism are controlled by the molecular clock. Carnitine
palmitoyltranserase (CPT) 1 and 2 control fatty acid β-oxidation by limiting the rate of fatty acyl
entry through the inner and outer mitochondrial membranes, respectively. The protein abundance
of both CPT1 and CPT2 follows circadian rhythms with CPT1 peaking at 17 h post-light
exposure and CPT2 peaking at 6 h post-light exposure (Neufeld-Cohen et al., 2016). Acetyl-CoA
carboxylase, the enzyme that catalyzes the rate-limiting reaction of de novo lipogenesis, also
follows a diurnal pattern, likely due to regulation by AMPK (Gnocchi et al., 2015). The master
lipid regulator sterol regulatory element-binding transcription factor 1 (SREBP1) also displays
circadian rhythms of activity, which appear to be directly driven by REV-ERBα (Le Martelot et
al., 2009). Recent research has suggested that circadian rhythms of both SREBP1 and another
lipogenic factor, thyroid hormone-responsive Spot 14 (S14), are entrained by the time of feed
intake in the mammary gland of mice (Ma et al., 2013).
Lipid trafficking from the intestine also appears to be regulated in a circadian manner.
Intestinal mRNA and protein expression and activity of microsomal triglyceride transfer protein
21
(MTP), an enzyme involved in the assembly of chylomicrons, follows a daily pattern with greatest
expression at 2400 h. (Pan and Hussain, 2007). This daily pattern is regulated directly by CLOCK
through its effects on the transcription factor called “small heterodimer partner (SHP)”, which
enhances expression of MTP (Pan et al., 2010). Pan and Hussain (2007) revealed that absorption
of 3H -labeled triolein cholesterol followed a daily pattern with greater absorption at 2400 h than
1200 h. Another potential link between the circadian clock and lipid transport is the protein
nocturnin. Nocturnin controls gene expression by deadenlyation of the poly(A) tail of mRNA which
leads to its degradation, and its expression is activated by BMAL/CLOCK (Stubblefield, 2012).
Global knockout of the nocturnin gene results in reduced export of chylomicrons out of enterocytes
(Douris et al., 2011).
Bile acid metabolism and cholesterol synthesis are also intimately linked with the circadian
clock. 3-hydroxy-3-methyl-glutaryl-Coenzyme A (HMG-CoA) reductase, the rate limiting enzyme
for the synthesis of cholesterol, exhibits a diurnal pattern of expression leading to a circadian
rhythm of cholesterol synthesis (Shefer et al., 1972). Additionally the rate-limiting enzyme in bile
acid synthesis, 7-alpha-hydroxylase (CYP7A1) follows a circadian rhythm (Chiang, 2009). This
rhythm is directly regulated by REV-ERBα, which binds to D-box elements on the CYP7A1
promoter and enhances its transcription. Circadian rhythms of bile acid synthesis allow for greater
amounts of bile acids to enter the small intestine and emulsify dietary lipids during the active phase
when food intake is more likely to occur. Knockout of the circadian clock has detrimental effects
on bile acid homeostasis. Ma et al. (2009) discovered that Per1-/- Per2-/- double knockout mice lost
their circadian rhythms of bile acid synthesis, leading to a greater synthesis and over accumulation
of bile acids in the liver.
Peroxisome proliferator-activated receptors (PPARs) are nuclear receptors that act as lipid
sensors and master regulators of lipid and glucose metabolism. The three primary isoforms include
PPARα, PPARγ, and PPARδ, which have different binding affinities to specific fatty acids and are
22
distributed at different expression levels across tissues. All three PPAR isoforms, along with
peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1-α), display circadian
rhythms of expression (Chen and Yang, 2014). An E-Box element has been identified in the
promoter of PPARα in mice, suggesting direct regulation by BMAL/CLOCK (Oishi et al., 2005).
The rhythms of PPARα are likely responsible for diurnal regulation of SREBP1 and fatty acid
synthase (FAS) and HMG-CoA because it activates expression of these genes (Chen and Yang,
2014). Nuclear translocation of PPARγ in bone-marrow stromal cells is stimulated by the circadian
polyadenylase nocturnin (Kawai et al., 2010). Furthermore, liver PPARδ is expression is regulated
in a circadian manner by miR-122, a cyclic microRNA that is activated by REV-ERBα (Gatfield
et al., 2009). The relationship between PPARs and the clock is bidirectional. The promoter regions
of Bmal1 and Rev-erbα possess PPAR response elements (PPREs) and are activated upon PPAR
binding (Ribas-Latre and Eckel-Mahan, 2016). Through this mechanism, lipids feed back and affect
the molecular clock.
Circadian Rhythms in Microorganisms
A growing body of evidence has suggested an important role for circadian rhythms in
modulating growth and metabolism of microorganisms. The idea that microorganisms are capable
of maintaining biological rhythms is relatively new. Prior to the 1980’s prokaryotes were
considered too “simplistic” and too short-living to express circadian rhythms. Grobbelaar et al.
(1986) first observed endogenous circadian rhythms of nitrogen fixation in the cyanobacteria.
These findings were later validated by Kondo et al. (1993), who used a luciferase reporter to
demonstrate that the cyanobacterial species Synechococcus sp. PCC7942 exhibits daily rhythms of
photosystem II protein (psbAI) gene expression under constant conditions. Their results challenged
the previous dogma that bacteria do not possess rhythms by demonstrating that S. elongatus possess
the three criteria necessary for circadian rhythmicity: entrainment by environmental stimuli,
23
maintenance of approximately 24 h cycles after removal of external cues, and maintenance of
rhythms under a wide range of environmental temperatures. Since the initial confirmation of
rhythms in cyanobacteria, over 100 different clock mutant strains of bacteria have been identified
with circadian periods ranging from 16 to 60 h (Kondo et al., 1994).
The structure of the cyanobacterial clock differs considerably from the molecular clock of
eukaryotes. Unlike the core eukaryotic clock, the core cyanobacterial clock operates exclusively
through a post-translational oscillator centered around the protein KaiC (Swan et al., 2018). Along
with KaiC, two members of the same gene family, KaiA and KaiB, are responsible for generating
cycles of phosphorylation and dephosphorylation. Binding of KaiA to unphosphorylated KaiC
during the day promotes a conformational change that leads to autophosphorylation of KaiC using
cellular ATP as a phosphorus donor. The phosphorylation of KaiC leads to the recruitment of KaiB,
which creates a separate conformational state that inhibits KaiA binding (Kageyama et al., 2006).
The lack of KaiA binding decreases the ATPase activity of KaiC and causes
autodephosphorylation. As phosphorylation state decreases, KaiB disassociates with KaiC,
allowing KaiA to bind again and the cycle to repeat. The KaiABC oscillator activates a signaling
cascade that activates the response regulator RpaA which regulates the transcription of circadian-
associated genes (Swan et al., 2018). Entrainment of the cyanobacterial clock occurs both through
sensing the ratio of ATP to ADP within the organism, as well as through the presence of oxidized
quinone metabolites which act as electron shuttles during photosynthesis (Williams et al., 2002;
Ivleva et al., 2006). Interestingly, the growth cycle of cyanobacteria actually lasts only about 6 h,
but the phase of the circadian clocks is passed onto daughter cells to maintain their rhythms
(Schibler and Sassone‐Corsi, 2002).
Besides Synechococcus, several other bacterial species exhibit circadian rhythmicity.
Circadian rhythms of bioluminescence have been exhibited in the purple photosynthetic bacterial
species Rhodobacter sphaerodies (Min et al., 2005). Rhodopseudomonas palustris TIE-1 show
24
self-sustained circadian rhythmicity of nitrogen-fixation that is entrained by oxygen (Ma, 2011).
Cycling patterns of anaerobic fermentation have been observed in continuous cultures of
Clostridium acetobutylicum, exhibiting 24-h cycles of acetate, propionate and butyrate, ethanol,
and biomass concentrations (Yerushalmi and Volesky, 1989). Additionally, transcriptomic analysis
uncovered 24-h temporal dynamics of metabolic gene expression in six oceanic bacterioplankton
(Ottesen et al., 2014).
Microorganisms may play an important role in modulating circadian rhythms in the host.
The intestinal microbiome and the host alimentary system have a bi-directional relationship through
which perturbations in either system can affect the circadian entrainment of the other. Estimates
suggest that 15 to 17% of the operational taxonomic units (OTU; proxy for species) and 23% of
the functional pathways of the colonic microbiome undergo daily rhythms (Thaiss et al., 2014;
Zarrinpar et al., 2014). These rhythms function to affect microbial metabolism, including growth,
energy metabolism, and amino acid metabolism, as well as regulate the spatial organization of
bacteria through affecting chemotaxis and mucosal adherence (Liang and FitzGerald, 2017).
Inhibition of the gut microbiome, either in germ-free models or through antibiotic treatment, leads
to damping of the circadian rhythms in the gut epithelium and liver (Leone et al., 2015; Thaiss et
al., 2016). Reciprocally, Per1/2 double knockout mice failed to demonstrate circadian rhythms of
microbial movement and mucosal attachment (Thaiss et al., 2016). The signaling mechanisms by
which the host and microbiome communicate temporal information are still being elucidated. Short
chain fatty acids, particularly butyrate, produced by hindgut fermentation follow daily rhythms,
and induce cyclicity in colonic epithelial cells, suggesting that they may be an important
entrainment signal derived from the microbiome (Leone et al., 2015). Murakami et al. (2016)
discovered using a fecal transplant model that PPARγ signaling may be responsible for
microbiome-entrainment of the circadian clock.
25
Humoral signals from the host also act directly to entrain the rhythms of the microbiome.
Paulose et al. (2016) discovered that the gastrointestinal bacterium Enterobacter aerogenes, which
displays daily patterns of motility, is responsive to treatment by exogenous melatonin. While it has
not yet been examined, cortisol may also be a mechanism whereby the host can entrain the
microbiome. As previously discussed, glucocorticoids follows a circadian rhythm that is directly
entrained by the master clock of the SCN. Feeding exogenous cortisol has been shown to modulate
the gut microbiome of pigs to promote the growth of coliform bacteria and suppress Salmonella
(Petrosus et al., 2018). Alverdy and Aoys (1991) implied that treatment with glucocorticoids can
cause increased bacterial adherence to the mucosal wall of the hindgut. Salivary cortisol
concentration follows a 24 h rhythm in humans (Katz and Shannon, 1969). Glucocorticoid signaling
through saliva may be particularly impactful for ruminants, because saliva enters the rumen
immediately after secretion and can directly impact the microbial community.
Annual Rhythms
In addition to circadian rhythms, many organisms have evolved a circannual timing system
to coordinate physiology across the year. This system uses environmental signals, primarily the
photoperiod, or length of daylight, to synchronize the organism’s physiology to the seasonal
changes in climate. Seasonal temperature changes occur due to the 23.5° angle of the earth’s axis
which causes the amount of solar radiation reaching different hemispheres of the earth’s surface to
change across the year (Foster and Kreitzman, 2009a). These changes in temperature are furthered
by albedo, the degree of sunlight reflectance from the earth’s surface back into space. Albedo is
greatest near the earth’s poles because ice reflects the greatest amount of sunlight, while water
reflects less solar energy than land (Sellers et al., 1997). Other annual changes such as jet streams,
26
ocean currents, tidal cycles, and cycles of rainfall create additional seasonal climates for organisms
to anticipate (Namias, 1976).
Many animals exhibit circannual rhythms regulating numerous physiological processes
including growth, reproduction, metabolism, and behavior. For example, migratory birds display
fattening and nocturnal activity (zugunrhue) coinciding with spring and fall migration. These
behaviors persist under constant light:dark (L:D) cycles, and can be modified by altering
photoperiod (Gwinner, 1996). Many mammals including horses, sheep, and hamsters exhibit
endogenous yearly rhythms of fertility (Gerlach and Aurich, 2000). Moreover, many mammals and
birds conserve energy in the winter through hibernation or through entering a state of winter torpor
(Geiser and Ruf, 1995).
Annual Rhythms and Reproduction
Many animals in seasonal environments have evolved to reproduce during a certain time
of the year. There is a clear advantage to birthing young during a time of high feed availability or
low predator abundance. In herbivorous mammals, births are typically timed to occur in the spring
when forage is readily available, providing the mother with ample calories to nurse her young
(Foster and Kreitzman, 2009b). Sheep (Ovis aries) are a classic example of seasonal breeders.
Domesticated ewes exhibit estrus and ovulation only during autumn and early winter resulting in
parturition in mid-spring (Robinson, 1951; Shelton and Morrow, 1965). Rams increase the
secretion of testosterone and production of sperm during the period of July through December,
corresponding to the time of estrus in the females (Lincoln, 1976; Dacheux et al., 1981).
Photoperiod is the major driver of the seasonality in sheep, but it is modulated by many
other factors such as nutritional status, previous lambing date and lactation (Lincoln, 1998).
Photoperiodic information is first sensed by the retina and transmitted to the suprachiasmatic nuclei
(SCN) of the hypothalamus by the photopigment melanopsin via photosensitive retinal ganglion
27
(Foster and Kreitzman, 2009b). Daylight increases the oscillation of the SCN (VanderLeest et al.,
2007). During periods of low SCN activity (i.e. dark photoperiod), norepinephrine signals the
release of melatonin from the pineal gland via efferent nerve signaling (Simonneaux and Ribelayga,
2003). The binding of norepinepherine causes an influx of calcium into pineal cells, activating
arylalkylamine N-acetyl transferase (AA-NAT), the rate limiting enzyme in melatonin production
(Klein et al., 1983).
The pituitary gland is responsible for integrating photoperiodic information with
reproductive signals. Thyrotroph cells in the pars tuberalis (PT) of the anterior pituitary gland are
rich in melatonin receptors (Morgan et al., 1989). These melatonin receptors are species-specific,
and are expressed at comparable levels throughout the year (Morgan et al., 1994). Briefly,
melatonin regulates the molecular clock of the PT by inducing Cry1 expression and repressing Per1
expression (Jilg et al., 2005; Johnston et al., 2006). This causes the relationship between abundance
of Per1 and Cry1 to expand and contract throughout the day, with Cry1 expression reaching a
maximum at dusk and Per1 peaking at dawn. The diurnal relationship of Per1 and Cry1 is an
example of an internal coincidence model that creates a yearly timekeeping mechanism implicated
in driving seasonal hormonal changes.
The seasonal effect of photoperiod modulates reproduction via thyroid hormone
production. Triiodothyronine (T3), the active form of thyroid hormone, affects the pulsatile release
of GnRH from the hypothalamus, subsequently affecting the release of LH and FSH from the
anterior pituitary gland (Misztal et al., 2002). The activity of thyroid hormone is dependent on the
synthesis of T3 from its inactive prohormone thyroxine (T4). The enzymatic conversion of T4 to T3
is dependent on the production of the enzyme thyroxine deiodinase 2 (DIO2), which cleaves an
iodine moiety from the outer ring of T4 (Watanabe et al., 2004). A second deiodinase, DIO3, can
inactivate T3 by cleaving a second iodine molecule, to synthesize diiodothyronine (T2).
Photoperiod-dependent melatonin release regulates the expression of both DIO2 and DIO3, as well
28
as the release of thyroid stimulating hormone (TSHB), which is responsible for the initial production
of T4 (Yoshimura, 2013). The production of DIO2 from specialized tanycyte cells in the ependyma
is increased, leading to greater T3 production (Dardente, 2015). Alternatively, DIO3 production is
increased during short-day photoperiods, leading to inactivation of T3 (Barrett et al., 2007). Both
mechanisms affect the season-dependent release of gonadotrophins, leading to the annual rhythm
of fertility (Figure 2-3).
Kisspeptin, a hypothalamic peptide that controls the release of GnRH via estrogen
feedback, may also play an intermediate role in modulating seasonal breeding pattern of sheep.
Short-day photoperiod downregulated kisspeptin gene expression, and subsequently reducing
negative feedback on GnRH secretion from the hypothalamus (Clarke et al., 2009; Castro, 2016).
In male hamsters, treatment with exogenous kisspeptin was able to restore reproductive function
during short-day lighting (Bailey et al., 2017). This effect appears to be melatonin-dependent
because the photoperiodic effect does not occur after pineal ablation (Revel et al., 2006). Bartzen‐
Sprauer et al. (2014) demonstrated that in addition to kisspeptin, its co-activators neurokinin B and
dynorphin are also inhibited during short-day lighting, potentially helping to direct the seasonality
of reproduction. While the relationship between photoperiodic melatonin signaling and kisspeptin
production is clear, the mechanism governing this relationship has yet to be discovered.
Seasonal Changes in Feeding Behavior and Metabolism
In the winter, reduced sunlight and cold temperatures lead to dormancy in most plants. As
a result, the availability of feed for herbivorous mammals is reduced. To avoid winter starvation,
these animals must alter their metabolism to prepare for food shortage. In modern livestock farming
systems, issues related to winter feed availability have been resolved through feed storage systems
such as ensiling forages, hay production, and the drying of cereal grains. However, the seasonal
rhythm of forage production is still relevant for rangeland animals, and many of the physiological
29
mechanisms that animals have adapted to cope with seasonal feed availability may still exist.
Furthermore, the seasonal changes in plant growth can have tremendous consequences on the
nutrient composition of feed harvested at different times of the year. In the northern United States,
digestibility of most legumes declines with the high-temperatures of mid-summer because of
increased lignification of stalks, causing second and third cuttings of alfalfa to be less digestible
than alfalfa harvested at first cutting (Van Soest, 1994). Understanding how yearly changes affect
feeding behavior and metabolism in domestic ruminants may provide management strategies to
improve animal efficiency.
In many wild animals, food intake and body weight exhibit dramatic fluctuations
throughout the year. The body weight of golden-mantled ground squirrels (Callospermophilus
lateralis) peaks from November to February and then declines to a nadir in June and July (Zucker
and Boshes, 1982). Pallid bats (Antrozous pallidus) have substantially greater body weight from
October to December than from April to September (Beasley et al., 1984). These seasonal changes
in feed intake appear to also be present in domesticated livestock. Recently, Ueda et al. (2016)
reported that total rumen dry matter and fiber pool size were greater in the autumn than in the spring
in dairy cows, with a reciprocal relationship of volatile fatty acid concentration within rumen fluid.
Alternatively, grazing sheep have been shown to increase feed intake in the spring (Milne et al.,
1978). While these changes may be related to changes in quality of pasture across the year, evidence
suggests that annual patterns of feed intake occur in non-pasture raised animals. Feedlot beef cattle
fed stored forage also have greater feed intake in the fall than in the spring (Mujibi et al., 2010).
Together, these results may indicate that cattle have an annual rhythm of feed intake.
Leptin may be one factor responsible for the seasonal changes in feed intake. Leptin is a
hormone produced by adipose tissue that acts as a satiety signal. Leptin secretion of male
woodchucks (Marmota monax) peaks in the spring, causing a concurrent minima of yearly feed
intake (Concannon et al., 2001). Sheep exhibit seasonal changes in leptin, with greater leptin
30
concentrations occurring in August compared to January (Henry et al., 2010). Rousseau et al.
(2003) discovered that the effect of leptin on adiposity in Siberian hamsters (Phodopus sungorus)
is affected by photoperiod, with long-day housed hamsters being more resistant to lipolysis-
inducing effects of leptin. However, administering exogenous melatonin appears to have no effect
on the annual rhythms of plasma leptin concentrations, suggesting that the photoperiodic effects on
leptin concentration may be melatonin-independent (Mustonen et al., 2005). Despite the intriguing
breadth of evidence correlating seasonal feed intake to leptin secretion, mechanisms governing this
relationship have not been explored.
Metabolic function is also under the influence of seasonality. An extreme example of the
seasonal role in metabolic rate occurs in animals living in polar climates. Many arctic animals adapt
to cold temperature by restricting blood flow to their extremities to maintain a higher core body
temperature. In species such as arctic wolves, foxes, muskrats, and reindeer this can mean that the
temperature of their feet can drop to near 0°C (Brix et al., 1990). Caribou maintain blood fluidity
in their cold extremities by increasing the concentration of fatty acids, especially unsaturated fatty
acids, in bone marrow (Meng et al., 1969). These changes are also often accompanied by a decrease
in metabolic rate and activity to conserve body fat reserves (Scholander et al., 1950). Many non-
arctic animal also exhibit seasonal changes in metabolic rate. The wild Przewalski horse (Equus
ferus przewalskii) displays an annual rhythm of basal metabolic rate with a peak occurring in
August (Arnold et al., 2006), and the North American moose (Alces alces) exhibits a 1.4-fold
increase in resting heat production during the summer compared to the winter (Regelin et al., 1985).
There has been a large amount of research examining the effect of photoperiod on milk
production dairy cattle. It has been known for several decades that a long light period increases dry
matter intake, milk production, and weight gain of dairy cattle (Peters et al., 1978, 1981). Dahl et
al. (1997) observed that exposing dairy cows to a long day photoperiod (18 h light: 6 h dark)
increased milk yield and concentrations of circulating IGF-1 compared to the natural winter
31
photoperiod of < 13 h of light. On the other hand, exposure to a short-day photoperiod in non-
lactating cows during the dry period before calving improved milk production in the subsequent
lactation (Miller et al., 2000). These production responses are likely due to the photoperiodic effects
on prolactin signaling. Crawford et al. (2015) demonstrated that the positive effects of long-day
lighting during lactation could be mimicked by the addition of exogenous prolactin, suggesting that
it is the primary driver of the effects of photoperiod on milk synthesis.
Despite many species demonstrating photoperiod-driven seasonal rhythms in feed intake
and metabolism in mammals, there have been few studies determining if these rhythms can be
maintained under constant conditions. Brinklow and Loudon, (1990) established that Red Deer
(Cervus elaphus) exhibit seasonal rhythms of feed intake even when placed in constant 18h light:6
h dark cycles and fed ad libitum. Conceivably, a circannual rhythm of feed intake and metabolism
could exist and be driven by endogenous yearly rhythms of prolactin. Prolactin plays a role in feed
intake as well as the initiation and maintenance of lactation in dairy cows (Bauman and Currie,
1980; Teyssot et al., 1981; Lawrence et al., 2000). Prolactin is released from the pars distalis (PD)
of the anterior pituitary in response to melatonin. An inherent yearly rhythm of prolactin
concentration is maintained in a constant photoperiod (Martinet et al., 1992). The signal that directs
seasonal prolactin release based on photoperiodic information has yet to be identified, but a
hypothetical ‘tuberalin’ protein is theorized to stimulate the release of prolactin (Foster and
Kreitzman, 2009b). Dupré et al. (2010) identified the tachykinins Substance P and Neurokinin A
as potential ‘tuberalin’ proteins. The Tac1 gene, which encodes both proteins was shown to increase
expression during long-day photoperiods, and both are secretagogues for prolactin.
Seasonal Changes in Immunity
Activation of the immune system is a highly energy-demanding process. It is estimated that
infection can lead to a 50% increase in basal metabolic rate (Lochmiller and Deerenberg, 2000).
32
Therefore, the animal must balance energy efficiency with the ability to mount an appropriate
immune response. Many infectious agents follow distinct yearly patterns. Seasonal variables such
as temperature, humidity, and sunlight, as well as host behavior, such as social interaction, all affect
the survivability of pathogens (Grassly et al., 2006). Influenza is one of the most well-known
examples of a seasonal disease in humans, with outbreaks occurring most frequently in the late fall
and winter of temperate climates (Tamerius et al., 2013). In pigs pneumonia infections are most
likely to occur during late winter and spring (Cowart et al., 1992). In the winter and spring, bovine
viral diarrhea (BVD) incidence is the greatest (Radostits and Littlejohns, 1988). The yearly pattern
of disease exposure creates predictable patterns that may be exploited by the immune system of
animals. By keeping the immune system ‘primed’ to respond to the pathogens that are most likely
to be encountered, it can most efficiently allocate resources without expending unnecessary energy.
Photoperiod can alter the growth and function of many immune cells. The number of
circulating leukocytes, B cells, T cells, and natural killer (NK) cells are increased by short day
photoperiod (SDPP) in Siberian hamsters (Bilbo et al., 2002). Furthermore, SDPP increases the
ability of NK cells to produce cytotoxicity via oxidative burst (Yellon et al., 1999). Non-lactating
dairy cattle increase neutrophil chemotaxis and lymphocyte proliferation in response to SDPP
(Auchtung et al., 2004). Additionally, season of the year can affect the functionality of specific
immune cells. Mann et al. (2000) observed a greater Th1: Th2 ratio in Rhesus macaques (Macaca
mulatta) in the summer versus winter. Alternatively, the response of polymorphonuclear blood
monocytes (PBMCs) to IL-2 stimulation was greater in the winter than summer.
One mechanism of photoperiodic changes in immune function may be melatonin signaling
in immune cells. Many immune cells including B cells, T cells, and PBMCs have membrane and
nuclear melatonin receptors (Pozo et al., 1997, 2004). The binding of melatonin to these receptors
can have multiple effects on the immune system. Treatment with exogenous melatonin increases
the systemic expression of IL-1α IL-1β, IL-5, IL-6, and IL-10 in endotoxemia-challenged mice
33
(Effenberger-Neidnicht et al., 2014). Removal of endogenous melatonin signaling via a
pinealectomy decreased the ability for splenocytes to react to antigen stimulation, and caused a near
depletion of lymphocytes (Maestroni et al., 1986). Melatonin itself can function as an antioxidant
by scavenging free radicals to ameliorate oxidative stress (Mayo et al., 2003).
The seasonal change in vitamin D may be another possible explanation for the effects of
photoperiod on the immune system. Vitamin D receptors (VDR) are present in nearly all immune
cells including macrophages, dendritic cells, neutrophils, T cells, and B cells (Baeke et al., 2010).
Photoperiod affects the function of vitamin D through sunlight-mediated vitamin D activation.
Briefly, ultraviolet radiation from the sun catalyzes the conversion of 7-dehydrocholesterol to
cholecalciferol (Vitamin D3) in the epidermal layer of the skin. After two successive hydroxylation
steps in the liver and kidney, respectively, the active form of vitamin D (1,25-dihydroxyvitamin
D3) is available for binding to the VDR (Horst et al., 2005). The role of sunlight in activation
implies a strong photoperiodic effect on vitamin D metabolism. Indeed, 1,25-dihydroxyvitamin D3
shows a strong yearly rhythm with plasma concentration peaking in the summer (Stumpf and
Privette, 1991).
While roles of melatonin and vitamin D may imply that yearly changes in immune function
are a result of strictly exogenous photoperiod, there is evidence to suggest an underlying
endogenous circannual rhythm of immune function. Brock detected a circannual rhythm in the
proliferation of B and T cells in C57BL/6 mice held in a constant photoperiod (12h light: 12h dark)
and temperature (22.5°C) (Brock, 1983). A similar experiment conducted in dogs under the same
photoperiod length also saw a circannual rhythm of total leukocytes and lymphocytes (Sothern et
al., 1993). These experiments provide some credence to the idea of an inherent circannual rhythm
of immune function, however, no mechanism has been described.
34
Epigenetic Control of Seasonality
Emerging evidence suggest that epigenetic modifications are important for the
maintenance and synchronization of circannual rhythms. Stevenson and Prendergast (2013)
demonstrated that SDPP increased the amount of DNA methylation at the promoter region of the
dio3 gene in the hypothalamus of Siberian hamsters. This led to decreased expression of diodinase
3, and greater inactivation of thyroid hormone. The authors concluded that hypermethylation at the
dio3 promoter helps precipitate seasonal reproduction by inhibition of reproductive activity during
summer months (Stevenson and Prendergast, 2013). Interestingly, DNA methyltransferase and
global methylation were actually lower in hamsters under SDPP, likely indicating a myriad of other
epigenetic changes are under photoperiodic control in the hypothalamus. The same authors
observed that the same mechanism of dio3 methylation can occur in a cell-autonomous manner
within immune cells. The increased methylation of deiodinase 3 caused an accumulation of T3
within lymphocytes under SDPP (Stevenson et al., 2014). Furthermore, supplementing with
exogenous T3 caused lymphocytes to exhibit greater sensitivity to inflammation.
Stoney et al. (2016) provided additional evidence that epigenetic programming may play a
role the seasonal changes of the immune system. They observed that expression of HDAC 4, 5, and
9 were found to be greater in mice under long-day (16h L: 8h D) versus short-day (8h L: 16h D)
photoperiod (Stoney et al., 2017). These histone deacetylases were shown to reduce the expression
of nuclear factor-κB (NF-κB), a master regulator of inflammation, during SDPP. Winter-related
behaviors also appear to be under the influence of epigenetic changes. Alvarado et al. (2015)
demonstrated that thirteen-lined ground squirrels (Ictidomys tridecemlineatus) exhibit a reduction
in global DNA methylation within muscle and liver at the onset of winter torpor.
Stevenson and Lincoln, (2017) have proposed a model of epigenetic control of circannual
rhythms, which postulates that circannual information is programmed into cells during early stem
cell development. They believe that this mechanism forms an ‘epigenetic memory’ of circannual
35
information, which can be reprogrammed in adult cells if seasonal inputs change. In insects, it
appears that seasonal information can be passed transgenerationally. Nasonia wasps (Nasonia
vitripennis) transmit seasonal signals of diapause from mother to an undifferentiated egg, which
was recently discovered to be a result of photoperiod-dependent DNA methylation (Pegoraro et al.,
2016). A similar mechanism appears to occur in Sarcophaga flies, which demonstrate the ability
to transmit photoperiodic signals from mother to embryo during early larval development.
Conclusions
Across multiple time-scales, biological rhythms serve to coordinate physiology with the
external environment. These biological rhythms are highly conserved across multiple organism,
illustrating their importance. Nutrient metabolism, in particular, is highly influenced by both
circadian and circannual rhythms. Moreover, nutrients themselves are important for entrainment
of biological clocks. Desynchronization of the day-light cycle from the time of feed intake can
lead to metabolic diseases such as obesity and type II diabetes in humans. Moreover, matching
the time of feed intake with the timing of the biological clock in livestock species may serve as an
opportunity to improve the efficiency of animal production.
36
Figures
Figure 2-1. Parameters used in the quantification of biological rhythms.
Period is the length of time to complete one cycle of the rhythm. Amplitude is the difference between peak
and mean. Double amplitude is the difference between peak and trough. Acrophase is the time at peak.
Bathyphase is the time at trough. Phase advance refers to shifting the rhythm curve so that the acrophase
occurs earlier than it did previously. Phase advance refers to shifting the rhythm curve so that acrophase
occurs later than it did previously.
37
Figure 2-2. Structure of the mammalian transcriptional circadian clock.
Key: BMAL1: Brain and muscles ARNT-like 1, CCG: Clock controlled gene, CK1ε: Casein kinase 1ε,
CLOCK: circadian locomotor output cycles kaput, CRY: Cryptochrome, E-Box: Enhancer-box
[canonically CANNTG], P: Phosphate group, PER: Period, RORα: Retinoic acid-related orphan receptor
α, RORE: Retinoic acid-related orphan receptor response element. Adapted from (Green et al., 2008).
38
Figure 2-3. Physiological mechanisms governing seasonal reproduction.
Key: DIO2: diodinase 2, GnRH: gonadotropin releasing hormone, MEL: Melatonin, NE: norepinephrine,
PT: Pars tuberalis, SCN: Suprachiasmatic nucleus, T3: Triiodothyronine, TSH:Thyroid stimulating
hormone, UV: Ultraviolet.
39
Chapter 3
The effects of day- versus night-restricted feeding of dairy cows on the daily
rhythms of feed intake, milk synthesis and plasma metabolites
Isaac J. Salfer and Kevin J. Harvatine
This chapter will be submitted as a manuscript to the Journal of Dairy Science.
40
ABSTRACT
The timing of feed intake can alter circadian rhythms of peripheral tissues. Dairy cows
have well-recognized daily rhythm of milk synthesis, but the effect of feeding time on these
rhythms is poorly characterized. The objective of this experiment was to determine if the time of
feed intake modifies the daily patterns of milk synthesis, key plasma metabolites, and body
temperature in dairy cows. Sixteen Holstein cows were randomly assigned to one of two
treatment sequences in a cross-over design with 17 d periods. Treatments included day-restricted
feeding (DRF) with feed available from 0700 to 2300 h and night-restricted feeding (NRF) with
feed available from 1900 to 1100 h. To determine a rhythm of milk synthesis, cows were milked
every 6 h on the last 7 d of each period and blood samples were collected to represent every 4 h
over the day and analyzed of glucose, insulin, non-esterified fatty acids (NEFA), and plasma urea
nitrogen (PUN). Body temperature was measured using indwelling vaginal temperature data
logger. The daily rhythms of milk yield and composition were modified by timing of feed
availability. Peak milk yield was shifted from morning in DRF to evening in NRF, while milk fat,
protein, and lactose concentration peaked in the evening in DRF and the morning in NRF. Plasma
glucose, insulin, NEFA, and PUN fit daily rhythms in all treatments. Night-feeding increased the
amplitude of glucose, insulin, and NEFA rhythms and shifted the daily rhythms by 8 to 12 h (P <
0.05). Night feeding also phase delayed the rhythm of core body temperature and DRF increased
its amplitude. Altering the time of feed availability shifts the daily rhythms of milk synthesis and
plasma hormone and metabolite concentrations and body temperature, suggesting that these
rhythms may be entrained by food intake.
Keywords: daily rhythm, milk synthesis, food entrainment, circadian
41
INTRODUCTION
Feeding behavior, milk synthesis, and plasma metabolites of dairy cows follow daily
patterns that may be regulated by endogenous circadian rhythms (Giannetto and Piccione, 2009).
Circadian rhythms are repeating cycles of about 24 h that govern many physiological functions
and allow organisms to anticipate changes in their environment. They are regulated at both the
systems level, through neural and humoral communication between tissues, and at the cellular
level, through transcriptional-translational feedback loops of ‘clock’ transcription factors and 24-
h protein phosphorylation cycles. In mammals, the master circadian pacemaker is located in the
suprachiasmatic nucleus (SCN) of the hypothalamus and is entrained by light through visual and
non-visual photoreceptors in the eye (Panda et al., 2002). However, food intake can entrain
circadian rhythms of peripheral tissues such as liver and adipose tissue, independent of the SCN
(Casey and Plaut, 2012).
Milk synthesis follows a daily rhythm in dairy cows with peak milk yield occurring in the
morning, and milk fat and protein concentration peaking in the evening (Rottman et al., 2014).
This adaptation may have evolved to provide neonates with energy-dense milk at night when
environmental temperature is lower and nursing and foraging activity is minimal. Evidence for a
role of the circadian system in lactation has been demonstrated in both rodents and humans, with
over 7% of the human mammary transcriptome oscillating in a circadian manner (Maningat et al.,
2009; Casey and Plaut, 2012).
Dairy cows also exhibit a crepuscular pattern of feed intake, consuming the majority of
their feed during the daylight hours with large bouts of intake in the morning and evening
(Albright, 1993). Furthermore, intake is stimulated by feed delivery and milking (DeVries et al.,
2005). Although most commercial dairy farms feed a total mixed ration (TMR) which is meant
provide cows with a consistent concentration of nutrients across the day, the daily pattern of
42
intake and sorting of feed cause nutrient intake and absorption to fluctuate diurnally (Rottman et
al., 2015). Moreover, the daily patterns of important peripheral hormones and metabolites such as
glucose, urea, nonesterfied fatty acids (NEFA), and insulin can be altered by the time of feed
intake (Nikkhah et al., 2008; Niu et al., 2014).
While there is evidence supporting a role of the mammary clock in lactation, little is
known about the relationship between feeding and daily rhythms of milk synthesis. Rottman et al.
(2014) observed that the amplitudes of milk fat and protein concentration are reduced by feeding
cows 4x/d compared to 1x/d. However, the response of the daily pattern of milk synthesis to
altering the timing of nutrient intake has not been examined. Time-restricted feeding is often
employed when studying food entrainment because it creates more robust changes in daily
rhythm compared to altering feeding time alone (Stokkan et al., 2001). The objectives of this
study were to determine the effects of the time of time-restricted feeding on daily rhythms of milk
synthesis and the daily rhythms of plasma hormones and metabolites and body temperature. Our
hypothesis was that restricting feed availability to 16 h over the dark period would invert the daily
rhythm of milk synthesis relative to feed available for 16 h during the light period and result in a
commensurate change in the daily pattern of plasma metabolites, with smaller changes in the
daily rhythm of body temperature.
MATERIALS & METHODS
Animals and Treatments
All experimental procedures were approved by the Pennsylvania State University
Institutional Care and Use Committee. Sixteen multiparous Holstein cows (183 ± 103 d
postpartum; mean ± SD) from the Pennsylvania State University Dairy Research and Teaching
Center were randomly assigned to one of two treatments sequences (n = 8 cows per sequence) in
43
a cross-over design. Treatment periods were 17 d and included 10 d of treatment adaptation with
2x/d milking and 7 d of 4x/d milking. Animals were housed in individual tie-stalls and had ad
libitum access to water. Cows were maintained in a 19 h light and 5 h dark photoperiod with
lights on from 0500 h to 0000 h. Treatments included (1) day-restricted feeding (DRF) with feed
available for 16 h/d from 0700 to 2300 and (2) night-restricted feeding (NRF) with feed available
for 16 h/d from 1900 to 1100 (Figure 3-1A). All cows were fed the same diet offered at 110% of
the previous day’s intake and intake was recorded daily. Feed was mixed once daily at 0700 and
immediately delivered to the DRF group. The remaining feed was compressed into plastic barrels,
covered with plastic, and stored at ambient temperature until feeding to NRF at 1900 h. Feed
samples were collected on d 7 and 14 of each period, composited by period and analyzed for dry
matter, crude protein, fiber, neutral detergent fiber, acid detergent fiber, and ash concentrations
according to Rico and Harvatine (2013) (Supplemental Table S3-1). The experiment was
conducted from February to March of 2016.
Milk Sampling and Analysis
Cows were milked 4x/d (0500, 1100, 1700 and 2300 h) for the last 7 d of each period to
observe the daily rhythm of milk synthesis. Milk collected at each time point represented milk
synthesis during previous 6 h, and data were expressed as the midpoint of the previous milking
interval (3 h prior to collection). Milk yield was measured at each milking using an integrated
milk meter (AfiMilk MPC Milk Meter; Afimilk Agricultural Cooperative Ltd., Kibbutz Afikim,
Israel) and corrected for the deviation of each individual stall according to Andreen et al. (2018).
Milk was sampled at all milkings on the last 2 d of each period. One subsample was stored at 4°C
with a preservative (Bronolab-WII; Advanced Instruments, Inc., Noorwood, MA) prior to
analysis of fat and protein concentration by Fourier transform infrared spectroscopy (Fossomatic
4000 Milko-Scan and 400 Fossomatic, Foss Electric; Dairy One DHIA). A second subsample was
44
stored without preservative at 4°C and centrifuged at 2300 x g within 12 h. The resulting fat
cakes were stored at -20°C and analyzed for concentrations of individual fatty acids (FA)
according to Baldin et al. (2018).
Feeding Behavior Observation and Analysis
The daily pattern of feed intake was monitored in nine of the 16 cows using an automated
feed observation system described by Rottman et al. (2015). Briefly, feed in hanging feed tubs
was weighed and recorded every 10 s from d 8 to 17 of each period by an electronic load cell
connected to a data acquisition system. To characterize feeding behavior, the number, length and
size of meals as well as the intermeal interval, eating time, and eating rate were determined as
described Niu et al. (2017). Hunger ratio was calculated as meal size divided by the previous
intermeal interval and satiety ratio was determined as meal size divided by the ensuing intermeal
interval. To characterize the rate of feed intake across the day, the data were smoothed by
calculating a running average over 180 s, the rate of feed intake was the determined over 10
minute intervals before averaging across 2 h intervals as described by Rottman et al. (2015).
Plasma Sampling and Analysis
Blood was collected by venipuncture of a coccygeal vessel using potassium EDTA
vacuum tubes (Greiner Bio-One North America, Inc., Monroe, NC). Samples were collected at
six time points on d 15 to 17 of each period to represent every 4 h across the day (0300, 0700,
1100, 1500, 1900, and 2300 h). Blood was immediately placed on ice and centrifuged within 1 h
at 2300 x g for 15 min at 4°C. Plasma was collected and stored at -20°C for analysis of glucose,
NEFA, plasma urea nitrogen (PUN), and insulin concentrations according to Rottman et al.
(2014). Briefly, plasma glucose concentration was analyzed using a glucose oxidase/peroxidase
enzymatic colorimetric assay (No. P 7119, Sigma-Aldrich, St. Louis, MO), NEFA concentration
45
was measured with an using acyl-CoA oxidase/peroxidase enzymatic colorimetric assay (NEFA-
HR (2), Wako Diagnostics, Richmond, VA), PUN was assayed using a modified Berthelot
methodology (Modified Enzymatic Urea Nitrogen Procedure No. 2050; Stanbio Laboratory,
Boerne, TX), and insulin was determined using a porcine 125I-insulin radioimmunoassay kit with
correction based on bovine insulin (PI-12K Porcine Insulin RIA, EMD Millipore, Billerica, MA).
Body Temperature Recording
Core body temperature was recorded every 10 min on d 12 to 17 of each period using an
intravaginal temperature data logger. A miniature plastic-coated thermometer (iBCod; Alpha
Mach Inc., Sainte-Julie, QC, Canada) was fastened to a progesterone-free CIDR vaginal implant
(Zoetis, Inc., Parsippany-Troy Hills, NJ) and placed centrally in the vagina of the cows using an
insertion tool. Raw data was averaged over 2 h intervals.
Statistical Analysis
All statistical analysis was performed using the MIXED procedure of SAS 9.4 (SAS
Institute, Cary, NC). Daily DMI, milk production, and FA yields were analyzed using the fixed
effects of treatment, period, and the interaction of treatment by period and the random effect of
cow. Repeated measures analysis was performed on data observed multiple times per day to test
the interaction of treatment and time of day on milk production, rate of feed intake, and plasma
metabolites and hormones. The model included the fixed effect of treatment, period, and the
interaction of treatment and period, the random effect of cow, and the repeated effect of time of
day. The AR(1) or ARH(1) covariance structure was selected based on convergence criteria, and
denominator degrees of freedom were adjusted using the Kenward-Roger method. Preplanned
contrasts tested the effect of treatment at each time point.
46
Time course data for milk production, plasma hormones and metabolites, and body
temperature were fit to the linear form of a cosine function using random regression in SAS 9.4.
The model included the fixed effect of treatment, cosine terms, and the interaction of treatment
with the cosine terms and the random effects of cow and period. A 12 h harmonic was also tested
for the daily patterns of plasma hormones and metabolites and body temperature but was removed
because it did not improve model fit according to Bayesian information criterion. The fit of the 24
h cosine was determined using a zero-amplitude test, an F-Test comparing a full model
containing the linear form of the cosine function to a reduced linear model. The acrophase (time
at peak) and the amplitude (difference between peak and mean) of the 24 h rhythm were
calculated and significance determined as described by Niu et al. (2014). In all analyses, points
with Studentized residuals outside of ± 3.5 were removed. Statistical significance was declared at
P < 0.05 and trends acknowledged at 0.05 < P < 0.10. High resolution figures were generated
using an add-in for Microsoft Excel (Kraus, 2014).
RESULTS
Eating Behavior
Total daily dry matter intake was not affected by the time of feed restriction (Table 1).
Both treatments consumed the greatest amount of feed in the first 2 h following feed delivery
(Figure 3-1B). However, the NRF treatment had a higher rate of intake after feed delivery,
consuming 21.6% (9.6 kg) of their feed per hour in the first two hours after feed delivery
compared to 15.6% (5.9 kg) in the DRF group (P < 0.0001).
The time of feed availability had no effect on meal size, number of meal bouts per day,
eating time/d, eating rate, or average intermeal interval (Supplemental Figure S3-1). However,
47
DRF increased average meal length 5.1 min/d (P = 0.02) and increased hunger ratio 40% (P =
0.01), while the satiety ratio was not modified.
Rhythm of Milk and Milk Components
The timing of feed restriction did not alter daily yield of milk and milk fat and protein or
daily average concentration of milk fat and protein (Table 1). However, milk yield fit a cosine
function with a 24 h period in both DRF (P = 0.04) and NRF (P < 0.0001; Figure 3-2A). The
acrophase of milk yield was delayed 8.1 h by NRF (P < 0.05), and the amplitude of the rhythm
was 32% greater in DRF than NRF.
Night-restricted feeding decreased total daily milk fat yield (P = 0.03; Table 1). Milk fat
yield fit a daily rhythm with an amplitude of 16.5 g and an acrophase at 0703 in NRF (P <
0.0001), but did not fit a 24 h rhythm in the DRF (Figure 3-2B). Milk fat concentration fit a 24 h
rhythm in NRF (P < 0.0001) and tended to fit a daily rhythm in DRF (P = 0.06; Figure 3-2C).
There was an effect of treatment on the rhythm, as milk fat concentration was phase shifted,
peaking at 1156 h in DRF compared to 0207 h in NRF (P < 0.05) and the amplitude was
increased by 64% in NRF compared to DRF (P < 0.05).
Milk protein yield fit a daily rhythm in both DRF and NRF (P < 0.001; Figure 3-2D)
with the peak of the rhythm occurring at 0211 h in DRF and at 0738 in NRF (P < 0.05). The
amplitude of the daily rhythm of protein yield was 30.1 g in DRF and was decreased to 10.4 g by
NRF (P < 0.05). Both DRF and NRF also exhibited daily rhythm of milk protein concentration (P
< 0.02; Figure 3-2E). Treatment modified the timing of the rhythm with DRF peaking at 1525 h
and NRF peaking at 0029 h (P < 0.05). The robustness of the rhythm was increased over two-fold
by NRF, which had an amplitude of 0.11% compared to 0.05% in DRF (P < 0.05).
48
Milk FA Yield and Profile
Compared to NRF, day-restricted feeding increased FA originating from de novo
synthesis in the mammary gland by 5% (Σ < 16C; P = 0.03) and mixed source FA that originate
both from de novo synthesis and preformed FA uptake from plasma by a 9% (Σ 16C; P = 0.001;
Figure 3-3A). However, no difference in yield of preformed FA (Σ >16C) was observed between
treatments.
Both de novo and mixed source FA fit a cosine function with a period of 24 h in both
treatments, but preformed FA only fit a cosine function in DRF (P < 0.05; Figure 3-3D-F). The
acrophase of the de novo FA rhythm was delayed 6.7 h by NRF (0108 vs 0750 h; P < 0.05) and
DRF increased the amplitude of the rhythm by 38% (P < 0.05). The acrophase and amplitude of
the mixed FA yield rhythms were also altered by treatments, with DRF peaking at 0032 h and
NRF peaking at 0721 (P < 0.05), and NRF having a 34% greater amplitude than DRF (P < 0.05).
Similar to the de novo FA rhythm, night-restricted feeding increased the robustness of the mixed
FA yield rhythm, increasing its amplitude 34% compared DRF (P < 0.05). Of the 44 individual
fatty acids quantified, 15 fit a daily rhythm in both DRF and NRF, 10 fit a rhythm in DRF only, 6
fit a rhythm in NRF only, and 13 did not fit a rhythm in either treatment (Supplemental Table S2).
Fatty acids generally followed a similar daily pattern as their source grouping (de novo, mixed
source, preformed).
To assess the potential cause of the reduced de novo fatty acid synthesis, milk fat
concentrations of trans-10 C18:1 (t10) and trans-11 C18:1 (t11) were determined. trans-10 C18:1
is produced by the alternate biohydrogenation pathways and is elevated during biohydrogenation-
induced milk fat depression and t11 is a product of the normal biohydrogention pathway and is
increased with slowing of the normal biohydrogenation pathway (Griinari et al., 1998). The
concentrations of t10 and t11 were analyzed as a percent of 18 carbon FA, rather than as a percent
of total FA to avoid bias of changes in de novo and mixed FA. Night-restricted feeding increased
49
t10 by 16.3% (P = 0.02) and t11 by 32.8% (P = 0.001) compared to NRF (Figure 3-3B). Both t10
and t11 fit daily rhythms in both treatments (P < 0.0001; Figure 3-3G&H). Milk fat t10 was
phase delayed 12.1 h by NRF, with NRF peaking at 1038 and DRF peaking at 2234 (P < 0.05).
The amplitude of t10 was increased by NRF compared to DRF (0.08% vs. 0.06%; P < 0.05). Like
t10, night-restricted feeding caused a complete inversion of the daily rhythms of t11
concentration, shifting the peak from 1816 in DRF to 0604 in NRF (P < 0.05). The amplitude of
t11 was not affected treatment (P > 0.10).
Plasma Hormones and Metabolites
Plasma glucose fit a 24 h cosine in both the DRF (P < 0.03) and NRF (P < 0.001), with
NRF increasing the amplitude of the rhythm by 149% (P < 0.05; Figure 3-4A). Night-restricted
feeding also phase shifted plasma glucose, with the acrophase occurring 10.2 h later in DRF than
NRF (P < 0.05).
Plasma insulin concentration fit a 24 h cosine function in both treatments (P < 0.001),
with no difference in mean insulin concentration between DRF and NRF (P = 0.69; Figure
3-4B). Night-restricted feeding shifted the acrophase of insulin production 8.0 h and increased the
amplitude 0.9 μIU/mL compared to DRF (P < 0.05).
Concomitant with the shift in plasma insulin rhythms, the rhythms of plasma NEFA
concentration were dramatically altered by treatment (Figure 3-4C). Both treatments exhibited 24
h rhythms in plasma NEFA (P < 0.01), but the acrophase of the rhythms were markedly different,
with DRF rhythms peaking at 0614 and NRF peaking at 1657 (P < 0.05). Additionally, NRF
increased the amplitude of the NEFA rhythm 3.3 fold (P < 0.05). This increase in amplitude in
the NRF treatment was due to a dramatic rise in NEFA concentration to 214 μEq/L at 1900 h, 6
hours into the fasting period.
50
Plasma urea nitrogen fit a 24 h rhythm in DRF (P = 0.02) and tended to fit a rhythm in
NRF (P = 0.08). The rhythm of NRF was completely inverted with respect to the DRF, with DRF
peaking at 0209 h and NRF peaking at 1420 (P < 0.05). Moreover, the amplitude of the PUN
rhythm was 20% greater in DRF than NRF (P < 0.05).
Body Temperature
Body temperature fit a 24 h cosine function in both treatments (P < 0.05) and was
modified by timing of feed restriction (Figure 3-5). NRF delayed the phase of the rhythm 15.5 h
(P < 0.05) and increased the amplitude 75% compared to DRF (P < 0.05).
DISCUSSION
The rate of feed intake was greater after food delivery in the NRF compared to DRF,
which is consistent with previous results reporting that feed delivery 1x/d at night increases intake
in the first the 2 to 3 h after feeding compared to feed delivery in the morning (Nikkhah et al.,
2008; Niu et al., 2014). Cows naturally exhibit a crepuscular pattern of intake, with highest intake
occurring at dusk and at dawn and a decline in intake overnight, and TMR-fed cows normally
have high intake at feed delivery and the late afternoon and early evening and low intake in the
overnight period (DeVries and von Keyserlingk, 2005). In NRF, feed was withheld during the
high-intake afternoon period of the day (1100 to 1900 h), whereas DRF cows were without feed
during the low-intake night. The greater rate of intake immediately after feeding in NRF suggests
greater hunger signaling presumably due to the circadian pattern of hunger and satiety. Other
species display 24 h rhythms of hunger. Humans, for example, exhibit the greatest appetite in the
evening, and lowest in the morning, independent of sleeping time or food intake (Scheer et al.,
2013). Moreover, rats display circadian rhythms of meal size and meal frequency under constant
51
illumination (Rosenwasser et al., 1981). Despite the difference in the daily pattern of feed intake,
total dry matter intake did not differ between treatments because DRF compensated by increasing
intake in the afternoon compared to the corresponding period in NRF (0100 to 0700 h).
Dairy cows generally have greatest milk yield in the morning and greater fat and protein
concentration in the evening when fed ad libitum with feed delivered in the morning (Quist et al.,
2008; Rottman et al., 2014). Our results confirmed these findings, but demonstrated that night-
restricted feeding shifts this daily pattern 8 hours later in the day. The change in the rhythms of
milk and milk component synthesis in response to altered feeding time is either due to a change in
available substrate for milk synthesis or entrainment of the mammary gland’s circadian clock. In
FVB mice, time-restricted feeding shifted the rhythms of clock genes in the mammary gland, and
affects the rhythmic gene expression of transcription factors (SREBP1c, Spot 14) and enzymes
(SCD1 and FASN) related to milk fat synthesis (Unpublished data). These results suggest that
alterations in the mammary molecular clock may mediate the response of feeding time on
rhythms of milk synthesis, but further research must be conducted to evaluate this effect.
The decrease in total milk fat yield and de novo and mixed fatty acids in NRF compared
to DRF indicates that de novo fatty acid synthesis was reduced by night-restricted feeding.
Concomitant with the decrease in apparent de novo fatty acid synthesis, the concentration of t10
C18:1 was elevated in NRF. Trans-10 C18:1 is highly correlated with reduced milk fat synthesis
in dairy cows as it is associated with production of bioactive intermediates of the trans-10
pathway (Lock et al., 2007). However, t11 C18:1, the intermediate of normal ruminal
biohydrogenation was also increased by NRF, which commonly occurs due to a slowing of the
normal pathway in conjunction with the shift to the alternate BH pathway (Rico and Harvatine,
2013). The decrease in ruminal biohydrogenation in NRF may be related to the daily pattern of
intake. The NRF group consumed a large percentage of their feed during the first 2 hours after
delivery, while DRF had more stable intake across the day. Previous research demonstrated that
52
stabilizing feed intake through 4x/d feeding increased milk fat yield and decreased t10 C18:0
(Rottman et al., 2014). The changes in daily patterns of de novo, mixed source, and preformed
FA closely reflected the daily pattern of milk yield. The daily pattern of t10 C18:1 appeared to be
highly impacted by time of feed restriction, peaking at the start of the fasting period in both
treatments. Trans-11 C18:1, meanwhile, peaked in the middle of the feeding period, 13 hours
after feed delivery in both treatments.
The effect of time of feed restriction on daily rhythms of fatty acids may also be due to
differences in entrainment of the mammary circadian clock. Previous research from other models
suggests circadian regulation of lipid metabolism. In lactating rats, mammary lipogenesis and
activity of acetyl-CoA carboxylase (ACC), the enzyme that catalyzes the rate-limiting step in FA
synthesis, follow clear daily pattern with a nadir at 1500 h (Munday and Williamson, 1983).
Furthermore, Hems et al. (1975) detected daily rhythms of FA synthesis in the livers of mice.
Similarly, the current study indicates that FA synthesis follows a daily pattern in the mammary
gland of cows. Several metabolic pathways linking the circadian clock to FA metabolism have
been characterized, including the peroxisome proliferator-activated receptor (PPAR) family of
nuclear receptors, which act as master regulators of lipid metabolism in multiple tissues and the
cellular energy sensor AMP-activated protein kinase (AMPK) influences the period of the
circadian clock while also increasing the expression of ACC (Jordan and Lamia, 2013; Chen and
Yang, 2014). The results of this experiment provide additional support that de novo FA synthesis
is controlled by the circadian rhythm of the mammary gland and suggest that this rhythm is
modified by timing of nutrient absorption.
Similar to previous reports in dairy cattle (Giannetto and Piccione, 2009) and other
species (Rudic et al., 2004), plasma glucose concentration fit a 24 h rhythm. Furthermore, both
glucose and insulin concentrations were phase shifted by NRF compared to DRF. Circadian
control over glucose metabolism in experimental models has been well-established. In mice, a
53
functioning circadian clock in pancreatic β-cells is required for maintenance of insulin sensitivity
(Sadacca et al., 2011). The amplitude of NEFA concentration was greatly increased in NRF
compared to DRF. This is likely a consequence of the daily pattern of hunger and satiety. Cattle
typically increase feed consumption during the afternoon, suggesting greater hunger signaling
during this time (Ray and Roubicek, 1971). The results of the current experiment suggest that
fasting during the high-intake afternoon period of the day causes greater lipid mobilization than
fasting during the low-intake overnight period. These results were similar to those reported in
rats, which displayed a shift in the daily pattern of insulin release when they were fasted during
the early part of their active period (Shimizu et al., 2018). Insulin in a potent inhibitor of hormone
sensitive lipase, the enzyme responsible for causing lipid mobilization from adipose tissue
(Locher et al., 2011). Peak NEFA concentration in both treatment groups coincided with the nadir
of insulin release. Shostak et al. (2013) reported that lipolysis from white adipose tissue follows a
daily pattern, and showed using ClockΔ19 and Bmal-/- mice that these patterns were controlled by
the molecular circadian clock. The current study provides evidence suggesting a similar
mechanism may occur in dairy cows.
The shift in body temperature by the time of time-restricted feeding likely demonstrates a
change in the central circadian rhythm. This result of the current study showed a more extreme
shift in the rhythm of body temperature than previously observed by Niu et al. (2014), who
reported a 3 hour phase delay after feeding at 2030 compared to 0830, without restricting the time
of feed availability. Our results also agreed with research performed in mice which showed shifts
in the body temperature rhythms when food was restricted to the inactive period (Damiola et al.,
2000). The change in body temperature may also be responsive to feeding pattern due to changes
in heat produced by rumen fermentation across the day. This may be a mechanism by which
ruminants can directly modify body temperature in response to feeding. Shifts in the rhythm of
body temperature may play a role in altering other daily rhythms within the animal because body
54
temperature has been shown to be capable of entraining peripheral circadian clocks in mammals
(Brown et al., 2002).
In conclusion, timing of feed intake dramatically altered daily rhythms of milk synthesis,
plasma hormones and metabolites, and body temperature. Modification of the phase of milk
synthesis by temporal changes in absorption of nutrients indicates possible changes in the
mammary molecular clock. Timing of nutrient intake also influences the central circadian clock,
evidenced by the shift in the body temperature rhythm. The fasting response was more dramatic
for cows on the NRF treatment, having a greater post-feeding feed consumption rate, and greater
pre-meal NEFA concentrations, which is likely due to the circadian rhythm of food intake. This
study supports the hypothesis that timing of nutrient absorption modified the daily rhythms of
milk synthesis, likely due to entrainment of the mammary circadian clock.
55
FIGURES
Figure 3-1. Effect of day- versus night-restricted feeding on the daily pattern of feed intake.
Panels show: (A) Daily schedule of feed availability and milking time for day-restricted feeding (DRF;
feed available for 16 h/d from 0700 to 2300) and night-restricted feeding (NRF; feed available for 16 h/d
from 1900 to 1100) treatments and milking times. Cows were adapted to feeding schedules for 10 d prior to
7 d of 4x milking. (B) Effects of day versus night feed availability on the rate of feed intake (kg DM/h).
Data are presented as the least square means with standard error bars for every 2 h period. Preplanned
contrasts of the effect of treatment at each time point are shown (*P < 0.05).
56
Figure 3-2. Effect of day- versus night-restricted feeding on daily rhythms of milk yield and milk
components.
Treatments were feed available for 16 h during the day [day-restricted feeding (DRF); feed from 0700 to
2300] or feed available for 16 during the night [night-restricted feeding (NRF); feed from 1900 to 1100].
Data are presented as LSM with SEM bars. Panels show the effect of day vs. night-restricted feeding on the
daily pattern of A. milk yield (kg), B. milk fat yield (g), C. milk fat concentration (%), D. milk protein yield
(g), E. milk fat concentration (%). 1Amplitude- difference between peak and mean. 2Acrophase-time at peak
of the rhythm. 3P-value of the zero-amplitude test. The black and white bars above the x-axis display the
light: dark cycle.
57
Figure 3-3. Effect of day- versus night-restricted feeding on the daily production and daily pattern
of milk fatty acids.
Treatments were feed available for 16 h during the day [day-restricted feeding (DRF); feed from 0700 to
2300] or feed available for 16 during the night [night-restricted feeding (NRF); feed from 1900 to 1100].
Panels show effect of day vs. night-restricted feeding on (A) total daily yield of de novo (Σ <16C FA), mixed
(Σ 16C FA), and preformed (Σ >16C FA) FA (g/d), (B) daily average milk fat concentration of trans-10
C18:1 (t10) trans-11 C18:1 (t11; % of total C18 FA), (C) daily patterns of milk denovo FA yield (g/d), (D)
daily patterns of milk t10 concentration (% of total C18 FA), (E) daily patterns of milk mixed source FA
yields (g/d), (F) daily patterns of milk t11 concentration (% of total C18 FA), (G) daily patterns of milk
preformed FA yields (g/d). 1Amplitude- difference between peak and mean. 2Acrophase-time at peak of the
rhythm. 3P-value of the zero-amplitude test. Data are presented as LSM with SEM bars. The black and white
bars above the x-axis display the light: dark cycle.
58
Figure 3-4. Effect of day- versus night-restricted feeding on daily rhythms of plasma hormones
and metabolites.
Treatments were feed available for 16 h during the day [day-restricted feeding (DRF); feed from 0700 to
2300] or feed available for 16 during the night [night-restricted feeding (NRF); feed from 1900 to 1100].
Panels show effect of day vs. night-restricted feeding on (A) daily patterns of plasma glucose concentration
(mg/dL), (B) daily patterns of plasma insulin concentration (μIU/mL), (C) daily patterns of plasma NEFA
concentration (μEq/L), (D) daily patterns of plasma urea nitrogen concentration (mg/dL), (E) daily patterns
of milk mixed source FA yields (g/d), (F) daily patterns of milk t11 concentration (% of total C18 FA), (G)
daily patterns of milk preformed FA yields (g/d). Data are presented as LSM with SEM bars. 1Amplitude-
difference between peak and mean. 2Acrophase-time at peak of the rhythm. 3P-value of the zero-amplitude
test. The black and white bars above the x-axis display the light: dark cycle.
59
Figure 3-5. Effect of day- versus night- restricted feeding on the circadian rhythm of body
temperature in dairy cows.
Treatments were feed available for 16 h during the day [day-restricted feeding (DRF); feed from 0700 to
2300] or feed available for 16 during the night [night-restricted feeding (NRF); feed from 1900 to 1100].
Data presented as 2 h means and SEM bars of body temperature collected every 10 min by a vaginal
temperature data logger. 1Amplitude- difference between peak and mean. 2Acrophase-time at peak of the
rhythm. 3P-value of the zero-amplitude test. The black and white bars above the x-axis display the light:
dark cycle.
60
TABLES
Table 3-1. Effect of day versus night-restricted feeding on total daily DMI, milk yield, and milk
composition.
1Treatments were feed available for 16 h during the day [day-restricted feeding (DRF); feed from 0700 to
2300] or feed available for 16 during the night [night-restricted feeding (NRF); feed from 1900 to 1100.
Treatment1
Item DRF NRF SEM2 P-value
DMI, kg/d 23.7 22.1 1.15 0.18
Milk yield, kg 33.7 33.2 1.34 0.68
Fat
Concentration, % 3.54 3.45 0.16 0.54
Yield, g/d 1125 1061 83.3 0.03
Protein
Concentration, % 3.28 3.23 0.10 0.64
Yield, g/d 1059 1010 59.5 0.43
61
SUPPLEMENTAL FIGURES
Supplemental Figure S3-1. Effect of the day versus night-restricted feeding on feeding behavior.
Treatments were feed available for 16 h during the day [day-restricted feeding (DRF); feed from 0700 to
2300] or feed available for 16 during the night [night-restricted feeding (NRF); feed from 1900 to 1100].
Data are presented as LSM with SEM bars. Panels show effect of day vs. night-restricted feeding on A.
average meal size (kg) and average number of meal bouts/d, B. Meal length and intermeal interval (min),
C. eating time (min/d) and eating rate (kg/min), D. hunger and satiety ratios. Hunger ratio=meal
size/preceding intermeal interval. Satiety ratio=meal size/post-meal interval.
62
SUPPLEMENTAL TABLES
Supplemental Table S3-1. Diet and nutrient composition of the experimental diet.
1Contained (% of DM): 6.8% CP, 36.4% NDF, 22.8% ADF.
2Contained (% of DM): 20.3% CP, 47.6% NDF, 41.6% ADF. 3Contained (% of DM): 9.2% CP, 70.9% NDF, 43.4% ADF. 4Contained (%, as-fed basis): 36.8% Calcium carbonate; 29.0% dried corn distillers grains; 24.9% salt;
4.2% magnesium oxide (54% Mg); 2.5% organic phosphorus (15% P); 1% zinc sulfate; 0.5% mineral oil.
Composition (DM-basis): 10.6% CP; 43.2% NDF; 7.0% ADF; 14.4% Ca; 0.75% P; 15.1% Cl; 0.28% K;
2.5% Mg; 0.5% S; 9.8% Na; 23.0 mg/kg Co; 651 mg/kg Cu; 796 mg/kg Fe; 54.0 mg/kg I; 1190 mg/kg Mn;
12.8 mg/kg Se; 3434 mg/kg Zn; 195,290 IU/kg vitamin A (retinyl acetate); 62,500 vitamin D (activated 7-
dehydrocholesterol); 1864 IU/kg vitamin E (dl-α tocopheryl acetate). 5Fed as coated urea (Optigen, Alltech Inc., Lexington, KY; 259% CP, DM basis).
Item Composition
Ingredients, % of DM
Corn silage1 40.3
Alfalfa haylage2 19.3
Canola meal 13.2
Ground corn 12.6
Roasted soybeans 4.6
Molasses 3.8
Grass hay/straw3 3.3
Vitamin/mineral mix4 2.5
Non-protein nitrogen5 0.40
Nutrients, % of DM
NDF 32.5%
ADF 22.2%
CP 16.8%
Ash 6.9%
63
Supplemental Table S3-2. Effect of day- versus night-restricted feeding on total daily and daily rhythms of individual milk fatty acids.
g/d1 g/6 h1 Amplitude2 Acrophase3 P-Value4
Name DRF NRF DRF NRF DRF NRF DRF NRF DRF NRF
C4:0 47.1 44.7 11.9 11.7 - 0.94 - 0702 0.13 0.003
C6:0 26.1 24.5 6.63 6.44 0.37b 0.58a 0114a 0741b 0.08 0.002
C8:0 15.1 14.2 3.83 3.73 0.22b 0.39a 0044a 0735b 0.07 0.003
C10:0 36.4 34.2 9.19 8.97 0.64b 1.06a 0002a 0740b 0.01 < 0.001
C10:1 3.10 2.97 0.79 0.73 0.069a 0.046b 0307a 1007b 0.002 0.07
C11:0 1.03 0.92 0.26 0.23 0.041 0.044 2243b 0726a < 0.001 < 0.001
C12:0 42.5 39.7 10.7 10.4 0.78b 1.18a 0019a 0750b 0.007 < 0.001
C13:0 1.59 1.47 0.40 0.36 0.05 0.052 2336b 0747a < 0.001 < 0.001
C13:0 iso 0.17 0.16 0.042 0.038 0.012a 0.0069b 1027b 0628a < 0.001 < 0.001
C13:0 anteiso 0.88 0.83 0.22 0.20 0.021a 0.016b 0246a 1011b < 0.001 0.01
C14:0 129.5 120.7 32.7 31.2 2.22b 2.75a 0104a 0806b 0.01 0.001
C14:0 iso 0.87 0.75 0.23 0.19 0.013 - 0258 - 0.03 0.36
C14:1 10.3a 9.63b 2.61a 2.44b 0.23 0.15 0311a 0959b < 0.001 0.007
C15:0 14.0 13.0 3.52 3.25 0.32b 0.36a 0016a 0757b 0.001 0.001
C15:0 iso 2.27 2.08 0.57 0.54 - - - - 0.18 0.45
C15:0 anteiso 4.99 4.80 1.26 1.22 0.097 - 0322 - 0.006 0.24
C16:0 320.2a 290.4b 80.5 75.4 4.97b 7.08a 0109a 0721b 0.05 < 0.001
C16:0 iso 2.50a 1.97b 0.61a 0.52b - - - - 0.14 0.96
C16:1 13.5 12.7 3.37 3.21 0.29 - 0507 - 0.003 0.63
C17:0 6.47 6.39 1.60 1.61 - - - - 0.23 0.53
C17:0 iso 3.49 3.75 0.89 0.96 0.057 0.061 0324a 2345b 0.04 0.05
C17:0 anteiso 5.01 5.10 1.26 1.29 0.089 - 0423 - 0.03 0.75
C17:1 2.06 2.12 0.50 0.52 0.044a 0.030b 1823b 0731a 0.001 0.04
C18:0 117.3 114.1 29.8 29.9 - - - - 0.61 0.11
64
trans-4 C18:1 0.21 0.25 0.048 0.056 0.018a 0.014b 1044a 1246b < 0.001 0.001
trans-5 C18:1 0.10 0.12 0.022 0.029 0.0073 0.0089 1004 0924 0.04 0.005
trans-6,8 C18:1 3.27 3.30 0.82 0.85 0.049 - 0449 - 0.07 0.28
trans-9 C18:1 2.61 2.63 0.65 0.68 - - - - 0.65 0.68
trans-10 C18:1 5.23 5.82 1.30 1.45 0.11a 0.095b 0417a 0937b 0.001 0.02
trans-11 C18:1 8.87 10.3 2.23b 2.64a - 0.15 - 0447 0.93 0.04
trans-12 C18:1 5.15 5.09 1.29 1.30 0.088 - 0403 - 0.04 0.66
trans-15 C18:1 2.70 2.64 0.68 0.68 0.043a 0.028b 0444a 0746b 0.06 0.02
cis-9 C18:1 196.4 191.7 48.5 47.8 4.58 - 0613 - 0.001 0.16
cis-11 C18:1 9.10 8.92 2.27 2.20 0.24 - 0504 - < 0.001 0.24
cis-12 C18:1 4.45 4.65 1.10 1.18 0.081a 0.055b 0402a 1008b 0.02 0.06
C18:2, n-6 29.1 28.4 7.33 7.16 0.70 - 0547 - 0.001 0.50
cis-9, trans-11 C18:2 4.59 4.56 1.28 1.15 0.077 - 0252 - 0.08 0.15
C18:3 n-3 6.09a 5.65b 1.53a 1.40b 0.045 - 0115 - 0.99 0.28
C20:0 1.60 1.49 0.41 0.39 - 0.029 - 0801 0.78 0.005
C20:1 0.43 0.41 0.12 0.11 - - - - 0.11 0.14
C20:2 0.62 0.61 0.15 0.15 - - - - 0.50 0.97
C20:3 1.38a 1.24b 0.35a 0.32b - - - - 0.87 0.50
C20:4 2.05a 1.88b 0.52a 0.48b - - - - 0.77 0.39
C22:0 0.33 0.33 0.074 0.08 - - - - 0.45 0.98 1Least squared means of day-restricted feeding (DRF) with feed available from 0700 h to 2300 h and night-restricted feeding (NRF) with feed available
from 1900 to 1100 h. Means with different superscripts are different at P < 0.05. 2Difference from mean to peak of 24-h rhythm, g. Amplitudes with different superscripts are different at P < 0.05. 3Time at peak of the 24-h rhythm hhmm. Acrophases with different superscripts are different at P < 0.05. 4P-value of the zero-amplitude test corresponding to the fit of a 24-h rhythm. The rhythm fit is significant if P < 0.05 with a tendency for a rhythm at 0.05 <
P < 0.10. Amplitude and acrophase are not presented if the P-value of the zero-amplitude test is greater than 0.10.
65
Chapter 4
The effects of night-restricted feeding on the molecular clock of the
mammary gland in lactating dairy cows
Isaac J. Salfer, Paul A. Bartell, and Kevin J. Harvatine
66
ABSTRACT
Circadian rhythms are generated within tissues through a molecular clock made of a set
of transcription factors that oscillate in a 24 h manner. These rhythms can be entrained both by
photoperiod and feeding time. Dairy cows display daily rhythms of milk synthesis that are altered
by feeding time, but the role of molecular clocks in these rhythms is poorly understood. To
determine the impact of feeding on the mammary molecular clock, 11 lactating Holstein cows
were used in a crossover design that tested two feeding regimes, either (1) feed available for the
entire 24 h day (CON) or (2) night-restricted feeding where feed availability was limited to 16 h/d
from 2000 h to 1200 h (NR). All cows were housed in the same 19:5 light:dark cycle. Milk
samples were collected at 0700 and 1900 h on d 11 and 17 of each period and analyzed for fat and
protein concentration. Mammary tissue representing four times across the day (0400, 1000, 1600,
and 2200 h) was collected by needle biopsy. Expression of clock genes including brain and
muscle ARNT-Like 1 (BMAL1), circadian locomoter output cycles kaput (CLOCK),
cryptochrome 1 (CRY1), period (PER1), period 2 (PER2), and REV-ERBα was determined at each
time point using Real-Time RT-PCR. Blood was sampled on d 15 to 17 of each period to
represent every 4 h across the day (0200, 0600, 1000, 1400, 1800, and 2200 h), and analyzed for
concentrations of glucose, nonesterified fatty acids (NEFA), and plasma urea nitrogen (PUN).
Cosinor rhythmometry was performed using the MIXED procedure of SAS 9.4 to determine if
expression of clock genes and the concentrations of plasma metabolites fit a 24 h rhythm and if
the amplitude and acrophase (time at peak) were different between treatments. Night-restricted
feeding tended to reduce dry matter intake (P = 0.06) and decreased milk (P = 0.001), fat (P =
0.02) and protein yield (P < 0.001). Milk protein concentration was also reduced by NR (P =
0.01), but fat concentration was not changed. The expression of CLOCK fit a 24 h rhythm in
CON (P < 0.001) and tended to fit a rhythm in NR (P = 0.06), with NR increasing the amplitude
67
75% and delaying the phase by 6.5 h compared to CON (P < 0.05). The expression of CRY1 fit a
24 h rhythm in both treatments and NR increased the amplitude by 53% and delayed the phase by
3 h. A 24 h rhythm of REV-ERBα expression was not present in CON, but a tendency for a
rhythm was induced by NR (P = 0.06). The expression of BMAL1, PER1, and PER2 did not fit a
24 h rhythm in either treatment. Plasma glucose, NEFA, and PUN concentrations fit a 24 h
rhythm (P < 0.005). The amplitudes of the daily rhythms of plasma glucose and PUN
concentration were increased 124% and 62% by NR, respectively (P < 0.05), but the amplitude of
NEFA was not altered. Night-restricted feeding phase delayed plasma glucose concentration by
12 h and PUN concentration by 9.9 h, but did not shift the phase of the daily NEFA rhythm.
Results indicate that key components of the mammary molecular clock are influenced by the
timing of feed intake.
Keywords: molecular clock, food entrainment, circadian rhythm.
68
INTRODUCTION
All organisms have adapted a circadian timing system that allows them to anticipate daily
changes in their environment and correspondingly alter their behavior, metabolism, and cell
signaling. This system is arranged in a hierarchical manner with a master clock located in the
suprachiasmatic nucleus (SCN) of the hypothalamus, which is normally entrained by photoperiod
and sends neural and hormonal signals to synchronize biological clocks located within individual
cells of peripheral tissues (Plaut and Casey, 2012). Circadian oscillations in both the master clock
and peripheral cells are driven by a core set of clock transcription factors that maintain rhythms of
about 24 h by transcriptional-translational feedback loops (Panda, 2016). These transcription
factors include positive transcriptional regulators brain and muscle ARNT-like 1 (BMAL1) and
circadian locomotor output cycles kaput (CLOCK), which heterodimerize and bind to E-box
elements [canonically CANNTG] to activate transcription, and the negative transcriptional
regulators period (PER) and cryptochrome (CRY), which feedback to inhibit the expression of
BMAL1 and CLOCK. Additionally, an ancillary loop consisting of the orphan nuclear receptors
REV-ERBα and RAR-related orphan receptor-α (RORα) stabilize the transcriptional-translational
feedback loop by repressing or activating BMAL1 and CLOCK (Takahashi, 2017). Together,
these circadian transcription factors and nuclear receptors regulate the expression of a wide host
of clock-controlled genes, which are involved in diverse functions including growth, metabolism,
and immunity.
In addition to receiving signals from the master clock in the SCN, various other local and
systemic signals influence the oscillations of peripheral clocks. One of the most prominent of
these signals is the timing of feed intake. In other animals, the daily rhythms of locomotor activity
and blood glucocorticoid concentrations are entrained by restricting the time of feed availability
to the inactive phase of the day and these rhythms can persist several days after complete food
69
removal, even with ablation of the SCN (Stephan, 2002). Stokkan et al. (2001) established that
time-restricted feeding shifts the phase of the circadian clock of the liver, without a concomitant
effect in the SCN. This desynchronization of the rhythms of master and peripheral clocks has
been consistently observed during time-restricted feeding experiments in model organisms (Asher
and Sassone-Corsi, 2015).
Circadian rhythms appear to be important for metabolism within the mammary gland. In
several species, including humans, rodents and cattle, milk synthesis follows a daily pattern with
greater milk yield in the morning, and higher milk fat and protein concentration in the evening
(Plaut and Casey, 2012). Maningat et al. (2009) observed that 7% of the mammary epithelial
transcriptome follows circadian rhythms in humans. The circadian rhythm of the mammary gland
appears to be related to stage of lactation. Casey et al. (2014) observed an increase in BMAL1
and CLOCK protein abundance as well as a decrease in PER2 in differentiated cells compared to
undifferentiated cells. This is substantiated by evidence that short photoperiod during the dry
period increases mammary epithelial cell proliferation and increases milk synthesis in the
subsequent lactation (Wall et al., 2005). Furthermore, rate-limiting enzymes for fat, protein, and
lactose synthesis follow circadian rhythms in the mammary gland (Suárez-Trujillo and Casey,
2016).
Recent evidence indicates that the daily pattern of milk synthesis is responsive to the
timing of feed intake. Rottman et al. (2014) determined that feeding four times per day in equal
meals altered the daily pattern of milk synthesis relative to feeding once per day. Furthermore, we
previously established that time-restricted feeding shifts the phase of daily rhythms of milk
synthesis and in dairy cows (Chapter 3). However, it is unknown if these shifts are due to
entrainment of the molecular clock of the mammary gland by feeding or by the timing of nutrient
absorption. In mice, night-restricted feeding altered the expression of circadian transcription
factors. Ma et al. (2013) observed that time-restricted feeding altered clock gene expression and
70
shifted the rhythms of expression of enzymes and transcription factors related to milk fat
synthesis in lactating mice. The objectives of this experiment were to investigate the effect of
time-restricted feeding on the molecular circadian clock of the mammary gland. The hypothesis
was that restricting the time of feeding to 16 h during the night, which is typically the lowest
period of feed intake, will cause a phase shift in the daily patterns of clock genes in mammary
tissue, demonstrating a change in the molecular clock of the mammary gland.
MATERIALS AND METHODS
Animals and Treatments
Eleven multiparous mid-lactation Holstein cows (190 ± 31 d postpartum; mean ± SD)
from the Pennsylvania State University Dairy Research and Teaching Center were randomly
assigned to one of two treatments sequences in a cross-over design. Treatment periods were 22 d
long with 10 d of diet adaptation followed by 12 d of sampling. Treatments included (1) control
(CON) with feed delivered at 0800 h and available 24 h/d and (2) night-restricted feeding (NR)
with feed available for 16 h/d from 2000 h to 1200 h (Figure 4-1). All cows were fed the same
total mixed ration (TMR) that was mixed once daily at 0800 h and fed immediately to the CON
group (Supplemental Table S4-1). The remaining feed was compressed into plastic barrels,
covered with plastic, and stored at ambient temperature until NF feeding at 2000 h. Feed intake
was recorded daily and feed was offered at 110% of the previous day’s intake.
Feed samples were collected on d 7, 14, and 21 of each period and composited by period
prior to analysis. Dry matter, ash, starch, and CP were analyzed according to the AOAC (2000)
Official Methods of Analysis, and NDF and ADF were analyzed according to Van Soest et al.
(1991). Dry matter intake was determined on d 15 to 18 of each period. Animals were housed in
individual tie-stalls with sawdust bedded rubber mattresses and had ad libitum access to water.
71
Cows were maintained in a 19 h light: 5 h dark photoperiod with lights on from 0500 to 0000 h.
The experiment was conducted from January to March of 2017. All experimental procedures
were approved by the Penn State University Institutional Care and Use Committee.
Milk Sampling and Analysis
Cows were milked 2x/d at 0700 and 1900 h and milk yield was measured at each milking
using an integrated meter (AfiMilk MPC Milk Meter; Afimilk Agricultural Cooperative Ltd.,
Kibbutz Afikim, Israel). Milk yields were corrected for stall deviations according to Andreen et
al. (2018). Milk was sampled on d 10 and 17 of each period. One subsample was stored at 4°C in
preservative (Bronolab-WII; Advanced Instruments, Inc., Noorwood, MA) prior to analysis of fat,
protein, lactose, and MUN by Fourier transform infrared spectroscopy (Fossomatic 4000 Milko-
Scan and 400 Fossomatic, Foss Electric; Dairy One Laboratory, Ithaca, NY). A second subsample
was immediately centrifuged at 2300 x g at 4°C for 15 min and fat cakes were stored at -20°C for
analysis of fatty acid (FA) composition. Milk FA were extracted using hexane:isopropanol,
transmethylated using sodium methoxide, and FA methyl esters were quantified by GLC with a
flame-ionization detector according to Rico and Harvatine (2013) as modified by Baldin et al.
(2018).
Mammary Tissue Collection
Mammary biopsies were collected to represent every 6 h across the day (0400, 1000,
1600, and 2200 h). On d 11 to 14, one of the rear quarters was randomly chosen for a biopsy
collection at 2200 h, followed by a second biopsy on the opposite quarter at 1000 h the following
day. Following 7 d of recovery, a third and fourth biopsy was similarly collected from each rear
quarter on d 19 to 22 at 1600 and 0400 h. Cows were sedated using intravenous xylazine (0.02
mg/100 kg of BW) and local anesthesia was applied by subdermal block using 10 mL of 2%
72
lidocaine HCL in a line above the incision site. The biopsy site was disinfected with providone-
iodine scrub (Betadine, Avrio Health L.P., Stamford, CT) and ethanol rinse and a surgical blade
was used to make a 0.2 to 0.5 cm incision near the midpoint of the rear quarter. Tissue was
collected from a depth of 8 to 12 cm using a vacuum-assisted electronic biopsy gun system with a
10 gauge needle (Vacora; Bard Biopsy Systems, Tempe, AZ). Tissue samples were rinsed with
0.9% saline solution, snap frozen in liquid nitrogen, and stored at −80°C until RNA extraction
(~80 mg tissue/biopsy). After biopsy collection, manual pressure was placed on the biopsy site
for 5 to 10 min until bleeding ceased. Minimal bleeding occurred and milk appeared normal
within 3 to 5 milkings after the biopsy. No apparent intramammary infections or production
declines occurred as a result of the surgery.
Plasma Sampling and Analysis
Blood was collected using potassium EDTA vacuum tubes (Greiner Bio-One North
America, Inc., Monroe, NC) by venipuncture of a coccygeal vessel on d 15 to 17 of each period
to represent every 4 h across the day (0200, 0600, 1000, 1400, 1800, and 2200 h). Blood was
immediately placed on ice and centrifuged at 2300 x g for 15 min at 4° C within 30 min of
collection. Plasma was collected and stored at -20°C for analysis of glucose, plasma urea nitrogen
(PUN), and non-esterified FA (NEFA) according to Rottman et al. (2014). Briefly, Plasma
glucose concentration was analyzed using a glucose oxidase/peroxidase enzymatic colorimetric
assay (No. P 7119, Sigma-Aldrich, St. Louis, MO), PUN was assayed using a modified Berthelot
methodology (Modified Enzymatic Urea Nitrogen Procedure No. 2050; Stanbio Laboratory,
Boerne, TX), and NEFA concentration was measured using a acyl-CoA oxidase/peroxidase
enzymatic colorimetric assay [NEFA-HR (2), Wako Diagnostics, Richmond, VA].
73
Gene Expression Analysis
Approximately 40 mg of mammary tissue was homogenized in 1.5 mL of RNA-Solv
(Omega Bio-Tek, Norcross, GA) and total RNA was isolated from approximately 40 mg of
mammary tissue using the E.Z.N.A. Total RNA Kit II with on-column DNase treatment (Omega
Bio-Tek). RNA concentration and quality were determined using a spectrophotometer (NanoDrop
1000, Thermo Scientific, Wilmington, DE). Total RNA was reverse transcribed with the High
Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Inc., Foster City, CA) using
random primers. Relative gene expression was performed using quantitative PCR (qPCR) using
an Applied Biosystems 7900HT Fast Real-Time PCR System (Applied Biosystems, Inc.) with
400 nM of gene-specific forward and reverse primers and SYBR Green reporter dye (PerfeCTa
SuperMix with ROX, Quanta Biosciences, Gaithersburg, MD). Primers were designed to span
exon boundaries and using Primer-BLAST (National Center for Biotechnology Information,
Bethesda, MD) to amplify a product length between 70 and 150 bp or were from previous
publications as indicated (Table 4-1). Reactions were performed in triplicate and transcript
abundance was determined relative to a dilution curve of pooled cDNA and normalized using the
geometric mean of reference genes [ribosomal protein 9 (RPS9), beta-2 microglobulin (B2M), and
beta-actin (ACTB)], which was included as a covariate in the statistical model. The presence of a
single product was verified by dissociation curve analysis.
Statistical Analysis
All statistical analysis was performed using the MIXED procedure of SAS 9.4 (SAS
Institute, Cary, NC). Daily DMI, milk, fat, and protein yield, fat and protein concentration, and
FA yields were analyzed as a mixed effects model with the fixed effects of treatment and period
and the random effects of cow and day nested within period. The autoregressive covariance
74
structure was used and denominator degrees of freedom were adjusted using the Kenward-Roger
method.
Cosinor rhythmometry was performed on time course gene expression, plasma
metabolites, and body temperature data. Briefly, data were fit to the linear form a cosine function
by random regression in SAS 9.4 with a model that tested the fixed effects of treatment, linear
form of the cosine function, and the interaction of treatment with cosine function, and the random
effects of cow and period. The fit of the 24 h cosine was tested using a zero-amplitude test, and
the amplitude and acrophase (time at peak of rhythm) were calculated and significance
determined according to Niu et al. (2014). The presence of a 12 h harmonic was tested for plasma
hormone and metabolite data and was included for glucose and NEFA because it improved model
fit. Secondly, a model testing the fixed effects of treatment, time, and their interaction, and the
random effects of cow with preplanned contrasts testing the interaction of treatment and time was
performed on time course data. In all analyses, points with Studentized residuals ± 3.5 were
removed as outliers. Significance was declared at P < 0.05 with trends recognized at P < 0.10.
High resolution figures were generated using an add-in for Microsoft Excel (Kraus, 2014).
RESULTS AND DISCUSSION
Dry Matter Intake
Night-restricted feeding tended to reduced DMI with NR cows consuming 1.7 kg (7%)
less feed per day than CON (P = 0.06; Figure 4-2A). The decrease in intake is likely because NR
cows were fasted during the high intake period of the day. Dairy cow naturally prefer to eat the
greatest amount of feed during the morning and afternoon, with limited feed intake overnight
(Albright, 1993). Previous research revealed that night-restricted feeding numerically decreased
DMI by 1.6 kg relative to restricting feed availability to 16/d during the day (0900 to 1100), but
75
this change was not significant (Chapter 3). Niu et al. (2014) also observed a 1.2 kg (4%)
decrease in DMI by delivering feed at night (2030) versus the morning (0830) without restricting
the time of feed availability. Taken together, fasting during the afternoon when intake is typically
high reduces the total amount of feed consumed per day.
Milk Yield, Milk Composition, and Milk Fatty Acid Yields
Night-restricted feeding reduced milk yield by 8% (P = 0.001; Figure 4-2B).
Furthermore, NR decreased milk fat and protein yield by 11% (P = 0.02) and 10% (P < 0.001),
respectively (Figure 4-2D). Milk fat concentration was not affected by treatment (P = 0.91),
while milk protein concentration was decreased 2% by NR (P = 0.01; Figure 4-2C). We
previously failed to detect a difference in milk yield and milk fat and protein concentration
between cows undergoing time-restricted feeding for 16 h/d either during the day or the night,
although night-restricted feeding decreased milk fat yield 64 g (6%) compared to day-restriction
(Chapter 3). Rottman et al. (2014) also observed that feeding 4 x/d in equal meals increased the
yields and concentrations of both fat and protein compared to feeding 1x/d.
The reduction in milk synthesis in the current experiment may be related to modification
of circadian clock of the mammary gland. In rodent models, restricting the time of feeding to the
inactive phase without altering feed intake reduces the accretion of adipose tissue (Hatori et al.,
2012; Sherman et al., 2012). This change has been associated with increased robustness of daily
rhythms of adipocyte clock genes. Moreover, time-restricted feeding reduces adiposity in humans
(Manoogian and Panda, 2017). A similar mechanism may occur in cows, whereby alterations in
the circadian clock reduce the uptake and/or synthesis of nutrients in the mammary gland. Hu et
al. (2017) observed that small interfering RNA (siRNA)-mediated knockdown of PER2 mRNA
expression was associated with a reduction in the synthesis of αs1- and αs2-casein in mammary
76
epithelial cell culture establishing a relationship between the mammary circadian clock and
protein synthesis.
The decrease in milk and milk fat and protein yield in the current experiment is possibly
confounded by the reduction in feed intake. While milk yield drives intake, unless another
limiting factor (rumen fill, energy excess) is present, fasting during the high-intake afternoon of
the day may have prevented cows from eating to meet their requirements for milk synthesis
(Allen, 2014). Future experiments examining the role of time of feed intake on the mammary
circadian clock may consider pair-feeding cows to eliminate this effect.
Night-restricted feeding did not alter the yield of FA originating from de novo synthesis
(Σ < 16C; P = 0.50), preformed FA originating from plasma (Σ >16C; P = 0.41), or mixed-source
FA (Σ 16C; P =0.37; Figure 4-3). These results differ from previous research, which observed a
5% reduction in de novo synthesized fatty acids and an 8% decrease in mixed-source fatty acids
in cows restricted to feeding for 16 h/d the night compared to the day (Chapter 3). However,
similar to the current paper, altering the timing of feed delivery without restricting the time of
feed availability did not affect FA profile (Niu et al., 2014). In the current experiment, individual
FA, iso-C14:1, iso-C15:1, iso-C16:1, C22:0 and C20:3-n6 were decreased by NR (P < 0.05),
C24:0 tended to be increased by NR feeding (P = 0.09), while NR tended to decrease both trans-
10 C18:1 and trans-11 C18:1 (0.05 < P < 0.10; Supplemental Table S4-2).
Molecular Clock Gene Expression
The expression of CRY1 fit a daily rhythm in both treatments (P < 0.05), CLOCK fit a
rhythm in CON (P < 0.001) and tended to fit a rhythm in NR (P = 0.06), and REV-ERBα failed
to fit a rhythm in CON (P = 0.62), but tended to fit a rhythm in NR (P = 0.06; Figure 4-4).
However, BMAL1, PER1, and PER2 did not fit daily rhythm in either treatment (P > 0.10). Night-
restricted feeding appeared to have a slight effect on the pattern of PER1 expression, with both
77
treatments having similar baseline expression with a noticeable rise of about 30% at 0400 in CON
that was shifted to 1600 in NR. However, this change was not significant. The daily rhythms of
CLOCK and CRY1 were affected by treatment. Night-restricted feeding increased the amplitude
of CLOCK by 75%, and phase-delayed the rhythm ~4.5 h. Night-restricted feeding did not alter
the amplitude of CRY1, but phase-delayed the rhythm by 3 h.
Time-restricted feeding can entrain circadian rhythms of other peripheral tissues in
rodent models. Damiola et al. (2000) detected a 12 h phase shift in the liver protein abundance of
PER1 and PER2 in mice that had feed availability restricted for 12 h/d during the light phase
(0600 to 1800) or the night phase (1800 to 0600), without a concomitant change in SCN clock
gene expression. These results were corroborated by Stokkan et al. (2001) in rats feed-restricted
to only 4 h during the inactive period. In addition to shifting the phase of the liver molecular
clock, they observed that the molecular clock in the lung was shifted by time-restricted feeding,
but the clock of the SCN was unaltered. The daily rhythms of mRNA abundance of Bmal1, Per1,
Per2, Cry2, and Rev-erbα in adipocytes of male C57BL/6 mice were phase shifted by restricting
the intake of a high-fat diet to 4 h/d during the inactive period of the day.
There has been limited research describing a role of the circadian clock in the mammary
gland. In human mammary epithelial cells, 7% of genes follow a 24 h rhythm (Maningat et al.,
2009). Moreover, Casey et al., (2009) suggested a potential role of the mammary clock in
modulating the homorhetic response to lactation by discovering an increases in expression and
phase shifts in rhythms of positive transcriptional regulators BMAL1, CLOCK, and RORα in
lactating compared to nonlactating mammary tissue. In mice, the amplitudes of BMAL1 and
CRY1 are increased after the start of lactation (Metz et al., 2006). Rhythms of mammary clock
gene expression can be entrained by altering the light-dark cycle (Plaut and Casey, 2012).
Furthermore, timing of nursing can entrain the mammary rhythms of PER1 in rabbits. The current
78
experiment demonstrates that, in addition to these cues, time-restricted feeding can modestly
entrain the molecular clock of mammary gland of dairy cows.
Plasma Hormones and Metabolites
There was a treatment by time interaction for plasma glucose (P = 0.006) but no main
effect of treatment was present (P = 0.94; Figure 4-5A). The fit of the 24 h rhythm was improved
in both treatments by including a 12 h harmonic (P < 0.001). A 24 h rhythm with a 12 h harmonic
was present in both CON (P = 0.004) and NR (P < 0.0001). The acrophase of the 24 h rhythm
peaked at 0137 in CON, but the rhythm with phase delayed 12 h by NR, which peaked at 1337 (P
< 0.05). Furthermore, night-restricted feeding increased the amplitude of plasma glucose
concentration by 124% from 6.5 mg/dL in CON to 14.6 mg/dL in NR (P < 0.05).
Previous research has demonstrated that time of feeding alters the daily rhythms of plasma
glucose concentration. We previously detected a phase delay of 10.2 h in plasma glucose
concentrations in cows with feed available for 16 h/d from 1900 to 1100 compared to those with
feed available for 16 h/d from 0700 to 2300 (Chapter 3). Similar to the current experiment, they
observed a 149% increase the amplitude of the plasma glucose rhythms during night-restricted
feeding. Moreover, Niu et al. (2014) detected that the daily pattern of plasma glucose
concentration was altered by changing the time of feed delivery from 0830 to 2030 h without
restricting the time of feed availability.
Unlike previous experiments, we observed a 12 h harmonic in the daily pattern of plasma
glucose, suggesting that glucose in our experiment followed a 12 h ultradian rhythm in addition to
the daily rhythm. In both treatments, the harmonic approximately corresponded to the acrophase
of the opposite treatment. In humans subjected to continuous feeding, glucose concentration has
been reported to follow ultradian rhythms of various period lengths, including 12 h (Simon et al.,
1987; Shea et al., 2005). To our knowledge, however, there have been no reports of a 12 h
79
harmonic in cattle. Lefcourt et al. (1993) observed that dairy cows express 3 h ultradian rhythms
of plasma cortisol concentration. Cortisol increases plasma glucose concentrations by increasing
gluconeogenesis and decreasing tissue glucose uptake, and may be partially responsible for the
ultradian rhythm of glucose concentration observed in the current experiment.
Night-restricted feeding did not affect average NEFA concentration (P = 0.74; Figure
4-5B). A single 24 h rhythm did not fit either treatment, but similar to glucose concentration, both
treatments fit a 24 h rhythm that included a 12 h harmonic (P < 0.01). The amplitude of the
primary 24 h rhythm was not altered by treatment (P > 0.05). Furthermore, the acrophase of the
main rhythm was not affected, with CON peaking at 1125 h and NR peaking at 1237 h (P >
0.05). These results suggest that night-restricted feeding did not alter the daily pattern of lipid
mobilization, despite fasting during the typically high-intake afternoon period of the day. Results
contrast with previous research from our lab which demonstrated that observed that night-
restricted feeding from 1900 to 1100 h shifted the rhythms of NEFA concentration by 10.7 h
relative to day-restricted feeding from 0700 to 2300 h (Chapter 3). Furthermore, we detected over
a 3-fold increase in the amplitude of the rhythm due to night-restricted feeding. The differences in
the current study and previous study may be related to the daily pattern of intake. The previous
experiment imposed day-restricted feeding which likely caused more extreme differences in the
pattern of feed intake. In the current experiment, maximum NEFA concentration in both
treatments corresponded to the period when the NR cows were fasted, with the harmonic
occurring at night, during which cows typically have reduced intake. Social interactions have
previously shown to have a strong influence on feeding behavior of dairy cattle (Grant and
Albright, 2001). Cows on CON may modified their daily pattern of feed intake to correspond to
the NR treatment. Moreover, cows on the current experiment may have been in more positive
energy balance, resulting in lower lipid mobilization in response to fasting. The results of our
experiment correspond more closely with experiments where the time of feed delivery was
80
altered without restricting the time of feed availability. Nikkhah et al. (2008) observed little
difference in the daily pattern of NEFA concentration when cows were fed at 0900 h versus 2100.
Furthermore, Niu et al. (2014) detected no difference in the daily pattern of NEFA among cows
fed 1x/d at 0830, 1x/d at 2030, or 2x/d at both 0830 and 2030.
Plasma urea nitrogen concentration was not affected by treatment (P = 0.25; Figure
4-5C). A 24 h rhythm was present in both treatments (P < 0.01), with treatment affecting the
daily rhythm. Night-restricted feeding increased the amplitude of the daily rhythm 62% compared
to CON (P < 0.05) and delayed the phase 9.9 h from 1247 to 2241h. A 24 h rhythm of PUN has
been well-established in previous experiments. Our results agreed with a previous experiment in
our lab which demonstrated a 24 h rhythm in cows under night-restricted feeding that was phase
delayed 12.2 h relative to day-restricted feeding for 16 h/d (Chapter 3). Furthermore, Nikkhah et
al. (2008) observed an approximately 12 h delay in the peak of PUN in cows fed at 2100 versus
0900, but only in cows fed low fiber (38:62 forage: concentrate; 28.6 NDF), but not high fiber
diets (51:49 forage: concentrate; 33.8% NDF), suggesting that the effect may be diet dependent.
Results suggest that night-restricted feeding shifts the daily rhythms of amino acid utilization
efficiency.
CONCLUSIONS
Restricting feed availability to 16 h during the night increased the amplitude and phase-
shifted the daily rhythms of expression of several molecular clock genes in mammary tissue,
suggesting that nutrients can entrain the mammary molecular clock. These shifts may be partially
mediated by changes in plasma metabolites, because night-restricted feeding shifted the daily
patterns of plasma glucose and plasma urea nitrogen. Results may be partially influenced by a
reduction in total dry matter intake by night-restricted feeding, which was likely responsible for
81
the decrease in milk, fat and protein yield observed in this treatment. Future studies examining
the role of the time of feeding on the molecular clock of the mammary gland may consider pair-
feeding cows to ensure intake is consistent among treatments. Additionally, this study was limited
to examining clock genes at the mRNA level. Additional post-transcriptional and post-
translational modifications of clock-related proteins also affect the manifestation of circadian
rhythms in tissues. Further research examining the role of food intake on the protein expression
and protein phosphorylation will provide a more complete depiction of the role of nutrients on the
mammary clock. However, this study provides initial compelling evidence to suggest that food
intake can entrain the molecular clock of the mammary gland.
82
FIGURES
Figure 4-1. Schedule of feeding, lighting and sampling during the experiment.
Treatments were feed available continuously for 22 h/d starting at 0800 h (CON) or night restricted feeding
with feed available for 16 h/d from 2000 to 1200 h (NR). Panels show (A) the times of the day when feed
was available, lights were on/off, mammary biopsies were collected, and blood was collected and (B) the
days of each period when milk was sampled, biopsies were collected, and blood was collected.
83
Figure 4-2. Effect of night-restricted feeding on average daily feed intake and milk synthesis.
Treatments were feed available continuously for 22 h/d starting at 0800 h (CON) or night restricted feeding
with feed available for 16 h/d from 2000 to 1200 h (NR). Panels show the effect of night-restricted feeding
on (A) dry matter intake (kg/d), (B) milk yield (kg/d), (C) milk fat and protein concentration (%), and (D)
milk fat and protein yield (g/d).
84
Figure 4-3. Effect of night-restricted feeding on milk fatty acid yields by origin.
Treatments were feed available continuously for 22 h/d starting at 0800 h (CON) or night restricted feeding
with feed available for 16 h/d from 2000 to 1200 h (NR). Fatty acids grouped by biosynthetic origin
including FA derived from de novo synthesis in the mammary gland (De novo; Σ < 16C), mixed source
from de novo synthesis and preformed uptake (Mixed; Σ 16C), and preformed FA taken up by the
mammary gland from plasma (Preformed; Σ >16C).
85
Figure 4-4. Effect of night-restricted feeding on the daily patterns of expression of molecular
clock genes in the mammary gland.
Treatments were feed available continuously for 22 h/d starting at 0800 h (CON) or night restricted feeding
with feed available for 16 h/d from 2000 to 1200 h (NR). Panels show mammary mRNA expression of (A)
brain and muscle ARNT-like 1 (BMAL1) (B) circadian locomotor output cycles kaput (CLOCK), (C) period
1 (PER1), (D) period 2 (PER2), (E) cryptochrome 1 (CRY1), and (F) REV-ERBα across the day. LSM and
SEM bars and cosine fit of the control (solid line) and NR (dotted line) are shown. Cosinor analysis is
shown below each panel and include the amplitude (Amp; difference between peak and mean), acrophase
(Acro; time at peak of the rhythm), and the P-value of the zero-amplitude test. The black and white bars
above the x axis display the light: dark cycle.
86
Figure 4-5. Effect of night-restricted feeding on the daily patterns of plasma metabolites.
Treatments were feed available continuously for 22 h/d starting at 0800 h (CON) or night restricted feeding
with feed available for 16 h/d from 2000 to 1200 h (NR). Panels show the effect of night-restricted feeding
on daily patterns of (A) glucose concentration (mg/dL), (B) nonesterified fatty acids (NEFA; μEq/L) and
(C) plasma urea nitrogen (mg/dL). Data are presented as relative expression LSM and SEM bars and cosine
fit of the control (solid line) and NR (dotted line) are shown. Cosinor analysis is shown below each panel
and include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of the
rhythm), and the P-value of the zero-amplitude test. NS: Reduced model fits data significantly better than
daily rhythm model. The black and white bars above the x-axis display the light: dark cycle.
87
TABLES
Table 4-1. Bovine primers used to quantify mammary gene expression with RT-qPCR.
1Gene names: ACTB = Beta-actin; B2M = Beta-2-microglobulin; BMAL1 = Brain and muscle ARNT-like protein-1; CLOCK = Circadian locomotor
output cycles kaput; CRY1 = Cryptochrome 1; PER1 = Period 1; PER2 = Period 2; RPS9 = Ribosomal protein S9. 2Encoded by the nuclear receptor subfamily 1 group d member 1 (NR1D1) gene.
Gene1
Accession
no. Forward Primer Reverse Primer
Amplicon
Size (nt) Reference
Genes of Interest
BMAL1 NC_037342.1 CAGCTCTCCTCCTCCGACAC TGGCCCAGGGGTTATATCTG 90
CLOCK NC_037333.1 TCAGTCTCAAGGAAGCGTTG GGAGTGCTAGTATCTGTTGGG 142
CRY1 NC_037332.1 TGCAAGTTGTACTCAAGGGAG CAATACTCTGCGTGTCCTCTTC 149
PER1 NC_037346.1 CCAGGAGTTCTACCAGCAATG GAGACAGCCACGGAGAAG 141
PER2 NC_037330.1 ACGTTTTCCAGTCTCGCTG CGTTTGGACTTCAGTTTTCGG 148
REV-ERBα2 NC_037346.1 TGGTGTTGCTGTGTAAGGTG CCTTTTGTACTGGATGTTCTGC 122
Reference Genes
ACTB NC_037352.1 GCGTGGCTACAGCTTCACC CTTGATGTCACGGACGATTTC 55 (Kadegowda et al., 2009)
B2M NC_037337.1 TTTGGAACAACCCCGGATAG AAACCTCGATGGTGCTGCTT 61 (Urrutia and Harvatine, 2017)
RPS9 NC_037345.1 CCTCGACCAAGAGCTGAAG CCTCCAGACCTCACGTTTGTTC 64 (Urrutia and Harvatine, 2017)
88
SUPPLEMENTAL TABLES
Supplemental Table S4-1. Diet and nutrient composition of experimental diet.
1Contained (% of DM): 33.0% NDF, 16.2% ADF, 42.3% Starch. 2Contained (% of DM): 42.5% NDF, 31.3% ADF, 0.7% Starch. 3Contained (% of DM): 73.2% NDF, 40.8% ADF, 1.2% Starch. 4Contained (%, as-fed basis): 37.0% Calcium carbonate; 29.9% dried corn distillers grains; 24.5% salt;
4.2% magnesium oxide (54% Mg); 2.4% organic phosphorus (15% P); 0.5% zinc sulfate; 0.2% mineral oil.
Composition (DM-basis): 7.2% CP; 7.1% NDF; 3.7% ADF; 15.0% Ca; 0.75% P; 0.33% K; 2.6% Mg; 0.5%
S; 9.8% Na; 23.0 mg/kg Co; 652 mg/kg Cu; 783 mg/kg Fe; 54.0 mg/kg I; 1190 mg/kg Mn; 12.8 mg/kg Se;
1718 mg/kg Zn; 88,700 IU/kg vitamin A (retinyl acetate); 28,400 vitamin D (activated 7-
dehydrocholesterol); 850 IU/kg vitamin E (dl-α tocopheryl acetate). 5Fed as coated urea (Optigen, Alltech Inc., Lexington, KY; 259% CP, DM basis).
Item Composition
Ingredients, % of DM
Corn silage1 43.6
Alfalfa haylage2 20.2
Canola meal 12.4
Ground corn 10.9
Roasted soybeans 5.7
Grass hay/straw3 2.7
Vitamin/mineral mix4 2.1
Molasses 2.0
Controlled-release N5 0.26
Nutrients, % of DM
NDF 30.6
ADF 17.9
Starch 27.8
Ash 8.6
89
Supplemental Table S4-2. Effect of night-restricted feeding on daily yields of individual milk
fatty acids.
Treatment
FA, g/d1 CON NR SEM P-value
C4:0 56.53 56.75 5.12 0.97
C6:0 36.45 33.95 2.93 0.52
C8:0 18.10 17.67 1.70 0.80
C10:0 44.14 42.65 3.76 0.70
cis-9 C10:1 3.89 3.58 0.30 0.32
C11:0 0.90 0.84 0.10 0.54
C12:0 52.99 50.40 3.57 0.48
C13:0 0.81 0.60 0.22 0.37
C13:0 iso 0.31 0.26 0.04 0.22
C13:0 anteiso 0.59 0.38 0.16 0.36
C14:0 161.87 153.98 6.10 0.21
C14:0 iso 1.55 1.03 0.16 0.004
cis-9 C14:1 12.73 12.07 0.83 0.43
C15:0 14.95 14.04 0.65 0.18
C15:0 iso 2.73 2.51 0.10 0.04
C15:0 anteiso 4.96 4.76 0.78 0.81
C16:0 375.38 362.77 13.68 0.37
C16:0 iso 3.51 2.68 0.21 0.0006
cis-9 C16:1 16.83 16.92 0.90 0.92
C17:0 6.73 6.59 0.25 0.58
C17:0 iso 2.62 2.59 0.58 0.96
C17:0 anteiso 5.64 6.01 0.23 0.12
cis-9 C17:1 2.19 2.40 0.16 0.22
OCFA2 42.22 41.15 1.93 0.59
BCFA3 21.82 19.98 1.86 0.33
OBCFA4 47.45 45.04 2.06 0.26
C18:0 139.31 127.95 7.04 0.12
trans-4 C18:1 0.35 0.37 0.05 0.60
trans-5 C18:1 0.23 0.23 0.04 0.92
trans-6,8 C18:1 4.36 4.46 0.18 0.57
trans-9 C18:1 3.54 3.55 0.14 0.92
trans-10 C18:1 5.81 6.61 0.46 0.09
trans-11 C18:1 12.33 13.56 0.71 0.10
t10:t11 C18:15 0.52 0.52 0.02 0.76
trans-12 C18:1 7.11 7.16 0.31 0.87
trans-15 C18:1 4.78 5.19 0.39 0.30
cis-9 C18:1 258.94 251.40 9.77 0.45
cis-11 C18:1 9.90 10.01 0.74 0.88
90
cis-12 C18:1 6.15 6.59 0.38 0.26
Total C18:1 315.12 310.14 12.05 0.68
C18:2 n-6 35.29 35.70 2.34 0.86
cis-9, trans-11 C18:2 6.53 7.01 0.48 0.33
C18:3 n-6 0.27 0.19 0.06 0.21
C18:3 n-3 4.02 3.52 1.21 0.78
Total C18 FA 505.08 489.19 17.84 0.38
C20:0 1.65 1.50 0.09 0.12
cis-11 C20:1 5.30 4.89 1.83 0.99
C20:3 n-6 1.80 1.63 0.07 0.02
C20:4 n-6 1.96 1.91 0.13 0.74
C20:5 n-6 0.43 0.36 0.04 0.14
C22:0 0.69 0.57 0.05 0.02
C22:4 n-3 0.33 0.25 0.07 0.27
C24:0 0.43 0.32 0.06 0.09 1Least squared means of control (CON) with feed available continuously for 22 h/d fed at 0800, and night-
restricted feeding (NR) with feed available 16 h/d from 1900 to 1100 h (n = 11 per treatment. 2Total odd-chain fatty acids (Σ 11:0, 13:0, 13:0 iso, 13:0 anteiso, 15:0, 15:0 iso, 15:0 anteiso, 17:0, 17:0
iso, 17:0 anteiso, cis-9 17:1). 3Total branched-chain fatty acids (Σ 13:0 iso, 13:0 anteiso, 14:0 iso, 15:0 iso, 15:0 anteiso, 16:0 iso, 17:0
iso, 17:0 anteiso). 4Total odd- and branched-chain fatty acids (Σ 11:0, 13:0, 13:0 iso, 13:0 anteiso, 14:0 iso, 15:0, 15:0 iso,
15:0 anteiso, 16:0 iso, 17:0, 17:0 iso, 17:0 anteiso, cis-9 17:1). 5Ratio of trans-10 18:1 to trans-11 18:1.
91
Chapter 5
The effects of time of fatty acid infusion on the daily rhythms of milk
synthesis and plasma hormones and metabolites in dairy cows
Isaac J. Salfer, P. A Bartell, and Kevin J. Harvatine
92
ABSTRACT
Dairy cows display daily rhythms of milk synthesis that appear to be driven by the
circadian clock of the mammary gland and are altered by the time of feed availability. Fatty acids
have been shown to entrain circadian rhythms in other peripheral tissues such as the liver and
adipose tissue in experimental models, but their role in the mammary gland has not been well-
investigated. The objective was to determine the effects of the timing of fatty acid absorption on
the daily rhythms of milk synthesis. Nine lactating Holstein cows were arranged in a 3 x 3 Latin
square design. Treatments were abomasal infusions of 350 g/d of an oil enriched in C18:1 either
continuously throughout the day (CON) or over 8 h/d from 0900 to 1700 h (DAY) or from 2100
to 0500 (NGT). Treatment periods were 12 d with a 5 d washout. Cows were milked every 6 h
during the final 7 d of each period to determine the daily patterns of milk synthesis. A 24 h
rhythm was fit to time-course data using cosine analysis and the amplitude and acrophase (time at
peak) were determined. Daily milk, fat, and protein yield and fat and protein concentration were
not affected by treatment. Milk yield fit a 24 h rhythm in CON and DAY, and tended to fit a
rhythm in NGT. The amplitude of the rhythm of milk yield was increased 29% in DAY compared
to CON (P < 0.05), but NGT was not affected. Furthermore, DAY delayed the acrophase of the
daily rhythm of milk yield by 2 h compared to CON and NGT (P < 0.05). Fat and protein
concentration exhibited a daily rhythms in CON and NGT (P < 0.05), but not DAY. The
amplitude of the rhythm of fat concentration was increased 29% by NGT compared to CON. Fat
yield tended to fit a 24 h rhythm in DAY (P = 0.07), but not in CON or NGT. Moreover, protein
yield fit a 24 h rhythm in all treatments and the amplitude of the rhythm was increased 30% by
DAY and decreased 47% by NGT. Both de novo and mixed source FA yield was reduced by
DAY and NGT, suggesting that restricting the time of FA infusion may reduce de novo fatty acid
synthesis. Glucose failed to fit a daily rhythm in any treatment, while nonesterified fatty acids fit
93
a rhythm in CON and NGT, but the rhythm was eliminated by DAY. Plasma urea nitrogen fit a
daily rhythm in all treatments, and both the mean and amplitude were increased by DAY. Fatty
acid infusion during the daytime modified the daily rhythms of milk synthesis by increasing the
amplitude of milk yield and decreasing the amplitude of milk fat and protein concentration
whereas infusion at night had little effect. Day-infusion also modified the daily rhythms of
plasma metabolites by reducing the amplitude of nonesterified fatty acids and increasing the
amplitude of plasma urea nitrogen.
Keywords: Daily rhythm, milk synthesis, nutrient entrainment, circadian rhythm
94
INTRODUCTION
Dairy cows exhibit a daily pattern of feed intake that is naturally crepuscular, with high
rates of intake during the morning and afternoon and a decline in intake overnight (Albright,
1993). Furthermore, intake is stimulated by delivery of fresh feed and milking (Menzi and Chase,
1994; DeVries et al., 2005). While most cows in the United States are fed a total mixed ration
(TMR), which provides a consistent composition of nutrients over the day, the daily pattern of
feed intake causes the consumption of nutrients to vary across the day (Ying et al., 2015). The
daily pattern of feed intake can be altered by modifying the time of feed delivery, as well as
increasing the frequency of feeding (e.g. Niu et al., 2014; Rottman et al., 2014).
In addition to a daily pattern of feed intake, cows display daily rhythms of milk and milk
component synthesis. Typically, milk yield is greatest in the morning, while milk fat and protein
concentration are greatest in the evening (Rottman et al., 2014; Ying et al., 2015). However,
night-restricted feeding shifts these rhythms so that milk yield peaks in the evening and milk fat
and protein concentration peak in the morning (Chapter 3). These daily rhythms appear to be
driven by the molecular circadian clock of the mammary gland (Chapter 4). The molecular clock
is composed of a transcriptional-translational feedback loop that includes a variety of
transcription factors that oscillate over 24 h and regulate the expression of gene in a circadian
manner (Takahashi, 2017). Moreover, a 24 h pattern post-translational regulation via
phosphorylation and dephosphorylation adds another level of regulation to the molecular clock
(Robles et al., 2017).
Light is the primary signal that entrains circadian rhythms, but an increasing body of
evidence has demonstrating a role for timing of food intake in circadian entrainment of peripheral
tissues. Fatty acids, in particular, play an important role in entrainment of circadian rhythms. The
lipid-sensing family of nuclear receptors known as peroxisome proliferator-activated receptors
95
(PPARs) display 24 h rhythms of expression, and possess E-box elements, suggesting direct
regulation by BMAL1/CLOCK (Oishi et al., 2005). Moreover, the expression of BMAL1 is
activated by PPARα though binding to a PPAR-response element present in the BMAL1 promoter
(Ribas-Latre and Eckel-Mahan, 2016). Feeding a high-fat diet has been shown to alter the
molecular clock and the daily pattern of adiponectin expression in the liver of male C57BL/6
mice (Barnea et al., 2009; Branecky et al., 2015). Moreover, CLOCK-/- and BMAL-/- double-
mutant mice have increased triglyceride accumulation in white adipose tissue, as well as
disrupted rhythms of free fatty acids in blood (Shostak et al., 2013).
Evidence suggests that time-restricted feeding alters the molecular circadian clock of the
mammary gland. We previously discovered that the daily rhythms of CLOCK and CRY1 mRNA
expression were shifted 4 h by restricting the time of feed availability from 7 PM to 11 AM
(Chapter 4). Furthermore, the daily rhythms of both clock genes and transcription factors
involved in milk fat synthesis were altered by time restricted feeding in mice (Ma et al., 2013).
While feed intake entrains the daily rhythms of milk synthesis, the role of individual nutrients,
such as FA, is unknown. The objective of this experiment was to determine the effect of time of
FA absorption on the daily rhythms of milk synthesis, plasma metabolites, and body temperature.
We hypothesize that altering the time of FA absorption through abomasally infusing FA either
continuously throughout the day or over 8 h during the day or the night will shift the phase of the
daily rhythms of milk fat synthesis. Furthermore, we expect the daily rhythm of plasma
nonesterified fatty acid concentration and core body to be shifted by time of fatty acid infusion.
MATERIALS & METHODS
96
Animals and Treatments
Nine multiparous mid-lactation (132 ± 90 d postpartum; mean ± SD) Holstein cows from
the Penn State University Dairy Research and Teaching Center fitted with rumen cannulas and
randomly assigned to treatment sequences in a 3x3 Latin Square design. Treatment periods were
12 d long with 10 d of treatment adaptation and 2 d of sampling. There was a 5 d washout
between periods. During the final 7 d of each period, cows were milked every 6 h (4 x/d) to allow
a daily rhythm to be fit to milk yield and components data.
Treatments were abomasal infusion of 350 g/d of an oil mixture either continuously
throughout the day (CON) or over 8 h/d from 0900 h to 1700 h (DAY) or from 2100 h to 0500 h
(NGT; Figure 5-1). The treatment consisted of a high oleate mixture of free fatty acids (KIC
Chemicals, Inc, New Paltz, NY; Supplemental Table S5-2) with 10% w/w Tween 80 added to
aid in emulsification. Cows in the CON treatment were infused a total of 22 h/d with no infusion
occurring for 30 min during each milking. All cows were provided the same basal TMR fed once
daily at 0600 h at 110% of the previous day’s intake (Supplemental Table S5-1). Feed samples
were collected on day 7 of each period and analyzed for dry matter, ash, neutral detergent fiber
(NDF), acid detergent fiber (ADF), starch, and crude protein according to Rico and Harvatine
(2013). The experiment was conducted from May 7 to June 22, 2018. All experimental
procedures were approved by the Penn State University Institutional Care and Use Committee.
Milk Sampling and Analysis
Cows were milked every 6 h on the final 7 d of each period (0100, 0700, 1300, and 1900
h) to observe the daily rhythm of milk synthesis. Milk collected at each time point represented the
sum of milk synthesis over the previous 6 h interval and is plotted as the midpoint of the previous
milking interval (3 h prior to collection). Milk yield was measured at each milking using an
97
integrated milk meter (AfiMilk MPC Milk Meter; Afimilk Agricultural Cooperative Ltd., Kibbutz
Afikim, Israel). Yields were corrected for the deviation of each individual stall according to
Andreen et al. (2018). Milk was sampled at each milking on the final 2 d of each period. One
subsample was stored at 4°C with preservative (Bronolab-WII; Advanced Instruments, Inc.,
Noorwood, MA) prior to analysis of fat and protein concentration by Fourier transform infrared
spectroscopy (Fossomatic 4000 Milko-Scan and 400 Fossomatic, Foss Electric; Dairy One DHIA,
Ithaca, NY). A second subsample was stored at 4°C and centrifuged at 2300 x g within 24 h. The
resulting fat cakes were stored at -20°C and analyzed for concentrations of individual FA
according to Baldin et al. (2018).
Plasma Sampling and Analysis
Blood was collected by veinipuncture of a coccygeal vessel using potassium EDTA
vacuum tubes (BD Vacutainer, Becton Dickinson, Franklin Lakes, NJ). Blood sampling occurred
6 times across d 11 to 12 representing every 4 h across the day (0300, 0700, 1100, 1500, 1900,
and 2300 h). Samples were immediately placed on ice and centrifuged within 30 min at 2300 x g
for 15 min at 4°C. Plasma was collected and stored at -20°C for analysis of glucose, nonesterified
fatty acids (NEFA), and plasma urea nitrogen (PUN) as described by Rottman et al. (2014).
Briefly, plasma glucose concentration was analyzed using a glucose oxidase/peroxidase
enzymatic colorimetric assay (No. P 7119, Sigma-Aldrich, St. Louis, MO), NEFA concentration
was measured with an using acyl-CoA oxidase/peroxidase enzymatic colorimetric assay (NEFA-
HR (2), Wako Diagnostics, Richmond, VA), and PUN was assayed using a modified Berthelot
methodology (Modified Enzymatic Urea Nitrogen Procedure No. 2050; Stanbio Laboratory,
Boerne, TX).
98
Body Temperature Analysis
An intravaginal temperature logger was used to record core body temperature every 10
min on d 10 to 12 of each period as described by Niu et al. (2017). Briefly, a miniature plastic
coated thermometer fastened to a vaginal implant was placed centrally in the vagina. Body
temperature was averaged over 2 h intervals.
Statistical Analysis
All data were analyzed using the MIXED procedure of SAS 9.4 (SAS Institute, Cary,
NC). Models testing the effect of treatment on daily dry matter intake, milk production, and FA
yields included the fixed effect of treatment and the random effects of cow and period. The AR(1)
or ARH(1) covariance structure was selected based on convergence criteria, and denominator
degrees of freedom were adjusted using the Kenward-Roger method. Individual treatment least
squared means were compared using preplanned contrasts of CON versus DAY, CON versus
NGT and DAY versus NGT.
Time course data for milk production, plasma metabolites, and body temperature were fit
to the linear form of a cosine function with a period of 24 h using random regression in SAS 9.4.
The model included the fixed effects of treatment, cosine terms, and the interaction between
treatment and cosine terms, as well as the random effects of cow and period. The presence of a 12
h harmonic was tested for body temperature and plasma hormone and metabolite data, and was
included if it improved model fit. The fit of the 24 h cosine was determined using a zero-
amplitude test, and the amplitude and acrophase (time at peak of rhythm) were determined
according to Niu et al. (2014). In all analyses, points with Studentized residuals outside ± 3.5
were removed. A separate model testing the fixed effects of treatment, time, and their interaction,
with random effects of cow and period, and preplanned contrasts testing the interaction of
99
treatment and time was also conducted on all time course data. Statistical significance was
declared at P < 0.05 and trends acknowledged at 0.05 < P < 0.10. High resolution figures were
generated using an add-in for Microsoft Excel (Daniel's XL Toolbox; Kraus, 2014).
RESULTS AND DISCUSSION
Milk and Milk Components Synthesis
Dry matter intake was not affected by treatment (P = 0.25; Figure 5-2). Furthermore,
treatment did not affect daily milk or milk fat and protein yield or milk fat and protein
concentration (P > 0.10; Figure 5-3). High oleic acid oils, like the one used in our experiment,
have previously been shown to increase milk yield and milk fat concentration (LaCount et al.,
1994). The results of our study suggest that the timing of FA absorption does not affect daily milk
or milk component production.
Treatment altered the daily rhythms of milk synthesis (Figure 5-4). Milk yield fit a 24 h
rhythm in CON and DAY (P < 0.002), and tended to fit a rhythm in NGT (P = 0.08). Relative to
CON, DAY increased the amplitude of milk yield by 29% (P < 0.05), while NGT was unaffected
(P > 0.10). Milk yield peaked in the morning in all treatments, similar to previous
characterizations of the daily rhythm of milk synthesis in dairy cows (Rottman et al., 2014). Day-
infusion phase-delayed the rhythm of milk yield by 2 h compared to CON and NGT (P < 0.05).
Fat yield did not fit a rhythm in CON or NGT, but a rhythm was induced by DAY (P = 0.02),
with the amplitude being 115% and 200% greater than CON and NGT, respectively (P < 0.05;
Figure 5-4B). The time of fat infusion did not alter the phase of fat yield. A daily rhythm of
protein yield was present in all treatments (Figure 5-4D). The amplitude was increased 30% by
DAY, and decreased 32% by NGT. The phase of the rhythm was delayed 2 h by DAY (P < 0.05),
but was not affected by NGT (P > 0.10).
100
Fat concentration fit a daily rhythm in CON and NGT (P < 0.05), but not DAY (P > 0.10;
Figure 5-4C). Corresponding with the loss of rhythm, the amplitude of fat concentration was
dramatically decreased by DAY (67%, P < 0.05), but was increased 29% by NGT, relative to
CON (P < 0.05). Similar to previous reports, fat concentration peaked in the middle of the day in
both CON and NGT (Rottman et al., 2014; Salfer et al., 2016). However, the phase of the rhythm
was altered by the timing of infusion, with DAY phase-advancing the rhythm 2.75 h and NGT
phase-delaying the rhythm 1 h relative to CON (P < 0.05). Similarly, protein concentration fit a
daily rhythm in CON and NGT (P < 0.05), but not DAY (P > 0.10; Figure 5-4E). The amplitude
was reduced by 50% in DAY (P < 0.05), but did not differ between CON and NGT (P > 0.10).
Furthermore, the acrophase of CON occurred at 1555 h and did not differ from DAY or NGT (P
> 0.10), but the phase occurred 1 h and 40 min earlier in DAY (1446 h) compared to NGT (1626
h; (P < 0.05).
Fat metabolism has been previously linked to circadian rhythms of milk synthesis. Ma et
al., (2013) observed differences in both the daily rhythms of expression of circadian clock and
lipid synthesis genes in wild type mice compared to mice lacking the gene encoding lipid-
metabolism regulator Thyroid-Hormone Responsive Spot 14. In the current experiment, phase
shifts in the daily rhythms of fat and protein yield and concentration occurred, these effects were
small (less than 3 hours) and likely not of biological significance. In contrast, altering the timing
of restricted feeding by 8 h caused an approximately 8 h shift in the phases of milk and protein
yield and fat and protein concentration (Chapter 3). The larger effects of FA infusion on the daily
rhythms of milk synthesis occurred through modification of the amplitude. Fatty acid infusion
during the day increased the robustness of the oscillations of milk, fat, and protein yield, while
reducing the amplitudes of fat and protein concentration. These changes in amplitude may occur
through altering amplitudes of molecular clock genes in the mammary gland. Dietary fat has been
shown to reduce the amplitude of daily rhythms of molecular clock expression and expression of
101
genes related to FA synthesis in the adipose tissue and liver of mice (Kohsaka et al., 2007).
However, to our knowledge, the direct effects of FA on mammary circadian clock gene
expression have not been investigated. Future research examining the role of the timing of FA
infusion on the molecular clock should be conducted.
Daily Yields and Daily Rhythms of Milk Fatty Acids
In addition to examining total fat yield, the effect of time of FA infusion on the origin of
milk FA was examined. The yield of FA originating from de novo synthesis in the mammary
gland (Σ < 16C) was affected by treatment (P = 0.001), with DAY and NGT reducing yields by
12% and 11% compared to CON (Figure 5-5A). Similarly, yields of mixed-source FA
originating both from de novo synthesis and preformed FA uptake from plasma (Σ 16C) were
decreased 9% by DAY and 6% by NGT compared to CON (P = 0.02). Yield of preformed FA (Σ
>16C) was not affected by treatment, however (P = 0.24). These results suggest that limiting the
time of FA absorption from the intestine reduced de novo FA synthesis in the mammary gland,
but that the time during which it is limited is irrelevant. Altering the timing or frequency of
feeding has previously been demonstrated to have no effect on apparent de novo FA synthesis
(Niu et al., 2014; Rottman et al., 2014). However, restricting feed intake to 16 h during the night
decreased de novo and mixed FA yields relative to 16 h feed restriction during the day,
suggesting a role for meal timing in mammary FA synthesis (Salfer et al., 2016).
Of the individual FA, three odd-chain FA (OCFA) emerged as having the greatest
decrease from CON in both DAY and NGT treatments (Figure 5-5C). These included C11:0
(28% decrease in DAY; 27% decrease in NGT; P = 0.03), C13:0 (28% decrease in DAY; 23%
decrease in NGT; P = 0.002), and C15:0 (23% decrease in DAY; 16% decrease in NGT; P =
0.0004). Along with this, OCFA and total odd- and branched-chain fatty acids (OBCFA) were
reduced in DAY and NGT relative to CON (P < 0.02; Figure 5-5B). Odd and branched chain FA
102
are derived primarily from microbial synthesis in the rumen, and found almost exclusively in
adipose tissue and milk of ruminants (Vlaeminck et al., 2006). Moreover, they may have a may
have a role in human heath because they have been shown to have anti-carcinogenic properties
(Yang et al., 2000). In the current experiment, oil was infused post-ruminally and major changes
in rumen fermentation would be unexpected. The effect may be indirectly modulated through gut
peptides. The gut peptides glucagon-like peptide 1 (GLP-1) and cholecystokinin (CCK) are
increased by intestinally-available unsaturated fatty acids and may modify rumen motility
(Bradford et al., 2008). Additionally, de novo synthesis of OBCFA can occur in the mammary
gland through elongation of the 3-carbon precursor propionyl -CoA, derived from ruminally-
produced propionate (Fievez et al., 2012). Similar to reducing total de novo FA synthesis,
restricting the time of FA availability appears to inhibit elongation of propionyl-CoA or other
OCFA. The effect appears to be fatty-acid specific because previous research demonstrates that
altering meal pattern through time-restricted feeding does not affect total or individual OBCFA
yields (Salfer et al., 2016; Salfer and Harvatine, 2018).
The daily rhythms of de novo, mixed, and preformed FA were affected by treatment.
Similar to previous reports (Rottman et al., 2014), de novo FA yield fit a daily rhythm in CON (P
= 0.01) and DAY (P = 0.03), but this rhythm was eliminated by NGT (P = 0.63; Figure 5-6A).
The amplitude of the rhythm was increased 19% by DAY and decreased 46% by NGT. The
acrophase of de novo FA yield occurred at 0603 h in CON and was phase-delayed 1.1 h by DAY
and phase-advanced 1.3 h by NGT. These small phase shifts are unlikely to be of biological
relevance. Similar changes due to time of FA infusion were observed for mixed FA yield, which
exhibited a daily rhythm in CON (P = 0.03) and DAY (P = 0.01), but not NGT (P = 0.68, Figure
5-6B). The amplitude of the daily rhythm of mixed source FA yield was increased 17% by DAY
and decreased 42% by NGT. Phase was also modified slightly by treatment with CON peaking at
0632, and being shifted 1 h later by DAY and 30 min earlier by NGT. These results suggest that
103
in addition to total de novo FA synthesis being affected by time of FA infusion, the daily rhythms
of de novo FA synthesis are also affected. Preformed FA failed to fit a rhythm in CON (P = 0.43)
or NGT (P = 0.68), but a rhythm was induced by DAY (P = 0.02; Figure 5-6C). Corresponding
with the induction of a rhythm according to the zero-amplitude test, DAY increased the amplitude
of the daily rhythm by 38% compared to CON and 33% compared to NGT. Moreover, the fitted
24 h rhythm was advanced 0.9 h by DAY and delayed 2.8 h by NGT, compared to CON which
peaked at 0812 h.
Daily Rhythms of Plasma Metabolites
Plasma glucose concentration has been demonstrated to follow a daily rhythm in cows
(Giannetto and Piccione, 2009; Niu et al., 2014). However, in our experiment, glucose failed to fit
a 24 h rhythm in any treatment according to the zero-amplitude test, nor did it fit a 24 h rhythm
with a 12 h harmonic (P > 0.20; Figure 5-7A). In model organisms, these rhythms are shown to
be influenced by peripheral circadian clocks in the liver and pancreas (Qian and Scheer, 2016).
However, nutrient intake has been shown to affect the daily rhythm of cows. Feeding 4x/d has
been shown to dampen the daily rhythm (Rottman et al., 2014), and the phase of the rhythm is
shifted by altering the time of feed availability (Chapter 3). Our results may suggest that FA
infusion reduces the amplitude of the daily pattern of glucose concentration, such that it fails to fit
a 24 h rhythm according to the zero-amplitude test, but a negative control was not included in this
experiment so this cannot be confirmed.
A daily rhythm of NEFA concentration was present in both CON (P = 0.003) and NGT
(P < 0.0001), but this rhythm was eliminated by DAY (P = 0.35; Figure 5-7B). In conjunction
with an elimination of the rhythm according to the zero-amplitude test, the amplitude of the 24 h
cosine function was reduced by DAY (P < 0.05). The phase of the 24 h rhythm was also affected
by treatment, with DAY phase advancing the rhythm 4.5 h and NGT phase delay the rhythm ~9 h
104
relative to CON. These results suggest that the time of fat absorption alters the daily patterns of
lipid mobilization from adipose tissue. Specifically, infusion during the morning dampens the
rhythm, while infusion overnight does not. Moreover, NEFA concentration decreased during
infusion in both DAY an NGT treatments, both reaching a nadir near the end of the infusion
period. The change in the daily pattern of NEFA concentration may be through entrainment of the
adipocyte circadian clock. In laboratory models, plasma free FA concentrations follow daily
rhythms that are under the control of the adipocyte molecular clock (Shostak et al., 2013).
Furthermore, lipids can entrain the molecular circadian clock of peripheral tissues through the
peroxisome proliferator-activated receptor (PPAR) family of nuclear receptors, which can bind to
response elements on the promoter regions of circadian clock genes BMAL1 and REV-ERBα and
enhance their expression. Interestingly, our study suggests that altering the time of FA absorption
may uncouple FA metabolism from glucose metabolism. While the phase of the NEFA rhythm is
shifted by the time of infusion during the day, glucose metabolism was not. Direct absorption of
FA may desynchronize the circadian clock of adipose tissue from the clocks of the liver and
pancreas that drive the daily rhythm of glucose concentration.
Treatment tended to affect concentration of plasma urea nitrogen (P = 0.06), with DAY
increasing PUN 20% compared to CON (P = 0.03) and 16% compared to NGT (P = 0.07; Figure
5-7C). This change in PUN was associated with a 16% increase in milk urea nitrogen in DAY
compared to control (Supplemental Figure S5-1), but no difference in milk protein
concentration or yield. Plasma and milk urea nitrogen concentrations are measures of nitrogen
utilization efficiency (Jonker et al., 1998; Kohn et al., 2005). Results suggest that FA infusion
during the day may decrease the efficiency of amino acid utilization by muscle and other non-
mammary tissues. A 24 h rhythm fit all treatments (P < 0.0005), and treatment affected the daily
rhythm. These results agree with previous research suggesting that cows exhibit daily rhythms of
plasma urea nitrogen concentration (Giannetto and Piccione, 2009; Niu et al., 2017). Day-
105
infusion increased the amplitude of the daily rhythm 26% compared to CON and 18% compared
to NGT (P < 0.05). This increase in amplitude may be related to the increase in average PUN by
NGT. The phase of the rhythm was slightly modified by treatment with the acrophase of NGT
occurring 29 min later than CON and 41 min later than DAY (P < 0.05), but this effect is likely
not biologically relevant. Salfer et al. (2016) previously observed that altering the time of time-
restricted feeding inverts the daily rhythms of PUN, but our results suggest that this effect was
not mediated by FA intake.
Daily Rhythm of Body Temperature
The time of FA infusion altered average body temperature (P = 0.02), with NGT
decreasing body temperature compared to CON and DAY (Figure 5-8). Body temperature fit a
24 h rhythm in all treatments (P < 0.0001) and the rhythm was altered by treatment. The
amplitude of the daily rhythm was increased 92% by DAY and 125% by NGT (P < 0.05).
Furthermore, the phase of the daily rhythm of body temperature was delayed 50 min by DAY, but
was not affected by NGT.
Previous research has suggested that altering the time of feed restriction dramatically
shifts the phase of the daily rhythm of body temperature. We previously observed a ~9 h phase
shift in the body temperature rhythm when feed was restricted to 16 h either during the day or
night (Chapter 4). Similar results were observed by Niu et al. (2014) who detected a 3 h shift in
the phase of body temperature when cows were fed at 0830 versus 2030 h, without restricting the
time of feed availability. The results of the current study contrast this previous work and suggest
that the time of fat absorption has very little effect on the phase of the rhythm of body
temperature. The discrepancy between the experiments may be related to rumen temperature.
Rumen temperature also oscillates in a circadian manner (Piccione et al., 2014). Moreover, rumen
temperature is directly correlated to core body temperature, and is related to dry matter intake
106
(Beatty et al., 2008). This relationship may explain why only slight differences in the daily
rhythm of body temperature occurred after abomasal infusion of fat, whereas changing the time
of feed intake has previously been shown to alter the rhythm, perhaps by changing the daily
rhythm of body temperature through altered rumen fermentation.
Interestingly, the amplitude of the daily rhythm of body temperature was dramatically
altered by time of feed restriction, with night-infusion increasing the rhythm to a greater extent
than day infusion. Similarly, we previously reported a 4-fold increase in the amplitude of the
daily rhythm when time of feed restricted was limited to the overnight period compared to the day
(Chapter 4). Therefore, the timing of fat infusion may influence the daily rhythm of body
temperature thorough altering the robustness, but not the phase of the rhythm. The increase in
amplitude of the rhythm may be important in modifying the circadian rhythms of other tissues
because core body temperature has been shown to be capable of circadian entrainment in
mammals (Brown et al., 2002).
CONCLUSIONS
The daily rhythms of milk synthesis were altered by modifying the time of FA infusion.
Most notably, infusion during the day increased the amplitude of milk yield, but decreased
concentration of fat and protein. However, FA do not appear to robustly entrain the mammary
clock because the phase of the rhythms of milk, fat, and protein yield and milk fat and protein
concentration were not markedly shifted by treatment. De novo FA synthesis was reduced by
restricting the time of FA infusion, regardless if time restriction occurred during the day or night.
Daily rhythms of glucose and nitrogen metabolism were not greatly affected by time of FA
infusion, but total nitrogen efficiency appeared to be decreased by day-infusion. The daily pattern
of lipid mobilization was altered by time of fat absorption, however, with day-infusion flattening
107
the daily rhythm. The amplitude of the daily body temperature was increased both by restricting
fat infusion to either the day or night, indicating that time-restricted fat absorption may increase
the robustness of the central circadian clock. The effects of time of FA infusion on the daily
rhythms of milk synthesis may be related to increasing the robustness of oscillations of the
mammary molecular clock. Moreover, the changes in the daily patterns of NEFA concentrations
may be related to the adipocyte clock. The role of FA on the molecular circadian clocks of
mammary and adipose tissue should be investigated in the future.
108
FIGURES
Figure 5-1. Schedule of feeding, lighting and sampling during the experiment.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to 0500].
109
Figure 5-2. Effect of time of fatty acid infusion on dry matter intake.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to 0500].
Data are presented as LSM with SEM bars.
110
Figure 5-3. Effect of time of fatty acid infusion on daily milk, fat and protein yield and fat and
protein concentration.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Panels the effect of time of fatty acid infusion on (A) milk yield (kg), (B) fat and protein yield (g)
and (C) fat and protein concentration. Data are presented as LSM with SEM bars.
111
Figure 5-4. The effects of time of fatty acid infusion on the daily rhythms of milk synthesis.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Panels show the effect of time of fatty acid infusion on the daily rhythms of milk (A) yield (kg), (B)
fat yield (g), (C) protein yield (g), (D) fat concentration (%), and (E) protein concentration (%). Data are
presented as relative expression LSM and SEM bars and cosine fit are shown. Cosinor analysis is shown
below each panel and include the amplitude (Amp; difference between peak and mean), acrophase (Acro;
time at peak of the rhythm), and the P-value of the zero-amplitude test. NS: Reduced model fits data
significantly better than daily rhythm model. The black and white bars above the x-axis display the light:
dark cycle.
112
Figure 5-5. Effect of time of fatty acid infusion on the daily yields of fatty acids.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Panels show the effect of time of fatty acid infusion on the average daily yields of (A) de novo (Σ <
16C), mixed (Σ 16C) and preformed (Σ >16C) sources of fatty acids (g/d), (B) odd and branched chain fatty
acids (OBCFA), odd chain fatty acids (OCFA), and branched chain fatty acid (BCFA) (g/d), and (C)
C11:0, C13:0 and C15:0 fatty acids (g/d). Data are presented as LSM with SEM bars. Means with differing
superscripts are different at P < 0.05.
113
Figure 5-6. Effect of time of fatty acid infusion on daily rhythms of fatty acids by source.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Panels show the effect of time of fatty acid infusion on the daily rhythms of (A) de novo synthesized
(Σ < 16C) FA yield (g/d) (B) mixed source (Σ 16C) FA yield (g/d), and (C) preformed (Σ >16C) FA yield
(g/d). Data are presented as relative expression LSM and SEM bars and cosine fit are shown. Cosinor
analysis is shown below each panel and include the amplitude (Amp; difference between peak and mean),
acrophase (Acro; time at peak of the rhythm), and the P-value of the zero-amplitude test. NS: Reduced
model fits data significantly better than daily rhythm model. The black and white bars above the x-axis
display the light: dark cycle.
114
Figure 5-7. The effects of time of fatty acid infusion on daily rhythms of plasma metabolites.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Panels include the effects of time of fatty acid infusion on daily rhythms of plasma (A) glucose
concentration (mg/dL), (B) nonesterified fatty acid concentration (NEFA; μEq/L), and (C) urea nitrogen
concentration (mg/dL). Data are presented as relative expression LSM and SEM bars and cosine fit are
shown. Cosinor analysis is shown below each panel and include the amplitude (Amp; difference between
peak and mean), acrophase (Acro; time at peak of the rhythm), and the P-value of the zero-amplitude test.
NS: Reduced model fits data significantly better than daily rhythm model. The black and white bars above
the x-axis display the light: dark cycle.
115
Figure 5-8. The effect of the time of fatty acid infusion on the daily rhythms of core body
temperature.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Data presented as 2 h means and SEM bars of body temperature collected every 10 min by a vaginal
temperature data logger, with cosine fit are shown. Cosinor analysis is shown below each panel and include
the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of the rhythm),
and the P-value of the zero-amplitude test. NS: Reduced model fits data significantly better than daily
rhythm model. The black and white bars above the x-axis display the light: dark cycle
116
SUPPLEMENTAL FIGURES
Supplemental Figure S5-1. Effect of time of fatty acid infusion on milk urea nitrogen
concentration.
Treatments were 350g/d of fatty acids abomasally infused continuously for 22 h/d (CON); for 8 h/d during
the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Data are presented as LSM with SEM bars. Means with differing superscripts are different at P <
0.05.
117
SUPPLEMENTAL TABLES
Supplemental Table S5-1. Diet and nutrient composition of the experimental diet.
1Contained (% of DM): 33.0% NDF, 16.9% ADF 2Contained (% of DM): 45.1% NDF, 34.7% ADF 3Contained (% of DM): 63.2% NDF, 33.5% ADF 4AminoPlus, Ag Processing Inc., Omaha, NE 5Contained (%, as-fed basis): 37.0% Calcium carbonate; 29.9% dried corn distillers grains; 24.5% salt;
4.2% magnesium oxide (54% Mg); 2.4% organic phosphorus (15% P); 0.5% zinc sulfate; 0.2% mineral oil.
Composition (DM-basis): 7.2% CP; 7.1% NDF; 3.7% ADF; 15.0% Ca; 0.75% P; 0.33% K; 2.6% Mg; 0.5%
S; 9.8% Na; 23.0 mg/kg Co; 652 mg/kg Cu; 783 mg/kg Fe; 54.0 mg/kg I; 1190 mg/kg Mn; 12.8 mg/kg Se;
1718 mg/kg Zn; 88,700 IU/kg vitamin A (retinyl acetate); 28,400 vitamin D (activated 7-
dehydrocholesterol); 850 IU/kg vitamin E (dl-α tocopheryl acetate). 5Fed as coated urea (Optigen, Alltech Inc., Lexington, KY; 259% CP, DM basis.
Item Composition
Ingredients, % of DM
Corn silage1 49.6
Alfalfa haylage2 15.0
Canola meal 12.5
Ground corn 11.6
Roasted soybeans 3.6
Grass hay/straw3 2.9
Mechanically-extracted soybean meal4 2.9
Vitamin/mineral mix5 1.9
Controlled-release N6 0.25
Nutrients, % of DM
NDF 33.1
ADF 18.3
Ash 5.6
118
Supplemental Table S5-2. Fatty acid profile of oil used for abomasal infusion of fatty acids.
Name g/100 g
14:0 0.22
15:0 0.01
16:0 5.11
cis-9 16:1 0.04
17:0 0.03
C18:0 1.39
trans-9 18:1 0.26
trans-10 18:1 0.02
trans-11 18:1 0.38
cis-9 18:1 79.6
cis-11 18:1 0.56
cis-9, cis-12 18:2 10.0
cis-6, cis-9, cis-12 18:2 0.15
cis-9, cis-12, cis-15 18:2 0.36
119
Chapter 6
The effect of timing of protein infusion on the daily rhythms of milk
production and plasma hormones and metabolites
Isaac J. Salfer, Rebecca A. Bomberger, Cesar I. Matamoros, P.A. Bartell, Kevin J. Harvatine
120
ABSTRACT
Dairy cows have a daily rhythm of milk synthesis that appears to be driven by the
molecular clock of the mammary gland and is modified the time of feed availability. Protein
metabolism is intimately linked to circadian rhythms in other species, but the effect of amino
acids on the daily rhythm of milk synthesis is not known. The objective of this experiment was to
determine the effects of intestinally-absorbed protein on the daily rhythms of milk and milk
component synthesis in dairy cows. Nine lactating Holstein cows were randomly assigned to one
of three treatment sequences in a 3 x 3 Latin square. Treatments were absomasal infusions of 500
g/d of sodium caseinate either continuous throughout the day (CON), for 8 h/d from 0900 to 1700
h (DAY), or for 8 h/d from 2100 to 0500 (NGT). Treatment periods were 14 d with 11 d of
treatment adaptation and 3 d of sampling. Cows were milked every 6 h during the final 8 d of
each period. A 24-h rhythm was fit to time-course production data using cosine analysis and the
amplitude and acrophase (time at peak) were determined. Daily milk and protein yield were
decreased 7.5% and 8% by NGT compared to CON, respectively, while milk fat yield was
increased by DAY (P < 0.05). Daily milk fat and protein concentration were not affected by
treatment. Milk yield failed to fit a 24 h rhythm in CON or DAY, but a rhythm was induced by
NGT (P = 0.03). Neither fat yield nor protein yield fit a rhythm in either treatment. However, fat
concentration fit a daily rhythm in all treatments (P < 0.05) and the amplitude of the rhythm was
decreased 57% by DAY and 26% by NGT (P < 0.05). The rhythm of milk fat concentration was
phase advanced ~2 h by DAY and phase delayed ~1 h by NGT (P < 0.05). Protein concentration
fit a daily rhythm in CON and DAY but not NGT. The phase of the rhythm of protein
concentration was delayed ~1.25 h by DAY and the amplitude was increased 2-fold relative to
CON (P < 0.05). The daily yields of de novo (Σ < C16), mixed (Σ 16C), and preformed (Σ >16C)
fatty acids in milk were not altered by treatment (P > 0.10), and none of these sources fit a daily
121
rhythm in any treatment (P > 0.10). Plasma glucose concentration fit a daily rhythm in CON and
NGT, but the rhythm was eliminated by DAY. Furthermore, the phase of glucose concentration
was not altered by treatment. Nonesterified fatty acid concentration did not fit a daily rhythm in
CON or NGT (P > 0.10), but a rhythm was induced by DAY (P < 0.05). A rhythm of plasma urea
nitrogen was present in CON and NGT (P < 0.0001) but not DAY, (P < 0.05), with no difference
in phase between CON and NGT. The time of protein infusion influenced the daily rhythms of
milk and milk protein synthesis, with night infusion abolishing rhythms of protein concentration
and inducing rhythms of milk yield, and day infusion increasing the amplitude of the rhythm of
protein concentration. Moreover, infusion of protein during the day damped the rhythms of
plasma glucose and PUN concentration, but increased the amplitudes of nonesterified FA,
suggesting that the daily rhythms of nutrient metabolism are also altered by the time of protein
infusion.
Keywords: Daily rhythm, milk synthesis, nutrient entrainment, circadian rhythm
122
INTRODUCTION
Dairy cows exhibit rhythms of milk synthesis, with milk yield typically peaking in the
morning and milk fat and protein concentration peaking in the evening (Rottman et al., 2014).
These daily rhythms appear to be related to changes in the molecular clock of the mammary gland
(Plaut and Casey, 2012). The molecular clock is responsible for generating circadian rhythms,
which are repeating ~24 h cycles that allow animals to anticipate daily changes in their
environment. This molecular clock is comprised of a set of ‘clock’ transcription factors and daily
rhythms of phosphorylation and dephosphorylation events that generate 24 h rhythms of gene
expression and intracellular signaling (Robles et al., 2017).
A master circadian oscillator located in the suprachiasmatic nucleus (SCN) of the
hypothalamus serves as the central control point for setting the circadian rhythms of peripheral
tissues. This master oscillator is entrained by the light-dark cycle and initiates neural and
hormonal signaling cascades to entrain the molecular clock of cells throughout the body. In
addition to receiving signals from the SCN, the molecular clock in peripheral tissues can be
entrained by other signals, including nutrient intake (Stokkan et al., 2001). Entrainment by
nutrients can occur independently of the SCN and can cause desynchronization of master and
peripheral circadian rhythms (Asher and Sassone-Corsi, 2015). Desynchronization of these
rhythms has been implicated in development of metabolic disorders such as obesity and type II
diabetes (Vetter and Scheer, 2017).
Nutrient intake appears to influence the circadian rhythms of milk synthesis in dairy
cows. Feeding cows four times per day in equal meals dampened the daily rhythms of milk fat
and protein concentration, as well as shifted peaks of milk, fat, and protein yield and fat and
protein concentration (Rottman et al., 2014). Salfer et al. (2016) discovered that restricting the
time of feed availability to 16 h/d at night (1900 to 1100) shifted the daily rhythms of milk yield
123
and fat and protein concentration approximately 8 h relative to restricting feed availability to the
same amount of time during the day (0700 to 2300). The change in these daily patterns of milk
synthesis appear to be at least partially modulated through modifying the molecular clock because
night-restricted feeding shifts the daily rhythms of some components of the molecular clock
(Chapter 4). Similar results were observed in mice under night-restricted feeding, which exhibited
shifts in the rhythms of milk fat synthesis, along with changes in the mammary molecular clock
(Ma et al., 2013). In addition to modifying the time of feed intake through time-restricted feeding,
individual nutrients have been shown to be capable of entraining molecular clocks (Oike, 2017).
In dairy cows, altering the time of fatty acid (FA) infusion shifts the amplitude and phase of milk,
fat, and protein synthesis in the mammary gland (Chapter 5).
While fatty acids appear to entrain daily rhythms of milk synthesis of dairy cows, to our
knowledge, other nutrients have not been examined. Amino acids may have a potential role in the
circadian rhythm of the mammary gland. Amino acid metabolism is intimately linked with the
molecular clock of mammals. In humans, plasma concentrations of total amino acids, and all
individual proteogenic amino acids follow circadian rhythms that are entrained by alterations to
the sleep-wake cycle (Feigin et al., 1968). Furthermore, Krüppel-like factor 15 (KLF15), a
transcription factor responsible for upregulation of pathways related to amino acid mobilization
from muscle and their use for gluconeogenesis during fasting, is regulated in a circadian manner
by glucocorticoid signaling (Fan et al., 2018). This pathway links amino acid metabolism with
circadian rhythms of plasma glucose concentrations.
Amino acids are important for milk protein synthesis in the mammary gland and play a
role in the synthesis of lactose because α-lactalbumin is a milk protein and is a regulatory
component of the lactose synthase enzyme complex. Altering the time of amino acid absorption
may affect the efficiency of protein synthesis in the mammary gland and/or alter the daily
rhythms of milk synthesis. The objective of this experiment was to examine the effect of time of
124
protein absorption on the daily rhythms of milk synthesis and plasma metabolites. We expect that
altering the timing of amino acid availability in the small intestine through abomasally infusing
sodium caseinate either continuously throughout the day, for 8 h during the day, or for 8 h during
the night, will shift the phase of the daily rhythms of milk yield and milk protein concentration
and yield and will shift the daily rhythms of plasma glucose concentrations.
MATERIALS & METHODS
Animals and Treatments
Nine cannulated multiparous mid-lactation (128 ± 46 d postpartum; mean ± SD) Holstein
cows from the Penn State University Dairy Research and Teaching Center were randomly
assigned to treatment sequences in a 3x3 Latin Square design. Treatment periods were 14 d with
11 d of treatment adaptation followed by 3 d of sampling. A 7 d washout was used between
treatment periods. During the final 7 d of each period, cows were milked 4 times per day to allow
observation of daily rhythm. Treatments included abomasal infusion of 500 g/d day of sodium
caseinate dissolved in 6 L of distilled water for either 24 h/d (CON), for 8 h/d from 0900 h to
1700 h (DAY), or for 8 h/d from 2100 h to 0500 h (NGT; Figure 6-1A). Cows on the CON
treatment received infusion for approximately of 22 h/d as infusion occurred for 30 min during
each milking. Sodium caseinate (Erie Foods International, Erie, IL) was used as a protein source
because of its high biological value for milk protein synthesis. Amino acid profile of sodium
caseinate was analyzed by The Experiment Station Chemical Laboratories at the University of
Missouri (Columbia, MO) according to (AOAC, 2006) (Supplemental Table S6-2). All cows
were provided the same basal TMR fed once daily at 0600 h at 110% of the previous day’s intake
(Supplemental Table S6-1). Feed samples were collected weekly throughout the course of the
experiment, composited by period, and analyzed for dry matter, crude protein, starch, neutral
125
detergent fiber (NDF), acid detergent fiber (ADF), and ash according to Rico and Harvatine
(2013). The experiment was conducted from July 30 to September 23, 2018 and all experimental
procedures were approved by the Penn State University Institutional Care and Use Committee.
Milk Sampling and Analysis
Cows were milked 4 x/d every 6 h (0500, 1100, 1700, and 2300) on the final 8 d of each
period to observe the daily pattern of milk synthesis. Milk collected at each time point
represented total milk synthesis over the previous 6 h and is plotted as the midpoint of the
previous milking interval (3 h prior to collection). Milk yield was measured at each milking using
an integrated milk meter (AfiMilk MPC Milk Meter; Afimilk Agricultural Cooperative Ltd.,
Kibbutz Afikim, Israel), and yields were corrected for the deviation of each individual stall
according to Andreen et al. (2018). Milk was sampled at all milkings on d 13 and 14 of each
period, with one subsample used for analysis of fat and protein concentration by Fourier
transform infrared spectroscopy (Fossomatic 4000 Milko-Scan and 400 Fossomatic, Foss
Electric; Dairy One DHIA, Ithaca, NY), and one subsample used for analysis of FA
concentrations according to Baldin et al. (2018).
Plasma Sampling and Analysis
Blood was collected via venipuncture of a coccygeal blood vessel into potassium-EDTA
vacuum tubes (BD Vacutainer, Becton Dickinson, Franklin Lakes, NJ) on d 13 to 14 of each
period. Sampling occurred at 6 time points representing every 4 h across the day (0300, 0700,
1100, 1500, 1900, and 2300 h). Samples were placed on ice immediately after collection and
centrifuged at 2300 x g for 15 min at 4°C within 30 min of collection. Plasma was collected and
stored at -20°C for analysis of glucose, nonesterified fatty acids (NEFA), and plasma urea
nitrogen concentrations (PUN) as described by Rottman et al. (2014). Briefly, glucose was
126
measured using a glucose oxidase/peroxidase enzymatic colorimetric assay (No. P 7119, Sigma-
Aldrich, St. Louis, MO), NEFA were measured using an acyl-CoA oxidase peroxidase enzymatic
colorimetric assay (NEFA-HR (2), Wako Diagnostics, Richmond, VA) and urea nitrogen was
measured using a modified Berthelot methodology (Modified Enzymatic Urea Nitrogen
Procedure No. 2050; Stanbio Laboratory, Boerne, TX).
Statistical Analysis
All data were analyzed using the MIXED procedure of SAS 9.4 (SAS Institute, Cary,
NC). Effects of time of protein absorption on daily dry matter intake, milk production, and milk
FA yields were tested using models that included the fixed effect of treatment and the random
effects of cow and period. Denominator degrees of freedom were adjusted using the Kenward-
Roger method, and the AR(1) or ARH(1) covariance structure was used based on convergence
criteria. Preplanned contrasts were used to compare least squared means among individual
treatments.
Time course data for milk production and plasma metabolites were fit to the linear form
of a cosine function with a 24 h period using random regression in SAS 9.4. Model parameters
included the fixed effects of treatment, cosine terms, and the interaction between treatment and
cosine terms as well as the random effects of cow and period. The daily patterns of plasma
metabolites were tested for the presence of a 12 h harmonic, and the harmonic was included in the
24 h rhythm if it improved model fit. The fit of the 24 h cosine function was determined using a
zero-amplitude test and the amplitude and acrophase (time at peak) and their significance were
determined according to Niu et al. (2014). Data points with Studentized residuals outside ± 3.5
were removed. Another linear model including the fixed effects of treatment, time, and their
interaction, with random effects of cow and period was also tested. Statistical significance was
127
declared at P < 0.05 and trends acknowledged at 0.05 < P < 0.10. High resolution figures were
generated using an add-in for Microsoft Excel (Kraus, 2014).
RESULTS AND DISCUSSION
Daily Milk Yield and Milk Components
Dry matter intake was not altered by treatment (P = 0.46; Figure 6-2). Daily milk yield
was decreased 7.4% by NGT relative to CON but NGT and DAY were not different (P = 0.02;
Figure 6-3A). Treatment tended to affect fat yield (P = 0.07), with DAY increasing fat yield
4.5% relative to CON and 4.6% relative to NGT (Figure 6-3B). The yield of protein was also
affected by treatment (P = 0.005; Figure 6-3B), with NGT decreasing protein yield 8.3% relative
to CON (P = 0.01) and 6.9% relative to DAY (P = 0.002). However, daily average concentration
of fat (P = 0.34) and protein (P = 0.61) were not affected by treatment (Figure 6-3C).
The results of the current experiment differ from previous research in our lab examining
the timing of FA infusion on milk synthesis, which exhibited no differences in milk or milk
components when FA were infused either continuously or over 8 h/d from 0900 to 1700 or 8 h/d
from 2100 to 0500 (Chapter 4). Postruminal sodium caseinate infusion consistently increases
milk protein yield in dairy cows (Clark, 1975; Mackle et al., 1999). This increase in milk protein
is attributed to the high biological value of sodium caseinate, with the amino acid profile of
sodium caseinate closely matching that required for milk protein synthesis. Furthermore, specific
amino acids, such as leucine and isoleucine can directly increase milk protein synthesis via
mechanistic target of rapamycin (mTOR) signaling (Appuhamy et al., 2012). Furthermore,
protein supply may increase milk lactose by increasing α-lactalbumin, a regulator of the lactose
synthesis pathway, or by providing glucogenic precursors that may increase lactose by increasing
plasma glucose availability (Kuhn et al., 1980). Lactose is an osmotic regulator of milk and is
128
drives milk yield (Schingoethe, 1996). These mechanisms by which protein can increase milk and
milk protein synthesis are variable, and depend on the availability of protein and glucogenic
precursors. In the current experiment, lactose yield tended to be decreased by NGT, but not DAY
(Supplemental Figure S 6-1). This reduction in lactose yield may be responsible for the
reduction in milk yield. The effect of time of protein infusion on milk, fat, and protein yield may
be related to the mammary clock. Hu et al. (2017) discovered that siRNA-mediated knockdown
of the PER2 gene in bovine mammary epithelial cells increased mRNA and protein abundance of
αs1- and αs2- casein. Restricting time of protein infusion to the night may similarly affect the
mammary clock, which may reduce milk yield.
Daily Rhythms of Milk Yield and Milk Components
Previous research has demonstrated that milk yield follows a 24 h rhythm (Gilbert et al.,
1972; Rottman et al., 2014). In the current experiment, milk yield fit a daily rhythm in NGT (P =
0.03), with an amplitude of 0.72 kg and an acrophase of 0330 (Figure 6-4A). However, milk
yield failed to fit a daily rhythm in CON (P = 0.12) or DAY (P = 0.25). These results imply that
infusion of protein throughout the day or during the morning dampened the rhythm of milk
synthesis, whereas infusion at night does not. A negative control treatment without protein
infusion was not included due to limitations in conducting the intensive experiment and the
hypothesis was focused on the timing of absorption.
Fat concentration fit daily rhythms with similar amplitudes in CON (P = 0.003) and NGT
(P = 0.01), but the rhythm was abolished by DAY (P = 0.26; Figure 6-4C). The acrophase of the
24 h rhythm was phase-advanced 1.5 h by DAY and phase advanced 1.75 h by NGT, relative to
CON (P < 0.05). Protein concentration fit a daily rhythm in CON (P = 0.0004) and DAY (P <
0.0001), but not NGT (P = 0.49; Figure 6-4E). Day-infusion increased the amplitude 2-fold
compared to CON, while NGT decreased the amplitude by 50% (P < 0.05). Similar to previous
129
reports (Rottman et al., 2014), the acrophase occurred in the afternoon in all treatments, with
DAY and NGT phase delaying the rhythm 1.25 and 4 h, respectively (P < 0.05).
Fat yield failed to fit a daily rhythm in any treatment (P > 0.38 Figure 6-4B). Despite the
failure to fit a rhythm via the zero-amplitude test, a post-hoc analysis comparing the amplitude
and acrophase of the fitted cosine function were performed. The acrophase of the fitted 24 h
cosine appeared to be phase-advanced 5.25 h from 0917 h in CON to 0402 h in DAY (P < 0.05).
Protein yield similarly failed to fit a daily rhythm in any treatment according to the zero-
amplitude test (P > 0.13; Figure 6-4D). The amplitude of the 24 h cosine function was increased
106% by NGT, compared to CON (P < 0.05). Moreover, the phase of the fitted cosine function
was shifted 4.25 earlier by DAY but was not affected by NGT.
The timing of nutrient intake has previously been shown to impact the daily rhythms of
milk synthesis in dairy cows. Rottman et al. (2014) observed a reduction in the amplitude of milk
fat and protein concentration when cows were fed 4x/d in equal meals compared to 1x/d.
Moreover, restricting the time of feed availability to 16 h/d during the night shifted the peak of
milk and milk protein yield to the afternoon, and the peak of fat and protein concentration to the
morning, relative to restricted feeding for 16 h during the day (Chapter 3). Furthermore, timing of
FA infusion has been shown to affect the daily rhythms of milk synthesis, with infusion during
the day causing a reduction in the amplitudes of milk and milk fat and protein yield, but increased
amplitudes of fat and protein concentration (Chapter 5). Results of the current study demonstrate
that the time of protein infusion also modifies the daily rhythms of milk synthesis. Specifically,
protein infusion during the day reduces the amplitude of milk fat concentration and increases the
amplitude of milk protein concentration, while protein infusion at night reduces the amplitude of
milk protein concentration.
130
Daily Yields and Daily Rhythms of Milk Fatty Acids
Because the timing of protein infusion tended to affect total daily fat yield, we examined
the effect of treatment on yields of FA by source. Time of protein infusion did not affect the daily
yields of FA originating from de novo synthesis in the mammary gland (Σ < 16C; P = 0.12),
preformed FA taken up from plasma (Σ >16C; P = 0.14), or mixed source FA (Σ 16C; P = 0.66;
Figure 6-5A). Day-infusion numerically increased the yields of de novo, mixed, and preformed
FA, suggesting that the increased total FA yield was due to the cumulative effects of slight
increases in both de novo FA synthesis and preformed FA uptake from plasma. We have
previously observed that the time of feed intake affects de novo FA synthesis, with night-
restricted feeding for 16 h/d (1900 to 1100) reducing de novo and mixed FA yields, but not
preformed FA, relative to day-restricted feeding for 16 h/d (0700 to 2300; Chapter 3).
Furthermore, we observed that restricting the time of FA infusion to the day (0900 to 1700) or
night (2100 to 0500) reduced de novo synthesis relative to continuous infusion across the day
(Chapter 5). Results of the current study suggest that unlike altering the time of feed restriction or
the time of FA infusion, the time of protein infusion does not alter de novo FA synthesis.
In addition to the daily yields of FA, we examined the effect of time of protein infusion
on the 24 h rhythms of de novo, mixed, and preformed FA yield. Previous research has
demonstrated that these sources of FA oscillate over 24 h (Rottman et al., 2014; Ma et al., 2015).
According to the zero-amplitude test, none of the treatments fit daily rhythms of de novo, mixed,
or preformed FA (P > 0.20; Figure 6-5B-D). The present study may indicate that infusion of
sodium caseinate eliminates these rhythms, but this cannot be definitively determined without a
negative control. Abomasal infusions of FA during the NGT have been shown to eliminate the
daily rhythms of de novo, mixed and preformed FA yields, whereas infusion during the day did
not (Chapter 4). Despite not fitting a rhythm in any treatment, there were detectable changes in
the fitted 24 h cosine functions of de novo, mixed, and preformed FA yield due to treatment. The
131
amplitude of the rhythm of de novo FA yield was increased over 6-fold by DAY and over 3-fold
by NGT, relative to CON (P < 0.05), with DAY having a 92% greater amplitude than NGT
(Figure 6-5B). Moreover, the phase of the daily rhythm was advanced 11.6 h by DAY and
delayed 10.2 h by NGT (P < 0.05). However, these results should be interpreted with caution
because their low amplitudes hinder the ability to fit a cosine function with an appropriate phase.
Relative to CON, the amplitude of mixed FA yield was increased nearly 4-fold by DAY, but was
reduced 52% by NGT (P < 0.05). Similar to de novo FA, the phase was advanced 10 h by DAY
and delayed 10.4 h by NGT, relative to CON (P < 0.05). Day-infusion decreased the amplitude of
preformed FA yield 48%, but NGT increased the amplitude 7% compared to CON (P < 0.05).
The phase of the rhythm was shifted 6.8 h earlier by DAY and 3.9 h earlier by NGT (P < 0.05).
Taken together, these results imply that although sodium caseinate seemingly causes dramatic
reductions in the amplitude of the daily rhythms of the three major sources of FA such that they
do not fit a 24 h rhythm, time of protein infusion still causes apparent changes in these low-
amplitude rhythms.
Daily Rhythms of Plasma Metabolites
Treatment affected the daily rhythms of plasma glucose concentration (Figure 6-6A). A
24 h rhythm was present in CON (P = 0.03) and NGT (P = 0.02), but was abolished in DAY (P =
0.85). Moreover, the amplitude of the rhythm was decreased 66% by DAY, but was increased
47% by NGT, relative to CON. Alternatively, the acrophase of the daily rhythm was not altered
by treatment. Furthermore, average plasma glucose concentration was not affected by treatment
(P = 0.14). Results were similar to previous research demonstrating that altering the time of FA
infusion had no effect on average glucose concentrations or the phase of the daily rhythm
(Chapter 4). However, altering the time of feed availability has previously been shown to entrain
daily rhythms of glucose concentration. Feeding 4x/d reduced the amplitude and phase-delayed
132
the rhythm 8 h relative to 1 x/d feeding (Rottman et al., 2014). Moreover, we previously
demonstrated that restricting the time of feed availability to 16 h/d during the night (1900 to
1100) increased the amplitude 2.5-fold and phase delayed the rhythm by 9.8 h relative to day-
restricted feeding (16 h /d from 0700 to 2300 h) (Chapter 3).
In dairy cows, plasma glucose is almost exclusively derived from gluconeogenesis in the
liver using propionate as a substrate (Aschenbach et al., 2010). Therefore, the circadian clock of
the liver may be linked to daily rhythms of glucose metabolism. In model organisms, the
circadian clock of the liver is highly responsive to food entrainment (Stokkan et al., 2001).
Maugeri et al. (2018) proposed that the molecular clock of the liver can be entrained by a protein-
only diet or treatment with cysteine, and that this effect was mediated through glucagon and IGF-
1 signaling. The changes in circadian amplitude due to the time of protein infusion may be
through increasing or decreasing the robustness of oscillations of the liver clock. The daily
rhythms of glucose metabolism may also be controlled by rhythms of insulin secretion. We have
previously demonstrated that insulin concentration follows a 24 h rhythm that is responsive to
feeding time (Chapter 3). In model organisms, the daily rhythms of insulin secretion are secreted
by a circadian clock in the pancreas, and this clock may be entrained by timing of nutrient
absorption (Allaman-Pillet et al., 2004). Further research examining the role of protein and amino
acids on the molecular clocks of the liver and pancreas should be conducted. The increase in
amplitude of the plasma glucose concentrations due to protein infusion at night may be
responsible for the increased amplitude of milk yield. Glucose is required for synthesis of lactose,
which is the osmotic driver of milk yield. The increased amplitude of glucose concentration may
increase the amplitude of conversion to lactose, and therefore the amplitude of milk yield.
Plasma NEFA concentration has previously been demonstrated to exhibit a 24 h rhythm
in dairy cows (Giannetto and Piccione, 2009; Niu et al., 2014). In the current experiment, NEFA
concentrated exhibited a daily rhythm in DAY (P = 0.003), but this rhythm was abolished by
133
CON (P = 0.93) and NGT (P = 0.69; Figure 6-6B). The elimination of the rhythms of CON and
NGT corresponded with a 73% and 61% decrease compared to DAY, respectively (P < 0.05).
The fitted daily rhythm peaked at 0313 in the CON group and was phase-advanced 10.5 h by
NGT (P < 0.05), but was not affected by DAY (P > 0.10). Average plasma NEFA concentration
was not affected by treatment (P = 0.19).
The time of feed intake has previously been demonstrated to alter the daily rhythm of
plasma NEFA concentration in dairy cows. Rottman et al. (2014) observed that the amplitude of
NEFA concentration was reduced by feeding 4x/d compared to 1x/d. Moreover, research from
our lab demonstrated that night-restricted feeding for 16 h from 1900 to 1100 delayed the phase
by 10 h and increased the amplitude over 3-fold compared to day-restricted feeding (Chapter 3).
Furthermore, plasma NEFA concentrations are affected by the time of fatty acid absorption. In a
previous experiment from our lab, the amplitude of NEFA was reduced 4–fold, and the phase was
delayed 9 h when the time of FA infusion was limited to 8 h during the day (0900 to 1700 h)
versus continuous infusion (Chapter 5). In contrast, the current experiment demonstrated that
protein infusion at during the day increased the amplitude of plasma NEFA nearly 4 fold.
Moreover, the previous experiment observed a 5 h phase delay due to day-restricted fat in fusion,
whereas the current experiment detected a phase-advance of 9.5 due to protein infusion during the
same time period. These disparate results suggest differential regulation of the daily rhythm of
lipid mobilization by fat and protein.
The daily rhythms of NEFA mobilization may be regulated by the molecular clock of
adipocytes. In model organisms, lipolysis follows a daily rhythm, and this rhythm is modulated
by the adipocyte clock (Shostak et al., 2013). However, to our knowledge, there has been no
research examining the role of amino acids on the circadian clock of adipose tissue. Future
research using explant or cell culture to examine the direct effects of protein on the adipocyte
clock should be conducted. Similar to previous research from our lab testing the time of fatty acid
134
infusion (Chapter 3), the time of protein infusion uncouples the daily rhythm of glucose
concentration from the daily rhythm of NEFA concentration. This suggests that protein infusion
during the night may desynchronize the peripheral circadian clocks governing glucose and fatty
acid metabolism.
Similar to previous reports (Giannetto and Piccione, 2009; Niu et al., 2014), a daily
rhythm of plasma urea nitrogen concentration was present in CON (P < 0.0001) and NGT (P <
0.0001), but the rhythm was ablated by DAY (P = 0.22; Figure 6-6C). The amplitude of the daily
rhythm was decreased 30% by DAY and increased 66% by NGT. Furthermore, relative to CON
the daily rhythm was phase-delayed 6.1 h and phase-advanced 1.4 h by DAY and NGT,
respectively. Treatment did not affect the concentration of PUN (P = 0.81). Unlike protein
infusion, previous unpublished research from our lab demonstrated that the time of FA infusion
alters PUN concentration, with 8 h of night-infusion causing a 20% increase compared to
continuous infusion (Chapter 5). In the same experiment, however, the phase of the rhythm of
PUN was only slightly (less than 1 h) shifted by treatment. Plasma urea nitrogen concentration is
an indicator of nitrogen utilization efficiency, with higher PUN values indicating poorer
efficiency (Kohn et al., 2005). Results of the current experiment suggest that the daily rhythm of
efficiency of nitrogen use is dependent on the time of protein absorption. Nitrogen homeostasis
has previously been shown to display endogenous circadian rhythmicity in humans (Minami et
al., 2009). Research conducted in model organisms revealed an integral role for Krüppel-like
factor 15 (KLF15), a zinc-finger DNA-binding protein, in circadian regulation of nitrogen
metabolism (Jeyaraj et al., 2012). Branched-chain amino acids suppress KLF15 expression, and
may be a potential avenue for entrainment of circadian rhythms of nitrogen utilization (Liu et al.,
2017).
135
CONCLUSIONS
The time of sodium caseinate absorption affected the daily rhythms of milk synthesis.
This effect appears to be primarily through affecting the robustness of the rhythm without
entraining the phase. In particular, restricting infusion to the night increased the amplitude of the
milk yield and decreased the amplitude of protein concentration, while day-infusion damped the
rhythms of milk yield and fat concentration. The changes in amplitude of milk yield may be
driven by changes in the daily rhythms of glucose concentration, because, similar to milk yield,
the amplitude of plasma glucose was decreased by day-infusion and increased by night-infusion.
The time of protein infusion also appears to affect the daily pattern of lipogenesis, with the peak
of the NEFA rhythm corresponding to the time of protein infusion, and day-infusion dramatically
increasing the amplitude. The nadir of PUN was shifted to the time when protein was infused,
suggesting that protein may entrain the daily rhythms of efficiency of nitrogen utilization. The
changes in the daily rhythms of milk synthesis and plasma metabolites may be mediated through
alterations in the circadian clock of peripheral tissues. Future research examining the role of
amino acids on the molecular clocks of the mammary gland, liver, and adipose should be
conducted.
136
FIGURES
Figure 6-1. Schedule of feeding, lighting and sampling during the experiment.
Treatments were 500g/d of sodium caseinate abomasally infused continuously for 22 h/d (CON); for 8 h/d
during the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500].
137
Figure 6-2. Effect of time of protein infusion on dry matter intake.
Treatments were 500g/d of sodium caseinate abomasally infused continuously for 22 h/d (CON); for 8 h/d
during the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Data are presented as LSM with SEM bars.
138
Figure 6-3. Effect of time of protein infusion on daily milk, fat and protein yield and fat and
protein concentration.
Treatments were 500g/d of sodium caseinate abomasally infused continuously for 22 h/d (CON); for 8 h/d
during the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Data are presented as LSM with SEM bars. Means with differing superscripts are different at P <
0.05.
139
Figure 6-4. The effects of time of protein infusion on daily rhythms of milk synthesis.
Treatments were 500g/d of sodium caseinate abomasally infused continuously for 22 h/d (CON); for 8 h/d
during the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Panels show the effect of time of fatty acid infusion on milk (A) yield (kg), (B) fat yield (g), (C)
protein yield (g), (D) fat concentration (%), and (E) protein concentration (%). Data are presented as
relative expression LSM and SEM bars and cosine fit are shown. Cosinor analysis is shown below each
panel and include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak
of the rhythm), and the P-value of the zero-amplitude test. The black and white bars above the x-axis
display the light: dark cycle.
140
Figure 6-5. Effect of time of protein infusion on daily yields and daily rhythms of milk FA by
source.
Treatments were 500g/d of sodium caseinate abomasally infused continuously for 22 h/d (CON); for 8 h/d
during the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Data are presented as LSM with SEM bars Panels show the effect of protein infusion on (A) daily
yields of de novo (Σ < 16C), mixed (Σ 16C) and preformed (Σ >16C) sources of fatty acids (g/d). (B) daily
rhythms of de novo synthesized (Σ < 16C) FA yield (g/d) (c) daily rhythms of mixed source (Σ 16C) FA
yield (g/d), and (C) daily rhythms of preformed (Σ >16C) FA yield (g/d). Data are presented as relative
expression LSM and SEM bars and cosine fit are shown. Cosinor analysis is shown below each panel and
include the amplitude (Amp; difference between peak and mean), acrophase (Acro; time at peak of the
rhythm), and the P-value of the zero-amplitude test. The black and white bars above the x-axis display the
light: dark cycle.
141
Figure 6-6. The effects of time of protein infusion on daily rhythms of plasma metabolites.
Treatments were 500g/d of sodium caseinate abomasally infused continuously for 22 h/d (CON); for 8 h/d
during the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Panels show the effects of time of fatty acid infusion on daily rhythms of plasma (A) glucose
concentration (mg/dL), (B) nonesterified fatty acid concentration (NEFA; μEq/L), and (C) urea nitrogen
concentration (mg/dL). Data are presented as relative expression LSM and SEM bars and cosine fit are
shown. Cosinor analysis is shown below each panel and include the amplitude (Amp; difference between
peak and mean), acrophase (Acro; time at peak of the rhythm), and the P-value of the zero-amplitude test.
The black and white bars above the x-axis display the light: dark cycle.
142
SUPPLEMENTAL FIGURES
Supplemental Figure S 6-1. Effect of time of protein infusion on lactose yield.
Treatments were 500g/d of sodium caseinate abomasally infused continuously for 22 h/d (CON); for 8 h/d
during the day [(DAY); infused from 0900 to 1700] or for 8 h/d during the night [(NGT); feed from 2100 to
0500]. Data are presented as LSM with SEM bars. Treatments with different superscripts tend to be
different at P < 0.10.
143
SUPPLEMENTAL TABLES
Supplemental Table S6-1. Diet and nutrient composition of the experimental diet.
1Contained (% of DM): 42.3% NDF, 21.7% ADF 2Contained (% of DM): 47.2% NDF, 33.5% ADF 3Contained (% of DM): 69.4% NDF, 35.8% ADF 4AminoPlus, Ag Processing Inc., Omaha, NE. 5Contained (%, as-fed basis): 37.0% Calcium carbonate; 29.9% dried corn distillers grains; 24.5% salt;
4.2% magnesium oxide (54% Mg); 2.4% organic phosphorus (15% P); 0.5% zinc sulfate; 0.2% mineral oil.
Composition (DM-basis): 7.2% CP; 7.1% NDF; 3.7% ADF; 15.0% Ca; 0.75% P; 0.33% K; 2.6% Mg; 0.5%
S; 9.8% Na; 23.0 mg/kg Co; 652 mg/kg Cu; 783 mg/kg Fe; 54.0 mg/kg I; 1190 mg/kg Mn; 12.8 mg/kg Se;
1718 mg/kg Zn; 88,700 IU/kg vitamin A (retinyl acetate); 28,400 vitamin D (activated 7-
dehydrocholesterol); 850 IU/kg vitamin E (dl-α tocopheryl acetate). 5Fed as coated urea (Optigen, Alltech Inc., Lexington, KY; 259% CP, DM basis).
Item Composition
Ingredients, % of DM
Corn silage1 45.5
Alfalfa haylage2 18.6
Canola meal 14.3
Ground corn 11.2
Grass hay/straw3 3.4
Roasted soybeans 2.6
Vitamin/mineral mix4 1.7
Controlled-release N5 0.35
Nutrients, % of DM
NDF 38.2
ADF 21.0
CP
Ash 4.7
144
Supplemental Table S6-2. Amino acid profile of sodium caseinate.
Amino Acid g/100 g AA
Glutamic Acid 21.1
Proline 10.1
Leucine 9.04
Lysine 7.59
Aspartic Acid 6.80
Valine 6.52
Tyrosine 5.32
Isoleucine 5.22
Phenylalanine 4.99
Serine 4.37
Threonine 3.85
Arginine 3.58
Alanine 2.91
Histidine 2.90
Methionine 2.71
Glycine 1.85
Tryptophan 1.38
Cysteine 0.36
145
Chapter 7
Annual rhythms of milk and milk fat and protein production in dairy cattle
in the United States
Isaac J. Salfer, Chad D. Dechow, and Kevin J. Harvatine
This chapter was published as a manuscript in the Journal of Dairy Science:
Salfer, I.J., C.D. Dechow, and K.J. Harvatine. 2019. Annual rhythms of milk and milk fat and
protein production in dairy cattle in the United States. J. Dairy Sci. 102:742–753.
doi:10.3168/jds.2018-15040.
146
ABSTRACT
An annual pattern of milk composition has been well appreciated in dairy cattle, with
highest milk fat and protein concentration observed during the winter and lowest occurring in the
summer. However, rhythms of milk yield and composition have not been well quantified.
Cosinor rhythmometry is commonly used to model repeating daily and annual rhythms and
allows determination of the amplitude (peak to mean), acrophase (time at peak), and period (time
between peaks) of the rhythm. The objective of this study was to use cosinor rhythmometry to
characterize the annual rhythms of milk yield and milk fat and protein concentration and yield
using both national milk market and cow-level data. First, ten years of monthly average milk
butterfat and protein concentration for each Federal Milk Marketing Order (FMMO) were
obtained from the USDA Agricultural Marketing Service database. Fat and protein concentration
fit a cosine function with a 12-month period in all milk markets. There was an interaction
between milk marketing order and milk fat and protein concentration (P < 0.01). The acrophase
(time at peak) of the fat concentration rhythm ranged from December 4th to January 19th in all
regions, while the rhythm of protein concentration peaked between December 27th and January
6th. The amplitude (peak to mean) of the annual rhythm ranged from 0.07 to 0.14 percentage
points for milk fat and from 0.08 to 0.12 percentage points for milk protein. The amplitude of the
milk fat rhythm generally was lower in southern markets and higher in northern markets.
Secondly, the annual rhythm of milk yield and milk fat and protein yield and concentration were
analyzed in monthly test day data from 1684 cows from eleven tie-stall herds in Pennsylvania. Fat
and protein concentration fit an annual rhythm in all herds, while milk and milk fat and protein
yield only fit rhythms in 8 of the eleven herds. On average, milk yield peaked in April, fat and
protein yield peaked in February, fat concentration peaked in January, and protein concentration
peaked in December. Amplitudes of milk, fat, and protein yield averaged 0.82 kg, 55.3 g, and
147
30.4 g, respectively. Milk fat and protein concentration had average amplitudes of 0.12 and 0.07,
respectively, similar to the results of the milk market data. Generally, milk yield and milk
components fit annual rhythm regardless of parity or DGAT1 K232A polymorphism, with only
cows of the low frequency AA genotype (5.2% of total cows) failing to fit rhythm of milk yield.
In conclusion, the yearly rhythms of milk yield and fat and protein concentration and yield
consistently occur regardless of region, herd, parity, or DGAT1 genotype, and supports
generation by a conserved endogenous annual rhythm.
Keywords: Annual rhythm, seasonal rhythm, milk synthesis, milk fat
148
INTRODUCTION
Milk production and milk component concentrations follow a seasonal pattern across the
year that is well-recognized by dairy producers and nutritionists. Milk yield and milk fat and
protein concentrations are typically greatest in the winter and reach a nadir in the summer.
Seasonal variation has long been quantified and accounted for by animal breeders, but has not
been well integrated into dairy management (Wood, 1970, 1976). The causes of these yearly
changes are not fully understood and are often attributed to environmental factors such as heat
stress or changes in forage quality. However, these yearly patterns may be the result of an
endogenous annual rhythm controlling milk synthesis.
In nature, annual rhythms occur as a mechanism for organisms to predict seasonal
environmental changes before they occur to allow proactive adaptations. Many animal species
exhibit yearly cycles of reproductive activity, hibernation, migration, hair growth, and feeding
behavior, allowing them to better prepare for changes in weather conditions and food supply
(Lincoln et al., 2006; Schwartz and Andrews, 2013). A classic example of annual rhythmicity in
production animals is seasonal breeding observed in ewes, which exhibit estrus only from early
autumn to late January, restricting lambing to the spring (Robinson, 1951; Shelton and Morrow,
1965). These annual rhythms of estrus are the result of interactions between photoperiod and an
endogenous physiological timekeeping mechanism (Malpaux et al., 1989). Feed intake of sheep
may also be regulated via annual rhythms, as the secretion of leptin, ghrelin, and orexin is
impacted by photoperiod length (Kirsz et al., 2012). In dairy cattle, the secretion of prolactin
(Chew et al., 1979) and serotonin (Philo and Reiter, 1980) follow melatonin-controlled annual
rhythms, with prolactin levels peaking in summer and serotonin peaking in winter. Furthermore,
Piccione et al. (2012) determined that circulating concentrations of β-hydroxybutyrate, bilirubin,
creatinine, and triglycerides followed annual rhythms in Italian Brown dairy cattle.
149
Biological rhythms, including annual rhythms, can be analyzed using cosinor
rhythmometry. This technique fits a linear form of the cosine function with a 12-month period to
multiple years of data to determine if it fits an annual rhythm. The degree to which the data
follow an annual rhythm is assessed using a zero-amplitude test, which determines if the linear
form of the cosine function models that data significantly better than the linear effect of time
(Went, 2006). Cosinor rhythmometry also determines the amplitude, difference from mean to
peak, and acrophase, or time at peak, of a biological rhythm (Bourdon et al., 1995). Quantifying
the annual rhythms of milk and milk component production can improve management decisions
and may provide insight into physiological mechanism governing annual rhythms. The objective
of this study was to quantify the annual rhythms of milk and milk component production in dairy
cattle in the USA and examine cow specific factors affecting these rhythms.
MATERIALS & METHODS
USDA Milk Market Data
Milk composition data from January 2000 through October 2015 were obtained from the
Agricultural Marketing Service (AMS) agency of the United States Department of Agriculture
(USDA). Monthly milk butterfat and protein concentration from 2000 through 2015 were
downloaded for each US federal milk marketing order (FMMO). Orders included: Northeast
(Order 1), Appalachian (5), Florida (6), Southeast (7), Upper Midwest (30), Central (32), Mideast
(33), Pacific Northwest (124), Southwest (126), Arizona-Las Vegas (131), and Western (135;
Figure 7-1). Butterfat data were available for all 11 orders; however, milk protein percentages
were not available for the Appalachian, Florida, Southeast, or Arizona-Las Vegas regions. Only
four years of data were available from the Western region because the order was terminated in
2004 (Tosi, 2004).
150
All statistical analysis was performed using the PROC MIXED procedure of SAS 9.4
(SAS Institute, Cary, NC). Monthly butterfat and protein concentration were fit to the linear form
of a cosine function with a 12-month period according to Bourdon et al. (1995) using random
regression (Seltman, 1997) as described by Niu et al. (2014). The model tested the random effect
of year and the fixed effects of marketing order, the linear form of a cosine function with a period
of 12-months, and the interaction of order and the linear form of cosine functions. Denominator
degrees of freedom were estimated using the Kenward-Roger method and the AR(1) covariance
structure was used. The fit of the 12-month rhythms was determined using a zero amplitude test,
which employs an F-Test to compare the full model containing the linear form of the cosine
function to a reduced linear model testing the linear effect of month (Went, 2006). The acrophase
and amplitude of each marketing order were calculated, with subsequent significance tests, using
a spreadsheet program developed in Microsoft Excel® (Bourdon et al., 1995).
Cow-Level Data
Data from a previously published experiment by Vallimont et al. (2010) were used to
examine the effect of seasonal rhythms on milk production at the individual cow level. Briefly,
test day milk yield and butterfat and protein concentration were collected from 1684 individual
cows in eleven Pennsylvania tie-stall dairy herds during the years 2002 to 2011. Of these cows,
731 had been tested for the diacylglycerol O-acyltransferase 1 (DGAT1) K232A polymorphism
using a panel of 121 candidate gene markers (Dekleva et al., 2012; Igenity, Neogen Corporation,
Lincoln, NE). Butterfat and protein yield were calculated as the product of milk yield and
butterfat and protein concentration. Test days when cows were less than 40 days in milk (DIM)
were removed from the analysis as high milk fat is expected during early lactation.
Milk, fat, and protein yield and milk fat and protein concentration were fit to the linear
form of a cosine function with a 12-month period using the MIXED procedure of SAS 9.4. Fixed
151
effects included herd, DGAT1 genotype, lactation group (1, 2, 3+ lactations), and days in milk,
along with the random effects of cow and test year. The levels of each main effect (herd, DGAT1
genotype, and lactation group), along with the average of all herds, were fit to a 12-month cosine
function and rhythm fit, amplitude and acrophase were determined using the methodology
described above for the milk market data.
RESULTS
USDA Milk Market Data
A cosine function with a 12-month rhythm was used to test for an annual rhythm in milk
composition using milk fat and protein concentration data from USDA federal milk marketing
orders. There was an interaction of milk marketing order and the cosine function on fat
concentration so the fit, amplitude, and acrophase of the 12-month cosine function were
determined for each milk market order (Table 7-1). Fat concentration in all marketing orders fit
the 12-month rhythm (P < 0.0001). Mean fat percent was lowest in the Arizona-Las Vegas region
(3.56%) and greatest in the Pacific Northwest (3.73%). The amplitude of the rhythm of fat
concentration ranged from 0.07 to 0.14 percentage units between orders. The Southwest, Central
(Figure 7-2C), Southeast, Mideast, Appalachian, and Western regions had 12-month rhythms
with the greatest amplitudes (0.13 to 0.14 percentage units) and did not differ between each other
(P > 0.10). The Northeast, Upper Midwest (Figure 7-2B), and Pacific Northwest marketing
orders all exhibited an amplitude of 0.11 percentage units, which was lower than the Southwest,
Central, Southeast, Mideast, Appalachian, and Western regions (P < 0.05). The amplitudes of fat
concentration rhythms were lowest in Arizona-Las Vegas (0.09 percentage units) and Florida
(0.07 percentage units), with Florida having a lower amplitude than Arizona-Las Vegas (P <
0.05; Figure 7-2A).
152
The acrophase, or time at peak, of butterfat concentration ranged from December 4th to
January 18th across milk marketing orders. In six of eleven marketing orders, including Arizona-
Las Vegas, Northeast, Upper Midwest, Mideast, Southwest, and Southeast, the 12-month rhythm
of fat concentration peaked within 4 days of January 1st (P > 0.10). No difference was observed
between the acrophases of the Western, Central, and Appalachian regions, and all peaked
between January 17th and January 19th (P > 0.10). Florida’s annual rhythm peaked on December
4th, markedly earlier than the other regions (P < 0.05).
An interaction between milk marketing order and the cosine function was observed for
protein concentration, so the fit, amplitude, and acrophase of the 12-month rhythm were
compared between marketing orders. Protein concentration fit the 12-month function in all seven
regions with protein data available (P < 0.001; Table 7-1). Mean protein concentration was
greatest in the Pacific Northwest, Southwest, and Western Regions, followed by Central and
Mideast (P < 0.05). The Northeast and Upper Midwest had the lowest mean protein
concentrations. The Northeast and Upper Midwest averaged 3.04% protein, the lowest of all
regions (P < 0.05). The amplitudes of annual rhythms in all regions were between 0.08 and 0.10
percentage units. The greatest amplitudes (0.10 percentage units) were observed in the Central
and Southwest (Figure 7-3B) regions. The Upper Midwest and Mideast had intermediate
amplitudes (0.09 percentage units), while the Northeast (Figure 7-3A), Pacific Northwest and
Western regions displayed amplitudes of 0.08 percentage units (P < 0.05). The acrophases of
annual protein concentration rhythms occurred between December 27th and January 6th in all
regions (P > 0.10). These results are consistent with those observed by Bailey et al. (2005), who
reported that three-year average milk production peaks were greatest in December and January in
the Mideast milk market.
153
Cow-Level Data
Data from 11 tie-stall herds in Pennsylvania were analyzed to characterize seasonal
patterns of yield of milk and milk fat and protein, determine how cow-level factors including
parity or DGAT1 genotype influence annual production rhythms, and examine variation in annual
rhythms between individual herds. Milk yield of the 11 herds fit an annual rhythm with an
amplitude of 1.26 kg and an acrophase of March 13st. A yearly rhythm of milk yield was
observed in ten of the eleven herds (P < 0.05; Supplemental Figure S7-3A&B). Eight of the ten
herds that exhibited a yearly rhythm of milk yield peaked between March 13th and May 5th, while
the remaining two herds peaked in mid-summer. These results agree with previous data
suggesting that the average amount of milk shipped from farms is greatest in March, April, and
May (Bailey et al., 2005). The amplitudes of the milk yield rhythms ranged from 0.84 to 1.64
kg/d, suggesting that annual rhythms may be responsible for as much as 3.3 kg/d difference in
milk production across the year.
Overall, the average fat yield of all herds fit a 12-month rhythm with a mean of 1313 g/d,
an amplitude of 55.3 g/d, and an acrophase occurring on February 23rd (P < 0.0001;
Supplemental Figure S7-3C&D). All but one of the eleven herds exhibited a yearly rhythm,
while one exhibited a rhythm with a drastically lower amplitude than the other nine herds (16.5
g/d; P < 0.0001). The nine remaining herds had amplitudes ranging from 97.2 to 47.3 g/d. The
non-rhythmic and low amplitude herds peaked on July 5th and 6th, respectively, in contrast to the
other 9 herds, which peaked between January 22nd and April 3rd. Average protein yield also fit a
yearly rhythm, with a mean of 1081 g, an amplitude of 30.4 g, and an acrophase occurring on
February 19th (Supplemental Figure S7-3E&F). A yearly rhythm was observed in 10 of the 11
herds. The amplitudes of annual rhythms of protein yield were highly variable between herds and
ranged from 14.3 to 47.0 g g/d. There was, however, low variability in the acrophases, with 8
herds peaking between January 31st and March 1st and one herd peaking on April 24th.
154
Fat and protein concentration fit 12-month rhythms in all herds as expected and exhibited
much more consistent yearly patterns than milk, fat, and protein yield (P < 0.0001; Supplemental
Figure S7-4A&B). The average rhythm of fat concentration for the 11 herds peaked on January
25th with an amplitude of 0.12 percentage units. Of the 11 herds, 9 peaked within 30 days of
January 25th, with the remaining two peaking on March 27th and April 7th. The amplitude of the
yearly rhythm averaged 0.12 percentage units, equivalent to a 0.24 unit change in fat percent
throughout the year and was highly variable between herds (0.07 to 0.19 units). Milk protein
concentration similarly fit an annual rhythm in all herds (P < 0.001), with an average amplitude
of 0.07 units and the average acrophase occurring on December 30th (Supplemental Figure
S7-4C&D). All herds peaked between December 8th and January 23rd with amplitudes ranging
from 0.04 to 0.11 percentage units, suggesting that annual rhythms of protein percent are highly
consistent between herds.
The effect of parity on fit of the cosine function and the mean, amplitude and acrophase
of annual production rhythms was evaluated (Figure 7-4). Groups included cows in first, second,
or third and greater lactations. Milk yield increased with increasing parity (P < 0.05) and fit an
annual rhythm in all three parity groups (P < 0.0001). The amplitude of the 12-month milk yield
rhythm was greater for cows in their third and greater lactations than 1st and 2nd lactation (P <
0.05), but was not different between 1st and 2nd lactation cows (P > 0.10). Furthermore, acrophase
did not differ between 1st and 2nd lactation, but occurred over a month earlier in third and greater
lactation cows (P < 0.05).
A yearly rhythm of fat yield was also observed in all lactation groups (P < 0.0001). Mean
fat yield increased with increasing parity, while cows in third and greater lactation had 80% and
55% greater amplitudes than 1st and 2nd lactation cows, respectively (P < 0.05). The acrophase of
fat yield did not differ between cows in second or third and greater lactations (P > 0.10), but
occurred 23 and 35 days later in 1st lactation cows compared with 2nd and 3rd lactation,
155
respectively (P < 0.05). Protein yield fit a 12-month rhythm in all lactation groups (P < 0.0001)
and mean protein yield increased from 936 g in 1st lactation to 1130 g in 2nd lactation and 1260 g
in 3rd and greater lactations (P < 0.05). Amplitude was greatest in third and greater lactation (73.1
g; P < 0.05), but did not differ between cows in first and second lactation (P > 0.10). The
acrophase of the yearly rhythm was similar between second and third and greater lactations, and
occurred on March 11th and March 1st, respectively (P > 0.10). Alternatively, the acrophase of the
annual rhythm of protein yield occurred later in first lactation cows (April 10th; P < 0.05).
Milk fat and protein concentration fit a 12-month rhythm in all three parity groups (P <
0.0001). Mean milk fat concentration decreased with increasing parity (P < 0.05), but no
difference in amplitude was observed between lactation groups (P > 0.10). Similarly, parity
exerted no effect on the acrophase of fat concentration rhythms, with all peaking between January
14th and January 25th (P > 0.10). Milk protein concentration fit a yearly rhythm in all three
lactation groups (P < 0.05). Mean protein concentration did not differ between 1st and 2nd
lactation (P > 0.10), but was 8% and 6% lower in 3rd and greater lactation cows than 1st and 2nd
lactation, respectively (P < 0.05). No difference occurred in the acrophase or amplitude of the
rhythms of protein concentration between lactation groups, with amplitudes ranging from 0.08 to
0.09 percentage units and all acrophases occurring between December 12th and December 26th (P
> 0.10).
DGAT1 genotype influenced average milk fat concentration, with the cows of the AA
genotype averaging 4.19% milk fat, KA genotype averaging 3.94% fat, and KK genotype
averaging 3.58% fat (P < 0.05; Figure 7-5). Fat yield did not differ between the AA (1296 g) and
KA genotype (1256 g; P > 0.10), but was lower the KK genotype (1178 g; P < 0.05). These
results are consistent with previous reports showing that the K232A polymorphism is associated
with milk fat percent and yield (Schennink et al., 2007). An annual rhythm of fat concentration
was observed regardless of DGAT1 genotype (P < 0.0001) and DGAT1 genotype did not affect
156
the amplitude or acrophase of the rhythm (P > 0.10). Similarly, fat yield fit a rhythm in all three
genotypes (P < 0.0001) with no difference in mean fat yield. Amplitude was similar between all
genotypes and ranged from 54.6 to 70.0 g (P > 0.10). Likewise, acrophase was unaffected by
DGAT1 genotype and all rhythms peaked between February 15th and March 1st. Average milk
yield did not differ between the AA and KA genotypes (P > 0.10), but was 2.4 and 1.2 kg greater
in cows with the KK genotype than AA and KA (P < 0.05). Milk yield fit a 12-month rhythm in
cows with the KA and KK genotype, but not AA. The failure of AA to fit a rhythm may have
been because of its low frequency within the population limited observations within the dataset
(only 5% of cows are AA). Between the KA and KK groups, no difference was observed for
either amplitude or acrophase.
DISCUSSION
Regional milk market and cow-level data both confirmed the presence of 12-month
rhythms of fat and protein concentration and cow level data demonstrated a rhythm in milk and
milk fat and protein yield, which is biologically and economically important. The consistency of
these rhythms between years, regions and herds suggests that the annual patterns of production
seen on dairy farms may be the result of a biological rhythm, rather than the outcome of
environmental factors. Heat stress is the most notable environmental factor correlated to seasonal
production declines, and begins to occur when the temperature-humidity index rises above 70
(Bohmanova et al., 2007). Cows in the United States are most likely to experience during July
and August. If the summer decrease in production were strictly due to the acute effects of heat
stress, yearly production would be expected to remain relatively stable, with a sharp decline in the
summer, followed by recovery in the fall when the stress is alleviated. However, the results of
this study suggest a consistent and predictable cosine decrease from the peak of production to the
157
nadir. For example, fat and protein concentration, which peak in January in the majority of
regions, begins to decrease in February and March, much earlier than would be expected if the
effect is simply a consequence of heat stress. Furthermore, milk yield was shown to reach a
minimum between September and November, rather than mid-summer when heat stress is
expected to be the greatest. Lastly, the rhythm is very consistent between years and the degree
heat stress is expected to differ between years, especially in more northern regions
(Supplemental Figure S7-5).
The annual rhythms of fat and protein concentration consistently peaked in mid-winter in
both the milk market and cow level datasets. These results are consistent with previous reports
showing that maximal fat and protein concentration occur during December and January (Bailey
et al., 2005). Research in grazing cattle has alternately suggested that the occurrence of milk fat
depression is greatest in May, however this may be influenced by changes in the fermentability of
grass throughout the year (Carty et al., 2017). Dunshea et al. (2008) observed that milk
concentrations of trans-10 C18:1 isomers, associated with biohydrogenation-induced milk fat
depression, were greatest in August and September which is consistent with the trough of fat
yield production observed in the current experiment.
The yearly change in photoperiod with lengthening and shortening days is a strong cue
for entraining annual rhythms and greater amplitude rhythms are often associated with greater
variation in the light-dark cycle across the year (Hut et al., 2013). In the Federal Milk Market
Order data the lowest amplitude annual rhythm occurred in Florida, followed by Arizona –Las
Vegas. Florida has the lowest latitude (Tallahassee, FL: 30.4° N) of all the regions investigated,
and consequently has the smallest change in photoperiod across the year. Similarly, Arizona-Las
Vegas has a lower latitude (Las Vegas, NV: 36.2° N) than the majority of the other marketing
orders. However, the amplitude of the other nine orders did not follow a pattern based on
photoperiod length. For example, the Southeast and Southwest regions had the greatest amplitude
158
rhythms, along with the Appalachian, Mideast and Western regions. The northernmost regions
including the Northeast, Upper Midwest, and Pacific Northeast had intermediate amplitudes. One
potential cause for the inconsistency between regions is the differences in housing systems among
various U.S. regions. In Florida and Arizona a large proportion of cows are housed in shaded dry
lots with access to natural lighting, while in northern states cows are more likely to be housed in
enclosed tie-stall or free-stall barns under the influence of artificial lighting. However, this does
not account for the high amplitudes observed in the Southeast and Southwest, where cows are
also commonly housed in open dry lots. Data from dairy herds in the southern hemisphere would
provide additional insight into the role of change in yearly photoperiod on annual rhythms of
production. If the effects is a direct effect of photoperiod, rhythms in the southern hemisphere
would be expected to be inverted compared to those in the northern hemisphere. Auldist et al.
(1998) demonstrated that in New Zealand, milk fat and protein concentration are greatest during
their winter (June-August), while milk yield peaks in their spring and early summer (September-
December). Results from data in the southern hemisphere are difficult to interpret, however,
because grazing systems are common in those countries, and effects of the annual rhythm may be
confounded by differences in the nutrient composition of pasture.
Herd level data suggested that, unlike fat and protein concentration, which peak between
December and February, fat and protein yield reach an acrophase between February and April
and milk yield reaches an acrophase between March and May in most herds. Between herds,
variability was present in the annual rhythms of milk, fat, and protein yield. Two herds in
particular stood out for having no annual rhythms, a low amplitude of fat and protein yield, and
acrophases of fat, protein, and milk yield that occurred during the summer, about four to six
months later than the other herds. One herd also notably failed to demonstrate a yearly rhythm of
milk yield. These herds possessed no identifiable feeding or management characteristics to
differentiate them from the other eight (Vallimont et al., 2010). Results suggest that management
159
factors such as changes in diet or feed intake across the year may mask an endogenous annual
rhythm driving yearly changes in milk production.
We examined the genotype at the diacylglycerol O-acyltransferase 1 (DGAT1) locus to
determine its influence on the yearly rhythms of fat concentration, fat yield, and milk yield. The
K2332A polymorphism in the DGAT1 gene is responsible for up to 50% of the genetic variation
in milk fat production (Schennink et al., 2007). The AA genotype is associated with high milk fat
concentration, the KK genotype is associated with low milk fat, and the KA heterozygote results
in an intermediate phenotype (Winter et al., 2002). The current experiment confirmed these
previous results, with AA cows producing milk with a 0.25% greater milk fat concentration than
KA cows, which was 0.36% greater than KK. Annual rhythms of nearly all production variables
persisted regardless of DGAT1 genotype, with a lone exception being milk yield in cows with the
AA genotype. The reason for the lack of an annual rhythm in this group is unclear, but may be a
consequence of a lower prevalence of this genotype. Only 1196 test day observations from 40
unique cows carried the AA genotype, which was equivalent to 5% of the total cows in the study.
This low number of cows, combined with management variability between herds may have
contributed to the inability to detect an annual rhythm. However, despite the low number of
observations, annual rhythms were detected for fat concentration and yield within the AA group.
The DGAT1 genotype of the animals did not affect the shape of the annual rhythms of fat percent
or fat yield, with similar amplitudes and phases in all groups. These results are similar to those
observed by Duchemin et al. (2013), who observed that milk fat concentration was not affected
by the interaction of DGAT1 genotype and season, although they did observe an interaction
between DGAT1 and summer vs winter milk for unsaturated fatty acid concentration.
Lactation number had little effect on the shape of annual production rhythms. Increased
amplitudes of milk, fat, and protein yield occurred in cows in their third or greater lactation.
These results suggest that as cows age, greater oscillation in the yearly rhythms of production
160
occurs. Furthermore, greater lactation number was associated with slightly earlier peaks in yearly
rhythms of milk protein and fat yield. The cause of changes is unclear, but may be related to
overall increase in production observed as lactation number increases.
Previous experiments have supported that dairy cattle are influenced by annual rhythms.
Although cows are not generally considered to be seasonal breeders, Hansen (1985) observed
modest effects of season and photoperiod length on reproduction in cattle. Furthermore, Kendall
and Webster (2009) observed greater daily fluctuations in body temperature in the summer
compared to winter. Several circulating hormones and metabolites also follow annual rhythms
across the year. Petitclerc et al. (1983) observed that circulating prolactin concentrations are over
six times greater in summer than winter, and that this effect occurs even when melatonin
signaling is blocked through blinding or pinealectomy. Plasma concentrations of bilirubin,
creatinine, triglycerides, and β-hydroxybutyrate (BHBA) follow 12-month rhythms, with BHBA
peaking on April 1st, bilirubin peaking on July 14th, creatinine peaking on June 12th, and
triglycerides peaking June 16th (Piccione et al., 2012). The annual rhythms observed in these
metabolites may be regulated by the same mechanism as the annual rhythms of production. The
annual patterns of production may also be a consequence of seasonal changes in feed intake.
Cattle typically increase feed intake in the winter and Ueda et al. (2016) reported dry matter and
fiber in the rumen of dairy cows is greater in autumn than spring, with a reciprocal relationship
for ruminal VFA concentration. Endocrine factors regulating intake have been reported to follow
an annual rhythm in other animals. Circulating leptin concentrations vary across the year in sheep
and Siberian hamsters (Rousseau et al., 2003; Henry et al., 2010) and leptin, orexin and ghrelin
appear to be modulated by photoperiod in sheep (Kirsz et al., 2012).
A seasonal rhythm in milk yield and composition may be adaptive to improve nutrition of
the calf. From a biological perspective, there is a clear advantage to providing nursing offspring
with high-energy milk during the winter months. In other organisms, seasonal rhythms function
161
to maximize reproductive success by scheduling parturition to allow neonates to be born during
periods of high food availability and favorable climatic conditions (Lincoln et al., 2003). As an
important component of mammalian reproduction, it is reasonable to expect lactation to similarly
follow an annual rhythm and provide more milk and higher fat and protein to neonates in the
winter when energetic demands are greater. Seasonal reproduction in sheep and hamsters appears
to be coordinated by interactions photoperiod and circadian clock genes located within the
hypothalamus. Melatonin is released from the pineal gland during the dark phase of the
photoperiod and binds with high affinity to the pars tuberalis (PT) of the hypothalamus (Johnston
et al., 2006). The duration of melatonin release affects the phase of the core clock genes Per and
Cry within the PT, and which acts downstream to influence the pulsatile frequency of GnRH, thus
affecting reproductive cyclicity (Misztal et al., 2002). Although this mechanism is well-
characterized for controlling seasonal reproduction, there is a dearth of research examining the
presence of mechanisms governing annual rhythms of lactation.
The amplitude, acrophase, and period length of annual rhythms can be manipulated
through photoperiod alterations. Gwinner (1981) demonstrated that yearly cycles of testes growth
and molting in migrating common starlings (Sturnus vulgaris) can be shorted by accelerating
changes in photoperiod. Furthermore, accelerated photoperiods can speed up the reproductive
cycles of rainbow trout (Oncorhynchus mykiss; Bon et al., 1997). Photoperiod manipulation has
been extensively studied for its effects on milk production (Peters et al., 1978; Dahl et al., 1997;
Miller et al., 1999). Cows consistently increase milk yield when placed in a photoperiod with
over 12 h of light, with maximal response occurring in a 16L:8D photoperiod (Reksen et al.,
1999). This effect of may be related to hormonal changes caused by long-day lighting. Dahl et al.
(1997) established that increased milk synthesis due to an artificial 18L: 6D photoperiod is
associated with increased concentrations of circulating IGF-1. While the direct effect of IGF-1 on
milk synthesis is unclear, its concentration is increased after exogenous treatment with
162
recombinant bovine somatotropin, suggesting it may increase milk yield (Bauman and Vernon,
1993). Plasma prolactin concentrations are also consistently elevated under long-day lighting
(Tucker et al., 1984; Stanisiewski et al., 1988; Lacasse et al., 2014). Prolactin is an important
driver of seasonal physiological changes in other mammalian species, drives changes in
reproduction and molting (Martinet et al., 1984; Gómez-Brunet et al., 2008). While an association
between prolactin and milk synthesis has been suggested, the effect does not appear to be direct
because no response to exogenous prolactin on milk yield has been observed in cattle (Plaut et al.,
1987; Lacasse et al., 2012).
There appears to be a paradox between the effects of long-day lighting and annual
rhythms of production. In natural conditions milk increases across the winter when the duration
of the light cycle is less than 12 h, whereas photoperiod research suggests that milk yield should
be greatest when more than 12 h of light is present. The effect may be explained by induction of
photorefractoriness to the annual rhythm by constant-long day photoperiods. In other mammalian
species, long-term exposure to a constant photoperiod causes spontaneous reversion, or
photorefractoriness, of a seasonal physiological response to the state expected in the opposite
photoperiod (Lincoln et al., 2005). Suffolk sheep exposed to a fixed short 8L: 16D photoperiod
exhibit summer-type physiology, characterized by a decrease in serum luteinizing hormone
concentrations and anestrous (Robinson and Karsch, 1984). In long-day breeding Syrian hamsters
(Mesocricetus auratus) gonadal degeneration spontaneously occurs after long-term administration
of a fixed 14L: 10D photoperiod. The mechanism for this phenomenon is thought to be through
dissociation of the endogenous annual timer in the PT with melatonin-based signals from the
light-dark cycle (Lincoln et al., 2005). The long-term treatment needed to elicit the effects of long
day photoperiod on milk synthesis provides additional support for this mechanism. In other
species, a fixed photoperiod must be applied for 4 to 12 weeks prior to induction of
photorefractoriness, which is consistent with data suggesting that the effect of long day lighting
163
on milk yield does not manifest until 4 weeks after administration (Dahl et al., 2000). While this
mechanism seems like a promising explanation for the incongruence between the annual rhythm
and photoperiod research, it has not been well-examined in cows and further research should be
performed to investigate this effect.
CONCLUSIONS
Milk yield and milk fat and protein concentration and yield fit annual rhythm at both the
regional and cow level. These rhythms appear to exist independent of environmental effects such
as heat stress and may be the result of endogenous circannual rhythms. Rhythms of fat and
protein concentration peak in the middle of winter, and reach a nadir in the summer, whereas
yields of milk, fat and protein peak in the spring. Region of the U.S. appears to have mild effects
on annual rhythms, with lower latitude regions having lower-amplitude rhythms. Moderate
variability in annual rhythms exists between herds, but DGAT1 genotype and lactation group
exert little effects. Dairy producers and consultants should consider the annual rhythms of
production when making management decisions. An understanding of these rhythms will allow
better benchmarking across the year.
164
FIGURES
Figure 7-1. Map of United States Federal Milk Marketing Orders.
Image provided courtesy of the USDA Agricultural Marketing Service. The Western (135) marketing order
was terminated in 2004.
165
Figure 7-2. Annual rhythms of fat concentration in selected Federal Milk Market orders.
Orders include (A) Florida (Order 6), (B) Upper Midwest (Order 30) and (C) Central (Order 32). Data
presented as least-squares means with error bars representing SEM of each month and the line representing
a fitted cosine function with a 12-month period. Characteristics of fitted rhythm are displayed including:
Mean, 1Amplitude (peak minus mean), 2Acrophase (time at peak), and 3P-value for the zero-amplitude test
corresponding to the fit of the 12-month cosine rhythm. Annual rhythms of fat concentration for the other
milk market orders are displayed in Supplemental Figure 1.
166
Figure 7-3. Annual rhythms of protein concentration in selected Federal Milk Market orders.
Orders include (A) Northeast (Order 1) and (B) Southwest (Order 126). Data presented as least-squares
means with error bars representing SEM of each month and the solid line representing a fitted cosine
function with a 12-month period. Characteristics of fitted rhythm are displayed including: Mean, 1Amplitude (peak minus mean), 2Acrophase (time at peak), and 3P-value for the zero-amplitude test
corresponding to the fit of the 12-month cosine rhythm. Annual rhythms of fat concentration for the other
milk market orders are displayed in Supplemental Figure 2.
167
Figure 7-4. The effect of parity on annual rhythms of milk yield and composition.
Annual rhythm by parity shown for (A) milk yield, (B) fat yield, (C) fat concentration, (D) protein yield,
and (E) protein concentration. Data presented as least-squares means with error bars representing SEM of
each month and the line representing a fitted cosine function with a period of 12-months. Mean, 1Amplitude (peak minus mean), 2Acrophase (time at peak), and 3P-value for the zero-amplitude test
corresponding to the fit of a 12-month cosine rhythm are displayed in tables below each corresponding
graph.
168
Figure 7-5. The effect of DGAT1 K232A polymorphism on annual rhythms of milk and milk fat
concentration and yield.
Effect of DGAT1 polymorphism on annual rhythms of (A) milk yield, (B) fat concentration, and (C) fat
yield are shown. Data presented as least-squares means with error bars representing SEM of each month
and the line representing a fitted cosine function with a 12-month period. Mean, 1Amplitude (peak minus
mean), 2Acrophase (time at peak), and 3P-value for the zero-amplitude test corresponding to the fit of a 12-
month cosine rhythm are displayed in tables below each corresponding graph.
169
TABLES
Table 7-1. The acrophase and amplitude of a cosine function with a 12-month period fit to the regional monthly averages of milk fat and
protein concentration.
1Federal Milk Marketing Order. 2Means that do not share a superscript differ P < 0.05 3Distance from mean to peak of the 12-month rhythm. Values that do not share a superscript differ P < 0.05. 4Date at peak of the 12-month fitted rhythm as month and day of month. Values that do not share a superscript differ P < 0.05. 5P-value for the zero-amplitude test corresponding to the fit of a 12-month cosine rhythm.
Item FMMO1 Region N Mean2 Amplitude3 (%) Acrophase4 (date) P-value5
Fat, % 1 Northeast 189 3.71ab 0.11b Dec 31a < 0.001
5 Appalachian 190 3.66cd 0.13a Jan 17bc < 0.001
6 Florida 189 3.61ef 0.07d Dec 4d < 0.001
7 Southeast 189 3.66d 0.14a Jan 3a < 0.001
30 Upper MW 189 3.72a 0.11b Dec 31a < 0.001
32 Central 189 3.67cd 0.14a Jan 19c < 0.001
33 Mideast 189 3.69bc 0.13a Dec 31a < 0.001
124 Pacific NW 189 3.73ab 0.11b Jan 12b < 0.001
126 Southwest 188 3.64de 0.14a Dec 31a < 0.001
131 AZ-LV 190 3.56f 0.09c Dec 29a < 0.001
135 Western 51 3.62de 0.13a Jan 18bc < 0.001
Protein, % 1 Northeast 190 3.04c 0.08c Dec 31 < 0.001
30 Upper MW 190 3.04c 0.09bc Dec 30 < 0.001
32 Central 189 3.07b 0.10ab Jan 6 < 0.001
33 Mideast 190 3.06bc 0.09ab Dec 30 < 0.001
124 Pacific NW 189 3.10a 0.08c Dec 27 < 0.001
126 Southwest 189 3.09ab 0.10a Dec 30 < 0.001
135 Western 51 3.09ab 0.08abc Jan 2 < 0.001
170
SUPPLEMENTAL FIGURES
Supplemental Figure S7-1. Annual rhythms of fat concentration in Federal Milk Market orders.
Orders include (A) Northeast (Order 1) (B) Appalachian (Order 5), (C) Southeast (Order 7), (D) Mideast
(Order 33), (E) Pacific Northwest (Order 124), (F) Southwest (Order 126), (G) Arizona-Las Vegas (Order
131), and (H) Western (Order 135). Data presented as least-squares means with error bars representing
SEM of each month and the solid line representing a fitted cosine function with a period of 12-months.
Characteristics of fitted rhythm are displayed including: Mean, 1Amplitude, 2Acrophase, and 3P-value for
the zero-amplitude test corresponding to the fit of a 12-month cosine rhythm.
171
Supplemental Figure S7-2. Annual rhythms of protein concentration in Federal Milk Market
orders.
Orders include (A) Upper Midwest (Order 30), (B) Central (Order 32), (C) Mideast (Order 33), (D) Pacific
Northwest (Order 124), and (E) Western (Order 135). Data presented as least-squares means with error bars
representing SEM of each month and the solid line representing a fitted cosine function with a period of 12-
months. Characteristics of fitted rhythm are displayed including: Mean, 1Amplitude, 2Acrophase, and 3P-
value corresponding to the zero-amplitude test of the 12-month rhythm.
172
Supplemental Figure S7-3. Variability in acrophase and amplitude of daily rhythms of milk, fat
and protein yield among 11 herds in Pennsylvania.
Panels show: (A&C) Acrophases (date at peak) of the 12-month (A) fat concentration and (C) protein
concentration rhythms among herds. The average acrophase of the 11 herds are shown at the top of the
figures in grey. Fit of a 12-month rhythm as determined by zero-amplitude test is shown on the right side of
the figures. (B&D). Amplitudes of the 12-month (B) fat concentration and (D) protein concentration
rhythms among herds. The average amplitude of the 11 herds is shown at the right side of the figures in
grey. Data are presented as means with error bars representing the 95% confidence interval.
173
Supplemental Figure S7-4. Variability in acrophase and amplitude of daily rhythms of milk fat
and protein concentration among 11 herds in Pennsylvania.
Panels show: (A & C) Acrophases (date at peak) of the 12-month (A) fat and (C) protein concentration
rhythms among herds. The average acrophase of the 11 herds are shown at the top of the figures in grey.
Fit of a 12-month rhythm as determined by zero-amplitude test is shown on the right side of the figures. (B
& D). Amplitudes of the 12-month (B) fat concentration and (D) protein concentration rhythms among
herds. The average amplitude of the 11 herds is shown at the right side of the figures in grey. Data are
presented as means with error bars representing the 95% confidence interval.
174
Supplemental Figure S7-5. Repeatability of annual rhythms across years.
Panels show monthly average fat and protein concentration in (A) the Northeast region (Order 1) and the
(B) Central region (Order 32) from 2000 to 2015.
175
Chapter 8
Annual rhythms of milk synthesis in dairy herds in four regions of the United
States and their relationships to environmental predictors.
Isaac J. Salfer, Paul A. Bartell, C.D. Dechow, Kevin J. Harvatine
176
ABSTRACT
Annual rhythms of milk fat and protein concentration are well described, but the rhythm
of milk and milk component yield is not well described and is important to dairy management.
Recent analysis of Federal Milk Marketing orders in the USA suggested that the amplitude and
time at peak (acrophase) of the rhythm differ between regions. Our objective was to determine the
annual rhythm of milk production and the effect of U.S. region at the herd level, and examine
potential environmental factors entraining this rhythm. Monthly DHIA records of all available
herds in Pennsylvania, Minnesota, Texas and Florida from the years 2003 to 2016 were obtained
from Dairy Records Managements Systems. Milk yield, fat and protein yield, and fat and protein
concentration were fit to the linear form of the cosine function with a 12-month period using a
linear mixed effects model. Model parameters included the fixed effects of state, cosine
parameters, the interaction of state and cosine parameters, and breed and the random effects of
herd and year. A zero-amplitude test was performed to determine cosine function. The fit of
models containing either the cosine function or environmental temperature were compared using
an F-test. Additionally, raw milk production responses and the predicted response from the cosine
function were correlated with daylength, change in daylength, and maximum temperature. Milk
yield and fat and protein yield and concentration fit a cosine function in all four states, indicating
an annual rhythm (P < 0.001). The amplitude of the rhythm of milk yield varied by state and was
lower in PA (2.8 kg) and MN (2.4 kg) compared to TX (6.9 kg) and FL (8.1 kg; P < 0.05). Fat
and protein yield similarly showed a greater amplitude in the southern versus northern states (P <
0.05). The fat and protein concentrations was opposite, with greater amplitudes occurring in MN
and PA than in TX and FL (P < 0.05). The acrophase of milk yield, fat and protein yield, and
concentration also varied by state, but all peaked between October and March (P < 0.05). The
annual rhythm fit the data better than environmental temperature for all responses in all states (P
177
< 0.001), except for fat and protein concentration in Florida. Both raw milk yield (r =0.19) and
the annual rhythm (r = 0.85) were highly correlated with the change in daylength (P < 0.0001),
while raw milk fat and protein concentration and their annual rhythms were highly correlated
with absolute daylength. Results suggest that region of the U.S. impacts annual production
rhythms, with a greater yearly variation in milk, fat and protein yield occurring in the south.
Moreover, the annual rhythms of milk yield and milk components appear to be controlled by two
separate oscillatory mechanisms with change in daylength controlling milk yield and absolute
daylength controlling milk fat and protein concentration.
Keywords: Annual rhythms, milk synthesis, yearly pattern
178
INTRODUCTION
Annual rhythms commonly occur in nature as a mechanism for animals to anticipate
seasonal changes in their environment before they occur and improve adaptation to environmental
changes. These rhythms drive remarkable biological changes including migration, hibernation,
and seasonal reproduction. The change in photoperiod across the year is a primary driver of
annual rhythms for organisms living in temperate climates. For example, in short day breeders
like sheep, estrus is triggered as photoperiod shortens resulting in mid-spring lambing (Woodfill
et al., 1991). True annual rhythms are endogenously generated and continue even if animals are
placed under constant photoperiods. The most extreme documented example of this occurred the
African stonechat (Saxicola torquatus axillaris), a species of thrush, which maintained annual
rhythms for 15 years when held in a constant 12.25 light: 11.75 dark photoperiod (Gwinner,
1996). Recent evidence suggests that specialized cells within the hypothalamus integrate changes
in photoperiod with a molecular circadian clock, causing time-of-year specific signals that can be
maintained even after the photoperiod becomes fixed (Hazlerigg et al., 2018).
Dairy cattle display seasonal changes in milk production that are well-appreciated by
dairy producers and nutritionists. We have previously characterized annual rhythms of milk and
milk components using USDA milk market data and test day data from 1684 cows in 11
Pennsylvania dairy herds (Salfer et al., 2019). Fat and protein concentration peaked near early
January in all regions within the USDA milk market dataset, although the amplitudes varied with
southern regions having a lower amplitude rhythm than northern regions. Moreover, cow-level
data revealed that milk yield peaked in late March and fat and protein yield peaked in late
February and that neither parity nor diacylglycerol O-acyltransferase 1 (DGAT1) K232A
polymorphism, associated a majority of the genetic variation in milk fat concentration, exert
substantial influence over annual rhythms of production (Salfer et al., 2019)
179
While U.S. milk market and cow-level data have been insightful for characterization of
annual rhythms, a more comprehensive dataset was desired to better quantify these rhythms. The
objective of this study was to model the annual rhythms of milk and milk fat and protein yield
and milk fat and protein concentration using a robust dataset containing a large number of herds
in northern and southern states. The hypothesis was that the annual rhythms of milk production
would vary by latitude, with Florida and Texas having lower amplitude rhythms than
Pennsylvania and Minnesota. Furthermore, the fit of a cosine function was compared to the
effects of environmental temperature, period, and change in photoperiod to determine which
factors correspond to the annual rhythms of milk production.
MATERIALS & METHODS
Data Collection
Monthly DHIA test day records from January 2004 to October 2016 were obtained from
Dairy Records Management Systems (DRMS) for all available herds in Florida (FL), Minnesota
(MN), Pennsylvania (PA), and Texas (TX). These states were selected because they represented
four distinct geographic regions of the U.S. and had a large amount of data available within the
DRMS system. Herd codes were blinded and records included test date, herd size, average DIM,
percent of cows in 1st, 2nd, and 3rd or greater lactation, milk yield, fat concentration, protein
concentration and somatic cell count. For comparisons between states, data were filtered to
include only herds designated as having Holsteins to eliminate the potentially confounding effect
of breed. A total of 764,196 records from 9757 Holstein herds were included in the dataset (5,021
records from 98 herds in FL, 184,521 records from 2708 herds in MN, 552,542 records from
6647 herds in PA, and 15,825 records from 303 herds in TX).
180
Cosinor rhythmometry
The fit, amplitude, and time at peak of a 12 month rhythm was determined among
selected states and breeds using cosinor rhythmometry as described by (Salfer et al., 2019).
Briefly, response variables were fit to the linear form of cosine functions with a 12 month period
in ASREML version 14 with the following model:
𝑦𝑖𝑗𝑘𝑙𝑚 = 𝜇 + 𝑆𝑖 + 𝐴 + 𝐵𝑘 + 𝑆𝑖 × 𝐴𝑗 + 𝑆𝑖 × 𝐵𝑘 + 𝑀𝑙 + 𝑌𝑚 + 𝜀𝑖𝑗𝑘𝑙𝑚
Where yijklm is the response variable of interest, μ is the overall mean, Si is the fixed effect of state,
Aj is the linear form of the sin function, Bk is the linear form of the cosine function, Ml is the fixed
effect of average days in milk, Ym is the random effect of year (I = 2004 to 2016) and εijklm is the
residual error. The fit of the linear form of cosine functions was compared to a reduced model
testing the linear effect of day of year:
𝑦𝑖𝑗𝑘𝑙 = 𝜇 + 𝑆𝑖 + 𝐷𝑗 + 𝑆𝑖 × 𝐷𝑗 + 𝑀𝑘 + 𝑌𝑙 + 𝜀𝑖𝑗𝑘𝑙
Where yijkl is the response variable of interest, μ is the overall mean, Si is the fixed effect of state,
Dj is the fixed effect of day of year (i= 1 to 365), Mk is the fixed effect of average days in milk, Yl
is the random effect of year (i= 2004 to 2016), and εijkl is the residual error. The fit of the annual
rhythms was determined by a zero amplitude test comparing the residual sums of squares between
the full model containing linearized cosine functions and the reduced model (Cornelissen, 2014).
The amplitude (peak minus mean) and acrophase (time at peak) and their significance tests were
determined using a script written in R ver. 3.4.3 according to equations of Bourdon et al. (1995).
Effect of breed on annual rhythms of production
The effect of breed was tested based on breed designation codes in the database. Breeds
within fewer than 50 herds were removed, leaving Holsteins (n = 8,790 herds, 675,068 records),
Jerseys (n = 398 herds, 21,510 records), Brown Swiss (n = 67 herds, 2,913 records), and
crossbreds (n = 1463 herds; 53,444 records). A full model containing the fixed effects of breed,
181
cosine terms, the interaction of breed with cosine terms, and average days in milk and the random
effect of year was compared to a reduced model containing the fixed effects of breed, day of year,
and average days in milk, and the random effect of year.
Comparison of annual rhythm with temperature
To compare the fit of the annual rhythm with the fit of a model testing the effect of
environmental temperature, daily climate data were collected from the U.S. National Weather
Service from representative weather stations centrally located near heavy populations of dairy
herds in each state (Tallahassee, FL; Litchfield, MN; Harrisburg, PA, Abilene, TX). The effect of
daily maximum temperature on milk and milk fat and protein yield and milk fat and protein
concentration was tested using the following model:
𝑦𝑖𝑗𝑘𝑙 = 𝜇 + 𝑆𝑖 + 𝑇𝑗 + 𝑆𝑖 × 𝑇𝑗 + 𝑀𝑘 + 𝑌𝑙 + 𝜀𝑖𝑗𝑘𝑙
Where yijkl is the response variable of interest, μ is the overall mean, Si is the fixed effect of state Tj
is the maximum daily temperature, Mk is the fixed effect of average days in milk, Yl is the random
effect of year (i= 2004 to 2016), and εijkl is the residual error. The coefficient of determination
(R2) was calculated for both the model containing cosine functions with a 12-month period
(Equation 1) and the model with maximum temperature (Equation 3). An F-test was performed to
compare the fit of the two models (Glatting et al., 2007).
Relationships among environmental variables and milk production responses
The relationship between daylength, change in daylength, and maximum temperature and
the annual rhythms of milk synthesis were tested to better understand which environmental
predictors best explain these annual rhythms. Daylength (min of daylight) for each day of the year
was determined using the latitude of the weather stations used for collection of temperature data
(Tallahassee, FL: 30.4°N; Litchfield, MN: 45.1°N; Harrisburg, PA: 40.3°N, Abilene, TX: 32.4°N)
182
of the from the United States Naval Observatory website
(http://aa.usno.navy.mil/data/docs/Dur_OneYear.php). The change in day length across the year
was computed for each region as described by Forsythe et al. (1995). These environmental
predictors were correlated with both the raw milk and milk component responses and with the
fitted 12 month cosine function as described by Roenneberg and Aschoff (1990). Pearson
correlations were performed using the cor.test() function of R ver. 3.4.3 (R Development Core
Team, 2013).
RESULTS
Annual Rhythms within Selected States
An interaction between state (PA, MN, TX, and FL) and the cosine function with a 12
month period occurred for milk and milk fat and protein yield and milk fat and protein
concentration so the fit, amplitude, and acrophase of the rhythm were determined for each state
(P < 0.001). Average milk yield was greatest in Pennsylvania (26.8 kg), followed by Minnesota
(26.7), Texas (25.8), and Florida (23.9; P < 0.001). In all four states, milk yield fit the 12 month
cosine function, indicating the presence of an annual rhythm (P < 0.001; Figure 8-1A). The
amplitude of the 12-month rhythm of milk yield differed by state, with Florida (3.31 kg) and
Texas (3.06) having amplitudes over twice that of Pennsylvania (1.35) and Minnesota (1.29; P <
0.001). The acrophase, or day at peak, occurred on April 11 for both Minnesota and
Pennsylvania, while Florida peaked 4 days earlier on April 7, and Texas peaked 8 days earlier on
April 3.
Milk fat and protein concentration also fit an annual rhythm in all states (P < 0.002;
Figure 8-1C&E). The amplitude of the 12-month rhythm of milk fat concentration was greatest
in PA (0.16%), followed by MN (0.14), TX (0.09), and FL (0.06). The acrophase of the annual
183
rhythm occurred near the first of the year and was similar between PA (Jan 7) and TX (Jan 8), but
occurred over one week earlier in MN (Dec 29), and nearly a month earlier in FL (Dec 3). The
annual rhythm of milk protein concentration differed between regions similar to milk fat
concentration, with FL having a lower amplitude (0.05%) than MN (0.10), TX (0.10), and PA
(0.09; P < 0.001). The acrophase of the protein rhythm was similar between MN, PA, and TX,
with all occurring within one day of Dec 21, while FL had an acrophase occurring 19 d earlier on
Dec 2.
The annual rhythms in milk yield and milk fat and protein concentration caused a rhythm
in milk fat and protein yield, which fit an annual rhythm regardless of state, with differences in
amplitude and acrophase among states (P < 0.001; Figure 8-1B&D). The amplitude of the fat
yield rhythm was greatest in TX (117 g), followed by FL (110), with much lower amplitudes
observed in PA (61.6) and MN (52.6). The acrophase of the 12-month rhythm of fat yield
occurred on April 2 in FL and was 10 days earlier in TX (March 23). The annual rhythm of fat
yield peaked about a month earlier in PA and MN, occurring on March 1 and March 2,
respectively (P < 0.001). The annual rhythm of protein yield did not differ in amplitude between
FL (94.3 g) and TX (93.1), which both oscillated more than MN (39.4) and PA (39.3; P < 0.001).
The acrophase of protein yield was different between all states, peaking on March 31 in FL,
March 17 in TX, March 7 in FL, and March 5 in MN (P < 0.001).
Annual Rhythms among Breeds
Annual rhythms were detected for milk and milk fat and protein yield and milk fat and
protein concentration regardless of breed (P < 0.001). As expected, average milk yield was
greatest for Holsteins (26.6 kg), followed by crossbreds (21.6), Brown Swiss (21.1), and Jerseys
(17.7; Figure 8-2A). The amplitude of milk yield varied by breed with Holsteins having a 106,
54, and 40%, greater amplitude than crossbreds, Jerseys, and Brown Swiss, respectively (P <
184
0.05). However, no differences were detected among the other three breeds (P > 0.05). The
greater amplitude in Holsteins is likely related to greater milk yield, and when amplitude was
examined as a percent of average milk yield, there was little difference among breeds (3% for
Brown Swiss, 5% for Holsteins, 5% for Jerseys, and 4% for crossbreds). Breeds also varied in
acrophase of the annual rhythm of milk yield with Holsteins (Apr. 1) and crossbreds (Apr. 22)
peaking earlier than Jerseys (May 11) and Brown Swiss (May 16). The annual rhythms of fat and
protein concentration also varied by breed (Figure 8-2D&E). Fat concentration was greatest in
Jerseys (4.71%), followed by Brown Swiss (4.07), crossbreds (3.95), and Holsteins (3.72). The
amplitude of the annual rhythm of fat concentration was greatest in Jerseys (0.27%) and lowest in
Holsteins (0.14%; P < 0.05), with Brown Swiss (0.23%) and Crossbreds (0.20%) having a similar
rhythm amplitude (P > 0.05). All breeds had an acrophase of milk fat concentration that occurred
between Jan 5th and 7th (P > 0.05). Similar to fat concentration, the amplitude of protein
concentration was greatest for Jerseys (0.15%), followed by Brown Swiss (0.14), crossbreds
(0.12), and Holsteins (0.08). The acrophase of the protein concentration rhythm occurred on Dec.
21 for both Holsteins and Jerseys, two days earlier than crossbreds and 8 d earlier than Brown
Swiss.
Milk fat yield was greatest in Holsteins (990 g; P < 0.05), followed by Brown Swiss (899
g) and crossbreds (850g), and was lowest in Jerseys (836 g; P < 0.05). Similar to fat yield, protein
yield was greatest in Holsteins (811 g), did not differ between Brown Swiss (745 g) and
crossbreds (689 g), and was lowest in Jerseys (635 g). The annual rhythm of milk fat yield had an
amplitude of 64.2 g in Holsteins 69, 55, and 42% greater than Brown Swiss, Jerseys and
crossbreds, respectively (P < 0.05; Figure 8-2B). The acrophase of fat yield did not differ among
Brown Swiss (Feb. 6), crossbreds (Feb. 23), or Holsteins (Feb. 24 P > 0.05), but the acrophase of
Jerseys (Mar 3) acrophase was delayed 25 d compared to Brown Swiss (P < 0.05). Holsteins also
had a larger amplitude of milk protein (42.8 g) than the other three breeds. Protein yield peaked
185
on Feb 12 in Brown Swiss, which was 16 d earlier than Holsteins, 18 d earlier than crossbreds
and 31 d earlier than Jerseys (P < 0.05; Figure 8-2C).
Comparison of Annual Rhythm and Temperature Models
To determine the influence of environmental temperature versus the annual rhythm, a
model containing the daily maximum temperature of a representative weather station in each state
was compared to the cosine function with a period of twelve months. Milk yield fit a 12 month
cosine function better than the model including the effect of maximum temperature (P < 0.0001;
Table 8-1). Furthermore, among the four states (Florida, Minnesota, Pennsylvania and Texas), all
fit a cosine function better than maximum temperature (P < 0.0001). Model R2 was increased
from 0.16 in the temperature model to 0.19 in the 12 month cosine model for the combined
dataset, and was also greater for the cosine function in all individual states. Florida (0.14 vs 0.32)
and TX (0.23 vs. 0.35) had the greatest improvements in model R2 from the temperature model to
the cosine function model.
Fat concentration fit a cosine function better than the temperature model in the combined
data set (P < 0.0001) and in MN, PA, and TX (P < 0.0001; Table 8-1). However, for FL, the
temperature model fit fat concentration better than the cosine function (P < 0.0001). Model R2 of
the combined dataset was improved from 0.30 in the temperature model to 0.31 in the cosine
model. Furthermore, the R2 of the cosine model was 0.01 units greater in MN, PA, and TX, but
were the same for both models in FL. Similarly, milk protein concentration fit a 12 month cosine
function better than maximum temperature in the combined dataset and in MN, PA, and TX (P <
0.0001). Like fat concentration, milk protein concentration fit temperature better than the cosine
function in FL (P < 0.0001). Model R2 of protein concentration was improved from 0.36 in the
temperature model to 0.39 in the cosine model in the combined dataset. Among the individual
186
states the R2 of the cosine model was greater than the temperature model for protein
concentration in MN, PA, and TX, but was not different between models in FL.
Fat and protein yield also both fit a cosine function better than temperature model for the
combined model (P < 0.0001) and each individual state (P < 0.0001). Model R2 of the combined
dataset was improved from 0.12 in the maximum temperature model to 0.15 in the cosine
function model for fat yield and improved from 0.12 to 0.14 for protein yield.
Relationships among Environmental Variables and Milk Production
Daylight and change in daylight change across the year, and appear to follow similar
daily patterns to milk component concentrations and milk yield, respectively (Figure 8-3B&C).
There was a significant relationship among daylength, change in daylength, and environmental
temperature with all production responses (milk yield, milk fat concentration, milk protein
concentration, milk fat yield, milk protein yield; P < 0.0001), with the exception of daylength and
milk protein concentration (P = 0.14; Figure 8-3A). Change in daylength was the environmental
predictor with the strongest relationship to both milk yield (r = 0.19) and the fitted cosine
function (r = 0.85). Maximum temperature had a negative correlation with both milk yield (r = -
0.06) and cosine-adjusted milk yield (r = -0.14), while daylength had a weak positive relationship
(raw: r = 0.04; fitted: r = 0.33).
For both raw data and the fitted cosine function, absolute day length had a strong
negative relationship with milk fat (raw; r = -0.31; fitted: -0.88) and protein (raw; r = -0.40; fitted:
-0.92) concentrations (Figure 8-3A). Maximum temperature also exhibited a strong negative
correlation with both milk fat (raw; r = -0.30; fitted: -0.85) and protein concentration (raw; r = -
0.36; fitted: -0.77). Alternatively, the correlation between change in daylength and raw and fitted
milk fat concentration was relatively low (raw; r = 0.07; fitted: 0.23). Milk protein concentration
187
had a weak negative relationship with change in daylength (r = -0.003) and the fitted cosine
function was not correlated (r = 0.002; P = 0.14).
Fat yield, like milk yield, was the most strongly correlated with change in daylength (r =
0.22). Moreover, milk fat yield was also negatively correlated with maximum temperature (r = -
0.19) and daylength (r = 0.10). The fitted cosine function of fat yield also had a strong positive
relationship with change in daylength (r = 0.83) and had a reasonably strong negative
relationships with maximum temperature (r = -0.62) and daylength (r = -0.28). Similarly, both
raw protein yield and the fitted cosine function had the strongest relationship with change in
daylength (raw: r = 0.19; fitted: 0.77), followed by a strong negative relationship with maximum
temperature (raw: r = -0.14; fitted: -0.58), and a weaker negative relationship with absolute
daylength (r = -0.07; -0.27).
Regression Equations to Adjust for Seasonal Changes in Production
The consistency of annual rhythms among years allowed for the derivation of regression
equations to adjust expected production for each month. Equations derived from the linear form
of the cosine function significantly predicted actual milk and milk fat and protein yield and milk
fat and protein concentration for each of the 4 regions (P < 0.001; Table 8-2). Furthermore, the
expected deviations from mean responses were computed from these equations for rapid
adjustment of milk and milk fat and protein yield and milk fat and protein concentration to
account for the seasonal rhythm (Supplemental Table S8-1 - Supplemental Table S8-4).
DISCUSSION
Milk synthesis followed an annual rhythm in all four states that were examined that
represent different regions of the U.S. In all states, milk yield peaked in early April, whereas fat
188
and protein concentration peaked at the start of the year. These results were similar to previous
accounts of seasonal variation in milk yield. Wood (1970) detected a ‘spring hump’ in milk yield
across multiple years. Moreover, Bailey et al. (2005) determined that the maximal quantity of
milk shipped per farm in the Mideast Federal Milk Market Order occurred in March, April, and
May. Salfer et al. (2019) observed that in 11 Pennsylvania dairy herds, the yearly rhythm of milk
yield peaked on March 13, while fat concentration peaked on January 25, and protein
concentration peaked on December 20.
The region of the U.S. had a clear effect on the robustness of the annual rhythms of milk
synthesis. States in the southern U.S. (FL and TX) had greater amplitudes of milk and milk fat
and protein yield, whereas states in the northern U.S. (MN and PA) had greater amplitudes of
milk fat and protein concentration. These results concurred with those observed by Salfer et al.
(2019), which demonstrated that the Milk Marketing Order located in Florida had a lower
amplitude of fat and protein concentration than other orders. Together, these results suggest that
the amplitude of annual rhythms is directly influenced by the latitude, perhaps related to
differences in photoperiod. While data from the southern hemisphere would be insightful for
better understanding the relationship between photoperiod and annual rhythms of milk synthesis,
these data are limited and confounded by management system. In New Zealand, milk yield peaks
in September through January, and fat and protein concentration peak from June through August
(Auldist et al., 1998). This would suggest that annual rhythms of milk synthesis are opposite in
the southern compared to the northern hemisphere. However, the cows in this experiment were
housed on pasture and results were likely affected by changes in pasture composition across the
year and perhaps by seasonal breeding.
Of the four breeds examined, all displayed annual rhythms of production with little
differences in the rhythms among breeds. Milk yield peaked approximately one month later in
Brown Swiss and Jerseys than Holsteins and crossbreds, but this difference is unlikely to be
189
biologically significant. All of major breeds studied are of the Bos taurus species and were
developed in northern Europe and are relatively genetically similar. An interesting comparison
would be between Bos taurus breeds and animals of the Bos indicus species, which were
developed in tropical climates where the change in photoperiod throughout the year is smaller,
but data is not available to our knowledge. Chital (Axis axis), a species of deer native to the
Indian subcontinent near the equator, show asynchronous annual rhythms of gonadal activity both
in the wild and captivity, whereas these rhythms are well-synchronized in related species of deer
native to Europe (Loudon and Curlewis, 1988).
The annual changes in milk production are often attributed to heat stress. Heat stress
occurs during mid-to late summer because of high ambient temperature and humidity (West,
2003). While the effects of heat stress on milk production are well-established, several factors
suggest that the annual rhythm of milk synthesis is influenced by factors beyond heat stress alone.
First, if the summer decrease in production was caused only by heat stress, yearly production
would be expected to remain stable during the fall through spring, with an acute decline in
production during the summer. Furthermore, the nadir of milk yield occurs during early October,
well past peak temperature, but could be a carry-over effect. Interestingly, average milk, fat and
protein yield fall below the fitted cosine function in June, July, and August, particularly in Texas
and Florida. This suggests that there is an additional depression in milk yield below the annual
rhythm during the summer, likely due to the additive effect of heat stress.
In the current study, the effect of heat stress was directly compared to the annual rhythm
using an F-test that tested the fit of a model testing the main effect of maximum daily
temperature, to a model testing the linear form of the cosine function. Milk, fat and protein yield
were better predicted by the annual rhythm than by maximum temperature in all regions
examined. Furthermore, the annual rhythm was a better predictor of fat and protein concentration
in Minnesota, Pennsylvania and Texas. In Florida, however, fat and protein concentration were
190
better predicted by maximum temperature. This result may be because Florida has a lower
amplitude annual rhythm than other states, so variability in environmental temperature may have
a greater influence on production. Maximum temperature was highly correlated with fat and
protein concentration and their respective fitted cosine functions. However, this effect appears to
be confounded with daylength. Daylength follows a similar yearly pattern to maximum
temperature (Figure 8-3B&D). Furthermore, daylength has a stronger negative relationship with
fat and protein concentration and their fitted cosine functions than maximum temperature.
Importantly, a summary of controlled experiments testing the high ambient temperature using
environmental chambers demonstrate that, while milk yield is consistently decreased by heat
stress, fat concentration is consistently increased, and the effects on protein yield are small and
inconsistent (Table 8-3). While milk yield was decreased by heat stress in all 6 experiments, with
an average decrease of 6.7 kg, fat concentration was increased by heat stress in 5 out of the 6
experiments and was increased 0.22 percentage units on average. Protein concentration was
decreased in 4 and increased in 2 experiments and was decreased 0.12 percentage units on
average. The decrease in milk yield is consistent with the annual rhythm of milk yield, but the
increase in milk fat during heat stress is opposite of the rhythm. These results suggest that the
yearly pattern of milk fat is regulated by a different mechanism.
The pattern of milk synthesis appears to represent an endogenous annual rhythm rather
than a consequence of heat stress. Annual rhythms occur in a wide variety of organisms as a
mechanism to predict seasonal changes in the environment before they occur (Lincoln, 2019).
Commonly, these rhythms are involved in optimizing reproductive processes to improve the
likelihood of offspring survival. For example, ewes express estrous only during late fall and early
winter, restricting the time of lambing to the spring (Shelton and Morrow, 1965). While dairy
cows are not generally considered to be seasonal breeders, Hansen (1985) detected modest effects
of season and photoperiod length on postpartum calving interval. As a component of
191
reproduction, annual rhythms of milk fat and protein concentration may exist to provide neonates
with nutrient-dense milk during the winter when energetic demands are higher. Evidence suggests
that other physiological responses follow annual rhythms in dairy cows. Piccione et al. (2012)
discovered that the concentrations of creatinine, triacylglycerides, and β-hydroxybutyrate follow
12-month rhythms. Moreover, Petitclerc et al. (1983) observed that circulating prolactin follows
an annual rhythm that persists even after blinding or pinealectomy. The core body temperature is
also influenced by season, with greater daily fluctuation in temperature during the summer
compared to winter (Kendall and Webster, 2009).
Annual rhythms are governed by endogenous circannual clocks that convert
environmental signals to time-of-year cues that influence physiology in a seasonal manner. In
mammals and birds, a central circannual pacemaker located in the pars tuberalis (PT) of the
pituitary gland has been described (Wood and Loudon, 2018). Notably, annual rhythms can
persist for long periods of time after organisms are placed in constant conditions (Pengelley et al.,
1976; Gwinner, 2003). When annual rhythms are allowed to freerun in the absence of
environmental stimuli, they typically have a period of only 10 to 11 months and require
entrainment of circannual clocks to maintain a 12 month period (Lincoln, 2019). Several
entrainment cues such as photoperiod, social stimuli, and nutritional cycles can synchronize
endogenous annual rhythms to the external environment (Lincoln et al., 2006).
To study the potential role of photoperiod on entrainment of annual rhythms of milk
synthesis, the relationships among milk production responses and both absolute photoperiod
(daylength) and change in daylength were tested. In addition to correlating these predictors with
raw production responses, they were correlated with the fitted response from the linear form of
the cosine function. Fat and protein concentration were most highly correlated with daylength,
and in particular the 12 month fitted cosine function was very highly correlated. These results
suggest that fat and protein concentration may be controlled by a circannual oscillator that is
192
entrained by photoperiod. A wealth of research has described the relationship between
photoperiod length and milk synthesis in dairy cattle. Placing cows in an long days (> 12 h
daylight) consistently stimulates milk yield, with the optimal response occurring with a 16 h light:
8 h dark (16L: 8D) photoperiod (Dahl et al., 2000). However, there are a few interesting
contrasts between the results of studies using artificial photoperiod manipulation and the results
of the current study examining the naturally-occurring annual rhythm. First, while long-day
lighting stimulates milk yield, it does not alter fat and protein concentration. In the current study,
not only are fat and protein concentration strongly correlated to daylength, the long days are
associated with decreases in fat and protein concentration. Moreover, milk yield is poorly
correlated with daylength, but maximal milk yield occurs during the spring, when daylength is
shorter than 12 h.
The paradox between controlled photoperiod manipulation trials and the annual rhythm
of milk yield may be caused by animals becoming photorefractory to artificial long-day lighting.
In seasonally-breeding mammals such as Suffolk sheep (Ovis aries) and Syrian hamsters
(Mesocricetus auratus), exposure to a fixed photoperiod for long periods of time causes a
spontaneous reversion to a physiological state expected during the opposite photoperiod
(Robinson and Karsch, 1984; Lincoln et al., 2005). In short-day breeding sheep, estrus is
inhibited by fixed 8L:16D photoperiods, while the estrus of long-day breeding hamsters is
conversely inhibited by fixed 14 light:10 dark photoperiods. Induction of photorefractoriness
requires 4 to 12 weeks of a fixed photoperiod prior to a physiological change, which is consistent
with results of long-day lighting experiments, which require 4 weeks of adaptation before
increased milk yield is observed (Dahl et al., 2000).
As mentioned above, milk, fat and protein yield were poorly correlated with absolute
daylength, but had a strong positive relationship with change in daylength. This suggests that,
rather than being entrained by the absolute daylength, a separate circadian oscillator controlled
193
change in daylength may be responsible for entraining the annual rhythms of milk yield.
Furthermore, the annual rhythm of milk yield may be related to the phase-relationship between an
oscillator entrained by photoperiod, and a daily circadian oscillator. Bartell and Gwinner (2005)
proposed that the degree of synchronization between a circannual and a circadian oscillator can
entrain seasonal responses.
The consistency of annual rhythms among years and herds allowed for the derivation of
regression equations to normalize milk and milk fat and protein yield and milk fat and protein
concentration. Furthermore, these regression equations were used to calculate adjustment factors
to standardize these responses across the year. Using these adjustment factors may allow dairy
producers to remove the annual variation in milk production to make better-controlled nutritional
and management decisions. A spreadsheet developed in Microsoft Excel that uses these
regression equations for rapid calculation of yearly rhythm-adjusted milk and milk fat and protein
yield and milk fat and protein concentration are included in the supplement.
CONCLUSIONS
The milk synthesis of dairy cows follows an annual rhythm that is affected by region of
the United States. The annual rhythm of milk yield appears to be related to, and is perhaps
entrained by the change in daylength. Fat and protein concentration have a strong relationship to
absolute daylength, suggesting that they may be entrained by photoperiod. While maximum
environmental temperature is also highly correlated to fat and protein concentration, the cosine
function fits the data better, suggesting it better explains their yearly pattern. Fat and protein
concentration closely align with the lengthening and shortening of days, while milk yield aligns
with the change in daylength. These annual rhythms may to be driven by endogenous circannual
oscillators, and may not be able to be influenced through management interventions. Further
194
research must be conducted to understand the physiological mechanisms governing the annual
rhythms of milk synthesis. Dairy producers can account for the annual rhythms of production by
standardizing milk production for this rhythm across the year.
195
FIGURES
Figure 8-1. Annual rhythms of milk, fat and protein yield and fat and protein concentration in
Holstein dairy herds in Florida (FL), Minnesota (MN), Pennsylvania (PA), and Texas (TX).
Data includes Dairy Herd Information Association (DHIA) test day results from individual herds in FL (98
herds, 5,021 records), MN (2708 herds, 184,521 records), PA (6647 herds, 552,542 records), and TX (303
herds, 15,825 records) acquired from Dairy Records Management Systems (DRMS). Panels show: (A)
Effect of state on annual rhythms of milk yield, (B) Effect of state on annual rhythms of fat yield, (C)
Effect of state on annual rhythms of fat concentration, (D) Effect of state on annual rhythms of protein
yield, (E) Effect of state on annual rhythms of protein concentration. Data are presented as the least
squares mean of each month with error bars representing the SEM and the line is the fitted cosine function
with a period of 12 months. Cosinor analysis is shown below each panel and include the amplitude (Amp;
difference between peak and mean), acrophase (Acro; time at peak of the rhythm), and the P-value of the
zero-amplitude test.
196
Figure 8-2. Annual rhythms of milk and milk fat and protein yield and milk fat and protein
concentration by breed.
Data includes Dairy Herd Information Association (DHIA) test day results from individual herds
designated to contain Brown Swiss (BS; 67 herds, 2,913 records), Holsteins (HO, 8790 herds, 675,071
records), Jerseys (JE; 398 herds, 21,510 records), and crossbreds (1463 herds, 53,444 records) acquired
from Dairy Records Management Systems (DRMS). Panels show: (A) Effect of breed on annual rhythms
of milk yield, (B) Effect of breed on annual rhythms of fat yield, (C) Effect of breed on annual rhythms of
fat concentration, (D) Effect of breed on annual rhythms of protein yield, (E) Effect of breed on annual
rhythms of protein concentration. Data are presented as the least squares mean of each month with error
bars representing the SEM and the line is the fitted cosine function with a period of 12 months. Cosinor
analysis is shown below each panel and include the amplitude (Amp; difference between peak and mean),
acrophase (Acro; time at peak of the rhythm), and the P-value of the zero-amplitude test.
197
Figure 8-3. Relationship among environmental predictors and annual rhythms of milk and milk
fat and protein yield and milk fat and protein concentration by breed.
Panels show: (A) Correlogram showing relationship among environmental variables [daylength, change in
daylength (ΔDaylength), and maximum temperature (Max. Temp)] and milk production responses (Milk
yield, fat concentration, protein concentration, fat yield and protein yield). For each relationship, the
correlation is listed first, followed by the P-value, (B) daylength, (C) change in daylength, and (D) average
maximum temperature across the year for Florida (FL), Minnesota (MN), Pennsylvania (PA), and Texas
(TX). Fitted correlations show relationship of milk and milk fat and protein yield and fat and protein
concentration with the response predicted by a the model: 𝑦 = 𝑦𝑖𝑗𝑘𝑙𝑚 = 𝜇 + 𝑆𝑖 + 𝐴 + 𝐵𝑘+. 𝑆𝑖 × 𝐴𝑗 +
𝑆𝑖 × 𝐵𝑘 + 𝑀𝑙 + 𝑌𝑚 + 𝜀𝑖𝑗𝑘𝑙𝑚 where yijklm is the response variable of interest, μ is the overall mean, Si is the
fixed effect of state, Aj is the linear form of the sin function, Bk is the linear form of the cosine function, Ml
is the fixed effect of average days in milk, Ym is the random effect of year (i= 2004 to 2016) and εijklm is the
residual error.
198
TABLES
Table 8-1. Comparison between models testing the fit of a cosine function versus maximum
temperature.
1Cosine model: 𝑦𝑖𝑗𝑘𝑙𝑚 = 𝜇 + 𝑆𝑖 + 𝐴 + 𝐵𝑘+. 𝑆𝑖 × 𝐴𝑗 + 𝑆𝑖 × 𝐵𝑘 + 𝑀𝑙 + 𝑌𝑚 + 𝜀𝑖𝑗𝑘𝑙𝑚where yijklm is the
response variable of interest, μ is the overall mean, Si is the fixed effect of state, Aj is the linear form of the
sin function, Bk is the linear form of the cosine function, Ml is the fixed effect of average days in milk, Ym
is the random effect of year (i= 2004 to 2016) and εijklm is the residual error. 2Temperature model: 𝑦𝑖𝑗𝑘𝑙 = 𝜇 + 𝑆𝑖 + 𝑇𝑗 + 𝑆𝑖 × 𝑇𝑗 + 𝑀𝑘 + 𝑌𝑙 + 𝜀𝑖𝑗𝑘𝑙where yijkl is the response variable
of interest, μ is the overall mean, Si is the fixed effect of state Tj is the maximum daily temperature, Mk is
the fixed effect of average days in milk, Yl is the random effect of year (i= 2004 to 2016) and εijkl is the
residual error. 3F-test of temperature model fits data better than the cosine model (P < 0.0001).
R2 Cosine
Model1
R2 Temperature
Model2 F-Value P-value
Milk yield, kg/d
Combined 0.19 0.16 9695.7 <0.0001
FL 0.32 0.14 558.5 <0.0001
MN 0.13 0.11 1468.8 <0.0001
PA 0.20 0.17 6972.2 <0.0001
TX 0.35 0.23 875.4 <0.0001
Fat yield, g/d
Combined 0.15 0.12 6885.0 <0.0001
FL 0.19 0.07 161.4 <0.0001
MN 0.10 0.09 860.9 <0.0001
PA 0.15 0.12 5274.4 <0.0001
TX 0.23 0.11 618.0 <0.0001
Protein yield, g/d
Combined 0.14 0.12 8152.6 <0.0001
FL 0.24 0.10 183.9 <0.0001
MN 0.10 0.09 1324.5 <0.0001
PA 0.14 0.12 6147.3 <0.0001
TX 0.27 0.17 515.7 <0.0001
Fat concentration, %
Combined 0.31 0.30 3030.7 <0.0001
FL 0.07 0.07 -55.6 NS3
MN 0.22 0.21 584.2 <0.0001
PA 0.33 0.32 2803.3 <0.0001
TX 0.30 0.29 7.38 0.0001
Protein concentration, %
Combined 0.39 0.36 10504.9 <0.0001
FL 0.06 0.06 -7.22 NS3
MN 0.34 0.31 2276.1 <0.0001
PA 0.41 0.38 8573.5 <0.0001
TX 0.21 0.19 138.9 <0.0001
199
Table 8-2. Equations for cosine regression equations for milk and milk fat and protein yield and
milk fat and protein concentration for Florida (FL), Minnesota (MN), Pennsylvania (PA), and
Texas (TX).
1Parameters derived from fitting Dairy Herd Information Association (DHIA) test day data acquired from
Dairy Records Management Systems (DRMS) to a 12 month cosine function with the following equation:
𝑦 = 𝑥 − 𝐴 ∗ cos(2𝜋𝑚
12) + 𝐵 ∗ sin(
2𝜋𝑚
12). A cosine by state interaction was included so parameters could be
derived within each state.
Cosine Term1
Response A B R2 P-value
Milk, kg
FL -0.381 3.30 0.04 0.001
MN -0.249 1.26 0.15 <0.0001
PA -0.261 1.33 0.20 <0.0001
TX -0.175 3.06 0.41 <0.0001
Fat, g
FL -3.90 110.4 0.38 <0.0001
MN 25.0 46.3 0.19 <0.0001
PA 30.2 53.6 0.26 <0.0001
TX 15.5 116.4 0.43 <0.0001
Protein, g
FL -0.151 94.3 0.47 <0.0001
MN 16.8 35.7 0.19 <0.0001
PA 15.3 36.2 0.23 <0.0001
TX 21.9 90.5 0.44 <0.0001
Fat, %
FL 0.049 -0.025 0.10 <0.0001
MN 0.135 -0.005 0.28 <0.0001
PA 0.155 0.017 0.20 <0.0001
TX 0.093 0.013 0.17 <0.0001
Protein, %
FL 0.046 -0.025 0.23 <0.0001
MN 0.095 -0.014 0.40 <0.0001
PA 0.091 -0.015 0.41 <0.0001
TX 0.107 -0.022 0.35 <0.0001
200
Table 8-3. A summary of milk yield and milk fat and protein concentration responses in
experiments testing the effects of heat stress in environmental chambers.
Reference Milk, kg Fat, % Protein, %
Rhoads et al. (2009) -10.6 0.34 -0.13
Schwartz et al. (2009) -10.1 0.06 -0.22
Wheelock et al. (2010) -9.6 0.60 -0.27
Zimbelman et al. (2010) -0.1 -0.17 0.13
Baumgard et al. (2011) -6.2 0.28 -0.12
Rungruang et al. (2014) -3.4 0.20 -0.10
201
SUPPLEMENTAL TABLES
Supplemental Table S8-1. Adjustment factors to correct for annual rhythms of milk production in Florida.
Milk,
kg/d
Milk lb/d Fat, % Protein, % Fat, g/d Fat, lb/d Protein, g/d Protein, lb/d
Jan 0.00 0.00 -0.05 -0.05 0.0 0.00 0.00 0.00
Feb -1.86 -3.79 -0.03 -0.03 -59.8 -0.12 -47.22 -0.10
Mar -3.32 -6.18 0.00 0.00 -104.6 -0.20 -76.00 -0.17
Apr -3.99 -6.55 0.03 0.03 -122.4 -0.20 -78.61 -0.17
May -3.70 -4.77 0.05 0.05 -108.5 -0.14 -54.36 -0.12
Jun -2.52 -1.35 0.05 0.05 -66.5 -0.02 -9.75 -0.02
Jul -0.76 2.82 0.04 0.04 -7.8 0.12 43.27 0.10
Aug 1.10 6.61 0.02 0.02 52.0 0.24 90.50 0.20
Sep 2.56 9.00 -0.01 -0.01 96.8 0.32 119.27 0.26
Oct 3.23 9.37 -0.04 -0.04 114.6 0.32 121.88 0.27
Nov 2.94 7.59 -0.06 -0.06 100.7 0.26 97.63 0.22
Dec 1.75 4.17 -0.06 -0.06 58.7 0.14 53.02 0.12
202
Supplemental Table S8-2. Adjustment factors to correct for annual rhythms of milk production in Minnesota.
Milk,
kg/d
Milk lb/d Fat, % Protein, % Fat, g/d Fat, lb/d Protein, g/d Protein, lb/d
Jan 0.00 0.00 -0.14 -0.11 0.0 0.00 0.0 0.00
Feb -0.62 -3.28 -0.16 -0.11 -59.8 -0.12 -47.2 -0.10
Mar -1.14 -5.68 -0.14 -0.09 -104.6 -0.20 -76.0 -0.17
Apr -1.42 -6.55 -0.08 -0.05 -122.4 -0.20 -78.6 -0.17
May -1.38 -5.65 0.00 0.00 -108.5 -0.14 -54.4 -0.12
Jun -1.05 -3.24 0.08 0.05 -66.5 -0.02 -9.7 -0.02
Jul -0.50 0.05 0.13 0.08 -7.8 0.12 43.3 0.10
Aug 0.12 3.33 0.15 0.09 52.0 0.24 90.5 0.20
Sep 0.64 5.73 0.13 0.07 96.8 0.32 119.3 0.26
Oct 0.92 6.60 0.07 0.03 114.6 0.32 121.9 0.27
Nov 0.89 5.70 -0.01 -0.02 100.7 0.26 97.6 0.22
Dec 0.55 3.29 -0.09 -0.07 58.7 0.14 53.0 0.12
203
Supplemental Table S8-3. Adjustment factors to correct for annual rhythms of milk production in Pennsylvania. Milk,
kg/d
Milk lb/d Fat, % Protein, % Fat, g/d Fat, lb/d Protein, g/d Protein, lb/d
Jan 0.00 0.00 -0.17 -0.11 -31.0 -0.07 -19.7 0.05
Feb -0.70 -1.49 -0.16 -0.09 -54.0 -0.12 -10.7 0.01
Mar -1.29 -2.61 -0.11 -0.05 -62.7 -0.13 0.0 0.00
Apr -1.60 -3.07 -0.03 0.00 -54.8 -0.11 9.5 0.01
May -1.55 -2.73 0.05 0.04 -32.5 -0.05 15.3 0.04
Jun -1.16 -1.69 0.11 0.07 -1.6 0.02 15.8 0.09
Jul -0.52 -0.22 0.14 0.08 29.4 0.09 10.9 0.14
Aug 0.18 1.27 0.13 0.06 52.3 0.14 1.9 0.17
Sep 0.77 2.39 0.08 0.02 61.0 0.15 -8.7 0.18
Oct 1.08 2.84 0.00 -0.03 53.2 0.13 -18.2 0.17
Nov 1.03 2.50 -0.08 -0.07 30.8 0.07 -24.0 0.14
Dec 0.63 1.46 -0.14 -0.10 0.0 0.00 -24.6 0.09
204
Supplemental Table S8-4. Adjustment factors to correct for annual rhythms of milk production in Texas. Milk,
kg/d
Milk lb/d Fat, % Protein, % Fat, g/d Fat, lb/d Protein, g/d Protein, lb/d
Jan 0.00 0.00 -0.10 -0.13 -31.1 -0.20 -43.7 -0.19
Feb -1.65 -3.92 -0.09 -0.10 -90.2 -0.34 -86.0 -0.31
Mar -2.91 -6.32 -0.06 -0.06 -129.3 -0.41 -111.1 -0.37
Apr -3.43 -6.55 -0.02 0.00 -137.9 -0.40 -112.3 -0.36
May -3.08 -4.55 0.03 0.05 -113.7 -0.31 -89.3 -0.28
Jun -1.96 -0.86 0.07 0.08 -63.3 -0.16 -48.2 -0.15
Jul -0.35 3.54 0.08 0.08 0.0 0.00 0.0 0.00
Aug 1.30 7.46 0.08 0.06 59.1 0.14 42.3 0.12
Sep 2.56 9.85 0.05 0.01 98.2 0.21 67.4 0.18
Oct 3.08 10.08 0.00 -0.04 106.8 0.20 68.6 0.17
Nov 2.73 8.08 -0.05 -0.09 82.7 0.11 45.6 0.09
Dec 1.60 4.39 -0.09 -0.13 32.2 -0.04 4.5 -0.04
205
Chapter 9
Integrative Discussion
Dairy producers are well aware of daily and seasonal changes in milk production.
Moreover, differences in milk and milk components between morning and evening milking
(Gilbert et al., 1972), and among different seasons of the year (Wood, 1976) have been
characterized in the literature. However, the regulation of these changes have been largely
unexplored and are not considered in modern dairy management. Developing a better
understanding of the biological rhythms governing milk synthesis may uncover opportunities to
improve the efficiency of dairy production. Therefore, the research conducted for this dissertation
sought to examine the nutritional and environmental factors affecting daily and annual rhythms of
milk synthesis.
The first four experiments focused on the relationship between feed intake and daily
rhythms of milk synthesis using an in vivo approach with lactating cows. Previous research has
provided groundwork to suggest that a relationship between feed intake and daily rhythms of milk
production exists. Rottman et al. (2014) demonstrated that increasing the frequency of feeding
from 1x to 4x/d damped the rhythms of milk yield, while Niu et al. (2014) observed differences in
the ratio of morning to evening milk production and daily patterns of plasma metabolites when
cows were fed at 8:30 AM versus 8:30 PM. Based on these initial reports, we wanted to
investigate the impact of the time of feed intake of the rhythms of milk synthesis more
completely. To do this, we employed time-restricted feeding to impose a more dramatic shift in
the feeding pattern of cows, while milking cows 4x/d to determine the daily rhythm of milk
synthesis (Chapter 3). Time-restricted feeding caused a substantial shift in the daily rhythms of
206
milk yield and milk components corresponding with the 12 h difference in feed delivery. These
results confirmed in a very clear fashion that feeding time does indeed affect the daily rhythms of
milk synthesis. Plasma hormone and metabolite concentrations were similarly shifted by the time
of feed intake, corresponding with the fasting cycle. Most interestingly, the response to fasting
was much greater in the night-restricted feeding group, who had much greater fasting non-
esterified fatty acid concentrations, indicating greater lipid mobilization. Cows typically increase
feed intake during the afternoon (Albright, 1993). Our results demonstrate that this daily pattern
of intake appears to be an inherent component of the cow’s physiology, because the cow appears
to continue to want to eat in the afternoon, even after being adapted to night-feeding. Although
not quantified, the cows on the night-restricted treatment also exhibited more food seeking
behavior and greater irritability than day-restricted cows.
In laboratory models, time-restricted feeding can shift physiological rhythms by
modifying the daily pattern of molecular clock gene expression (Stokkan et al., 2001). We wanted
to determine if this mechanism was responsible for the changes in the daily rhythms of milk
synthesis in Chapter 3. We used mammary biopsies to collect tissue from the mammary gland at
four times across the day in cows that were either fed 24 h/d or 16 h/d at night. Of the 6 clock
genes measured, 3 followed a daily rhythm. Time-restricted feeding shifted the daily rhythms of
expression of two clock genes CLOCK and CRY1, and induced a rhythm in REV-ERBα.
To our knowledge, this is the first demonstration that feeding can entrain the mammary
molecular circadian clock in cattle, and one of the first to demonstrate this in any species. On
many large-scale modern dairy farms, milking parlors are operating 24 h/d, leading to cows being
milked at all hours of the day and night. However, feeding of all cows is usually performed
within the same timeframe. Differences in the relationship between feeding and milking time
among individual cows may cause differences in the entrainment of their mammary gland. Future
research should be done to better elucidate this relationship. The results of this experiment not
207
only has exciting implications for dairy production, but human lactation as well. In humans 7% of
the transcriptome follows a daily rhythm (Maningat et al., 2009). Because humans have a less
consistent daily eating pattern than cows, they may be even more responsive to food entrainment.
Therefore, nutritional composition of breastmilk may be affected by meal pattern of the mother.
Research into this area may develop potential avenues to improve post-natal nutrition.
Unfortunately, while we observed compelling evidence to suggest the circadian clock of
the mammary gland is entrained by food intake, the results were not as clear cut as we hoped.
Unexpectedly, 3 of the measured clock genes failed to follow a daily rhythm according to the
zero-amplitude test. One potential cause of this is that we had an unexpectedly high variability in
the samples. Visually, all of the genes analyzed appear to be somewhat rhythmic. Moreover,
PER1 seems to be influenced by treatment with a large spike in expression at 0400 in control that
occurs 12 h later under night-restricted feeding, but again this change is not statistically
significant. The high sample variability may have been due to several potential factors. First, the
mammary tissue collected may have been more heterogeneous than anticipated. The mammary
gland is a complex organ containing multiple cell types including epithelial cells, smooth muscle
cells, and adipocytes (Inman et al., 2015). While biopsy collection was performed as consistently
as possible, variation in the proportions of individual cell types and their potential entrainment
may have contributed to the variation in the gene expression results. More frequent sampling
could have been employed to develop a better characterization of the daily pattern of gene
expression and perhaps helped us detect a rhythm, but mammary biopsy is an invasive species
and we wanted to avoid excess stress on the cows for both animal welfare and research purposes.
An increased number of experimental subjects could have also been used to increase the power of
the experiment, and this will likely be done in the future.
Recent research has illustrated the importance of post-transcriptional and post-
translational regulation of circadian rhythms (Robles et al., 2014, 2017). In Chapter 4, the
208
molecular circadian clock was only examined at the mRNA level, limiting its scope. In the future,
research should consider using proteomic and phosphoproteomic analyses, or antibody-based
techniques to understand these additional levels of regulation.
In Chapters 5 and 6, the effects of the time of fat and protein availability on the daily
rhythms of milk synthesis were studied by infusing cannulated cows with these nutrients post-
ruminally. The time of protein infusion had considerable effects on daily milk yield. If maintained
in commercial herds, the 7 lb reduction in milk yield by restricting protein infusion to the night
would have momentous implications for dairy farmers. This experiment provides compelling
evidence that if protein availability is limited during the day, milk synthesis will also be limited.
Alternatively, the time of fat infusion did not affect milk or milk components. Day infusion of fat
increased the robustness of milk yield and decreased the robustness of fat and protein
concentration (Chapter 5). Alternately, night infusion of protein increased the robustness of milk
yield and decreased the robustness of protein concentration (Chapter 6). These experiments were
the first to study individual nutrients on daily rhythms in the mammary gland. Interestingly, the
response of the daily rhythms of milk synthesis differed between infusions of fat and protein,
suggesting that these nutrients differentially affect the clock. Future research looking at the direct
effect of these nutrients on the mammary circadian clock should be performed.
In Chapters 5 and 6, mixtures of fatty acids or amino acids were used. Presumably,
individual fatty acids and/or amino acids may regulate the mammary clock differently. The
choice to use nutrient mixtures was made partially because no previous characterizations of the
role of nutrients on the mammary clock had been made. Providing a mixture of nutrients allowed
for detection of a general effect of each nutrient on the daily rhythms. Furthermore, performing in
vivo experiments with cows is expensive and time-consuming, and obtaining quantities of pure
fatty acids or amino acids large enough to perform these experiments would be very difficult and
costly. Therefore, experiments using readily available sources of fat and protein were used.
209
Because we have demonstrated that these nutrients do affect daily rhythms, future research
examining specific fatty acids and amino acids would be compelling. The most cost-effective
approach would be to test individual nutrients in a cell culture system first, before developing an
in vivo experiment.
Altering the time of fatty acid and protein infusion also influenced the daily rhythms of
plasma glucose, nonesterified fatty acid and urea nitrogen concentrations. These changes may
have been due to entrainment of peripheral tissues besides the mammary gland. Future research
should consider rhythms in other tissues. The liver of dairy cows is an incredible metabolic organ
that is responsible for producing nearly all of the glucose used by the animal via gluconeogenesis.
Additionally, adipose tissue and muscle are important metabolic organs for dairy cows. Currently
the molecular clocks in these tissues are poorly characterized but developing a better
understanding of them and their relationship to metabolism could reveal opportunities to improve
dairy cow efficiency. Another unexplored aspect of circadian biology in dairy cows is circadian
rhythms of the rumen microbiome. As ruminants, dairy cows are dependent on microbial
fermentation to meet their nutritional requirements. Recent research has shown that microbes in
the hindgut of laboratory models display daily rhythms that are related to the circadian rhythms of
the host (Heath-Heckman, 2016). However, this has been unexplored in ruminants.
Understanding the daily rhythms of microbial metabolism and their relationship to the host could
uncover promising nutritional strategies to improve dairy cow efficiency.
In Chapter 6, milk synthesis failed to exhibit a daily rhythm in the control treatment with
continuous infusion of sodium caseinate. This may suggest that sodium caseinate for 24 h/d
abolishes the rhythm of milk synthesis, but this cannot be confirmed without a negative control
where no sodium caseinate was infusion. Unfortunately in both Chapters 5 and 6, the availability
of cannulated cows was limited so no negative controls were included.
210
Importantly, the daily rhythms of milk synthesis observed in these experiments do not
represent true circadian rhythms. Detection of a true circadian rhythm requires that subjects be
placed in constant lighting conditions. Maintaining lactating cows in a constant lighting
environment is technically challenging because they would need to be housed in an environment
with no access to outside light with the necessary equipment to milk cows in place. At the
Pennsylvania State University, all cows must be moved through an alleyway exposed to natural
light on their way to the milking parlor. Development of a facility where cows could be placed
under constant lighting conditions would open up interesting avenues of research. Not only could
the presence or absences of a true circadian rhythm of milk synthesis be determined, research into
the relationships between light entrainment and food entrainment could be performed.
The characterizations of the rhythms of milk synthesis in Chapters 3, 5, and 6 were
limited by the frequency of milking. Practically, at the Penn State facility it was only possible to
milk cows 4x/d without drastically altering behavior. A facility where cows are milked in-place
would allow more frequent milking and better characterization of the rhythm.
Chapters 7 and 8 of this dissertation characterize annual rhythms of milk and milk
component production. In the Chapter 7, results demonstrated that annual rhythms are highly
conserved among individual cows, but vary by geographic region. To our knowledge, this
experiment was the first to describe yearly changes of milk production as an annual rhythm.
Using a much more robust dataset, we discovered that the annual rhythms differed between the
northern and southern U.S, with the amplitude of milk yield being greater in the south and the
amplitude of milk fat and protein concentration being greater in the north. Moreover, milk yield
was found to be highly related to change in daylength, while milk component concentrations were
highly correlated with absolute daylength, suggesting that two separate circannual oscillators may
regulate these responses. The observations made in these experiments represent a dogmatic shift
in the way that the dairy industry views seasonal changes in milk production. Previously, the
211
changes were thought to be strictly a consequence of environmental conditions, such as heat
stress. Our results clearly suggest that there is an additional effect of an annual rhythm. So far in
the dairy industry, this work has received significant attention. We hope that it will allow dairy
farmers and consultants to better manage cows across the year.
While characterizing the annual rhythms of milk synthesis is a strong initial step,
mechanistic work really must be performed in order to both better understand these rhythms and
develop approaches to modify them. Across all species, there is limited research examining the
mechanisms governing annual rhythms due to the long time-course required to conduct
experiments. In dairy cows, this is complicated even further because of the 9 month lactation
period, which would conflict with the annual rhythm. In other model organisms, ‘accelerated
year’ experiments with controlled lighting have been conducted that allow an entire annual cycle
to be compressed into about 3 months. This approach could be used to better understand the
annual rhythms governing milk production in dairy cows. Additionally, if the mechanisms
causing these rhythms become more well-understood, lighting strategies could be developed to
modify them and potentially improve milk production.
212
REFERENCES
Albright, J.L. 1993. Feeding behavior of dairy cattle. J. Dairy Sci. 76:485–498.
doi:10.3168/jds.S0022-0302(93)77369-5.
Allaman-Pillet, N., R. Roduit, A. Oberson, S. Abdelli, J. Ruiz, J.S. Beckmann, D.F. Schorderet,
and C. Bonny. 2004. Circadian regulation of islet genes involved in insulin production and
secretion. Mol. Cell. Endocrinol. 226:59–66. doi:10.1016/j.mce.2004.06.001.
Allen, M.S. 2014. Drives and limits to feed intake in ruminants. Anim. Prod. Sci. 54:1513.
doi:10.1071/AN14478.
Alvarado, S., T. Mak, S. Liu, K.B. Storey, and M. Szyf. 2015. Dynamic changes in global and
gene-specific DNA methylation during hibernation in adult thirteen-lined ground squirrels,
Ictidomys tridecemlineatus. J. Exp. Biol. 218:1787–1795. doi:10.1242/jeb.116046.
Alverdy, J., and E. Aoys. 1991. The effect of glucocorticoid administration on bacterial
translocation. Ann. Surg. 214:719–723. doi:10.1097/00000658-199112000-00012.
Andreen, D.M., I.J. Salfer, Y. Ying, and K.J. Harvatine. 2018. T150 Method for calibrating parlor
milk meters and adjusting milk weights for stall effects. J. Dairy Sci. 101:268.
AOAC. 2000. Official Methods of Analysis. W. Horwitz and G. Latimer, ed.
AOAC. 2006. Official Methods of Analysis. 18th ed. Arlington, VA.
Appuhamy, J.A.D.R.N., N.A. Knoebel, W.A.D. Nayananjalie, J. Escobar, and M.D. Hanigan.
2012. Isoleucine and leucine independently regulate mTOR signaling and protein synthesis
in MAC-T cells and bovine mammary tissue slices. J. Nutr. 142:484–491.
doi:10.3945/jn.111.152595.
Arnold, W., T. Ruf, and R. Kuntz. 2006. Seasonal adjustment of energy budget in a large wild
mammal, the Przewalski horse (Equus ferus przewalskii) II. Energy expenditure. J. Exp.
Biol. 209:4566–4573. doi:10.1242/jeb.02536.
Aschenbach, J.R., N.B. Kristensen, S.S. Donkin, H.M. Hammon, and G.B. Penner. 2010.
Gluconeogenesis in dairy cows: The secret of making sweet milk from sour dough. IUBMB
Life 62:869–877. doi:10.1002/iub.400.
Aschoff, J. 1955. Jahresperiodik der fortpflanzung bei warmblütern. Stud. Gen. 8:742–776.
Aschoff, J. 1963. Comparative physiology: Diurnal rhythms. Annu. Reviow Physiol. 25:581–600.
doi:10.1146/annurev.ph.25.030163.003053.
Asher, G., and P. Sassone-Corsi. 2015. Time for food: The intimate interplay between nutrition,
metabolism, and the circadian clock. Cell 161:84–92. doi:10.1016/j.cell.2015.03.015.
Auchtung, T.L., J.L. Salak-Johnson, D.E. Morin, C.C. Mallard, and G.E. Dahl. 2004. Effects of
photoperiod during the dry period on cellular immune function of dairy cows. J. Dairy Sci.
87:3683–9. doi:10.3168/jds.S0022-0302(04)73507-9.
Auldist, M.J., B.J. Walsh, and N. a Thomson. 1998. Seasonal and lactational influences on bovine
213
milk composition in New Zealand. J. Dairy Res. 65:401–411.
doi:10.1017/S0022029998002970.
Azzi, A., R. Dallmann, A. Casserly, H. Rehrauer, A. Patrignani, B. Maier, A. Kramer, and S.A.
Brown. 2014. Circadian behavior is light-reprogrammed by plastic DNA methylation. Nat.
Neurosci. 17:377–382. doi:10.1038/nn.3651.
Baeke, F., T. Takiishi, H. Korf, C. Gysemans, and C. Mathieu. 2010. Vitamin D: Modulator of
the immune system. Curr. Opin. Pharmacol. 10:482–496. doi:10.1016/j.coph.2010.04.001.
Bailey, A.M., S.J. Legan, and G.E. Demas. 2017. Exogenous kisspeptin enhances seasonal
reproductive function in male siberian hamsters. Funct. Ecol. 31:1220–1230.
doi:10.1111/1365-2435.12846.
Bailey, K.W., C.M. Jones, and A.J. Heinrichs. 2005. Economic returns to Holstein and Jersey
herds under multiple component pricing. J. Dairy Sci. 88:2269–2280.
doi:10.3168/jds.S0022-0302(05)72903-9.
Baldin, M., G.I. Zanton, and K.J. Harvatine. 2018. Effect of 2-hydroxy-4-(methylthio)butanoate
(HMTBa) on risk of biohydrogenation-induced milk fat depression. J. Dairy Sci. 101:376–
385. doi:10.3168/jds.2017-13446.
Balsalobre, A., S.A. Brown, L. Marcacci, F. Tronche, C. Kellendonk, H.M. Reichardt, G. Schutz,
and U. Schibler. 2000. Resetting of circadian time in peripheral tissues by glucocorticoid
signaling. Science (80-. ). 289:2344–2347. doi:10.1126/science.289.5488.2344.
Barnea, M., Z. Madar, and O. Froy. 2009. High-fat diet delays and fasting advances the circadian
expression of adiponectin signaling components in mouse liver. Endocrinology 150:161–
168. doi:10.1210/en.2008-0944.
Barrett, P., F.J.P. Ebling, S. Schuhler, D. Wilson, A.W. Ross, A. Warner, P. Jethwa, A. Boelen,
T.J. Visser, D.M. Ozanne, Z.A. Archer, J.G. Mercer, and P.J. Morgan. 2007. Hypothalamic
thyroid hormone catabolism acts as a gatekeeper for the seasonal control of body weight and
reproduction. Endocrinology 148:3608–3617. doi:10.1210/en.2007-0316.
Bartell, P.A., and E. Gwinner. 2005. A separate circadian oscillator controls nocturnal migratory
restlessness in the songbird Sylvia borin. J. Biol. Rhythms 20:538–549.
doi:10.1177/0748730405281826.
Bartzen‐Sprauer, J., P. Klosen, P. Ciofi, J.D. Mikkelsen, and V. Simonneaux. 2014.
Photoperiodic co‐regulation of kisspeptin, neurokinin B and dynorphin in the hypothalamus
of a seasonal rodent. J. Neuroendocrinol. 26:510–520. doi:10.1111/jne.12171.
Bauman, D.E., and W.B. Currie. 1980. Partitioning of nutrients during pregnancy and lactation: A
review of mechanisms involving homeostasis and homeorhesis. J. Dairy Sci. 63:1514–1529.
doi:10.3168/jds.S0022-0302(80)83111-0.
Bauman, D.E., and R.G. Vernon. 1993. Effects of exogenous bovine somatotropin on lactation.
Annu. Rev. Nutr. 13:437–461. doi:10.1146/annurev.nu.13.070193.002253.
Beasley, L.J., K.M. Pelz, and I. Zucker. 1984. Circannual rhythms of body weight in pallid bats.
Am. J. Physiol. 246:R955-8. doi:10.1152/ajpregu.1984.246.6.r955.
Beatty, D.T., A. Barnes, E. Taylor, and S.K. Maloney. 2008. Do changes in feed intake or
214
ambient temperature cause changes in cattle rumen temperature relative to core
temperature?. J. Therm. Biol. 33:12–19. doi:10.1016/j.jtherbio.2007.09.002.
Bellinger, L.L., V.E. Mendel, and G.P. Moberg. 1975. Circadian insulin, GH, prolactin,
corticosterone and glucose rhythms in fed and fasted rats. Horm. Metab. Res. 7:132–135.
doi:10.1055/s-0028-1093763.
Bilbo, S.D., F.S. Dhabhar, K. Viswanathan, A. Saul, S.M. Yellon, and R.J. Nelson. 2002. Short
day lengths augment stress-induced leukocyte trafficking and stress-induced enhancement
of skin immune function. Proc. Natl. Acad. Sci. U. S. A. 99:4067–72.
doi:10.1073/pnas.062001899.
Blum, I.D., Z. Patterson, R. Khazall, E.W. Lamont, M.W. Sleeman, T.L. Horvath, and A.
Abizaid. 2009. Reduced anticipatory locomotor responses to scheduled meals in ghrelin
receptor deficient mice. Neuroscience 164:351–359.
doi:10.1016/j.neuroscience.2009.08.009.
Bohmanova, J., I. Misztal, and J.B. Cole. 2007. Temperature-humidity indices as indicators of
milk production losses due to heat stress. J. Dairy Sci. 90:1947–1956. doi:10.3168/jds.2006-
513.
Bon, E., B. Breton, M.S. Govoroun, and F. Le Menn. 1999. Effects of accelerated photoperiod
regimes on the reproductive cycle of the female rainbow trout: II Seasonal variations of
plasma gonadotropins (GTH I and GTH II) levels correlated with ovarian follicle growth
and egg size. Fish Physiol. Biochem. 20:143–154. doi:http://dx.doi.org/10.1016/S0300-
9629(96)00415-X.
Bourdon, L., A. Buguet, M. Cucherat, and M.W. Radomski. 1995. Use of a spreadsheet program
for circadian analysis of biological/physiological data. Aviat. Sp. Environ. Med. 66:787–
791.
Bowen, A.J., and R.L. Reeves. 1967. Diurnal variation in glucose tolerance. Arch. Intern. Med.
119:261. doi:10.1001/archinte.1967.00290210093007.
Bradford, B.J., K.J. Harvatine, and M.S. Allen. 2008. Dietary unsaturated fatty acids increase
plasma glucagon-like peptide-1 and cholecystokinin and may decrease premeal ghrelin in
lactating dairy cows. J. Dairy Sci. 91:1443–1450. doi:10.3168/jds.2007-0670.
Branecky, K.L., K.D. Niswender, and J.S. Pendergast. 2015. Disruption of daily rhythms by high-
fat diet is reversible. PLoS One 10:1–12. doi:10.1371/journal.pone.0137970.
Brinklow, B.R., and A.S.I. Loudon. 1990. Development of seasonal rhythms in a long-lived
ungulate: the red deer (Cervus elaphus). J Interdiscip Cycle Res 21:173–175.
Brix, O., A. Bårdgard, S. Mathisen, N. Tyler, M. Nuutinen, S.G. Condo, and B. Giardina. 1990.
Oxygen transport in the blood of arctic mammals: Adaptation to local heterothermia. J.
Comp. Physiol. 159:655–60. doi:10.1007/bf00691710.
Brock, M.A. 1983. Seasonal rhythmicity in lymphocyte blastogenic responses of mice persists in
a constant environment. J. Immunol. 130:2586–2588.
Brown, S.A., G. Zumbrunn, F. Fleury-Olela, N. Preitner, and U. Schibler. 2002. Rhythms of
mammalian body temperature can sustain peripheral circadian clocks. Curr. Biol. 12:1574–
1583. doi:10.1016/S0960-9822(02)01145-4.
215
Buhr, E.D., S.-H. Yoo, and J.S. Takahashi. 2010. Temperature as a universal resetting cue for
mammalian circadian oscillators. Science (80-. ). 330:379–386.
doi:10.1126/science.1195262.
De Candolle, M. 1832. De l’influence de la lumiere sur les végétaux. Physiol. Vég 4:1069.
doi:10.3406/linly.1947.8346.
Carty, C.I., A.G. Fahey, M.R. Sheehy, S. Taylor, I.J. Lean, C.G. McAloon, L. O’Grady, and F.J.
Mulligan. 2017. The prevalence, temporal and spatial trends in bulk tank equivalent milk fat
depression in Irish milk recorded herds. Ir. Vet. J. 70:1–14. doi:10.1186/s13620-017-0092-
y.
Casey, T., O. Patel, K. Dykema, H. Dover, K. Furge, and K. Plaut. 2009. Molecular signatures
reveal circadian clocks may orchestrate the homeorhetic response to lactation. PLoS One
4:e7395. doi:10.1371/journal.pone.0007395.
Casey, T.M., J. Crodian, E. Erickson, K.K. Kuropatwinski, A.S. Gleiberman, M.P. Antoch, and
C.E.T. Al. 2014. Tissue-specific changes in molecular clocks during the transition from
pregnancy to lactation in mice. Biol. Reprod. 90:1–15. doi:10.1095/biolreprod.113.116137.
Casey, T.M., and K. Plaut. 2012. Lactation biology symposium: Circadian clocks as mediators of
the homeorhetic response to lactation. J. Anim. Sci. 90:744–754. doi:10.2527/jas.2011-
4590.
Castro, C.U. 2016. Hypothalamic concentration of kisspeptin and GnRH during breeding season
(BS) and non breeding season (NBS) in sheep. Colorado State University,.
Van Cauter, E., J.D. Blackman, D. Roland, J.-P. Spire, S. Refetoff, and K.S. Polonsky. 1991.
Modulation of glucose regulation and insulin secretion by circadian rhythmicity and sleep.
J. Clin. Invest. 88:934–942. doi:doi.org/10.1172/jci115396.
Van Cauter, E., K.S. Polonsky, and A.J. Scheen. 1997. Roles of circadian thythmicity and sleep
in human glucose regulation. Endocr. Rev. 18:716–738. doi:10.1210/edrv.18.5.0317.
Chen, L., and G. Yang. 2014. PPARs integrate the mammalian clock and energy metabolism.
PPAR Res. 2014:1–6. doi:10.1155/2014/653017.
Cheng, H.Y.M., J.W. Papp, O. Varlamova, H. Dziema, B. Russell, J.P. Curfman, T. Nakazawa,
K. Shimizu, H. Okamura, S. Impey, and K. Obrietan. 2007. microRNA modulation of
circadian-clock period and entrainment. Neuron 54:813–829.
doi:10.1016/j.neuron.2007.05.017.
Chew, B.P., P.V. Malven, R.E. Erb, C.N. Zamet, M.F. D’Amico, and V.F. Colenbrander. 1979.
Variables associated with peripartum traits in dairy cows. IV. Seasonal relationships among
temperature, photoperiod, and blood plasma prolactin. J. Dairy Sci. 62:1394–1398.
doi:10.3168/jds.S0022-0302(79)83435-9.
Chiang, J.Y.L. 2009. Bile acids: Regulation of synthesis. J. Lipid Res. 50:1955–1966.
doi:10.1194/jlr.R900010-JLR200.
Chou, T.C., T.E. Scammell, J.J. Gooley, S.E. Gaus, C.B. Saper, and J. Lu. 2003. Critical role of
dorsomedial hypothalamic nucleus in a wide range of behavioral circadian rhythms. J.
Neurosci. 23:10691–702. doi:10.1523/jneurosci.23-33-10691.2003.
216
Clark, J.H. 1975. Lactational responses to postruminal administration of proteins and amino
acids. J. Dairy Sci. 58:1178–1197. doi:10.3168/jds.s0022-0302(75)84696-0.
Clarke, I.J., J.T. Smith, A. Caraty, R.L. Goodman, and M.N. Lehman. 2009. Kisspeptin and
seasonality in sheep. Peptides 30:154–163. doi:10.1016/j.immuni.2010.12.017.Two-stage.
Concannon, P., K. Levac, R. Rawson, B. Tennant, and A. Bensadoun. 2001. Seasonal changes in
serum leptin, food intake, and body weight in photoentrained woodchucks. Am. J. Physiol.
Regul. Integr. Comp. Physiol. 281:R951-9. doi:10.1152/ajpregu.2001.281.3.r951.
Cornelissen, G. 2014. Cosinor-based rhythmometry. Theor. Biol. Med. Model. 11:1–24.
doi:10.1186/1742-4682-11-16.
Cowart, R.P., C.R. Boessen, and J.B. Kliebenstein. 1992. Patterns associated with season and
facilities for atrophic rhinitis and pneumonia in slaughter swine. J. Am. Vet. Med. Assoc.
200:190–193.
Crawford, H.M., D.E. Morin, E.H. Wall, T.B. McFadden, and G.E. Dahl. 2015. Evidence for a
role of prolactin in mediating effects of photoperiod during the dry period. Animals 5:803–
820. doi:10.3390/ani5030385.
Czeisler, C.A., J.F. Duffy, T.L. Shanahan, E.N. Brown, J.F. Mitchell, D.W. Rimmer, J.M. Ronda,
E.J. Silva, J.S. Allan, J.S. Emens, D. Dijk, and R.E. Kronauer. 1999. Stability, precision,
and near–24-hour period of the human circadian pacemaker. Science (80-. ). 284:2177–
2182. doi:10.1126/science.284.5423.2177.
Daan, S. 2000. Colin Pittendrigh, Jürgen Aschoff, and the natural entrainment of circadian
systems. J. Biol. Rhythms 15:195–207.
Dacheux, J.L., C. Pisselet, M.R. Blanc, M.-T. Hochereau-de-Reviers, and M. Courot. 1981.
Seasonal variations in rete testis fluid secretion and sperm production in different breeds of
ram. J. Reprod. Fertil. 61:363–371. doi:10.1530/jrf.0.0610363.
Dahl, G.E., B.A. Buchanan, and H.A. Tucker. 2000. Photoperiodic effects on dairy cattle: A
review. J. Dairy Sci. 83:885–93. doi:10.3168/jds.S0022-0302(00)74952-6.
Dahl, G.E., T.H. Elsasser, A. V Capuco, R.A. Erdman, and R.R. Peters. 1997. Effects of a long
daily photoperiod on milk yield and circulating concentrations of insulin-like growth factor-
I. J. Dairy Sci. 80:2784–2789. doi:10.3168/jds.S0022-0302(97)76241-6.
Damiola, F., N. Le Minh, N. Preitner, B. Kornmann, F. Fleury-Olela, and U. Schibler. 2000.
Restricted feeding uncouples circadian oscillators in peripheral tissues from the central
pacemaker in the suprachiasmatic nucleus. Genes Dev. 14:2950–61.
doi:10.1101/GAD.183500.
Dardente, H. 2015. Circannual biology: The double life of the seasonal thyrotroph. Curr. Biol.
25:R988–R991. doi:10.1016/j.cub.2015.09.002.
Davidson, A.J. 2009. Lesion studies targeting food-anticipatory activity. Eur. J. Neurosci.
30:1658–1664. doi:10.1111/j.1460-9568.2009.06961.x.
Davies, S.P., D. Carling, M.R. Munday, and D.G. Hardie. 1992. Diurnal rhythm of
phosphorylation of rat liver acetyl–CoA carboxylase by the AMP‐activated protein kinase,
demonstrated using freeze‐clamping: Effects of high fat diets. Eur. J. Biochem. 203:615–
217
623.
Dekleva, M.W., C.D. Dechow, J.M. Daubert, W.S. Liu, G.A. Varga, S. Bauck, and B.W.
Woodward. 2012. Short communication: Interactions of milk, fat, and protein yield
genotypes with herd feeding characteristics. J. Dairy Sci. 95:1559–1564.
doi:10.3168/jds.2011-4758.
DeVries, T.J., and M.A.G. von Keyserlingk. 2005. Time of feed delivery affects the feeding and
lying patterns of dairy cows. J. Dairy Sci. 88:625–631. doi:10.3168/jds.S0022-
0302(05)72726-0.
DeVries, T.J., M.A.G. von Keyserlingk, and K.A. Beauchemin. 2005. Frequency of feed delivery
affects the behavior of lactating dairy cows. J. Dairy Sci. 88:3553–3562.
doi:10.3168/jds.S0022-0302(05)73040-X.
Dickmeis, T. 2009. Glucocorticoids and the circadian clock. J. Endocrinol. 2000:3–22.
doi:10.1677/JOE-08-0415.
Doi, M., J. Hirayama, and P. Sassone-Corsi. 2006. Circadian regulator CLOCK is a histone
acetyltransferase. Cell 125:497–508. doi:10.1016/j.cell.2006.03.033.
Douris, N., S. Kojima, X. Pan, A.F. Lerch-gaggl, S.Q. Duong, M.M. Hussain, and C.B. Green.
2011. Nocturnin regulates circadian trafficking of dietary lipid in intestinal enterocytes.
Curr. Biol. 21:1347–1355. doi:10.1016/j.cub.2011.07.018.
Drouyer, E., C. Rieux, R.A. Hut, and H.M. Cooper. 2007. Responses of suprachiasmatic nucleus
neurons to light and dark adaptation: Relative contributions of melanopsin and rod–cone
Inputs. J. Neurosci. 27:9623–9631. doi:10.1523/JNEUROSCI.1391-07.2007.
Duchemin, S., H. Bovenhuis, W.M. Stoop, A.C. Bouwman, J.A.M. van Arendonk, and M.H.P.W.
Visker. 2013. Genetic correlation between composition of bovine milk fat in winter and
summer, and DGAT1 and SCD1 by season interactions. J. Dairy Sci. 96:592–604.
doi:10.3168/jds.2012-5454.
Dunshea, F.R., G.P. Walker, E. Ostrowska, and P.T. Doyle. 2008. Seasonal variation in the
concentrations of conjugated linoleic and trans fatty acids in milk fat from commercial dairy
farms is associated with pasture and grazing management and supplementary feeding
practices. Aust. J. Exp. Agric. 48:1062–1075. doi:10.1071/EA07286.
Dupré, S.M., K. Miedzinska, C. V. Duval, L. Yu, R.L. Goodman, G.A. Lincoln, J.R.E. Davis,
A.S. McNeilly, D.D. Burt, and A.S.I. Loudon. 2010. Identification of Eya3 and TAC1 as
Long-Day Signals in the Sheep Pituitary. Curr. Biol. 20:829–835.
doi:10.1016/j.cub.2010.02.066.
Edgar, R., E. Green, Y. Zhao, V.O. G, M. Olmedo, X. Qin, Y. Xu, M. Pan, U. Valekunja, K.
Feeney, E. Maywood, M. Hastings, N. Baliga, M. Merrow, A. Millar, C. Johnson, C.
Kyriacou, J. O’Neill, and A. Reddy. 2012. Peroxiredoxins are conserved markers of
circadian rhythms. Nature 485:459–464. doi:10.1038/nature11427.
Effenberger-Neidnicht, K., L. Brencher, M. Broecker-Preuss, T. Hamburger, F. Petrat, and H. De
Groot. 2014. Immune stimulation by exogenous melatonin during experimental
endotoxemia. Inflammation 37:738–744. doi:10.1007/s10753-013-9792-y.
Etchegaray, J., C. Lee, P. a Wade, and S.M. Reppert. 2003. Rhythmic histone acetylation
218
underlies transcription in the mammalian circadian clock. Nature 421:177–182.
doi:10.1038/nature01282.1.
Fan, L., P.N. Hsieh, D.R. Sweet, and M.K. Jain. 2018. Krüppel-like factor 15: Regulator of
BCAA metabolism and circadian protein rhythmicity. Pharmacol. Res. 130:123–126.
doi:10.1016/j.phrs.2017.12.018.
FAO. 2012. World agriculture towards 2030/2050: the 2012 revision.
Feigin, R.D., A.S. Klainer, and W.R. Beisel. 1968. Factors affecting circadian periodicity of
blood amino acids in man. Metabolism 17:764–775. doi:10.1016/0026-0495(68)90026-7.
Fievez, V., E. Colman, J.M. Castro-Montoya, I. Stefanov, and B. Vlaeminck. 2012. Milk odd-
and branched-chain fatty acids as biomarkers of rumen function-An update. Anim. Feed Sci.
Technol. 172:51–65. doi:10.1016/j.anifeedsci.2011.12.008.
Forsythe, W.C., E.J. Rykiel, R.S. Stahl, H. Wu, and R.M. Schoolfield. 1995. A model comparison
for daylength as a function of latitude and day of year. Ecol. Modell. 80:87–95.
doi:10.1016/0304-3800(94)00034-f.
Foster, R.G., and L.J. Kreitzman. 2009a. The generation of seasons. 1st ed. Yale University Press,
New Haven, CT, CT.
Foster, R.G., and L.J. Kreitzman. 2009b. Seasonal reproduction in mammals and birds. 1st ed.
Yale University Press, New Haven, CT.
Foster, R.G., I. Provencio, D. Hudson, S. Fiske, W. De Grip, and M. Menaker. 1991. Circadian
photoreception in the retinally degenerate mouse (rd/rd). J. Comp. Physiol. A 169:39–50.
doi:10.1007/BF00198171.
Gallego, M., and D.M. Virshup. 2007. Post-translational modifications regulate the ticking of the
circadian clock. Nat. Rev. Mol. Cell Biol. 8:139–148. doi:10.1038/nrm2106.
Gatfield, D., G. Le Martelot, C.E. Vejnar, D. Gerlach, O. Schaad, M. Oresic, C.C. Esau, F.
Fleury-olela, and A. Ruskeepa. 2009. Integration of microRNA miR-122 in hepatic
circadian gene expression. Genes Dev. 23:1313–1326. doi:10.1101/gad.1781009.anisms.
Geiser, F., and T. Ruf. 1995. Hibernation versus daily torpor in mammals and birds:
Physiological variables and classification of torpor patterns. Physiol. Zool. 68:935–966.
Gerlach, T., and J.E. Aurich. 2000. Regulation of seasonal reproductive activity in the stallion,
ram and hamster. Anim. Reprod. Sci. 58:197–213. doi:10.1016/S0378-4320(99)00093-7.
Giannetto, C., and G. Piccione. 2009. Daily rhythms of 25 physiological variables in Bos taurus
maintained under natural conditions. J. Appl. Biomed. 7:55–61.
doi:doi.org/10.32725/jab.2009.005.
Gil-Lozano, M., E.L. Mingomataj, W.K. Wu, S.A. Ridout, and P.L. Brubaker. 2014. Circadian
secretion of the intestinal hormone, glucagon-like peptide-1, by the rodent L-cell. Diabetes
1–46. doi:10.2337/db13-1501.
Gilbert, G.R., G.L. Hargrove, and M. Kroger. 1972. Diurnal variations in milk yield, fat yield,
milk fat percentage, and milk protein percentage of holstein-friesian cows. J. Dairy Sci.
56:409–410. doi:10.3168/jds.S0022-0302(73)85187-2.
219
Glatting, G., P. Kletting, S.N. Reske, K. Hohl, and C. Ring. 2007. Choosing the optimal fit
function: Comparison of the Akaike information criterion and the F-test. Med. Phys.
34:4285–4292. doi:10.1118/1.2794176.
Gnocchi, D., M. Pedrelli, E. Hurt-Camejo, and P. Parini. 2015. Lipids around the clock: Focus on
circadian rhythms and lipid metabolism. Biology (Basel). 4:104–132.
doi:10.3390/biology4010104.
Gómez-Brunet, A., J. Santiago-Moreno, A. del Campo, B. Malpaux, P. Chemineau, D.J.
Tortonese, A. Gonzalez-Bulnes, and A. López-Sebastián. 2008. Endogenous circannual
cycles of ovarian activity and changes in prolactin and melatonin secretion in wild and
domestic female sheep maintained under a long-day photoperiod. Biol. Reprod. 78:552–
562. doi:10.1095/biolreprod.107.064394.
Gooley, J.J., A. Schomer, and C.B. Saper. 2006. The dorsomedial hypothalamic nucleus is critical
for the expression of food-entrainable circadian rhythms. Nat. Neurosci. 9:398–407.
doi:10.1038/nn1651.
Grant, R.J., and J.L. Albright. 2001. Effect of animal grouping on feeding behavior and intake of
dairy cattle. J. Dairy Sci. 84:E156–E163. doi:10.3168/jds.S0022-0302(01)70210-X.
Grassly, N.C., C. Fraser, N.C. Grassly, and C. Fraser. 2006. Seasonal infectious disease
epidemiology. Proc. R. Soc. B Biol. Sci. 273:2541–2550. doi:10.1098/rspb.2006.3604.
Green, C.B. 2018. Circadian posttranscriptional regulatory mechanisms in mammals. Cold Spring
Harb. Perspect. Biol. 10:1–12. doi:10.1101/cshperspect.a030692.
Griinari, J.M., D.A. Dwyer, M.A. McGuire, D.E. Bauman, D.L. Palmquist, and K. V Nurmela.
1998. Trans-octadecenoic acids and milk fat depression in lactating dairy cows. J. Dairy Sci.
81:1251–61. doi:10.3168/jds.S0022-0302(98)75686-3.
Grobbelaar, N., T.C.C. Huang, H.Y.Y. Lin, and T.J.J. Chow. 1986. Dinitrogen-fixing endogenous
rhythm in Synechococcus RF-1. FEMS Microbiol. Lett. 37:173–177. doi:10.1111/j.1574-
6968.1986.tb01788.x.
Grosso, A.R., A.Q. Gomes, N.L. Barbosa-Morais, S. Caldeira, N.P. Thorne, G. Grech, M. von
Lindern, and M. Carmo-Fonseca. 2008. Tissue-specific splicing factor gene expression
signatures. Nucleic Acids Res. 36:4823–4832. doi:10.1093/nar/gkn463.
Guillaumond, F., H. Dardente, V. Giguère, and N. Cermakian. 2005. Differential control of
Bmal1 circadian transcription by REV-ERB and ROR nuclear receptors. J. Biol. Rhythms
20:391–403. doi:10.1177/0748730405277232.
Gwinner, E. 1968. Circannuale periodik als grundlage des jahreszeitlichen funktionswandels bei
zugvögeln. J. für Ornithol. 109:70–95.
Gwinner, E. 1981. Circannuale rhythmen bei tieren und ihre photoperiodische synchronisation.
Naturwissenschaften 68:542–551. doi:10.1007/BF00401662.
Gwinner, E. 1996. Circadian and circannual programmes in avian migration. J. Exp. Biol.
199:39–48. doi:10.1098/rspb.2009.2112.
Gwinner, E. 2003. Circannual rhythms in birds. Curr. Opin. Neurobiol. 770–778.
doi:10.1016/j.conb.2003.10.010.
220
Halberg, F., P.G. Albrecht, and C.P. Barnum. 1960. Phase shifting of liver-glycogen rhythm in
intact mice. Am. J. Physiol. 199:400–402. doi:10.1152/ajplegacy.1960.199.3.400.
Halberg, F., E. Halberg, C.P. Barnum, and J.J. Bittner. 1959. Physiologic 24-hour periodicity in
human beings and mice, the lighting regimen and daily routine. Photoperiod. Relat.
Phenom. plants Anim. 55:803–878.
Hansen, P.J. 1985. Seasonal modulation of puberty and the postpartum anestrus in cattle: A
review. Livest. Prod. Sci. 12:309–327. doi:10.1016/0301-6226(85)90131-9.
Hatori, M., C. Vollmers, A. Zarrinpar, L. DiTacchio, E.A. Bushong, S. Gill, M. Leblanc, A.
Chaix, M. Joens, J.A.J. Fitzpatrick, M.H. Ellisman, and S. Panda. 2012. Time-restricted
feeding without reducing caloric intake prevents metabolic diseases in mice fed a high-fat
diet. Cell Metab. 15:848–60. doi:10.1016/j.cmet.2012.04.019.
Hattar, S., H.-W. Liao, M. Takao, D.M. Berson, and K.-W.Y. Yau. 2002. Melanopsin-containing
retinal ganglion cells: Architecture, projections, and intrinsic photosensitivity. Science (80-.
). 295:1065–1071. doi:10.1126/science.1069609.
Hazlerigg, D., D. Lomet, G. Lincoln, and H. Dardente. 2018. Neuroendocrine correlates of the
critical day length response in the Soay sheep. J. Neuroendocrinol. e12631.
doi:10.1111/jne.12631.
Heath-Heckman, E.A.C. 2016. The metronome of symbiosis: Interactions between microbes and
the host circadian clock. Integr. Comp. Biol. 56:776–783. doi:10.1093/icb/icw067.
Hems, D.A., E.A. Rath, and T.R. Verrinder. 1975. Fatty acid synthesis in liver and adipose tissue
of normal and genetically obese (ob/ob) mice during the 24-hour cycle. Biochem. J.
150:167–173. doi:doi.org/10.1042/bj1500167.
Henry, B.A., D. Blache, F.R. Dunshea, and I.J. Clarke. 2010. Altered “set-point” of the
hypothalamus determines effects of cortisol on food intake, adiposity, and metabolic
substrates in sheep. Domest. Anim. Endocrinol. 38:46–56.
doi:10.1016/j.domaniend.2009.07.006.
Horst, R.L., T.A. Reinhardt, and G.S. Reddy. 2005. Vitamin D metabolism. Academic Press, Inc.:
New York.
Hu, L.Y., M.Z. Wang, J.L. Ouyang, P.F. Li, and J.J. Loor. 2017. Rapid Communication: Period2
gene silencing increases the synthesis of αs–casein protein in bovine mammary epithelial
cells. J. Anim. Sci. 95:4510–4513. doi:10.2527/jas2017.1938.
Huang, R.C. 2018. The discoveries of molecular mechanisms for the circadian rhythm: The 2017
Nobel Prize in Physiology or Medicine. Biomed. J. 41:5–8. doi:10.1016/j.bj.2018.02.003.
Hut, R.A., S. Paolucci, R. Dor, C.P. Kyriacou, and S. Daan. 2013. Latitudinal clines: An
evolutionary view on biological rhythms. Proc. R. Soc. London B Biol. Sci. 280:20130433.
doi:10.1098/rspb.2013.0433.
Inman, J.L., C. Robertson, J.D. Mott, and M.J. Bissell. 2015. Mammary gland development: Cell
fate specification, stem cells and the microenvironment. Development 142:1028–1042.
doi:10.1242/dev.087643.
Ivleva, N.B., T. Gao, A.C. LiWang, and S.S. Golden. 2006. Quinone sensing by the circadian
221
input kinase of the cyanobacterial circadian clock. Proc. Natl. Acad. Sci. 103:17468–17473.
doi:10.1073/pnas.0606639103.
Jeyaraj, D., F.A.J.L. Scheer, J.A. Ripperger, S.M. Haldar, Y. Lu, D.A. Prosdocimo, S.J. Eapen,
B.L. Eapen, Y. Cui, G.H. Mahabeleshwar, H.G. Lee, M.A. Smith, G. Casadesus, E.M.
Mintz, H. Sun, Y. Wang, K.M. Ramsey, J. Bass, S.A. Shea, U. Albrecht, and M.K. Jain.
2012. Klf15 orchestrates circadian nitrogen homeostasis. Cell Metab. 15:311–323.
doi:10.1016/j.cmet.2012.01.020.
Jilg, A., J. Moek, D.R. Weaver, H. Korf, J.H. Stehle, and C. Von Gall. 2005. Rhythms in clock
proteins in the mouse pars tuberalis depend on MT1 melatonin receptor signalling. Eur. J.
Neurosci. 22:2845–2854. doi:10.1111/j.1460-9568.2005.04485.x.
Johnston, J.D., B.B. Tournier, H. Andersson, M. Masson-Pévet, G.A. Lincoln, and D.G.
Hazlerigg. 2006. Multiple effects of melatonin on rhythmic clock gene expression in the
mammalian pars tuberalis. Endocrinology 147:959–965. doi:10.1210/en.2005-1100.
Jonker, J.S., R.A. Kohn, and R.A. Erdman. 1998. Using milk urea nitrogen to predict nitrogen
excretion and utilization efficiency in lactating dairy cows. J. Dairy Sci. 81:2681–2692.
doi:10.3168/jds.s0022-0302(98)75825-4.
Jordan, S.D., and K.A. Lamia. 2013. AMPK at the crossroads of circadian clocks and
metabolism. Mol. Cell. Endocrinol. 366:163–169. doi:10.1016/j.mce.2012.06.017.
Kadegowda, A.K.G., M. Bionaz, B. Thering, L.S. Piperova, R.A. Erdman, and J.J. Loor. 2009.
Identification of internal control genes for quantitative polymerase chain reaction in
mammary tissue of lactating cows receiving lipid supplements. J. Dairy Sci. 92:2007–2019.
doi:10.3168/jds.2008-1655.
Kageyama, H., T. Nishiwaki, M. Nakajima, H. Iwasaki, T. Oyama, and T. Kondo. 2006.
Cyanobacterial circadian pacemaker: Kai protein complex dynamics in the KaiC
phosphorylation cycle in vitro. Mol. Cell 23:161–171. doi:10.1016/j.molcel.2006.05.039.
Katz, F.H., and I.L. Shannon. 1969. Parotid fluid cortisol and cortisone. J. Clin. Invest. 48:848–
855. doi:https://doi.org/10.1172/JCI106042.
Kawai, M., C.B. Green, B. Lecka-czernik, N. Douris, M.R. Gilbert, and S. Kojima. 2010.
Adipogenesis by stimulating PPAR-γ nuclear translocation. Proc. Natl. Acad. Sci.
107:10508–10513. doi:10.1073/pnas.1000788107.
Kendall, P.E., and J.R. Webster. 2009. Season and physiological status affects the circadian body
temperature rhythm of dairy cows. Livest. Sci. 125:155–160.
doi:10.1016/j.livsci.2009.04.004.
Kessler, K., and O. Pivovarova-Ramich. 2019. Meal timing, aging, and metabolic health. Int. J.
Mol. Sci. 20:1911. doi:10.3390/ijms20081911.
Ki, Y., H. Ri, H. Lee, E. Yoo, J. Choe, and C. Lim. 2015. Warming up your tick-tock:
Temperature-dependent regulation of circadian clocks. Neuroscientist 21:503–518.
doi:10.1177/1073858415577083.
Kirsz, K., M. Szczesna, E. Molik, T. Misztal, A.K. Wojtowicz, and D.A. Zieba. 2012. Seasonal
changes in the interactions among leptin , ghrelin , and orexin in sheep. J. Anim. Sci.
90:2524–2531. doi:10.2527/jas2011-4463.
222
Klein, D.C., R. Smoot, J.L. Weller, S. Higa, S.P. Markey, G.J. Creed, and D.M. Jacobowitz.
1983. Lesions of the paraventricular nucleus area of the hypothalamus disrupt the
suprachiasmatic→ spinal cord circuit in the melatonin rhythm generating system. Brain Res.
Bull. 10:647–652. doi:10.1016/0361-9230(83)90033-3.
Kohn, R.A., M.M. Dinneen, and E. Russek-Cohen. 2005. Using blood urea nitrogen to predict
nitrogen excretion and efficiency of nitrogen utilization in cattle, sheep, goats, horses, pigs,
and rats. J. Anim. Sci. 83:879–889. doi:10.2527/2005.834879x.
Kohsaka, A., A.D. Laposky, K.M. Ramsey, C. Estrada, C. Joshu, Y. Kobayashi, F.W. Turek, and
J. Bass. 2007. High-fat diet disrupts behavioral and molecular circadian rhythms in mice.
Cell Metab. 6:414–421. doi:10.1016/j.cmet.2007.09.006.
Kondo, T., C.A. Strayer, R.D. Kulkarni, W. Taylor, M. Ishiura, S.S. Golden, and C.H. Johnson.
1993. Circadian rhythms in prokaryotes: Luciferase as a reporter of circadian gene
expression in cyanobacteria. Proc. Natl. Acad. Sci. USA 90:5672–6.
doi:10.1073/pnas.90.12.5672.
Kondo, T., N.F. Tsinoremas, S.S. Golden, C.H. Johnson, S. Kutsuna, and M. Ishiura. 1994.
Circadian clock mutants of cyanobacteria. Science (80-. ). 266:1233–1236.
doi:10.1126/science.7973706.
Konopka, R.J., and S. Benzer. 1971. Clock mutants of Drosophila melanogaster. Proc. Natl.
Acad. Sci. 68:2112–2116. doi:10.1073/pnas.68.9.2112.
Kraus, D. 2014. Consolidated data analysis and presentation using an open-source add-in for the
Microsoft Excel® spreadsheet software. Med. Writ. 23:25–28.
doi:10.1179/2047480613Z.000000000181.
Krieger, D.T., H. Hauser, and L.C. Krey. 1977. Suprachiasmatic nuclear lesions do not abolish
food-shifted circadian adrenal and temperature rhythmicity. Science (80-. ). 197:398–399.
Kuhn, N.J., D.T. Carrick, and C.J. Wilde. 1980. Lactose synthesis: The possibilities of regulation.
J. Dairy Sci. 63:328–336. doi:10.3168/jds.s0022-0302(80)82934-1.
Kuo, T., A. McQueen, T.-C. Chen, and J.-C. Wang. 2015. Regulation of Glucose Homeostasis by
Glucocorticoids. J.-C. Wang and C. Harris, ed. Springer New York, New York, NY.
Lacasse, P., V. Lollivier, F. Dessauge, R.M. Bruckmaier, S. Ollier, and M. Boutinaud. 2012. New
developments on the galactopoietic role of prolactin in dairy ruminants. Domest. Anim.
Endocrinol. 43:154–160. doi:10.1016/j.domaniend.2011.12.007.
Lacasse, P., C.M. Vinet, and D. Petitclerc. 2014. Effect of prepartum photoperiod and melatonin
feeding on milk production and prolactin concentration in dairy heifers and cows. J. Dairy
Sci. 97:3589–98. doi:10.3168/jds.2013-7615.
LaCount, D.W., J.K. Drackley, S.O. Laesch, and J.H. Clark. 1994. Secretion of oleic acid in milk
fat in response to abomasal infusions of canola or high oleic sunflower fatty acids. J. Dairy
Sci. 77:1372–1385. doi:10.3168/jds.s0022-0302(94)77076-4.
Lamia, K.A., U.M. Sachdeva, L. DiTacchio, E.C. Williams, J.G. Alvarez, D.F. Egan, D.S.
Vasquez, H. Juguilon, S. Panda, and R.J. Shaw. 2009. AMPK regulates the circadian clock
by cryptochrome phosphorylation and degradation. Science (80-. ). 326:437–440.
223
Landgraf, D., A.H. Tsang, A. Leliavski, C.E. Koch, J.L. Barclay, D.J. Drucker, and H. Oster.
2015. Oxyntomodulin regulates resetting of the liver circadian clock by food. Elife
4:e06253. doi:10.7554/eLife.06253.
Landry, G.J., B.A. Kent, D.F. Patton, M. Jaholkowski, E.G. Marchant, and R.E. Mistlberger.
2011. Evidence for time-of-day dependent effect of neurotoxic dorsomedial hypothalamic
lesions on food anticipatory circadian rhythms in rats. PLoS One 6:e24187.
doi:10.1371/journal.pone.0024187.
Landry, G.J., M.M. Simon, I.C. Webb, and R.E. Mistlberger. 2006. Persistence of a behavioral
food-anticipatory circadian rhythm following dorsomedial hypothalamic ablation in rats.
Am. J. Physiol. Integr. Comp. Physiol. 290:R1527–R1534.
doi:10.1152/ajpregu.00874.2005.
Landry, G.J., G.R. Yamakawa, I.C. Webb, R.J. Mear, and R.E. Mistlberger. 2007. The
dorsomedial hypothalamic nucleus is not necessary for the expression of circadian food-
anticipatory activity in rats. J. Biol. Rhythms 22:467–478. doi:10.1177/0748730407307804.
Lawrence, C.B., F. Celsi, J. Brennand, and S.M. Luckman. 2000. Alternative role for prolactin-
releasing peptide in the regulation of food intake. Nat. Neurosci. 3:645–646.
doi:10.1038/76597.
Lefcourt, A.M., J. Bitman, S. Kahl, and D.L. Wood. 1993. Circadian and ultradian rhythms of
peripheral cortisol concentrations in lactating dairy cows. J. Dairy Sci. 76:2607–2612.
doi:10.3168/jds.s0022-0302(93)77595-5.
Leone, V., S.M. Gibbons, K. Martinez, A.L. Hutchison, E.Y. Huang, C.M. Cham, J.F. Pierre,
A.F. Heneghan, A. Nadimpalli, N. Hubert, E. Zale, Y. Wang, Y. Huang, B. Theriault, A.R.
Dinner, M.W. Musch, K.A. Kudsk, B.J. Prendergast, J.A. Gilbert, and E.B. Chang. 2015.
Effects of diurnal variation of gut microbes and high-fat feeding on host circadian clock
function and metabolism. Cell Host Microbe 17:681–689. doi:10.1016/j.chom.2015.03.006.
LeSauter, J., N. Hoque, M. Weintraub, D.W. Pfaff, and R. Silver. 2009. Stomach ghrelin-
secreting cells as food-entrainable circadian clocks. Proc. Natl. Acad. Sci. U. S. A.
106:13582–7. doi:10.1073/pnas.0906426106.
Liang, X., and G.A. FitzGerald. 2017. Timing the microbes: The circadian rhythm of the gut
microbiome. J. Biol. Rhythms 32:505–515. doi:10.1177/0748730417729066.
Lim, C., and R. Allada. 2013. Emerging roles for post-transcriptional regulation in circadian
clocks. Nat. Neurosci. 16:1544–1550. doi:10.1038/nn.3543.
Lincoln, G. 2019. A brief history of circannual time. J. Neuroendocrinol. e12694.
doi:10.1111/jne.12694.
Lincoln, G., I. Clarke, R. Hut, and D. Hazlerigg. 2006. Characterizing a mammalian circannual
pacemaker. Sci. Mag. 314:1941–1944. doi:10.1126/science.1132009.
Lincoln, G.A. 1976. Seasonal variation in the episodic secretion of luteinizing hormone and
testosterone in the ram. J. Endocrinol. 69:213–26. doi:10.1677/JOE.0.0690213.
Lincoln, G.A. 1998. Reproductive seasonality and maturation throughout the complete life-cycle
in the mouflon ram (Ovis musimon). Anim. Reprod. Sci. 53:87–105. doi:10.1016/s0378-
4320(98)00129-8.
224
Lincoln, G.A., H. Anderson, and A. Loudon. 2003. Clock genes in calendar cells as the basis of
annual timekeeping in mammals - a unifying hypothesis. J. Endocrinol. 179:1–13.
doi:10.1677/joe.0.1790001.
Lincoln, G.A., J.D. Johnston, H. Andersson, G. Wagner, and D.G. Hazlerigg. 2005.
Photorefractoriness in mammals: Dissociating a seasonal timer from the circadian-based
photoperiod response. Endocrinology 146:3782–3790. doi:10.1210/en.2005-0132.
Liu, Y., W. Dong, J. Shao, Y. Wang, M. Zhou, and H. Sun. 2017. Branched-chain amino acid
negatively regulates KLF15 expression via PI3K-AKT pathway. Front. Physiol. 8:1–9.
doi:10.3389/fphys.2017.00853.
Locher, L.F., N. Meyer, E.-M. Weber, J. Rehage, U. Meyer, S. Dänicke, and K. Huber. 2011.
Hormone-sensitive lipase protein expression and extent of phosphorylation in subcutaneous
and retroperitoneal adipose tissues in the periparturient dairy cow. J. Dairy Sci. 94:4514–
4523. doi:10.3168/jds.2011-4145.
Lochmiller, R.L., and C. Deerenberg. 2000. Trade-offs in evolutionary immunology: Just what is
the cost of immunity?. Oikos 88:87–98. doi:doi:10.1034/j.1600-0706.2000.880110.x.
Lock, A.L., C. Tyburczy, D.A. Dwyer, K.J. Harvatine, F. Destaillats, Z. Mouloungui, L. Candy,
and D.E. Bauman. 2007. Trans-10 octadecenoic acid does not reduce milk fat synthesis in
dairy cows. J Nutr 137:71–6. doi:10.1093/jn/137.1.71.
Loudon, A.S.I., and J.D. Curlewis. 1988. Cycles of antler and testicular growth in an aseasonal
tropical deer (Axis axis). J. Reprod. Fertil. 729–738. doi:10.1530/jrf.0.0830729.
Ma, K., R. Xiao, H. Tseng, L. Shan, L. Fu, and D.D. Moore. 2009. Circadian dysregulation
disrupts bile acid homeostasis. PLoS One 4:e6843. doi:10.1371/journal.pone.0006843.
Ma, L., K.L. Cook, D.E. Bauman, and K.J. Harvatine. 2015. Short communication: Milk fat
depression induced by conjugated linoleic acid and a high-oil and low-fiber diet occurs
equally across the day in Holstein cows. J. Dairy Sci. 98:1851–1855. doi:10.3168/jds.2014-
8614.
Ma, L., Y. Ying, A.R. Clarke, P.A. Bartell, and K.J. Harvatine. 2013. Feeding entrainment of the
mammary circadian rhythm in FVB mice. J. Dairy Sci 96:154.
Ma, P. 2011. Searching for a circadian clock in Rhodopseudomonas palustris Strain TIE-1 by
oxygen entrainment. Vanderbilt University,.
Mackle, T.R., D.A. Dwyer, and D.E. Bauman. 1999. Effects of branched-chain amino acids and
sodium caseinate on milk protein concentration and yield from dairy cows. J. Dairy Sci.
82:161–171. doi:10.3168/jds.s0022-0302(99)75220-3.
Maestroni, G.J.M., A. Conti, and W. Pierpaoli. 1986. Role of the pineal gland in immunity:
Circadian synthesis and release of melatonin modulates the antibody response and
antagonizes the immunosuppressive effect of corticosterone. J. Neuroimmunol. 13:19–30.
de Mairan, J.J. d’Ortous. 1729. Observation botanique. Hist. l’Academie R. des Sci. 35–36.
Malpaux, B., J.E. Robinson, N.L. Wayne, and F.J. Karsch. 1989. Regulation of the onset of the
breeding season of the ewe: Importance of long days and of an endogenous reproductive
rhythm. J. Endocrinol. 122:269–78. doi:10.1677/JOE.0.1220269.
225
Maningat, P.D., P. Sen, M. Rijnkels, A.L. Sunehag, D.L. Hadsell, M. Bray, and M.W. Haymond.
2009. Gene expression in the human mammary epithelium during lactation: the milk fat
globule transcriptome. Physiol. Genomics 77030:12–22.
doi:10.1152/physiolgenomics.90341.2008.
Mann, D.R., M.A. Akinbami, K.G. Gould, and A.A. Ansari. 2000. Seasonal variations in
cytokine expression and cell-mediated immunity in male rhesus monkeys. Cell. Immunol.
200:105–115. doi:10.1006/cimm.2000.1623.
Manoogian, E.N.C., and S. Panda. 2017. Circadian rhythms, time-restricted feeding, and healthy
aging. Ageing Res. Rev. 39:59–67. doi:10.1016/j.arr.2016.12.006.
Le Martelot, G., T. Claudel, D. Gatfield, O. Schaad, B. Kornmann, G. Lo Sasso, A. Moschetta,
and U. Schibler. 2009. REV-ERBa participates in circadian SREBP signaling and bile acid
homeostasis. PLoS Biol. 7:1–12. doi:10.1371/journal.pbio.1000181.
Martinet, L., D. Allain, and C. Weiner. 1984. Role of prolactin in the photoperiodic control of
moulting in the mink (Mustela vison). J. Endocrinol. 103:9–15. doi:10.1677/joe.0.1030009.
Martinet, L., M. Mondain-Monval, and R. Monnerie. 1992. Endogenous circannual rhythms and
photorefractoriness of testis activity, moult and prolactin concentrations in mink (Mustela
vison). J. Reprod. Fertil. 95:325–338. doi:10.1530/jrf.0.0950325.
Maugeri, A., J.A. Oben, and M. Vinciguerra. 2018. Protein-rich or amino-acid only diets entrain
the liver clock: Time to scrap insulin?. EBioMedicine 28:9–10.
doi:10.1016/j.ebiom.2018.01.020.
Mayo, J.C., D.X. Tan, R.M. Sainz, M. Natarajan, S. Lopez-Burillo, and R.J. Reiter. 2003.
Protection against oxidative protein damage induced by metal-catalyzed reaction or
alkylperoxyl radicals: Comparative effects of melatonin and other antioxidants. Biochim.
Biophys. Acta - Gen. Subj. 1620:139–150. doi:10.1016/S0304-4165(02)00527-5.
McGlincy, N.J., A. Valomon, J.E. Chesham, E.S. Maywood, M.H. Hastings, and J. Ule. 2012.
Regulation of alternative splicing by the circadian clock and food related cues. Genome
Biol. 13. doi:10.1186/gb-2012-13-6-r54.
McMillen, I., G. Thorburn, and D. Walker. 1987. Diurnal variations in plasma concentrations of
cortisol, prolactin, growth hormones and glucose in the fetal sheep and pregnant ewe during
late gestation. J. Endocrinol. 114:65–72. doi:doi.org/10.1677/joe.0.1140065.
Mehta, N., and H.Y.M. Cheng. 2013. Micro-managing the circadian clock: The role of
microRNAs in biological timekeeping. J. Mol. Biol. 425:3609–3624.
doi:10.1016/j.jmb.2012.10.022.
Meng, M.S., G.C. West, and L. Irving. 1969. Fatty acid composition of caribou bone marrow.
Comp. Biochem. Physiol. 30:187–91. doi:10.1016/0010-406x(69)91314-0.
Menzi, W., and L.E. Chase. 1994. Feeding behavior of cows housed in free stall barns. Dairy
Syst. 21st Century. Am. Soc. Agric. Eng. St. Joseph, MI 829–831.
Metz, R.P., X. Qu, B. Laffin, D. Earnest, and W.W. Porter. 2006. Circadian clock and cell cycle
gene expression in mouse mammary epithelial cells and in the developing mouse mammary
gland. Dev. Dyn. 235:263–271. doi:10.1002/dvdy.20605.
226
Mieda, M., S.C. Williams, J.A. Richardson, K. Tanaka, and M. Yanagisawa. 2006. The
dorsomedial hypothalamic nucleus as a putative food-entrainable circadian pacemaker.
Proc. Natl. Acad. Sci. 103:12150–12155. doi:10.1073/pnas.0604189103.
Miller, A.R.E., R.A. Erdman, L.W. Douglass, and G.E. Dahl. 2000. Effects of photoperiodic
manipulation during the dry period of dairy cows. J. Dairy Sci. 83:962–967.
doi:http://dx.doi.org/10.3168/jds.S0022-0302(00)74960-5.
Miller, A.R.E., E.P. Stanisiewski, R.A. Erdman, L.W. Douglass, and G.E. Dahl. 1999. Effects of
long daily photoperiod and bovine somatotropin (Trobest®) on milk yield in cows. J. Dairy
Sci. 82:1716–1722. doi:10.3168/jds.S0022-0302(99)75401-9.
Milne, J.A., J.C. Macrae, A.M. Spence, and S. Wilson. 1978. A comparison of the voluntary
intake and digestion of a range of forages at different times of the year by the sheep and the
red deer (Cervus elaphus). Br. J. Nutr. 40:347–357. doi:10.1079/bjn19780131.
Min, H., H. Guo, and J. Xiong. 2005. Rhythmic gene expression in a purple photosynthetic
bacterium, Rhodobacter sphaeroides. FEBS Lett. 579:808–12.
doi:10.1016/j.febslet.2005.01.003.
Minami, Y., T. Kasukawa, Y. Kakazu, M. Iigo, M. Sugimoto, S. Ikeda, A. Yasui, G.T.J. van der
Horst, T. Soga, and H.R. Ueda. 2009. Measurement of internal body time by blood
metabolomics. Proc. Natl. Acad. Sci. 106:9890–9895. doi:10.1073/pnas.0900617106.
Minh, N. Le, F. Damiola, Ë. Tronche, and È. Schu. 2001. Glucocorticoid hormones inhibit food-
induced phase-shifting of peripheral circadian oscillators. EMBO J. 20:7128–7136.
doi:10.1093/emboj/20.24.7128.
Mistlberger, R.E. 1994. Circadian food-anticipatory activity: Formal models and physiological
mechanisms. Neurosci. Biobehav. Rev. 18:171–195. doi:10.1016/0149-7634(94)90023-X.
Mistlberger, R.E., and B. Rusak. 1988. Food-anticipatory circadian rhythms in rats with
paraventricular and lateral hypothalamic ablations. J. Biol. Rhythms 3:277–291.
doi:10.1177/074873048800300306.
Misztal, T., K. Romanowicz, and B. Barcikowski. 2002. Melatonin-a modulator of the GnRH/LH
axis in sheep. Reprod. Biol. 2:267–275.
Mohawk, J.A., C.B. Green, and J.S. Takahashi. 2012. Central and peripheral circadian clocks in
mammals. Annu. Rev. Neurosci. 35:445–462. doi:10.1146/annurev-neuro-060909-153128.
Moore-Ede, M. 1982. Sleeping as the world turns. Nat. Hist. 91:28–36.
Morgan, P.J., P. Barrett, H.E. Howell, and R. Helliwell. 1994. Melatonin receptors: localization,
molecular pharmacology and physiological significance. Neurochem. Int. 24:101–146.
doi:10.1016/0197-0186(94)90100-7.
Morgan, P.J., L.M. Williams, G. Davidson, W. Lawson, and E. Howell. 1989. Melatonin
receptors on ovine pars tuberalis: characterization and autoradiographicai localization. J.
Neuroendocrinol. 1:1–4. doi:10.1111/j.1365-2826.1989.tb00068.x.
Moriya, T., R. Aida, T. Kudo, M. Akiyama, M. Doi, N. Hayasaka, N. Nakahata, R. Mistlberger,
H. Okamura, and S. Shibata. 2009. The dorsomedial hypothalamic nucleus is not necessary
for food-anticipatory circadian rhythms of behavior, temperature or clock gene expression
227
in mice. Eur. J. Neurosci. 29:1447–1460. doi:10.1111/j.1460-9568.2009.06697.x.
Mujibi, F.D.N., S.S. Moore, D.J. Nkrumah, Z. Wang, and J.A. Basarab. 2010. Season of testing
and its effect on feed intake and efficiency in growing beef cattle. J. Anim. Sci. 88:3789–
3799. doi:10.2527/jas.2009-2407.
Munday, M.R., and D.H. Williamson. 1983. Diurnal variations in food intake and in lipogenesis
in mammary gland and liver of lactating rats. Biochem J 214:183–187.
doi:10.1042/bj2140183.
Murakami, M., P. Tognini, Y. Liu, K.L. Eckel‐Mahan, P. Baldi, and P. Sassone‐Corsi. 2016. Gut
microbiota directs PPARγ‐driven reprogramming of the liver circadian clock by nutritional
challenge. EMBO Rep. 17:1292–1303. doi:10.15252/embr.201642463.
Mustonen, A.-M., T. Pyykönen, J. Asikainen, S. Hänninen, J. Mononen, and P. Nieminen. 2005.
Circannual leptin and ghrelin levels of the blue fox (Alopex lagopus) in reference to
seasonal rhythms of body mass, adiposity, and food intake. J. Exp. Zool. Part A Comp. Exp.
Biol. 303A:26–36. doi:10.1002/jez.a.125.
Nagel, R., L. Clijsters, and R. Agami. 2009. The miRNA-192/194 cluster regulates the Period
gene family and the circadian clock. FEBS J. 276:5447–5455. doi:10.1111/j.1742-
4658.2009.07229.x.
Nakahata, Y., S. Sahar, G. Astarita, M. Kaluzova, and P. Sassone-Corsi. 2009. Circadian control
of the NAD+ salvage pathway by CLOCK-SIRT1. Science (80-. ). 324:654–657.
doi:10.1126/science.1170803.
Namias, J. 1976. Negative ocean–air feedback systems over the North Pacific in the transition
from warm to cold seasons. Mon. Weather Rev. 104:1107–1121. doi:10.1016/0146-
6291(77)90138-2.
Neufeld-Cohen, A., M.S. Robles, R. Aviram, G. Manella, Y. Adamovich, B. Ladeuix, D. Nir, L.
Rousso-Noori, Y. Kuperman, M. Golik, M. Mann, and G. Asher. 2016. Circadian control of
oscillations in mitochondrial rate-limiting enzymes and nutrient utilization by PERIOD
proteins. Proc. Natl. Acad. Sci. U. S. A. 1519650113-. doi:10.1073/pnas.1519650113.
Nikkhah, A., C.J. Furedi, A.D. Kennedy, G.H. Crow, and J.C. Plaizier. 2008. Effects of feed
delivery time on feed intake, milk production, and blood metabolites of dairy cows. J. Dairy
Sci. 91:4249–60. doi:10.3168/jds.2008-1075.
Nisembaum, L.G., N. de Pedro, M.J. Delgado, and E. Isorna. 2014. Crosstalking between the
“gut-brain” hormone ghrelin and the circadian system in the goldfish. Effects on clock gene
expression and food anticipatory activity. Gen. Comp. Endocrinol. 205:287–295.
doi:10.1016/j.ygcen.2014.03.016.
Niu, M., Y. Ying, P.A. Bartell, and K.J. Harvatine. 2014. The effects of feeding time on milk
production, total-tract digestibility, and daily rhythms of feeding behavior and plasma
metabolites and hormones in dairy cows. J. Dairy Sci. 97:7764–7776. doi:10.3168/jds.2014-
8261.
Niu, M., Y. Ying, P.A. Bartell, and K.J. Harvatine. 2017. The effects of feeding rations that differ
in fiber and fermentable starch within a day on milk production and the daily rhythm of feed
intake and plasma hormones and metabolites in dairy cows. J. Dairy Sci. 100:187–198.
228
doi:10.3168/jds.2016-11129.
O’Neill, J., and A. Reddy. 2011. Circadian clocks in human red blood cells. Nature 469:498–503.
doi:10.1038/nature09702.Circadian.
O’Neill, J.S., G. Van Ooijen, L.E. Dixon, C. Troein, F. Corellou, F.Y. Bouget, A.B. Reddy, and
A.J. Millar. 2011. Circadian rhythms persist without transcription in a eukaryote. Nature
469:554–558. doi:10.1038/nature09654.
Oike, H. 2017. Modulation of circadian clocks by nutrients and food factors. Biosci. Biotechnol.
Biochem. 81:863–870. doi:10.1080/09168451.2017.1281722.
Oishi, K., H. Shirai, and N. Ishida. 2005. CLOCK is involved in the circadian transactivation of
peroxisome-proliferator-activated receptor α (PPARα) in mice. Biochem J. 581:575–581.
doi:10.1042/bj20041150.
Ottesen, E.A., C.R. Young, S.M. Gifford, J.M. Eppley, R. Marin, S.C. Schuster, C.A. Scholin,
and E.F. DeLong. 2014. Multispecies diel transcriptional oscillations in open ocean
heterotrophic bacterial assemblages. Science 345:207–12. doi:10.1126/science.1252476.
Pan, X., and M.M. Hussain. 2007. Diurnal regulation of microsomal triglyceride transfer protein
and plasma lipid levels. J. Biol. Chem. 282:24707–24719. doi:10.1074/jbc.M701305200.
Pan, X., Y. Zhang, L. Wang, and M.M. Hussain. 2010. Diurnal regulation of MTP and plasma
triglyceride by CLOCK is mediated by SHP. Cell Metab. 12:174–186.
doi:10.1016/j.cmet.2010.05.014.
Panda, S. 2016. Circadian physiology of metabolism. Science (80-. ). 354:317–322.
doi:10.1126/science.aah4967.
Panda, S., I. Provencio, D.C. Tu, S.S. Pires, M.D. Rollag, A.M. Castrucci, M.T. Pletcher, T.K.
Sato, T. Wiltshire, M. Andahazy, S.A. Kay, R.N. Van Gelder, and J.B. Hogenesch. 2003.
Melanopsin is required for non–image-forming photic responses in blind mice. Science (80-.
). 301:525–528. doi:10.1126/science.1086179.
Panda, S., T.K. Sato, A.M. Castrucci, M.D. Rollag, W.J. Degrip, J.B. Hogenesch, I. Provencio,
and S.A. Kay. 2002. Melanopsin (Opn4) requirement for normal light-induced circadian
phase shifting. Science (80-. ). 298:2213–2217. doi:doi.org/10.1126/science.1076848.
Papazyan, R., Y. Zhang, and M.A. Lazar. 2016. Genetic and epigenomic mechanisms of
mammalian circadian transcription. Nat. Struct. Mol. Biol. 23:1045–1052.
doi:10.1038/nsmb.3324.
Partch, C.L., B. Green, Carla, and J.S. Takahashi. 2014. Molecular architecture of the mammalian
circadian clock. Trends Cell Biol. 24:90–99. doi:10.1016/j.pmrj.2014.02.014.Lumbar.
Paulose, J.K., J.M. Wright, A.G. Patel, and V.M. Cassone. 2016. Human gut bacteria are sensitive
to melatonin and express endogenous circadian rhythmicity. PLoS One 1–13.
doi:10.1371/journal.pone.0146643.
Peek, C.B., K.M. Ramsey, B. Marcheva, and J. Bass. 2012. Nutrient sensing and the circadian
clock. Trends Endocrinol. Metab. 23:312–318. doi:10.1016/j.tem.2012.02.003.
Pegoraro, M., A. Bafna, N.J. Davies, D.M. Shuker, and E. Tauber. 2016. DNA methylation
changes induced by long and short photoperiods in Nasonia. Genome Res. 26:203–10.
229
doi:10.1101/gr.196204.115.
Pengelley, E.T., S.J. Asmundson, B. Barnes, and R.C. Aloia. 1976. Relationship of light intensity
and photoperiod to circannual rhythmicity in the hibernating ground squirrel, Citellus
lateralis. Comp. Biochem. Physiol. 273–277. doi:10.1016/s0300-9629(76)80035-7.
Peters, R.R., L.T. Chapin, R.S. Emery, and H.A. Tucker. 1981. Milk yield, feed intake, prolactin,
growth hormone, and glucocorticoid response of cows to supplemented light. J. Dairy Sci.
64:1671–1678. doi:10.3168/jds.s0022-0302(81)82745-2.
Peters, R.R., L.T. Chapin, K.B. Leining, and H.A. Tucker. 1978. Supplemental lighting
stimulates growth and lactation in cattle. Science (80-. ). 199:911–912.
doi:10.1126/science.622576.
Petitclerc, D., R.R. Peters, L.T. Chapin, W.D. Oxender, K.R. Refsal, R.K. Braun, and H.A.
Tucker. 1983. Effect of blinding and pinealectomy on photoperiod and seasonal variations
in secretion of prolactin in cattle. Exp. Biol. Med. 174:205–211. doi:10.3181/00379727-
174-41726.
Petrosus, E., E.B. Silva, D.L. Jr, and S.D. Eicher. 2018. Effects of orally administered cortisol
and norepinephrine on weanling piglet gut microbial populations and Salmonella passage. J.
Anim. Sci. 96:1–9. doi:10.1093/jas/sky312.
Philo, R., and R.J. Reiter. 1980. A circannual rhythm in bovine pineal serotonin. Experientia
36:664–665. doi:10.1007/BF01970125.
Piccione, G., V. Messina, S. Scianó, A. Assenza, T. Orefice, I. Vazzana, and A. Zumbo. 2012.
Annual changes of some metabolical parameters in dairy cows in the Mediterranean area.
Vet. Arh. 82:229–238.
Piccione, G., M. Rizzo, S. Casella, S. Marafioti, G. Piccione, M. Rizzo, S. Casella, and S.
Marafioti. 2014. Application of the iButton® for measurement of the rumen temperature
circadian rhythms in lambs. Biol. Rhythm Res. 45:375–381.
doi:10.1080/09291016.2013.830507.
Plaut, K., D.E. Bauman, N. Agergaard, and R.M. Akers. 1987. Effect of exogenous prolactin
administration on lactational performance of dairy cows. Domest. Anim. Endocrinol. 4:279–
290. doi:10.1016/0739-7240(87)90024-5.
Plaut, K., and T. Casey. 2012. Does the circadian system regulate lactation?. Animal 6:394–402.
doi:10.1017/S1751731111002187.
Pozo, D., M. Delgado, J.M. Fernandez-Santos, J.R. Calvo, R.P. Gomariz, I. Martin-Lacave, G.G.
Ortiz, and J.M. Guerrero. 1997. Expression of the Mel1a-melatonin receptor mRNA in T
and B subsets of lymphocytes from rat thymus and spleen. FASEB J. 11:466–473.
doi:10.1096/fasebj.11.6.9194527.
Pozo, D., S. García‐Mauriño, J.M. Guerrero, and J.R. Calvo. 2004. mRNA expression of nuclear
receptor RZR/RORα, melatonin membrane receptor MT1, and hydroxindole‐O‐
methyltransferase in different populations of human immune cells. J. Pineal Res. 37:48–54.
doi:10.1111/j.1600-079x.2004.00135.x.
Preußner, M., G. Goldammer, A. Neumann, T. Haltenhof, P. Rautenstrauch, M. Müller-McNicoll,
and F. Heyd. 2017. Body temperature cycles control rhythmic alternative splicing in
230
mammals. Mol. Cell 67:433-446.e4. doi:10.1016/j.molcel.2017.06.006.
Qian, J., and F.A.J.L. Scheer. 2016. Circadian system and glucose metabolism: Implications for
physiology and disease. Trends Endocrinol. Metab. 27:282–293.
doi:10.1016/j.tem.2016.03.005.
Quist, M.A., S.J. LeBlanc, K.J. Hand, D. Lazenby, F. Miglior, and D.F. Kelton. 2008. Milking-
to-milking variability for milk yield, fat and protein percentage, and somatic cell count. J.
Dairy Sci. 91:3412–3423. doi:10.3168/jds.2007-0184.
R Development Core Team. 2013. R: A language and environment for statistical computing. R
Found. Stat. Comput. 1:409. doi:10.1007/978-3-540-74686-7.
Radostits, O.M., and I.R. Littlejohns. 1988. New concepts in the pathogenesis, diagnosis and
control of diseases caused by the bovine viral diarrhea virus. Can. Vet. J. 29:513–28.
Ramsey, K.M., J. Yoshino, C.S. Brace, D. Abrassart, Y. Kobayashi, B. Marcheva, H.-K. Hong,
J.L. Chong, E.D. Buhr, C. Lee, J.S. Takahashi, S. Imai, and J. Bass. 2009. Circadian clock
feedback cycle through NAMPT-mediated NAD+ biosynthesis. Science (80-. ). 324:651–
654. doi:10.1126/science.1171641.
Ray, D.E., and C.B. Roubicek. 1971. Behavior of feedlot cattle during two seasons. J. Anim. Sci.
33:72–76. doi:10.2527/jas1971.33172x.
Reddy, A.B., N.A. Karp, E.S. Maywood, E.A. Sage, M. Deery, J.S. O’Neill, G.K.Y. Wong, J.
Chesham, M. Odell, K.S. Lilley, C.P. Kyriacou, and M.H. Hastings. 2006. Circadian
orchestration of the hepatic proteome. Curr. Biol. 16:1107–1115.
doi:10.1016/j.cub.2006.04.026.
Regelin, W.L.., C.C. Schwartz, A.W. Franzmann, and C.C. Schwartz. 1985. Seasonal energy
metabolism of adult moose. Wildl. Soc. 49:388–393. doi:10.2307/3801539.
Reimann, F., and A.B. Reddy. 2014. Clocking up GLP-1: Considering intestinal rhythms in the
incretin effect. Diabetes 63:3584–3586. doi:10.2337/db14-0781.
Reksen, O., A. Tverdal, K. Landsverk, E. Kommisrud, K.E. Bøe, and E. Ropstad. 1999. Effects of
photointensity and photoperiod on milk yield and reproductive performance of Norwegian
Red cattle. J. Dairy Sci. 82:810–816. doi:http://dx.doi.org/10.3168/jds.S0022-
0302(99)75300-2.
Rensing, L., and P. Ruoff. 2002. Temeprature effect on entrainment, phase shifting and amplitude
of circadian clocks and its molecular bases. Chronobiol. Int. 19:807–864. doi:10.1081/CBI-
120014569.
Revel, F.G., M. Saboureau, M. Masson-Pévet, P. Pévet, J.D. Mikkelsen, and V. Simonneaux.
2006. Kisspeptin mediates the photoperiodic control of reproduction in hamsters. Curr. Biol.
16:1730–1735. doi:10.1016/j.cub.2006.07.025.
Rhee, S.G., H.Z. Chae, and K. Kim. 2005. Peroxiredoxins: A historical overview and speculative
preview of novel mechanisms and emerging concepts in cell signaling. Free Radic. Biol.
Med. 38:1543–1552.
Ribas-Latre, A., and K. Eckel-Mahan. 2016. Interdependence of nutrient metabolism and the
circadian clock system: Importance for metabolic health. Mol. Metab. 5:133–152.
231
doi:10.1016/j.molmet.2015.12.006.
Richter, C.P. 1922. A behavioristic study of the activity of the rat. W.S. Hunter, ed.
Rico, D.E., and K.J. Harvatine. 2013. Induction of and recovery from milk fat depression occurs
progressively in dairy cows switched between diets that differ in fiber and oil concentration.
J. Dairy Sci. 96:6621–30. doi:10.3168/jds.2013-6820.
Robinson, J.E., and F.J. Karsch. 1984. Refractoriness to inductive day lengths terminates the
breeding season of the Suffolk ewe. Biol. Reprod. 31:656–663.
doi:10.1095/biolreprod31.4.656.
Robinson, T. 1951. Reproduction in the ewe. Biol. Rev. 26:121–157. doi:10.1111/j.1469-
185X.1951.tb00644.x.
Robles, M.S., J. Cox, and M. Mann. 2014. In-vivo quantitative proteomics reveals a key
contribution of post-transcriptional mechanisms to the circadian regulation of liver
metabolism. PLoS Genet. 10:e1004047. doi:10.1371/journal.pgen.1004047.
Robles, M.S., S.J. Humphrey, and M. Mann. 2017. Phosphorylation is a central mechanism for
circadian control of metabolism and physiology. Cell Metab. 25:118–127.
doi:10.1016/j.cmet.2016.10.004.
Roenneberg, T., and J. Aschoff. 1990. Annual rhythm of human reproduction: environmental
coorelations. J. Biol. Rhythms 5:217–239. doi:10.1177/074873049000500304.
Rosenwasser, A.M., Z. Boulos, and M. Terman. 1981. Circadian organization of food intake and
meal patterns in the rat. Physiol. Behav. 27:33–39. doi:10.1016/0031-9384(81)90296-1.
Rottman, L.W., Y. Ying, K. Zhou, P. a. Bartell, and K.J. Harvatine. 2015. The effects of feeding
rations that differ in neutral detergent fiber and starch concentration within a day on
production, feeding behavior, total-tract digestibility, and plasma metabolites and hormones
in dairy cows. J. Dairy Sci. 98:4673–4684. doi:10.3168/jds.2014-8859.
Rottman, L.W., Y. Ying, K. Zhou, P.A. Bartell, and K.J. Harvatine. 2014. The daily rhythm of
milk synthesis is dependent on the timing of feed intake in dairy cows. Physiol. Rep. 2:1–
12. doi:10.14814/phy2.12049.
Rousseau, K., Z. Atcha, and A.S.I. Loudon. 2003. Leptin and seasonal mammals. J.
Neuroendocrinol. 15:409–14. doi:10.1046/j.1365-2826.2003.01007.x.
Rudic, R.D., P. McNamara, A.-M.M. Curtis, R.C. Boston, S. Panda, J.B. Hogenesch, G.A.
FitzGerald, N.R. Biermasz, K.W. van Dijk, J.A. Romijn, and J.H. Meijer. 2004. BMAL1
and CLOCK, two essential components of the circadian clock, are involved in glucose
homeostasis. PLoS Biol. 2:e377. doi:10.1371/journal.pbio.0020377.
Sadacca, L.A., K.A. Lamia, A.S. DeLemos, B. Blum, and C.J. Weitz. 2011. An intrinsic circadian
clock of the pancreas is required for normal insulin release and glucose homeostasis in mice.
Diabetologia 54:120–124. doi:10.1007/s00125-010-1920-8.
Saini, C., J. Morf, M. Stratmann, P. Gos, and U. Schibler. 2012. Simulated body temperature
rhythms reveal the phase-shifting behavior and plasticity of mammalian circadian
oscillators. Genes Dev. 26:567–580. doi:10.1101/gad.183251.111.
Salfer, I.J., C.D. Dechow, and K.J. Harvatine. 2019. Annual rhythms of milk and milk fat and
232
protein production in dairy cattle in the United States. J. Dairy Sci. 102:742–753.
doi:10.3168/jds.2018-15040.
Salfer, I.J., and K.J. Harvatine. 2018. 480 The effect of night restricted feeding on the molecular
circadian clock of the mammary gland. J. Dairy Sci. 101:405.
Salfer, I.J., J.Y. Ying, and K.J. Harvatine. 2016. 1512 The timing of feed availability entrains the
circadian rhythm of milk synthesis in dairy cattle. J. Anim. Sci. 94:734.
Scheer, F.A.J.L., C.J. Morris, and S.A. Shea. 2013. The internal circadian clock increases hunger
and appetite in the evening independent of food intake and other behaviors. Obesity 21:421–
423. doi:10.1002/oby.20351.
Schennink, A., W.M. Stoop, M.H.P.W. Visker, J.M.L. Heck, H. Bovenhuis, J.J. Van Der Poel,
H.J.F. Van Valenberg, and J.A.M. Van Arendonk. 2007. DGAT1 underlies large genetic
variation in milk-fat composition of dairy cows. Anim. Genet. 38:467–473.
doi:10.1111/j.1365-2052.2007.01635.x.
Schibler, U., and P. Sassone‐Corsi. 2002. A web a circadian pacemakers. Cell 111:919–922.
doi:10.1053/j.jfas.2016.11.020.
Schingoethe, D.J. 1996. Dietary influence on protein level in milk and milk yield in dairy cows.
Anim. Feed Sci. Technol. 60:181–190. doi:10.1016/0377-8401(96)00975-3.
Scholander, P.F., R. Hock, V. Walters, and L. Irving. 1950. Adaptation to cold in arctic and
tropical mammals and birds in relation to body temperature, insulation, and basal metabolic
rate. Biol. Bull. 99:259–271. doi:10.2307/1538742.
Schwartz, C., and M.T. Andrews. 2013. Circannual transitions in gene expression: Lessons from
seasonal adaptations. Curr. Top. Dev. Biol. 105:247–73. doi:10.1016/B978-0-12-396968-
2.00009-9.
Sellers, P.J., R.E. Dickinson, D.A. Randall, A.K. Betts, F.G. Hall, J.A. Berry, G.J. Collatz, A.S.
Denning, H.A. Mooney, and C.A. Nobre. 1997. Modeling the exchanges of energy, water,
and carbon between continents and the atmosphere. Science (80-. ). 275:502–509.
doi:10.1126/science.275.5299.502.
Seltman, H. 1997. A primer on using SAS mixed models to analyze biorhythm data 1–11.
Shea, S.A., M.F. Hilton, C. Orlova, R. Timothy Ayers, and C.S. Mantzoros. 2005. Independent
circadian and sleep/wake regulation of adipokines and glucose in humans. J. Clin.
Endocrinol. Metab. 90:2537–2544. doi:10.1210/jc.2004-2232.
Shefer, S., S. Hauser, V. Lapar, and E.H. Mosbach. 1972. Diurnal variation of HMG CoA
reductase activity in rat intestine. J. Lipid Res. 13:571–573.
Shelton, M., and J.T. Morrow. 1965. Effect of season on reproduction of Rambouillet ewes. J.
Anim. Sci. 24:795–799. doi:10.2527/jas1965.243795x.
Sherman, H., Y. Genzer, R. Cohen, N. Chapnik, Z. Madar, and O. Froy. 2012. Timed high-fat
diet resets circadian metabolism and prevents obesity. FASEB J. 26:3493–3502.
doi:10.1096/fj.12-208868.
Shi, S.Q., T.S. Ansari, O.P. McGuinness, D.H. Wasserman, and C.H. Johnson. 2013. Circadian
disruption leads to insulin resistance and obesity. Curr. Biol. 23:372–381.
233
doi:10.1016/j.cub.2013.01.048.
Shimizu, H., F. Hanzawa, D. Kim, S. Sun, T. Laurent, M. Umeki, S. Ikeda, S. Mochizuki, and H.
Oda. 2018. Delayed first active-phase meal, a breakfast- skipping model, led to increased
body weight and shifted the circadian oscillation of the hepatic clock and lipid metabolism-
related genes in rats fed a high-fat diet. PLoS One 13:1–17.
doi:10.1371/journal.pone.0206669.
Shostak, A., J. Meyer-Kovac, and H. Oster. 2013. Circadian regulation of lipid mobilization in
white adipose tissues. Diabetes 62:2195–2203. doi:10.2337/db12-1449.
Simon, C., G. Brandenberger, and M. Follenius. 1987. Ultradian oscillations of plasma glucose,
insulin, and C-peptide in man during continuous enteral nutrition. J. Clin. Endocrinol.
Metab. 64:669–674. doi:10.1210/jcem-64-4-669.
Simonneaux, V., and C. Ribelayga. 2003. Generation of the melatonin endocrine message in
mammals: A review of the complex regulation of melatonin synthesis by norepinephrine,
peptides, and other pineal transmitters. Pharmacol. Rev. 55:325–95. doi:10.1124/pr.55.2.2.
So, A.Y.-L., T.U. Bernal, M.L. Pillsbury, K.R. Yamamoto, and B.J. Feldman. 2009.
Glucocorticoid regulation of the circadian clock modulates glucose homeostasis. Proc. Natl.
Acad. Sci. U. S. A. 106:17582–7. doi:10.1073/pnas.0909733106.
Van Soest, P.J. Van. 1994. Nutritional Ecology of the Ruminant. 2nd ed. Cornell University
Press, Ithaca, NY.
Van Soest, P.J. van, J.B. Robertson, and B.A. Lewis. 1991. Methods for dietary fiber, neutral
detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci.
74:3583–3597. doi:10.3168/jds.s0022-0302(91)78551-2.
Sothern, R.B., M.S. Farber, and S.A. Gruber. 1993. Circannual variations in baseline blood values
of dogs. Chronobiol. Int. 10:364–382. doi:10.3109/07420529309064491.
Stanisiewski, E.P., L.T. Chapin, N.K. Ames, S.A. Zinn, and H.A. Tucker. 1988. Melatonin and
prolactin concentrations in blood of cattle exposed to 8, 16 or 24 hours of daily light. J.
Anim. Sci. 66:727–734. doi:10.2527/jas1988.663727x.
Stephan, F.K. 2002. The “other” circadian system: Food as a zeitgeber. J. Biol. Rhythms 17:284–
292. doi:10.1177/074873040201700402.
Stevenson, T.J., and G.A. Lincoln. 2017. Epigenetic mechanisms regulating circannual rhythms.
V. Kumar, ed. Springer India, Dehli, India, India.
Stevenson, T.J., K.G. Onishi, S.P. Bradley, and B.J. Prendergast. 2014. Cell-autonomous
iodothyronine deiodinase expression mediates seasonal plasticity in immune function.
Brain. Behav. Immun. 36:61–70. doi:10.1016/j.bbi.2013.10.008.
Stevenson, T.J., and B.J. Prendergast. 2013. Reversible DNA methylation regulates seasonal
photoperiodic time measurement. Proc. Natl. Acad. Sci. U. S. A. 110:16651–6.
doi:10.1073/pnas.1310643110.
Stokkan, K.-A., S. Yamazaki, H. Tei, Y. Sakaki, and M. Menaker. 2001. Entrainment of the
circadian clock in the liver by feeding. Science 291:490–3.
doi:10.1126/science.291.5503.490.
234
Stoney, P.N., D. Rodrigues, G. Helfer, T. Khatib, A. Ashton, E.A. Hay, R. Starr, D. Kociszewska,
P. Morgan, and P. McCaffery. 2017. A seasonal switch in histone deacetylase gene
expression in the hypothalamus and their capacity to modulate nuclear signaling pathways.
Brain. Behav. Immun. 61:340–352. doi:10.1016/j.bbi.2016.12.013.
Stubblefield, J.J. 2012. Nocturnin: at the crossroads of clocks and metabolism. Trends
Endocrinol. Metab. 23:326–333. doi:10.1016/j.tem.2012.03.007.
Stumpf, W.E., and T.H. Privette. 1991. The steroid hormone of sunlight soltriol (vitamin D) as a
seasonal regulator of biological activities and photoperiodic rhythms. J. Steroid Biochem.
Mol. Biol. 39:283–289. doi:10.1016/0960-0760(91)90074-F.
Suárez-Trujillo, A., and T.M. Casey. 2016. Serotoninergic and circadian systems: Driving
mammary gland development and function. Front. Physiol. 7:1–15.
doi:10.3389/fphys.2016.00301.
Swan, J.A., S.S. Golden, A. LiWang, and C.L. Partch. 2018. Structure, function, and mechanism
of the core circadian clock in cyanobacteria. J. Biol. Chem. 293:5026–5034.
doi:10.1074/jbc.TM117.001433.
Takahashi, J.S. 2017. Transcriptional architecture of the mammalian circadian clock. Nat. Rev.
Genet. 18:164–179. doi:10.1038/nrg.2016.150.
Takai, N., M. Nakajima, T. Oyama, R. Kito, C. Sugita, M. Sugita, T. Kondo, and H. Iwasaki.
2006. A KaiC-associating SasA-RpaA two-component regulatory system as a major
circadian timing mediator in cyanobacteria. Proc. Natl. Acad. Sci. U. S. A. 103:12109–
12114. doi:10.1073/pnas.0602955103.
Tamerius, J.D., J. Shaman, W.J. Alonso, K. Bloom-Feshbach, C.K. Uejio, A. Comrie, and C.
Viboud. 2013. Environmental predictors of seasonal influenza epidemics across temperate
and tropical climates. PLOS Pathog. 9:e1003194. doi:10.1371/journal.ppat.1003194.
Teyssot, B., L.-M. Houdebine, and J. Djiane. 1981. Prolactin induces release of a factor from
membranes capable of stimulating β-casein gene transcription in isolated mammary cell
nuclei. Proc. Natl. Acad. Sci. 78:6729–6733. doi:doi.org/10.1073/pnas.78.11.6729.
Thaiss, C.A., M. Levy, T. Korem, L. Dohnalová, H. Shapiro, D.A. Jaitin, E. David, D.R. Winter,
M. Gury-BenAri, E. Tatirovsky, T. Tuganbaev, S. Federici, N. Zmora, D. Zeevi, M. Dori-
Bachash, M. Pevsner-Fischer, E. Kartvelishvily, A. Brandis, A. Harmelin, O. Shibolet, Z.
Halpern, K. Honda, I. Amit, E. Segal, and E. Elinav. 2016. Microbiota diurnal rhythmicity
programs host transcriptome oscillations. Cell 167:1495-1510.e12.
doi:10.1016/j.cell.2016.11.003.
Thaiss, C.A., D. Zeevi, M. Levy, G. Zilberman-Schapira, J. Suez, A.C. Tengeler, L. Abramson,
M.N. Katz, T. Korem, N. Zmora, Y. Kuperman, I. Biton, S. Gilad, A. Harmelin, H. Shapiro,
Z. Halpern, E. Segal, and E. Elinav. 2014. Transkingdom control of microbiota diurnal
oscillations promotes metabolic homeostasis. Cell 159:514–529.
doi:10.1016/j.cell.2014.09.048.
Tong, X., D. Zhang, B. Arthurs, P. Li, L. Durudogan, N. Gupta, and L. Yin. 2015. Palmitate
inhibits SIRT1-dependent BMAL1/CLOCK interaction and disrupts circadian gene
oscillations in hepatocytes. PLoS One 10:1–20. doi:10.1371/journal.pone.0130047.
235
Tosi, G.M. 2004. Milk in the western marketing area; Termination of the order. Federal Register,
Washington, DC, United State of America.
Tucker, H.A., D. Petitclerc, and S.A. Zinn. 1984. The influence of photoperiod on body weight
gain, body composition, nutrient intake and hormone secretion. J. Anim. Sci. 59:1610–1620.
doi:10.2527/jas1984.5961610x.
Ueda, K., T. Mitani, and S. Kondo. 2016. Relationship of rumen fill and fermentation to diurnal
and seasonal variation of herbage intake in dairy cows grazed on perennial ryegrass pasture.
Anim. Sci. J. 87:1148–1156. doi:10.1111/asj.12563.
Urrutia, N., and K.J. Harvatine. 2017. Effect of conjugated linoleic acid and acetate on milk fat
synthesis and adipose lipogenesis in lactating dairy cows. J. Dairy Sci. 100:5792–5804.
doi:10.3168/jds.2016-12369.
Vallimont, J.E.E., C.D.D. Dechow, J.M.M. Daubert, M.W.W. Dekleva, J.W.W. Blum, C.M.M.
Barlieb, W. Liu, G.A.A. Varga, A.J.J. Heinrichs, and C.R.R. Baumrucker. 2010. Genetic
parameters of feed intake, production, body weight, body condition score, and selected type
traits of Holstein cows in commercial tie-stall barns. J. Dairy Sci. 93:4892–4901.
doi:http://dx.doi.org/10.3168/jds.2010-3189.
VanderLeest, H.T., T. Houben, S. Michel, T. Deboer, H. Albus, M.J. Vansteensel, G.D. Block,
and J.H. Meijer. 2007. Seasonal encoding by the circadian pacemaker of the SCN. Curr.
Biol. 17:468–473. doi:10.1016/j.cub.2007.01.048.
Vetter, C., and F.A.J.L. Scheer. 2017. Circadian biology: Uncoupling human body clocks by food
timing. Curr. Biol. 27:R656–R658. doi:10.1016/j.cub.2017.05.057.
Vielhaber, E., E. Eide, A. Rivers, Z.-H. Gao, and D.M. Virshup. 2000. Nuclear entry of the
circadian regulator mPER1 is controlled by mammalian casein kinase Iε. Mol. Cell. Biol.
20:4888–4899. doi:10.1128/MCB.20.13.4888-4899.2000.
Vlaeminck, B., V. Fievez, A.R.J. Cabrita, A.J.M. Fonseca, and R.J. Dewhurst. 2006. Factors
affecting odd- and branched-chain fatty acids in milk: A review. Anim. Feed Sci. Technol.
131:389–417. doi:10.1016/j.anifeedsci.2006.06.017.
Wall, E.H., T.L. Auchtung, G.E. Dahl, S.E. Ellis, and T.B. McFadden. 2005. Exposure to short
day photoperiod during the dry period enhances mammary growth in dairy cows. J. Dairy
Sci. 88:1994–2003. doi:10.3168/jds.S0022-0302(05)72875-7.
Watanabe, M., S. Yasuo, T. Watanabe, T. Yamamura, N. Nakao, S. Ebihara, and T. Yoshimura.
2004. Photoperiodic regulation of type 2 deiodinase gene in djungarian hamster: Possible
homologies between avian and mammalian photoperiodic regulation of reproduction.
Endocrinology 145:1546–1549. doi:10.1210/en.2003-1593.
Went, F. 2006. Chronobiometry: Analyzing for Rhythms. 1st ed. W.L. Koukkari and R.B.
Sothern, ed. Springer Netherlands, Dordrecht.
West, J.W. 2003. Effects of heat-stress on production in dairy cattle. J. Dairy Sci. 86:2131–2144.
doi:10.3168/jds.S0022-0302(03)73803-X.
Williams, S.B., I. Vakonakis, S.S. Golden, and A.C. LiWang. 2002. Structure and function from
the circadian clock protein KaiA of Synechococcus elongatus: A potential clock input
mechanism. Proc. Natl. Acad. Sci. 99:15357–15362. doi:10.1073/pnas.232517099.
236
Winter, A., W. Krämer, F.A.O. Werner, S. Kollers, S. Kata, G. Durstewitz, J. Buitkamp, J.E.
Womack, G. Thaller, and R. Fries. 2002. Association of a lysine-232/alanine polymorphism
in a bovine gene encoding acyl-CoA:diacylglycerol acyltransferase (DGAT1) with variation
at a quantitative trait locus for milk fat content. Proc. Natl. Acad. Sci. U. S. A. 99:9300–5.
doi:10.1073/pnas.142293799.
Wood, P.D.P. 1970. A note on the repeatability of parameters of the lactation curve in cattle.
Anim. Prod. 12:535–538. doi:10.1017/S0003356100029135.
Wood, P.D.P. 1976. Algebraic models of lactation curves for milk, fat and protein production,
with estimates of seasonal variation. Anim. Prod. 22:35–40.
doi:10.1017/s000335610003539x.
Wood, S., and A. Loudon. 2018. The pars tuberalis: The site of the circannual clock in
mammals?. Gen. Comp. Endocrinol. 258:222–235. doi:10.1016/j.ygcen.2017.06.029.
Woodfill, C.J., J.E. Robinson, B. Malpaux, and F.J. Karsch. 1991. Synchronization of the
circannual reproductive rhythm of the ewe by discrete photoperiodic signals. Biol. Reprod.
45:110–121. doi:10.1095/biolreprod45.1.110.
Yang, Z., S. Liu, X. Chen, H. Chen, M. Huang, and J. Zheng. 2000. Induction of apoptotic cell
death and in vivo growth inhibition of human cancer cells by a saturated branched-chain
fatty acid, 13-methyltetradecanoic acid. Cancer Res. 60:505–509.
Yellon, S.M., O.R. Fagoaga, and S.L. Nehlsen-Cannarella. 1999. Influence of photoperiod on
immune cell functions in the male Siberian hamster. Am. J. Physiol. 276:R97–R102.
doi:10.1152/ajpregu.1999.276.1.r97.
Yerushalmi, L., and B. Volesky. 1989. Circadian rhythmicity in a fermentation process?. Appl.
Microbiol. Biotechnol. 30:460–474. doi:10.1007/bf00263850.
Yin, L., N. Wu, J.C. Curtin, M. Qatanani, N.R. Szwergold, R.A. Reid, G.M. Waitt, D.J. Parks,
K.H. Pearce, and G.B. Wisely. 2007. Rev-erbα, a heme sensor that coordinates metabolic
and circadian pathways. Science (80-. ). 318:1786–1789. doi:10.1126/science.1150179.
Ying, Y., L.W. Rottman, C. Crawford, P.A. Bartell, and K.J. Harvatine. 2015. The effects of
feeding rations that differ in neutral detergent fiber and starch concentration within a day on
rumen digesta nutrient concentration, pH, and fermentation products in dairy cows. J. Dairy
Sci.. doi:10.3168/jds.2014-8873.
Yoshimura, T. 2013. Thyroid hormone and seasonal regulation of reproduction. Front.
Neuroendocrinol. 34:157–166. doi:10.1016/j.yfrne.2013.04.002.
Zarrinpar, A., A. Chaix, S. Yooseph, and S. Panda. 2014. Diet and feeding pattern affect the
diurnal dynamics of the gut microbiome. Cell Metab. 20:1006–1017.
doi:10.1016/j.cmet.2014.11.008.
Zucker, I., and M. Boshes. 1982. Circannual body weight rhythms of ground squirrels: role of
gonadal hormones. Am. J. Physiol. 243:R546–R551. doi:10.1152/ajpregu.1982.243.5.r546.
VITA
Isaac James Salfer was born on February 6, 1991 in Maplewood, Minnesota. He began
showing dairy heifers at the age of 11 and working on commercial dairy farms during high
school, which ignited his passion for the dairy industry. Isaac enrolled in the University of
Minnesota in the fall of 2009. As an undergraduate he was highly active within the campus
community, including being a member of the university dairy judging team, the dairy challenge
team, serving as the treasurer of Gopher Dairy Club, and being the president of FarmHouse
Fraternity. He completed his bachelor’s degree in Animal Science in spring of 2013 and stayed at
the University of Minnesota to start master’s degree program in the laboratory of Dr. Marshall
Stern. His master’s research used dual-flow continuous culture fermenters to study potential
modifiers of rumen fermentation and the rumen microbiome. Isaac finished his master’s degree in
Animal Science in the summer of 2015, and immediately began a Ph.D. at The Pennsylvania
State University. At Penn State, his Ph.D. dissertation focused on the nutritional and
environmental factors influencing daily and annual rhythms of milk synthesis in dairy cows. After
graduation, Isaac will remain in academia as an assistant professor of dairy science at South
Dakota State University.