Disturbed energy balance regulation and links to human metabolic diseases such as obesity – general introduction, discussion of phenotyping test assays for body composition
Jan Rozman
Helmholtz Zentrum München
Fatty acid oxidation
Mitochondrial metabolism
Porphyrin metabolism
Purine or pyrimidine metabolism
Metabolism – chemical transformations to provide energy
and matter for life sustaining structures and processes
Major categories of inherited metabolic disorders
rare diseases – widespread diseases
Steroid metabolism
Mitochondrial function
Peroxisomal function
Lysosomal storage disorders
Carbohydrate metabolism
Amino acid metabolism
Urea Cycle Defects
Organic acid metabolism
Energy transformation
Basics
SI unit energy: Joule [J] 4.1868 J is the energy amount required to heat up 1 g of water from 14.5°C to 15.5°C at standard conditions (1013 hPa barometric pressure).
Energy turnover of a 70 kg man: 11300 kJ.
SI unit „power“: Watt [W] Power is defined as energy per time (1 Watt = 1 Joule/s). Power of a 70 kg man: 131 W (11300 kJ *1000 / 86400).
Heterotrophy and oxidative processes
source: www.hcrowder.com
Sun
Max Kleiber: Fire of Life (1961)
Max Kleiber (1893 –1976), Swiss
agricultural biologist, born and
educated in Zurich, Switzerland.
Thesis The Energy Concept in the
Science of Nutrition. Research at
UC Davies. Kleiber contributed
fundamental concepts to better
understand principles of metabolic
regulation (Kleiber’s Law)
Obesity and diabetes are serious public
health problems
Energy balance regulation and disease
adapted from:
Several human diseases are related to a positive
energy balance e.g. obesity and its sequelae
from: WHO 2013 (http://apps.who.int/bmi/index)
Energy balance regulation
Rozman et al. 2014
Overweight and obesity – a showcase
Body mass development of a 67 y/o man
In 1998: BMI 23.2 in 2015: BMI 28.9 (plus of 19.6 kg in 17 years)
Partly, daily body mass recordings together with major abnormalities in life
style factors (food intake, stress, etc.)
with kind permission of B. Meier
Precise energy regulation:
.... an average person actually gains 11 kg between the ages of 25 and 65; during this period he/she eats roughly 20 tons of food; his/her weight change corresponds to an average daily error of only 350 mg of food, as compared with the exact amount of food required for energy balance .
Reginald Passmore (c.f. Hervey, Nature 1969)
Energy balance regulation
Rozman et al. 2014
Implications of the „scaling problem“ on metabolic phenotyping
Rough estimate of the ratio of daily energy flux and total body energy
content:
Humans: 1:70 (10 MJ : 700 MJ)
Mice: 1:7 (40 kJ : 280 kJ)
In smaller animals fluctuations of daily energy flux have a higher
impact on energy regulation compared with larger animals:
Advantage of mouse models
Implications of the „scaling problem“ on metabolic phenotyping
Rough estimate of the ratio of daily energy flux and total body energy
content:
Humans: 1:70 (10 MJ : 700 MJ)
Mice: 1:7 (40 kJ : 280 kJ)
In smaller animals fluctuations of daily energy flux have a higher
impact on energy regulation compared with larger animals:
Advantage of mouse models
Kleiber‘s Law
How can we measure energy balance?
dietary energy
assimilated
energy
metabolizable
energy
energy storage (fat)
feces
urine
glucose fatty acids amino acids
basal metabolism
activity thermogenesis
diet induced
thermo-
genesis
growth
bomb calorimetry &
FT/IR spectrometry
metabolic cages clinical chemistry
body composition analysis
indirect calorimetry &
activity monitoring
metabolic cages
chemical analysis &
FT/IR spectrometry
monitoring of body temperature
Read outs for energy flux
Purpose of primary metabolic phenotyping:
Detection of disturbed regulation of energy balance in mice in search of new animal models for human metabolic disorders.
Let‘s start with obesity!
Difference in body weight (g)
-20 -15 -10 -5 0 5 10 15 20
Dnmt1Abcb4_2
IFIT-2VMP1 KO cond_het
TOM40AEY069
Traf7ADAR2
Tyk2Camp1
Cln3_delta_ex7/8_mutEmp3
LucPLKMFAP4
EPD0089_4_F11mir34 koBCH002
EUC FP00052B10_1Cln3_delta_ex7/8_het
EUC FP00100C08MEGAPBAP005RNtre2
MVD013HdhQ111
COX4-2 KoALI034
EUC BL6-FP00042B07IRP2
Fermt2EPD0028_5_G01
LrbaTmem45b
AEY17EPS8L2
EPD0156_1_B01 EPD0105_5_E01
HST012_hetNox4
Trdmt1Fra2
Cited4 KO condEPD0051_2_D09
PRDM11Csemp1
EmoryPept1
SAP007Afamin_2
FLG/HRN1Them 2
Abcb4_1HST012_mut
NfyaBAP002BAP004BAP012HST001 A
Difference in body weight (g)
-20 -15 -10 -5 0 5 10 15 20
SAP007IRP2
Afamin_2Abcb4_2
TOM40Csemp1
Nox4Cln3_delta_ex7/8_het
EUC FP00100C08MFAP4
Abcb4_1EPD0051_2_D09EPD0028_5_G01
Tyk2VMP1 KO cond_mut
Cln3_delta_ex7/8_mutPept1Fra2
EPS8L2mir34 ko
Mir221 KO convS248A KI_mutS248A KI_het
RNtre2Traf7
EPD0089_4_F11BAP005
NfyaHdhQ111
Dnmt1Trdmt1
Tmem45bEUC BL6-FP00042B07
Camp1MEGAP
PRDM11EUC FP00052B10_1
EPD0105_5_E01IFIT-2
Fermt2EPD0156_1_B01
Emp3COX4-2 Ko
ADAR2MVD013
PTBP2Lrba
ALI034FLG/HRN1
BAP012AEY069AEY17
HST012_hetHST012_mut
Cited4 KO condEmory
BAP002LucPLKHST001 B
male female
unpublished data
Body mass
Body mass is an easy to obtain parameter
at least it should be – calibration of electronic
balances, control time of the day, treatment of mice,
correct identification, etc.
Important parameter – first indication that growth or
energy allocation are affected
Important confounder for many other parameters
like energy expenditure, etc.
0
5
10
15
20
25
0 5 10 15 20 25 30 35
Body mass (g)
Lean
mass (
g)
m 44 wt
m 44 mut
0
5
10
15
0 10 20 30 40
Body mass (g)
Fat
mass (
g)
m 44 wt
m 44 mut
Good to know - big or fat?
NMR
Micro CT
lean visc
sc
Prim
ary
phenoty
pin
g
Secondary
phenoty
pin
g
unpublished data
Primary Screen Results
Body mass and body composition phenotypes
Males Females
Parameter Reduced Increased cohorts Reduced Increased cohorts
Body mass 19 (31.1%) 2 (3.3%) 61 16 (27.1%) 5 (8.5%) 59
Lean mass 13 (26.5%) 1 (2.0%) 49 10 (21.7%) 6 (13.0%) 46
Fat mass 13 (26.5%) 4 (8.2%) 49 7 (15.2%) 2 (4.3%) 46
Body composition analysis normalized to body mass
Lean mass 2 (4.1%) 4 (8.2%) 49 2 (4.3%) 5 (10.9%) 46
Fat mass 4 (8.2%) 1 (2.0%) 49 6 (13.0%) 2 (4.3%) 46
Analysis based on 1655 mice from 73 different wildtype ~ mutant cohorts (data: M. Willershäuser)
Energy balance regulation
Energy assimilation
Food intake Separation of spillage and feces Bomb calorimetry
Workflow
day 1 day 8
Mo Tu We Thu Fr Sa Mo Su Tu We Thu Fr Sa Su Mo
Food ad libitum
pill
Tu We
Make sure enough material is collected (> 2 g of feces)
Quality control
Energy balance regulation
Indirect calorimetry
Indirect calorimetry in humans
from: www.cosmed.com
Clinical application
• Therapy of adiposity
• substrate utilization and
metabolic flexibility
• control of wasting
syndrome in cancer
patients
• control of energy balance
during post surgery
therapy or total parenteral
nutrition
Also important in large-scale cohort studies
monitoring metabolic functions and disease
(e.g. obesity), addressing basal research
questions (contribution of BEIGE/BRITE
cells to daily energy expenditure,
contribution of microbiota to metabolic
rate).
Challenges & Aims: Explain Sources of Variation in Metabolic Rate to Finally Identify
Factors Playing a Role in Energy Balance Regulation.
PhenoScale: Development and refinement of new metabolic and behavioural phenotyping assays
www.phenoscale.com
Partners:
Medical Research Council – Mammalian Genetics Unit – Harwell, UK (Coordinator Roger Cox)
National Research Council, CNR – Institute of Cell Biology, IBC – Monterotondo, Italy
Helmholtz Zentrum München – German Mouse Clinic, GMC – Munich, Germany
The Italian Institute of Technology, IIT – Genova, Italy
TSE Systems GmbH - Bad Homburg, Germany
•Measurement: 21 hours
•Food and water ad libitum
change
T-5 T0 T12 T16
Light on Light on Light off
T-5
measurement
HMGU 19:00 CET
drafted by J Rozman (HMGU) & M. F. Champy (ICS)
further contributions from all EUMODIC partners
Careful evaluation of the protocol during PhenoScale
Metabolic rate
unpublished data
Are 21 hours sufficient for a robust measurement?
20
40
60
80
100
120
14:00 19:00 00:00 05:00 10:00 15:00 20:00 01:00 06:00 11:00 16:00 21:00
Time (hrs:min)
Oxyg
en
co
nsu
mp
tio
n (
ml/h
)
DAY 1 DAY 2
20
40
60
80
100
120
Day 1 Day 2
Ox
yg
en
co
ns
um
pti
on
(m
l/h
)
n =
14
r2 =
Comparison of 14 mice over more than
two days
small differences in the temporal pattern
(e.g. adaptation)
mean VO2 not different but highly
correlated (r2 = 0.87)
unpublished data
• SOP is rather flexible and takes into account capacity limitations (possible to do only males).
• Duration is flexible (minimum 21 hours, acclimation can be added).
• Equipment differs between centers. • Calorimetry still seems to be a bottleneck in the pipeline. • Data analysis challenging because of complex nature of
the results (display on www.mousephenotype.org needs improvement).
Read outs of the indirect calorimetry test
• Oxygen consumption VO2 [ml O2*h-1*animal-1]
• Carbon dioxide production VCO2 [ml CO2*h-1*animal-1]
• Respiratory exchange ratio RER [VCO2/ VO2]
• Lipid and carbohydrate oxidation rates [mg min-1]
• Heat production HP [mW*animal-1]
• Food consumption and water uptake [g and ml]
• Physical activity (locomotor activity [cm 15 min-1] and rearing [counts 15 min-1])
• Body mass [g]
Read outs are evaluated in the literature meaningful to characterize metabolic functions in our experience identify interesting new phenotypes
Several high ranking papers address how metabolic data should be analysed (e.g. Tschoep et al, Speakman papers, Kayala/Schwarz group).
Metabolic rate plotted versus body mass
Body mass (g)
0 10 20 30 40 50 60
VO
2 (
ml O
2*h
-1)
0
20
40
60
80
100
120
140
160
WT and MUT all lines
WT line 05
MUT line 05
Analysis by linear regression models
including genotype, sex and body mass:
Body mass V
O2
VO2 ~ sex * genotype + body mass
Read outs of the indirect calorimetry test
Hypometabolism in Myoz1 (IMPC)
The reduction in energy turnover
likely exceeds the reduction in
physical activity.
Further studies are conducted to
dissect primary and secondary
effects on energy balance
regulation.
Update GMC indirect calorimetry in IMPC
Project Sex Wildtype Mutant Sum
EUMODIC females 168 218 386
males 201 243 444
(sub-total 369 461 830)
IMPC females 371 640 1011
males 392 646 1038
(sub-total 763 1286 2049)
Total 1132 1747 2879
Data available (reporting date 22/04/2015): • 99 mutant lines IMPC • 36 mutant lines EUMODIC 2879 mice in total For pilot evaluation only mean oxygen consumption (VO2 mean over 21 hours) was used. First time, cross-project analysis of gene effects on metabolic regulation.
Histogram
cohort size [n]
0 5 10 15 20 25 30
Count
[n]
0
10
20
30
40
50
60
Count
Meta analysis of indirect calorimetry data
EUMODIC IMPC
SM-MARS-8A, 8 cages, Sable Systems, Las Vegas, US
PhenoMaster, 32 cages, TSE Systems GmbH, Bad Homburg, Germany
Main source of variation: body mass
Amazingly good fit of the two data sets from EUMODIC and IMPC. We modelled mean VO2 in a linear regression model to obtain residual and adjusted VO2.
Adjusted mean oxygen consumption seems to be independent of body mass.
But is related to rearing and physical activity
Residual and adjusted VO2: independent of BM
Mean oxygen consumption in EUMODIC & IMPC
VO2
unadjusted
Body mass
.
Mutant lines
Mean oxygen consumption in EUMODIC & IMPC
VO2
unadjusted
Body mass
.
Mutant lines VO2 residual .
Mutant lines
0
5
10
15
20
25
30
35
0,7
500
0,7
750
0,8
000
0,8
250
0,8
500
0,8
750
0,9
000
0,9
250
0,9
500
0,9
750
1,0
000
1,0
250
1,0
500
1,0
750
1,1
000
1,1
250
1,1
500
1,1
750
1,2
000
1,2
250
1,2
500
Freq
uen
cy [
co
un
ts]
Out of 134 mutant lines form IMPC
and EUMODIC analysed here:
• 10 were hypometabolic (7.5%)
• 3 hypermetabolic (2.2%)
In almost 10% of the mutant line
effects on metabolic regulation could
be detected. The additional data (e.g.
locomotor activity) will help to
distinguish between primary and
secondary effects on energy turnover.
mean +/- 1 S.D. of
controls (n=1132
mice)
New gene candidates involved in metabolic regulation?
Not only monitoring of metabolic rate, but also energy
allocation and on-line monitoring of metabolic functions (e.g. in
combination with challenges like HFD, food deprivation,
treadmill, or running wheel)
Additional value – further read outs of the indirect calorimetry
• Oxygen consumption VO2 [ml O2*h-1*animal-1]
• Carbon dioxide production VCO2 [ml CO2*h-1*animal-1]
• Heat production HP [mW*animal-1]
• Food consumption and water uptake [g and ml]
• Physical activity - locomotor activity, rearing, link between metabolic & behavioral phenotyping, biorhythms (?), sleep (?)
• Respiratory exchange ratio RER [VCO2/ VO2], metabolic flexibility
• Lipid and carbohydrate oxidation rates [mg min-1], substrate utilization, in vivo monitoring of metabolic functions
Additional value – further read outs of the indirect calorimetry
control mutant
A
ctivity
S
ubstr
ate
use
RE
R
[cm
/20
min
]
[g/m
in]
[VC
O2
/VO
2]
Time [h:min]
One step further:
Indirect calorimetry – Substrate oxidation
C6H12O6 + 6 H2O + 6 O2 6 CO2 + 12 H2O H0 = -2813 kJ mol-1
CH3(CH2)14COOH + 23 O2 16 CO2 + 16 H2O H0 = -10025 kJ mol-1
Glucose oxidation: RER = 6 CO2 / 6 O2 = 1.0
oxicaloric equivalent =
2813 kJ mol-1 / 6 mol O2 = 468.8 kJ mol O2-1 = 20.9 kJ liter-1 O2
Palmitate oxidation: RER = 16 CO2 / 23 O2 = 0.70
oxicaloric equivalent =
10025 kJ mol-1 / 23 mol O2 = 435.9 kJ mol O2-1 = 19.5 kJ liter-1 O2
Net glucose oxidation: c [g/min] = 0.746 VCO2 – 3.21 VO2 Net lipid oxidation: f [g/min] = 1.67 VO2 – 1.67 VCO2
Calculations do not consider nitrogen metabolism and are only valid for standard conditions
RER
1.00
0.70
© J Rozman HMGU
Substrate utilization depending on energy expenditure
AKR/J and SWR/J mice, fed
CD, 11-12 weeks old, n=29-
30, by indirect calorimetry for
24 h.
(submitted manuscript: Kless,
Rink, Rozman & Klingenspor
2016)
More lipid oxidation in SWR/J.
Different pattern in substrate
utilization.
Energy balance regulation – a showcase
-25 -20 -15 -10 -5 0 5 10 15 20 25
KTA041
ABE17 hom
Sepp hom
ABE17 het
AGA002
A008A01
Popdc2
Ali18
ATE2
Ali35
HST009
ABE012
ABE 1
ATE001
Medane1
Drasic
SIP
Mag
Dea3
ABE 2
NCAM
R1-KO
EYL
Delta 1
Insl 5
PK
Ptdsr
DKK3
M076C04
DLG3
NADH
Glut8
Cin85
Tp53
Miz1
Ali027
Fin13
Vimentin
EPS15
p0071
SUMO 1
Neuch het
UBB+1/3413
CIN85
FoxP2
DMBT1
Elastin
Sepp het
FFM1017
Mchr1
Ptpg
Lrba
Arl4
DNAseX
ALI22
Nbea
ESCP
Neurobeachin heterozygous mice are
moderately obese:
What is known about the gene?
member of BEACH domain family of proteins
role intracellular targeting of membrane proteins
peripheral & central neurons as well as in
endocrine cells
essential for synaptic neurotransmission
In humans, the gene was supposed to be
associated with autism
Nbea has not been associated with obesity
before
Neurobeachin: A new candidate gene for obesity?
Nbea
Energy balance regulation
Rozman et al. 2014
Nbea +/- were moderately overweight, have increased body fat at 12 weeks of
age, and were slightly more susceptible to feeding a high fat diet (60en% fat)
Olszewski et al. (PLoS Genetics 2012)
Only subtle effects in behavioural and neurological screening, clinical
chemistry and gene expression. No obvious link to energy regulation.
Even differences in body mass and body composition are moderate.
Olszewski et al. (PLoS Genetics 2012)
Nbea heterozygotes
• overeat when offered an energy dense
and palatable diet
• overeat after fasting
• overconsume palatable caloric liquids
• but not palatable non-caloric liquids
• are leptin sensitive
• hyper-responsive to the anorexigenic
opioid receptor antagonist naltrexone
(NTX)
In depth analysis of
physiological functions -
specific feeding
paradigms:
Links to human studies:
Olszewski et al. (PLoS Genetics 2012)
Comparison of Nbea+/- and WT
mice under different feeding
paradigms
Ad libitum-fed - differential
expression of dynorphin (DYN)
mRNA in the hypothalamus.
16-h food deprivation differently
affected expression of DYN,
proopiomelanocortin (POMC),
opioid-like receptor-1 (ORL1) and
corticotropin releasing hormone
(CRH) in the hypothalamus
Differential expression of feeding-related genes
Monitoring of energy balance – indirect calorimetry
The exact determination of food intake, energy assimilation and energy
expenditure in mice 8 weeks of age (early onset of obesity)
Olszewski et al. (PLoS Genetics 2012)
Ad libitum access to
food and water
Home cage conditions
Single caged mice
Determination of energy balance
Olszewski et al. (PLoS Genetics 2012)
Polymorphisms in the Nbea gene are linked to human
obesity
Collaborators at the National Childhood Obesity Centre, Karolinska
Institute, Stockholm, Sweden detected a linkage of polymorphisms in the
Neurobeachin gene with human obesity in two clinical studies.
- adult man (normal weight and obese) older than 60 years of age
- overweight adolescents
Olszewski et al. (PLoS Genetics 2012)
The new metabolic phenotype linked to Nbea heterozygotes was detected in
the primary screen.
In depth second-line phenotyping provided a functional explanation for the
development of the phenotype (hyperphagia).
A direct link to human obesity was found in clinical studies.
Nbea is a new candidate gene for human adiposity
Acknowledgement
Molekulare Ernährungsmedizin
Technische Universität München
Martin Hrabé de Angelis (Director)
GMC Coordination Team
Helmut Fuchs (sci.-tech. Head)
Valerie Gailus-Durner (sci.-adm. Head)
Energy Metabolism Monja Willershäuser, Martin Kistler Ann-Elisabeth Schwarz, Anna Dewert, Brigitte Hermann
Nicole Ehrhard
Clinical Chemistry Birgit Rathkolb Kateryna Butuzova
Elfi Holupirek
Eckhard Wolf, LMU München
Martin Klingenspor
Hannelore Daniel
Diabetes Susanne Neschen