qtl analysis of phosphorus utilization-efficiency (pute
TRANSCRIPT
QTL Analysis of Phosphorus Utilization-Efficiency
(PUtE) in Arabidopsis thaliana
MSc Thesis in Genetics (GEN-80436)
Submitted by
Nakai Goredema
Registration Number: 850604270100
ii
QTL Analysis of Phosphorus Utilization-Efficiency (PUtE) in Arabidopsis
thaliana
MSc Thesis in Genetics
Submitted by Nakai Goredema
In partial fulfillment of Master of Science in Plant Breeding and Genetic Resources
Wageningen University
The Netherlands
Supervisors: MSc. Charles N. Moreira and Dr. Mark Aarts
Examiners: Dr. Mark Aarts and Prof. Maarten Koornneef
iii
Table of Contents List of Figures ............................................................................................................................................ v
List of Tables ............................................................................................................................................ vi
Abstract .................................................................................................................................................. vii
Acknowledgements ............................................................................................................................... viii
Dedication ............................................................................................................................................... ix
1. INTRODUCTION ................................................................................................................................... 1
Phosphorus, a major nutrient in plant growth .................................................................................... 1
P-deficiency alleviation in cropping systems ....................................................................................... 1
Plant response mechanisms to P-stress .............................................................................................. 1
Arabidopsis as a model plant species .................................................................................................. 2
Linkage mapping versus Genome-wide association mapping (GWAS) ............................................... 2
Genes responsive to P-deficiency ....................................................................................................... 3
Study objectives .................................................................................................................................. 4
2. MATERIALS AND METHODS................................................................................................................. 5
RIL Population.......................................................................................................................................... 5
Plant material ...................................................................................................................................... 5
Plant growth ........................................................................................................................................ 5
Shoot Fresh and Dry weight ................................................................................................................ 5
Determining phosphate concentration ............................................................................................... 5
Calculating P-content and PUtE .......................................................................................................... 6
QTL mapping and Collocation ............................................................................................................. 6
Gene expression analysis ........................................................................................................................ 7
Plant material ...................................................................................................................................... 7
Harvesting samples for RNA extraction .............................................................................................. 7
RNA extraction and purification .......................................................................................................... 7
cDNA synthesis .................................................................................................................................... 7
Gene Primers ....................................................................................................................................... 8
Real-time qPCR .................................................................................................................................... 8
Summary of Gene expression analysis ................................................................................................ 9
3. RESULTS ............................................................................................................................................. 10
Descriptive statistics: RIL population ................................................................................................ 10
Symptoms of P-deficiency ................................................................................................................. 13
QTL detection for SFW, SDW, PCONC, PCONT, PUtE, sulphate and oxalate concentration ................. 14
iv
QTL detection using Interval mapping .............................................................................................. 14
QTL detection using Multiple QTL mapping (MQM) ......................................................................... 17
Collocating QTLs for SFW .................................................................................................................. 18
Collocating QTLs for PCONC .............................................................................................................. 19
Collocating QTLs for Sulphate and Oxalate ....................................................................................... 19
Co- expression analysis ...................................................................................................................... 20
Gene expression analysis ...................................................................................................................... 21
Determining Primer efficiency .......................................................................................................... 21
Expression of IPS1, Pht1;4, SULTR2;1, ARF19, MYB111 and RHD6 ................................................... 23
4. DISCUSSION ....................................................................................................................................... 26
QTL Detection: SFW, SDW, PCONC, PCONT, PUtE, sulphate and oxalate concentration ................. 26
Gene expression analysis .................................................................................................................. 28
5. CONCLUSIONS AND RECOMENDATIONS ........................................................................................... 30
REFERENCES .......................................................................................................................................... 31
APPENDICES ........................................................................................................................................... 34
v
List of Figures
Figure 1: Summary of materials and methods used for QTL detection ..................................... 6
Figure 2: Summary of materials and methods used for gene expression analysis .................... 9
Figure 3: Frequency distribution curves of traits phenotyped under sub-optimal Pi............. 11
Figure 4: Frequency distribution curves of traits phenotyped under optimal Pi.................... 12
Figure 5: Symptoms of Pi deficiency ....................................................................................... 13
Figure 6: QTL detection using interval mapping. ....................................................................16
Figure 7: QTL detection using multiple QTL mapping............................................................. 17
Figure 8: Collocating SFW QTLs ............................................................................................... 18
Figure 9: Collocating QTLs for PCONC ...................................................................................... 19
Figure 10: Co-expression analysis of MGD3 . ........................................................................... 20
Figure 11: Primer efficiencies: 10 times dilution. .................................................................... 21
Figure 12: Primer efficiencies: 2 times dilution ....................................................................... 22
Figure 13: Relative expression of GWAS candidate genes for PUtE ........................................ 24
vi
List of Tables
Table 1: DNA primers used for qPCR ......................................................................................... 8
Table 2: Trait means for Sha x Col RILs under optimal Pi ......................................................... 10
Table 3: Trait means for Sha x Col RILs under sub-optimal Pi .................................................. 10
Table 4: Trait means for Col and Sha ....................................................................................... 10
Table 5: QTL detection using interval mapping ....................................................................... 15
Table 6: QTL detection using multiple QTL mapping ............................................................... 17
vii
Abstract
Phosphorus is an essential nutrient for plant growth but is limiting in many cropping
systems. It is absorbed by plants roots as inorganic phosphate (Pi). Fertilisers used to
alleviate Pi deficiency are costly and environmentally unfriendly. It is paramount to have
plant genotypes that utilize less Pi to produce an equal amount of biomass. Here I report a
QTL analysis to unravel the genetic basis of Phosphate Utilization Efficiency (PUtE) in
Arabidopsis thaliana using linkage mapping and subsequent collocation with previous
genome-wide association studies (GWAS). A Shahdara x Columbia core set population of 164
F10 lines was grown on optima and sub-optimal Pi. The phenotyped traits were shoot fresh
and dry weight, Pi concentration and content, PUtE, sulphate and oxalate concentration.
Several QTLs that suggest direct involvement in PUtE were mapped for shoot fresh weight
and Pi concentration mainly on chromosomes 1, 2, 3 and 5. Candidate genes from a
previous GWAS study (IPS1, Pht1;4, ARF19, MYB111, SULTR2;1 and RHD6) were analysed for
gene expression using accessions of contrasting phenotypes, Col-0 and Sha harvested at
different time points. IPS1 and Pht1;4 were up-regulated under sub-optimal Pi supply.
SULTR2;1 was down regulated and Sha showed higher ARF19 expression than Col-0.
viii
Acknowledgements
I would like to thank my supervisors Charles Neris Moreira and Mark Aarts for guidance,
commitment, patience and encouragement. I also want to thank Diaan Jamar in the
Laboratory of Plant Physiology, and all members of the Genetics Department for their
technical assistance and advice. A special thanks goes to Ana Carolina Campos and Mina
Carvalho for their practical assistance during the experiments.
ix
Dedication
To my family, relatives and friends.
1
1. INTRODUCTION
Phosphorus, a major nutrient in plant growth
Phosphorus (P) is one of the most important mineral nutrients required for plant growth and
development (Jain et al. 2007). It plays an important role in essential molecules such as
nucleic acids, adenosine triphosphate (ATP) and phospholipids. It is therefore involved in an
array of metabolic processes including energy metabolism, photosynthesis, membrane
synthesis and stability, glycolysis, nucleic acid synthesis, redox reactions, respiration,
nitrogen (N) fixation and enzyme activation and inactivation (Vance et al. 2002). Although
most soils have large reserves of P, little is available for absorption by plants. P is absorbed
by plants roots as inorganic phosphate (Pi) dissolved in solution mainly as phosphate ions
(HPO4-2
and H2PO4-) (Jain et al. 2007). Organic forms are unavailable to plants and have to be
hydrolysed to release Pi. Furthermore it is one of the slowly diffusing nutrients in soil
solutions, thus creating a zone of depletion around the rhizosphere. It was also discovered
that poor Pi availability is further aggravated as Pi gets fixed by Ca in calcareous soils and by
Fe and Al in acid soils, which reduces solubility and availability to plants (Jain et al. 2007).
P-deficiency alleviation in cropping systems
Pi deficiency in many cropping systems is alleviated by fertilizer applications. However this
approach is financially costly and not sustainable as it estimated to deplete the easily mined
P rock by 2060 (Zhang et al. 2009). Excessive fertilizer application can result in leaching of Pi
into deeper soil zones and eventually reach the water table, which subsequently results in
water pollution and eutrophication (Alva et al. 1999). Eutrophication reduces oxygen in
aquatic ecosystems and may have huge adverse effects on aquatic life. Pi fertilizer
application alone is inefficient because up to 80% of the applied Pi fertilizer can be
immobilized into forms other than orthophosphate (monophosphate) before plant uptake
(Ma et al. 2011) and only 15-20% of the fertilizer available to plants is assimilated, the rest
remains in the soil (Syers et al. 2008). All these constraints, elucidate that Pi fertilizer
application alone cannot fully alleviate Pi unavailability and more so is environmentally
unfriendly. This has led to a search for more environmentally friendly and economically
feasible strategies to improve crop production in low phosphorus soils (Narang et al. 2000).
P is a non-renewable resource thus should be used sustainably. Therefore the development
or availability of P-efficient genotypes in all plants is inevitable. This strategy sounds ideal as
the plants could efficiently use Pi that is already present in the soil.
Plant response mechanisms to P-stress
Plants adapt to low Pi conditions by morphological and metabolic changes that include
increased lateral roots, reduction in primary root length, increased root to shoot ratio,
secretion of phosphatases, exudation of organic acids and enhanced expression of Pi
transporters (Vance et al. 2002). Most of these changes are controlled by many genes, each
2
contributing either in small or large effect towards overall Pi homeostasis. The regions along
the genome that contain such genes are known as quantitative trait loci (QTLs). QTL mapping
studies done on phosphate stress response in A. thaliana focused on changes in root
architecture (Svistoonoff et al. 2007; Vance et al. 2002; Reymond et al. 2006; Rouached et al.
2010; Pérez-Torres et al. 2008; Jiang et al. 2007), where in most cases plants were
completely deprived (starved) of Pi at a certain developmental stage. Moreover most of the
previous research work was only centered on Pi acquisition or phosphate acquisition
efficiency (PAE) (Narang et al. 2000; Rose & Wissuwa 2012; Vance et al. 2002). PAE is given
as the amount of Pi in the plant per unit of Pi supplied. Phosphate utilization efficiency
(PUtE) is defined as the amount of biomass produced per unit of Pi acquired (Wang et al.
2010). Hence PAE cannot represents how efficiently plants utilize the acquired Pi.
Comparative studies of PUtE have often suffered from confusion of variation in P-acquisition
efficiency (Veneklaas et al. 2012). Genotypes with high PAE are not necessarily P-efficient
because very high concentrations of Pi in the cytosol may become toxic to plants when they
cannot tolerate it. Research focus on PUtE can be useful since PUtE complements PAE in
increasing the overall phosphorus use efficiency (PUE) in plants. Thus PUE is comprised of
two components: PAE and PUtE. Aiming at increasing Pi uptake alone will benefit yields in
some cases but also increases the total amount of Pi exported from the soils (Veneklaas et
al. 2012). Ideally, plants should produces high amount of biomass per unit Pi taken up or
acquired. Here I focus on PUtE and the study is based on mild Pi stress, rather than
complete Pi starvation. This is because in most cases crop plants are grown under limiting Pi
conditions but are rarely completely starved. Therefore mild Pi stress conditions seem to be
a better representation of low Pi conditions in nature, as such will be mimicking field
conditions.
Arabidopsis as a model plant species
A. thaliana is favored for many studies on plant response to Pi stress because of its short life
cycle, naturally occurring genotypic variations and the availability of its complete genome
sequence (Koornneef et al. 2011). The species can be self-fertilized and does not suffer from
inbreeding depression and hence recombinant inbred lines (RILs) for QTL mapping can be
obtained relatively easy (Alonso-Blanco et al. 2000). Most Brassicaceae including
Arabidopsis, do not form symbiosis with mycorrhizal fungi which increase the surface area of
Pi absorption. These species have to develop other mechanisms to acquire Pi on their own.
Hence suitable for PUE studies.
Linkage mapping versus Genome-wide association mapping (GWAS)
Linkage mapping (also QTL mapping) and genome-wide association mapping (GWAS) are the
most commonly used strategies to dissect the genetic basis of polygenic traits (Kloth et al.
2012). For QTL mapping, a RIL population is developed by crossing accessions of contrasting
phenotypes and repeated selfing of progeny to attain homozygosity. Recombination
between genetic markers is the key to success since the linkage map is constructed from
recombination frequencies. Although this tool has been used for decades, it has limited
3
resolution because of limited recombination events that can be studied (Kloth et al. 2012).
However its advantages include relatively small population size requirement, low number of
genetic markers and absence of population structure effects (Kloth et al. 2012). A Hapmap
population that consists of numerous genetically diverse accessions collected from different
habitats around the world is used for GWAS. GWAS uses linkage disequilibrium to measure
the association of genetic markers with the trait in question. This tool has recently gained
preference to linkage mapping since mapping resolution is often increased by the ability to
analyse recombination events that have accumulated over thousands of years, thus fine
mapping made possible (Kloth et al. 2012). However GWAS often requires a large population
size and has population structure effects. Because of these contrasting attributes, QTL
mapping and GWAS are in principle complementary tools rather than substitutes.
Furthermore, the combination of these mapping tools allows the identification of true
positives, false positives and false negatives (Bergelson and Roux, 2010).
Genes responsive to P-deficiency There are several genes that were reported to play a role in plant response to early, mid and
long-term Pi deficiency as reported by Misson et al. 2005; Morcuende et al. 2007). Some of
these genes are INDUCED BY PHOSPHATE STARVATION 1 (IPS1) and PHOSPHATE
TRANSPORTER 1;4 (Pht1;4), PHOSPHATE TRANSPORTER 3;2 (Pht3;2), RIBONUCLEASE 1
(RNS1). IPS1 is induced by phosphate starvation and is expressed in both roots and shoots
(Martin et al. 2000). Pht1;4 and Pht3;2 are phosphate transporters, with Pht1,4 involved in
Pi acquisition (Shin et al. 2004) and Pht3;2 encoding ATP/ADP antiporter (Misson et al 2005).
RNS1, a regulator of RNA metabolism, is actively involved in anthocyanin biosynthetic
pathway as a response to Pi limitation (Bariola et al. 1994). PHO1 and PHO2 also have been
reported to play an important role in Pi homeostasis. PHO1 is involved in loading and
transfer of Pi from roots to shoots (Morcuende et al. 2006; Secco et al. 2010). The gene is
known to encode a transporter that channels Pi ions, along an electrochemical gradient,
from the root stellar cells to the xylem (Hamburger et al. 2002). A mutant of PHO1, pho1-2
exhibits severe phosphorus deficiency in leaves, producing dark green leaves with purple
petioles (Poirier et al, 1991; Hamburger et al. 2002). PHO2 encodes ubiquitin-conjugase, an
enzyme that conjugates ubiquitin and unneeded proteins in the endomembrane (Lui et al.
2012). Ubiquitin labelled proteins are directed to the proteasome where they are degraded.
Hence PHO2 plays a role in maintaining Pi homeostasis in Arabidopsis. The pho2-1 mutant is
defective in this PHO2 regulation, and has been found to express PHO1 to high levels, with
unlimited loading of Pi into plant tissues up to toxic levels (Lui et al. 2012).
Moreira (unpublished data) previously identified several genes that seem to be involved in
phosphorus utilisation efficiency using genome wide association mapping. These are: AUXIN
RESPONSE FACTOR 19 (ARF19), MYB DOMAIN PROTEIN 111 (MYB111), ROOT HAIR
DEFECTIVE 6 (RHD6), and SULFATE TRANSPORTER 2;1 (SULTR2;1). ARF19 is a transcriptional
4
regulator of auxin-mediated genes, whose expression enable lateral root development, a
typical response of phosphate limitation (Okushima et al. 2007). MYB111 regulates the
production of flavonol glycosides, that have an anti-oxidative activity when plants are under
stress (Stracke et al. 2010). RHD6 regulates transcription of genes encoding root hair
development (Singh et al. 2008). SULTR2;1, a sulphate transporter, has been shown to be
up-regulated under Pi deprivation (Misson et al. 2005), suggesting sulphate use to cushion
anion/cation balance when Pi is limiting.
Study objectives
• To map QTLs for PUtE in Arabidopsis by linkage mapping so as to identify genomic
regions that are responsible for the variation of this trait. This can help breeders to
efficiently select for P-efficient genotypes.
• To perform a collocation analysis by comparing Moreira’ s GWAS results and my
results from linkage mapping to confirm true positives.
• To check the expression of IPS1, Pht1;4, ARF19, SULTR2;1, MYB111, and RHD6 (as
PUtE candidates) in response to sub-optimal Pi.
5
2. MATERIALS AND METHODS
RIL Population
Plant material
A Sha x Col-0 core set population of 164 F10 lines selected as an optimal set for QTL mapping
experiments was used for this analysis. The RILs are derived from a cross between two
contrasting lines for PUE, Colombia (Col-0) and Shahdara (Sha). Narang et al. 2000 reported
Col-0 as poor in PAE, a component of PUE. Shahdara is highly P-efficient (Reymond et al.
2006). The Sha x Col-0 RIL population has been developed at INRA, Versailles (France) and
has been used widely in research on abiotic stress. More information about this population
can be found at the Arabidopsis thaliana Resource Centre in Versailles
(http://dbsgap.versailles.inra.fr/vnat/Documentation/13/DOC.html).
Plant growth
Seeds were sown on rock wool blocks. Two seeds were sown per block and thinned few days
after emergence to allow only one plant to grow. The population was grown under optimal
and sub-optimal Pi concentrations of 1.2 mM KH2PO4 and 100 µM KH2PO4, respectively.
Seeds were subjected to cold treatment, a process known as stratification to break
dormancy and enable uniform seed germination. The seeds in Petri dishes containing water-
soaked filter paper were put in cold room at 4oC for 3 days. The stratified seeds were sown
in a climate cell with temperatures of 21oC, 70% humidity, day length of about 10 hours and
200 µmoles/m2/s of light intensity. Hyponex nutrient solution was used (see Appendix I for
composition) and refreshed after 14 days after plant emergence. Irrigation was done
automatically and uniformly every second day by an ebb-and-flood system. General
monitoring on the phenotypic changes on plants was done during growth and development.
Harvesting was done at five weeks after sowing.
Shoot Fresh and Dry weight
Immediately after cutting the above ground parts, shoot fresh weight (SFW) was measured
using a sensitive balance. The leaf rosettes were put each in envelopes and cast into liquid
nitrogen for short-term preservation during harvesting. The plant material was stored at -
80oC for long term storage. The shoots were freeze-dried for 60 hours at a temperature of -
40oC, at a pressure of about 60 kPa and shoot dry weight (SDW) was determined.
Determining phosphate concentration
Shoot Pi concentration (PCONC) and organic acids such as malate, oxalate, nitrate, phytate,
sulphate and citrate were determined simultaneously by a Dionex® High-Performance
Anion-Exchange Chromatography (HPAE). The main purpose of freeze-drying plant material
was to preserve the integrity of the phosphate anions. Leaf material weighing between 6 to
9 milligrams was ground to almost fine powder in eppendorf tubes and boiled at 100oC for
6
15 minutes, with 1 ml of 0.5 N HCl containing 50 mg/l t-aconitate. The samples were
centrifuged at maximum speed for 4 minutes and 200 µl of the supernatant from each
sample was transferred into 300 µl glass vials and analysed with HPAE. Deionised autoclaved
water was used as negative control and 50 mg/l t-aconitate solution as positive control.
Calculating P-content and PUtE
Shoot phosphate content (PCONT) was calculated by the formulae (SDW x Pi concentration)
and PUtE was calculated as SDW produced per unit phosphate used or acquired. Hence PUtE
equals SDW divided by Pi content in the shoots. As Pi content is equal to Pi concentration x
SDW, the final formula which I used was: PUtE = 1/Pi concentration (Good et al. 2007; Zhang
et al. 2009). The SPSS statistical package was used for analysis of variance within the RIL
population and descriptive statistics for all the phenotypic traits.
QTL mapping and Collocation
QTL analysis was done using MapQTL 6 software by interval mapping (IM) and multiple QTL
mapping (MQM). A Sha x Col-0 linkage map, spanning all the five Arabidopsis chromosomes,
and a total of 86 markers were used for QTL mapping. A LOD score threshold of 2.5 was
determined (α = 0.05) by performing a permutation test, to identify significant QTLs
associated with the trait being analyzed. Collocation was achieved by firstly performing a
GWAS, using HapMap phenotypic data previously collected, and finally compare regions that
assumed presence of segregating QTLs . Figure 1 is an illustration of the QTL mapping and
collocation procedures followed in this study.
Figure 1: Summary of materials and methods used to identify QTLs involved in PUtE
Grow RIL population:
Treatment: 100 µM KH2PO4
Control: 1.2 mM KH2PO4
Plant Material:
Sha x Col-0 RIL
population
QTL
confirmation
Linkage
mapping
Collocation
(GWAS) QTLs
Phenotypic traits:
SFW, SDW, PCONC, PCONT,
PUtE, Sulphate & Oxalate
concentration
7
Gene expression analysis
Plant material
Sixteen contrasting lines were grown for gene expression analysis. These lines had been used
in a preceding GWAS study for PUtE in Arabidopsis (Moreira, unpublished data). Five were
found to be highly P-efficient (high biomass yielders per unit of Pi): Br-0, Hi-0, Bla-1, Sq-8,
Kelsterbach-4, four moderately P-efficient: Gie-0, Utrecht, Gel-1, Ga-0, and five to be low:
Bur-0, Gr-1, Tha-1, Wei-0, Ca-0. Col-0, Sha and two mutant lines (pho1-2 and pho2-1) were
also included for this purpose.
Harvesting samples for RNA extraction
Samples were harvested at three different time-points, signifying different Pi stress terms:
14 (short), 21 (medium) and 28 (long) days after sowing (DAS). This was done to check the
changes of gene expression levels. However pho2-1 mutant did not germinate at all and was
omitted for the rest of the analysis. Col-0 and Sha did not germinate in all replications and
were only harvested at 21 and 28 DAS. Few plants of pho1-2 mutant germinated but had
severe stunted growth and therefore samples were only taken at 28 DAS, but plants were
still very small. Plants were cut, put in envelopes and immediately cast into liquid nitrogen.
Samples were stored at -80oC.
RNA extraction and purification
RNA was extracted only from the above ground parts. Frozen plant material was transferred
from envelopes to 2 ml eppendorf tubes and ground on a tissue lysing machine with the aid
of 2 mm sterile glass beads. The Qiagen RNeasy Plant Mini Kit extraction protocol was used
(see Appendix I). In summary, a maximum of 100 mg ground tissue was lysed in 450 µl Buffer
RLT by vortexing vigorously followed by incubating at 56oC for 3 minutes in a thermal mixer.
The lysate was transferred to a Qiashredder spin column and centrifuged for 2 minutes at
maximum speed. The lysate was separated from the plant debris and transferred to an
RNeasy spin column (Qiagen). After several purification steps (see Appendix I), the RNA was
further purified by on-column DNase digestion, and washed using Qiagen buffer RW1 and
RPE. The RNA was finally eluted from the RNeasy spin column using 40 µl of RNase-free
water. 1 µl of each sample was used to check the RNA quality using Nanodrop ® ND-1000
spectrophotometer. 2 µl of RNA was used for confirming RNA quality on 1.2% agarose gel
electrophoresis. Overally the quality ranged from about 10 ng/ µl to over 1000 ng/ µl. I
proceeded with samples that confirmed good quality on agarose gel electrophoresis.
Unfortunately the pho1-2 mutant did not yield good quality RNA because the plants were
very small.
cDNA synthesis
cDNA synthesis was done using iScript™ cDNA synthesis kit. The complete reaction mix
contained: 1 µg purified RNA, 4 µl iScript reaction mix, 1 µl iScript reverse transcriptase and
varying volumes of nuclease-free water to add up to 20 µl total volume. The mixture was
incubated under the following reaction protocol: 5 minutes at 25oC, 30 minutes at 42
oC, 5
8
minutes at 85oC and finally held at 4
oC. Finally 80 µl of nuclease free water was added to
each sample to obtain 100µl cDNA. This volume was considered as undiluted and cDNA was
stored at -22oC.
Gene Primers
Gene oligonucleotides (primers) were ordered from Bio Rad, and Table 1 shows a brief
description of loci of genes, primers and the duration of Pi stress under which they are
usually expressed. The primer sequences were not designed but taken from literature
(Czechowski et al. 2004; Misson et al. 2005; Kawashima et al. 2011). Each primer was tested
for efficiency before use using a cDNA supermix prepared by pooling 5 µl cDNA from each
sample (Ahmed, unpublished data). 5 µl of the pooled cDNA was diluted with nuclease free
water four times as follows: 10, 100, 1000 and 10000 times. Each dilution had four technical
replications.
Table 1: DNA primers used for qPCR
Gene ID Name Forward primer Reverse primer Pi Stress
duration*
AT3G09922 IPS1 AGGGGATGGCCTAAATACAAAATG GGGAGATAAACAAAACTCGCAGTC S
AT2G38940 Pht1;4 CACTTTCTTGGTACCTGAATC GTTTCTAGAGACAAGGAGAAAGAA S, M & L
AT1G19220 ARF19 TCCAACGAAGGAGAGAAGAAGCCA TGAAACTAAAGGCCCTGCACAAGC M & L
AT5G49330 MYB111 GCATTCCCTTCTCGGCAACAGAT GGAAACGGCAGTGAAGGCATAGAT L
AT1G66470 RHD6 TTTCCGGTCACACCGCTACTCA TTCTCACACGGGAGAGAGCACTCA -
AT5G10180 SULTR2;1 CAGAGAGTTTTGAATCTCTCTCACATC CCATCTGGATCATGTGTGTTG -
*Key: S, M & L means short, medium and long-term Pi stress respectively.
Real-time qPCR
For real-time qPCR, I used iQ™ SYBR® Green Supermix (Bio Rad), which contains SYBR Green
I dye, iTaq DNA polymerase, reaction buffer and dNTPs. The CFX_2stepAmp+Melt reaction
protocol which amplifies the target and produces a melting curve was used and is outlined
below. The total reaction volume was 20 µl with 3 µl cDNA, 10 µl SYBR Green dye, 1 µl
primer: forward and reverse(0.5 µl each) and 6 µl of nuclease free water. The reaction
protocol is as follows:
1. 3 min at 95oC
2. 15 sec at 95oC
3. 1 min at 60oC
4. Go to Step 2 and repeat 39 times
5. 10 sec at 95oC
6. Melt Curve 60oC to 95
oC, for 5 sec
The reactions were performed in quadruplicate in 96-well white PCR plates using CFX96
thermal cycler (Bio-Rad). At least two biological replications for most of the sample were
used. Quantification of gene expression was performed by the Cq method (also Ct), which
9
uses the threshold cycle to quantify gene expression. In fact, I used the comparative Cq
method which is also known as the 2-∆∆Cq
(Schmittgen and Livak, 2008). Following this
protocol, I normalised the target gene expression with the housekeeping gene; UBIQUITIN
10 (UBQ10). Further normalisation was done using a calibrator sample and is elaborated as:
∆Cq (Expression) = Cq (target gene ) – Cq (UBQ10)
∆∆Cq (Expression) = ∆Cq (sample) - ∆Cq (calibrator)
The calibrator sample was Col grown on normal Pi harvested at 21 days after sowing (DAS).
The ∆∆Cq values were used to calculate the normalised expression ratios using formula: 2-
∆∆Cq. Normalisation was done to correct for the differences in the initial amount of cDNA in
different samples. The 2-∆∆Cq
is preferred by most researchers to absolute quantification
because in most cases the required quantification is relative, and also saves the effort to
generate standard curves as with the latter (Schmittgen and Livak, 2008). The melting curve
analysis (Tm) was used to confirm a single amplicon.
Summary of Gene expression analysis
Figure 2 gives a summary of the activities that were done for analysing gene expression of
candidate genes identified in a previous GWAS study.
Figure 2: Summary of materials and methods used to analyse the changes in expression of
IPS1, Pht1;4, SULTR2;1, ARF19, MYB111 and RHD6 under normal and low Pi conditions.
Data
normalization
qPCR
Harvesting
samples RNA
extraction
Growing extreme lines
Treatment:100µM KH2PO4
Control: 1.2mM KH2PO4
cDNA
synthesis
cDNA RNA
Cq values
∆Cq values Data
normalization
∆∆Cq values
Expression
analysis
Expression
ratios
Data
normalization
10
3. RESULTS
Descriptive statistics: RIL population
Table 2 and 3 list the quantitative trait means for 164 (F10) Sha x Col-0 RIL population grown
under optimal and sub-optimal conditions respectively. The traits are shoot fresh and dry
weight (SFW and SDW), Pi concentration (PCONC), Pi content (PCONT), Pi utilization
efficiency (PUtE), sulphate and oxalate concentration. Table 4 depicts the means for Col-0
and Sha. The frequency distribution curves are shown in Figure 3 and 4.
Table 2: SFW, SDW, PCONC, PCONT, PUtE, sulphate and oxalate concentration mean,
minimum and maximum values for 164 Sha x Col-0 RIL population grown under optimal Pi
conditions
SFW (g) SDW (g) PCONC
(mg/gDW)
PCONT
(mg)
PUtE
Sulphate
(mg/gDW)
Oxalate
(mg/gDW)
Minimum 0.17 0.004 2.33 0.032 0.148 4.02 0.66
Maximum 0.95 0.097 13.53 0.988 0.606 28.75 1.68
Mean 0.492 0.037 8.71 0.331 0.124 11.78 1.10
Table 3: SFW, SDW, PCONC, PCONT, PUtE, sulphate and oxalate concentration mean,
minimum and maximum values for 164 Sha x Col-0 RIL population grown under sub-
optimal Pi conditions
SFW (g) SDW (g) PCONC
(mg/gDW)
PCONT
(mg)
PUtE Sulphate
(mg/gDW)
Oxalate
(mg/gDW)
Minimum 0.063 0.006 0.8 0.017 0.153 2.89 0.92
Maximum 0.764 0.082 6.64 0.213 1.273 26.26 2.43
Mean 0.356 0.039 2.11 0.079 0.548 10.42 1.52
Table 4: SFW, SDW, PCONC, PCONT, PUtE, sulphate and oxalate concentration mean
values for Sha and Col-0 grown under optimal and sub-optimal Pi conditions
SFW (g) SDW (g) PCONC
(mg/gDW)
PCONT
(mg)
PUtE Sulphate
(mg/gDW)
Oxalate
(mg/gDW)
Sub-optimal Pi
Col-0 0.411 0.025 1.55 0.038 0.646 8.31 1.42
Sha 0.578 0.037 1.71 0.063 0.595 10.76 1.06
Optimal Pi
Col-0 0.517 0.061 7.57 0.458 0.132 13.50 1.25
Sha 0.729 0.018 6.27 0.117 0.161 7.69 0.82
11
Figure 3: Frequency distribution curves of SFW, SDW, PCONC , PCONT and PUtE for 164
Sha x Col-0 RILs grown on sub-optimal Pi conditions
12
Figure 4: Frequency distribution curves of SFW, SDW, PCONC , PCONT and PUtE for 164
Sha x Col-0 RILs grown on optimal Pi conditions
13
Symptoms of P-deficiency
Several recombinant inbred lines planted on sub-optimal Pi showed severe symptoms of P-
deficiency. These include, stunting, chlorosis, dark green leaves with purple petioles (Figure
5). Plants that were grown on optimal Pi did not show any of these symptoms. Early
flowering was also observed as a response to nutrient stress and is also shown in Figure 5.
Figure 5: Symptoms of Pi deficiency on some of the Sha x Col-0 RILs: a & b showing
anthocyanin and stunted growth, c & d depicting early flowering, a typical response to
nutrient stress.
a b
d c
14
QTL detection for SFW, SDW, PCONC, PCONT, PUtE, sulphate and
oxalate concentration
Significant QTLs were mapped only on SFW, PCONC, sulphate and oxalate . There were no
significant QTLs mapped for SDW, PCONT and PUtE. These traits had heritabilities of 65, 61,
50% (under optimal Pi) and 68, 55, 50% (under sub-optimal Pi) respectively. Appendix III and
IV give details of the genetic and physical positions of all markers used. Linkage groups are
symbolic of five A. thaliana chromosomes.
QTL detection using Interval mapping
Shoot fresh weight QTL was mapped using interval mapping (see Table 5 and Figure 6). The
QTL was detected on linkage group 1 using data from sub-optimal Pi with a LOD score of
2.65 above the 2.5 threshold. This position explained 7.2% of the phenotypic variation with a
narrow-sense heritability (h2) of 71%. The C1_23381 genetic marker that mapped this QTL
resides on chromosome 1 (23381469 bp). This SFW QTL contained several genes that
suggests involvement in Pi homeostasis such as MYB DOMAIN PROTEIN 113 (MYB113) and
114 (MYB114), ROOT HAIR DEFECTIVE6 (RHD6) and a negative regulator of gibberellic acid
responses (RGL1). MYB113 and MYB114 are transcription regulators in anthocyanin
biosynthesis. RGL1 regulates transcription of superoxide dismutases that scavenge reactive
oxygen species that accumulate in response to stress (TAIR website, AT1G66350). No
significant QTLs were found on SFW data from optimal Pi. AFR19 (GWAS candidate) was
found on a small peak at the beginning of chromosome 1 but the LOD score was
considerably far from the threshold.
A QTL for phosphate concentration was found on linkage group 5 with a LOD score of 2.59,
slightly above the threshold (Table 5 and Figure 6). This was found using data from optimal
Pi conditions. C5_20318 mapped this QTL position which had an explained variance of 7.1%.
The heritability of PCONC was 56% for both optimal and sub-optimal Pi. A small QTL peak
(LOD score of 2.35) was observed on linkage group 3 mapping with data from sub-optimal Pi.
Although this position is not above the threshold, this region mapped by C3_01901 marker is
in proximity to IPS1 locus. More so GALACTINOL SYNTHASE 9 (GATL4) was found very close
to C3_01901. However the association of IPS1 is thinly regarded as C3_02968 which is
exactly at the IPS1 locus (3050989- 3051530 bp) had a very weak LOD score of 0.62.
Sulphate concentration QTL was found on the same position for both optimal and sub-
optimal Pi. This QTL was mapped by C1_22181 as only the nearest marker with LOD scores
of 2.93 and 8.92 for sub-optimal and optimal Pi conditions respectively. This position
explained 7.9 and 22.3% of the total variation with heritability of 66 and 89% respectively.
C1_23381, a marker next to C1_22181, was also found to be significantly associated with the
QTL with LOD scores of 2.61 (sub-optimal Pi) and 8.03 (optimal Pi). The latter marker also
mapped SFW as shown in Table 5. A small QTL peak was also mapped by C3_20729 with a
LOD score of 2.31. This marker is associated with SULFATE UTILIZATION EFFICIENCY 4 (SUE4),
which is involved in response to sulphate starvation (Wu et al. 2010). A QTL for oxalate was
15
mapped by C5_05319 with 3.75 LOD score. The QTL explained 10.1% of the total variance
with a highest heritability of 89%.
Table 5: QTL detection for SFW, PCONC, sulphate and oxalate concentrations using
Interval mapping.
Trait LOD score LGa %ExpVar
b Additive Gen.Pos
c Locus Phys.Pos
d h
2(%)
SFW
(-Pi)
2.65 1 7.2 0.0388 73.200 C1_23381 23381469 71
PCONC
(+Pi)
2.59 5 7.1 -0.5119 63.600 C5_20318 20318239 56
PCONC
(-Pi)
2.35e 3 6.5 0.1675 3.800 C3_01901 1901320 56
Sulphate
(-Pi)
2.93 1 7.9 -1.3291 71.300 **C1_22181 22181333 66
2.61 1 7.1 -1.2163 73.200 C1_23381 23381469
2.31 3 6.3 -1.1398 63.400 C3_20729 20728878
Sulphate
(+Pi
8.92 1 22.3 -2.7719 71.300 **C1_22181 22181333 89
8.03 1 20.3 -2.5578 73.200 C1_23381 23381469
Oxalate
(+Pi)
3.75 5 10.1 0.0700 15.800 **C5_05319 5319287 68
*Cofactors **closest marker a is linkage group,
b is % explained variance,
c is genetic position,
d is physical position (bp),
e is a small peak
near GWAS candidate gene locus
a. b.
c. d.
0.0
1.0
2.0
3.0
0 20 40 60 80 100
LOD
sco
re
Genetic Pos ition
Group 1
0.01.02.03.0
c1_0059
3c1
_0221
2c1
_0299
2c1
_0417
6c1
_0559
3
c1_0838
5c1
_0978
2
c1_1229
5c1
_1386
9c1
_1563
4c1
_1687
5
c1_1843
3c1
_1947
8c1
_2038
4c1
_2218
1c1
_2338
1c1
_2479
5c1
_2569
8c1
_2699
3c1
_2845
4c1
_2989
8LO
D s
core
Markers
Group 1
0.0
1.0
2.0
3.0
0 10 20 30 40 50 60 70 80 90
LO
D s
core
Genetic Pos ition
Group 5
0.01.02.03.0
c5_
00
576
c5_
01
587
c5_
02
900
c5_
04
011
c5_
05
319
c5_
06
820
c5_
07
442
c5_
09
082
c5_
10
428
c5_
13
614
c5_
16
368
c5_
19
316
c5_
20
318
c5_
21
319
c5_
22
415
c5_
23
116
c5_
24
997
c5_
26
671L
OD
sco
re
Markers
Group 5
16
e. f.
g. h.
i. j.
k. l.
Figure 6: QTL detection for SFW (a, b), PCONC(c to f), sulphate (g to j) and oxalate (k, l)
concentration using interval mapping. On the left are genetic positions of QTLs and on the
right are markers associated with QTLs. Marked by a dashed line is the 2.5 LOD score
threshold.
0.0
1.0
2.0
3.0
0 10 20 30 40 50 60 70
LO
D s
core
Genetic Pos ition
Group 3
0.01.02.03.0
c3_
00
580
c3_
00
885
c3_
01
901
c3_
02
968
c3_
06
631
c3_
08
042
c3_
09
748
c3_
11
192
c3_
12
647
c3_
15
714
c3_
17
283
c3_
18
180
c3_
20
729
c3_
22
147L
OD
sco
re
Markers
Group 3
0.0
1.0
2.0
3.0
0 20 40 60 80 100
LO
D s
core
Genetic Pos ition
Group 1
0.01.02.03.0
c1_
00
593
c1_
02
212
c1_
02
992
c1_
04
176
c1_
05
593
c1_
08
385
c1_
09
782
c1_
12
295
c1_
13
869
c1_
15
634
c1_
16
875
c1_
18
433
c1_
19
478
c1_
20
384
c1_
22
181
c1_
23
381
c1_
24
795
c1_
25
698
c1_
26
993
c1_
28
454
c1_
29
898L
OD
sco
re
Markers
Group 1
0
5
10
0 20 40 60 80 100
LO
D s
core
Genetic Pos ition
Group 1
0
5
10
c1_
005
93
c1_
022
12
c1_
029
92
c1_
041
76
c1_
055
93
c1_
083
85
c1_
097
82
c1_
122
95
c1_
138
69
c1_
156
34
c1_
168
75
c1_
184
33
c1_
194
78
c1_
203
84
c1_
221
81
c1_
233
81
c1_
247
95
c1_
256
98
c1_
269
93
c1_
284
54
c1_
298
98LO
D s
core
Markers
Group 1
0
1
2
3
4
0 10 20 30 40 50 60 70 80 90
LO
D s
core
Genetic Pos ition
Group 5
01234
c5_
005
76
c5_
015
87
c5_
029
00
c5_
040
11
c5_
053
19
c5_
068
20
c5_
074
42
c5_
090
82
c5_
104
28
c5_
136
14
c5_
163
68
c5_
193
16
c5_
203
18
c5_
213
19
c5_
224
15
c5_
231
16
c5_
249
97
c5_
266
71LO
D s
core
Markers
Group 5
17
QTL detection using Multiple QTL mapping (MQM)
MQM analysis detected an additional QTL for PCONC close to 2.5 LOD score as shown in
Table 6. This was achieved by using C5_20318 as a cofactor. This marker had the highest LOD
score in mapping PCONC with interval mapping. Marking C5_20318 as a cofactor increased
the LOD score of C5_20318 from 2.59 to 2.97 with an increased explained variance of 7.7%.
Moreover the LOD score of C2_05588 improved from 1.93 to 2.32. C2_05588 is located on
chromosome 2 (5587702 bp) and is the only nearest marker to GLYCEROL-3-PHOSPHATE
PERMEASE 5 (G3PP5) a gene reported to be induced by Pi starvation (Ramaiah et al, 2011).
Figure 7 shows the QTLs that were found using MQM.
Table 6: QTL detection for PCONC using Multiple QTL mapping
Trait LOD score LGa %ExpVar
b Additive Gen.Pos
c Locus Phys.Pos
d h
2(%)
PCONC
(+Pi)
2.97 5 7.7 -0.5329 63.600 C5_20318* 20318239 56
2.32e 2 6.6 0.4634 8.300 C2_05588 5587702 56
*Cofactors a is linkage group,
b is % explained variance,
c is genetic position,
d is physical position (bp),
e is a small peak
near GWAS candidate gene locus.
a. b.
c. d.
Figure 7: QTL detection for Pi concentration (PCONC) using multiple QTL mapping. a & b
show genetic positions and markers associated with the QTL. b & c show an additional QTL
for PCONC identified by MQM. Marked by a dashed line is the 2.5 LOD score threshold.
0.0
1.0
2.0
3.0
0 10 20 30 40 50 60 70 80 90
LOD
sco
re
Genetic pos ition
Group 5
0.01.02.03.0
c5_
00
576
c5_
01
587
c5_
02
900
c5_
04
011
c5_
05
319
c5_
06
820
c5_
07
442
c5_
09
082
c5_
10
428
c5_
13
614
c5_
16
368
c5_
19
316
c5_
20
318
c5_
21
319
c5_
22
415
c5_
23
116
c5_
24
997
c5_
26
671L
OD
sco
re
Markers
Group 5
0.0
1.0
2.0
3.0
0 10 20 30 40 50 60 70 80 90
LOD
sco
re
Genetic Pos ition
Group 2
0.01.02.03.0
c2_
00
593
c2_
02
365
c2_
03
041
c2_
05
588
c2_
06
655
c2_
07
650
c2_
10
250
c2_
11
457
c2_
12
435
c2_
13
472
c2_
15
252
c2_
16
837
c2_
17
606
c2_
18
753L
OD
sco
re
Markers
Group 2
18
Collocating QTLs for SFW
The QTL that had a LOD score of 2.65 had a peak demarcated by markers c1_18433 and
c1_24795 (see Figure 6). The peak of this QTL collocated with a SNP of -log10(P-value) score
of 4 (Figure 8). There were also two SNPs of -log10(P-value) of 3.8 and 3.63, close to the 4.0
threshold. Two genes that can be interesting for this study collocated with this QTL peak
and these are 5’- ADENYLYLPHOSPHOSULFATE REDUCTASE (APSR) (AT1G62180) and
GALACTINOL SYNTHASE 7 (AT1G60450) (Figure 8). There were also several genes towards
the end of QTL peak (c1_24795) though highest SNP in this region had a -log10(P-value)
score of 3.5. This region contain several genes such as RHD6, (AT1G66470), MYB90
(AT1G66390), MYB113 (AT1G66370), and RGL1 (AT1G66350). However these genes were
not associated with SNPs of significant -log10(P-value) scores (Figure 8c).
a. 5- adenylylphosphosulfate reductase b. Galactinol synthase 7
c. RHD6
Figure 8: Collocating SFW QTL demarcated by c1_18433 and c1_24795 markers with GWAS
output. The y-axis is the -log10(P-value) score with a score of 4 as threshold. The x-axis is
a zoom-in of chromosomal positions in base pairs. Green dots represent single nucleotide
polymorphisms (SNPs).
19
Collocating QTL for PCONC
The small QTL peak of 2.35 LOD score mapped by C3_01901 collocated with GWAS with a
SNP of -log10(P-value) score of 4. This SNP was found to be in the GATL4 locus (Figure 9).
One protein phosphatase (AT3G06270) also associated with this SNP. MGD3 was associated
with a SNP of 3.7 score close to 4 threshold. G3PP5 was not associated with a SNP of
significance. However, MGD3 and G3PP5 are relatively close to each other, separated by
about 2000 kilo base pairs (Figure 9). UDP-sugar pyrophosphorylase (USP) was found in
collocation with a SNP of -log10(P-value) score of 4.
a. GATL4 b. MGD3
c. G3PP5 d. USP
Figure 9: Collocating PCONC QTLs with GWAS output. The y-axis is the -log10(P-value)
score with a score of 4 as threshold. The x-axis is a zoom-in of chromosomal positions in
base pairs. Blue and red dots represent single nucleotide polymorphisms (SNPs).
Collocating QTLs for Sulphate and Oxalate
The LOD scores of QTLs for sulphate and oxalate concentrations were relatively higher than
for SFW and PCONC. However it was not possible to pin-point the exact QTL position as
C1_22181 and C5_05319 were not associated with the QTL peak. However collocation was
done using gene loci close to these markers. SNPs in these loci were not significant.
20
Co- expression analysis
Co-expression analysis using GeneMania (www. genemania.org) was done to check the
simultaneous expression of genes, and may suggest the involvement of mapped QTLs in
phosphate utilization efficiency. MONOGALACTOSYL-DIACYLGLYCEROL SYNTHASE 3 (MGD3)
is highly co-expressed with several genes that are involved in Pi homeostasis (Figure 10). It is
co-expressed with an acid phosphatase (AT1G67600), SPX DOMAIN GENE 3 (ATSPX3),
phospholipase (PLDZETA2), SULFOQUINOVOSYLDIACYLGLYCEROL 1 and 2 (SQD1 and SQD2)
that are involved in cellular response to Pi starvation (Liu et al. 2011; Morcuende et al. 2007;
Misson et al. 2005). ATSPX3 is involved in Pi ion transport and galactolipid biosynthesis (Liu
et al. 2011) where as SQD1 and SQD2 in sulfo-, galacto- and glycolipid biosynthesis. These
genes are expressed in shoots, suggesting role in phosphate utilization than in acquisition.
Other candidates such as GATL4, Pht1;4 and SULTR2;1 are highly co-expressed with other
genes in their families (data not shown). For instance, Pht1;4 is co-expressed with other
phosphate transporters such as Pht3;2, GATL4 with GATL7, SULTR2;1 with SULTR1;1.
Figure 10: Co-expression of MGD3 (MGDC) with AT1G67600, PLDZETA2, ATSPX3 and SQD1,
SQD2. Co-expression is indicated by purple links. The thicker the line, the more co-
expressed the genes in the link are. The co-expression and the role of these genes in
cellular response to Pi deficiency suggest the involvement of MGD3 in phosphate
utilization efficiency.
21
Gene expression analysis
Determining Primer efficiency
As described in Materials and Methods, the primer efficiencies were determined using a
cDNA supermix prepared by pooling 5 µl cDNA from each sample. Five microliters of the
pooled cDNA was diluted with nuclease free water four times as follows: 10, 100, 1000 and
10000 times. Each dilution had four technical replications. However the 10 times dilution
factor worked only for IPS1, Pht 1;4 and SULTR2;1 primers with efficiencies of 122% , 97.2%
and 121.2% respectively (Fig 11). ARF19, MYB111 and RHD6 primers were tested again using
a lower dilution factor: undiluted, 2, 4, and 8 times. The efficiencies improved to 112.8%,
132% and 111.6% respectively (Figure 12).
a. b.
c.
Figure 11: Primer efficiencies of a. IPS1 b. Pht1;4 c. SULTR2;1 using 10 times dilution. Cq
values (y- axis) plotted against dilution series (x- axis). The green dots inside the graph
indicate the technical replications of each cDNA dilution.
22
a. b.
c.
Figure 12: Primer efficiencies of a. ARF19 b. MYB111 c. RHD6 using 2 times dilution. Cq
values (y- axis) plotted against dilution series (x- axis). The green dots inside the graph
indicate the technical replications of each cDNA dilution.
23
Expression of IPS1, Pht1;4, SULTR2;1, ARF19, MYB111 and RHD6
Given in Table 7 are non-normalized expression means in Cq values directly from the qPCR.
UBQ10 is the reference and all comparisons are between Col-0 and Sha grown on both
optimal and sub-optimal conditions. Comparisons are also between two different time
points (21 and 28 days after sowing (DAS). Cq values are inversely proportional to the
quantity of gene expression but to check for errors incurred during qPCR, the final result is
only relative to the expression of UBQ10, hence normalized. Figure 13 shows the normalized
expression ratios of GWAS candidate genes for PUtE calculated using the 2 -∆∆Cq
method. The
expression ratios are relative to Col-0 grown under optimal Pi, harvested at 21 days
(calibrator sample), which is shown at the beginning of the x-axis in (Figure 13). The data
presented are means of 2-∆∆Cq
of two biological replicates ± standard error.
Table 7: Non-normalized gene expression data (in Cq Values ) for Col-0 and Sha under
optimal and sub-optimal conditions
Accession Growing conditions Cq Values
IPS1 Pht1;4 ARF19 MYB111 SULTR2;1 RHD6 UBQ10
Col 21DAS optimal Pi 34.15 27.82 29.95 34.32 30.90 37.27 32.30
Sha 21DAS optimal Pi 35.93 27.18 28.64 31.16 30.08 37.87 31.67
Col 21DAS sub-optimal Pi 22.67 21.55 29.14 29.38 29.49 33.80 30.57
Sha 21DAS sub-optimal Pi 29.58 25.49 29.26 31.28 30.58 34.54 30.99
Col 28DAS optimal Pi 32.68 26.84 28.71 28.35 27.61 36.62 30.10
Sha 28DAS optimal Pi 36.16 26.93 30.38 28.83 28.49 37.27 31.92
Col 28 DAS sub-optimal Pi 26.53 22.72 29.96 29.50 27.95 36.20 30.67
Sha 28DAS sub-optimal Pi 21.62 22.47 30.27 31.10 30.17 36.23 31.36
1.060.19
861.08
9.58
0.74 0.20
103.34
3082.75
0.10
1.00
10.00
100.00
1000.00
10000.00
Col21DSha21DCol21DSha21DCol28DSha28DCol28DSha28D
EX
pre
ssio
n r
ati
o
IPS1 Expressiona
24
Figure 13: Relative expression of IPS1, Pht1;4, SULTR2;1, ARF19, MYB111 and RHD6 in Col-
0 and Sha grown on both optimal and sub-optimal Pi conditions. Expression is relative to
Col grown on optimal Pi, harvested at 21 DAS. Data presented with error bars are means
of 2-∆∆Cq
of two biological replicates ± standard error. Blue and yellow bars indicate
optimal and sub-optimal Pi respectively. 21 and 28D are short for 21 and 28 days after
sowing respectively.
1.02 1.14 0.80 0.51
2.18
4.38
3.52
0.87
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
Ex
pre
ssio
n r
ati
o
SULTR2;1 Expressionc
1.02
1.60
0.530.65
0.52 0.570.35 0.42
0.00
0.50
1.00
1.50
2.00
Ex
pre
ssio
n r
ati
o
ARF19 Expressiond
1.925.78
9.25
3.32
13.76
34.54
14.35
4.86
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
Ex
pre
ssio
n r
ati
o
MYB111 Expressione
1.05
0.43
3.35
2.69
0.34
1.230.69
1.08
0.00
1.00
2.00
3.00
4.00
Ex
pre
ssio
n r
ati
o
RHD6 Expressionf
1.01 1.01
23.26
2.030.43 1.43
11.65
21.26
0.00
5.00
10.00
15.00
20.00
25.00
Ex
pre
ssio
n r
ati
oPht1;4 Expressionb
25
The difference in expression of IPS1 between Col-0 and Sha was smaller than the difference
between optimal and sub-optimal Pi, with bars marked in blue and yellow respectively. This
was observed across all time points. Col-0 and Sha accessions did not germinate in all
replications, otherwise it would be more informative to see the trend of IPS1 expression at
14 DAS. There was a large amount of variation in the expression of IPS1. However the trend
remained unchanged; in plants that were grown on sub-optimal Pi, IPS1 was expressed
almost 10-fold and above (see Figure 12). Data labels for expression ratios are given on bar
tops. At 21 DAS, IPS1 was expressed more in Col-0 than in Sha and vice versa at 28 DAS.
The expression pattern of Pht1;4 assumed the same trend as of IPS1. Pht1;4 was expressed
almost 25-fold in Col-0 on sub-optimal Pi than the same accession on optimal Pi at 21 DAS.
Sha on the same conditions showed 2-fold Pht1;4 expression than in the calibrator sample.
Pht1;4 expression increased considerably in Sha on low Pi, attaining almost double the
expression of Col-0 at 28 DAS. However both accessions showed higher expression of Pht1;4
on sub-optimal Pi than on optimal Pi.
Plants grown on optimal Pi assumed more expression of SULTR2;1 than plants on sub-
optimal Pi. The trend remained unchanged across all time points. However the expression
ratios of SULTR2;1 at 28 DAS (low Pi) are slightly larger than at 21 DAS considering the
endpoint of the negative error bar on Col-0 28 DAS, sub-optimal Pi.
ARF19 expression declined as plants grew from 21 DAS to 28 DAS. Although the difference in
expression ratios were not large, plants grown on optimal Pi always had higher expression
ratios than of those grown on sub-optimal Pi. At 28 DAS, Col-0 and Sha on sub-optimal Pi
assumed about 150 percent less expression than Col-0 at 21 DAS. It is also shown in all
cases, that Sha had higher levels of ARF19 transcripts than Col-0.
MYB111 assumed no clear pattern between either accessions or time points. More so with a
large error bar on Col-0 28 DAS for sub-optimal Pi, nothing much can be said on this gene.
RHD6 was expressed more than 2.5 fold on sub-optimal Pi in both accessions, than optimal
Pi at 21 DAS. However at 28 DAS the expression ratios declined. It is important to note that
the Cq values of RHD6 were the lowest amoung all genes analysed.
26
4. DISCUSSION
QTL Detection: SFW, SDW, PCONC, PCONT, PUtE, sulphate and oxalate
concentration
Fewer than expected QTLs were mapped for all the phenotyped traits. This could have been
caused by several factors. Firstly, quantitative traits are polygenic with each QTL contributing
in small effect to the overall trait. High heritability is essential to efficiently map the QTLs
and the heritabilities for this study are shown in Chapter 3. PUtE had the lowest heritability
of 50% and might have resulted in failure to directly map the PUtE QTLs. However high
heritability with few or no QTLs (for instance SDW), may indicate that many genes are
involved in regulation of the trait, with each gene contributing a considerably small effect to
the overall trait variation. More so the frequency distribution curve for PUtE under optimal
Pi was not normal (see Figure 4). I suppose that the formula for calculating PUtE needs more
attention. The assumption that the amount of Pi measured in a gram of dry weight, is the
amount that is being utilized to produce one gram of dry weight sounds improper. More
focus should be channelled towards devising a method or formula to correctly measure
PUtE, because when one calculates PUtE as 1/PCONC, then all the highly efficient genotypes
become the less efficient ones. I feel that most of the genotypes that are regarded highly Pi-
efficient today were defined based on phosphate acquisition efficiency (PAE), which
somehow contradicts with the goal of this study. Although the difference between PAE and
PUtE as determinants of PUE can be dissected, the difference seems to be small. For
instance, QTLs that are found to be involved in PAE are also found in PUtE. Another reason
that could have reduced the mapping resolution is of less replicates. Although four replicates
were sown, only two replicates were phenotyped for QTL mapping. For PCONC
measurements, the genotypes differed into which extend one could grind the leaves. For
instance, genotypes that had mealy-type leaves, could be ground to fine powder but some
were so plastic that fine powder could not be attained.
Several QTLs mapped for SFW, SDW and PCONC manifest involvement in phosphate
utilization efficiency. These traits are indispensable as they explain to which extend a
genotype is efficient in phosphate utilization. SFW and SDW are yield components and in any
form of biotic or abiotic stress, the overall goal is to produce high yield. This enables
extrapolation of the essence of the mapped QTLs to those of PUtE as described below:
The QTL mapped for SFW under sub-optimal Pi conditions was relatively large with so many
genes. Although this QTL is for SFW, there are several genes that suggest a role in Pi
homeostasis. RHD6, MYB113 and RGL1, were found towards the end of the QTL region. As
described in Chapter 2, RHD6 is quite promising as it promotes root hair development.
MYB113 and MYB114 are transcription factors in the anthocyanin biosynthetic pathway
(Warner, unpublished data) and are indispensable as anthocyanin production is one of the
signs of Pi deficiency. It is however not clear to me whether they are directly involved in
regulation of MYB111, one of the GWAS candidates. RGL1, a negative regulator of
27
gibberellic acid responses, is also interesting for this study as it is suggested to take part in
reduction of reactive oxygen species (ROS) that are produced when plants are under stress.
This is achieved by up-regulation of superoxide dismutases that scavenge the ROS.
Superoxide dismutases were found to be up-regulated in Pi-starved Arabidopsis (Mission et
al. 2005). AFR19 was found in the small peak at the beginning of chromosome 1. Considering
the insignificant LOD score, its role in phosphate utilization efficiency is inconclusive. 5’-
adenylylphosphosulfate reductase (APSR) and GALACTINOL SYNTHASE 7 collocated with this
SFW QTL. The former is involved in sulphate assimilation and is known to influence the
variation of sulphate in shoots (Lee et al. 2011). GALACTINOL SYNTHASE 7 plays a role in
carbohydrate metabolism by encoding galactinol which was reported to accumulate when
plants are subjected to drought, salinity and cold-treatment stress (Taji et al. 2002). Recently
galactinol was reported to function as an osmoprotectant in the above mentioned abiotic
stress conditions (Nishizawa et al. 2008). The remaining question is whether these two genes
are important for phosphate utilization efficiency and its correlation with other candidates.
Mutation of these genes probably by Agrobacterium-mediated T-DNA gene knockout
resulting in disruption of the coding DNA sequence may elaborate whether they are key
players in phosphate utilization or not.
Several genes were found on chromosome 2, 3, and 5, on mapped QTLs for PCONC.
GALACTINOL SYNTHASE 9 (GATL4), IPS1, G3PP5, MGD3, UDP-SUGAR PYROPHOSPHORYLASE
(USP). GATL4, IPS1, G3PP5, MGD3 have been reported in several studies to be involved in Pi
homeostasis (Nishizawa et al. 2008; Ramaiah et al. 2011; Kobayashi et al. 2008; Misson et al.
2005). GALT4 has similar function as GALACTINOL SYNTHASE 7 described above. G3PP5 is
known to be involved in Pi transport (Ramaiah et al. 2011). MGD3 has been shown to be up-
regulated when plants are subjected to Pi deficiency (Misson et al. 2005; Kobayashi et al.
2008; Li et al. 2006). It is in involved in galactolipid metabolism suggesting the use of
galactolipids in the interim, to replace phospholipids when Pi is insufficient. MGD3 suggests
an important role in phosphate utilization efficiency because efficient genotypes may enable
the up-regulation of these genes so as to synthesize non-phosphorus metabolites for use in
lipid biosynthesis. More so MGD3 is expressed only in shoots and highly co-expressed with
SQD1 and 2, ATSPX3 and PLDZETA2 that are induced in response to Pi starvation (Liu et al.
2011; Morcuende et al. 2007; Misson et al. 2005). USP functions in nucleotide sugar
metabolism as shown by Kotake et al. 2007. However its direct involvement in phosphate
utilisation efficiency is still unclear.
28
Gene expression analysis
The large difference in expression of IPS1 shown in Chapter 3 confirm the up-regulation of
this gene in plants grown on sub-optimal Pi conditions. IPS1 is shown to be induced even
under mild Pi stress following studies that showed IPS1 induction when plants are subjected
to complete Pi starvation (Misson et al. 2005). The clear-cut between expression ratios of
optimal and sub-optimal Pi on Col-0 and Sha is suggesting that IPS1 is up-regulated when the
plants signal Pi-limiting conditions. At 21 DAS, IPS1 is seen to have a large transcriptome
already and further increases from 21 DAS to 28 DAS. Although no Col-0 and Sha samples
harvested on 14 DAS, the trend suggests early involvement of this gene when Pi is limiting.
IPS1 resides on chromosome 3, (3050989 - 3051530 bp) and is within a small QTL peak
mapped in Chapter 3. Although this region is not seen to be significantly associated with
IPS1, its induction on sub-optimal Pi, suggests its influence on phosphate utilization
efficiency. The induction of IPS1 may signals impending Pi stress conditions, and the
subsequent metabolic changes prepare efficient genotypes for this predicament. Pht1;4
expression follow a similar expression trend as IPS1. Pht1;4 belongs to a Pht1 family of
phosphate transporters, that are involved in phosphate ion transport. As phosphate is
essential for plant growth, Pht1;4 remains essential even when Pi is adequate. However the
gene seems to be up-regulated in Pi stress conditions. Shin et al. 2004 suggests the
involvement of Pht1;4 in Pi acquisition as confirmed by high transcript levels in roots. Since
no root tissues were analysed for Pht1;4, I suggests its role in remobilization of Pi from
storage organs to growing tissues. In this regard, it could have been interesting to know
whether Pht1;4 was expressed differently in leaves of different ages.
SULTR2;1 assumed more expression in adequate Pi than in Pi limiting conditions. Recent
studies have shown up-regulation of SULTR2;1 in Pi-starvation (Rouached et al. 2011; Misson
et al. 2005). These studies revealed the flux of sulphate from shoots to roots when Pi is
limiting. However the expression of this gene in the shoot is not symbolic of sulphate stress
(TAIR website, ATG10180). If sulphate cushions Pi when limiting, it sounds ideal to suggest
that SULTR2;1 up-regulation in shoots has nothing to do with Pi-stress response. Analysis of
SULTR2;1 expression pattern on both roots and shoots of Col-0 and Sha subjected to Pi
limiting conditions may help to determine whether SULTR2;1 is actually down-regulated in
shoots of plants subjected to P-stress. The differences between Col-0 and Sha in expression
levels of ARF19 in all cases can be an indication of differences in nutrient stress perception
between these accessions. Sha seems to have a higher ARF19 expression than Col-0. This
tallies with its reported high PUE efficiency since ARF19 expression enable lateral root
development. This further evidence the proposition that genotypes that are regarded as
highly Pi efficient today were annotated only based on Pi acquisition efficiency. However
ARF19 expression declined from 21 DAS to 28 DAS which contradicts with its induction under
long term nutrient stress conditions. Unexpectedly plants grown on optimal Pi retain more
ARF19 transcripts than on sub-optimal Pi.
29
MYB111 expression assumes no clear expression pattern and in this regard nothing can be
deduced. In comparison to the calibrator sample, RHD6 was expressed more than 2.5-fold in
Col-0 and Sha grown on sub-optimal than on optimal Pi conditions. This indicates the
induction of RHD6 in promoting root hair development when Pi is limiting. Although RHD6
transcripts declined on 28 DAS, the large error bar on Sha 28 DAS (optimal Pi) suggests a
similar trend of more RHD6 expression on sub-optimal than on optimal Pi. Nevertheless the
transcriptome of RHD6 was the least amoung all gene analysed. Probably this is because
RHD6 is involved in root hair development and could have been expressed more in roots.
Considering the overall PUE, RHD6 remains an interesting candidate as it is centred on root
development, the only medium for Pi acquisition.
30
5. CONCLUSIONS AND RECOMENDATIONS
PUtE is polygenic and dissecting its genetic basis is cumbersome. Although no QTLs were
mapped directly on PUtE data, its manifestation in traits such as SFW, SDW and PCONC
enables direct extrapolation of the essence of the mapped QTLs to those of PUtE. In this
regard, several QTLs for SFW and PCONC mapped suggests involvement in phosphate
utilization efficiency. Genome wide association mapping (GWAS) candidates such as IPS1,
ARF19 and RHD6 were mapped, although the LOD scores were not convincingly high.
Moreover several QTLs mapped in this study are also related to phosphate acquisition
efficiency. Disrupting the functions of these QTLs help to confirm their involvement in
phosphate utilization efficiency. Although the genetic basis of PAE and PUtE can be
dissected, these two components of PUE seem to be strongly correlated. To efficiently
determine PUtE, the 1/PCONC formula needs to be scrutinized.
IPS1 is induced even in mild Pi-stress conditions. It is important in PUtE since metabolic
changes brought by its induction prepare highly Pi efficient genotypes to thrive under Pi
deficient conditions. Pht1;4 is up-regulated in shoots on sub-optimal Pi conditions. It might
be involved in Pi remobilization from storage organs to growing points. SULTR2;1 is down
regulated in shoots in Pi-stress. Sha expresses ARF19 more than Col-0 and is still unclear
whether MYB111 is directly involved in PUtE. RHD6 is slightly expressed in shoots.
Physiological differences brought by differences in expression of these PUtE promising
candidates is the missing link.
31
REFERENCES Alonso-Blanco C. & Koornneef M. (2000) Naturally occuring variation in Arabidopsis: an
underexploited resource for plant genetics. Trends in Plant Science 5: 22–29
Alva A.K., He Z.L., Li Y.C., Calvert D.V. & Banks D.J. (1999) Sorption-Desorption and Solution
Concentration of Phosphorus in a Fertilized Sandy Soil. Journal of Environmental Quality 28: 1804-1810
Bariola P.A., Howard C.J., Taylor C.B., Verburg M.T., Jaglan V.D. & Green P.J. (1994) The
Arabidopsis ribonuclease gene RNS1 is tightly controlled in response to phosphate limitation. The Plant
Journal 6: 673-685
Bergelson J. & Roux F. (2010) Towards identifying genes underlying ecologically relevant traits in
Arabidopsis thaliana. Nature Reviews Genetics 11: 867-879
Czechowski T., Bari R.P., Stitt M., Scheible W.-R. & Udvardi M.K. (2004) Real-time RT-PCR
profiling of over 1400 Arabidopsis transcription factors, unprecedented sensitivity reveals novel root- and
shoot-specific genes. Plant Journal 38: 366–379
Good A.G., Shrawat A.K., & Muench D.G. (2007) Can less yield more? Is reducing nutrient input into
the environment compatible with maintaining crop production? Trends in Plant Science 9: 597-605
Hamburger D., Rezzonico E., Petétot J.M.C., Somerville C. & Poirier Y. (2002) Identification and
characterization of the Arabidopsis PHO1 gene involved in phosphate loading to the xylem. The Plant Cell
14: 889–902
Jain A., Vasconcelos M. J., Raghothama K. G. & Sahi S. V. (2007) Molecular Mechanisms of Plant
Adaptation to Phosphate Deficiency. In: Plant Breeding Reviews 29: 359-419, John Wiley & Sons, USA
Jiang C., Gao X., Liao L., Harberd N.P., & Fu X. (2007) Phosphate Starvation Root Architecture and
Anthocyanin Accumulation Responses Are Modulated by the Gibberellin-DELLA Signaling Pathway in
Arabidopsis. Plant Physiology 145: 1460–1470
Kawashima C.G., Matthewman C.A., Huang S., Lee B.R., Yoshimoto N., Koprivova A., Rubio-
Somoza I., Todesco M., Rathjen T., Saito K., Takahashi H., Dalmay T. & Kopriva S. (2011)
Interplay of SLIM1 and miR395 in the regulation of sulfate assimilation in Arabidopsis. Plant Journal 66:
863-76.
Kobayashi K., Awai K., Nakamura M., Nagatani A., Masuda T.& Ohta H. (2008) Type-B
monogalactosyldiacylglycerol synthases are involved in phosphate starvation-induced lipid remodeling,
and are crucial for low-phosphate adaptation. The Plant Journal 57: 322–331
Kloth K.J., Thoen M.P.M., Bouwmeester H.J., Jongsma M.A. & Dicke M. (2012) Association
mapping of plant resistance to insects. Trends in Plant Science 17: 311-319
Koornneef M., Alonso-Blanco C., & Reymond M. (2011) Natural Variation in Arabidopsis thaliana.
In: Genetics and Genomics of the Brassicaceae. Plant Genetics and Genomics: Crops and Models 9, New
York, USA. Pp. 123-151
Kotake T., Hojo S., Yamaguchi D., Aohara T., Konishi T. & Tsumuraya Y.(2007) Properties and
Physiological Functions of UDP-Sugar Pyrophosphorylase in Arabidopsis. Bioscience, Biotechnology and
Biochemistry 71: 761-771
Lee B. R., Koprivova A. & Kopriva S. (2011) The key enzyme of sulfate assimilation, adenosine 5'-
phosphosulfate reductase, is regulated by HY5 in Arabidopsis. The Plant Journal 67:1042-54
Li M., Welti R., & Wang X. (2006) Quantitative Profiling of Arabidopsis Polar Glycerolipids in Response
to Phosphorus Starvation. Roles of Phospholipases Dz1 and Dz2 in Phosphatidylcholine Hydrolysis and
Digalactosyldiacylglycerol Accumulation in Phosphorus-Starved Plants. Plant Physiology 142: 750–761
32
Liu T.Y., Huang T.K., Tseng C.Y., Lai Y.S., Lin S.I., Lin W.Y., Chen J.W. & Chiou T.J. (2012) PHO2-
Dependent Degradation of PHO1 Modulates Phosphate Homeostasis in Arabidopsis. The Plant Cell 24:
2168–2183
Ma X.F., Tudor S., Butler T., Ge Y., Xi Y., Bouton J., Harrison M. & Wang Z.Y. (2011) Transgenic
expression of phytase and acid phosphatase genes in alfalfa (Medicago sativa) leads to improved
phosphate uptake in natural soils. Molecular Breeding 30: 377–391
Martin A.C., del Pozo J.C., Iglesias J., Rubio V., Solano R., de La Peña A, Leyva A. & Paz-Ares J.
(2000) Influence of cytokinins on the expression of phosphate starvation responsive genes in
Arabidopsis. The Plant Journal 24: 559-567
Misson J., Raghothama K.G., Jain A., Jouhet J., Block M.A., Bligny R., Ortet P., Creff A.,
Somerville S., Rolland N., Doumas P., Nacry P., Herrerra-Estrella L., Nussaume L. & Thibaud M.
(2005) A genome-wide transcriptional analysis using Arabidopsis thaliana Affymetrix gene chips
determined plant responses to phosphate deprivation. Plant Biology 102: 11934-11939
Morcuende R., Bari R., Gibon Y., Zheng W., Pant B.D., Bläsing O., Usadel B., Czechowski B.,
Udvardi M.K., Stitt M. & Scheible W.R. (2007) Genome-wide reprogramming of metabolism and
regulatory networks of Arabidopsis in response to phosphorus. Plant, Cell and Environment 30: 85–112
Narang R. A., Bruene A., & Altmann T. (2000) Analysis of Phosphate Acquisition Efficiency in Different
Arabidopsis Accessions. Plant Physiology 124: 1786–1799
Nishizawa A., Yabuta Y, & Shigeoka S. (2008) Galactinol and Raffinose Constitute a Novel Function to
Protect Plants from Oxidative Damage. Plant Physiology 147: 1251-1263
Okushima Y., Fukaki H., Onoda M., Theologis A. & Tasaka M. (2007) ARF7 and ARF19 Regulate
Lateral Root Formation via Direct Activation of LBD/ASL Genes in Arabidopsis. Plant Cell: 19: 118–130
Pérez-Torres A.C., López-Bucio J., Cruz-Ramírez A., Ibarra-Laclette E., Dharmasiri S., Estelle
M. & Herrera-Estrella L. (2008) Phosphate Availability Alters Lateral Root Development in Arabidopsis
by Modulating Auxin Sensitivity via a Mechanism Involving the TIR1 Auxin Receptor. Plant Cell: 20:
3258–3272
Poirier Y., Thoma S., Somerville C.& Schiefelbein J. (1991) A mutant of Arabidopsis deficient in
xylem loading of phosphate. Plant Physiology 97:1087–109
Ramaiah M., Jain A., Baldwin J.C., Karthikeyan A. S., & Raghothama K. G. (2011) Characterization
of the Phosphate Starvation-Induced Glycerol-3-phosphate permease Gene Family in Arabidopsis. Plant
Physiology 157: 279-291
Reymond M., Svistoonoff S., Loudet O. & Nussaume L. (2006) Identification of QTL controlling root
growth response to phosphate starvation in Arabidopsis thaliana. Plant, Cell and Environment 29: 115–
125
Rose T.J. & Wissuwa M. (2012) Rethinking Internal Phosphorus Utilization Efficiency: A New Approach
Is Needed to Improve PUE in Grain Crops. Advances in Agronomy 116: 185-217
Rouached H., Arpat B.A. & Poirier Y. (2010) Regulation of Phosphate Starvation Responses in Plants:
Signaling Players and Cross-Talks. Molecular Plant 3: 288–299
Rouached H., Secco D., Arpat B. & Poirier Y. (2011) The transcription factor PHR1 plays a key role in
the regulation of sulfate shoot-to-root flux upon phosphate starvation in Arabidopsis. BMC Plant Biology
11: 1471-2229
Secco D., Baumann A., & Poirier Y.(2010) Characterization of the Rice PHO1 Gene Family Reveals a
Key Role for OsPHO1;2 in Phosphate Homeostasis and the Evolution of a Distinct Clade in Dicotyledons.
Plant Physiology 152: 1693–1704
33
Schmittgen T. D. & Livak K.J. (2008) Analyzing real-time PCR data by the comparative CT method.
Nature Protocols 3: 1101 - 1108
Shin H., Shin H.S., Dewbre G.R. & Harrison M.J. (2004) Phosphate transport in Arabidopsis: Pht1;1
and Pht1;4 play a major role in phosphate acquisition from both low- and high-phosphate environments.
The Plant Journal. 39: 629-642
Singh S.K., Fischer U., Singh M., Grebe M. & Marchant A. (2008) Insight into the early steps of root
hair formation revealed by the procuste1 cellulose synthase mutant of Arabidopsis thaliana. BMC Plant
Biology 8: 57
Stracke R., Jahns O., Keck M., Tohge T., Niehaus K., Fernie A.R. & Weisshaar B. (2010) Analysis
of PRODUCTION OF FLAVONOL GLYCOSIDES-dependent flavonol glycoside accumulation in Arabidopsis
thaliana plants reveals MYB11-, MYB12- and MYB111-independent flavonol glycoside accumulation. New
Phytologist 188: 985-1000
Syers J. K., Johnston A. E. & Curtin D. (2008) Efficiency of soil and fertilizer phosphorus use:
Reconciling changing concepts of soil phosphorus behaviour with agronomic information. Food and
Agriculture Organisation, Italy
Taji T., Ohsumi C., Luchi S., Seki M., Kasuga M., Kobayashi M., Yamaguchi-Shinozaki K. &
Shinozaki K. (2002) Important roles of drought- and cold-inducible genes for galactinol synthase in
stress tolerance in Arabidopsis thaliana. The Plant Journal 29:417-26
Vance C.P., Uhde-Stone C. & Allan D.L. (2002) Phosphorus acquisition and use: critical adaptations
by plants for securing a non-renewable resource. New Phytologist 157: 423-447
Veneklaas E.J., Lambers H., Bragg J., Finnegan P.M., Lovelock C.E., Plaxton W.C., Price C.A.,
Scheible W.R., Shane M.W., White P.J. & Raven J.A. (2012) Opportunities for improving
phosphorus-use efficiency in crop plants. New Phytologist 195: 306–320
Wang X., Wang Y., Tian J., Lim B.L., Yan X. & Liao H.(2009) Overexpressing AtPAP15 Enhances
Phosphorus Efficiency in Soybean. Plant Physiology 151: 233–240
Warner N. M. (2008) Unravelling the roles of two senescent enhanced MYB transcription factors in the
regulation of anthocyanin biosynthesis in Arabidopsis thaliana. PHD thesis, University of Warwick, UK
Wu Y., Zhao Q., Gao L., Yu X. M., Fang P., Oliver D.J. & Xiang C. B. (2010) Isolation and
characterization of low-sulphur-tolerant mutants of Arabidopsis. Experimental Botany 61: 3407-3422.
Zhang D., Cheng H., Geng L., Kan G., Cui S., Meng Q., Gai J. & Yu D. (2009) Detection of
quantitative trait loci for phosphorus deficiency tolerance at soybean seedling stage. Euphytica
167:313–322
34
APPENDICES
I. Nutrient composition used for growing plants
Nutrient Composition Unit
NH4
1.4
mmol/l
K 5.7
Na 0.2
Ca 1.9
Mg 1.2
Si < 0.01
NO3
5.7
Cl 0.2
S 3.0
HCO3 0.4
P normal/ P low 1.24 mmol/ 100 μmol
Fe
19
μmol/l
Mn 11
Zn 6.3
B 22
Cu 9.2
Mo 0.4
II. RNA extraction from Arabidopsis leaves with Qiagen ® RNeasy Plant Mini Kit
β-Mercaptoethanol (β-ME) must be added to Buffer RLT before use. For every 1 ml of Buffer
RLT add, 10 µl of β-ME.
Buffer RPE is supplied as a concentrate. For the first time of use, add 4 volumes of 96-100%
ethanol.
• Add 450 µl of Buffer RLT to a maximum of 100 mg tissue powder. Vortex vigorously.
Incubate for 3 minutes at 56oC.
• Transfer the lysate to Qiashredder spin column (lilac) placed in a 2 ml collection
tube, and centrifuge for 2 minutes at full speed.
• Carefully transfer the supernatant to a 2 ml tube without disturbing the pellet
• Add 200 µl of ethanol (96-100%) and mix immediately by pipetting
• Transfer the sample to an RNeasy spin column (pink). Centrifuge for 15 s at full
speed. Discard the flow -through
35
• Add 350 µl of Buffer RW1 and centrifuge for 15 s at full speed. Discard the flow -
through
• Perform on column DNase digestion procedure by mixing 70 µl Buffer RDD and 10 µl
DNase and pipetting it directly to the spin membrane. Incubate for 15 minutes at
room temperature
• Add 350 µl of Buffer RW1 and centrifuge for 15 s at full speed. Discard the flow -
through
• Add 500 µl of Buffer RPE and centrifuge for 15 s at full speed. Discard the flow -
through
• Add 500 µl of Buffer RPE and centrifuge for 2 minutes at full speed. Discard the flow-
through
• Place the RNeasy spin column in a new 2 ml collection tube and centrifuge for 1
minute at full speed
• Place the RNeasy spin column in a new 1.5 ml collection tube. Add 40 µl of RNase-
free water directly to the spin column membrane. Centrifuge at maximum speed for
1 minute to elute the RNA.
III. Linkage map: Genetic Positions
Group 1 cM Group 2 cM Group 3 cM Group 4 cM Group 5 cM
c1_00593 0 c2_00593 0 c3_00580 0 c4_00012 0 c5_00576 0
c1_02212 4.5 c2_02365 7.2 c3_00885 0.9 c4_00641 7 c5_01587 6.4
c1_02992 7.2 c2_03041 7.7 c3_01901 3.8 c4_02133 15.6 c5_02900 9
c1_04176 11.7 c2_04263 7.7 c3_02968 5.9 c4_03833 15.6 c5_04011 11.8
c1_05593 15.2 c2_05588 8.3 c3_04141 5.9 c4_04877 15.9 c5_05319 17.2
c1_08385 27.9 c2_06655 12.2 c3_05141 5.9 c4_05850 22.2 c5_06820 21.7
c1_09782 31.9 c2_07650 19.3 c3_06631 9.4 c4_06923 27.7 c5_07442 24.9
c1_12295 43.2 c2_10250 30.7 c3_08042 12.9 c4_07549 30.8 c5_09082 36.1
c1_13869 49.1 c2_11457 35.1 c3_09748 24.7 c4_08930 35.5 c5_10428 41.7
c1_15634 49.4 c2_12435 41.5 c3_11192 35.5 c4_10609 42.4 c5_13614 43.1
c1_16875 51.2 c2_13472 45 c3_12647 38.7 c4_11878 47.9 c5_16368 53
c1_18433 59.3 c2_15252 48.5 c3_14097 38.7 c4_13171 55.1 c5_19316 62.1
c1_19478 61.5 c2_16837 54.4 c3_15714 41 c4_14819 59.4 c5_20318 63.6
c1_20384 65.7 c2_17606 56.7 c3_17283 48.9 c4_15765 60.8 c5_21319 66.1
c1_22181 70.3 c2_18753 61.4 c3_18180 51.8 c4_18056 68.3 c5_22415 72.5
c1_23381 73.2 c3_20729 63.4 c5_23116 74.7
c1_24795 79.7 c3_22147 67.7 c5_24997 80.9
c1_25698 83.1 c5_26671 87.3
c1_26993 87.7
c1_28454 91.5
c1_29898 96.4
36
IV. Linkage map: Physical Positions (PP)
Marker chromosome PP (bp) Marker chromosome PP (bp)
c2_00593 2 592610
c1_00593 1 592939 c2_02365 2 2365070
c1_02212 1 2211522 c2_03041 2 3041384
c1_02992 1 2991901 c2_04263 2 4263347
c1_04176 1 4175761 c2_05588 2 5587702
c1_05593 1 5592874 c2_06655 2 6654616
c1_08385 1 8385006 c2_07650 2 7650461
c1_09782 1 9781624 c2_10250 2 10249585
c1_12295 1 12294768 c2_11457 2 11457144
c1_13869 1 13869179 c2_12435 2 12435201
c1_15634 1 15633764 c2_13472 2 13471565
c1_16875 1 16874540 c2_15252 2 15251700
c1_18433 1 18433370 c2_16837 2 16837474
c1_19478 1 19477618 c2_17606 2 17606352
c1_20384 1 20384267 c2_18753 2 18753024
c1_22181 1 22181333 c4_00012 4 11691
c1_23381 1 23381469 c4_00641 4 641363
c1_24795 1 24795166 c4_02133 4 2133235
c1_25698 1 25697965 c4_03833 4 3833156
c1_26993 1 26993011 c4_04877 4 4877120
c1_28454 1 28454383 c4_05850 4 5850298
c1_29898 1 29898175 c4_06923 4 6923164
c3_00580 3 580137 c4_07549 4 7549125
c3_00885 3 884847 c4_08930 4 8929959
c3_01901 3 1901320 c4_10609 4 10608576
c3_02968 3 2968064 c4_11878 4 11878132
c3_04141 3 4141103 c4_13171 4 13170826
c3_05141 3 5140894 c4_14819 4 14819373
c3_06631 3 6631399 c4_15765 4 15765120
c3_08042 3 8041859 c4_18056 4 18056361
c3_09748 3 9747530 c5_00576 5 576456
c3_11192 3 11192190 c5_01587 5 1586683
c3_12647 3 12647316 c5_02900 5 2900008
c3_14097 3 14097228 c5_04011 5 4010649
c3_15714 3 15714156 c5_05319 5 5319287
c3_17283 3 17282688 c5_06820 5 6820442
c3_18180 3 18180380 c5_07442 5 7442381
c3_20729 3 20728878 c5_09082 5 9081741
c3_22147 3 22146585 c5_10428 5 10428114
c5_13614 5 13614326
c5_16368 5 16368385
c5_19316 5 19316305
c5_20318 5 20318239
c5_21319 5 21319032
c5_22415 5 22414701
c5_23116 5 23115566
c5_24997 5 24996754
c5_26671 5 26671472