Intraspecific variation in the Populus balsamifera drought response: A systems biology approach
by
Erin T. Hamanishi
A thesis submitted in conformity with the requirementsfor the degree of Doctor of Philosophy
Faculty of ForestryUniversity of Toronto
© Copyright by Erin T. Hamanishi 2013
ii
Intraspecific variation in the Populus balsamifera drought
response: A systems biology approach
Erin T. Hamanishi
Doctor of Philosophy
Faculty of Forestry
University of Toronto
2013
Abstract
As drought can impinge significantly on forest health and productivity, the mechanisms by which
forest trees respond to drought is of interest. The research presented herein examined the intra-
specific variation in the Populus balsamifera drought response, examining the potential role of the
transcriptome to configure growth, metabolism and development in response to water deficit.
Amassing evidence indicates that different species of Populus have divergent mechanisms and Three
lines of inquiry were pursued to investigate the intraspecific variation the drought response in P.
balsamifera.
First, the transcriptome responses of six genotypes of P. balsamifera were examined using
Affymetrix Poplar GeneChips under well-watered and water-deficit conditions. A core species-
level transcriptome response was identified. Significantly, intraspecific variation in the drought
transcriptome was also identified. The data support a role for genotype-derived variation in the
magnitude of P. balsamifera transcriptome remodelling playing a role in conditioning drought
iii
responsiveness.
Second, the impact of drought-stress induced declines in stomatal conductance, as well as an
alteration in stomatal development in two genotypes was examined. Patterns of transcript
abundance of genes hypothesised to underpin stomatal development had patterns congruent with
their role in modulation of stomatal development. These results suggest that stomatal development
may play a role as a long-term mechanism to limit water loss from P. balsamifera leaves under
conditions of drought-stress.
Finally, the drought-induced metabolome of six P. balsmaifera genotypes was interrogated.
Metabolite profiling reveled amino acids such as isoleucine and proline and sugars such as galactinol
and raffinose were found with increased abundance, whereas TCA intermediates succinic and malic
acid were found with decreased abundance in response to drought. Comparative analysis of the
metabolome and the transcriptome revealed genotypic-specific variation in energy and carbohydrate
metabolism.
Taken together, the findings reported in this thesis form a foundation to understand the basis
of intraspecific variation in the drought response in P. balsamifera. Transcripts and metabolites
that contribute to within-species differences in drought tolerance were defined. These molecular
components are useful targets for both future study, as well as efforts aimed at protecting and
growing trees of this important species under challenging environmental conditions.
iv
Acknowledgments
First and foremost, I would like to extend my heartfelt gratitude to my supervisor, Dr. Malcolm
Campbell, for accepting me as his graduate student all those many years ago. I am forever grateful
for his support and guidance throughout this academic journey. Whether it was in the lab, or
walking through field trials in Oregon, or basking in the beauty of network plots, Malcolm’s
enthusiasm has been of a great inspiration to me. I would also like to thank my supervisory
committee, Dr. Nick Provart and Dr. Sean Thomas, for all their support and excellent academic
advice over the years.
The research presented in this thesis would not have been possible without help from so many
people. Over the years I have grown hundreds of poplar trees, and I am so thankful for the
technical support of Bruce Hall, John McCarron and Andrew Petrie. I am also thankful to
Matthew Hoskins for spending many hours counting epidermal and stomatal cells with me. I have
benefited from collaborating with many colleagues, including Genoa Barchet, Dr. Shawn Mansfield,
Dr. Aine Plant and Dr. Barb Thomas. I thank them for their advice, support and collaborations in
my scientific endeavors.
I would like to express my appreciation for my fellow graduate students, many of which have been
important sources of inspiration and support. Most notably, I would like to thank Liz Nelson
and Agnieszka Sztaba for their unconditional friendship and support, both academically and non-
academically. My fellow Campbell lab-mates have always been willing to provide feedback and have
been a source of excellent support and friendship over the years. To Katharina Bräeutigam, Thomas
Canam, Michael Prouse, Sherosha Raj, Julia Romano, Joseph Skaf, Michael Stokes, Heather
Wheeler and Olivia Wilkins, I extend my sincere gratitude. I am particularly thankful to Joan
Ouellette for always ensuring everything ran smoothly in the lab, and for always lending a helping
hand when needed.
Lastly, I am indebted to my family, for everything. I am grateful to my sister, Sarah, for always
being there when I needed her and for our countless Skype study dates. Finally, I would like to
thank my parents, Joan and Russel, who have provided never-ending support and encouragement
over the years, without which I would likely not be where I am today. In addition, I must also
thank them for undoubtedly instilling in me a keen sense of curiosity and love for all things
“science.”
v
Table of Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiAcknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ixList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xviiList of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xixChapter 1: Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Research Hypotheses and Aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Chapter 2: Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.1 Responses of forest trees to drought . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2 Modulation of stomatal development in response to environmental change . . . . 7
2.3 Plant perception of water status and downstream signalling pathways . . . . . . . . 8
2.4 Molecular outputs in response to water-deficit signalling . . . . . . . . . . . . . . . . 10
2.5 Early identification of drought-responsive genes in forest trees . . . . . . . . . . . . 12
2.6 Genome-wide dissection of forest tree drought responses—whole transcriptome
analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.7 From drought transcriptome to drought proteome . . . . . . . . . . . . . . . . . . . . . 17
2.8 The metabolic drought response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.9 Recent advances in genome analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.10 Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.11 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Chapter 3: Intraspecific variation in the Populus balsamifera drought transcriptome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3.1 Plant Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3.2 Physiological and growth traits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.3.3 RNA extraction, microarray hybridisation and analysis . . . . . . . . . . . . . . . . 28
3.3.4 Single-feature polymorphism (SFP) analysis . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3.5 DNA extraction and simple-sequence repeat (SSR) analysis . . . . . . . . . . . . . 29
vi
3.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.4.1 There is intraspecific variation in the productivity and physiological responses in
Populus balsamifera following water deficit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.4.2 Water deficit conditions elicit significant responses within the P. balsamifera
transcriptome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.4.3 There is a common P. balsamifera drought transcriptome . . . . . . . . . . . . . . . 38
3.4.4 There is a notable significant variation in the drought transcriptome across P.
balsamifera genotypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.4.5 Time of day shapes the P. balsamifera drought transcriptome . . . . . . . . . . . . . 44
3.4.6 The extent of transcriptome-wide transcript abundance change enables the P.
balsamifera to sustain growth under water-deficit conditions . . . . . . . . . . . . . . . . . . 44
3.4.7 The extent of differences in drought-responsive transcriptomes between P.
balsamifera clones positively correlated with the extent of intraspecific DNA sequence
variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.6 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.7 Supplementary Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.8 Supplementary Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Chapter 4: Drought induces alterations in the stomatal development program in Populus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.3 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.3.1 Plant material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.3.2 Physiological measurements and stomatal quantification . . . . . . . . . . . . . . . 68
4.3.3 Gene selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.3.4 Targeted transcript abundance analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.3.5 Statistical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.4.1 Stomatal conductance (gs) and relative water content (RWC) in response to water-
deficit stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
vii
4.4.2 Stomatal quantification following water-deficit stress . . . . . . . . . . . . . . . . . . 72
4.4.3 Populus homologues of genes implicated in stomatal development . . . . . . . . . 72
4.4.4 Developmental variation in transcript abundance of stomatal development genes
after water deficit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.4.5 Genes acting as positive regulators in stomatal development have correlated
transcript profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.5.1 Drought response varied between Populus balsamifera genotypes over time . . . 81
4.5.2 Transcript abundance of the Populus homologues of key stomatal development
regulatory genes varied through leaf development in a manner consistent with their
proposed molecular functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.5.3 Elevated Populus ERECTA (ER) transcript abundance early in development
corresponded with decreased stomatal indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.5.4 STOMATAL DENSITY AND DISTRIBUTION 1 (SDD1) and genotype-
specific control of stomatal development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.5.5 Stomatal development and the regulation of Populus STOMAGEN and FAMA
transcript abundance in response to water deficit . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.7 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.8 Supplementary Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.9 Supplementary Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Chapter 5: Integrated analysis of the drought metabolome and transcriptome in Populus balsamifera . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.3 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.3.1 Plant material and experimental design . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.3.2 Non-targeted metabolic profiling by gas chromatography/mass spectrometry . 106
5.3.3 Metabolome: data processing and statistics . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.3.4 RNA isolation and transcriptome analysis . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
viii
5.4.1 Populus balsamifera genotypes were subjected to water withdrawal to induce a
drought response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.4.2 Variation in Populus balsamifera metabolite profiles was evident . . . . . . . . 109
5.4.3 A Populus balsamifera drought metabolome was identifiable . . . . . . . . . . . 114
5.4.4 The drought metabolome varied among P. balsamifera genotypes . . . . . . . . . 121
5.4.5 There were correlations between drought-responsive metabolites and specific
components of transcriptome remodelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.4.6 Energy metabolism and secondary metabolite biosynthesis varied in a genotypic-
dependant manner in response to drought . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
5.4.7 Network analysis illuminated the nature of genotype-specific responses to drought
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
5.4.8 AP-1006 had a genotype-specific transcriptome response to drought . . . . . . . 137
5.4.9 There were strong correlates between specific transcript-metabolite pairs in response
to drought in AP-1006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Chapter 6: Conclusion and Future Directions . . . . . . . . . . . . . . . . . . . . . . . . 1421.1 Major Findings and Significance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Literature Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
ix
Figure 3.1 Source of origin of the six P. balsamifera genotypes examined in this study . . . . . . . . 27
Figure 3.2 Above ground biomass (a), plant height (b) and stem circumference (c) of six genotypes
of P. balsamifera were calculated 15 d after the onset of the water-withdrawal experiment for both
well watered (blue bars) and water deficit treated (orange bars) plants. Significant differences
between genotypes and treatments (P < 0.05) are denoted by small letters for all variables. Mean
values and SE bars are represented. Figure originally published in black and white. . . . . . . . . . . 31
Figure 3.3 Box plot of the variation in midday leaf stomatal conductance for six P. balsamifera
genotypes: (a) AP-947 (b) AP-1005 (c) AP-1006 (d) AP-2278 (e) AP-2298, and (f ) AP-2300.
Midday stomatal conductance for well watered plants (blue boxes) and plants grown under water-
deficit conditions (orange boxes) are represented. Asterisks indicate significant difference between
well-watered and water-deficit-treated plants: *P < 0.1; **P < 0.05; ***P < 0.001. WD, Water-deficit
treatment.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Figure 3.4 Heat maps representing transcript abundance of all drought responsive probe sets in
six P. balsamifera genotypes: AP-947, AP-1005, AP-1006, AP-2278, AP-2298 and AP-2300. Only
probe sets that are significant for treatment main effect, irrespective of time of day or genotype,
and are differentially expressed relative to a given threshold are represented (n = 280; FDR = 0.05,
log2(fold-change)-cutoff = 2.0) for both time points: (a, b) mid day, and (c, d) pre-dawn. Row
normalized, transcript abundance for all drought responsive probe sets at (a) mid day and (c) pre-
dawn. Each column represents a biological sample, and all treatments are represented in triplicate
replicates. Red indicates increased transcript abundance; blue indicates decreased transcript
abundance. Data are row normalized. Heat maps representing mean relative fold-change transcript
abundance for all genotypes at (b) mid day and (d) pre-dawn. Dark blue indicates increased
mean transcript abundance in water-deficit treated samples relative to well-watered samples; white
indicates decreased mean transcript abundance in water-deficit treated samples relative to well-
watered samples. Rows are clustered using Pearson correlation for all heat maps. . . . . . . . . . . . 37
Figure 3.5 Pearson correlation coefficient (PCC) heat map representing the P. balsamifera drought
List of Figures
x
transcriptome responses. Differential transcript abundance between well watered and water-deficit
samples for the six genotypes for the drought responsive probe sets are represented (Treatment
main effect; FDR = 0.05, log2(fold-change) cutoff = 2.0, n = 280 probe sets). Differential transcript
abundance was calculated as the mean log2(fold-change) between well watered and water-deficit
samples for a given probe-set. The PCC was determined for each pair-wise comparison, and is
represented by the colour in the corresponding cell. All samples are represented on both the x- and
y-axis, in the same order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Figure 3.6 Box plot illustrating the interplay of genotype and treatment in shaping the drought
transcriptome of six P. balsamifera genotypes. The average log2(fold-change) between well watered
and water-deficit treated samples for all genes identified as significantly differentially expressed
for treatment main effect (FDR = 0.05, log2(fold-change)-cutoff = 2.0, n = 280 probe sets) for
probe sets with (a) decreased transcript abundance in response to WD and (b) increased transcript
abundance in response to WD at the midday time point. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Figure 3.7 The relationship between the magnitude change in gene expression and the difference
in plant height between well watered and water-deficit treated P. balsamifera trees. Linear regression
analysis revealed a significant relationship between these two variables (P = 0.02033). The
coefficient of determination (R2) is shown in the figure panel. . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Supplementary Figure S3.1 (a) Correlation between historic climatic variables and observed
phenotypic traits for the six. balsamifera genotypes. (b) Correlation between absolute magnitude
change in gene expression of probe sets identified as significant for treatment main effect in response
to WD conditions and historic climatic variables, phenotypic traits and physiological response. 52
Supplementary Figure S3.2 Bar graphs representing the functional categories represented by genes
that are differentially expressed for treatment main effect (n = 280, FDR = 0.05, log2(fold-change)
cutoff = 2.0) for (a) increased transcript abundance, and (b) decreased transcript abundance in
response to water-limitation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Supplementary Figure S3.3 Bar graphs representing the functional categories represented by genes
that are significantly differentially expressed between WD and WW conditions. The proportion of
probe sets identified classified for each GO biological process functional category is represented as
the percentage of total genes differentially expressed for treatment main effect increased transcript
xi
abundance and decreased transcript abundance; FDR = 0.05, log2(fold-change) cutoff = 2.0, n =
280), and each individual genotype when analysed individually as a 2 x 2 factorial (FDR = 0.05). 54
Supplementary Figure S3.4 Quantitative reverse transcription PCR validation of transcript
abundance levels of selected genes from microarray data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Figure 4.1 The stomatal development signalling network, based on current literature. Arrows
represent positive regulation; whereas, blocked lines represent negative regulation. Question marks
represent unknown interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Figure 4.2 Variation in the physiological response to drought stress in genotype AP-1005 and AP-
1006. Box plot of the variation in midday stomatal conductance for (a) AP-1005 and (b) AP-1006
for well-watered (blue boxes) and water-deficit-treated (orange boxes) samples. Response of intrinsic
water use efficiency (WUEi; A/gs) across well-watered and water-deficit-treated samples for (c) AP-
1005 and (d) AP-1006 and photosynthesis (A) for (e) AP-1005 and (f ) AP-1006 at days 0, 5, and
15 after the onset of water withdrawal. Error bars represent the standard error of the mean. . . . 71
Figure 4.3 Variation in leaf epidermis between genotype AP-1005 and AP-1006 under (a, b)
well-watered and (c, d) water-deficit conditions, on day 30. White scale bar=50 µm. (e, f ) Box
plot of variation in stomatal index for well-watered (blue box) and water-deficit-treated (orange
box) samples for leaves that were fully developed prior to the onset of the drought experiment (leaf
A) and for those that developed during the drought experiment (leaf B). A significant reduction
in stomatal index is observed in leaves that developed during the experimental period (leaf B) for
each genotype (e) AP-1005 and (f ) AP-1006; however, no significant variation in stomatal index
is observed for leaves that developed prior to the experiment (leaf A), and the onset of water-
deficit conditions. The midline of the box represents the median value for stomatal index (e-f ) or
stomatal density (g-h), the upper and lower bounds of the box represent the interquartile range,
and the whiskers extend to the most extreme values that are not outliers. No signficant change in
stomatal density in response to water-deficit conditions for genotype (g) AP-1005 and (h) AP-1006.
Asterisks represent significant variation between well-watered and water-deficit-treated plants. *P
<0.05 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Figure 4.4 Heat map of transcript abundance across a range of tissues for Populus homologues of
genes implicated in stomatal differentiation and patterning. Transcript accumulation for the 14
xii
Populus homologues that had differential transcript abundance across the dataset, was derived from
the PopGenExpress microarray compendium made available via http://bar.utoronto.ca (Wilkins
et al., 2009a). As per the scale provided, elevated transcript abundance is represented by red and
diminished transcript abundance is represented by green. The highest levels of transcript abundance
for this group of genes are in young leaves in contrast to other tissue types. Each column represents
a discrete biological sample, and data are represented as biological triplicate replicates for each tissue
type. Data are row normalized. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Figure 4.5 Variation in transcript abundance between well-watered and water-deficit-treated trees
at six time points (days 5, 10, 15, 20, 25, and 30) for genotype AP-1005 (yellow) and AP-1006
(green) represented by the log2(fold change transcript abundance) for genes involved in stomatal
development. A positive log2(fold change transcript abundance) value is an indicator of higher
transcript abundance in water-deficit-treated samples, whereas a negative log2(fold change transcript
abundance) value is an indicator of lower transcript abundance in water-deficit-treated samples. 76
Figure 4.6 Pearson correlation coefficient (PCC) heat map representing the transcript abundance
profiles across AP-1005 and AP-1006 and six time-points. The PCC was determined for each pair-
wise comparison (gene–gene), and is represented by the colour in the corresponding cell. All genes
are represented in the same order on the x- and y-axes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Supplementary Figure S4.1 Experimental design to test the change in transcript abundance
through time and across a developmental series in Populus balsamifera. The first fully expanded
leaf (red circle) and first expanding leaf (red arrow) was marked at the onset of the water-deficit
experiment (day 0), these leaves were subsequently followed throughout the experimental period
(30 days). This enabled collection of leaf tissue that developed throughout the water-deficit
experimental at day 5, 10, 15, 20, 25 and 30. The first fully expanded leaf at day 0 represented a
leaf that was fully developed prior to the onset of water-deficit treatment. . . . . . . . . . . . . . . . . . 89
Supplementary Figure S4.2 Variation in the relative transcript accumulation of aPopulus TMM
homologue as determined by qRT-PCR. Transcript abundance calculated relative to ACT-7
transcript abundance levels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Supplementary Figure S4.3 Variation in the relative transcript accumulation of a Populus YODA
homologue as determined by qRT-PCR. Transcript abundance calculated relative to ACT-7
xiii
transcript abundance levels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Supplementary Figure S4.4 Variation in the relative transcript accumulation of aPopulus ERECTA
homologue as determined by qRT-PCR. Transcript abundance calculated relative to ACT-7
transcript abundance levels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Supplementary Figure S4.5 Variation in the relative transcript accumulation of aPopulus
STOMATAL DENSITY AND DISTRIBUTION-1 homologue as determined by qRT-PCR.
Transcript abundance calculated relative to ACT-7 transcript abundance levels. . . . . . . . . . . . . . 93
Supplementary Figure S4.6 Variation in the relative transcript accumulation of aPopulus FAMA
homologue as determined by qRT-PCR. Transcript abundance calculated relative to ACT-7
transcript abundance levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Supplementary Figure S4.7 Variation in the relative transcript accumulation of aPopulus
STOMAGEN homologue as determined by qRT-PCR. Transcript abundance calculated relative to
ACT-7 transcript abundance levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Supplementary Figure S4.8 Pearson correlation coefficient (PCC) heat map representing the
transcript accumulation profiles and stomatal indices in P. balsamifera at (a) 5 d, (b) 10 d, and (c)
15 d after the imposition of water-deficit stress. The Pearson correlation coefficient (r; top) and
P-value (bottom) are indicated within each square. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Figure 5.1 Box-plot representing net photosynthetic rate (µmol CO2 m-2 s-1) for genotype AP-947,
AP-1005, AP-1006, AP-2278, AP-2298 and AP-2300. Well-watered samples represented by blue
boxes; water-deficit-treated samples represented by orange boxes (n=3 per treatment per genotype).
The midline of the box represents the median value for photosynthesis, the upper and lower bounds
of the box represent the interquartile range, and the whiskers extend to the most extreme values that
are not outliers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Figure 5.2 Dendrogram obtained after hierarchal clustering analysis (HCA) of the metabolic
profiles of the six P. balsamifera genotypes under well-watered and water-deficit conditions at mid-
day and pre-dawn time point. Rows represent specific metabolites (n=87). Columns represent
mean intensity of all replicates for each genotype, treatment and time of day sample. Plotted values
are the mean of n = 4-10 replicates for each sample. Metabolite classes: AA = Amino Acid; C =
xiv
Carbohydrate; P = Phenolic, SA = Sugar Alcohol. NI = Not Identified. . . . . . . . . . . . . . . . . . . 112
Figure 5.3 HCA reveals 13 significant clusters (P < 0.05). Significant clusters are labeled with
unique colours and numbered (I through XIII) for identification. Hierarchical clustering was done
using pvclust (Suzuki & Shimodaira 2006), with a correlation distance measure and a complete
agglomerative clustering method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Figure 5.5 Metabolite accumulation levels for treatment main effect and treatment x genotype
interaction. (a) Hierarchally clustered metabolites that are significant for treatment main effect
across all genotypes at two different time-points [pre-dawn (PD) and mid-day (MD)]. (b) Venn
diagram demonstrating the number of metabolites that are significant for treatment main effect or
a 2-way interaction. (c) Mean log2(fold-change) of metabolite abundance for metabolites that are
significant for treatment main effect only. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Figure 5.6 Variation in the drought metabolome among six genotypes of P. balsamifera represented
by a Pearson correlation coefficient (PCC) heatmap. Differential abundance [log2(fold-change)] . .
for metabolites significant for treatment main effect (ANOVA, P < 0.05) are represented. The PCC
value was calculated for each pair-wise comparison among genotypes, and is represented by the
colour in the given cell. All genotypes are represented on both the x- and y-axis in the same order.
123
Figure 5.7 Box-plot illustrating the interplay of genotype and treatment in shaping the drought
metabolome and transcriptome of six P. balsamifera genotypes. The average absolute log2(fold-
change) between well-watered and water-deficit-treated samples for all (a) metabolites (n=40; P <
0.05) and (b) transcripts (n = 2636; P < 0.05) with significant variation in their abundance between
treatment conditions at the mid-day (MD) time point. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Figure 5.8 Heatmap of drought responsive transcript and metabolite correlations. Of all the
drought responsive transcripts, 747 unique transcripts were correlated with at least one metabolite
(|r| > 0.6; P<0.05). The rows in the heatmap represent metabolites, and the columns represent
transcripts. The columns are clustered based on their expression across samples, and the metabolites
are grouped according to functional categories. Pearson correlation coefficient (r) are represented
for each pair-wise M:T comparison, and were calculated using R. . . . . . . . . . . . . . . . . . . . . . . 127
xv
Figure 5.9 A heatmap of representative functional classes (transcripts) from the correlation data.
The averaged Spearman correlation value is represented for significant functional class: metabolite
comparisons (coloured squares). Red indicates positive correlation, whereas blue indicates negative
correlation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Figure 5.10 Pathway analysis related to the galactose metabolism. (a) Pathway map displays
selected steps from galactose metabolism pathway. Colours indicate fold-change in transcript
or metabolite abundance between water-deficit and well-watered treated samples for all six
genotypes; red indicates higher abundance in water-deficit-treated samples and blue indicates lower
abundance in water-deficit-treated samples. Enzymes are given as EC numbers. EC 2.4.1.123,
galactinol synthase; EC:2.4.1.82, raffinose synthase; EC:2.4.1.67, stachyose synthase. (b) Heatmap
representing Spearman correlation values among transcripts related to galactose metabolism and
raffinose or galactinol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
Figure 5.11 Pathway analysis related to the citric cycle (TCA cycle). (a) Correlation among select
transcripts and metabolites from the KEGG pathway pop00020 ‘Citrate cycle (TCA cycle)’ for
genotype AP-1006. Colors represent Pearson correlation value. Red indicates positive correlation
and blue represents negative correlation values. (b) Map displays selected steps from citrate cycle
pathway. Colours indicate fold-change in transcript or metabolite abundance between water-
deficit and well-watered treated samples for genotype AP-1006; red indicates higher abundance in
water-deficit-treated samples and blue indicates lower abundance in water-deficit-treated samples.
Enzymes are given as EC numbers. EC 1.1.1.37, malate dehydrogenase; EC:1.1.1.41, isocitrate
dehydrogenase (NAD+); EC:1.3.5.1, succinate dehydrogenase; EC:2.3.3.1, citrate synthase;
EC:5.2.1.2, fumarate hydratase, EC: 5.2.1.3, aconitate hydratase, EC: 6.2.1.5, succinate-CoA ligase,
beta subunit. Pearson correlation and pathway maps for other genotypes can be found in Appendix
Figure A.4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Figure 5.12 Transcript correlation networks obtained from WGCNA for (a) AP-947, (b) AP-1005,
(c) AP-1006, (d) AP-2278, (e) AP-2298 and (f ) AP-2300. The top 1000 interactions for each
genotype are represented. Nodes in the graphs represent individual transcripts that connect via
edges to other transcripts. Each node is colored according to the modules defined in Table 5.3. 136
Figure 5.13 Overrepresentation of GO terms associated with transcripts that have (a) decreased
xvi
transcript abundance in AP-1006 and (b) increased transcript abundance in AP-1006. Figures
generated using AgriGO (http://bioinfo.cau.edu.cn/agriGO). Significant overrepresentation is
represented by darker coloured boxes (P < 0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
xvii
Table 3.1 Location and climate variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Table 3.2 Total number of single-feature polymorphisms (SFPs) were identified in all probe sets
on the Affymetrix Poplar GeneChip using SNEP (P < 0.05; Fujisawa et al. 2009). Genes that were
identified as significantly differentially expressed (FDR = 0.05; log2(FC) cutoff = 2.0) and genes
whose expression is not significantly different among genotypes were surveyed for SFPs and the
proportion was calculated based on the total number of probe sets examined, respectively. . . . . 48
Supplementary Table S3.1 The microsatellite loci used to fingerprint the six P.
balsamifera genotypes in this study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Supplementary Table S3.2 Relative Water Content (RWC) calculated for each of the six P.
balsamifera genotypes after 15 days of water-deficit treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Supplementary Table S3.3 Probe sets with significant main effects or interactions for (a) all
genotypes (FDR = 0.05, log2(fold-change) cutoff = 2.0) (b) all genotypes (FDR = 0.05, no
minimum threshold) and (c) all pair-wise genotype comparisons (FDR = 0.05). . . . . . . . . . . . . 59
Table 4.1 Mean cumulative transcript abundance across all time-points for genotype AP-1005 and
AP-1006 in well-watered and water-deficit-treated samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Table 4.2 Mean plant height (in cm) on day 30 ±standard error of the mean, n ≥6 . . . . . . . . . . 84
Supplementary Table S4.1 Primers used for qRT-PCR analysis.. . . . . . . . . . . . . . . . . . . . . . . . 98
Supplementary Table S4.2 Relative water content (RWC) on day 30. . . . . . . . . . . . . . . . . . . . 99
Supplementary Table S4.3 ANOVA results: transcript abundance. . . . . . . . . . . . . . . . . . . . . 100
Table 5.1 Number of metabolites with significant main effects or interactions (n=87 metabolites).
Padj-value cutoff = 0.05 (Benjamini-Hochberg). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Table 5.2 Metabolites with significantly different abundance levels in response to drought
List of Tables
xviii
(ANOVA, Padj-value < 0.05). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Table 5.3 Module membership in the drought transcriptome network of AP-1006 and preservation
of drought modules among the other genotypes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Table 5.4 Module-treatment or -time of day relationships of the P. balsamifera (AP-1006) drought
transcriptome. Columns 2 and 3 represent the correlation between the mean expression of the
module and the experimental factor (T or D). Significant values are in bold (P-value < 0.05). 135
xix
List of Abbreviations
A Carbon assimilation
AA Amino acid
AB Alberta
ABA Abscisic Acid
ABF ABRE-binding factor
ABRE ABA-responsive element
AFLP Amplified fragment length polymorphism
ANOVA Analysis of variance
AP Alberta Pacific
At Arabidopsis thaliana
BCAA Branched chain amino acid
bHLH Basic-helix-loop-helix
BLAST Basic local alignment search tool
C Carbohydrate
cDNA Complementary DNA
cm Centimeter
CO2 Carbon dioxide
d Days
D Time-of-day
DNA Deoxyribonucleic acid
DRE Drought-responsive element
DW Dry weight
EPF EPERDIMAL PATTERNING FACTOR
ER ERECTA
ERD EARLY RESPONSE TO DROUGHT
EST Expressed sequence tag
FDR False discovery rate
Fs Fagus sylvatica
FW Fresh weight
xx
G Genotype
GC Gas chromatography
GO Genome ontology
gs Stomatal conductance
HAB1 HYPERSENSITIVE TO ABA 1
HCA Hierarchal clustering analysis
HIC HIGH CARBON DIOXIDE
HKT1 A high-affinity K transporter
HTP High-throughput
Ile Isoleucine
IRGA Infrared gas analyzer
KEGG Kyoto encyclopedia of genes and genomes
LD Linkage disequilibrium
LEA Late embryogenesis-abundant
LI Licor
LPI Leaf plastochron index
LRR Leucine rich repeat
MAP Mitogen activated protein
MD Midday
MK Mitogen activated protein kinase
MKK Mitogen activated protein kinase kinase
MS Mass spectrometry
MYB Myeloblastoma
NE Nebraska
NI Not identified
P Phenolic
P. Populus or Pinus
PCC Pearson correlation coefficient
PD Pre-dawn
PHYB PHYTOCHROME B
PIF4 PHYTOCHROME INTERACTION FACTOR 4
PIP Plasma membrane intrinsic protein
xxi
PP2C Protein phosphatase 2C
Pro Proline
PYR/PRL PYRABACTIN RESISTANT / PYR-like
PYR1 PYRABACTIN RESISTANCE 1
qPCR Quantitative polymerase chain reaction
QTL Quantative trait loci
RD22 RESPONSIVE TO DESICCATION 22
RFO Raffinose family oligosaccharides
RNA Ribonucleic acid
RT-PCR Reverse transcriptase polymerase chain reaction
RWC Relative water content
SA Sugar alcohol
SDD1 STOMATAL DENSITY AND DISTRIBUTION 1
SFP Single-feature polymorphism
SNEP Simultaneous detection of nucleotide and expression polymorphisms
SSR Simple-sequence repeat
T Treatment
TCA Tricarboxylic acid cycle
TMM TOO MANY MOUTHS
Tp Time point
Tx Treatment
USA United States of America
Val Valine
WA Washington
WD Water-deficit
WGCNA Weighted gene correlation network analysis
WUE Water use efficiency
WW Well-watered
YDA YODA
1
Chapter 1: Overview
2
Chapter 1: Overview
As long-lived sessile organisms, forest trees are exposed to a variety of deleterious environmental
conditions and must develop mechanisms in order to avoid, tolerate or adapt to the consequences
of these stresses. Although trees have evolved to contend with the adverse environmental conditions
that occur over their lifetimes, forests are threatened by rapid changes in climate that are occurring
on a global scale. Widespread forest mortality has been recently been attributed to global climate
change, including drought stress (Allen & Breshears 1998; Breshears et al. 2005; Bonan 2008; van
Mantgem et al. 2009; Allen et al. 2010). The frequency and severity of future droughts is predicted
to increase (Intergovernmental Panel on Climate Change 2007; Allen et al. 2010), negatively
impacting many facets of forest productivity and ecosystem function (Fischlin et al. 2007; Adams et
al. 2010).
In Canada, forest trees, including those of the genus Populus are ecologically and economically
important. The continued survival and productivity of such trees is highly correlated with
water availability. In hybrid poplar plantations, decreased water availability negatively impacts
productivity (Silim et al. 2009). Due to the negative impact of drought stress on forest health and
productivity, it is becoming increasingly important to understand the mechanisms that underpin the
drought response in trees.
1.1 Research Hypotheses and Aims
The main aim of the research presented in this thesis is to explore various aspects of the intraspecific
variation in the drought response among Populus balsamifera genotypes. In addition, this research
aims to increase our understanding of the molecular underpinnings that allow sessile organisms,
such as Populus trees to contend with the impact of drought stress. When the research presented
herein began, very little was known about the extent of intraspecific variation at the transcriptome-
level in a single species of Populus. Early studies focused on transcriptome-level responses to water-
deficit stress in individual or between different Populus hybrids (Street et al. 2006; Wilkins et al.
2009). More specifically, the research presented in this thesis focused on a single species, Populus
balsamifera, and tested the following hypotheses:
1. Intraspecific variation in phenotypic responses to drought stress among six genotypes
3of Populus balsamifera are underpinned by significant differences in their drought-responsive
transcriptomes.
2. Drought-induced modification of the transcription of genes implicated in the stomatal
development regulatory network are linked to changes in stomatal density.
3. There are transcript-metabolite relationships that occur in response to drought in P.
balsamifera that vary in a genotype-dependant manner.
In order to test these hypotheses three experiments were undertaken. First, as described in Chapter
3, six genotypes of P. balsamifera were grown in a common garden experiment and exposed to
a period of drought stress (15 days). The transcriptomes of the six genotypes in response to
well-watered and water-deficit conditions were interrogated using Affymetrix Poplar GeneChip
technology at both a pre-dawn and mid-day time point. Second, as described in Chapter 4, the
expression of genes involved in the stomatal development pathways were examined in two P.
balsamifera genotypes (AP-1005 and AP-1006) in conjunction with phenotypic analysis of stomatal
development under well watered and water deficit conditions throughout the experimental period
(exposure to drought stress; days 0 through 30). Finally, as described in Chapter 5, a non-targeted
metabolome analysis was performed on the six P. balsamifera genotypes to identify the drought
responsive metabolites. This was then compared to the drought transcriptomes in P. balsamifera to
identify relationships between the transcriptome and metabolome.
As summarized in Chapter 6, the results of the research described herein contribute to our
understanding of the intraspecific variation in the drought response in P. balsamifera at the
molecular-level.
4
Chapter 2: Literature Review
Contents of this chapter have been published in Forestry an International Journal of Forest
Research: Erin T. Hamanishi and Malcolm M. Campbell. 2011. Genome-wide responses to
drought in forest trees. Forestry. 84: 273-283
The published paper can be found online at
http://forestry.oxfordjournals.org/content/84/3/273.full?sid=c51c6426-0679-4a21-890c-
8bc3e06b96f6
The material in this chapter is © by Oxford University Press, 2011.
5
Chapter 2: Literature Review
Alterations in global climate and precipitation regimes strongly influence forest distribution and
survival (Allen & Breshears 1998; Shaw et al. 2005; Engelbrecht et al. 2007). While forest trees
are sessile organisms, they possess many attributes that allow them to contend with variable water
availability, within limits (Ingram & Bartels 1996; Chaves et al. 2003). Nevertheless, expected
rates of global climate change are unprecedented, with longer and more severe periods of drought
predicted (Intergovernmental Panel on Climate Change 2007). This may have a profound effect on
forest health, as water limitation is one of the leading contributors to forest declines globally (Bigler
et al. 2006; van Mantgem et al. 2009; Allen et al. 2010). For example, in western Canadian aspen
forests, drought negatively impacted growth and survival after a severe regional drought during
the 2001–2002 growing season (Hogg et al. 2008), with similar reductions in forest productivity
observed in Europe (Ciais et al. 2005).
Drought is a multidimensional environmental factor; affecting tree responses from the molecular
level to the forest stand level. Interpretation of the drought response at the stand and tree level is
complex because it involves consideration of the stress effects and responses (Yordanov et al. 2000).
Nevertheless, negative impacts of drought are observed in many facets of forest health, including
seedling recruitment, productivity, susceptibility to pathogen or insect attack and fire susceptibility
(Hogg & Wein 2005; van Mantgem et al. 2009; Zhao & Running 2010). Consequently, there is
considerable incentive to better understand the means by which forest trees respond to drought,
so as to develop strategies for preservation of forest tree growth and survival against this particular
environmental threat.
Given recent advances in genome biology, there is great scope to develop a more comprehensive
mechanistic understanding of forest tree drought responses. In keeping with this, over the past
decades, advances made in genetics, molecular biology, genomics, proteomics and bioinformatics
have provided ever-growing insights into how forest trees respond to drought. This review considers
the emergence of those insights and how they might shape the protection of forest trees from
drought in the future. Readers interested in a broader consideration of the genomics of forest tree
responses to abiotic stress or the modification of specific genes as a means by which to achieve
drought tolerance are directed to excellent reviews elsewhere [for review, see: Polle et al. (2006) and
Fischer and Polle (2010)].
6
2.1 Responses of forest trees to drought
A reduction of available water impinges on trees’ ability to grow and transpire by affecting the soil–
plant water continuum. Often, in response to declines in plant water potential, in order to reduce
water loss under drought stress, forest trees reduce transpiration by closing their stomatal pores, at
the expense of CO2 assimilation (Jarvis & Jarvis 1963; Cowan & Farquhar 1977). Additionally,
water fluxes in a tree can further be disrupted through cavitation or xylem embolism at high xylem
tensions induced by water stress. Cavitation, or xylem embolism, is the filling of xylem with air or
water vapour instead of water, leading to a reduction in the water conductivity of the plant. The
reduction of water conductance within a tree can, in turn, limit growth (Tyree et al. 1994; Rice et al.
2004).
Forest trees utilize a range of strategies to contend with drought. A series of molecular, biochemical,
physiological and morphological changes underpin plant response to water deprivation, and the
extents of these responses are highly variable and complex [for review, see Chaves et al. (2003) and
Ingram & Bartels (1996)]. The variability in drought responses is a function of the severity and
duration of the drought stress (Chaves et al. 2003; Yan et al. 2012), superimposed upon genetic
variation at the individual, population and species levels (Zhang et al. 2004; Wilkins, Waldron, et
al. 2009b; Hamanishi et al. 2010). Foresters have known for many years that different tree species
have variable responses to drought. Almost 50 years ago, in a cross-species examination of tree
seedlings under drought conditions, Jarvis and Jarvis (1963) concluded that Pinus spp. were the
most drought-tolerant species, whereas Populus tremuloides was the most susceptible.
Variation in the drought response is not only seen between forest tree species but also seen
intraspecifically. For example, Pseudotsuga menziesii (Douglas fir) seedlings, from distinct geographic
regions, exhibited variable drought resistance when grown under water-deficit conditions in a
greenhouse (Ferrell & Woodard 1966). Similarly, variation in response to water availability has
been observed in progeny and provenance testing in Pinus taeda (loblolly pine) (Teskey et al. 1987).
These early genetic studies, aimed at examining drought tolerance, focused on the relationship
between genotype and environment and established the importance of genetic variation in drought
tolerance both inter- and intraspecifically among forest trees. Nevertheless, the mechanisms that
underpinned drought tolerance and resistance in forest trees and their molecular basis were much
less well understood.
7Forest trees posses a wide array of traits that confer drought tolerance. The ability to avoid drought
stress is dependent on the trees’ ability to minimize water loss and maximize water uptake (Chaves
et al. 2003). For example, some forest trees can increase water uptake through more extensive and
deeper roots (Nguyen & Lamant 1989). In order to minimize water loss under drought conditions,
forest trees can utilize a variety of traits including altered leaf morphology [e.g. cuticular wax
(Hadley & Smith 1990)], reduction in leaf area [e.g. increased leaf abscission (Munne-Bosch &
Alegre 2004)] and reduction in stomatal conductance. In response to drought, stomatal closure
is an effective mechanism to limit water loss and prevent desiccation. For example, in potatoes,
leaf water potential was maintained under a period of slowly developing drought stress, and
the transpiration rate was regulated by stomatal aperture (Kopka et al. 1997). Within a plant,
stomatal control plays a particularly important role in regulating water balance and, in turn, CO2
assimilation.
2.2 Modulation of stomatal development in response to environmental change
The development of stomata on the leaf surface is regulated by developmental and environmental
cues. Much is known about the stomatal development in Arabisopsis thaliana, including many
of the regulatory components and networks that underlie stomatal differentiation [for review, see
Bergmann and Sack (2007), Casson and Heatherington (2010)]; however, the modulation of this
pathway in response to environmental cues is largely unknown. In Arabidopsis, stomatal density in
new, developing leaves is adjusted with the environment sensed by mature leaves (Lake et al. 2001;
Miyazawa et al. 2006). Increasing light availability is correlated with higher stomatal density, and
is modulated through the function of PHYTOCHROME B (PHYB) and a transcription factor,
PHYTOCHROME INTERACTION FACTOR 4 (PIF4; Casson et al. 2009). Whereas, elevated
atmospheric CO2 levels are associated with a decline in stomatal density (Woodward 1987). HIGH
CARBON DIOXIDE (HIC) modulates stomatal density in response to changing atmospheric CO2
in Arabidopsis (Gray et al. 2000). In response to drought stress, modification to stomatal density is
variable among plant species and is dependant on the severity of drought. In Arabidopsis mutants
with increased decreased stomatal density demonstrated improved drought tolerance (Yu et al. 2008;
Yoo et al. 2010). Alterations to stomatal development, including the reduction of stomatal density,
resulting from the exposure to drought stress may represent a long-term strategy to contend with
8water-deficit stress; however, it may also be a limiting factor to future productivity.
Although stomatal regulation is an efficient mechanism to contend with water shortages by limiting
water loss through the stomatal pores (Froux et al. 2005), it cannot solely prevent water balance
decline in trees. Consequently, forest tree growth and survival under drought stress is frequently
dependent on many of the aforementioned strategies acting in concert. Although inroads have been
made in dissecting the physiological responses to drought in forest trees, understanding the depths
of the molecular underpinnings is scant in forest trees relative to herbaceous annual plant species.
2.3 Plant perception of water status and downstream signalling pathways
Some of the means by which plants sense water-deficit conditions and subsequently induce
downstream molecular signalling response cascades have been elucidated [for review, see Shinozaki
and Yamaguchi-Shinozaki (2007)]. In order to mount any response to drought, plants must first
sense water-deficit conditions. Although the precise mechanism underlying drought perception
in plants is not well understood, there are multiple hypotheses related to how roots sense drought
conditions in the soil. Under conditions of decreased soil water, the plant phytohormone abscisic
acid (ABA) accumulates in the soil solution. The increase in soil ABA concentration may act as
a mechanism by which roots sense reduced soil water (Slovik et al. 1995; Hartung et al. 1996).
Drought may also be perceived through a reduction in turgor by osmosensors. Urao et al. (1999)
identified an Arabidopsis thaliana transmembrane histidine kinase, AtHK1, which has putative
function as an osmosensor. AtHK1 senses osmotic changes and transmits a stress signal to
downstream mitogen-activated protein kinase signalling cascades, which in turn induce drought-
responsive gene expression (Urao et al. 1999). In red river gum (Eucalyptus camaldulensis), Liu et al.
(2001) identified two HKT1 homologues that can sense changes in solute concentration, similar to
AtHK1. Eucalyptus HKT1 homologues altered sodium and potassium transport in Xenopus oocytes,
suggesting a role in osmoperception and osmoregulation. These homologues are strong candidates
for tree proteins that play a key role in the perception of water limitation leading to drought
response signalling.
Downstream of water-deficit sensing, the steps in the drought-responsive pathway proceed via one
of two signalling pathways: the ABA-dependent and the ABA- independent pathways (Shinozaki &
9Yamaguchi-Shinozaki 1996). Under drought conditions, increasing levels of ABA are observed in
the roots and shoots; ABA is thought to play an important role in root to shoot signalling (Walton
et al. 1976; Zeevaart & Creelman 1988; Davies & Zhang 1991). The ABA signal is modified
through changes to xylem or apoplastic pH, influencing the signalling process by moderating
sensitivity and availability of ABA in planta (Wilkinson et al. 1998; Bahrun et al. 2002; Sobeih
et al. 2004). ABA not only acts in the signalling of drought stress but also plays a central role in
regulating drought response in plants.
One of the most prominent roles of ABA is in the regulation of stomatal movement in response to
drought [for review, see Belin et al. (2010), Wilkinson and Davies (2002) and Popko et al. (2010)].
Mediating stomatal aperture under drought conditions allows plants to limit water loss and regulate
water balance during periods of water deficit. Another important role of ABA-mediated drought
response in plants is the maintenance of root growth under mild or moderate drought stress,
whereas leaf growth under drought conditions is restricted (Hsiao & Xu 2000). Under drought
conditions, growth allocation patterns in plants are altered and the variable growth rates observed
in roots and shoots are correlated with ABA levels; however, the regulation of growth rates may be
mediated through another plant hormone, ethylene (Sharp et al. 2004).
Despite what is known about the role of ABA in mediating plant response to drought stress,
little was known about ABA perception by the plant cell until recently. Initial reports of putative
ABA receptors were considered controversial because of limited evidence of central roles in ABA
perception and response (McCourt & Creelman 2008). More recently, using a chemical screen
technique, Park et al. (2009) identified a protein, PYRABACTIN RESISTANT 1 (PYR1), which is
involved in ABA signalling. PYR1 and PYR1-like receptors are necessary for many plant responses
to ABA. Members of the PYRABACTIN RESISTANT / PYRABACTIN RESISTANT-LIKE
(PYR/PRL) family of receptor proteins interact downstream with HAB1, a 2C protein phosphatase
(PP2C). PP2Cs negatively regulate ABA signalling(Saez et al. 2004). Through proteomic
approaches, another group simultaneously identified the ABA receptor, RACR1, which belongs to
the same PYR/PYL family of receptors (Ma et al. 2009). This family of receptor proteins appears
to be highly conserved across crop plants, and recent work is aimed at elucidating members in tree
species [for review, see Klingler et al. (2010)]. Saavedra et al. (2010) identified a PP2C homologue
from beech tree (Fagus sylvatica) that is a negative regulator of ABA signalling and showed that
10FsPP2C interacts with Arabidopsis PYR7 and PYR8. Identification of ABA receptor protein
homologues is important for understanding the perception of ABA and its involvement in the
drought response in trees.
2.4 Molecular outputs in response to water-deficit signalling
Whole-plant responses to ABA are underpinned by ABA- dependent changes in gene expression
that are mediated through the action of ABA-inducible transcription factors controlling the
expression of genes containing cis-acting ABA response elements (ABREs) [for review, see Ingram
and Bartels (1996)]. The ABRE is stereotypically found in the upstream regulatory regions of
drought-responsive genes (Giuliano et al. 1988; Bray 1994). ABRE-like sequences have also been
identified in the upstream regulatory regions of drought-inducible genes, including the G-box
sequence (Williams et al. 1992; Shen et al. 1996). Members of the bZIP protein family are known
to bind to ABRE and ABRE-like sequences and, in turn, activate ABA-dependent gene expression
(Guiltinan et al. 1990; Choi et al. 2000; Uno et al. 2000).
Among the best-characterized drought-induced genes, RESPONSIVE TO DESSICATION 22
(RD22) has ABA-regulated transcription. ABA-mediated regulation of RD22 transcription requires
the synthesis of an MYC (rd22BP1/AtMYC2) and an MYB (AtMYB2) transcription factor, both
of which are induced by ABA. AtMYC2 and AtMYB2 act as transcriptional activators and bind
cis- elements in the promoter of RD22 (Abe et al. 1997; 2003). AtMYC2 and AtMYB2 also are
involved in ABA-dependent gene expression of other ABA-inducible genes (Abe et al. 2003).
While many drought responses are mediated by ABA, plants also have ABA-independent responses
to drought conditions. Several genes are induced under drought conditions that are not dependent
on ABA (Shinozaki & Yamaguchi-Shinozaki 1996). Often these genes contain a conserved
dehydration-responsive element (DRE) in their upstream gene regulatory region, which functions
to recruit transcription factors that are not regulated by ABA. Many of the non-ABA abiotic stress
signalling pathways are complex, and it is hypothesized that the DRE cis-acting element plays a role
mediating different stress-signalling cascades, resulting in an overall plant response to abiotic stress
(Knight & Knight 2001).
Integration of the ABA-dependent and ABA-independent signalling cascades also occurs through
11downstream gene regulation. For example, the gene RD29A contains both an ABRE and a DRE
within its cis-acting upstream gene regulatory region. In the initial stages of drought stress,
expression of RD29A is independent of ABA but later is dependent on ABA for gene expression
(Shinozaki & Yamaguchi-Shinozaki 2000).
Through molecular analysis of multiple plant species, insights have been gained into a range of
proteins that are induced under water-deficit conditions [for review, see Ramanjulu and Bartels
(2002)]. For example, the hydrophobic late-embryogenesis-abundant (LEA) proteins accumulate
under drought stress and are commonly associated with tolerance to water-deficit conditions
(Welin et al. 1994). Recent evidence suggests that LEA proteins may have an important role in
the stabilization of other proteins and membranes, as well as the prevention of protein aggregation
during periods of water deficit (Close 1996; Goyal et al. 2005). In a poplar clone (P. euramericana
cv Dorskamp), the rapid induction of a LEA family protein, dehydrins, gene expression was
observed after osmotic stress was imposed on the clones (Caruso et al. 2002). Similar increases in
transcript or protein levels of LEA family proteins in other forest trees, such as spruce, have been
observed (Blodner et al. 2007).
Aquaporins are another major class of proteins that play a key role in the water-deficit response.
Aquaporins are channel proteins that are found in cellular membranes and are responsible for water
flux and are crucial for maintaining proper water balance [for review, see Maurel et al. (2008)].
There are two major groups of aquaporins: those found specifically in plasma membranes are plasma
membrane intrinsic proteins (PIPs), whereas those found in the tonoplast are known as tonoplast
membrane intrinsic proteins. Both classes of water transport proteins are important for maintaining
water status in the plant, which is vital for photosynthesis and subsequently growth. In Eucalyptus,
PIPs are essential for normal growth; a reduction in PIPs resulted in a suppression of growth
(Tsuchihira et al. 2010). The expression of aquaporins is dynamic in response to plant water status.
Under drought stress conditions, the expression of Plasma-membrane-Intrinsic-Protein (PIP)-type
aquaporins was reduced in tobacco plants, hypothesized to decrease water transport (Mahdieh et al.
2008). In white poplar (Populus alba L.), Berta et al. (2010) identified five transcripts for aquaporin
proteins that were up-regulated in following re-watering in trees that experienced drought stress.
Accumulation of aquaporins following re-watering may be integral to restoration of water transport
of plants under well-watered conditions. In poplar trees, members of the PIP1 family of aquaporins
12are important for recovery from xylem embolism Populus (Secchi & Zwieniecki 2010). The
functional variability and importance of aquaporins in a trees’ response to drought is reflected in the
distinct responses of aquaporins in trees with different drought response strategies. Under drought
conditions, the drought responsiveness of specific aquaporin family members varied between two
poplar clones (P. balsamifera and P. simonii × P. balsamifera) that had contrasting drought response
strategies (Almeida-Rodriguez et al. 2010). The variability in aquaporin response may reflect the
different roles of aquaporins, with respect to water transport, in trees.
Over the past decades, examination of the drought response in herbaceous annual plant species has
revealed many details about the molecular pathways involved in the drought response. This had
led to the identification of many important proteins that accumulate under drought conditions,
including transporter proteins, messenger RNA- binding proteins, proteases and many others
involved in regulation and signal transduction [for review, see Ingram and Bartels (1996)]. More
recently, progress has been made uncovering the molecular mechanisms underpinning these
pathways and, in turn, drought tolerance and resistance in forest trees. Currently, genomic
approaches are being brought to bear the drought responses that enable the integration of
knowledge of tree-level responses with gene expression and function.
2.5 Early identification of drought-responsive genes in forest trees
Prior to the genomic era, foresters and plant biologists alike were limited to studying the function
of one or a few genes at a time. Early studies in trees revealed that drought- responsive genes
initially identified in herbaceous annual plants, such as dehydrins and heat-shock proteins, and had
homologues expressed in the bark tissue of various woody plants (Wisniewski et al. 1996). Insights
into the molecular response of trees to drought began to improve through the identification of
genes induced by drought in trees. For example, a number of drought-induced genes were first
identified in Pinus taeda through comparative analysis of complementary DNA (cDNA) clones
whose expression was induced under water-deficit conditions (Chang et al. 1996). Chang et al.
(1996) were able to further characterize four water-deficit-induced cDNAs, providing insight into
their sequences and patterns of expression. Based on sequence similarity to characterized genes in
other plant species, the majority of these genes were thought to function in cell wall reinforcement
and hypothesized to participate in the adaptation of the cells to water-deficit stress (Chang et al.
1996). To identify larger numbers of drought-responsive genes without basing discovery on a
13priori knowledge, Dubos and Plomion (2003) pioneered the use of cDNA-Amplified Fragment
Length Polymorphism (AFLP) to identify drought-responsive genes in roots and then the needles of
Pinus pinaster (maritime pine). Dubos et al. (2003) identified 48 putative genes that were drought
responsive in Pinus pinaster seedlings. Of these 48 genes, many corresponded to proteins of known
function with roles in photosynthesis, carbohydrate metabolism, cell wall synthesis and plant
defence; however, a relatively high proportion corresponded to genes of unknown function(Dubos
et al. 2003). Similar experiments using a cDNA-AFLP technique with almond (Prunus amygdalus)
identified drought-responsive genes in young leaves of different cultivars with variable drought
response (Campalans et al. 2001).
The breadth and depth of gene discovery in forest trees was expanded through partial sequencing
of transcribed cDNA libraries of loblolly pine (Allona et al. 1998) and poplar (Sterky et al. 1998)
to generate compendia of expressed sequence tags (ESTs). These pioneering EST efforts focused
almost exclusively on genes involved in wood formation but generated information about gene
expression and coding sequences in what were, at that time, almost completely uncharacterized
genomes. These initial efforts in gene discovery, although modest by today’s standards, were ground
breaking and provided important foundations for future studies.
Following the initial efforts in pine and poplar, the number of reported ESTs from forest tree
species, including birch, pine and eucalyptus, increased year by year (Strabala 2004; Li et al. 2009;
Wang et al. 2010). Nevertheless, many of these efforts continued to focus ESTs that were related
to wood formation. With an increasing desire to gain insights into stress responses in forest trees,
biologists carried out EST analysis on other tissues under various treatment regimes. Ujino-Ihara et
al. (2000) identified ~1400 ESTs from the inner bark from a sugi tree (Cryptomeria japonica), which
was felled 2 days prior to EST analysis in order to enrich the EST library in drought, wounding
and other stress-related genes. In order to directly identify water-responsive genes in loblolly pine,
Lorenz et al. (2006) subjected seedlings to various watering regimes and generated an EST library
from the root tissues. In these studies, some of the ESTs identified were homologous to genes
previously identified as drought responsive in herbaceous plants, such as LEAs and dehydrins.
Although there was some degree of similarity between ESTs identified in trees and previous attempts
in other plant species, many of the transcripts identified in trees were of unknown function.
Although the identification of drought-induced ESTs in forest trees was important for early gene
14identification and transcriptome studies, the EST libraries generated from these studies played key
roles in founding broader analyses of tree transcriptome activity (Nagaraj et al. 2007).
2.6 Genome-wide dissection of forest tree drought responses—whole transcriptome analyses
During the course of EST discovery efforts, it was clear that the identification of subsets of drought-
responsive genes was insufficient to fully understand the complexities of the drought response in
forest trees. cDNA-AFLP techniques and EST frequencies revealed basic data with respect to gene
expression; however, genome-wide analysis techniques, such as microarray analysis, had the potential
to reveal global gene expression patterns. Early microarray platform experiments investigating the
gene expression patterns in hybrid aspen, based on a small set of ESTs identified in wood formation
(Sterky et al. 1998), revealed unique tissue-specific transcript profiles in differentiating xylem
(Hertzberg et al. 2001). Some of the earliest insights into transcriptome responses to drought in
forest trees was determined using cDNA microarray based on ESTs libraries from specific tissues,
such as xylem, shoot tips or pollen. Heath et al. (2002) used a 384 pine cDNA microarray to
investigate the adaptation to mild drought in pine seedlings. Although the number of genes being
investigated was limited, the importance of molecular chaperones and membrane transport proteins
was revealed. These particular proteins are postulated to be vital in cell maintenance and repair and
therefore necessary for forest trees to cope with mild drought stress (Heath et al. 2002).
Early microarray experiments, aimed at investigating the molecular basis of a given trait or response,
had potential; however, they were constrained by the number of genes under investigation. As
development of genetic resources for forest trees continued, more comprehensive microarrays were
established that enabled relationships between physiological responses and genome-wide gene
expression profiles to be investigated. Watkinson et al. (2003) used a microarray consisting of
~2100 cDNA clones to examine the gene expression in drought-stressed loblolly pine that revealed
that alterations in gene expression patterns in response to drought in loblolly pine were not only
qualitative but also quantitative. The increased number of transcripts examined, representing 15
functional categories, allowed the authors to correlate patterns of expression with acclimation to
mild or severe drought and define roles for specific groups of genes (Watkinson et al. 2003).
Although EST sequencing and early microarray experiments provided significant insights into
15groups of drought- responsive genes, the sequencing of the complete black cottonwood (P.
trichocarpa Torr. & Gray) genome represented a significant milestone in the ability to explore the
drought transcriptome in its entirety in a tree (Tuskan et al. 2006). The Populus genome sequence
was not only integral in the development of more comprehensive whole-genome microarray
platforms but also important for comparative analyses with other plant genomes, such as
Arabidopsis. In the ‘post-genomic’ era, a number of microarray resources were developed for poplar
species using the available sequence data. POP2, a spotted cDNA array representing more than
100 000 ESTs and ~40 per cent of predicted gene models from the Populus genome, was used to
investigate the global drought response in black cottonwood (P. trichocarpa) and eastern cottonwood
[P. deltoides Bart.; (Sterky et al. 2004; Street et al. 2006)]. Street et al. (2006) identified genes with
contrasting responses to drought in the two Populus species and hypothesized that the control of
gene expression may be an important process in species divergence.
In addition to cDNA microarrays, two short oligo-nucleotide microarrays were developed for
Populus: the Affymetrix GeneChip Poplar Genome Array (www.affymetrix.com) and the Nimblegen
Populus whole-genome array (http://www.nimblegen.com/products/exp/ eukaryotic.html). Both
of these microarrays were designed based on the gene model sequences from the poplar genome
sequence (Tuskan et al. 2006), as well as available publicly available EST sequences. Using
Affymetrix GeneChip Poplar Genome Arrays, Wilkins et al. (2009b) were able to identify divergent
responses in gene expression profiles in response to drought between two poplar hybrids, suggesting
that it is difficult to capture a genome-wide drought response with one or a few Populus genotypes.
They also showed that transcriptional responses to drought are time of day dependent in hybrid
poplars, indicating that any investigation into the molecular-level responses to drought should factor
time of day in order to fully grasp the molecular basis of such a response.
The ability to identify a Populus-specific drought transcriptome is becoming increasingly more
difficult as whole-genome arrays uncover variation in the transcriptomes within a given species,
P. balsamifera (balsam poplar; Hamanishi et al. 2010; Chapter 3, this volume). Hamanishi et al.
(2010) compared phenotypic traits with gene expression profiles, showing that balsam poplar
genotypes that exhibited increased magnitude change in gene expression were also able to sustain
growth under drought conditions. Differences in the drought-responsive transcriptomes among
balsam poplar genotypes was related to the extent of intra-specific DNA sequence variation,
16suggesting that genetic relatedness is likely an indicator of a shared drought response.
Until recently, many gene expression studies on forest trees focused on leaves or roots provided
insights into transcriptome response in a given organ; however, the subtleties of the drought
response at the cellular level could not be identified from these experiments. Berta et al. (2010)
examined the transcriptome response to water deficit in the wood-forming tissue in white poplar
(Populus alba). This study investigated the transcriptome response to drought, as well as the
interplay between wood formation and drought response in trees. Many drought-responsive gene
networks were shared between different tissues (e.g. leaves and roots), but some transcripts were
identified that may have had specific roles in modulating wood formation under drought stress
(Berta et al. 2010). Together, these studies on Populus reveal the complexities in the genome-
wide drought response. In potato, Kopka et al. (1997) investigated guard cell-specific patterns of
transcript abundance. Guard cell-specific transcript levels indicated a systemic drought response
signal, leading to long-term changes in transcript abundance (Kopka et al. 1997). In forest trees,
the variability in the drought transcriptomes on a temporal and spatial scale, as well as the variability
that is present among various individuals, can be exploited for breeding and selection of drought-
resistant stock.
Genome-wide dissection of forest tree drought responses—quantitative trait locus mapping and association studies
The response of forest trees to drought stress at the morphological and transcriptome level is
complex and highly variable, both intra- and inter-specifically. In order to dissect such a complex
trait, biologists have often employed quantitative genetic techniques to reveal genetic intervals
to which variability in the trait can be ascribed. Quantitative trait locus (QTL) analysis is
advantageous for analysing traits, where a priori knowledge of molecular underpinnings or genes
is elusive. Genetic maps of many forest trees have been generated for forest trees, and many QTLs
have been identified for drought-related traits. In rapidly growing willow hybrids (Salix dasyclados
× Salix viminalis), Ronnberg-Wästljung et al. (2005) identified a few QTLs that had a significant
effect on water use efficiency (WUE), and the authors noted that the analysis revealed the complex
nature of drought tolerance in willow. Similarly, in pedunculate oak (Quercus robur L.), Brendel
et al. (2008) identified 10 QTL for WUE, where only a few QTL were responsible for the larger
proportion of the clonal variation. Generally, in forest tree species, the number of QTL identified,
17and the amount of variation that is explained by any given QTL reveals the complexity of drought
tolerance or WUE. The knowledge gained from QTL analysis in tree species is useful for tree
breeding; however, these gains have been restricted by difficulties and time-consuming nature
of identifying genes or genes located at a given QTL for species with limited genomic sequence
availability.
In order to overcome the obstacles associated with QTL mapping experiments, biologists have
used association or linkage disequilibrium (LD) mapping in order to determine the genetics
underpinning complex traits, as it is proposed to be more efficient than QTL mapping (Hall et al.
2010). Association mapping relies on the association of genomic regions containing genetic markers
with complex traits. With time, the availability of genetic markers has improved allowing the use
of association or LD mapping to become more prevalent. Using a candidate gene loci approach,
Gonzalez-Martinez et al. (2006) estimated the LD estimate for 18 drought-tolerance candidate
genes in loblolly pine (P. taeda). A majority of the drought-tolerance candidate genes showed
neutral selection, with the exception of CCoAOMT-1 and EARLY RESPONSE TO DROUGHT 3
(ERD3) (Gonzalez-Martinez et al. 2006).
An alternative association mapping approach is a whole-genome scan. In white spruce (Picea
glauca), single- nucleotide polymorphisms (SNPs) were identified in expressed genes and used
as genetic markers for mapping purposes (Namroud et al. 2008). Although the authors noted
limitations in their methods, Namroud et al. (2008) found potential associations between local
adaptation of candidate genes and phenotypic attributes of populations. The benefits of identifying
genes under potential selection for drought tolerance in non-model tree species through association
mapping has the potential to be very useful for tree breeding strategies in the future.
2.7 From drought transcriptome to drought proteome
Studies unveiling the drought-responsive transcriptome in trees have provided a wealth of
information regarding the molecular underpinnings of the drought response; however, analysis of
the proteome reveals the abundance of final gene products that may be important in understanding
the down stream drought response. The drought proteome has been examined in several tree
species, including poplar (Plomion et al. 2006) and spruce (Blodner et al. 2007). Initial drought
studies examining both gene and protein expression in poplar revealed limited overlap between
18drought-related transcripts and proteins, suggesting the need for complementary approaches to
unveil the mechanisms and molecular plasticity that control drought responses in trees (Plomion et
al. 2006).
Many proteome studies focused on drought response in forest trees assess the role of genotype
in shaping the response. Proteome analysis of eight P. x euramericana genotypes with varying
intrinsic water use efficiencies revealed a number of proteins with significant genotype by treatment
interaction. A large majority of these proteins were found to be chloroplastic in nature and
involved in control of carbon fixation (Bonhomme et al. 2009). While comparing two native
poplar species from China, Yang et al. (2010) examined the combined effect of physiological
and proteome response to drought stress. Although the two species of poplar differed in their
responses, to drought stress, it is evident that physiological and proteomic processes are important
for maintenance of cellular homeostasis under drought conditions. Responses to drought stress are
not limited to the species level; Zhang et al. (2010) identified sex-specific variation in the expression
of drought-responsive proteins in P. cathayana. Many photosynthetic-related and stress-responses
proteins have a significant sex by drought interaction effect (Zhang et al. 2010). The differences
in the proteomic response to drought observed between the sexes may provide insight into the
variability in their productivity as well as their response to drought stress.
Investigations into proteomic variation among genotypes provide excellent insights into the
variability in drought- stress response among trees; however, investigations within a given
individual will shed light on the dynamic nature of plant stress responses. Pechanova et al. (2010)
examined the proteome within apoplastic continuum of P. deletoides. Using a systems approach,
the dynamic nature of the proteome between leaves and stems was investigated and many stress-
responsive proteins were identified. Interestingly, a large constituent of diverse peroxidases thought
to play a role in cell wall modifications were identified. Complexities in the proteome among and
within individual trees highlight the diversity in the drought response. Using system approaches,
combining genomic and proteomic methods of investigation will increase our ability to understand
the complex and dynamic responses of forest trees to drought stress.
2.8 The metabolic drought response
Like the drought-responsive proteome, the complement of metabolites that are identified under
19drought conditions likely play an important role in acclimation process, helping trees tolerate
or avoid stress. Non-targeted profiling of metabolites in biological samples is therefore highly
complementary to transcriptome and proteomic methods in the study of drought responses in
plants (Weckwerth 2003). Examination of the metabolome provides a snapshot of the patterns
of accumulation of metabolites in response to a given biological condition, providing insight into
variation due to treatment or genotype.
Osmotic adjustment is one of the initial responses to water-deficit stress, and metabolites such as
polyols manitol and sorbitol, sugars such as sucrose or raffinose family oligosaccharides (RFOs),
or amino acids such as proline are thought to be of particular importance, serving as osmolytes
or osmoprotectants under drought stress (Seki et al. 2007; Krasensky & Jonak 2012). Small
molecules such as anthocyanins and carotenoids, which accumulate under conditions of drought,
are hypothesized to protect plant tissue from damage caused by reactive oxygen species (Shulaev et
al. 2008). Other small organic molecules, including ABA [see: ‘Plant perception of water status and
downstream signaling pathways’] and jasmonic acid (Ollas et al. 2012)accumulate under drought
stress and serve as signalling molecules activating downstream drought responses.
Plant metabolism is complex and highly dynamic; regulation of metabolic processes occurs at many
different levels (Sweetlove & Fernie 2005). Understanding the regulatory networks underpinning
any aspect of plant metabolism will undoubtedly require detailed analyses of all cellular processes.
Although characterization of the whole plant metabolome is not yet possible, the integration of
comprehensive metabolomic data with other whole-genome platforms will help uncover important
metabolic pathways involved in the drought response (Sweetlove & Fernie 2005; Schauer &
Fernie 2006; Guy et al. 2008). In Arabidopsis, an integrated approach was used to evaluate the
effects of nutritional stress on gene-metabolite networks(Hirai 2005). Correlation networks
revealed specific responses to nutritional deficiencies, including the coordinated modulation of
genes and metabolites in the glucosinolate metabolic pathway (Hirai 2005). Similar approaches
in Populus, uncovered some of the perturbations to molecular networks preceding the onset of
winter (Hoffman et al. 2010). Pathway analysis has also proven fruitful identifying variations in
stress tolerance mechanisms between two poplar species with varying salt-tolerances with respect to
both metabolites and patterns of transcript accumulation (Janz et al. 2010). Although integrated
approaches will ultimately provide a more holistic view of the complex and dynamic metabolic
20networks in trees, improved metabolite profiling techniques and data integration methods are
required for a deeper understanding of these complex networks.
2.9 Recent advances in genome analysis
In the field of genomics, many milestones have been passed, including the sequencing of whole
genomes, which have opened many doors to our understanding of tree molecular biology [for
review, see Deschamps and Campbell (2009)]. The next-generation high-throughput (HTP)
sequencing technologies offer more opportunities to ask other, more complex, biological questions
(Mardis 2008). While the use of poplar as a model species has been highly beneficial to the
understanding of biology and molecular underpinnings of trees to stress, technology is moving
at such a pace that forest biologists can now investigate different tree species with many of the
genomic and technological advantages of a model species. The ability to rapidly sequence genomes
at increased depth and speed allows for the rapid increase in available genomic resources, including
sequence data, physical maps and molecular genetic markers. All these advantages will improve
marker-aided tree breeding and tree improvement methods.
Next-generation HTP sequencing technology not only provides better insight into sequence
variation but also gives us the ability to investigate epigenetic modifications. Epigenetic
modifications, such as DNA or histone modifications, play key roles in regulating gene expression
and, therefore, plant growth and development. Under stress conditions, epigenetic modifications
play important roles regulating the expression of stress-induced genes (Boyko & Kovalchuk 2008;
Chinnusamy & Zhu 2009). Divergent drought transcriptomes and differences in global DNA
methylation in Populus trees of the same genotype is observed between clones propagated in
different geographic locations (Raj et al. 2011). Variation in epigenome reprogramming resulting
in altered gene expression may enable long-lived organisms, such as trees, to better acclimate to
environmental fluctuations (Raj et al. 2011). Some epigenetic modifications are heritable and
may provide a sort of ‘stress-memory’ to plants, allowing them to better cope with future stress
conditions; however, the benefit of being better able to cope with these conditions may be at the
expense of growth (Chinnusamy & Zhu 2009). While genotypic variation for DNA methylation
is observed among poplar hybrids under drought stress and is correlated to productivity under
non-stress conditions (Gourcilleau et al. 2010), understanding the role and the mechanisms by
which epigenetic modifications regulate gene expression under stress conditions is of increasing
21importance.
Synergistic approaches, including the integration of whole-genome expression data with genotyping
data, such as SNP marker analysis, similarly offer the opportunity to link sometimes apparently
disparate biological mechanisms to derive a more holistic description of tree responses to a drought
stimulus. Advances in the next-generation sequencing technology opens doors for the ability to
examine genomic sequences for non-model tree species, as well as many individuals of the same
species to gain insight into sequence variation, as well as epigenetic modifications. These new
technologies create opportunities to increase our wealth of information about stress adaptation and
to delineate the relationships among phenotypic, genetic and epigenetic variation in forest trees.
2.10 Perspectives
As climate and precipitation regimes change and impinge on forest productivity, it is becoming
increasingly clear that understanding how trees adapt and survive under adverse conditions is
important for many reasons. Increased periods of drought stress may limit the ability for trees to
survive; however, with new found knowledge of molecular responses to drought in forest trees, we
may be able to equip ourselves with the ability to plant more resilient stock that is best suited for
future conditions.
Over the past few decades, we have accelerated from the initial discovery of individual genes
involved in a drought response to variations observed at whole transcriptome level. Microarray
studies and other HTP transcriptome analyses have revealed many complexities in the drought
response among forest trees. Links between phenotypic observations and transcriptome responses
reveal potential mechanisms for adaptation to drought. With further efforts in other “-omics”
platforms, investigators had the ability to also examine proteomic and metabolic responses to
drought in trees. Although the response at any given molecular level reveal much information about
how trees respond to drought, the integration of the many various high-throughput platforms may
uncover many complex molecular mechanisms and pathways that underpin the drought response.
A more holistic or systems biology approach will be important in order to understand the relative
importance of various pathways and mechanisms. For example, transcriptome studies reveal many
genes involved in the synthesis of raffinose and galactinol sugars are found with higher transcript
abundance in drought-treated trees (Shinozaki & Yamaguchi-Shinozaki 2007; Hamanishi et al.
222010). These metabolites are thought to have an osmoprotectant role under drought conditions.
Understanding the mechanisms and molecular plasticity of each level of this pathway could be
important to exploit this innate drought protection mechanism in trees.
The ability to capitalize on the new genomics technologies has the potential to lead to strategies
to better preserve existing tree populations, as well as improve the productivity of new stands and
plantations under changing climates. One might make use of these technologies, based on whole-
genome approaches or multi-pronged systems biology approaches in order to identify genes and
gene products related to drought responses. The identified genes, or gene products, can be used for
strategies for selection or directed modification of trees with enhanced capacity to tolerate drought.
For example, identification of drought-resistant QTLs in rice (Bernier et al. 2009) has played an
important role in the marker-aided selection of drought-tolerant rice varieties (Steele 2009). The
identification of genes, such as homologues of the AtMYB61 gene in Arabidopsis thaliana involved in
the closure of stomata, and therefore regulation of water loss (Liang et al. 2005), can be used for the
future modification of tree stocks with enhanced drought tolerance. Using bioinformatic methods,
the relationship of genes from the herbaceous annual Arabidopsis thaliana can be transferred to forest
trees, such as Populus (Wilkins et al. 2009a), and the roles of genes, such as MYB61, can be inferred.
Using these tools, we can engineer trees with enhanced drought tolerance through combination of
genomic, bioinformatics and prior knowledge of drought responses in plants. Improved planting
stock helps improve or maintain productivity in areas influenced by increasing levels of drought.
As well, knowledge of the molecular responses to drought will facilitate in the identification of
naturally occurring variation in the drought response. This variation can be selected for or used as
a focus for conservation in forests. In this period of uncertainty about our climatic future, genomic
approaches that enable us to enhance and increase precision and improve rates of identification of
resilient individuals will be of paramount importance.
2.11 Acknowledgements
We are most grateful for very useful comments on the draft manuscript provided by two anonymous
reviewers. Research in the Campbell laboratory is generously supported by the Natural Science and
Engineering Research Council of Canada (NSERC), the Canada Foundation for Innovation (CFI),
the Ontario Research Fund (ORF), Genome Canada and the University of Toronto.
23
Chapter 3: Intraspecific variation in the Populus balsamifera drought transcriptome
Contents of this chapter have been published in Plant, Cell and Environment: Erin T. Hamanishi,
Sherosha Raj, Olivia Wilkins, Barb R. Thomas, Shawn D. Mansfield, Aine L. Plant, Malcolm M.
Campbell. 2010. Intraspecific variation in the Populus balsamifera drought transcriptome. Plant,
Cell and Environment. 33: 1742-1755
Contributions: ETH, SDM, ALP and MMC designed research; ETH, SR, BT, SDM, ALP and
MMC organised experimental logistics including transfer and establishment of biological materials;
ETH, SR, and OW performed research; ETH, OW, and MMC analysed data; ETH and MMC
wrote manuscript with editorial assistance from SR, OW, BT, SDM, ALP and MMC.
The published paper and supplementary files can be found online at
http://onlinelibrary.wiley.com/doi/10.1111/j.1365-3040.2010.02179.x/suppinfo
Supplemental tables and figures are numbered in the order in which they appear online and in the
published paper. Supplemental figures S3.1 through S3.3 and supplemental tables S3.1 through
S3.3 are also included at the end of this chapter.
The material in this chapter is © by the Wiley-Blackwell Publishing Limited.
24
Chapter 3: Intraspecific variation in the Populus balsamifera drought transcriptome
3.1 Abstract
Drought is a major limitation to the growth and productivity of trees in the ecologically and
economically important genus Populus. The ability of Populus trees to contend with drought is
a function of genome responsiveness to this environmental insult, involving reconfiguration of
the transcriptome to appropriately remodel growth, development and metabolism. Here we test
hypotheses aimed at examining the extent of intraspecific variation in the drought transcriptome
using six different Populus balsamifera L. genotypes and Affymetrix GeneChip technology. Among
genotypes there was a positive correlation between the magnitude of water-deficit induced changes
in transcript abundance across the transcriptome, and the capacity of that genotype to maintain
growth following water deficit. Genotypes that had more similar drought-responsive transcriptomes
also had fewer genotypic differences, as determined by microarray-derived single feature
polymorphism (SFP) analysis, suggesting that responses may be conserved across individuals that
share a greater degree of genotypic similarity. This work highlights the fact that a core species-level
response can be defined; however, the underpinning genotype-derived complexities of the drought
response in Populus must be taken into consideration when defining both species- and genus-level
responses.
3.2 Introduction
Trees of the genus Populus, which include poplars, aspen and cottonwoods (herein collectively
referred to as poplars), are found primarily throughout the northern hemisphere (Dickmann 2001),
and have many favourable attributes which have lead to their widespread use in both forestry and
agriculture (Brunner et al. 2004). Occupying both a large geographical region and a diverse array of
habitats, poplar trees must contend with a variety of environmental conditions in order to survive.
Along with other environmental stresses, such as insect defoliation and rust cankers, drought is a
major factor impinging on poplar growth, productivity and survival throughout its range (Hogg et
al. 2002; van Mantgem et al. 2009). The drought sensitivity of poplar trees, which are a prominent
species in many temperate forests, has posed increased concern for the future amidst predictions of
changing climate and water shortages, as well as during periods of episodic drought (Schindler &
25Donahue 2006).
In response to water deficit, poplar trees generally display alterations in plant water status resulting
in suppression of stomatal conductance and declines in productivity; however, considerable
variation in drought response and tolerance has been observed within the genus Populus (Gebre
& Kuhns 1991; Tschaplinski et al. 1994; Chen et al. 1997; Monclus et al. 2006; Bonhomme et al.
2009). The ability of trees to adapt and survive diverse environmental variables, such as drought,
is a consequence of a variety of biochemical and physiological processes, many of which are the
result of stress signal perception leading to alterations in the transcriptome, resulting in an adaptive
response (Dickmann 2001; Kozlowski & Pallardy 2002; Lei et al. 2006). The variability in response
to drought observed at the morphological and physiological level suggests that poplar trees are an
excellent organism to study the molecular underpinnings of the drought response and the variation
in this adaptive response to such an environmental insult.
Variation in gene expression observed among populations is heritable (Oleksiak et al. 2002; Brunner
et al. 2004; Whitehead & Crawford 2006) and, therefore, examination of the variability in the
transcriptome response to drought may provide insight into the diversity and adaptation to such a
response. Previous investigations dissecting the molecular underpinnings of the drought response
in the genus Populus have focused on transcriptional differences among various poplar species or
hybrids (Street et al. 2006; Wilkins et al. 2009b). This indicates that poplar trees, regardless of
species or hybrid, likely have a variety of mechanisms governing the drought response, and that this
response is highly dependent on genotype. Although there is evidence that intra-specific variation
in drought response can be observed among growth rates and physiological traits within a species of
Populus (Schindler & Donahue 2006; Lu et al. 2009), drought-induced variation in gene expression
within a given species of Populus has not yet been investigated. Here, we explore the variation
in transcriptome responses to drought within the species Populus balsamifera L. spp. balsamifera
(balsam poplar). P. balsamifera is a dominant tree species within North America’s boreal ecosystems
whose range is transcontinental, and can be found growing on upland riparian sites (Gebre &
Kuhns 1991; Tschaplinski et al. 1994; Chen et al. 1997; Monclus et al. 2006; Bonhomme et al.
2009; USDA-NRCS 2009). On account of its similarity to black cottonwood (Populus trichocarpa
Torr. & Gray) and the wealth of available genomic tools (Tuskan et al. 2006), P. balsamifera
represents an ideal species for studying intra-specific variation in the drought response.
26In this study, Affymetrix GeneChip technology was used to study the drought responsive
component of the Populus transcriptome. Six P. balsamifera genotypes from various geographical
regions (Figure 3.1) were grown in a common growth chamber environment and the variation in
gene expression in trees responding to drought was examined. These experiments aimed to test the
hypothesis that there are significant differences in the trancriptomes of P. balsamifera genotypes
in response to water-deficit conditions; however a common species-level response could also be
assessed. We expect that the knowledge about transcriptome variation within a given species of
poplar will contribute to our understanding of the adaptive responses to drought and the molecular
underpinnings of these responses.
3.3 Materials and Methods
3.3.1 Plant Material
Dormant, 25 cm, un-rooted hardwood cuttings of six P. balsamifera genotypes (AP-947, AP-1005,
AP-1006, AP-2278, AP-2298, AP-2300) were obtained from Alberta-Pacific [Forest Industries
Inc. (Al-Pac), Boyle, AB, Canada]. Cuttings were imbibed for 48 h prior to planting (Desrochers
& Thomas 2003) into Sunshine Mix-1 (Sun Gro Horticulture Inc, Bellevue, WA, USA; http://
www.sungro.com) in 1 m opaque pots (10.5 cm diameter). The plants were grown in a climate-
controlled growth chamber under long day conditions (16 h photoperiod, light intensity: 178-
220 µmol m-2 s-1), with a maximum day temperature of 22 °C and a minimum night temperature
of 17 °C throughout the experiment. All plants were watered every 2 to 3 d to field capacity and
fertilized (20:20:20, N-P-K, 1.5 g L-1, 600 mL plant-1) every 3 weeks. All plants were grown for 9
weeks prior to the onset of the water-withholding experiment, at which point they were divided
into two groups, well watered (WW; n = 27–35 per genotype) and water deficit (WD; n = 27–35
per genotype). Water deficit conditions were imposed on dry plants by withholding water; wet
plants were regularly watered every 2 to 3 d to maintain water status. Fifteen days following the
water withholding, the first fully expanded leaf was harvested from three individual trees from each
genotype for both well watered and water-deficit treatments at two time points: midday (MD;
middle of the light period) and pre-dawn (PD; 1 h before the lights were turned on). Leaves were
pooled and immediately flash frozen in liquid nitrogen for subsequent analysis. This was repeated
three times in order to achieve triplicate replicates for each genotype-treatment combination at each
27
AP 1005, AP 1006
AP 947
AP 2298AP 2278
AP 2300
Figure 3.1 Source of origin of the six P. balsamifera genotypes examined in this study
28time point.
3.3.2 Physiological and growth traits
A portable infrared gas analyser (IRGA; LI-6400, LI-COR Biosciences Inc., Lincoln, NE, USA) was
used for measuring photosynthesis, stomatal conductance (gs) and transpiration. Beginning at the
start of the water-withdrawal experiment, 9 weeks after planting, measurements were taken daily
throughout the experimental period (n = 4–7 individuals per genotype/treatment). Measurements
were on mature, fully expanded leaves at the midday time-point. Productivity and relative water
content (RWC) measurements were made 15d after the onset of the water- withdrawal experiment
on both well-watered and water-deficit-treated plants. Productivity was measured by determination
of tree height, stem circumference and total aboveground dry-weight biomass. Data analysis was
performed using R (R Development Core Team 2009). Means were calculated with their standard
error (SE), and compared using a two-way ANOVA. Genotype and treatment were considered as
the main factors; differences between treatments and among genotypes were determined using a
TukeyHSD test. Leaf RWC was calculated on a mature fully expanded leaf (n = 5 individuals per
genotype/treatment). Fresh weight (FW) was recorded, and the leaf was allowed to rehydrate in
distilled H2O for 24 h in the dark in order to obtain turgor weight (TW). Leaf dry weight (DW)
was obtained after the leaf was dried at 70 °C for 48 h. RWC was calculated according to Barrs &
Weatherley (1962) as: RWC (%) = (FW - DW) * 100%/(TW - DW).
3.3.3 RNA extraction, microarray hybridisation and analysis
Total RNA was isolated from fully expanded leaves of P. balsamifera and hybridized to an Affymetrix
Poplar GeneChip (Affymetrix, Santa Clara, CA, USA) at the Center for the Analysis of Genome
Evolution & Function (CAGEF) at the University of Toronto as described by Wilkins et al.
(2009b). GeneChip expression analysis was performed using the Bioconductor (Gentleman et
al. 2004) software package AFFY (Gautier et al. 2004) in (R Development Core Team 2009) as
described in Wilkins et al. (2009b). All 72 arrays were pre-processed together using GC-robust
multi-array analysis (gcrma; Wu et al. 2004). Expression data was filtered to eliminate probe sets
with low levels of variation across samples and low levels of expression according to Wilkins et al.
(2009b). The preprocessed data was analysed as a 6x2x2 factorial ANOVA design (six genotypes,
two treatments and two time points) using the linear models for microarray package (LIMMA;
29Smyth 2004) package in R (R Development Core Team 2009). Treatment, genotype and time point
were considered the main factors. Differential expression in response to water deficit was determined
using an empirical-Bayes moderated t-statistic with a Benjamini and Hochberg adjustment to
control the false discovery rate (adjusted P value cut-off of 0.05; (Smyth 2004). In order to take into
consideration the magnitude of differential expression for genes that are significantly differentially
expressed for treatment main effect only, probe sets were filtered according to a t-test threshold,
which corresponds to a minimum fold-change of 2.0 (TREAT; McCarthy & Smyth 2009). Genes
were annotated using the Annotation for Probe Sets in PLEXdb (Wise et al. 2007) and Annotation
Batch Function in PopGenie (Sjodin et al. 2009). All samples were uploaded to Gene Expression
Omnibus (http://www.ncbi.nlm.nih.ov/geo/); accession number GSE21171.
3.3.4 Single-feature polymorphism (SFP) analysis
SFPs were identified using pair-wise comparisons between genotypes using Affymetrix Poplar
GeneChip arrays for well-watered, midday poplar samples according to Fujisawa et al. (2009). The
number of SFPs were identified for all probe sets that passed through the filtering criteria, as well as
for probe sets that were either significantly differentially expressed for genotype main effect, or not.
3.3.5 DNA extraction and simple-sequence repeat (SSR) analysis
Total DNA was extracted according to Doyle & Doyle (1990). Seven SSR microsatellite loci were
used to fingerprint the six P. balsamifera genotypes. Five of the seven loci mapped to distinct linkage
group in the Populus genome (Tuskan et al. 2004); however, the remaining two have no informative
mapping information. Electrophoresis-based SSR genotyping was performed by The Centre for
Applied Genomics, The Hospital for Sick Children, Toronto, Canada.
3.4 Results and Discussion
3.4.1 There is intraspecific variation in the productivity and physiological responses in Populus balsamifera following water deficit
To investigate the intraspecific variation of P. balsamifera in response to water deprivation, a multi-
factorial experiment was conducted using six P. balsamifera genotypes. The six genotypes examined
originated from five distinct geographic regions in Western Canada (Figure 3.1), with varying
30climatic histories (Table 3.1). Each P. balsamifera genotype was genetically unique based on SSR
microsatellite fingerprinting (Supplementary Table S3.1).
After 15 d of withholding water, decreased productivity (Figure 3.2) and stomatal conductance
(Figure 3.3) were observed between well-watered plants and those deprived of water. Multi-
factor ANOVA analysis for aboveground biomass, plant height and stem circumference revealed
a significant genotype effect for all three variables; whereas, only significant treatment effect for
aboveground biomass (P = 0.05, data not shown). In response to water-deficit, all six genotypes
exhibited significant differences in midday stomatal conductance (gs); however, genotype AP-1006
had significant differences as early as five days after the onset of the water-deficit treatment (P =
0.1), whereas genotype AP-2300 did not exhibit significant differences until 11 d after the onset of
water-deficit conditions (Figure 3.3). Genotypes AP-1006 and AP-2278 showed striking differences
in gs, between well watered plants and plants grown under water-deficit conditions at day 15. The
differences in gs between well-watered and water-limited plants observed in other genotypes, such
as AP-947 and AP-1005, was less marked. By fifteen days after the onset of the water-withholding
experiment, relative water content (RWC) was significantly lower in genotypes AP-1005, AP-
1006, AP-2298 and AP-2300 (P = 0.05, Supplementary Table S3.2). Phenotypic responses, both
physiological and morphological, to water-deficit treatment in P. balsamifera showed no significant
correlation with historic climatic conditions and geographic origin (Supplementary Figure S3.1).
This may reflect the high level of variation among these genotypes in their ability to respond to
environmental stimuli regardless of their origins. Although their responses do not reflect historic
origins, the variability observed may an important trait for survival in fluctuating environments
(Clark 2010). Moreover, the variation in such adaptive traits may be particularly important in P.
balsamifera as population genotypic variation and effective population size is considered low (Olson
et al. 2010).
Similar reductions in gs and RWC, as well as photosynthetic capacity, have been demonstrated in
poplar and other plant species under drought stress conditions (Duan et al. 2005; Giovannelli et al.
2007); however, in poplar, suppression of gs and photosynthesis occur well before changes in whole
leaf water status are observed (Tardieu & Simonneau 1998; Giovannelli et al. 2007). The regulation
of stomatal conductance and water status in poplar may represent an important trait for survival
under fluctuating environments, particularly the extreme stresses induced by episodic drought.
31
0.00
1.00
2.00
3.00
4.00
5.00
6.00
AP-947 AP-1005 AP-1006 AP-2278 AP-2298 AP-2300
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
AP-947 AP-1005 AP-1006 AP-2278 AP-2298 AP-2300
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
AP-947 AP-1005 AP-1006 AP-2278 AP-2298 AP-2300
Abov
egro
und
biom
ass
(g D
W)
Plan
t hei
ght (
cm)
Stem
circ
umfe
renc
e (m
m)
abc abc
ac
abc
abcabc abc
abd
b b
d
c
e efefg efg efg efg
ef fef ef
gg
hi hi
hihi
hihi
hi
h hh
h
i
Figure 3.2 Above ground biomass (a), plant height (b) and stem circumference (c) of six genotypes
of P. balsamifera were calculated 15 d after the onset of the water-withdrawal experiment for both
well watered (blue bars) and water deficit treated (orange bars) plants. Significant differences
between genotypes and treatments (P < 0.05) are denoted by small letters for all variables. Mean
values and SE bars are represented. Figure originally published in black and white.
32
0.0
0.1
0.2
0.3
0.4
0.5 AP-947
0.0
0.1
0.2
0.3
0.4
0.5 AP-2278
0 1 3 5 7 9 11 13 15
AP-1005
AP-2298
AP-1006
AP-2300
0 1 3 5 7 9 11 13 15 0 1 3 5 7 9 11 13 15
*
*
**
*
*
*
** * **
*******
***
******
******
**** *
Con
duct
ance
(mol
H2O
m-2 s
-1)
Days since the onset of WD
Figure 3.3 Box plot of the variation in midday leaf stomatal conductance for six P. balsamifera
genotypes: (a) AP-947 (b) AP-1005 (c) AP-1006 (d) AP-2278 (e) AP-2298, and (f ) AP-2300.
Midday stomatal conductance for well watered plants (blue boxes) and plants grown under water-
deficit conditions (orange boxes) are represented. Asterisks indicate significant difference between
well-watered and water-deficit-treated plants: *P < 0.1; **P < 0.05; ***P < 0.001. WD, Water-deficit
treatment.
33Table 3.1 Location and climate variables
Clone Lattitude LongitudeElevation (m)
Mean Annual Temperature (°C)
Mean Annual Precipitation (mm)
Degree Days (>5°C)
AP-947 55° 38' 3 .13" 113° 23' 53 .68" 786 1 .8 565 1193AP-1005 55° 24' 25 .95" 114° 36' 19 .67" 630 1 .8 538 1258AP-1006 55° 24' 25 .95" 114° 36' 19 .67" 630 1 .8 538 1258AP-2278 58° 46' 13" 123° 4' 21" NA* -0 .2 533 1294AP-2298 59° 11' 19" 122° 46' 35" NA* -0 .9 468 1255AP-2300 58° 51' 14" 122° 31' 28" NA* -1 .1 439 1241
Location details and historic climatic variables adjusted for specific location and elevation using the
Climate BC model described by Wang et al. (2006). * no elevation data available.
34The marked differences in gs at day 15 suggest variability in the regulation of stomatal control and
physiological response to drought. Intraspecific variation in acclimation strategies in response to
drought has been observed for many other tree species (Beikircher & Mayr 2009). Regulation of
morphological and physiological parameters in response to water-deficit may provide insight into
the hydraulic strategies of the various P. balsamifera genotypes. The observed variation in stomatal
conductance, and other physiological and morphological variables in response to water-deficit,
suggested that these genotypes deployed various drought tolerance and acclimation strategies. To test
this hypothesis, the variation in the transcriptome response to water-deficit conditions among the
six P. balsamifera genotypes was examined.
3.4.2 Water deficit conditions elicit significant responses within the P.
balsamifera transcriptome
Transcriptome analysis can provide insights into similarities and differences in the mechanisms
underpinning the response to water-deficit between groups of individuals. Variation in the
molecular mechanism regulating the drought response in Populus suggests that genotype plays an
important role in shaping the drought transcriptome (Street et al. 2006; Wilkins et al. 2009b).
Comparison of the drought transcriptome of two Populus hybrid genotypes indicate that there
is indeed a level of conservation in the transcriptome response; however, the variable response of
a given genotype cannot be overlooked (Wilkins et al. 2009a). In this study we hypothesize that
conserved transcriptome level responses to drought will be observed, and that the differences
observed in the drought transcriptomes that are specific to an individual genotype may provide
valuable insight into the molecular basis of ecologically important variation in the drought response.
Using Affymetrix Poplar GeneChip microarrays, we investigated the transcript-level response to
water-deficit among six P. balsamifera. Employing a multi-factorial ANOVA design (adjusted P
< 0.05), 280 probe sets reported on genes with significant differential transcript accumulation in
response to water-deficit conditions, irrespective of the effect of genotype or sampling time (Figure
3.4; Supplementary Table S3.4). Many more probe sets were considered differentially expressed
when no minimum threshold cutoff was applied; however, many of those probe sets had very low
levels of differential accumulation of transcripts in response to water deficit. Filtering significant
probe sets using a minimum threshold cutoff allows the identification of genes that are consistently
differentially expressed and may prove to be more biologically meaningful (McCarthy & Smyth
35
AP947.Wet.MDAP947.Wet.MDAP947.Wet.MDAP947.Dry.MDAP947.Dry.MDAP947.Dry.MDAP1005.Wet.MDAP1005.Wet.MDAP1005.Wet.MDAP1005.Dry.MDAP1005.Dry.MDAP1005.Dry.MDAP1006.Wet.MDAP1006.Wet.MDAP1006.Wet.MDAP1006.Dry.MDAP1006.Dry.MDAP1006.Dry.MDAP2278.Wet.MDAP2278.Wet.MDAP2278.Wet.MDAP2278.Dry.MDAP2278.Dry.MDAP2278.Dry.MDAP2298.Wet.MDAP2298.Wet.MDAP2298.Wet.MDAP2298.Dry.MDAP2298.Dry.MDAP2298.Dry.MDAP2300.Wet.MDAP2300.Wet.MDAP2300.Wet.MDAP2300.Dry.MDAP2300.Dry.MDAP2300.Dry.MD
AP2300.MD
AP2298.MD
AP947.MD
AP1005.MD
AP1006.MD
AP2278.MD
−20
2R
ow Z
−Sco
re
Col
or K
ey
Valu
e1
23
4
Col
or K
ey
(b) Relative fold-change transcript abundance at mid day
(a) Relative transcript abundance at mid day
36
AP2300.PD
AP2298.PD
AP947.PD
AP1005.PD
AP1006.PD
AP2278.PD
−20
2R
ow Z
−Sco
re
Col
or K
ey
Valu
e1
23
4
Col
or K
ey
AP947.Wet.PDAP947.Wet.PDAP947.Wet.PDAP947.Dry.PDAP947.Dry.PDAP947.Dry.PDAP1005.Wet.PDAP1005.Wet.PDAP1005.Wet.PDAP1005.Dry.PDAP1005.Dry.PDAP1005.Dry.PDAP1006.Wet.PDAP1006.Wet.PDAP1006.Wet.PDAP1006.Dry.PDAP1006.Dry.PDAP1006.Dry.PDAP2278.Wet.PDAP2278.Wet.PDAP2278.Wet.PDAP2278.Dry.PDAP2278.Dry.PDAP2278.Dry.PDAP2298.Wet.PDAP2298.Wet.PDAP2298.Wet.PDAP2298.Dry.PDAP2298.Dry.PDAP2298.Dry.PDAP2300.Wet.PDAP2300.Wet.PDAP2300.Wet.PDAP2300.Dry.PDAP2300.Dry.PDAP2300.Dry.PD
(d) Relative fold-change transcript abundance at pre-dawn
(c) Relative transcript abundance at pre-dawn
37
Figure 3.4 Heat maps representing transcript abundance of all drought responsive probe sets in
six P. balsamifera genotypes: AP-947, AP-1005, AP-1006, AP-2278, AP-2298 and AP-2300. Only
probe sets that are significant for treatment main effect, irrespective of time of day or genotype,
and are differentially expressed relative to a given threshold are represented (n = 280; FDR = 0.05,
log2(fold-change)-cutoff = 2.0) for both time points: (a, b) mid day, and (c, d) pre-dawn. Row
normalized, transcript abundance for all drought responsive probe sets at (a) mid day and (c) pre-
dawn. Each column represents a biological sample, and all treatments are represented in triplicate
replicates. Red indicates increased transcript abundance; blue indicates decreased transcript
abundance. Data are row normalized. Heat maps representing mean relative fold-change transcript
abundance for all genotypes at (b) mid day and (d) pre-dawn. Dark blue indicates increased
mean transcript abundance in water-deficit treated samples relative to well-watered samples; white
indicates decreased mean transcript abundance in water-deficit treated samples relative to well-
watered samples. Rows are clustered using Pearson correlation for all heat maps.
382009). Consistent with previous findings among species and hybrids of Populus (Street et al. 2006;
Wilkins et al. 2009b), a larger proportion of genes (~5000 probe sets) had significant differences in
transcript abundance for the main effect of genotype relative to the effect of water-deficit treatment
alone or the genotype x treatment interaction together. Genotype was known to play an integral role
in shaping the drought response in Populus between species or hybrids (Street et al. 2006; Wilkins
et al. 2009b). The current study extends this finding, and highlights the importance of genotype in
shaping the water-deficit response within a given Populus species.
3.4.3 There is a common P. balsamifera drought transcriptome
Previously, identification of a common drought transcriptome in the genus Populus was challenging
because of extensive variation in the transcriptome between hybrid genotypes, as well as variation in
the transcriptome-level water-deficit response that is time of day dependent (Wilkins et al. 2009b).
By contrast, comparison of the drought transcriptome across six genotypes of P. balsamifera, at two
time points, revealed a common transcriptome-level response to water-deficit treatment within this
species (Figure 3.4, Supplementary Tables S3.3 and S3.4). The common response genes that were
identified in this comparison had a significant change in transcript abundance in response to water
deficit that was genotype independent (i.e. corresponded to main effect of treatment irrespective of
genotype in ANOVA).
The functional roles of the probe sets that are significant for the treatment main effect for all
genotypes (FDR = 0.05) and also show differential transcript abundance in response to water deficit
according to a minimum threshold cutoff [log2 (fold-change) of 2.0] were classified using GO
categories (Berardini et al. 2004; Supplementary Figure S3.2). Overall, the largest proportion of
probe sets with increased transcript abundance under water deficit conditions were those categorized
as ‘other cellular processes’. By contrast, probe sets with decreased transcript abundance under water
deficit conditions largely fell into the ‘protein metabolism’ category. Interestingly, for both probe
sets, with increased and decreased transcript abundance under water deficit conditions, a large
proportion were categorized as ‘response to abiotic or biotic stimulus or stress’, in keeping with their
involvement in the water deficit response.
Many of the genes comprising the common water-deficit response, genotype-independent P.
balsamifera transcriptome corresponded to genes previously identified as drought responsive in other
39plants (Kreps et al. 2002; Bray 2004; Street et al. 2006; Bogeat-Triboulot et al. 2007; Wilkins et al.
2009a). Of the 280 probe sets that were significant for the main effect of treatment irrespective of
genotype, with a minimum log2 fold-change of 2.0, 29% corresponded to water-deficit-responsive
genes identified in a similar experiment conducted with two hybrid poplar genotypes (Wilkins et
al. 2009b). In comparison to other high-throughput experiments examining the drought response
in Populus (for example, see: Brosche et al. 2005; Street et al. 2006; Bogeat-Triboulot et al. 2007) a
much more limited shared response was identified for P. balsamifera, as was observed by Wilkins et
al. (2009b).
Of the 98 probe sets that reported increased transcript abundance in response to water-deficit in
this study, several of particular interest include those involved in the production of galactinol and
stachyose, including GALACTINOL SYNTHASE and STACHYOSE SYNTHASE. The expression
of genes encoding enzymes involved in the production of sugars from the raffinose family of
oligosaccharides was also water-deficit responsive in hybrid poplar (Wilkins et al. 2009b). Raffinose-
derived oligosaccharides are believed to function as osmoprotectants during drought stress (Taji et
al. 2002; Nishizawa et al. 2008). Increased transcript abundance of genes encoding enzymes that
produce these compounds is consistent with a plant that is attempting to counteract water deficit.
In keeping with a plant mounting a water-deficit response, the common P. balsamifera water-deficit
transcriptome also included genes homologous to gene families in Arabidopsis with key roles in
adjusting water balance in response to water-deficit including ABA RESPONSIVE ELEMENT
BINDING FACTOR 4 (ABRE4) and EARLY RESPONSIVE TO DEHYDRATION 7 (Bray 2004).
The phytohormone abscisic acid (ABA) has an extremely well established role in plant drought
signalling (Bray 2004; Mahajan & Tuteja 2005). The increased transcript abundance of genes
implicated into the ABA signalling pathway in P. balsamifera in response to water deficit emphasizes
the central role of this compound in the drought response across diverse taxa.
A large number of probe sets with decreased transcript abundance in response to water-deficit in
the common P. balsamifera water-deficit transcriptome were homologous to genes involved in cell
wall modification, including pectin esterases and endoxyloglucan transferases. Decreased transcript
abundance of these classes of genes is thought to decrease cell wall extensibility by promoting cell
wall loosening or stiffening (Micheli 2001) and by controlling the cleavage of xyloglucan chains
(Hyodo et al. 2003), respectively. Genes involved in cell wall modification have previously been
40identified as drought responsive in A. thaliana (Bray 2004). A large proportion of genes identified in
the conserved P. balsamifera drought response have unknown function based on homology to other
plant species. These genes that appear to be drought responsive in P. balsamifera may represent a
specific water-deficit response for this species.
3.4.4 There is a notable significant variation in the drought transcriptome across
P. balsamifera genotypes
While a core set of probe sets representing a common species-level response to water-deficit
treatment was identified, many probe sets reported differential expression between P. balsamifera
genotypes. That is, when the water-deficit transcriptomes of the six P. balsamifera genotypes
were compared using a Pearson correlation based on all drought-responsive probe sets (FDR =
0.05), some genotypes were more closely related than others with respect to their water-deficit
transcriptome (Figure 3.5). The similarity between all genotypes was still quite high, with minimum
Pearson correlation coefficient values for any given pair-wise comparison > 0.6. Genotypes AP-
1005, AP-1006 and AP-2278 had the most similar drought transcriptomes; whereas, genotype AP-
2300 was the most distinct with respect to the transcriptome response to water-deficit relative to the
other genotypes.
Various patterns of gene expression in response to drought were identified: genes with increased
levels of transcript abundance for all genotypes, those with decreased transcript abundance across
all genotypes and those with differential drought responsiveness between the six P. balsamifera
genotypes. However, in the conserved set of probe sets that were drought responsive regardless
of genotype (treatment main effect), notable differences in the mean log2 fold-change between
well watered and water-deficit treated samples were observed. The magnitude of change in gene
expression in response to water-deficit treatment varied considerably between genotypes (Figure
3.6). For example, genotypes AP-1006 and AP-2278 had significantly larger fold-changes in
transcript abundance levels relative to other genotypes for those genes with significant differences in
transcript abundance in response to water deficit. This suggests that there was significant variation in
not only transcript abundance of P. balsamifera genes that exhibited a significant treatment-genotype
interaction, but also in the magnitude of intraspecific differences in gene expression for those genes
whose change in transcript abundance was attributable to treatment main effect alone. This is to say
that the variation in P. balsamifera water-deficit transcriptomes across the species is attributable to
41
AP-2
300
AP-9
47
AP-2
298
AP-1
005
AP-1
006
AP-2
278
AP-2300
AP-947
AP-2298
AP-1005
AP-1006
AP-2278
0.6 0.7 0.8 0.9 1Value
Colour Key
Figure 3.5 Pearson correlation coefficient (PCC) heat map representing the P. balsamifera drought
transcriptome responses. Differential transcript abundance between well watered and water-deficit
samples for the six genotypes for the drought responsive probe sets are represented (Treatment
main effect; FDR = 0.05, log2(fold-change) cutoff = 2.0, n = 280 probe sets). Differential transcript
abundance was calculated as the mean log2(fold-change) between well watered and water-deficit
samples for a given probe-set. The PCC was determined for each pair-wise comparison, and is
represented by the colour in the corresponding cell. All samples are represented on both the x- and
y-axis, in the same order.
42
AP-947 AP-1006 AP-2298
−8
−6
−4
−2
0
2
Genotype
−2
0
2
4
6
8
Genotype
log 2
(FC
)
AP-2300AP-2278AP-1005 AP-947 AP-1006 AP-2298 AP-2300AP-2278AP-1005
log
(FC
)2
(a) (b)
Figure 3.6 Box plot illustrating the interplay of genotype and treatment in shaping the drought
transcriptome of six P. balsamifera genotypes. The average log2(fold-change) between well watered
and water-deficit treated samples for all genes identified as significantly differentially expressed
for treatment main effect (FDR = 0.05, log2(fold-change)-cutoff = 2.0, n = 280 probe sets) for
probe sets with (a) decreased transcript abundance in response to WD and (b) increased transcript
abundance in response to WD at the midday time point.
43both qualitative and quantitative changes in transcript abundance.
Qualitative variation was observed among the genotypes in the representation of genes in different
functional categories in the drought transcriptomes. Qualitative differences (i.e. individual gene
identities) in the transcriptomes of each genotype were analysed as a two-factor ANOVA, where
treatment and time point were the two main factors. Genotype-specific responses to water deficit
treatment emerged from these analyses (Supplementary Table S3.3). Notably, across all genotypes, a
large number of genes were predicted to be involved in protein metabolism, or response to biotic or
abiotic stimulus in each genotype, similar to the results observed for the treatment main effect across
all six genotypes (i.e. the common transcriptome). Nevertheless, across genotypes there was variation
in the representation of given GO functional categories, with some GO categories more populated
by drought transcriptome genes of some genotypes relative to other genotypes. This underscores
the fact that there were qualitative differences in the nature of the drought transcriptomes across
genotypes, with each genotype having a ‘GO fingerprint’ that was broadly similar to the other
clones, but still relatively unique.
While natural variation in the transcriptome response to various environmental stimuli has not been
documented in poplar, it has previously been described in A. thaliana (Kreps et al. 2002; Hannah et
al. 2006; van Leeuwen et al. 2007). Variation in the transcriptome response to cold stress between
various accessions of A. thaliana highlighted the complexities of such a response. These stress-
induced alterations in gene expression suggested that not only differential expression of genes, but
also the variation in the magnitude of expression is likely to influence the variation in acclimation
capacity of these accessions (Hannah et al. 2006). Consistent with this, analysis of the A. thaliana
salicylic acid response revealed extensive transcriptome variation, where relatively few genes
responded similarly across the A. thaliana accessions (van Leeuwen et al. 2007). These findings
suggested that A. thaliana ecotypes differentiate to a greater extent in terms of environmental
responsiveness, such that each ecotype is well matched to local environmental conditions. The
higher level of commonalities in the P. balsamifera water-deficit transcriptome response is likely a
consequence of the relatively low population genetic variation found in P. balsamifera (Olson et al.
2010). The higher level of commonalities in the P. balsamifera water-deficit transcriptome response
may be reinforced by the fact that a broad range, long-lived species, like P. balsamifera, must retain
a more generalist response to environmental stimuli across its range. Future studies could test this
44hypothesis by examining the intraspecific variation in P. balsamifera transcriptome responses to a
wider variety of environmental stimuli.
3.4.5 Time of day shapes the P. balsamifera drought transcriptome
The P. balsamifera water-deficit transcriptome is not only shaped by genotype, but also by the
time of day. The interaction between time point and water-limitation revealed 129 probe sets with
significant differences in transcript abundance (Supplementary Table S3.3); however, when the
interaction between the time of day and treatment was assessed within each individual genotype,
particular genotypes, such as AP-1006, had a larger cohort of probe sets with differential transcript
abundance relative to others, such as AP-2298. As previously observed with hybrid poplar
genotypes, the transcriptome-level response to water-deficit conditions were influenced by time of
day, and time of day was an important factor when considering the conserved drought response in
Populus (Wilkins et al. 2009b). However, in P. balsamifera, the time of day treatment interaction
was less significant than that observed previously between the hybrid poplar genotypes (Wilkins et
al. 2009b). Hybrid phenotypes that are more extreme than the parental phenotypes is defined as
transgression (deVicente and Tanksley 1993). The magnitude of transcriptome differences observed
with the hybrid poplar genotypes relative to that observed between the P. balsamifera genotypes
may be attributable to transgressive effects. Transgressive effects have been observed in interspecific
hybrids in other plant genera (Lai et al. 2006), and might be expected in the hybrid poplars, but
would be lacking in the pure P. balsamifera genotypes.
3.4.6 The extent of transcriptome-wide transcript abundance change enables the
P. balsamifera to sustain growth under water-deficit conditions
This study demonstrates the complexities of the drought response within a given species of Populus.
Genotypes with strong physiological responses to water-deficit conditions tended to have increased
magnitude change in expression of genes that were significant for the treatment main effect (Figure
3.6, Supplementary Figure S3.1b). Genotype AP-1006 had the most rapid decline in stomatal
conductance in response to the imposition of water-deficit conditions, and also showed the largest
mean log2(fold-change) between well-watered and water-deficit treated samples for all probe
sets significant for treatment main effect. Correlation between magnitude of cold tolerance and
amplitude of gene expression [log2(fold-change)] has been observed among Arabidopsis accessions
45(Hannah et al. 2006). It has been hypothesized that increased capacity for cold acclimation may be
related to the observed changes in the transcriptome; however, supporting evidence revealed that
reduced cold acclimation in accessions with reduced capacity for cold tolerance was not supported
by metabolic activity. This suggests that the mechanisms that form the foundations of complex
phenotypic traits, such as cold tolerance, or drought acclimation are likely controlled by a large
number of transcriptome changes, rather than individual genes.
Tree growth, another complex phenotypic trait, is also underpinned by genetic factors that respond
to environmental stimuli (Grattapaglia et al. 2009). Consistent with this, the capacity of P.
balsamifera to sustain growth during drought was positively correlated (R2 = 0.776, P = 0.02) with
the magnitude of change in transcript abundance across the remodeled transcriptomes (Figure 3.7).
This suggests that it is not merely the nature of genes that enables plant growth during drought,
but also the magnitude of change in transcript abundance for genes that are drought responsive.
While most studies emphasise the importance of changes in the specific ‘cohort’ of genes expressed
in response to a stress stimulus, the results presented here indicate that the magnitude of change in
transcript abundance for all genes across the transcriptome is every bit as important in buffering
the response. It is noteworthy that sustained growth under drought conditions and the magnitude
of drought-induced, transcriptome-wide changes transcript abundance were the only two factors
that showed a significant correlation in this study (Supplementary Figure S3.1). This underlines
the often-overlooked role of magnitude of transcriptome-wide changes in transcript abundance
as a capacitor for growth in response to key environmental stimuli, and provides a balanced
counterpoint to the focus on the role of individual genes.
3.4.7 The extent of differences in drought-responsive transcriptomes between P. balsamifera clones positively correlated with the extent of intraspecific DNA sequence variation
The differences and commonalities in water-deficit-induced transcript abundance patterns may be
attributable to sequence variants from one P. balsamifera genotype to another. One advantage of
Affymetrix GeneChip data of the sort described herein is that probe-level data provide a relatively
simple means by which to assess sequence polymorphisms between pairs of genotypes. Genome-
wide sequence polymorphisms, known as single feature polymorphisms (SFPs) can be identified
using Affymetrix GeneChip data (Luo et al. 2007; Gupta et al. 2008; Fujisawa et al. 2009). When
46
Difference in Plant Height (cm)
Abso
lute
mag
nitu
de c
hang
e in
gene
exp
ress
ion
(Log
2(fol
d-ch
ange
))
0.0
1.0
2.0
3.0
4.0
0 1 2 3 4 5
AP-947AP-1005AP-1006AP-2278AP-2298AP-2300
Genotype
R2 = 0 .776
Figure 3.7 The relationship between the magnitude change in gene expression and the difference
in plant height between well watered and water-deficit treated P. balsamifera trees. Linear regression
analysis revealed a significant relationship between these two variables (P = 0.02033). The
coefficient of determination (R2) is shown in the figure panel.
47pair-wise comparisons of the single-probe level hybridisation data for the six P. balsamifera genotypes
were examined, across all probe sets on the Affymetrix Poplar whole-genome GeneChip, the number
of SFPs found in any given pair-wise comparison between genotypes varied from approximately
3100 to 13 000 (Table 3.2). Genotypes that appeared to be most divergent based on increased SFP
occurrence between the genotypes also showed decreased commonalities in drought-responsive
genes. For example, genotype AP-947 and AP-2300 were divergent with respect to their drought
transcriptomes (Figure 3.5) and also had >10 000 SFPs. Conversely, genotype AP-1006 and AP-
2278 were more closely related with a high degree of similarity for genes significantly expressed in
response to water-deficit, and also had the least number of SFPs between them.
Notably and importantly, transcript abundance differences observed between genotypes were not
attributable to the number of SFPs identified for a given pair-wise comparison. The proportion
of SFPs identified for any given pair-wise comparison was the same for genes with significant
differences in transcript abundance in response to water-deficit in combination with genotype (i.e.
they had a significant genotype-treatment interaction) and for those where genotype played no role
in water-deficit-induced changes in transcript abundance (i.e. determined by treatment main effect
only; Table 3.2). These data are important in that they reveal that the differences in transcriptome
observed between two genotypes were not attributable to differences in hybridization on account
of sequence polymorphism, but rather most of the intraspecific differences in water-deficit-induced
changes to the P. balsamifera transcriptome were likely attributable to non-coding cis-acting
sequences.
Intriguingly, the degree of relatedness between P. balsamifera genotypes, as defined by frequency of
SFPs, did not correspond to the geographical origin of the genotypes. That is, pairs of genotypes
that were acquired nearby had as many pair-wise SFP differences as pairs that were acquired
from two very different locations. This finding has two important implications. First, inasmuch
as the six genotypes reported here were representative of P. balsamifera, SFP-inferred relatedness
does not reflect geographic distance between genotypes. Second, and more importantly, SFP-
inferred relatedness corresponded to transcriptome-level relatedness for the water-deficit-induced
transcriptome. That is, pairs of genotypes with fewer SFPs had more closely related transcriptome
profiles; whereas, pairs with greater numbers of SFPs had more distinct transcriptomes. These
findings suggest that genetic relatedness is likely to be an indicator of a shared water-deficit
48
Number of ProbeSets with 1 SFP
All Probesets (n=61313)
Filtered probesets not significant for Genotype main effect (n=10794)
Filtered probesets significant for Genotype main effect (n=5123)
Genotype 1 Genotype 2AP-947 AP-1005 11663 40 .40 40 .52AP-947 AP-1006 12110 45 .59 45 .81AP-947 AP-2278 13657 46 .90 47 .18AP-947 AP-2298 10943 43 .16 63 .01AP-947 AP-2300 10602 39 .22 39 .72AP-1005 AP-1006 7185 37 .81 38 .16AP-1005 AP-2278 5052 23 .55 24 .54AP-1005 AP-2298 6934 36 .71 37 .81AP-1005 AP-2300 6759 34 .32 34 .86AP-1006 AP-2278 3158 13 .23 13 .65AP-1006 AP-2298 3785 19 .18 20 .11AP-1006 AP-2300 3815 18 .83 19 .13AP-2278 AP-2298 5066 28 .67 29 .22AP-2278 AP-2300 5786 29 .01 29 .77AP-2298 AP-2300 4650 20 .11 20 .03
Table 3.2 Total number of single-feature polymorphisms (SFPs) were identified in all probe sets
on the Affymetrix Poplar GeneChip using SNEP (P < 0.05; Fujisawa et al. 2009). Genes that were
identified as significantly differentially expressed (FDR = 0.05; log2(FC) cutoff = 2.0) and genes
whose expression is not significantly different among genotypes were surveyed for SFPs and the
proportion was calculated based on the total number of probe sets examined, respectively.
49response. Moreover, the findings suggest that while local environments must play a role in the
selection of specific phenotypic responses in P. balsamifera, the responses of some genotypes appear
to be relatively robust across a large geographical distance for this species. Moreover, resent research
suggests most present day populations of P. balsamifera are the result of large-scale range expansions
that occurred since the last glacial maximum (Keller et al. 2010). Three unique sub-populations
have been identified; samples analyzed for this research originate from one sub-population, and
future studies should include sampling across the whole geographic range for P. balsamifera.
However, taken together, this suggests that both local adaptation and phenotypic plasticity might be
underlying factors determining the wide geographical range of P. balsamifera.
3.5 Conclusion
Although there was a common, shared water-deficit induced transcriptome level response for P.
balsamifera, the amplitude of gene expression for the shared water-deficit transcriptome varied
among genotypes. Larger changes in the absolute magnitude of transcript abundance for probe sets
that were significant for treatment main effect were observed for genotypes that had more rapid
declines in their physiological status in response to drought. Phenotypic traits, such as growth, are
correlated with genetic responsiveness to drought. Genotypes that had the capacity to sustain growth
under water-limitation also exhibited increased magnitude change in the remodelled transcriptome.
Genotypes that had greater commonalities in their drought transcriptomes in response to water-
deficit also had fewer SFP differences, suggesting that responses may be conserved across individuals
that share a greater degree of genotypic similarity. Moreover, the lack of correspondence between
pair-wise SFP differences and geographical origin between genotypes suggests that some genotype-
derived responses are locally adapted, while others are spread widely on the landscape. Together,
these findings better define within-species variation in the response of an important genus to a key
environmental challenge, and raise testable hypotheses regarding the mechanisms underpinning the
drought response in poplars, and how these shape the distribution of poplars on the landscape.
3.6 Acknowledgements
We are most grateful to Bruce Hall and Andrew Petrie for excellent greenhouse assistance, John
McCarron for experimental set up, Joan Ouellette for technical assistance, and Dave Kamelchuk
(Al-Pac) for collecting all the plant materials. We are also most grateful for incredibly useful
50comments on the draft manuscript provided by two anonymous reviewers. Research infrastructure
and technical support was generously provided by the Centre for Analysis of Genome Evolution
& Function at University of Toronto. OW was generously supported by a Natural Science and
Engineering Research Council of Canada (NSERC) Canadian Graduate Scholarship (CGSD).
SDM is a Canada Research Chair. This work was generously supported by funding from NSERC,
the Canada Foundation for Innovation (CFI), and the University of Toronto to SDM, ALP and
MMC.
51
3.7 Supplementary Figures
R = 0.267410
2
4
6
8
10
12
14
0 100 200 300 400 500 600
R = 0.140410
2
4
6
8
10
12
14
-3 -2 -1 0 1 2 3
R = 0.137310
2
4
6
8
10
12
14
1000 1100 1200 1300 1400 1500
R = 0.030310.0
0.5
1.0
1.5
2.0
0 100 200 300 400 500 600
R = 0.037230.0
0.5
1.0
1.5
2.0
-3 -2 -1 0 1 2 3
R = 0.160880.0
0.5
1.0
1.5
2.0
1000 1100 1200 1300 1400 1500
R = 0.354260.0
1.0
2.0
3.0
4.0
5.0
0 100 200 300 400 500 600
R = 0.418220.0
1.0
2.0
3.0
4.0
5.0
-3 -2 -1 0 1 2 3
R = 0.230720.0
1.0
2.0
3.0
4.0
5.0
1000 1100 1200 1300 1400 1500
R = 0.23323-0.5
0.0
0.5
1.0
1.5
2.0
2.5
0 100 200 300 400 500 600
R = 0.20315-0.5
0.0
0.5
1.0
1.5
2.0
2.5
-3 -2 -1 0 1 2 3
R = 0.35953-0.5
0.0
0.5
1.0
1.5
2.0
2.5
1000 1100 1200 1300 1400 1500
Mean Annual Precipitation (mm) Mean Annual Temperature (ºC) Degree Days (> 5ºC)
Num
ber o
f day
s to
sig
nific
ant
diffe
renc
e in
gs
Cha
nge
in A
bove
grou
nd
biom
ass
(g D
W)
Cha
nge
in P
lant
heig
ht (c
m)
Cha
nge
in S
tem
ci
rcum
fere
nce
(mm
)
(a)
52
R = 0.149420.0
1.0
2.0
3.0
4.0
0 100 200 300 400 500 600
R = 0.089710.0
1.0
2.0
3.0
4.0
-3 -2 -1 0 1 2 3
R = 0.572370.0
1.0
2.0
3.0
4.0
1000 1100 1200 1300 1400 1500
R = 0.068940.0
1.0
2.0
3.0
4.0
0.0 0.5 1.0 1.5 2.0
R = 0.058510.0
1.0
2.0
3.0
4.0
-0.5 0.0 0.5 1.0 1.5 2.0 2.5
R = 0.405850.0
1.0
2.0
3.0
4.0
0 2 4 6 8 10 12 14
Mean Annual Temperature (ºC) Degree Days (> 5ºC)
Change in Abovegroundbiomass (g DW)
Change in Stem circumference (mm)
Number of days to significantdifference in gs
Mean Annual Precipitation (mm)
Abso
lute
mag
nitu
de c
hang
e in
gene
exp
ress
ion
(Log
2(FC
))Ab
solu
te m
agni
tude
cha
nge
inge
ne e
xpre
ssio
n (L
og2(F
C))
AP-947AP-1005AP-1006AP-2278AP-2298AP-2300
(b)
Supplementary Figure S3.1 (a) Correlation between historic climatic variables and observed
phenotypic traits for the six. balsamifera genotypes. (b) Correlation between absolute magnitude
change in gene expression of probe sets identified as significant for treatment main effect in response
to WD conditions and historic climatic variables, phenotypic traits and physiological response.
53
0 10 20 30
unknown biological processes transport
transcription signal transduction response to stress
response to abiotic or biotic stimulus protein metabolism
other metabolic processes other cellular processes
other biological processes electron transport or energy
DNA or RNA metabolism developmental processes
cell organization and biogenesis
0 10 20 30
unknown biological processes transport
transcription signal transduction response to stress
response to abiotic or biotic stimulus protein metabolism
other metabolic processes other cellular processes
other biological processes electron transport or energy
DNA or RNA metabolism developmental processes
cell organization and biogenesis
Relative proportion of probe sets
(a)
(b)
Supplementary Figure S3.2 Bar graphs representing the functional categories represented by genes
that are differentially expressed for treatment main effect (n = 280, FDR = 0.05, log2(fold-change)
cutoff = 2.0) for (a) increased transcript abundance, and (b) decreased transcript abundance in
response to water-limitation.
54
Supplementary Figure S3.3 Bar graphs representing the functional categories represented by genes
that are significantly differentially expressed between WD and WW conditions. The proportion of
probe sets identified classified for each GO biological process functional category is represented as
the percentage of total genes differentially expressed for treatment main effect increased transcript
abundance and decreased transcript abundance; FDR = 0.05, log2(fold-change) cutoff = 2.0, n =
280), and each individual genotype when analysed individually as a 2 x 2 factorial (FDR = 0.05).
0
10
20
30
40
50
60
unkn
own b
iologic
al pro
cesse
s
trans
port
trans
cripti
on
signa
l tran
sduc
tion
resp
onse
to st
ress
resp
onse
to ab
iotic o
r biot
ic stim
ulus
prote
in meta
bolism
othe
r meta
bolic
proce
sses
othe
r cellu
lar pr
oces
ses
othe
r biolo
gical
proce
sses
elec
tron t
ransp
ort or
energ
y path
ways
DNA or R
NA meta
bolism
deve
lopmen
tal pr
oces
ses
cell o
rganiz
ation
and b
iogen
esis
DNA or R
NA bind
ing
hydro
lase a
ctivity
kinas
e activ
ity
nucle
ic acid
bind
ing
nucle
otide
bind
ing
othe
r bind
ing
othe
r enz
yme a
ctivity
othe
r mole
cular
func
tions
prote
in bin
ding
rece
ptor b
inding
or ac
tivity
struc
tural
molecu
le ac
tivity
trans
cripti
on fa
ctor a
ctivity
trans
feras
e activ
ity
trans
porte
r activ
ity
unkn
own m
olecu
lar fu
nctio
ns
AP947AP1005AP1006AP2278AP2298AP2300
0
10
20
30
40
50
60
unkn
own b
iologic
al pro
cesse
s
trans
port
trans
cripti
on
signa
l tran
sduc
tion
resp
onse
to st
ress
resp
onse
to ab
iotic o
r biot
ic stim
ulus
prote
in meta
bolism
othe
r meta
bolic
proce
sses
othe
r cellu
lar pr
oces
ses
othe
r biolo
gical
proce
sses
elec
tron t
ransp
ort or
energ
y path
ways
DNA or R
NA meta
bolism
deve
lopmen
tal pr
oces
ses
cell o
rganiz
ation
and b
iogen
esis
DNA or R
NA bind
ing
hydro
lase a
ctivity
kinas
e activ
ity
nucle
ic acid
bind
ing
nucle
otide
bind
ing
othe
r bind
ing
othe
r enz
yme a
ctivity
othe
r mole
cular
func
tions
prote
in bin
ding
rece
ptor b
inding
or ac
tivity
struc
tural
molecu
le ac
tivity
trans
cripti
on fa
ctor a
ctivity
trans
feras
e activ
ity
trans
porte
r activ
ity
unkn
own m
olecu
lar fu
nctio
ns
IncreasedTranscript abundance in response to WD
Decreased transcript abundance in response to WD
55
-15
-10
-5
0
5
10
15
-20 -15 -10 -5 0 5 10 15 20
-10.0
-5.0
0.0
5.0
10.0
-10.0 -5.0 0.0 5.0 10.0
-10
-5
0
5
10
-15.00 -10.00 -5.00 0.00 5.00 10.00 15.00
-6
-4
-2
0
2
4
6
-10.00 -5.00 0.00 5.00 10.00
TMMER
FAMASTOMAGEN
Supplementary Figure S3.4 Quantitative reverse transcription PCR validation of transcript
abundance levels of selected genes from microarray data.
56
3.8 Supplementary Tables
Supplementary Table S3.1 The microsatellite loci used to fingerprint the six P.
balsamifera genotypes in this study.
Clone Locus Allele 1 Allele 2AP- 947 PMGC 2501 234 -AP- 1005 PMGC 2501 208 264AP- 1006 PMGC 2501 218 234AP- 2278 PMGC 2501AP- 2298 PMGC 2501 300 307AP- 2300 PMGC 2501 347 -AP- 947 PMGC 2818 169 176AP- 1005 PMGC 2818 300 209AP- 1006 PMGC 2818 351 351AP- 2278 PMGC 2818AP- 2298 PMGC 2818AP- 2300 PMGC 2818AP- 947 PMGC 2328 266AP- 1005 PMGC 2328 300AP- 1006 PMGC 2328 348AP- 2278 PMGC 2328 299AP- 2298 PMGC 2328 299AP- 2300 PMGC 2328 299AP- 947 ORMP 356 164 177AP- 1005 ORMP 356 210 246AP- 1006 ORMP 356 300AP- 2278 ORMP 356 300AP- 2298 ORMP 356 350AP- 2300 ORMP 356 246 246AP- 947 PMGC 93 328 354AP- 1005 PMGC 93 328 354AP- 1006 PMGC 93 350 354AP- 2278 PMGC 93 186AP- 2298 PMGC 93 350 354AP- 2300 PMGC 93 350
(a)
57
(b)
Locus Reference Linkage group in P. trichocar-pa
Repeat Motif
Size Range (bp)
PMGC 2501 http://ornl .gov/sci .ipgc/ssr_resource .htm III (GA) 208-347PMGC 2818 http://ornl .gov/sci .ipgc/ssr_resource .htm NA (GA) 169-351PMGC 2328 http://ornl .gov/sci .ipgc/ssr_resource .htm NA (GA) 266-348PMGC 93 http://ornl .gov/sci .ipgc/ssr_resource .htm I (CTT) 186-350ORMP 356 Tuskan et al . 2004 IV (AT) 164-350
58Supplementary Table S3.2 Relative Water Content (RWC) calculated for each of the six P.
balsamifera genotypes after 15 days of water-deficit treatment.
Relative water content (%)Genotype Well Watered Water Deficit pAP-947 80 .18 73 .18AP-1005 81 .4 71 .79 *AP-1006 88 .85 70 .03 *AP-2278 87 .87 76 .54AP-2298 81 .86 74 .84 *AP-2300 83 .7 75 .61 *
*p<0 .05, Students t-test
59Supplementary Table S3.3 Probe sets with significant main effects or interactions for (a) all
genotypes (FDR = 0.05, log2(fold-change) cutoff = 2.0) (b) all genotypes (FDR = 0.05, no
minimum threshold) and (c) all pair-wise genotype comparisons (FDR = 0.05).
Number of Significant Probe setsDecreased Transcript Abundance
Increased Transcript Abundance
G .947 .1005 600 741G .947 .1006 630 1224G .947 .2278 735 1121G .947 .2298 343 288G .947 .2300 322 402G .1005 .1006 368 533G .1005 .2278 343 370G .1005 .2298 735 563G .1005 .2300 622 785G .1006 .2278 404 248G .1006 .2298 1423 810G .1006 .2300 1264 1144G .2278 .2298 1155 658G .2278 .2300 1027 1044G .2298 .2300 147 294Treatment 182 98Tx .947 .1005 90 52Tx .947 .1006 155 55Tx .947 .2278 42 51Tx .947 .2298 0 6Tx .947 .2300 0 2Tx .1005 .1006 861 225Tx .1005 .2278 511 225Tx .1005 .2298 102 70Tx .1005 .2300 35 36Tx .1006 .2278 761 230Tx .1006 .2298 175 65Tx .1006 .2300 67 46Tx .2278 .2298 42 60Tx .2278 .2300 15 48Tx .2298 .2300 0 5Time Point 245 481
(a)
60
G .Tx .947 .1005 9 52G .Tx .947 .1006 51 386G .Tx .947 .2278 33 50G .Tx .947 .2298 0 0G .Tx .947 .2300 0 0G .Tx .1005 .1006 16 24G .Tx .1005 .2278 13 4G .Tx .1005 .2298 21 4G .Tx .1005 .2300 189 18G .Tx .1006 .2278 13 4G .Tx .1006 .2298 329 28G .Tx .1006 .2300 686 51G .Tx .2278 .2298 41 19G .Tx .2278 .2300 115 41G .Tx .2298 .2300 0 0G .Tp .947 .1005 129 21G .Tp .947 .1006 217 32G .Tp .947 .2278 154 40G .Tp .947 .2298 26 1G .Tp .947 .2300 64 3G .Tp .1005 .1006 2 0G .Tp .1005 .2278 1 0G .Tp .1005 .2298 22 42G .Tp .1005 .2300 29 11G .Tp .1006 .2278 0 1G .Tp .1006 .2298 42 122G .Tp .1006 .2300 52 30G .Tp .2278 .2298 45 52G .Tp .2278 .2300 58 27G .Tp .2298 .2300 5 1
61
Number of Significant Probe sets
AP-947 WD Response 122Time Point 3040WD x Time Point 86
AP-1005 WD Response 1356Time Point 1207WD x Time Point 11
AP-1006 WD Response 1949Time Point 1685WD x Time Point 10
AP-2278 WD Response 1001Time Point 1369WD x Time Point 0
AP-2298 WD Response 36Time Point 985WD x Time Point 5
AP-2300 WD Response 27Time Point 1380WD x Time Point 40
(b)
62
Chapter 4: Drought induces alterations in the stomatal devel-opment program in Populus
Contents of this chapter have been published in the Journal of Experimental Botany: Erin T.
Hamanishi, Barb R. Thomas and Malcolm M. Campbell. 2012. Drought induces alterations in
the stomatal development program in Populus. Journal of Experimental Botany. 63(13): 4959-4971
The published paper and supplementary files can be found online at
http://jxb.oxfordjournals.org/content/63/13/4959.long
The material in this chapter is © Oxford University Press, 2012.
63
Chapter 4: Drought induces alterations in the stomatal devel-opment program in Populus
4.1 Abstract
Much is known about the physiological control of stomatal aperture as a means by which plants
adjust to water availability. By contrast, the role played by the modulation of stomatal development
to limit water loss has received much less attention. The control of stomatal development in
response to water deprivation in the genus Populus is explored here. Drought induced declines in
stomatal conductance as well as an alteration in stomatal development in two genotypes of Populus
balsamifera . Leaves that developed under water-deficit conditions had lower stomatal indices than
leaves that developed under well-watered conditions. Transcript abundance of genes that could
hypothetically underpin drought-responsive changes in stomatal development was examined, in two
genotypes, across six time points, under two conditions, well-watered and with water deficit. Populus
homologues of STOMAGEN, ERECTA (ER), STOMATA DENSITY AND DISTRIBUTION 1
(SDD1), and FAMA had variable transcript abundance patterns congruent with their role in the
modulation of stomatal development in response to drought. Conversely, there was no significant
variation in transcript abundance between genotypes or treatments for the Populus homologues of
YODA (YDA) and TOO MANY MOUTHS (TMM). The findings highlight the role that could be
played by stomatal development during leaf expansion as a longer term means by which to limit
water loss from leaves. Moreover, the results point to the key roles played by the regulation of the
homologues of STOMAGEN, ER, SDD1, and FAMA in the control of this response in poplar.
4.2 Introduction
Water availability is a key determinant of plant growth and survival. In keeping with this, plants
have evolved mechanisms to modulate physiological and developmental processes so as to match
water use and retention with water availability. Stomata, the pores found on plant surfaces, play a
key role in regulating water movement and retention in response to the prevailing environmental
conditions. For example, episodic water deficit can invoke a decrease in stomatal aperture with a
concomitant decrease in water loss from the plant body (Cowan and Farquhar, 1977; Chaves et
al., 2003). Although reduction in stomatal aperture in response to drought limits photosynthesis
and affects water-use efficiency, it is a short-term response that enables plants to contend with
64fluctuating water supply (Chaves et al., 2003). Under drought conditions, guard cell-specific
signal transduction in potato modulated short-term stomatal movements as well as long-term gene
expression (Kopka et al. 1997). Plants can also mount more lasting stomatal-based responses to
persistent water deficit (i.e. drought) by controlling stomatal density during development. Lower
stomatal density restricts the number of sites for water loss, with an attendant decrease in water loss.
Changes in stomatal density are brought about by modulating stomatal development during leaf
formation.
Much is known about stomatal development in Arabidopsis thaliana. Stomatal development
proceeds from the asymmetric division of epidermal meristemoid mother cells to the final terminal
differentiation of the guard cells that will form the stomate early in leaf development. Many of
the components of the regulatory network underlying this terminal differentiation pathway have
been characterized (for a review see Bergmann and Sack (2007), Casson and Hetherington (2010).
Intracellular signalling peptides belonging to the EPIDERMAL PATTERNING FACTOR-LIKE
(EPFL) family, such as, EPF-1 and EPF-2, enforce correct stomatal patterning by acting as negative
regulators of stomatal development (Hara et al., 2007; 2009). By contrast, STOMAGEN acts as
a positive signalling factor in stomatal patterning (Kondo et al., 2010; Sugano et al., 2010). The
positive and negative signalling ligands act antagonistically with cell surface receptors, including
members of the ERECTA family of leucine rich repeat (LRR) -receptor like kinases (ER, ERL-1,
ERL-2) to regulate asymmetric divisions at the onset of stomatal development and spacing divisions
(Nadeau and Sack, 2002; Shpak et al., 2005). TOO MANY MOUTHS (TMM) is another LRR-
receptor-like protein involved in the modulation of stomatal patterning, which acts synergistically
as a signal modulator through interactions with the ER-family of receptors (Lee et al., 2012). In
addition, the subtilisin-like protease, STOMATAL DENSITY AND DISTRIBUTION-1 (SDD-
1) acts independently of EPF-1 and EPF-2, but also negatively regulates asymmetric cell division
(Berger and Altmann, 2000; von Groll et al., 2002). Downstream of the aforementioned receptors,
a mitogen activated protein (MAP) kinase signalling cascade is implicated. Activation of the TMM-
ER family complex leads to the stimulation of the MAP kinase signalling cascade starting with
YODA (YDA), a MAP kinase kinase kinase (Bergmann et al., 2004), which in turn activates MKK4
and MKK5, and finally, MK3 and MK6 (Wang et al., 2007). This signalling cascade negatively
regulates stomatal development through three important basic-helix-loop-helix transcription factors,
SPEECHLESS (SPCH), MUTE, and FAMA (Figure 4.1).
65
Figure 4.1 The stomatal development signalling network, based on current literature. Arrows
represent positive regulation; whereas, blocked lines represent negative regulation. Question marks
represent unknown interactions.
TMM ERECTA-family
STOMAGENEPF1 or 2
YDA
kina
se
Plasma Membrane
Apoplast
MAPK Cascade
Nucleus
SPCHMUTEMYB 88
Stomatal Development
SDD1?
MKK 4/5MPK 3/6
FAMA
66The commitment to stomatal development begins with asymmetric cell division of the meristemoid
mother cell regulated by a basic-helix-loop-helix transcription factor, SPEECHLESS (MacAlister
et al., 2007). Following asymmetric division, cells destined to become guard cells change into a
guard mother cell under the control of MUTE (Pillitteri et al., 2007). Each additional amplifying
asymmetric division results in the creation of a new meristemoid cell and a larger neighbouring
cell. These additional divisions result in the formation of more pavement and stomatal cells. Final
differentiation of the stomatal lineage is controlled by another basic-helix-loop-helix transcription
factor, FAMA (Pillitteri et al., 2007). Furthermore, an additional class of bHLH transcription
factors, SCREAM/ICE1 and SCREAM2 that interact directly with SPCH, MUTE, and FAMA, act
to promote the sequential steps in stomatal differentiation (Kanaoka et al., 2008).
In response to environmental change, plants can modulate stomatal development in new leaves
(Casson and Hetherington, 2010). As mature leaves sense environmental conditions, stomatal
density is adjusted in developing leaves (Lake et al., 2001; Miyazawa et al., 2006). An increase in
light quantity positively influences stomatal numbers through the action of PHYTOCHROME
B (PHYB) and the downstream transcription factor phytochrome-interacting FACTOR 4 (PIF4).
Elevated concentration of carbon dioxide leads to a decline in stomatal density, a phenomenon that
has been observed over geological time (Woodward, 1987). In response to CO2, the gene HIGH
CARBON DIOXIDE (HIC) modulates stomatal development in Arabidopsis (Gray et al., 2000).
Loss-of-function hic mutants exhibit elevated stomatal numbers when grown under elevated CO2
(Gray et al., 2000).
Modification of stomatal density in response to drought varies between plant species, and is
contingent on the severity of water deficit. For example, a drought-induced reduction in stomatal
numbers was observed in wheat (Quarrie and Jones, 1977), squash cotyledons (Sakurai et al., 1986),
and umbu trees (Silva et al., 2009). By contrast, increased stomatal density was observed in grass
with moderate drought stress; although, this increase was reversed under conditions of more severe
drought stress (Xu and Zhou, 2008). Variation in stomatal density was observed in response to
drought in Mediterranean plants (Galmés et al., 2007). No significant alteration to stomatal density
in groundnut was observed under drought (Clifford et al., 1995).
The impact of drought on stomatal density in the ecologically and economically important genus
Populus is examined here. Focusing on Populus balsamifera, the aim was to determine the impact
67of drought on stomatal development during leaf formation by testing the hypothesis that drought-
induced modification of the transcription of genes implicated in the stomatal development
regulatory network are linked to changes in stomatal density. More specifically, we set out to test
the hypothesis that the transcript accumulation of positive regulators of stomatal development will
be lower in the developing foliar tissue of water-deficit-treated trees and, conversely, that transcript
accumulation of negative regulators will be higher in the developing foliar tissue of water-deficit-
treated trees. Making use of a transcriptome database for leaf development, and a time-course
series during leaf formation in the presence and absence of drought, genes involved in the Populus
stomatal development network were identified and a subset shown to show a pattern of transcript
abundance in keeping with a role in modifying stomatal numbers in response to drought.
4.3 Materials and Methods
4.3.1 Plant material
Two Populus balsamifera genotypes (AP-1005 and AP-1006) were propagated from unrooted
cuttings (Alberta Pacific, Boyle, Alberta, Canada) in Sunshine mix-1 (Sun Gro Horticulture
Inc, Bellevue, WA, USA). The cuttings used in this experiment were obtained from the research
stoolbeds at the Alberta-Pacific mill site (Alberta, Canada); however, genotype AP-1005 historically
originates from Slave Lake, Alberta, Canada whereas, genotype AP-1006 originates from Smith,
Alberta, Canada. A more detailed description of the two P. balsamifera genotypes can be found in
Hamanishi et al. (2010; Chapter 3, this volume). Trees were grown in a climate-controlled growth
chamber at the University of Toronto (Toronto, Ontario, Canada) with conditions described by
Hamanishi et al. (2010; Chapter 3, this volume). After nine weeks of growth under well-watered
conditions, half of the trees were placed under water-deficit conditions by withholding water, while
temperature and light conditions remained constant. At the onset of the water-deficit experiment,
the first fully expanded leaf on day 0 of the experiment was marked with a red thread, and the
position of the first expanding leaf relative to the first fully expanded leaf on day 0 was recorded.
Fully expanded P. balsamifera leaves were at leaf plastochron index (LPI) 7–8 (Larson and Isebrands,
1971); whereas the developing leaves on day 5 were often at LPI = 2 and at LPI = 4–5 on day 15
after the onset of the water-deficit experiment.
Plant material was harvested at day 0, and every 5 d thereafter until the completion of the 30
68d experiment (days 0, 5, 10, 15, 20, 25, and 30; see Supplementary Figure S4.1). Using three
replicates from each treatment–genotype combination, at the harvesting time-point, the first fully
expanded leaf marked on day 0 from two trees was collected, pooled, and flash-frozen using liquid
nitrogen. This represented a single sample from a single genotype, treatment, and time-point
combination. Similarly, the first expanding leaf from day 0 was collected from two trees, pooled,
and flash-frozen for future analysis. For each sample collection, only two leaves were removed from
each tree: the first fully expanded leaf, and the first expanding leaf from day 0. Once leaves were
sampled from a given tree, the tree was no longer included in the experiment.
4.3.2 Physiological measurements and stomatal quantification
For each genotype, physiological responses to drought conditions were monitored every 2 d starting
from the onset of the water-withholding experiment. Stomatal conductance (gs) measurements
were taken using an infrared gas analyser (LI-6400XT Portable Photosynthesis System, Li-Cor
Biosciences Inc., Lincoln, NE, USA). Measurements of gs were taken on the mature, fully expanded
leaves at the experimental midday time point (n = 3–5 per genotype–treatment group). Temperature
and relative humidity were maintained at 21.3 ± 0.6°C and 62.6 ± 2.17%, respectively, for gas
exchange measurements. Productivity and relative water content (RWC) was assessed periodically
throughout the 30 d experiment. Height and stem diameter were recorded 5, 10, 20, and 30 d after
the onset of the water-withholding experiment. Above-ground biomass was determined at the end
of the experiment (day 30); plants (n=10) were harvested and above-ground biomass (fresh weight
and dry weight) was measured. Leaf RWC was determined using methods described by Hamanishi
et al. (Hamanishi et al., 2010) 15 and 30 d after the onset of the water-withholding experiment.
Impressions of the abaxial epidermis were taken 30 d after the onset of the water withholding
experiment for two classes of leaves. The two classes of leaves included (a) leaves that were fully
developed prior to the onset of the experiment (the first fully expanded leaves at day 0) and (b)
leaves that expanded during the water-withholding experiment (leaves that were marked as the
first emerging leaf at day 0). Impressions of 10 leaves from each class, genotype, and treatment
combination were assessed. Abaxial impressions were taken using clear nail polish and cellophane
tape, as described by Ceulemans et al. (1995) at the widest point of the leaf (approximately 3cm
wide). A minimum of 5 microscopic fields were randomly selected per sample leaf impression, and
the epidermal cell density and stomatal density were calculated. Stomatal index was defined as:
69Stomatal Index = (s / (s+e)) x 100 (Equation 4.1)
where s is the number of stomata and e is the number of epidermal cells per unit area (Radoglou
and Jarvis, 1990).
4.3.3 Gene selection
Putative homologues of genes known to be involved in stomatal development were identified
from Populus using P. trichocarpa sequence data available on phytozome v7.0 [http://phytozome.
net; Tuskan et al. (2006)]. The protein sequences fromArabidopsis were used as a query for BLAST
(BLASTp/PtPEPv2.0) searches against the databases. FAMA (MacAlister and Bergmann, 2011)
and STOMAGEN (Kondo et al., 2010) orthologues in Populus have previously been reported.
Poplar GeneChip (Affymetrix, Santa Clara, CA, USA) probe sets were identified using the NetAffx
resource (http://www.affymetrix.com/analysis/index.affx). Transcript abundance for homologues of
genes implicated in stomatal development was assessed through interrogation of Populus balsamifera
transcript abundance data available in the PopGenExpress compendium of the Bio-Array Resource
(BAR; http://bar.utoronto.ca/; Wilkins et al. 2009a).
4.3.4 Targeted transcript abundance analysis
Three samples were collected from each genotype (AP-1005 and AP-1006), treatment (well-
watered and water-deficit-treated) and developmental stage (first fully expanded leaf from day
0). Flash-frozen plant material collected throughout the experiment (days 5, 10, 15, 20, 25, and
30) was ground to a fine powder under liquid nitrogen. Starting with 1–2g frozen ground leaf
tissue per sample, total RNA was isolated according to Chang et al. (1993). RNA quality was
assessed electrophoretically and spectrophotometrically. 3 µg total RNA was reverse-transcribed
using oligio(dT)18 primers and SuperScript II Reverse Transcriptase (Invitrogen) according to
manufacturer’s instructions. cDNA was diluted 4-fold prior to qRT-PCR analysis. Using iQ SYBR
Green Supermix (Bio-Rad), qRT-PCR was performed for three biological replicates in triplicate
according to published protocols (Raj et al., 2011). Relative transcript abundance was determined
using the method described by Pfaffl (2001). Primer sequences can be found in Supplementary
Table S4.1.
70
4.3.5 Statistical analysis
Significant variation in relative transcript abundance was analysed using a general linear model. The
general linear model for the 2×2×6 factorial experiment (2 genotypes, 2 treatments, and 6 time-
points) is represented by:
y ijk= u + A i + B j + C k + (AB)ij + (AC)ik + (BC)jk + (ABC)ijk + εijk (Equation 4.2)
where A corresponds to genotype with i levels, B corresponds to treatment with j levels, and C
corresponds to time-point with k levels. Four possible interactions between genotype and treatment
are represented by ij, 12 interactions between genotype and time-point are represented by ik, 12
interactions between treatment and time-point are represented by jk, 24 three-way interactions
are represented by ijk, and the random error is εijk. The α-level was set to 0.05 for all analyses.
Analysis of variance (ANOVA) was determined for all relative transcript abundance profiles for each
gene. All analyses were performed using R (R Development Core Team, 2011).
Correlation between relative transcript abundance profiles was calculated using Pearson correlation
coefficient analysis in R (R Development Core Team, 2009). Pair-wise analysis of transcript profiles
across all samples was compared for each given transcript.
4.4 Results
4.4.1 Stomatal conductance (gs) and relative water content (RWC) in response to
water-deficit stress
In the two Populus balsamifera genotypes, AP-1005 and AP-1006, gs was significantly lower after
30 d of water withdrawal (t test; P <0.05; Figure 4.2 a, b). Genotype AP-1006 experienced the
largest gs decline in water-deficit-treated plants relative to well-watered samples by day 30 (88%
decrease; Figure 4.2b). A gs decline in water-deficit-treated samples was observed throughout the
drought period for both genotypes; however, AP-1006 exhibited the earliest significant differences
between water-deficit and well-watered samples. A significant difference in gs was observed between
water-deficit and well-watered samples as early as 7 d after water-withdrawal in AP-1006; whereas,
AP-1005 did not show any significant differences in gs until 15 d after the onset of the drought
experiment (P <0.05, Figure 4.2a, b). In AP-1005 and AP-1006, RWC was significantly lower in
71
Figure 4.2 Variation in the physiological response to drought stress in genotype AP-1005 and AP-
1006. Box plot of the variation in midday stomatal conductance for (a) AP-1005 and (b) AP-1006
for well-watered (blue boxes) and water-deficit-treated (orange boxes) samples. Response of intrinsic
water use efficiency (WUEi; A/gs) across well-watered and water-deficit-treated samples for (c) AP-
1005 and (d) AP-1006 and photosynthesis (A) for (e) AP-1005 and (f ) AP-1006 at days 0, 5, and
15 after the onset of water withdrawal. Error bars represent the standard error of the mean.
0.0
0.1
0.2
0.3
0.4
0.5 AP-1005
Con
duct
ance
(mol
H2O
m-2 s
-1)
Days0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
(a)
AP-1006(b)
0.0
0.1
0.2
0.3
0.4
0.5
0 5 15
WUE i
(µm
ol C
O2 m
mol
−1 H
2O)
WetDrought
(c) AP-1005 (d) AP-1006
0 5 15DaysDays
050
100
150
050
100
150
02
46
810
1412
02
46
810
1412
0 5 15Days
0 5 15Days
A (µ
mol
CO
2 m-2
s-1) (e) AP-1005 (f) AP-1006
72water-deficit-treated plants after 30 d of water withdrawal when compared with well-watered trees
(see Supplementary Table S4.2). Similar to the declines observed in gs, genotype AP-1006 had a
more severe reduction in RWC when compared with AP-1005. Overall, there was an increase in the
intrinsic water use efficiency [WUE = A/gs; (Seibt et al., 2008)] in drought-treated samples on day 15
(Figure 4.2c, d), where a larger increase in intrinsic WUE was observed in genotype AP-1006.
4.4.2 Stomatal quantification following water-deficit stress
Abaxial leaf stomatal density and index was lower in leaves that developed under water-deficit
conditions when compared with leaves that developed under well-watered conditions for both
genotypes (Figure 4.3). Leaves that were fully developed prior to the onset of the drought
experiment had no significant variation between treatments in their stomatal indices. Under well-
watered conditions, AP-1006 had the highest stomatal index; however, the leaves of genotype
AP-1006 that developed under water-deficit conditions had the lowest stomatal indices. The
significant reductions in stomatal index were observed for AP-1005 and AP-1006 at 12% and 25%
respectively. Notably, significant variation between treatment and genotype were observed with
respect to stomatal index (P <0.05). No significant difference in photosynthesis (A) was observed at
days 0, 5 or 15 after onset of water withdrawal between well-watered and water-deficit-treated plants
for genotype AP-1005 or AP-1006. Photosynthetic rates throughout the experimental period for
genotype AP-1005 and AP-1006 were significantly lower than previously observed in field grown
P. balsamifera (Silim et al. 2010). The lower photosynthetic rates observed in the chamber-grown
seedlings may be attributable to the lower light levels in the growth-chamber.
4.4.3 Populus homologues of genes implicated in stomatal development
Stomatal development is a function of the integration of many different endogenous and exogenous
signals. Many of the genes involved in the underlying pathways regulating stomatal development
in Arabidopsis have been identified [for a review see Bergmann and Sack (2007)]. Homologues
of genes underlying stomatal development in Arabidopsis thaliana are found in Populus. The
PopGenExpress transcript abundance compendium (Wilkins et al., 2009a) shows that many of these
homologues have relatively greater transcript abundance in young leaves compared with other plant
organs, consistent with their role in modulating stomatal development (Figure 4.4).
73
AP-1005 AP-1006
Wel
l-Wat
ered
Wat
er-D
efic
it
WetDry
**
0
5
10
15
20
25
Stom
atal
Inde
x
Leaf A Leaf B
Wet Dry Wet Dry
Leaf A Leaf B
Wet Dry Wet Dry
(a) (b)
(c) (d)
(e) (f)
Stom
atal
Den
sity
(no.
mm
-2)
Leaf A Leaf B
Wet Dry Wet Dry
Leaf A Leaf B
Wet Dry Wet Dry
WetDry
0
100
200
300
400
500
600
700 (g) (h)
74Figure 4.3 Variation in leaf epidermis between genotype AP-1005 and AP-1006 under (a, b)
well-watered and (c, d) water-deficit conditions, on day 30. White scale bar=50 µm. (e, f ) Box
plot of variation in stomatal index for well-watered (blue box) and water-deficit-treated (orange
box) samples for leaves that were fully developed prior to the onset of the drought experiment (leaf
A) and for those that developed during the drought experiment (leaf B). A significant reduction
in stomatal index is observed in leaves that developed during the experimental period (leaf B) for
each genotype (e) AP-1005 and (f ) AP-1006; however, no significant variation in stomatal index
is observed for leaves that developed prior to the experiment (leaf A), and the onset of water-
deficit conditions. The midline of the box represents the median value for stomatal index (e-f ) or
stomatal density (g-h), the upper and lower bounds of the box represent the interquartile range,
and the whiskers extend to the most extreme values that are not outliers. No signficant change in
stomatal density in response to water-deficit conditions for genotype (g) AP-1005 and (h) AP-1006.
Asterisks represent significant variation between well-watered and water-deficit-treated plants. *P
<0.05
75
Figure 4.4 Heat map of transcript abundance across a range of tissues for Populus homologues of
genes implicated in stomatal differentiation and patterning. Transcript accumulation for the 14
Populus homologues that had differential transcript abundance across the dataset, was derived from
the PopGenExpress microarray compendium made available via http://bar.utoronto.ca (Wilkins
et al., 2009a). As per the scale provided, elevated transcript abundance is represented by red and
diminished transcript abundance is represented by green. The highest levels of transcript abundance
for this group of genes are in young leaves in contrast to other tissue types. Each column represents
a discrete biological sample, and data are represented as biological triplicate replicates for each tissue
type. Data are row normalized.
Figure 5
−4 0 4Row Z−Score
Color Key
Young Leaf
MatureLeaf
Roots Xylem FemaleCatkin
MaleCatkin
SPCHMUTEFAMASCRMCDKB1PIF4YDAEPF1SDD1ERL1TMMCHALERSTOMAGEN
76
Figure 4.5 Variation in transcript abundance between well-watered and water-deficit-treated trees
at six time points (days 5, 10, 15, 20, 25, and 30) for genotype AP-1005 (yellow) and AP-1006
(green) represented by the log2(fold change transcript abundance) for genes involved in stomatal
development. A positive log2(fold change transcript abundance) value is an indicator of higher
transcript abundance in water-deficit-treated samples, whereas a negative log2(fold change transcript
abundance) value is an indicator of lower transcript abundance in water-deficit-treated samples.
Figure 5
5 10 15 20 25 30
Days of water withdrawal
Log 2(f
old
chan
ge tr
ansc
ript a
bund
ance
)
5 10 15 20 25 30
Days of water withdrawal
Log 2(f
old
chan
ge tr
ansc
ript a
bund
ance
) AP−1005AP−1006
5 10 15 20 25 30
Days of water withdrawal
Log 2(f
old
chan
ge tr
ansc
ript a
bund
ance
)
−3−1
12
3
5 10 15 20 25 30
Days of water withdrawal
Log 2(f
old
chan
ge tr
ansc
ript a
bund
ance
)
5 10 15 20 25 30
Days of water withdrawal
Log 2(f
old
chan
ge tr
ansc
ript a
bund
ance
)
5 10 15 20 25 30
Days of water withdrawal
Fold
cha
nge
trans
crip
t abu
ndan
ce
−20
−3−1
12
3−2
0
−3−1
12
3−2
0−3
−11
23
−20
−3−1
12
3−2
0
−3−1
12
3−2
0(a) ERECTA (b) FAMA
(c) SDD1 (d) STOMAGEN
(d) TMM (e) YDA
Dry
Wet
Dry
Wet
Dry
Wet
Dry
Wet
Dry
Wet
Dry
Wet
77
4.4.4 Developmental variation in transcript abundance of stomatal development genes after water deficit
The transcript abundance profiles of six genes with putative roles modulating stomatal development
in Populus were examined through development under well-watered and water-deficit conditions
using qPCR. Based on previous studies (Bergmann and Sack, 2007; Casson and Hetherington,
2010), the genes selected are believed to play roles in the development pathway, ranging from
receptors in the signalling cascade to transcription factors regulating the final differentiation step
in the stomatal lineage. The fold change variation in transcript abundance between well-watered
and water-deficit-treated samples revealed variation between time-points, treatments or genotypes
(Figure 4.5). Total cumulative transcript abundance after 30 d of water-withdrawal varied
considerably among genotypes for all genes analysed in this experiment (Table 4.1). A factorial
ANOVA analysis revealed significant variation among days for all genes analyzed, with peaks in
transcript abundance observed early in the experiment, corresponding to earlier in leaf development
(see Supplementary Table S4.3).
TMM (see Supplementary Figure S4.2c ) and YDA (see Supplementary Figure S4.3c) only had
significant differential transcript abundance among days (ANOVA; P <0.05). The highest transcript
levels of TMM were observed on day 10 in the experimental period; whereas, YDA had peak
transcript accumulation on day 5, with no significant difference in transcript accumulation among
the first three time points (see Supplementary Figure S4.3c). Both TMM and YDA exhibited
gradual declines in transcript abundance after the peak in transcript abundance was observed,
with mean transcript levels declining below a 1-log2(fold-change) by the end of the experimental
timeframe (see Supplementary Figures S4.2 and S4.3).
The transcript abundance profiles for ERECTA (ER), FAMA, STOMAGEN, and STOMATAL
DENSITY AND DISTRIBUTION 1 (SDD1) exhibited significant genotype × day interactions
(ANOVA; P <0.05). Transcript abundance profiles of ER for both genotypes had many differences
(see Supplementary Figure S4.4). The highest levels of ER transcript abundance for well-watered
samples in AP-1005 were observed later in the experimental period (day 15 and day 20) when
compared with water-deficit-treated samples; however, the shift in transcript abundance between
treatments was not as evident in AP-1006. For ER (see Supplementary Figure S4.4c) and SDD1
(see Supplementary Figure S4.5c), a more rapid and severe decline in transcript abundance was
78Table 4.1 Mean cumulative transcript abundance across all time-points for genotype AP-1005 and
AP-1006 in well-watered and water-deficit-treated samples.
1005 1006Wet Dry Wet Dry
ER 13 .48 ± 1 .66 10 .21 ± 1 .07 6 .46 ± 1 .18 8 .90 ± 1 .64FAMA 8 .04 ± 0 .81 5 .81 ± 0 .27 10 .82 ± 1 .67 6 .55 ± 0 .43SDD1 7 .67 ± 0 .29 8 .34 ± 0 .19 8 .39 ± 1 .61 6 .57 ± 0 .65STOMAGEN 11 .84 ± 1 .94 9 .55 ± 1 .03 14 .78 ± 2 .25 9 .03 ± 1 .57TMM 7 .67 ± 0 .58 5 .22 ± 0 .34 8 .21 ± 0 .67 8 .91 ± 0 .31YDA 7 .47 ± 0 .38 7 .12 ± 0 .34 10 .39 ± 1 .99 7 .04 ± 0 .53
79observed in AP-1006; however, the rapid decline in transcript abundance was not as evident in AP-
1005.
Significant variation in ER and SDD1 transcript abundance was observed between genotypes
(ANOVA; P <0.05). An overall reduction in transcript abundance for ER and SDD1 was observed
in AP-1006 (see Supplementary Figures S4.5d and S4.6d). ER had a 55% reduction in total
transcript abundance in AP-1006 relative to AP-1005. The reduction observed in genotype AP-
1006 for SDD1 was not as severe.
FAMA and STOMAGEN also had significant treatment × day interactions (ANOVA; P <0.05).
Genotypic influences on transcript abundance variation were less evident for FAMA and
STOMAGEN. Significantly higher transcript abundance was found on days 5 and 10 for FAMA
in well-watered samples (see Supplementary Figure S4.6). Although there was reduced variation
in FAMA transcript abundance in the later part of the experiment (days 15 through 30), FAMA
had a significant treatment main effect. Like FAMA, STOMAGEN had significantly higher
transcript abundance in well-watered samples on day 10 compared with the low variation observed
between treatments for all other days examined (see Supplementary Figure S4.7c). Peak transcript
abundance of STOMAGEN on day 10 is observed for both genotypes; however, the highest
STOMAGEN transcript abundance for water-deficit-treated samples was day 10 for AP-1005 and
day 15 for AP-1006 (see Supplementary Figure S4.7). Significant variation between treatments was
also observed for STOMAGEN ( ; P <0.1; see Supplementary Figure S4.4).
4.4.5 Genes acting as positive regulators in stomatal development have
correlated transcript profiles
The transcript abundance profiles of FAMA and STOMAGEN are more highly correlated (r=0.62;
Figure 4.6) than the negative regulators of stomatal development. The homologuous FAMA and
STOMAGEN show inverse correlation with other negative regulators of stomatal development
analysed in this experiment.
4.5 Discussion
Optimization of carbon uptake and water loss, regulated through the modulation of stomatal
function and development, is critical for plant survival against a background of fluctuating
80
Figure 4.6 Pearson correlation coefficient (PCC) heat map representing the transcript abundance
profiles across AP-1005 and AP-1006 and six time-points. The PCC was determined for each pair-
wise comparison (gene–gene), and is represented by the colour in the corresponding cell. All genes
are represented in the same order on the x- and y-axes.
Figure 6
YDA
EREC
TA
TMM
SDD1
STOMAG
EN
FAMA
YDA
ERECTA
TMM
SDD1
STOMAGEN
FAMA
0 0.5 1
Color Key
Correlation coefficient0.5
1.00
1.00
1.00
1.00
1.00
1.00
0.62
0.440.35
-0.02 0.620.21
0.02 -0.16-0.24
-0.31 -0.15-0.49
0.01
0.42 -0.32
81environmental conditions. Environmental cues such as light and CO2 concentrations have been
shown to modulate stomatal development in Arabidopsis thaliana (Casson and Hetherington, 2010).
There is relatively scant information about the role that water availability plays in the control of
stomatal development in terrestrial plants (Casson and Hetherington, 2010). The modulation
of stomatal development in response to water deficit in Populus was explored here and the role of
candidate genes in the regulation of this process was examined.
4.5.1 Drought response varied between Populus balsamifera genotypes over time
An increase in intrinsic WUE was observed in water-deficit-treated trees, is attributable to the
decline in stomatal conductance observed in the water-deficit-treated plants (Figure 4.2). No
significant difference in photosynthesis was observed between well-watered and water-deficit-
treated plants at day 0, 5 or 15 after the onset of water withdrawal (Figure 4.2 e, f ). After 30 d of
water-deficit, AP-1006 had the most severe reduction in stomatal conductance, yet it also had the
largest reduction in RWC. Variation between poplar genotypes in their physiological response to
drought stress is often observed, and hypothesized to be a result of various strategies to contend with
drought-stress (Marron et al., 2002; Zhang et al., 2004). The changes in physiological status of the
trees in response to water-deficit stress, including an increase in intrinsic WUE, and the decline in
RWC and stomatal conductance in both genotypes after the imposition of water-deficit conditions
may have been responsible for drought-induced modifications to leaf development.
In response to changes in water availability, leaf morphology can vary considerably (Pena-Rojas
et al., 2005). Specifically, with respect to stomatal numbers, modification to environmental
factors that influence mature leaf gs will have lasting effects on the stomatal differentiation of new
leaves (Miyazawa et al., 2006). Two P. balsamifera genotypes, genotype AP-1005 and AP-1006,
had a reduction in stomatal index in response to drought stress, indicating a potential impact on
leaf development by a reduction in water availability. A greater reduction in stomatal index was
observed in AP-1006 (Figure 4.3). The anatomical variation observed between these two genotypes
with respect to stomatal numbers supports the greater reduction in gs observed in AP-1006 at the
end of the experiment (Figure 4.2). The greater reduction in stomatal index in AP-1006 may
be a result of the reduction in water availability and its influence on whole plant water status. A
reduction in stomatal index would result in a long-term strategy to regulate water loss during
82prolonged periods of drought stress, which would ultimately be reflected in larger declines in gs
between well-watered and water-deficit-treated samples. As integral regulators of carbon uptake and
plant water relations, modulation of stomatal differentiation will influence long-term plant WUE
and productivity (Casson and Hetherington, 2010).
4.5.2 Transcript abundance of the Populus homologues of key stomatal
development regulatory genes varied through leaf development in a manner consistent with their proposed molecular functions
Transcript abundance for the Populus TMM orthologue was highly variable between days (Figure
4.5d; see Supplementary Figure S4.2). TMM, a LRR receptor like protein, functions primarily in
the modulation of stomatal number and regulation of stomatal patterning in A. thaliana (Nadeau
and Sack, 2002). In A. thaliana, a single loss of function mutation in TMM results in an excess
of stomata arranged in clusters (Nadeau and Sack, 2002). Although limited variation in TMM
transcript abundance was observed in P. balsamifera, peak transcript abundance was observed on day
10 of the experimental period, early in leaf development (see Supplementary Figure S4.2c). In A.
thaliana, the transcript abundance of TMM is highest in the early stages of the stomatal lineage, in
the young meristemoid mother cell; whereas transcripts were absent in mature guard cells (Nadeau
and Sack, 2002). Transcript abundance for the P. balsamifera TMM homologue peaked at the early
stages of leaf development, and may be congruent with its role previously described in A. thaliana.
Like TMM, there was significant variation in Populus YDA transcript abundance throughout
leaf development; however, peak transcript abundance for P. balsamifera YDA occurred on day
5, presaging the TMM peak (see Supplementary Figure S4.3). YDA encodes a mitogen-activated
protein (MAP) kinase signalling cascade that is involved in the regulation of stomatal differentiation
downstream of the TMM–ERF receptors. In A. thaliana, loss-of-function mutations in the YDA-
encoded MAP kinase kinase kinase result in excess stomatal proliferation (Bergmann et al., 2004).
Although transcript abundance for the Populus YDA homologue did not have a significant treatment
or genotype effect in the experiments described here, it had the highest transcript abundance early in
the experiment, and development, congruent with its early role in stomatal development.
4.5.3 Elevated Populus ERECTA (ER) transcript abundance early in development
83
corresponded with decreased stomatal indices
Transcript abundance analysis revealed significant genotype, day, treatment × day and genotype
× day effects for a Populus ER homologue (see Supplementary Figure S4.4) over the course of the
experiment. In A. thaliana, ER appears to regulate the initial decision of cells to enter the stomatal
lineage and is important for correct stomatal differentiation. Consistent with this, a single loss-
of-function mutation, er, in A. thaliana results in a higher number of stomatal-lineage ground
cells as well as guard cells (Shpak et al., 2005). In the experiment described here, higher transcript
abundance in water-deficit-treated samples, relative to well-watered samples, was observed 5 d
after the onset of the drought experiment for both poplar genotypes (Figure 4.5a). In both poplar
genotypes, a negative relationship between the stomatal indices of leaves that developed under
water-deficit stress and ER transcript abundance (see Supplementary Figure S4.8a). Declines in
stomatal numbers have been observed in response to increased ER transcript abundance in A.
thaliana (Masle et al., 2005). Similarly, over-expression in Arabidopsis of a Populus ER orthologue
(PdERECTA) conferred decreased stomatal numbers. Thus, transcript abundance patterns early in
the experiments are consistent with ER playing a role in the suppression of stomatal numbers in
Populus under water withdrawal conditions.
Although declines in stomatal index were observed in samples with high ER transcript abundance
on day 5, this pattern was not consistent throughout the developmental period. Increased
variability in transcript abundance patterns between well-watered and water-deficit samples
were observed after day 10 for both genotypes. The variation observed for the day × genotype
interaction (see Supplementary Figure S4.4c) highlights the variation observed between AP-1005
and AP-1006. A gradual decline in ER transcript abundance was observed in AP-1006, earlier in
development; whereas a decline in transcript abundance was not observed until after day 15 in AP-
1005. This could be a result of genotypic plasticity or the influence of the redundancy of ER and
functional paralogues, ERECTA-LIKE 1 (ERL1) and ERECTA-LIKE 2 (ERL2), that are known to
act together in the negative regulation of stomatal development (Shpak et al., 2005). However, the
significant decline in overall ER transcript abundance as determined by the genotypic variation (see
Supplementary Figure S4.4c, d) may indicate a more fundamental role of ER in plant development.
ER plays an important role in regulating leaf and whole plant development, and is not restricted to
its involvement in stomatal development (Tisne et al., 2011). Elevated ER transcript abundance
was observed in AP-1005 that demonstrated significantly more height growth than AP-1006 (Table
84Table 4.2 Mean plant height (in cm) on day 30 ±standard error of the mean, n ≥6
AP-1005 AP-1006 P-valueWell-watered 78 .81 ± 2 .73 67 .70 ± 3 .03 0 .0083 *Water-deficit 69 .00 ± 3 .81 63 .06 ± 2 .98 0 .1811
854.2).
4.5.4 STOMATAL DENSITY AND DISTRIBUTION 1 (SDD1) and genotype-specific
control of stomatal development
Loss of SDD1 function in A. thaliana leads to significant increases in stomatal density. SDD1 is a
subtilisin-like Ser protease that is predominantly expressed in stomatal precursor cells, and activates
ER and TMM to repress stomatal development (von Groll et al., 2002). Despite its prominent
role in A. thaliana stomatal development, no significant variation with respect to treatment was
observed in the transcript abundance pattern of the Populus SDD1 homologue over the course of
the experiment in either genotype (see Supplementary Figure S4.5); however, significant variation
between genotypes was observed. SDD1 transcript abundance in AP-1005 was significantly higher
than in AP-1006 (see Supplementary Figure S4.5c). The reduction in transcript abundance in
AP-1006 may reflect an alteration in signalling mechanisms to the underlying signalling cascade
regulating stomatal development in this genotype.
4.5.5 Stomatal development and the regulation of Populus STOMAGEN and
FAMA transcript abundance in response to water deficit
STOMAGEN encodes a peptide that positively regulates stomatal density. The STOMAGEN
peptide is thought to act through antagonistic competition with other peptide signalling molecules,
EPIDERMAL PATTERNING FACTOR 1 and 2 (EPF1 and EPF2), through the LRR-receptor
like protein, TMM (Kondo et al., 2010; Sugano et al., 2010). In A. thaliana, STOMAGEN, is
derived from the mesophyll-tissue, unlike EPF1 and EPF2 that are primarily expressed in the leaf
epidermis, specifically the early meristemoid cells, guard mother cells and guard cells (Kondo et
al., 2010; Sugano et al., 2010). In A. thaliana, over-expression of STOMAGEN increased stomatal
density; whereas loss of STOMAGEN function decreased stomatal density (Kondo et al., 2010;
Sugano et al., 2010).
In the P. balsamifera drought experiment described here, there was significant variation in
STOMAGEN transcript abundance among days and between treatments (see Supplementary Figure
S4.6 and Table S4.4). ANOVA revealed significant day × treatment and day × genotype interactions
for STOMAGEN transcript abundance, suggesting a role for changes in STOMAGEN transcript
86abundance in genotype- and treatment-dependent differences in the regulation of stomatal
development. The most severe log2(fold-change) reduction in STOMAGEN transcript abundance
between well-watered and water-deficit-treated samples were observed in AP-1006 on day 10
(Figure 5d). Notably, the genotype with the lowest stomatal index in the water-deficit-treated
samples was AP-1006. A positive relationship between STOMAGEN transcript abundance on day
10 and stomatal index was observed (see Supplementary Figure S4.8b), consistent with the Populus
homologue of STOMAGEN playing a role in the control of stomatal density.
In Populus hybrids, the stomatal index of new leaves is highly correlated with stomatal conductance
and the physiological status of fully developed leaves suggesting that a long-distance signalling
mechanism is used to regulate stomatal development (Miyazawa et al., 2006). STOMAGEN is
expressed in the mesophyll tissue of immature Arabidopsis leaves (Sugano et al., 2010); however,
stomata are derived from cells on the epidermal layer of leaves. STOMAGEN may play a role in this
long-distance signalling mechanism by acting as a signalling molecule between the mesophyll and
the epidermal layer in leaves. In the Populus drought experiment, reduced stomatal conductance
was observed in response to water-deficit conditions and, similarly, plants exposed to water-deficit
conditions had reduced transcript accumulation of the STOMAGEN homologue together with
reduced stomatal indices. It may be that, in response to water-deficit stress in Populus, STOMAGEN
optimizes long-term carbon uptake and water loss through its role as a mesophyll-derived signalling
factor modulating stomatal development.
Similar to STOMAGEN, FAMA transcript accumulation was highest in the early stages of
development, with peak transcript abundance at day 5 and 10 (see Supplementary Figure S4.6).
In A. thaliana, FAMA is required for the final stages of stomatal differentiation, exhibiting the
greatest transcript abundance in differentiating guard cells, with declining transcript accumulation
as guard cells mature (Ohashi-Ito and Bergmann, 2006). In the Populus drought experiment,
significant variation in FAMA transcript accumulation was observed between days and treatment
(see Supplementary Figure S4.6). The lower levels of transcript accumulation in water-deficit-
treated samples on days 5 and 10 for both genotypes (Figure 5b) and the lower stomatal numbers
observed in water-deficit-treated P. balsamifera samples (see Supplementary Figure S4.8a, b) are
consistent with the role described for FAMA in A. thaliana. Variation among water-deficit-treated
samples throughout the experimental period is considerably less than in the well-watered samples.
87As a transcription factor that is both required and sufficient for the final stages of differentiation,
a minimum accumulation of FAMA transcript may be required for correct stomatal development.
Regardless, the elevated FAMA transcript abundance suggests a role for modulation of this gene to
control stomatal development in Populus under drought conditions.
Despite our knowledge of other genes that influence stomatal development, it is not yet known how
their expression is influenced by water-deficit stress or how they may influence the stomatal index
in Populus. Further exploration of these known players in the stomatal development pathway may
provide an increased insight into the long-term modulation of stomatal development, including
genotypic variation, in response to water-deficit stress. Although it is evident that some players
in the basal stomatal development pathway show altered transcript abundance under water-deficit
conditions, an important question to consider is how water-deficit cues are perceived and integrated
into the stomatal developmental pathway. Such foci will undoubtedly provide fertile grounds for
future research.
4.6 Conclusion
In response to water-deficit stress, P. balsamifera demonstrated significant declines in stomatal
conductance and RWC. Intraspecific variation in physiological responses between two P. balsamifera
genotypes was observed. The largest declines in physiological status were observed in AP-1006 in
response to the imposition of water-deficit conditions. Reductions in stomatal indices were also
observed in both genotypes; however, declines in stomatal index in AP-1006 were markedly larger
than AP-1005. Quantification of transcript abundance profiles of a subset of genes involved in
stomatal development under well-watered and water-deficit-treated conditions revealed variation
between genotypes, as well as between treatments. Notably, STOMAGEN, a mesophyll-derived
signalling peptide, had significantly higher transcript abundance in well-watered samples on days 5
and 10 for both genotypes that corresponded with higher stomatal indices, congruent with its role
as a positive regulator in stomatal development. ERECTA transcript abundance was reduced in
well-watered samples on day 5 for both genotypes; however, variation in transcript abundance later
in development may be a result of the other roles of ERECTA in plant development. Variation in
transcript accumulation of ER and SDD1 between genotypes may indicate variation in drought-
response strategies, specifically with respect to the modulation of development in response to
water-deficit stress. TMM and YDA may not have notable roles in the regulation of stomatal
88differentiation in response to drought.
4.7 Acknowledgements
We are most grateful to Bruce Hall and Andrew Petrie for excellent greenhouse assistance, John
McCarron for the experimental set-up, Joan Ouellette for technical assistance, and Dave Kamelchuk
(Al-Pac) for collecting all the plant materials. We would also like to extend our gratitude for
helpful comments and feedback received by two anonymous reviewers. Research infrastructure and
technical support was generously provided by the Centre for Analysis of Genome Evolution and
Function at the University of Toronto. ETH was supported by an Ontario Graduate Scholarship in
Science and Technology. This work was supported by generous funding from the Natural Sciences
and Engineering Research Council of Canada (NSERC) and the University of Toronto to MMC.
89
4.8 Supplementary Figures
Supplementary Figure S4.1 Experimental design to test the change in transcript abundance
through time and across a developmental series in Populus balsamifera. The first fully expanded
leaf (red circle) and first expanding leaf (red arrow) was marked at the onset of the water-deficit
experiment (day 0), these leaves were subsequently followed throughout the experimental period
(30 days). This enabled collection of leaf tissue that developed throughout the water-deficit
experimental at day 5, 10, 15, 20, 25 and 30. The first fully expanded leaf at day 0 represented a
leaf that was fully developed prior to the onset of water-deficit treatment.
First expanding leaf at day zero
First fully expanded leaf at
day zero
Day 0 Day 30
Time
90
Supplementary Figure S4.2 Variation in the relative transcript accumulation of aPopulus TMM
homologue as determined by qRT-PCR. Transcript abundance calculated relative to ACT-7
transcript abundance levels.
Days of water withdrawal
(a)
(b)
5 10 15 20 25 30
0.0
0.5
1.0
1.5
2..5(c)
Days of water withdrawal
AP-1005
AP-1006
5 10 15 20 25 30
01
23
4
5 10 15 20 25 30
01
23
4
Days of water withdrawal
TMM
: Rel
ativ
e tra
nscr
ipt a
bund
ance
well-wateredwater-deficit
TMM
: Rel
ativ
e tra
nscr
ipt a
bund
ance
TMM
: Rel
ativ
e tra
nscr
ipt a
bund
ance
91
Supplementary Figure S4.3 Variation in the relative transcript accumulation of a Populus YODA
homologue as determined by qRT-PCR. Transcript abundance calculated relative to ACT-7
transcript abundance levels.
(a)
(b)
(c)
Days of water withdrawal
AP-1005
AP-1006
5 10 15 20 25 30
Days of water withdrawal0.
00.
51.
01.
52.
02.
53.
0
5 10 15 20 25 30
Days of water withdrawal
0.0
0.5
1.0
1.5
2.0
2.5
3.0
5 10 15 20 25 30
0.0
0.5
1.0
1.5
2.0
2.5
well-wateredwater-deficit
YDA:
Rel
ativ
e tra
nscr
ipt a
bund
ance
YDA:
Rel
ativ
e tra
nscr
ipt a
bund
ance
YDA:
Rel
ativ
e tra
nscr
ipt a
bund
ance
92
Supplementary Figure S4.4 Variation in the relative transcript accumulation of aPopulus ERECTA
homologue as determined by qRT-PCR. Transcript abundance calculated relative to ACT-7
transcript abundance levels.
5 10 15 20 25 30
Days of water withdrawal
01
23
45
6
5 10 15 20 25 30
Days of water withdrawal
01
23
45
6
AP−1005 AP−1006
0.0
0.5
1.0
1.5
2.0
2.5
5 10 15 20 25 30
01
23
45
AP−1005AP−1006
Days of water withdrawal Genotype
(d)
AP-1005 AP-1006 well-wateredwater-deficit
Genotype x Day Interaction Genotype main-effect
ER: R
elat
ive
trans
crip
t abu
ndan
ceER
: Rel
ativ
e tra
nscr
ipt a
bund
ance
ER: R
elat
ive
trans
crip
t abu
ndan
ceER
: Rel
ativ
e tra
nscr
ipt a
bund
ance
(c)
(a) (b)
93
Supplementary Figure S4.5 Variation in the relative transcript accumulation of aPopulus
STOMATAL DENSITY AND DISTRIBUTION-1 homologue as determined by qRT-PCR.
Transcript abundance calculated relative to ACT-7 transcript abundance levels.
(a) (b)
(c)
0.0
0.5
1.0
1.5
2.0
2.5
Days of water withdrawal Genotype
(d)
AP-1005 AP-1006
5 10 15 20 25 30
Days of water withdrawal
01
23
4
5 10 15 20 25 30
Days of water withdrawal
01
23
4
AP−1005 AP−10065 10 15 20 25 30
0.0
0.5
1.0
1.5
2.0
2.5
3.0
AP−1005AP−1006
Genotype x Day InteractionGenotype main-effect
well-wateredwater-deficit
SDD
1: R
elat
ive
trans
crip
t abu
ndan
ceSD
D1:
Rel
ativ
e tra
nscr
ipt a
bund
ance
SDD
1: R
elat
ive
trans
crip
t abu
ndan
ceSD
D1:
Rel
ativ
e tra
nscr
ipt a
bund
ance
94
Supplementary Figure S4.6 Variation in the relative transcript accumulation of aPopulus FAMA
homologue as determined by qRT-PCR. Transcript abundance calculated relative to ACT-7
transcript abundance levels
5 10 15 20 25 30
Days of water withdrawal
01
23
4
5 10 15 20 25 30
Days of water withdrawal
01
23
4
(a)
(b)
5 10 15 20 25 30
01
23
4
Days of water withdrawal
(c)
AP-1005
AP-1006
well-wateredwater-deficit
Treatment x Day Interaction
FAM
A: R
elat
ive
trans
crip
t abu
ndan
ceFA
MA:
Rel
ativ
e tra
nscr
ipt a
bund
ance
FAM
A: R
elat
ive
trans
crip
t abu
ndan
ce
95
Supplementary Figure S4.7 Variation in the relative transcript accumulation of aPopulus
STOMAGEN homologue as determined by qRT-PCR. Transcript abundance calculated relative to
ACT-7 transcript abundance levels
well-wateredwater-deficit
5 10 15 20 25 30
Days of water withdrawal
STO
MAG
EN: F
old
Indu
ctio
n0
12
34
56
5 10 15 20 25 30
Days of water withdrawal
STO
MAG
EN: F
old
Indu
ctio
n0
12
34
56
(a)
(b)
(c)
05 10 15 20 25 3
01
23
45
67
Days of water withdrawal
STO
MAG
EN: F
old
Indu
ctio
n
AP-1005
AP-1006
Treatment x Day Interaction
96
ER
TMM
Stom
atal
Inde
x
FMA
STO
MAG
EN
SDD
1
YDA
ER
TMM
Stomatal Index
FMA
STOMAGEN
SDD1
YDA
-1 0 0.5 1
Color Key
-0.5Correlation coefficient
0.2750.725
0.4360.564
0.4740.526
0.8870.113
0.9890.011
0.9990.001
0.2880.712
0.4170.583
0.4960.531
0.8820.118
0.9870.013
0.1300.870
0.4980.502
0.5950.405
0.9460.054
-0.1930.807
0.5930.407
0.8100.190
-0.6310.369
0.3920.608
-0.4310.569
ER
SDD
1
YDA
TMM
FMA
Stom
atal
Inde
x
STO
MAG
EN
ER
SDD1
YDA
TMM
FMA
Stomatal Index
STOMAGEN0.1390.861
-0.2370.763
0.0370.963
0.5390.461
0.7470.253
0.9780.02
0.0150.985
-0.1200.880
0.0820.918
0.5640.436
0.5920.408
0.5090.491
-0.5170.483
-0.0960.904
0.2650.735
-0.6880.312
-0.6970.303
-0.7440.256
0.6580.342
0.8970.103
0.3000.700
ER
Stom
atal
Inde
x
TMM
SDD
1
YDA
FMA
STO
MAG
EN
ER
Stomatal Index
TMM
SDD1
YDA
FMA
STOMAGEN0.4430.557
-0.2770.723
0.1670.833
0.5560.444
0.6840.316
0.9310.069
-0.7080.292
-0.2720.728
-0.4890.511
0.3580.642
0.8980.102
-0.8560.144
-0.0770.923
0.7320.268
-0.0350.965
-0.0800.920
-0.8170.183
-0.6950.305
-0.4890.511
0.5640.436
0.4430.557
(a) Day 5
(b) Day 10
(c) Day 15
97
Supplementary Figure S4.8 Pearson correlation coefficient (PCC) heat map representing the
transcript accumulation profiles and stomatal indices in P. balsamifera at (a) 5 d, (b) 10 d, and (c)
15 d after the imposition of water-deficit stress. The Pearson correlation coefficient (r; top) and
P-value (bottom) are indicated within each square.
98
Supplementary Table S4.1 Primers used for qRT-PCR analysis.
Forward Primer Reverse PrimerSTOMAGEN TTGTTATTCAAGGATCCAGA CTGTGGAGCCAATCATCAATERECTA ATCCAGGGCTGATGACAACA ACAGTAATGCAAGTTGGAAAFAMA ATCAGTGCCAAGCTTGAAGA AACACAGGGCAGTTGCTTCCSDD1 AATATCATGTCAGGTACATC TTGTCACTAACATAGGCAGTTMM CCTGATGGCGAAGAAAAGGC CTGGGAGCGTTAGGCGAGTGYODA CAAGCGTGATGCGACAGGATC AAATTGCCCGTTTTGGTGGTGACT7 GCATCACACCTTCTACAATGAGC CCTGGATAGCGACATACATTGC
4.9 Supplementary Tables
99Supplementary Table S4.2 Relative water content (RWC) on day 30.
AP-1005 AP-1006Well-watered 93 .13 ± 0 .51 93 .98 ± 0 .76Water-deficit 82 .88 ± 1 .92 78 .57 ± 2 .33
100Supplementary Table S4.3 ANOVA results: transcript abundance.
ERECTA Df SS MS F-value p-valueTreatment 1 0 .219 0 .219 0 .171 0 .68Genotype 1 13 .98 13 .98 10 .892 0 .001 *Day 5 55 .99 11 .198 8 .725 6 .57E-6 *Treatment: Genotype 1 7 .213 7 .213 7 .213 0 .021 *Treatment:Day 5 24 .125 4 .825 4 .825 0 .006 *Genotype:Day 5 7 .42 1 .484 1 .156 0 .344Treatment:Genotype:Day 5 14 .38 2 .876 2 .876 0 .65*data shown in supplemental figure 4
FAMA Df SS MS F-value p-valueTreatment 1 1 .217 1 .217 3 .998 0 .049 *Genotype 1 0 .002 0 .002 0 .007 0 .936Day 5 9 .838 1 .968 25 .792 3 .53E-6 *Treatment: Genotype 1 0 .229 0 .229 0 .828 0 .367Treatment:Day 5 3 .653 0 .731 2 .642 0 .035 *Genotype:Day 5 3 .533 0 .707 2 .555 0 .060 .Treatment:Genotype:Day 5 1 .015 0 .203 0 .877 0 .601*data shown in supplemental figure 6
SDD1 Df SS MS F-value p-valueTreatment 1 0 .469 0 .469 1 .414 0 .240Genotype 1 2 .153 2 .153 6 .495 0 .014 *Day 5 3 .753 0 .751 2 .226 6 .29E-2 *Treatment: Genotype 1 0 .046 0 .046 0 .137 0 .713Treatment:Day 5 0 .502 0 .100 0 .303 0 .909Genotype:Day 5 4 .909 0 .982 2 .98 0 .021 *Treatment:Genotype:Day 5 1 .629 0 .326 0 .983 0 .438*data shown in supplemental figure 5
101
STOMAGEN Df SS MS F-value p-valueTreatment 1 2 .689 2 .689 3 .084 0 .086 .Genotype 1 0 .301 0 .301 0 .345 0 .560Day 5 130 .33 26 .066 29 .897 2 .16E-13 *Treatment: Genotype 1 0 .188 0 .188 0 .216 0 .644Treatment:Day 5 14 .03 2 .806 3 .218 0 .014 *Genotype:Day 5 9 .875 1 .975 2 .265 0 .064 .Treatment:Genotype:Day 5 3 .965 0 .793 0 .91 0 .483*data shown in supplemental figure 7
TMM Df SS MS F-value p-valueTreatment 1 0 .031 0 .031 0 .066 0 .798Genotype 1 1 .163 1 .163 2 .451 0 .123Day 5 14 .085 2 .817 5 .938 1 .77E-2 *Treatment: Genotype 1 2 .131 2 .131 4 .491 0 .038 *Treatment:Day 5 0 .79 0 .158 0 .333 0 .566Genotype:Day 5 1 .23 0 .246 0 .519 0 .474Treatment:Genotype:Day 5 2 .04 0 .408 0 .86 0 .357*data shown in supplemental figure 2
YDA Df SS MS F-value p-valueTreatment 1 0 .001 0 .001 0 .002 0 .961Genotype 1 0 .115 0 .115 0 .296 0 .588Day 5 34 .09 6 .818 17 .535 8 .80E-5 *Treatment: Genotype 1 0 .081 0 .081 0 .209 0 .649Treatment:Day 5 0 .015 0 .003 0 .007 0 .932Genotype:Day 5 4 .64 0 .928 2 .387 0 .127Treatment:Genotype:Day 5 0 0 0 0 .992*data shown in supplemental figure 3
* p-value <0 .05 . p-value <0 .1
102
Chapter 5: Integrated analysis of the drought metabolome and transcriptome in Populus balsamifera
Erin T. Hamanishi, Genoa Barchet, Shawn D. Mansfield and Malcolm M. Campbell.
Contributions: ETH, GB, SDM and MMC designed research and organized experimental
logistics; ETH established biological materials, collected samples and analyzed data; GB and SDM
performed metabolic profiling by GC-MS; ETH wrote chapter with editorial assistance from SDM
and MMC.
103
Chapter 5: Integrated analysis of the drought metabolome and transcriptome in Populus balsamifera
5.1 Abstract
Drought has a major impact on tree growth and survival. Understanding tree responses to this
stress can have important application in both conservation of forest health, and in production
forestry. Trees of the genus Populus provide an excellent opportunity to explore the mechanistic
underpinnings of forest tree drought responses, given the growing molecular resources that are
available for this taxon. Here, foliar tissue of six water-deficit stressed P. balsamifera genotypes
was analysed for variation in the metabolome in response to drought and time of day by using
an untargeted metabolite profiling technique, gas chromatography/mass-spectrometry (GC/
MS). Significant variation in the metabolome was observed in response the imposition of water-
deficit stress. Notably, organic acid intermediates such as succinic and malic acid decreased in
accumulation in response to drought, whereas galactinol and raffinose increased in accumulation.
A significant proportion of metabolites with significant difference in accumulation under water-
deficit conditions exhibited intraspecific variation in metabolite accumulation. Larger magnitude
fold-change accumulation was observed in genotype AP- 947, AP-1005 and AP-2278. In order
to understand the interaction between the transcriptome and metabolome, an integrated analysis
of the drought responsive transcriptome and the metabolome was performed. Genotype AP-1006
demonstrated a lack of congruence between the magnitude of the drought transcriptome response
and the magnitude of the metabolome response. More specifically, metabolite profiles in AP-1006
demonstrated the smallest changes in response to water-deficit conditions. Pathway analysis of the
transcriptome and metabolome revealed specific genotypic responses with respect to primary sugar
accumulation, citric acid metabolism and raffinose family oligosaccharide biosynthesis.
5.2 Introduction
Water limitation, particularly periods of drought, impinge on plant growth and survival. There is a
wealth of evidence underscoring the importance of changes in plant biochemistry as a mechanism
to contend with drought. For example, amino acids including proline (Pro), valine (Val) and
isoleucine (Ile)], carbohydrates [such as sucrose, raffinose family oligosaccharides (RFO) and
sorbitol], polyols, and organic acids vary in abundance in response to drought (Krasensky & Jonak
1042012). Elevated levels of sucrose were observed in leaf tissue of water-stressed Populus tomentosa
(Nishizawa et al. 2008); whereas a combination of glucose, fructose and sucrose accumulated in
Populus hybrids in response to drought (Kozlowski & Pallardy 2002). Some of these compounds
are thought to function as osmolytes, maintaining cell turgor and stabilisation of cellular proteins
(Seki et al. 2007). Raffinose, and other raffinose-type oligosaccharides, accumulate in response
to water-stress, and are hypothesised to be osmoprotectants, with the capacity for membrane and
enzyme stability (Taji et al. 2002; Krasensky & Jonak 2012), along with a putative role as hydroxyl
radical scavengers.
Proline accumulation has long been associated with stress tolerance in plants, and is likely one of the
most widely distributed osmolytes among plants and animals (Bartels & Sunkar 2005; Nishizawa
et al. 2008). Similar to carbohydrates, proline is hypothesised to aid in the osmotic adjustment in
response to drought; however, proline is also hypothesised to have roles in reactive oxygen species
(ROS) scavenging and membrane stability. Proline accumulated in severely water-stressed mature
Populus nigra leaves (Kozlowski & Pallardy 2002; Cocozza et al. 2010); whereas no significant
increase in proline accumulation was observed in field-grown, drought treated Populus hybrids .
Proline accumulation in response to drought is variable in among the long-lived, woody Populus
trees and is hypothesised to be the result of the varied role of amino acids in the osmotic adjustment
process, whereby carbohydrates may play a more predominant role in osmotic adjustment in Populus
than amino acids.
Organic acids have also been implicated in the biochemical response to drought. For example,
malic acid increased in abundance under mild periods of water-stress (Tschaplinski et al. 1994;
Escobar-Gutiérrez et al. 1998; Seki et al. 2007). Unlike carbohydrate and amino acid accumulation,
malic acid accumulation may be a function of the stomatal system in plants rather than being
osmotically active (Wilkinson & Davies 2002).
Unsurprisingly, changes in metabolites in response to drought appear to be underpinned by changes
in transcript abundance of specific genes. Large-scale microarray experiments studying water-
deficit stress have identified many transcripts involved in stress tolerance, including key enzymes
involved in the biosynthesis of osmotically-important metabolites in grape (Cramer et al. 2006),
Arabidopsis (Seki et al. 2002; Kreps et al. 2002; Harb et al. 2010), rice (Rabbani et al. 2003) and
poplar (Brosche et al. 2005; Wilkins et al. 2009; Hamanishi et al. 2010; Krasensky & Jonak 2012;
105Yan et al. 2012). For example, galactinol synthase frequently has increased transcript abundance
in response to drought in plants, including Arabidopsis (Taji et al. 2002; Nishizawa et al. 2008) and
Populus (Kozlowski & Pallardy 2002; Wilkins et al. 2009; Hamanishi et al. 2010; Yan et al. 2012).
As the response to drought-stress is not simply the product of the drought-responsive transcriptome,
complexity in the whole plant response to drought is the result of the interactions between genes,
transcripts, proteins, metabolites and the environment. The model plant genus Populus provides
an opportunity to explore the relationship between the drought transcriptome and the drought
metabolome. In keeping with this, the relationship between the transcriptome and metabolome for
specific metabolic pathways in Populus has also been characterised in response to salt-stress, revealing
the importance of control mechanisms for osmotic adjustment (Seki et al. 2007; Janz et al. 2010).
In order to test hypotheses related to intra-specific variation in drought responses in Populus, the
transcriptomes and metabolomes of six genotypes of P. balsamifera were examined. Shared versus
genotype-specific P. balsamifera drought transcriptomes were identified (Hamanishi et al. 2010)
and superimposed onto metabolome variation. This approach identified important pathways in the
drought response, and highlighted genotypic-specific responses that provide insight into different
mechanisms of acclimation to water-limiting conditions.
5.3 Materials and Methods
5.3.1 Plant material and experimental design
Populus balsamifera ramets were grown in a climate controlled growth chamber at the University
of Toronto using conditions as described by Hamanishi et al. (2010). Un-rooted cuttings of six
P. balsamifera genotypes (Alberta Pacific, Boyle, Alberta) were propagated and grown under well-
watered conditions for 9 weeks, at which point, water-deficit stress was imposed on half the trees by
withholding water, while temperature, light and relative humidity remained constant.
Foliar tissue was harvested for metabolite and transcriptome analysis 15 days after the onset of the
water-withdrawal experiment. For the transcriptome analysis, the first fully-expanded, mature
leaf was collected from each tree; three leaves were pooled to create a single replicate. Triplicate
replicates were collected for each genotype and treatment combination at pre-dawn (PD; 1 hour
before the light period)) and mid-day (MD; middle of the light period). Leaves were immediately
106flash frozen, and then ground to a fine powder in preparation for RNA isolation, as described by
Hamanishi et al (2010). For the metabolite analysis, a single mature, fully-expanded leaf (leaf
plastochron index (LPI) = 5-7) was collected from each tree (n=10 per genotype per treatment at
MD and PD). Harvested foliar tissue was weighed to determine fresh weight (FW), subsequently
freeze-dried, and weighed again to determine dry weight (DW). Foliar tissue harvested and
prepared for metabolite analysis was shipped to the laboratory of Professor Shawn Mansfield in the
Faculty of Forestry at the University of British Columbia for analysis.
5.3.2 Non-targeted metabolic profiling by gas chromatography/mass spectrometry
Metabolite extraction was performed using a methanol/chloroform-based extraction protocol as
described by Robinson et al. (2005) and Barchet (2010) at the University of British Columbia.
Approximately 0.5mL of sample was extracted in 1300 uL 97% methanol with the internal standard
ortho-anisic acid (0.62mg mL-1) for 15 minutes at 70°C prior to centrifugation at 17,000 g for 10
minutes. The supernatant was transferred to a new 1.5 mL tube. 130 uL chloroform and 270 uL
distilled, deonized water was added and the tube was gently shaken prior to centrifugation at 17,000
g for 5 minutes. A 400 uL aliquot of the upper polar phase was transferred to a new 1.5mL tube
and dried overnight at 30°C in a Vacufuge (Eppendorf ).
Samples were then derivatized for GC/MS analysis by resuspension in 50uL methoxyamine
hydrochloride solution (20 mg mL-1 in pyridine) and incubated at 37°C for two hours. 10 uL of
n-alkane standard and 70 uL of N-methyl-N-trimethylsilytriflouroacetamide (MSTFA) was added,
and incubated at 37°C for 30 minutes with constant agitation. Samples were then filtered through
filter paper and allowed to rest at room temperature until GC/MS analysis.
GC/MS analysis was conducted on a ThermoFinnigan Trace GC-PolarisQ ion trap MS, fitted with
an AS2000 auto-sampler and a split-injector (Thermo Electron Co., Waltham, MA, USA). The GC
was equipped with a Restek Rtx-5MS column (fused silica, 30m, 0.25 mm ID, stationary phase:
5% diphenyl 95% dimethyl polysiloxane). The GC conditions were set with an inlet temperature
of 250 °C, helium carrier gas at a constant flow rate of 1 mL min-1, injector split ratio 10:1, resting
oven temperature at 70°C and a GC/MS transfer line temperature of 300°C. After a sample
injection of 1 uL, the oven temperature was held at 70°C for two minutes prior to ramping to
107325°C at a rate of 8°C min-1. The temperature was held at 325°C for six minutes before cooling to
the initial resting oven temperature, prior to the next run.
For MS analysis in the positive electron ionization mode an ionization potential of 70eV was used
and the foreline was evacuated to 40 mTorr with helium gas flow in to the chamber set at a rate of
0.3 mL min-1 and the source temperature was held at 230°C. Detector signal was recorded from
3.35-35.5 minutes after the injection, and, with a total scan time 0f 0.58 s, ions were scanned across
the range of 50-650 mass units.
5.3.3 Metabolome: data processing and statistics
The raw metabolite data generated by GC/MS for each metabolite was normalised through
comparison to internal standards and normalised to freeze-dried DW for each tissue sample.
Raw-data was processed using XCMS as described by Barchet (2010; Krasensky & Jonak 2012).
Descriptive statistics were calculated using R 2.14.1 (R Development Core Team 2009). For
subsequent analyses, the metabolite data were log10 transformed. The dataset comprised 87
metabolites, 181 samples (n=4-10 per genotype per treatment per time of day).
Metabolic profiles for all samples were subjected to hierarchical cluster analysis (HCA) using
Pearson correlation coefficient (Eisen et al. 1998; Rabbani et al. 2003) to search for metabolic
similarities and differences among samples and metabolites. The uncertainty associated with HCA
was assessed generating a consensus dendrogram on 1000 bootstrap replicates using the R package
pvclust (Suzuki & Shimodaira 2006). Over-representation of a given metabolite class within a
cluster was determined using Fisher’s exact test in R (R Development Core Team 2009). Statistical
significance was calculated using a three-way ANOVA. The P-values were corrected for multiple
hypothesis testing using the false discovery rate (FDR) procedure of Benjamini and Hochberg
(Benjamini & Hochberg 1995). A P-value of < 0.05 was considered statistically significant.
5.3.4 RNA isolation and transcriptome analysis
RNA isolation and microarray analysis was performed as described by Hamanishi et al. (2010;
Chapter 3 this volume); however, the global drought transcriptome was considered to include
all transcripts significant for a treatment-main effect (P<0.05) with no log2(fold-change) cutoff.
Weighted co-expression network analysis (WGCNA) was performed using the R statistical package
108WGCNA with a power of 7 (Langfelder & Horvath 2008). Functional annotations were assigned
based on the most recent version probe-set annotations from Affymetrix (NetAffx build 32).
Networks generated with WGCNA were plotted using Cytoscape (Lopes et al. 2010). Analysis of
GO term enrichment was calculated by comparing the number of annotations within the list of
query transcripts to all annotated transcripts on the Poplar Affymetrix Genome Array. Statistical
significance was calculated using Fisher’s exact test in R (R Development Core Team 2009), and
applying the Benjamini-Hochberg correction to adjust for FDR . Overrepresentation of GO Slim
terms was confirmed and plotted using AgriGO (Du et al. 2010). Molecular pathways relevant to
the drought transcriptome/metabolome were previously characterised in Kyoto Encyclopedia of
Genes and Genomes (KEGG: Kanehisa & Goto 2000; Masoudi-Nejad et al. 2007; Kanehisa et al.
2011).
5.4 Results and Discussion
5.4.1 Populus balsamifera genotypes were subjected to water withdrawal to induce a drought response
To investigate the impact of drought-like conditions on the abundance of Populus balsamifera
metabolites, six genotypes (AP-947, AP-1005, AP-1006, AP-2278, AP-2298 and AP-2300)
were exposed to a prolonged period of water-withdrawal. All plants were grown under the same
controlled growth conditions for 9 weeks, after which half of the plants continued to receive water
(well-watered) and the other half received no water (water deficit). This divergence in treatment
continued for15 days, at which point plant physiology was recorded, and foliar tissue for metabolic
and transcriptome analysis was collected at pre-dawn (PD) and mid day (MD).
Under conditions of water deficit, declines in above-ground biomass and relative water content
(RWC) were observed in all genotypes (Seki et al. 2002; Kreps et al. 2002; Hamanishi et al. 2010;
Harb et al. 2010). Stomatal conductance significantly decreased in all genotypes, with the greatest
decline observed in AP-1006; whereas genotype AP-2278 had the smallest reduction in stomatal
conductance in response to the imposition of water-deficit conditions (Hamanishi et al. 2010,
Chapter 3, this volume). Net photosynthetic rate also decreased in response to water-deficit
conditions after 15 days of water-withdrawal (treatment main effect; ANOVA, P < 0.05); however, a
significant decline only occurred in genotype AP-1005, AP-1006 and AP-2298 (Welch’s two-sample
109t-test, P < 0.05; Figure 5.1). The net photosynthetic rates were lower in all six genotypes 15 days
after onset of water withdrawal than net photosynthetic rates observed in field-grown P. balsamifera
(Silim et al. 2010). Reduced photosynthetic rates observed in the chamber grown seedlings may
be attributable to the lower light levels in the growth-chamber at the University of Toronto as
compared to ambient levels in field-grown seedlings or trees.
5.4.2 Variation in Populus balsamifera metabolite profiles was evident
To differentiate between genotypic (G), treatment (T) and time-of-day (D) effects, metabolic
profiles of P. balsamifera were analysed using gas chromatography/mass spectrometry (GC/MS).
Trend analysis was restricted to 87 metabolites that were identified across all samples (n=4-10 per
genotype per treatment per time of day), which represented both known and unknown metabolites
(Table 5.1). Hierarchical clustering analysis using the Pearson correlation coefficient revealed the
relative changes in metabolite abundance across samples (Figure 5.2). The dendrogram suggested
that there was a large degree of variation in the metabolite abundance for a given metabolite among
samples, yet both genotype and treatment appeared to play an important role in the segregation
of samples. Notably, the metabolite profiles from water-deficit samples of AP-1005 and AP-2278
appeared most different from the other metabolomes, whereas samples of genotype AP-947 and AP-
2300 clustered in a genotype-wise fashion regardless of treatment or time of day.
Although the metabolomes were highly variable among samples, further investigation of the
relationship among metabolites revealed 13 significant clusters of metabolites that had a high degree
of similarity in their abundance profiles across all samples (Figure 5.3) as determined by hierarchical
cluster analysis using Pearson correlation coefficient. These clusters may be indicative of different
mechanisms of regulation for these metabolites. For example, three of the 13 clusters had significant
over-representation of a given metabolite class (Fisher’s exact test; Padj < 0.05). Specifically, cluster
II comprised predominantly carbohydrates (Padj = 0.00366), cluster IX was all organic acids (Padj =
0.00245) and cluster XII was primarily amino acids (Padj = 0.000251; Figure 5.2).
A three-way factorial analysis of variance (ANOVA) identified metabolites that had significantly
different abundance in response to drought [Treatment (T) main effect], genotype (G main effect),
time of day (D main effect), as well as any interaction between the three experimental factors (Table
5.2). Similar to HCA results, significant variation in the metabolic profiles was attributable to
110
Figure 5.1 Box-plot representing net photosynthetic rate (µmol CO2 m-2 s-1) for genotype AP-947,
AP-1005, AP-1006, AP-2278, AP-2298 and AP-2300. Well-watered samples represented by blue
boxes; water-deficit-treated samples represented by orange boxes (n=3 per treatment per genotype).
The midline of the box represents the median value for photosynthesis, the upper and lower bounds
of the box represent the interquartile range, and the whiskers extend to the most extreme values that
are not outliers.
45
67
89
AP-947 AP-1005 AP-1006 AP-2278 AP-2298 AP-2300
TreatmentWell wateredWater-deficit
Phot
osyn
thes
is (µ
mol
CO
2 m-2 s
-1)
111
Number of
Metabolites
Percent (%) of
MetabolitesGenotype 79 91 .95%Treatment 40 45 .98%Time of Day 11 12 .64%Genotype:Treatment 41 47 .13%Genotype:Time of Day 6 5 .75%Treatment:Time of Day 15 17 .24%Genotype:Treatment:Time of Day 0 0 .00%
Table 5.1 Number of metabolites with significant main effects or interactions (n=87 metabolites).
Padj-value cutoff = 0.05 (Benjamini-Hochberg).
112
Figure 5.2 Dendrogram obtained after hierarchal clustering analysis (HCA) of the metabolic
profiles of the six P. balsamifera genotypes under well-watered and water-deficit conditions at mid-
day and pre-dawn time point. Rows represent specific metabolites (n=87). Columns represent
mean intensity of all replicates for each genotype, treatment and time of day sample. Plotted values
are the mean of n = 4-10 replicates for each sample. Metabolite classes: AA = Amino Acid; C =
Carbohydrate; P = Phenolic, SA = Sugar Alcohol. NI = Not Identified.
−2 0Row Z-Score
Colour Key
AP947AP1005AP1006AP2278AP2298AP2300
GenotypeMid-dayPre-dawn
Time of Day
WetDry
Treatment
2
GenotypeTreatment
Time of Day
NI (8)MaltoseAdenosineSalicinNI (22)NI (Amino Acid; 1)NI (28)NI (3)Shikimic AcidQuercitinKaempferolNI (24)NI (18)L-Ascorbic AcidPyroglutamic AcidL-AlanineL-TheronineL-SerineL-GlutamateButyric_acid-4-amino-n- GlycineL-Proline (1)NI (Amino Acid; 3)L-Proline (2)NI (19)L-Phenylalanine SucroseMyo-inositolQuinic AcidFumaric AcidAspartic AcidMalonic AcidNI (5C Sugar)NI (5C Sugar; 4)NI (2)NI (5C Sugar; 2)NI (Amino Acid; 2)GlycerolThreonic acid 1,4-lactoneThreonic acidGlycolic AcidSuccinic AcidMalic AcidNI (25)NI (7)Thymidine-5'-monophophateNI (17)NI (29)NI (12)Digalactosyl glycerolNI (4)NI (20)NI (11)NI (13)Citric AcidRaffinoseGalactinolNI (27)NI (10)NI (5)NI (15)Ni (9)MelibioseNI(1)NI (5C Sugar; 1)NI (Sugar Alcohol)L-IsoleucineBenzoic AcidCatecholNI (6C Sugar; 2)NI (16)NI (Organic Acid)Fructose (2)NI (6C Sugar; 1)Glucose (2)Glucose (1) Glucose (3)Fructose (1) NI (26)NI (5C Sugar; 3)L-TyrosineSalicyl_alcoholNI (23)NI (6)NI (21)NI (14)Catechin
113
Figure 5.3 HCA reveals 13 significant clusters (P < 0.05). Significant clusters are labeled with
unique colours and numbered (I through XIII) for identification. Hierarchical clustering was done
using pvclust (Suzuki & Shimodaira 2006), with a correlation distance measure and a complete
agglomerative clustering method.
0.0
0.5
1.0
1.5
Cat
echi
nN
I (14
)N
I (21
)N
I (6)
NI (
23)
Salic
yl A
lcoh
olL-
Tyro
sine
NI (
5C S
ugar
; 3)
NI (
26)
Fruc
tose
(1)
Glu
cose
(3)
Glu
cose
(1)
Glu
cose
(2)
NI (
6C S
ugar
; 1)
Fruc
tose
(2)
NI (
Org
anic
Aci
d)N
I (16
)N
I (6C
Sug
ar; 2
)C
atec
hol
Benz
oic
Acid
L−Is
oleu
cine
NI (
Suga
r Alc
ohol
)N
I (5C
Sug
ar; 1
)N
I(1)
Mel
ibio
seN
I (9)
NI (
15)
NI (
5)N
I (10
)N
I (27
)G
alac
tinol
Raf
finos
eC
itric
Aci
dN
I (13
)N
I (11
)N
I (20
)N
I (4)
Dig
alac
tosy
l gly
cero
lN
I (12
)N
I (29
)N
I (17
)Th
ymid
ine−
5’−m
onop
hoph
ate
NI (
7)N
I (25
)M
alic
Aci
dSu
ccin
ic A
cid
Gly
colic
Aci
dTh
reon
ic a
cid
Thre
onic
aci
d 1,
4-la
cton
eG
lyce
rol
NI (
Amin
o Ac
id; 2
)N
I (5C
Sug
ar; 2
)N
I (2)
NI (
5C S
ugar
; 4)
NI (
5C S
ugar
; 5)
Mal
onic
Aci
dAs
parti
c Ac
idFu
mar
ic A
cid
Qui
nic
Acid
Myo
−ino
sito
lSu
cros
eL−
Phen
ylal
anin
eN
I (19
)L−
Prol
ine
(2)
NI (
Amin
o Ac
id; 3
)L−
Prol
ine
(1)
Gly
cine
Buty
ric_a
cid−
4−am
ino−
n−L−
Glu
tam
ate
L-Se
rine
L−Th
eron
ine
L−Al
anin
ePy
rogl
utam
ic A
cid
L−As
corb
ic A
cid
NI (
18)
NI (
24)
Kaem
pfer
olQ
uerc
itin
Shik
imic
Aci
dN
I (3)
NI (
28)
NI (
Amin
o Ac
id; 1
)N
I (22
)Sa
licin
Aden
osin
eM
alto
seN
I (8)
[I] [II] [III] [IV] [V] [VI] [VII] [VIII] [IX] [X] [XI] [XII] [XIII]
Hei
ght
Metabolite Clusters
114genotype. A large proportion of metabolites had differential abundance among genotypes (n=79;
P < 0.05; Table 5.2). Of these 79 metabolites, 38 had no significant two- (i.e., TxG- or DxG-
interaction) or three-way interactions.
ANOVA analysis also revealed a small subset (n=11) of metabolites that varied significantly in
abundance in response to time of day (main effect; Figure 5.4); however, a larger number of
metabolites (n = 15) had abundance that varied significantly in response to water-deficit treatment
in a time-of-day dependent fashion (TxD interaction; Table 5.2; Supplementary Figure 5.4).
Notably, proline had a significant increase at the PD relative to MD (Figure 5.4a). Conversely,
sucrose had increased abundance at the MD time point (Figure 5.4b). Sucrose is a diurnally
regulated metabolite that has increased abundance during light conditions, as seen in potato
(Urbanczyk-Wochniak et al. 2005) and in Populus (Hoffman et al. 2010).
5.4.3 A Populus balsamifera drought metabolome was identifiable
Water withdrawal induced significant changes in metabolite abundance. Forty metabolites had
different abundance levels in response to drought, regardless of genotype or time of day. Twenty-
one metabolites increased in abundance and 19 decreased in abundance (ANOVA; P < 0.05; Table
5.3; Figure 5.5a). No general class of metabolites responded to drought. For example, the amino
acid (AA) class had variable response to drought. The contribution of amino acids in Populus clones
is thought to be small relative to the effect of carbohydrates and other osmolytes (Tschaplinski &
Tuskan 1994); however, isoleucine had the largest fold increase in abundance in response to drought
of any metabolite assessed, but was the only branched chain amino acid (BCAA) to be analysed,
whereas aspartic acid and threonine decreased in abundance in the drought-treated samples.
Increased accumulation of BCAAs has been observed in other organisms including Arabidopsis
(Joshi & Jander 2009) and various wheat cultivars (Bowne et al. 2012). Although increased
accumulation of BCAAs has frequently been observed in response to abiotic stress, little is known
about their role in stress tolerance; however, accumulated BCAAs may serve as a substrate for the
synthesis of other stress-induced proteins and may act as signalling molecules in response to drought
stress (Nambara et al. 1998).
Two organic acids, representative of TCA cycle intermediates, succinic and malic acid, had a
general decline in abundance; whereas raffinose and galactinol, a trisaccharide and sugar alcohol
115Table 5.2 Metabolites with significantly different abundance levels in response to drought
(ANOVA, Padj-value < 0.05).
Group Metabolite Fold-changeAmino Acid Aspartic Acid -0 .68
L-Isoleucine 3 .32L-Threonine -0 .39NI (Amino Acid; 2) -0 .26
Carbohydrate Fructose (2) -0 .16Glycerol -0 .58Melibiose 0 .30NI (5C Sugar; 2) -1 .18NI (6C Sugar; 1) -0 .16Raffinose 1 .13Sucrose 0 .28
Organic Acid Benzoic Acid 1 .13Citric Acid 1 .07Fumaric Acid -1 .30Glycolic Acid -0 .39Malic Acid -0 .12Malonic Acid -1 .58Quinic Acid -0 .37Shikimic Acid -0 .55Succinic Acid -0 .97Threonic acid -0 .25Threonic acid 1,4-lactone -0 .57
Phenolic Catechol 0 .45Quercitin -0 .24Salicin 0 .83Salicyl_alcohol 1 .56
Sugar Alcohol Galactinol 0 .52Myo-inositol 0 .15
Not Identified NI (2) -0 .14NI (3) -0 .43NI (4) 0 .36NI (5) 0 .69NI (7) 0 .15Ni (9) 0 .24NI (10) 1 .36NI (11) 0 .53NI (15) 0 .97NI (18) 0 .39NI (20) 0 .55NI (25) 0 .53
116
Figure 5.4 Time of Day (D) main effect observed for (a) proline (2) and (b) sucrose between mid-
day (MD) and pre-dawn (PD).
MD PD
1.0
1.5
2.0
2.5
Proline (2)
MD PD
5.7
5.9
6.1
6.3
Sucrose
Abun
danc
e
Abun
danc
e
117
MD
AP-947 AP-1005 AP-1006 AP-2278 AP-2298 AP-2300
PDM
DPD
MD
PDM
DPD
MD
PDM
DPD
MD
AP-947 AP-1005 AP-1006 AP-2278 AP-2298 AP-2300
PDM
DPD
MD
PDM
DPD
MD
PDM
DPD
15
(a)
(b) (c
)
Trea
tmen
t m
ain
effe
ctTr
eatm
ent:G
enot
ype
Inte
ract
ion
1517
55
4
1
Trea
tmen
t:Tim
e of
Day
Inte
ract
ion
-0.2
-0.1
5-0
.1-0
.05
00.
050.
10.
15
Rel
ativ
e Ab
unda
nce
[log 2(F
old
Cha
nge)
]
NI
NI
NI
NI
NI
Mel
ibio
seSu
cros
eM
yo-in
osito
lN
IN
I(Am
ino
Acid
)L-
Thre
onin
e NI
Thre
onic
Aci
dG
lyce
rol
Que
rciti
n
−20
12
Row
Z−S
core
Col
or K
ey
−1
118Figure 5.5 Metabolite accumulation levels for treatment main effect and treatment x genotype
interaction. (a) Hierarchally clustered metabolites that are significant for treatment main effect
across all genotypes at two different time-points [pre-dawn (PD) and mid-day (MD)]. (b) Venn
diagram demonstrating the number of metabolites that are significant for treatment main effect or
a 2-way interaction. (c) Mean log2(fold-change) of metabolite abundance for metabolites that are
significant for treatment main effect only.
119Table 5.3 Module membership in the drought transcriptome network of AP-1006 and preservation
of drought modules among the other genotypes.
120
Col
our
Mod
ule
AP-
1006
Num
ber
of m
odul
e m
embe
rs
(gen
es)
Mod
ule
AP-
947
Mod
ule
AP-
1005
Mod
ule
AP-
2278
Mod
ule
AP-
2298
Mod
ule
AP-
2300
Ove
rlap
P-va
lue
Ove
rlap
P-va
lue
Ove
rlap
P-va
lue
Ove
rlap
P-va
lue
Ove
rlap
P-va
lue
yello
wM
1_10
0619
2-
6812
1-
-bl
ueM
2_10
0635
549
387
242
--
blac
kM
3_10
0693
--
58-
-tu
rquo
ise
M4_
1006
399
4030
914
044
244
brow
nM
5_10
0619
948
283
183
-38
gree
nM
6_10
0618
8-
140
149
9751
pink
M7_
1006
72-
201
81-
-re
dM
8_10
0611
174
128
53-
-m
agen
ta-
M9-
947
(27)
--
--
purp
le-
--
M10
_227
8 (5
9)-
-
121respectively, were some of the most highly accumulated metabolites in response to water-deficit
conditions (Table 5.3). Although a general decline was observed in abundance of malic acid,
patterns of accumulation in response to drought in Populus are often varied; both increased and
decreased levels of accumulation in response to drought have been observed (Tschaplinski & Tuskan
1994; Koussevitzky et al. 2008). Malic acid is a very abundant organic acid in plants, and its role
is likely not restricted to the citric acid cycle (Maclennan et al. 1963). Sugars have previously
been shown to increase in abundance in response to water-stress, having an important role in
the osmotic adjustment (Chaves et al. 2003; Regier et al. 2009). Raffinose and galactinol have
been hypothesised to be osmoprotectants in drought-stress conditions, and have frequently been
implicated in the drought response in plants (Taji et al. 2002; Nishizawa et al. 2008).
Notably, of the 40 metabolites that were significant for T main effect, only 15 did not show any
significant interactions (i.e., TxG or TxD; Figure 5.5b). Carbohydrates, a sugar alcohol, and some
unknown metabolites had increased abundance in water-deficit conditions, whereas decreased
abundance was exhibited by a variety of metabolites representative of different metabolite classes
(Figure 5.5c). As indicated by the large proportion of metabolites significant for TxG or TxD
interactions, the accumulation of metabolites was not simply due to the imposition water-deficit
stress, rather, metabolite accumulation was a complex response shaped by genotype and time of day.
The variation in metabolite accumulation across genotypes and at different time-points could be
exploited to further investigate the unique responses of P. balsamifera genotypes.
5.4.4 The drought metabolome varied among P. balsamifera genotypes
While a large proportion of metabolites had significant response to water-deficit treatment,
many of these varied in a genotype- (G) or a time-of-day- (D) dependent manner (Figure
5.5b). The abundance of 41 metabolites was significantly impacted by TxG interaction (Table
5.2; Supplemental Figure A.2). Certain metabolites had opposite patterns of accumulation in
response to drought (i.e., higher abundance in one genotype and lower abundance in another
genotype). Of note, glucose had elevated abundance levels in AP-947 and AP-1006, but decreased
abundance levels in the remaining four genotypes in response to water-deficit conditions. Similarly,
galactinol was significant for a G x T interaction (P = 0.0259), but the highest level of galactinol
accumulation was observed in drought treated samples of genotype AP-947 and AP-2278. Other
metabolites that had a significant TxG interaction also had consistent response to water-deficit stress
122among the six genotypes. Notably, considerable variation in the magnitude of change was observed.
For example, glycolic and threonic acids, two metabolites belonging to cluster IX (Figure 5.3)
decreased significantly in abundance in response to water-deficit conditions, with notable reductions
observed in genotype AP-1005 and AP-2278. Moreover, half of the metabolites that had significant
differences in abundance between treatments (T main effect) also varied in response to genotype
(n=20; Figure 5.5b) confirming the importance of genotype in defining the drought response
observed among samples.
Ten drought-responsive metabolites also had significant differences in abundance for a TxD
interaction, indicative of the variation in metabolite level observed between pre-dawn and mid day.
Raffinose abundance was significant for a TxD interaction, having ~2-fold increase in accumulation
in response to water-deficit at MD (P = 0.0122), but no significant change in abundance at PD
(Supplemental Figure 1).
A notable feature of the P. balsamifera drought metabolome was the magnitude of variation observed
between samples. On average, peak signal intensity (non-transformed data) varied ~3000-fold
between minimum and maximum peak intensity for any given metabolite (Supplementary Figure
5). Similarly, the magnitude of variation in metabolite accumulation between water-deficit and
well-watered samples varied considerably. Among the metabolites whose accumulation had a
significant T main effect, the fold-change variation ranged from ~3 fold decrease in malonic acid
accumulation to ~10 fold increase in isoleucine accumulation. Overall variation in the drought
metabolome was examined by Pearson correlation comparison of the log2(fold-change) of the
water-deficit metabolome of the six P. balsamifera genotypes. This analysis revealed which genotypes
had metabolome responses that were more equivalent to others (Figure 5.6). Notably, genotypes
AP-1005 and AP-2278 had not only the most similar metabolomes (Figure 5.2) but also the most
similar drought metabolomes (r = 0.845; P < 0.05), whereas genotypes AP-2300 (r < 0.550) and
AP-2298 (r < 0.606) were most divergent when compared to all other genotypes (Figure 5.6).
The magnitude of drought-induced changes in metabolite abundance among the six P. balsamifera
genotypes had a high degree of variation (Figure 5.4a). Notably, the largest absolute magnitude
change in drought responsive metabolites occurred in genotype AP-1005 (mean = 0.306, standard
deviation = 0.229) and AP-2278 (mean = 0.290; standard deviation = 0.234); whereas the smallest
magnitude change was observed in genotype AP-1006 (mean = 0.149; standard deviation = 0.150).
123
AP-2298
AP-2300
AP-1006
AP-947
AP-1005
AP-2278
0.4 0.6 0.8 1
Color Key
AP-2
298
AP-2
300
AP-1
006
AP-9
47
AP-1
005
AP-2
278
Pearson correlation coefficient, r
Figure 5.6 Variation in the drought metabolome among six genotypes of P. balsamifera represented
by a Pearson correlation coefficient (PCC) heatmap. Differential abundance [log2(fold-change)]
for metabolites significant for treatment main effect (ANOVA, P < 0.05) are represented. The PCC
value was calculated for each pair-wise comparison among genotypes, and is represented by the
colour in the given cell. All genotypes are represented on both the x- and y-axis in the same order.
124
5.4.5 There were correlations between drought-responsive metabolites and specific components of transcriptome remodelling
To assess relationships between drought-responsive metabolites and transcripts, the metabolomes
and transcriptomes of P. balsamifera were compared. These analyses made use of previously-reported
drought-responsive transcriptome data for P. balsamifera (Hamanishi et al. 2010). Quantitatively,
there was a high level of congruence between the metabolome and the transcriptome, where larger
magnitude changes in the transcriptome corresponded with larger magnitude changes in the
metabolome, with the notable exception for genotype AP-1006 (Figure 5.7). More specifically,
genotype AP-1006 and AP-2278 had significantly larger magnitude change in the drought
transcriptome relative to all other genotypes (Bonferroni’s P < 0.001; Supplementary Figure 7;
Figure 5.7b); whereas the absolute magnitude change observed in the metabolome for AP-1006 and
AP-2278 was among the smallest and largest, respectively. This suggests that coordination of the
transcriptome and metabolome is variable among genotypes, and that the overall magnitude change
in metabolite abundance does not necessarily reflect the magnitude of transcriptome variation
resulting from water-deficit treatment.
A correlation matrix of all pair-wise comparisons among drought responsive metabolites and
transcripts revealed 747 transcripts that were significantly correlated with at least one metabolite
(Pearson correlation coefficient, |r| > 0.60, P < 0.05) based on the similarity of abundance profiles
across all samples (Figure 5.8; Supplementary Figure 8). Correlation patterns between metabolites
and transcripts were similar among a majority of the organic acids with the exception of citric,
benzoic and shikimic acid. A significant proportion of organic acids were previously shown
(5.4.2 Behavior of metabolites in Populus balsamifera) to share similar patterns of abundance
across samples; however, citric, benzoic and shikimic acid did not. Similarly, three amino acids
(aspartic acid, threonine and an unidentified amino acid) had similar correlation patterns; whereas
the correlation pattern for isoleucine was distinct. Unlike the other three amino acids, isoleucine
increased significantly in abundance in response to water-deficit with a more pronounced increase at
the mid-day time point (Supplemental Figure A.1). These results suggest that the regulatory control
of the metabolites with similar patterns of expression may be shared; whereas the metabolites with
distinct correlation patterns are likely influenced by distinct molecular mechanisms.
Among the transcripts significantly correlated with at least one metabolite, enrichment for GO
125
Figure 5.7 Box-plot illustrating the interplay of genotype and treatment in shaping the drought
metabolome and transcriptome of six P. balsamifera genotypes. The average absolute log2(fold-
change) between well-watered and water-deficit-treated samples for all (a) metabolites (n=40; P <
0.05) and (b) transcripts (n = 2636; P < 0.05) with significant variation in their abundance between
treatment conditions at the mid-day (MD) time point.
01
23
AP-947 AP-1005 AP-1006 AP-2278 AP-2298 AP-2300
Log 2(f
old-
chan
ge) t
rans
crip
t acc
umul
atio
n
Transcriptome MD
Metabolome MD
0.0
0.2
0.4
0.6
Log 2(f
old-
chan
ge) m
etab
olite
acc
umul
atio
n
0.8 (a)
(b)
126
−0.5
00.
5Pe
arso
n C
orre
latio
n C
oeffi
cien
t, r
Col
or K
ey
NI
NI
NI
NI
NI
NI
NI
NI
NI
NI
NI
NI
Aspa
rtic
Acid
NI (
Amin
o Ac
id; 2
)L-
Ther
onin
eL-
Isol
euci
neG
lyce
rol
NI (
5C S
ugar
; 2)
NI (
6C S
ugar
; 1)
Fruc
tose
(2)
Mel
ibio
seSu
cros
eR
affin
ose
Gly
colic
Aci
dFu
mar
ic A
cid
Thre
onic
aci
d 1,
4-la
cton
eM
alic
Aci
dTh
reon
ic a
cid
Mal
onic
Aci
dSu
ccin
ic A
cid
Qui
nic
Acid
Shik
imic
Aci
dC
itric
Aci
dBe
nzoi
c Ac
idSa
licyl
alc
ohol
Que
rciti
nSa
licin
Cat
echo
lM
yo-in
osito
lG
alac
tinol
127
Figure 5.8 Heatmap of drought responsive transcript and metabolite correlations. Of all the
drought responsive transcripts, 747 unique transcripts were correlated with at least one metabolite
(|r| > 0.6; P<0.05). The rows in the heatmap represent metabolites, and the columns represent
transcripts. The columns are clustered based on their expression across samples, and the metabolites
are grouped according to functional categories. Pearson correlation coefficient (r) are represented
for each pair-wise M:T comparison, and were calculated using R.
128terms among transcripts was determined. For transcripts with increased transcript abundance in
response to drought and correlated with at least one metabolite (n=404), four significant enriched
GO biological process terms were identified: ‘proline metabolic process’ (GO:0006560), ‘arginine
metabolic process’ (GO:0006525), ‘galactose metabolic process’ (GO: 0006012) and ‘serine family
amino acid metabolic process’ (GO:0009069). A total of 13 significant GO terms were identified
(for complete list see Supplementary Figure 9a). Among transcripts that had decreased transcript
abundance in response to drought and were correlated with at least one metabolite (n=343),
15 significantly enriched GO terms were identified (Supplementary Figure 9b). For GO terms
associated with biological process, ‘serine family amino acid metabolic process’ (GO:0009069),
‘tyrosine metabolic process’ (GO:0006570) and ‘aromatic amino acid family metabolic process’
(GO:0009072) were significantly enriched.
Functional annotation of the correlated transcripts and metabolites revealed pathways that were
perturbed by water withdrawal (Figure 5.9). A functional class related to starch and sucrose
metabolism (pop00500) was overrepresented among the transcripts that are correlated with two
identified 5C sugars and glucose (Figure 5.9). Photosynthesis-related categories (pop00195 and
pop 00196) were highly associated with malic acid, raffinose and galactinol (Figure 5.9).
In spinach, raffinose accumulation reduced electron and cyclic photophosphorylation in
photosynthesis (Santarius 1973), and it has been hypothesised that raffinose and other RFOs play
an important role in the protection of cellular metabolism, especially with respect to photosynthesis
in chloroplasts in Arabidopsis (Nishizawa et al. 2008). Evidence herein suggests there may be a
functional relationship in P. balsamifera between raffinose accumulation and transcripts associated
with photosynthesis. An association between photosynthetic metabolic processes and RFO
accumulation may highlight unique relationships that can be garnered from transcriptome-
metabolome relationships in Populus.
5.4.6 Energy metabolism and secondary metabolite biosynthesis varied in a
genotypic-dependant manner in response to drought
Galactinol accumulation varied in response to water-deficit stress in genotype AP-1006 [log2(fold-
change) = -0.4526]; whereas galactinol accumulated consistently in the other genotypes. Raffinose
accumulation was significant in water-deficit-treated plants, with the exception of trees of the
129
Phenylalanine, tyrosine and tryptophan biosynthesis
Amino sugar and nucleotide sugar metabolism
Steroid biosynthesis
Glycerolipid metabolism
Photosynthesis ~ antenna proteins
Plant hormone signal transduction
Starch and sucrose metabolism
Photosynthesis
Porphyrin and chlorophyll metabolism
Oxidative phosphorylation
biquinone and other terpenoid~quinone biosynthesis
Glycolysis / Gluconeogenesis
Phenylalanine metabolism
Ribosome
Ubiquitin mediated proteolysis
Arginine and proline metabolism
Galactose metabolism
Valine, leucine and isoleucine degradation
RNA transport
Cysteine and methionine metabolism
Purine metabolism
NI
Salic
inN
IN
I_Su
gar_
alco
hol
NI
NI
NI
Cat
echo
lN
IN
IN
IL.
Isol
euci
neC
atec
hin
Benz
oic
Acid
Mel
ibio
seN
I_5C
_sug
arAd
enos
ine
NI
Mal
tose
NI
NI
Sucr
ose
Succ
inic
_aci
dN
I_Am
ino_
acid
NI_
5C_s
ugar
Thre
onic
.aci
d.1.
4.la
cton
eM
alon
ic_a
cid
Qui
nic_
acid
Fum
aric
Aci
dN
I_5C
_sug
arN
I_5C
_sug
arL.
Thre
onin
eTh
ymid
ine
5 m
onop
hosh
ate
NI_
Org
anic
_aci
dL.
Tyro
sine
L.As
parti
c Ac
idL.
Alan
ine
Gly
cero
lN
IM
alic
_aci
dG
lyco
lic_a
cid
Thre
onic
_aci
dM
yo in
osito
lL.
Prol
ine_
2L.
Prol
ine_
1N
IN
IC
itric
Aci
dN
I_5C
_sug
arN
IN
IN
IN
IN
IN
ISa
licyl
_alc
ohol
Glu
cose
_2Ka
empf
erol
Que
rciti
nN
I_Am
ino_
acid
NI_
6C_s
ugar
NI
L.Ph
enyl
alan
ine
NI
Dig
alac
tosy
lgly
cero
lG
lyci
neL.
Glu
tam
ate
L.Se
rine
NI
L.As
corb
ic_a
cid
Buty
ric a
cid
4 am
ino
n-N
IN
IPy
rogl
utam
ic_a
cid
NI
NI
Fruc
tose
_1G
luco
se_1
Glu
cose
_3Sh
ikim
ic_a
cid
NI_
6C_s
ugar
Fruc
tose
2R
affin
ose
Gal
actin
ol
−0.5
00.
5Sp
earm
an R
ank
Cor
rela
tion
Col
or K
ey
130
Figure 5.9 A heatmap of representative functional classes (transcripts) from the correlation data.
The averaged Spearman correlation value is represented for significant functional class: metabolite
comparisons (coloured squares). Red indicates positive correlation, whereas blue indicates negative
correlation.
131genotype AP-2300. There was drought-responsive variation in transcript accumulation of genes
hypothesised to be involved in the galactose metabolism pathway. All genotypes showed increased
abundance of transcripts corresponding to galactinol synthase [EC:3.4.1.123], raffinose synthase
[EC:2.4.1.82] and stachyose synthase [EC: 2.4.1.67] (Figure 5.10). Galactinol synthase transcript
accumulation varied in magnitude in response to water-deficit conditions among the six genotypes,
with the largest increase in transcript accumulation observed in genotypes AP-2278 and AP-1006.
Elevated levels of RFOs in Arabidopsis plants increased drought tolerance, highlighting the
importance of these oligosaccharides in the response to osmotic-stress (Taji et al. 2002). Increased
accumulation of raffinose has been observed in desiccation tolerant seeds (Castillo et al. 1990),
chloroplasts of frost-hardy Brassica oleracea leaves (Santarius 1973), and in Populus tremula leaves
exposed to osmotic stress (Janz et al. 2010). Increased transcript abundance of galactinol synthase
and raffinose synthase has been observed in response to drought in Arabidopsis (Taji et al. 2002;
2004) and Populus (Janz et al. 2010; Hamanishi et al. 2010).
Mounting evidence suggests that the role of raffinose and other RFOs is consistent across species;
however, the magnitude of change is variable, as was observed among the six P. balsamifera
genotypes reported here. This suggests the existence of genotypic specific metabolite profile related
to these oligosaccharides, and that the level of accumulation may influence the overall drought
response. Moreover, the data suggest that AP-1006 may not produce elevated levels of galactinol
in response to drought, galactinol may be metabolised more quickly in AP-1006, or that the RFO
metabolites may not be as important for osmotic protection in certain genotypes.
Unique relationships were also observed in the citrate cycle (TCA) pathway (KEGG, pop00020).
TCA cycle intermediates, such as succinic and malic acid, were significant for T and TxG or
TxD interactions. The metabolic rate of the TCA cycle is known to be influenced by drought
(Rodríguez-Calcerrada et al. 2010). Notably, similar variations among genes involved in the TCA
metabolic pathway were observed in genotype AP-1006 (Figure 5.11a; for other genotypes, see
Supplemental Figure A.4). Pair-wise comparisons within the TCA cycle for select transcripts and
metabolites found weak relationships among transcripts and malic and citric acid accumulation
profiles for AP-1006 (Figure 5.11b); however, succinic acid and malate dehydrogenase
(EC:1.1.1.37) were significantly negatively correlated (r = -0.67, P = 0.0204) in genotype AP-1006
(Figure 5.11b).
132
EC 2.4.1.123
EC 2.4.1.82
EC 2.4.1.67
UDP-Galactose
Galactinol
Raffinose
Stachyose
Sucrose
myo-inositol
UDP
sucrose
myo-inositol
galactinol
myo-inositol
AP-9
47AP
-100
5AP
-100
6AP
-227
8AP
-229
8AP
-230
0
AP-9
47AP
-100
5AP
-100
6AP
-227
8AP
-229
8AP
-230
0
AP-9
47AP
-100
5AP
-100
6AP
-227
8AP
-229
8AP
-230
0
AP-9
47AP
-100
5AP
-100
6AP
-227
8AP
-229
8AP
-230
0
AP-9
47AP
-100
5AP
-100
6AP
-227
8AP
-229
8AP
-230
0
Figure 5.10 Pathway analysis related to the galactose metabolism. (a) Pathway map displays
selected steps from galactose metabolism pathway. Colours indicate fold-change in transcript
or metabolite abundance between water-deficit and well-watered treated samples for all six
genotypes; red indicates higher abundance in water-deficit-treated samples and blue indicates lower
abundance in water-deficit-treated samples. Enzymes are given as EC numbers. EC 2.4.1.123,
galactinol synthase; EC:2.4.1.82, raffinose synthase; EC:2.4.1.67, stachyose synthase. (b) Heatmap
representing Spearman correlation values among transcripts related to galactose metabolism and
raffinose or galactinol.
Raf
finos
e
Gal
actin
ol
EC:2.4.1.67
EC:3.4.1.123
EC:2.4.1.82
Spearman Correlation
Color Key
0.1 0.3 0.5 0.7
(a)
(b)
133
Citrate
Acetyl CoA
cis-Aconitate
Isocitrate
2-ketoglutarateSuccinyl-CoA
Succinate
Fumarate
Malate
Pyruvate
Oxaloacetate
4.2.1.2
1.3.5.1
glycolysis IV
1.1.1.37
6.2.1.5
1.1.1.41
2.3.3.1
4.2.1.3
−2 −1 0 1 2Fold Ratio [Log2(Fold-change)]
Color K ey Populus balsamifera, genotype AP-1006
Malic AcidCitric Acid Succinic Acid
EC
:1.3
.5.1
EC
:1.1
.1.3
7
EC
:4.2
.1.3
EC
:2.3
.3.1
EC
:1.1
.1.4
1
EC
:6.2
.1.5
EC
:4.2
.1.2
(a)
(b)
−1 −0.5 0 0.5 1Pearson Correlation Coefficient
Color Key
Figure 5.11 Pathway analysis related to the citric cycle (TCA cycle). (a) Correlation among select
transcripts and metabolites from the KEGG pathway pop00020 ‘Citrate cycle (TCA cycle)’ for
genotype AP-1006. Colors represent Pearson correlation value. Red indicates positive correlation
and blue represents negative correlation values. (b) Map displays selected steps from citrate cycle
pathway. Colours indicate fold-change in transcript or metabolite abundance between water-
deficit and well-watered treated samples for genotype AP-1006; red indicates higher abundance in
water-deficit-treated samples and blue indicates lower abundance in water-deficit-treated samples.
Enzymes are given as EC numbers. EC 1.1.1.37, malate dehydrogenase; EC:1.1.1.41, isocitrate
dehydrogenase (NAD+); EC:1.3.5.1, succinate dehydrogenase; EC:2.3.3.1, citrate synthase;
EC:5.2.1.2, fumarate hydratase, EC: 5.2.1.3, aconitate hydratase, EC: 6.2.1.5, succinate-CoA ligase,
beta subunit. Pearson correlation and pathway maps for other genotypes can be found in Appendix
Figure A.4.
134The magnitude change between well-watered and water-deficit-treated samples for transcripts
associated with the TCA cycle varied among genotypes. Citrate synthase (EC:2.3.3.1) had increased
transcript accumulation in water-deficit-treated samples of AP-947, AP-1006, AP-2278 and AP-
2300; however, decreased transcript accumulation was observed in the other genotypes. Similarly,
malate dehydrogenase (EC:1.1.1.37), had <1 log2(fold-change) in response to drought in AP-1006
and AP-2298, whereas >1 log2(fold-change) increase was observed in AP-947 and AP-2278. In
Arabidopsis, malate dehydrogenase demonstrated increased transcript accumulation in response
to drought, cold or high-salinity stress (Seki et al. 2007); however, the variation in the genotypic
response in P. balsamifera highlights the complexity in this response. This highlights the influence of
genotype on the drought-induced modifications to the TCA cycle in P. balsamifera.
Comparative pathway analysis among genotypes has proved useful in Populus. In two different
genotypes of Populus with varying salt-tolerance, pathway analysis revealed different mechanisms
of tolerance between the two genotypes. Janz et al. (2010) found that the salt-tolerant Populus
eupharatica demonstrated moderate transcriptome changes in response to stress when compared
to a salt-sensitive Populus hybrid. However, stress tolerance in P. eupharatica was not dependant
on transcriptome modification under conditions of stress; instead, it was linked to greater energy
requirements for cellular metabolism (Janz et al. 2010). In P. balsamifera it is evident that there are
varying degrees of transcriptional remodeling in response to drought among genotypes; however,
further analysis is required to understand the subtleties in these differences.
5.4.7 Network analysis illuminated the nature of genotype-specific responses to
drought
Some drought transcriptome alterations were specific to given genotype. To identify genotype-
specific transcriptome alterations, a network was created including all genes that were deemed
significantly expressed in a T-main effect manner for each genotype using weighted gene co-
expression network analysis (WGCNA; Langfelder & Horvath 2008). Weighted Pearson
correlation matrices were calculated and used to determine topological overlap (TO) among genes.
The TO calculated in WGCNA measured connectivity of a gene within a network relative to its
neighbours. HCA based on the TO scores for all genes in the drought transcriptome grouped genes
with equivalent transcript abundance profiles across all samples (Langfelder & Horvath 2008).
135Table 5.4 Module-treatment or -time of day relationships of the P. balsamifera (AP-1006) drought
transcriptome. Columns 2 and 3 represent the correlation between the mean expression of the
module and the experimental factor (T or D). Significant values are in bold (P-value < 0.05).
Module AP-1006 Treatment (T) Time of Day (D)M1_1006 -0 .833 -0 .484M2_1006 -0 .662 0 .709M3_1006 -0 .77 0 .432M4_1006 -0 .952 0 .164M5_1006 0 .759 -0 .638M6_1006 0 .983 -0 .038M7_1006 0 .928 0 .342M8_1006 0 .699 0 .708M9_947 0 .87 0 .025M10_2278 -0 .85 0 .145
136
(A) AP-947
(B) AP-1005
(C) AP-1006
(D) AP-2278
(E) AP-2298
(F) AP-2300
Figure 5.12 Transcript correlation networks obtained from WGCNA for (a) AP-947, (b) AP-1005,
(c) AP-1006, (d) AP-2278, (e) AP-2298 and (f ) AP-2300. The top 1000 interactions for each
genotype are represented. Nodes in the graphs represent individual transcripts that connect via
edges to other transcripts. Each node is colored according to the modules defined in Table 5.3.
Colour Module yellow M1_1006blue M2_1006black M3_1006turquoise M4_1006brown M5_1006green M6_1006pink M7_1006red M8_1006magenta M9-947purple M10_2278
137Overall, 10 network modules with equivalent transcript abundance patterns were identified.
Many network modules were similar across genotypes - 80% of the network modules significantly
overlapped, with respect to their gene membership, in at least two genotypes (Table 5.4). Not
surprisingly, all of the modules were highly correlated with treatment; whereas only three were
significantly correlated with time of day (M2, M5 and M8; Table 5.4). Notably, M3 was only
shared between AP-1006 and AP-2278, whereas M9_947 and M10_2278 were unique to AP-947
and AP-2278, respectively. M3 (black) had a 62% overlap with respect to gene membership, and
among those transcripts, there was an overrepresentation of transcripts involved in ‘intracellular
signalling cascade (GO: 0007242)’. M5 (brown) demonstrated a high degree of overlap among
genotypes, with the exception of AP-2298. M5 was functionally characterised by a general drought
response that is similar to the overall drought response, with overrepresented GO terms which
include: ‘response to abiotic stimulus (GO:0009628)’, ‘cellular catabolic process (GO: 0044248)’
and ‘response to water deprivation (GO: 0009414)’. The high degree of overlap between modules
identified in each genotype validated the presence of a highly conserved drought transcriptome in P.
balsamifera.
Although there was a high degree of network module preservation among genotypes (Table 5.4),
organisation within modules varied among genotypes. When visualising the top (n=1000) network
connections of each genotype, and labeling the nodes according to their module membership within
the drought transcriptome, two general observations could be made (Figure 5.12). First, transcript
connectivity varied among genotypes. In certain genotypes, there was a higher degree of topological
overlap between individual genes (nodes), as indicated by the colour of the edges (higher TO
indicated with a red/purple colour, whereas lower TO is indicated with blue). For example, a higher
degree of TO was observed in genotype AP-1005 as compared to genotype AP-2278. Second, the
importance of any given module varied among genotypes. For example, the nodes of top network
connections in genotype AP-1005 were from module M4_1006 and M6_1006, whereas genotype
AP-947 had nodes that belonged to many other modules. More specifically, genes that played a
more central (“hub”) role in the drought transcriptome networks varied among genotypes.
5.4.8 AP-1006 had a genotype-specific transcriptome response to drought
To interrogate the uniqueness of the transcriptome of AP-1006, genes central to the network
modules in this genotype were identified. Transcript abundance of hub genes in AP-1006 revealed
138
(a)
GO:0044042 (0.00523)glucan metabolic
proces s8/149 | 329/36285
GO:0006073 (0.00314)cellular glucan
metabolic proces s8/149 | 298/36285
GO:0044238 primary metabolic
proces s
GO:0005975 carbohydrate metabolic
proces s
GO:0044264 (0.0357)cellular polys accharide
metabolic proces s8/149 | 457/36285
GO:0005976 (0.0234)polys accharide metabolic
proces s10/149 | 639/36285
GO:0009311 (0.00314)oligos accharide metabolic
proces s6/149 | 151/36285
GO:0044262 cellular carbohydrate
metabolic proces s
GO:0005982 (0.000229)s tarch metabolic
proces s6/149 | 79/36285
GO:0044260 cellular macromolecule
metabolic proces s
GO:0005984 (0.000582)dis accharide metabolic
proces s6/149 | 101/36285
GO:0005985 (5.82e-05)s ucros e metabolic
proces s6/149 | 56/36285
GO:0009987 cellular proces s
GO:0044237 cellular metabolic
proces s
GO:0043170 macromolecule metabolic
proces s
GO:0008150 biological_proces s
GO:0008152 metabolic proces s
GO:0016137 (0.000582)glycos ide metabolic
proces s7/149 | 159/36285
139
Figure 5.13 Overrepresentation of GO terms associated with transcripts that have (a) decreased
transcript abundance in AP-1006 and (b) increased transcript abundance in AP-1006. Figures
generated using AgriGO (http://bioinfo.cau.edu.cn/agriGO). Significant overrepresentation is
represented by darker coloured boxes (P < 0.05).
GO
:001
6137
(0.0
0109
)gl
ycos
ide
met
abol
icpr
oces
s8/
218
| 159
/362
85
GO
:000
5985
(3.3
8e-0
5)su
cros
e m
etab
olic
proc
ess
7/21
8 | 5
6/36
285
GO
:000
9311
(0.0
0538
)ol
igos
acch
arid
e m
etab
olic
proc
ess
7/21
8 | 1
51/3
6285
GO
:000
5984
(0.0
0065
5)di
sacc
harid
e m
etab
olic
proc
ess
7/21
8 | 1
01/3
6285
GO
:001
9752
ca
rbox
ylic
aci
dm
etab
olic
pro
cess
GO
:000
6520
ce
llula
r am
ino
acid
met
abol
ic p
roce
ss
GO
:003
4641
ce
llula
r nitr
ogen
com
poun
d m
etab
olic
pro
cess
GO
:004
4106
ce
llula
r am
ine
met
abol
ic p
roce
ss
GO
:000
6807
ni
troge
n co
mpo
und
met
abol
ic p
roce
ss
GO
:000
9308
am
ine
met
abol
icpr
oces
s
GO
:004
4281
sm
all m
olec
ule
met
abol
ic p
roce
ss
GO
:004
2180
ce
llula
r ket
one
met
abol
ic p
roce
ss
GO
:000
6082
or
gani
c ac
idm
etab
olic
pro
cess
GO
:000
6519
ce
llula
r am
ino
acid
and
der
ivat
ive
met
abol
ic p
roce
ss
GO
:000
9069
(0.0
268)
serin
e fa
mily
amin
o ac
id m
etab
olic
pro
cess
5/21
8 | 1
00/3
6285
GO
:004
4264
ce
llula
r pol
ysac
char
ide
met
abol
ic p
roce
ss
GO
:000
6073
(0.0
111)
cellu
lar g
luca
nm
etab
olic
pro
cess
9/21
8 | 2
98/3
6285
GO
:004
4262
ce
llula
r car
bohy
drat
em
etab
olic
pro
cess
GO
:004
4260
ce
llula
r mac
rom
olec
ule
met
abol
ic p
roce
ss
GO
:000
5982
(0.0
0018
6)st
arch
met
abol
icpr
oces
s7/
218
| 79/
3628
5
GO
:000
9072
(0.0
268)
arom
atic
am
ino
acid
fam
ily m
etab
olic
pro
cess
5/21
8 | 1
01/3
6285
GO
:000
9987
ce
llula
r pro
cess
GO
:004
4237
ce
llula
r met
abol
icpr
oces
s
GO
:000
6725
ce
llula
r aro
mat
icco
mpo
und
met
abol
ic p
roce
ss
GO
:000
8150
bi
olog
ical
_pro
cess GO
:000
8152
m
etab
olic
pro
cess
GO
:004
4238
pr
imar
y m
etab
olic
proc
ess
GO
:004
3170
m
acro
mol
ecul
e m
etab
olic
proc
ess
GO
:004
3436
ox
oaci
d m
etab
olic
proc
ess
GO
:004
4042
(0.0
197)
gluc
an m
etab
olic
proc
ess
9/21
8 | 3
29/3
6285
GO
:000
5975
ca
rboh
ydra
te m
etab
olic
proc
ess
GO
:000
5976
po
lysa
ccha
ride
met
abol
icpr
oces
s
(b)
140samples clustered according to treatment; however, greater transcript abundance profiles were more
similar between well-watered samples, regardless of time of day. The absolute magnitude change in
abundance of transcripts central to the network modules in AP-1006 was significantly higher than
the magnitude change of the transcripts in the other genotypes [absolute log2(fold-change) AP-
1006 = 2.36]. Many hub transcripts had significant changes in transcript abundance in response to
drought in AP-1006. There are 195 hub transcripts (TO Network Ratio > 0.5) that have decreased
abundance in response to drought in AP-1006; whereas there are 104 hub transcripts that had
increased abundance in response to drought (Table 5.5).
Enrichment of GO terms within the set of central network hub transcripts from genotype AP-
1006 revealed the components of the genotype-specific drought transcriptome. For example,
transcripts implicated in the response to stress and stimulus were enriched. Of the hub transcripts
with significant declines in abundance, genes implicated in carbohydrate metabolism were
enriched, including those with GO terms for: sucrose (GO:0005985), starch (GO: 0005982) and
disaccharide (GO: 0005984) metabolic processes (Figure 5.9a), among others. Conversely, core
hub transcripts with increased accumulation in response to drought in AP1006 were enriched for
biological processes, including response to stimulus (GO:0050896) and stress (GO: 0006950) as
well as transport (GO:0006910) and regulation of cellular processes (GO:0050794; Figure 5.13b);
however, it should be noted that there was a large proportion of transcripts that had unknown
function. The transcripts that played a central role in the network organisation of the drought
transcriptome in AP-1006 were likely important regulators of the drought response, and the analysis
of transcript co-expression relationships may help with functional annotation; albeit, not with
immediate interpretation.
5.4.9 There were strong correlates between specific transcript-metabolite pairs in response to drought in AP-1006
Pathways with significant correlations between metabolites and transcripts that had significant
differences in abundance in response to drought in AP-1006 included ‘plant hormone signal
transduction’, ‘arginine and proline metabolism’, and ‘glycolysis/gluconeogenesis’. As previously
noted, the largest magnitude change in transcript abundance was observed in AP-1006 (Figure
5.7). Transcripts, including those encoding genes homologues to Arabidopsis thaliana RAC-like 2
protein (ARAC2) and IRREGULAR XYLEM 9 (IRX9) had significantly larger fold-change decrease
141in transcript abundance in response to water-deficit conditions as compared to other genotypes.
Conversely, several transcripts annotated as universal stress proteins or those involved in hormone
signalling had significantly higher transcript abundance in AP-1006 in response to water-deficit
stress. Correlation network analysis revealed core transcripts that might have played a role in the
underlying mechanisms regulating metabolite accumulation in AP-1006. Transcripts most strongly
correlated with metabolite levels were identified (Supplementary Figure 10). Although no particular
class of metabolites or transcripts appeared specific to AP-1006, a large number of transcripts highly
correlated with succinic acid, raffinose and galactinol accumulation were observed (Supplementary
Figure 10). For example, strong positive correlations were observed between raffinose, galactinol
and a photosystem II reaction center PsbP family protein (r = 0871 and 0.835, respectively;
Supplemental 5.6). Strong correlations between drought responsive metabolites and transcripts
reveal pathways that may be of importance in the drought tolerance mechanisms in a genotype.
5.5 Conclusion
The complexity of the metabolomic response to drought in Populus balsamifera was highlighted
by variation among genotypes and between time-of-day responses. Although common drought-
responsive metabolites could be identified across all six P. balsamifera genotypes, a significant
proportion of metabolites varied in a genotype or time-of-day dependant manner. Notably, the
magnitude of drought-induced changes in the metabolite abundance among the six P. balsamifera
genotypes varied significantly. The complexity of the genotype-metabolite relationship was notable,
and likely attributable to the function of many genes, the environment and their interaction.
Integrating transcriptome-and metabolome data identified significant metabolite-gene correlation
whereby biologically meaningful correlations were derived. Metabolite-transcript relationships
from the same and different pathways were identified, and may be useful for future elucidation of
important drought response mechanisms. Integration of the transcriptome and metabolome data at
individual pathway levels revealed variation in metabolite flux and transcript accumulation among
genotypes in energy and galactose metabolism.
142
Chapter 6: Conclusion and Future Directions
143
Chapter 6: General Conclusions and Future Directions
This thesis investigated the intra-specific variation in the Populus balsamifera drought response.
Specifically, the research addressed how transcriptome responses varied among P. balsamifera
genotypes, and how growth, development and metabolism were remodeled in response to drought
stress. The results represent significant contributions to our knowledge about intraspecific variation
within the Populus balsamifera drought responses.
1.1 Major Findings and Significance
This thesis comprised a test of three overarching hypotheses. Those hypotheses and the major
findings that were derived in testing them are reported below.
1. There are significant differences in the transcriptomes of Populus balsamifera trees in response
to drought stress.
The Populus drought transcriptome is a highly dynamic and complex system in which genetic and
environmental cues combine, resulting in various tolerance mechanisms and adaptations that allow
tree to contend with drought stress. A common, shared water-deficit induced transcriptome-level
response for Populus balsamifera was identified using the Affymetrix Poplar GeneChip microarray
platform; however, the amplitude of gene expression varied significantly among genotypes. This
highlights the importance of genotype in shaping the drought-response among Populus species
and genotypes. Selection of single genotypes or hybrid clones for gene expression studies should
proceed with caution, as there can be notable difference between genotypes, and a single genotype
or clone may not be representative of the species in question. Future studies could aim to study
extensive cohorts of genotypes in order to grasp the diversity in responses both within a species, and
among individuals of the same genus. Variation in the physiological and morphological responses
to drought among trees of the genus Populus is frequently observed (Yin et al. 2004; Monclus et
al. 2006; Giovannelli et al. 2007), and the variation in gene expression could be exploited to study
the various mechanisms that underpinning responses among individuals of the genus Populus.
Furthermore, there is a long-term goal of determining how these molecular mechanisms underpin
the drought phenotype in Populus. Genome-wide gene expression data can be used to generate
gene networks, with the hopes of understanding the gene regulatory networks that may explain
144the genetic causes of a given phenotype, or in this case, a drought response. The data generate
in chapter 3 can be used to begin the search for these underlying genetic regulatory networks in
P. balsamifera. Future studies may also begin to integrate DNA structural variants (e.g., single
nucleotide polymorphisms, or SNPs) and gene expression data in order to better understand how
SNPs and genes influence one another, as well as predicting phenotype (Chang and McGeachie
2011).
Physiological and phenotypic drought responses in P. balsamifera also varied significantly among
genotypes, where certain genotypes rapidly modified physiological status at the onset of water-
deficit stress compared to other genotypes that had gradual declines in physiological status.
Notably, phenotypic traits, such as growth, correlated with genetic responsiveness to drought.
Among the six genotypes reported in Chapter 3, genotypes that had increased magnitude change
in their drought responsive transcriptome sustained growth under conditions water-limitation.
Although the sample size for this relationship was limited, the evidence presented herein suggests
a relationship that warrants future investigation. Future studies could be directed at sampling
larger numbers of Populus genotypes with variable growth rates in response to drought. Assessing
variation in their drought transcriptomes and comparing responsiveness to the ability to sustain
growth under drought conditions could lead to better insights into the molecular mechanisms that
define growth under stress conditions in Populus. Specifically, future experiments could test whether
a larger magnitude change in the transcriptome buffers the plant against the negative growth and
development impacts of drought-stress. This would be important knowledge for the maintenance
and improvement of productivity of hybrid poplars in bio-fuel crops, for example, under conditions
of changing climate.
In chapter 3, genotypes with a higher degree of similarity among their drought transcriptomes
also had fewer single feature polymorphism (SFP) differences, suggesting that individuals sharing
a greater degree of genotypic congruence may have conserved drought responses. However, there
was no correspondence between SFP differences and geographic origin. This suggests that some
genotype specific responses may be locally adapted, although others may be spread widely across the
landscape. Moreover, the population structure of P. balsamifera is such that there are three distinct
sub-populations, or demes (Keller et al. 2010). The research presented herein was performed on
genotypes that originated from a single deme; future research should consider sampling across the
145entirety of the range. Sampling within multiple sub-populations across the range would provide
further insight into the genetic and transcriptomic variation within and among different demes
across the range of P. balsamifera.
2. Drought-induced modification of the transcription of genes implicated in the stomatal
development regulatory network are linked to changes in stomatal density.
When two genotypes (AP-1005 and AP-1006) of P. balsamifera were grown under water-deficit
conditions, a significant reduction in stomatal index was observed in leaves that developed under
water-deficit stress, as compared to those that developed under well-watered conditions. Alteration
to the stomatal development program in Populus may be indicative of a long-term strategy to
minimize water loss and contend with drought-stress. Reductions in stomatal conductance was
also observed in P. balsamifera, with the greatest reduction observed in AP-1006, which was also the
genotype that had the largest decline in stomatal index in water-deficit treated samples.
Quantification of transcript abundance of genes hypothesised to be involved in the stomatal
development pathway in Populus were interrogated throughout development, and specific genes
demonstrated transcript abundance profiles congruent with their hypothesised role in stomatal
development. For example, STOMAGEN, a positive regulator of stomatal development had
significantly higher transcript abundance in well-watered samples early in development (day 5 and
10). Other genes, such as ERECTA and STOMATAL DENSITY and DISTRIBUTION 1 had
variable transcript accumulation between genotypes. These findings suggest that there may be
variable drought-response strategies amongst genotypes.
Modifications of stomatal development under conditions of changing environments may represent
long-term water-deficit tolerance strategies. Although limiting water loss in conditions of water-
deficit stress would likely be beneficial in the short term, a reversion to non-drought conditions
may result in suboptimal conditions for the plant in question. More specifically, a tree grown under
water deficit stress with fewer stomatal pores and reduced water-loss during drought will also have
limited capacity for gas exchange under well watered conditions as compared to an individual that
did not have reduced stomatal development under conditions of water-deficit stress.
As stomatal development is regulated by both environmental and endogenous factors, such as ABA,
it is also important to consider the former. In Arabidopsis thaliana cotyledons ABA plays a role in
146both plant physiology and development under water-deficit conditions (Tanaka et al. 2013). ABA
regulation of stomatal development is upstream of both SPCH and MUTE; however, the influence
of ABA on stomatal development is dependent on the presence of stomatal lineage cells. More
specifically, ABA limits stomatal initiation and the enlargement of pavement cells (Tanaka et al.
2013). Future studies in Populus should consider the relationship between ABA and genes known
to regulate stomatal development, such as SPCH and MUTE.
Understanding the variation in the developmental responses to drought-stress in Populus may lead
to a better ability to select or breed plants with improved drought tolerance. Future interrogations
of the stomatal development pathway in Populus will likely improve our knowledge surrounding
the evolution of gene networks underpinning development in various plant species, and the role
stomatal development plays in long-term drought tolerance. The molecular underpinnings of
stomatal development in Arabidopsis are relatively well characterised compared to other plant
species, and experimental evidence in Arabidopsis supports that alterations in stomatal density results
in more drought tolerant individuals (Yoo et al. 2010). For example, Yu et al. (2008) overexpressed
the cDNA of a key transcriptional regulator in transgenic tobacco which lead to increased drought
tolerance associated with decreased stomatal density. In Populus, by understanding the molecular
mechanisms and indentifying the key regulators of stomatal development, transgenic Poplars with
increased drought tolerance can be created.
3. A Populus balsamifera drought metabolome can be identified, and there are transcript-
metabolite relationships that vary in a genotype-dependent manner.
In order to understand the changes that occur at the metabolic level in Populus, the perturbation
of the soluble metabolome was assessed in six genotypes of P. balsamifera under well watered and
water-deficit conditions. In chapter 5, metabolites that had significant differences among genotypes,
treatment and time-of-day were identified. Metabolites from a broad range of functional groups
were included, such as: amino acids (e.g., proline and isoleucine), representatives from the citric
acid (TCA) cycle (e.g., succinic and malic acid), photorespiration (e.g., glycolic acid), phenolic
species (e.g., catechin) and diverse metabolites thought to have roles in osmotic adjustment (e.g.,
galactinol and raffinose). Specifically, TCA cycle intermediates malic and succinic acid had reduced
abundance under conditions of water deficit stress and both exhibited variability among genotypes.
147Assessing the impact of water-deficit stress on metabolite levels is complicated by the involvement
of a given metabolite in multiple pathways. For example, succinic acid is a constituent of the
TCA cycle and also plays a role in γ-aminobutyric acid (GABA) metabolism. Under periods of
drought-stress citric acid metabolism is hypothesized to decrease, whereas GABA metabolism is
thought be up regulated (Allan et al. 2008). Raffinose and galactinol had increased abundance in
response to drought. A significantly larger fold-change increase in raffinose accumulation occurred
at mid-day, whereas galactinol accumulation exhibited variability among genotypes. Both raffinose
and galactinol are thought to play important roles as osmoprotectants in plants under drought
stress, and it has been suggested that raffinose plays an important role protecting the membranes of
chloroplasts (Santarius 1973).
The non-targeted metabolome was compared to the microarray-based transcriptome profiles of six
P. balsamifera genotypes. A total of 747 metabolites were significantly correlated with at least one
metabolite. Notably, transcripts functionally annotated to photosynthesis related categories were
highly associated with malic acid, raffinose and galactinol. The correlation between metabolite and
transcript levels identified transcripts that may be influenced by particular metabolites or indentified
future targets for analysis of metabolites that regulate gene expression under conditions of water-
deficit stress.
Pathway analysis comparing transcript and metabolite abundance reveals variation in the metabolite
flux and gene expression among genotypes. Galactose metabolism is hypothesised to be impacted
by water-deficit stress. Genotypic variability in metabolite abundance and gene expression was
observed. Increased transcript abundance for galactinol synthase was not always congruent with
increased galactinol abundance among genotypes. Variation in the TCA cycle is also notable
among genotypes. This suggests that the genotypic influence on both metabolite and transcript
accumulation in response to drought is a highly complex and dynamic process. Along these lines,
focus on other specific drought-related metabolic pathways may provide insight into the drought
responses among genotyeps of Populus. For example, ABA is known to play an important role in
signalling in response to drought in plants (Shinozaki & Yamaguchi-Shinozaki 1996). Focused
analysis of metabolites involved in the ABA biosynthetic pathways, including precursors such as
zeaxanthin, may identify differences in the biosynthesis and metabolism of an important drought-
related plant horomone.
148Although the number of metabolites analysed in this chapter were limited, they provided insight
into the variation of the drought metabolome in P. balsamifera and provided evidence for treatment,
genotype and time-of day variations in the metabolome. Intraspecific variation at the metabolite-
and transcript-level was underscored by diverse patterns of accumulation. Interestingly, genotype
AP-1006 demonstrated notable differences between the magnitude changes observed at the
transcriptome as compared to its metabolome. To this end, genotype AP-1006 may represent a case
study for diverse metabolite-transcript relationships as compared to the other five genotypes studied
in chapter 5.
Integration of multiple ‘omics platforms is an incredibly powerful tool to assess the interactions
between metabolites and gene expression. A systems biology approach allows a better view of
the dynamics and the complex relationships that occur under drought-stress. However, future
studies in Populus would be aided by the inclusions of an increased number of metabolites, and
the identification of unknown metabolites. Alternative methods of metabolite profiling could be
deployed in order to broaden the characterization of the drought metabolome.
The complexities of the drought response in Populus is tremendous. In order to gain a better
understanding of the dynamics of the response at the molecular level, future studies may wish to
include comprehensive time-series experimental set-up. As previously shown, the Populus drought
transcriptome is shaped by time-of day (Wilkins et al. 2009a), and many metabolites demonstrated
time-of-day variation. Both the metabolome and transcriptome analyses capture the abundance
levels at a specific time; these analyses provide no insight into the flux of metabolites or transcripts
that may occur before or after sampling. A time-series experimental set-up may begin to shed
light into these dynamic fluxes, and highlight relationships between changes in metabolite or gene
expression levels, rather than simply analyzing abundance levels at a given time-point.
The results presented herein were collected from potted grown grown seedlings in a climate-
controlled environment. Extrapolating results to mature, field-grown trees is difficult due to many
confounding factors. These include the impacts of age-dependant variation in physiology; variable
biotic and abiotic stimuli; and, tree-to-tree competition. Although the experiments presented in
this thesis do not take into consideration the combined effects of the various stimuli that a tree
could encounter in its natural environment, the experiments enabled the controlled manipulation of
water-deficit stress for a large number of poplar seedlings. Future experiments could be established
149to explore the intraspecific variation in the molecular underpinnings of the drought responses in P.
balsamifera under field-grown conditions in both young and mature trees in order to understand the
differences attributable to tree age, as well as growth conditions.
The impacts of environmental stress on forest health and productivity are becoming of increasing
concern. The results presented herein demonstrate that future experiments aimed at understanding
the complexities of the responses of forest trees to environmental stimuli must take into
consideration the intraspecific variation in these responses. Although common drought responses
among genotypes of P. balsamifera could be identified, significant intraspecific variation was
observed. Genotype shapes the genome-wide transcriptome and metabolome responses to water-
deficit stress, as well as the modulation of stomatal development in P. balsamifera. The intraspecific
variation in the molecular strategies that underpin the responses to drought among genotypes
may have an important role in the maintenance of forest health and productivity amidst future
predictions of changing climate.
150
Abe H., Urao T., Ito T., Seki M., Shinozaki K., Yamaguchi-Shinozaki K. (2003) Arabidopsis AtMYC2 (bHLH) and AtMYB2 (MYB) function as transcriptional activators in abscisic acids signaling. The Plant Cell 15:63–78.
Abe H., Yamaguchi-Shinozaki K., Urao T., Iwasaki T., Hosokawa D., Shinozaki K. (1997) Role of Arabidopsis MYC and MYB homologs in drought-and abscisic acid-regulated gene expression. The Plant Cell 9:1859–1868.
Adams H.D., Macalady A.K., Breshears D.D., Allen C.D. (2010) Climate-induced tree mortality: earth system consequences. Eos
Allan W.L., Simpson J.P., Clark S.M., Shelp B.J. (2008) Gamma-hydroxybutyrate accumulation in Arabidopsis and tobacco plants is a general response to abiotic stress: putative regulation by redox balance and glyoxylate reductase isoforms. Journal of Experimental Botany 59:2555–2564.
Allen C.D., Breshears D.D. (1998) Drought-induced shift of a forest-woodland ecotone: rapid landscape response to climate variation. Proceedings of the National Academy of Sciences of the United States of America 95:14839–14842.
Allen C.D., Macalady A.K., Chenchouni H., Bachelet D., McDowell N., Vennetier M., Kitzberger T., Rigling A., Breshears D.D., Hogg E.H.T., Gonzalez P., Fensham R., Zhang Z., Castro J., Demidova N., Lim J.-H., Allard G., Running S.W., Semerci A., Cobb N. (2010) A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. Forest Ecology and Management 259:660–684.
Allona I., Quinn M., Shoop E., Swope K., Cyr S.S., Carlis J., Riedl J., Retzel E., Campbell M.M., Sederoff R.R., Whetten R.W. (1998) Analysis of xylem formation in pine by cDNA sequencing. Proceedings of the National Academy of Sciences of the United States of America 95:9693–9698.
Almeida-Rodriguez A.M., Cooke J.E.K., Yeh F., Zwiazek J.J. (2010) Functional characterization of drought-responsive aquaporins in Populus balsamifera and Populus simonii × balsamifera
Literature Cited
151clones with different drought resistance strategies. Physiologia Plantarum 140:321–333.
Bahrun A., Jensen C.R., Asch F., Mogensen V.O. (2002) Drought-induced changes in xylem pH, ionic composition, and ABA concentration act as early signals in field-grown maize (Zea mays L.). Journal of Experimental Biology 53:251–263.
Barchet G. (2010) Effects of drought stress on metabolite profiles of hybrid and pure lines of Populus spp. University of British Columbia, Vancouver, BC
Barrs H.D., Weatherley P.E. (1962) A re-examination of the relative turgidity technique for estimating the water deficits in leaves. Australian Journal of Biological Sciences 15: 413:428.
Bartels D., Sunkar R. (2005) Drought and salt tolerance in plants. Critical Reviews in Plant Sciences 24:23–58.
Beikircher B., Mayr S. (2009) Intraspecific differences in drought tolerance and acclimation in hydraulics of Ligustrum vulgare and Viburnum lantana. Tree Physiology 29:765–775.
Belin C., Thomine S., Schroeder J.I. (2010) Water balance and the regulation of stomatal movements. In: Pareek A, Sopory SK, Bohnert HJ, Govindjee (eds) Abiotic stress adaptation in plants. Springer Netherlands, pp 283–305.
Benjamini Y., Hochberg Y. (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society, Series B 57:289–300.
Berardini T.Z., Mundodi S., Reiser L., Huala E., Garcia-Hernandez M., Zhang P., Mueller L.A., Yoon J., Doyle A., Lander G., Moseyko N., Yoo D., Xu I., Zoeckler B., Montoya M., Miller N., Weems D., Rhee S.Y. (2004) Functional annotation of the Arabidopsis genome using controlled vocabularies. Plant Physiology 135:745–755.
Berger D., Altmann T. (2000) A subtilisin-like serine protease involved in the regulation of stomatal density and distribution in Arabidopsis thaliana. Genes & Development 14:1119–1131.
Bergmann D., Sack F.D. (2007) Stomatal development. Annual Review of Plant Biology 58:163–
152181.
Bergmann D.C., Lukowitz W., Somerville C.R. (2004) Stomatal development and pattern controlled by a MAPKK Kinase. Science 304:1494–1497.
Bernier J., Serraj R., Kumar A., Venuprasad R., Impa S., R P V.G., Oane R., Spaner D., Atlin G. (2009) The large-effect drought-resistance QTL qtl12.1 increases water uptake in upland rice. Field Crops Research 110:139–146.
Berta M., Giovannelli A., Sebastiani F., Camussi A., Racchi M.L. (2010) Transcriptome changes in the cambial region of poplar Populus alba L. in response to water deficit. Plant Biology 12:341–354.
Bigler C., Bräker O., Bugmann H., Dobbertin M., Rigling A. (2006) Drought as an inciting mortality factor in scots pine stands of the Valais, Switzerland. Ecosystems 9:330–343.
Blodner C., Majcherczyk A., Kues U., Polle A. (2007) Early drought-induced changes to the needle proteome of Norway spruce. Tree Physiology 27:1423–1431.
Bogeat-Triboulot M.-B., Brosche M., Renaut J., Jouve L., Le Thiec D., Fayyaz P., Vinocur B., Witters E., Laukens K., Teichmann T., Altman A., Hausman J.-F., Polle A., Kangasjarvi J., Dreyer E. (2007) Gradual soil water depletion results in reversible changes of gene expression, protein profiles, ecophysiology, and growth performance in Populus euphratica, a poplar growing in arid regions. Plant Physiology 143:876–892.
Bonan G.B. (2008) Forests and climate change: forcings, feedbacks, and the climate benefits of forests. Science 320:1444–1449.
Bonhomme L., Monclus R., Vincent D., Carpin S., Claverol S., Lomenech A.-M., Labas V., Plomion C., Brignolas F., Morabito D. (2009) Genetic variation and drought response in two Populus × euramericana genotypes through 2-DE proteomic analysis of leaves from field and glasshouse cultivated plants. Phytochemistry 70:988–1002.
Boyko A., Kovalchuk I. (2008) Epigenetic control of plant stress response. Environmental and Molecular Mutagenesis 49:61–72.
153Bray E.A. (1994) Stress Induced Gene Expression in Plant. In: Basra AS (ed) Hardwood Academic
Publishers, pp 1–23.
Brendel O., Le Thiec D., Scotti-Saintagne C., Bodénès C., Kremer A., Guehl J.-M. (2008) Quantitative trait loci controlling water use efficiency and related traits in Quercus robur L. Tree Genetics & Genomes 4:263–278.
Breshears D.D., Cobb N.S., Rich P.M., Price K.P., Allen C.D., Balice R.G., Romme W.H., Kastens J.H., Floyd M.L., Belnap J., Anderson J.J., Myers O.B., Meyer C.W. (2005) Regional vegetation die-off in response to global-change-type drought. Proceedings of the National Academy of Sciences of the United States of America 102:15144–15148.
Brosche M., Vinocur B., Alatalo E., Lamminmaki A., Teichmann T., Ottow E.A., Djilianov D., Afif D., Bogeat-Triboulot M.-B., Altman A., Polle A., Dreyer E., Rudd S., Paulin L., Auvinen P., Kangasjarvi J. (2005) Gene expression and metabolite profiling of Populus euphratica growing in the Negev desert. Genome Biology 6:R101.
Brown G.R., Kadel E.E., Bassoni D.L., Kiehne K.L., Temesgen B., van Buijtenen J.P., Sewell M.M., Marshall K.A., Neale D.B. (2001) Anchored reference loci in loblolly pine (Pinus taeda L.) for integrating pine genomics. Genetics 159:799–809.
Brunner A.M., Busov V.B., Strauss S.H. (2004) Poplar genome sequence: functional genomics in an ecologically dominant plant species. Trends in Plant Science 9:49–56.
Campalans A., Pages M., Messeguer R. (2001) Identification of differentially expressed genes by the cDNA-AFLP technique during dehydration of almond (Prunus amygdalus). Tree Physiology 21:633–643.
Caruso A., Morabito D., Delmotte F., Kahlem G., Carpin S. (2002) Dehydrin induction during drought and osmotic stress in Populus. Plant Physiology and Biochemistry 40:1033–1042.
Casson S.A., Franklin K.A., Gray J.E., Grierson C.S., Whitelam G.C., Hetherington A.M. (2009) phytochrome B and PIF4 Regulate Stomatal Development in Response to Light Quantity. Current Biology 19:229–234.
154Casson S.A., Hetherington A.M. (2010) Environmental regulation of stomatal development.
Current Opinion in Plant Biology 13:90–95.
Castillo E.M., De Lumen B.O., Reyes P.S. (1990) Raffinose synthase and galactinol synthase in developing seeds and leaves of legumes. Journal of agricultural 38: 351–355.
Ceulemans R., van Praet L., Jiang X.N. (1995) Effects of CO2 enrichment, leaf position and clone on stomatal index and epidermal cell density in Poplar (Populus). New Phytologist 131:99–107.
Chang H.-H., McGeachie M. (2011) Phenotype prediction by integrative network analysis of SNP and gene expression microarrays. Conference proceedings: Annual International Conference of the IEEE Engineering in Medicine and Biology Society IEEE Engineering in Medicine and Biology Society Conference 2011:6849–6852.
Chang S., Puryear J.D., Cairney J. (1993) A simple and efficient method for isolating RNA from Pine trees. Plant Mol Biology Rep 11:113–116.
Chang S., Puryear J.D., Dias M.A.D.L., Funkhouser E.A., Newton R.J., Cairney J. (1996) Gene expression under water deficit in loblolly pine (Pinus taeda): Isolation and characterization of cDNA clones. Physiologia Plantarum 97:139–148.
Chaves M.M., Maroco J.P., Pereira J.S. (2003) Understanding plant responses to drought - from genes to the whole plant. Functional Plant Biology 30:239–264.
Chen S., Wang S., Altman A., Huttermann A. (1997) Genotypic variation in drought tolerance of poplar in relation to abscisic acid. Tree Physiology 17:797–803.
Chinnusamy V., Zhu J.-K. (2009) Epigenetic regulation of stress responses in plants. Current Opinion in Plant Biology 12:133–139.
Choi H.-I., Hong J.-H., Ha J.-O., Kang J.-Y., Kim S.Y. (2000) ABFs, a Family of ABA-responsive Element Binding Factors. Journal of Biological Chemistry 275:1723–1730.
155Ciais P., Reichstein M., Viovy N., Granier A., Ogee J., Allard V., Aubinet M., Buchmann N.,
Bernhofer C., Carrara A., Chevallier F., De Noblet N., Friend A.D., Friedlingstein P., Grunwald T., Heinesch B., Keronen P., Knohl A., Krinner G., Loustau D., Manca G., Matteucci G., Miglietta F., Ourcival J.M., Papale D., Pilegaard K., Rambal S., Seufert G., Soussana J.F., Sanz M.J., Schulze E.D., Vesala T., Valentini R. (2005) Europe-wide reduction in primary productivity caused by the heat and drought in 2003. Nature 437:529–533.
Clark J.S. (2010) Individuals and the variation needed for high species diversity in forest trees. Science 327:1129.
Clifford S.C., Black C.R., Roberts J.A., Stronach I.M., Singleton-Jones P.R., Mohamed A.D., Azam-Ali S.N. (1995) The effect of elevated atmospheric CO2 and drought on stomatal frequency in groundnut (Arachis hypogaea(L.)). Journal of Experimental Botany 46:847–852.
Close T.J. (1996) Dehydrins: Emergence of a biochemical role of a family of plant dehydration proteins. Physiologia Plantarum 97:795–803.
Cocozza C., Cherubini P., Regier N., Saurer M., Frey B., Tognetti R. (2010) Early effects of water deficit on two parental clones of Populus nigra grown under different environmental conditions. Functional Plant Biology 37:244–254.
Cowan I.R., Farquhar G.D. (1977) Stomatal function in relation to leaf metabolism and environment. Symposia of the Society for Experimental Biology 31:471–505.
Cramer G.R., Ergül A., Grimplet J., Tillett R.L., Tattersall E.A.R., Bohlman M.C., Vincent D., Sonderegger J., Evans J., Osborne C., Quilici D., Schlauch K.A., Schooley D.A., Cushman J.C. (2006) Water and salinity stress in grapevines: early and late changes in transcript and metabolite profiles. Functional & Integrative Genomics 7:111–134.
Davies W.J., Zhang J. (1991) Root signals and the regulation of growth and development of plants in drying soil. Annual Review of Plant Physiology and Plant Molecular Biology 42:55–76.
Deschamps S., Campbell M. (2009) Utilization of next-generation sequencing platforms in plant
156genomics and genetic variant discovery. Molecular Breeding 25:553–570.
Desrochers A., Thomas B.R. (2003) A comparison of pre-planting treatments on hardwood cuttings of four hybrid poplar clones. New Forests 26:17–23.
deVicente M.C., Tanksley S.D. (1993) QTL analysis of transgressive segregation in an interspecific tomato cross. Genetics 134:585–596.
Dickmann D.I. (2001) An Overview of the genus Populus. In: Dickmann DI, Isebrands JG, Eckenwaler JE, Richadson J (eds) Poplar Culture in North America (Part 2). NRC Research Press, Ottawa, pp 1–42.
Doyle J.J., Doyle J.L. (1990) A rapid total DNA preparation procedure for fresh plant tissue. Focus 12:13–15.
Du Z., Zhou X., Ling Y., Zhang Z., Su Z. (2010) agriGO: a GO analysis toolkit for the agricultural community. Nucleic Acids Research 38 Suppl:W64–W70.
Duan B., Lu Y., Yin C., Junttila O., Li C. (2005) Physiological responses to drought and shade in two contrastingPicea asperatapopulations. Physiologia Plantarum 124:476–484.
Dubos C., Le Provost G., Pot D., Salin F., Lalane C., Madur D., Frigerio J.-M., Plomion C. (2003a) Identification and characterization of water-stress-responsive genes in hydroponically grown maritime pine (Pinus pinaster) seedlings. Tree Physiology 23:169–179.
Dubos C., Plomion C. (2003b) Identification of water-deficit responsive genes in maritime pine (Pinus pinaster Ait.) roots. Plant Molecular Biology 51:249–262.
Eisen M.B., Spellman P.T., Brown P.O., Botstein D. (1998) Cluster analysis and display of genome-wide expression patterns. Proceedings of the National Academy of Sciences of the United States of America 95:14863–14868.
Engelbrecht B.M.J., Comita L.S., Condit R., Kursar T.A., Tyree M.T., Turner B.L., Hubbell S.P. (2007) Drought sensitivity shapes species distribution patterns in tropical forests. Nature
157447:80–82.
Escobar-Gutiérrez A.J., Zipperlin B., Carbonne F., Moing A., Gaudillére J.P. (1998) Photosynthesis, carbon partitioning and metabolite content during drought stress in peach seedlings. Australian Journal of Plant Physiology 25:197.
Ferrell W.K., Woodard E.S. (1966) Effects of seed origin on drought resistance of Douglas-Fir (Pseudotsuga menziesii) (Mirb.) Franco. Ecology 47:499–503.
Fischer U., Polle A. (2010) Populus responses to abiotic stress. In: Jansson S, Bhalerao R, Groover A (eds) Plant Genetics and Genomics: Crops and Models. Springer New York, pp 225–246.
Fischlin A., Midgley G.F., Price J.T., Leemans R. (2007) Ecosystems, their properties, goods, and services. In: van der Linden PJ, Hanson E (eds) Climate change 2007: impacts, adaptation and vulnurability. contribution of working group ii to the fourth assessment report of the intergovernmental panel on climate change, Cambridge University Press, Cambridge, pp 211–272.
Froux F., Ducrey M., Dreyer E., Huc R. (2005) Vulnerability to embolism differs in roots and shoots and among three Mediterranean conifers: consequences for stomatal regulation of water loss? Trees - Structure and Function 19:137–144.
Fujisawa H., Horiuchi Y., Harushima Y., Takada T., Eguchi S., Mochizuki T., Sakaguchi T., Shiroishi T., Kurata N. (2009) SNEP: Simultaneous detection of nucleotide and expression polymorphisms using Affymetrix GeneChip. BMC Bioinformatics 10:131.
Galmés J., Flexas J., Savé R., Medrano H. (2007) Water relations and stomatal characteristics of Mediterranean plants with different growth forms and leaf habits: responses to water stress and recovery. Plant and Soil 290:139–155.
Gautier L., Cope L., Bolstad B.M., Irizarry R.A. (2004) affy--analysis of Affymetrix GeneChip data at the probe level. Bioinformatics 20:307–315.
Gebre G.M., Kuhns M.R. (1991) Seasonal and clonal variations in drought tolerance of Populus deltoides. Canadian Journal of Forest Research 21:910–916.
158Gentleman R., Carey V., Bates D., Bolstad B., Dettling M., Dudoit S., Ellis B., Gautier L., Ge
Y., Gentry J., Hornik K., Hothorn T., Huber W., Iacus S., Irizarry R.A., Leisch F., Li C., Maechler M., Rossini A., Sawitzki G., Smith C., Smyth G.K., Tierney L., Yang J., Zhang J. (2004) Bioconductor: open software development for computational biology and bioinformatics. Genome Biology 5:R80.
Giovannelli A., Deslauriers A., Fragnelli G., Scaletti L., Castro G., Rossi S., Crivellaro A. (2007) Evaluation of drought response of two poplar clones (Populus x canadensis Monch ‘I-214’ and P. deltoides Marsh. “Dvina”) through high resolution analysis of stem growth. Journal of Experimental Botany 58:2673–2683.
Giuliano G., Pichersky E., Malik V.S., Timko M.P., Scolnik P.A., Cashmore A.R. (1988) An evolutionarily conserved protein binding sequence upstream of a plant light-regulated gene. Proceedings of the National Academy of Sciences of the United States of America 85:7089–7093.
Gonzalez-Martinez S.C., Ersoz E., Brown G.R., Wheeler N.C., Neale D.B. (2006) DNA sequence variation and selection of tag single-nucleotide polymorphisms at candidate genes for drought-stress response in Pinus taeda L. Genetics 172:1915–1926.
Gourcilleau D., Bogeat-Triboulot M.-B., Le Thiec D., Lafon-Placette C., Delaunay A., El-Soud W.A., Brignolas F., Maury S. (2010) DNA methylation and histone acetylation: genotypic variations in hybrid poplars, impact of water deficit and relationships with productivity. Annals of Forest Science 67:208.
Goyal K., Walton L.J., Tunnacliffe A. (2005) LEA proteins prevent protein aggregation due to water stress. Biochemical Journal 388:151–157.
Grattapaglia D., Plomion C., Kirst M., Sederoff R.R. (2009) Genomics of growth traits in forest trees. Current Opinion in Plant Biology 12:148–156.
Gray J.E., Holroyd G.H., van der Lee F.M., Bahrami A.R., Sijmons P.C., Woodward F.I., Schuch W., Hetherington A.M. (2000) The HIC signalling pathway links CO2 perception to stomatal development. Nature 408:713–716.
159Guiltinan M.J., Marcotte W.R.J., Quatrano R.S. (1990) A plant leucine zipper protein that
recognizes an abscisic acid response element. Science 250:267–271.
Gupta P.K., Rustgi S., Mir R.R. (2008) Array-based high-throughput DNA markers for crop improvement. Heredity 101:5–18.
Guy C., Kaplan F., Kopka J., Selbig J., Hincha D.K. (2008) Metabolomics of temperature stress. Physiologia Plantarum 132:220–235.
Hadley J., Smith W. (1990) Influence of leaf surface wax and leaf area to water content ratio on cuticular transpiration in western conifers, U.S.A. Canadian Journal of Forest Research 20:1306–1311.
Hall D., Tegstrom C., Ingvarsson P.K. (2010) Using association mapping to dissect the genetic basis of complex traits in plants. Briefings in Functional Genomics 9:157–165.
Hamanishi E.T., Raj S., Wilkins O., Thomas B.R., Mansfield S.D., Plant A.L., Campbell M.M. (2010) Intraspecific variation in the Populus balsamifera drought transcriptome. Plant, Cell & Environment 33:1742–1755.
Hannah M.A., Wiese D., Freund S., Fiehn O., Heyer A.G., Hincha D.K. (2006) Natural genetic variation of freezing tolerance in Arabidopsis. Plant Physiology 142:98–112.
Hara K., Kajita R., Torii K.U., Bergmann D.C., Kakimoto T. (2007) The secretory peptide gene EPF1 enforces the stomatal one-cell-spacing rule. Genes & Development 21:1720–1725.
Hara K., Yokoo T., Kajita R., Onishi T., Yahata S., Peterson K.M., Torii K.U., Kakimoto T. (2009) Epidermal cell density is autoregulated via a secretory peptide, EPIDERMAL PATTERNING FACTOR 2 in Arabidopsis leaves. Plant and Cell Physiology 50:1019–1031.
Harb A., Krishnan A., Ambavaram M.M.R., Pereira A. (2010) Molecular and physiological analysis of drought stress in Arabidopsis reveals early responses leading to acclimation in plant growth. Plant Physiology 154:1254–1271.
160Hartung W., Sauter A., Turner N.C., Fillery I., Heilmeier H. (1996) Abscisic acid in soils: What is
its function and which factors and mechanisms influence its concentration? Plant and Soil 184:105–110.
Heath L.S., Ramakrishnan N., Sederoff R.R., Whetten R., Chevone B., Struble C., Jouenne V., Chen D., vanZyl L., Grene R. (2002) Studying the functional genomics of stress responses in loblolly pine with expresso microarray experiment management system. Computational Functional Genomics 3:226–243.
Hertzberg M., Aspeborg H., Schrader J., Andersson A., Erlandsson R., Blomqvist K., Bhalerao R., Uhlen M., Teeri T.T., Lundeberg J., Sundberg B., Nilsson P., Sandberg G. (2001) A transcriptional roadmap to wood formation. Proceedings of the National Academy of Sciences of the United States of America 98:14732–14737.
Hirai M.Y. (2005) Elucidation of gene-to-gene and metabolite-to-gene networks in arabidopsis by integration of metabolomics and transcriptomics. Journal of Biological Chemistry 280:25590–25595.
Hoffman D.E., Jonsson P., Bylesjo M., Trygg J., Antti H., Eriksson M.E., Moritz T. (2010) Changes in diurnal patterns within the Populus transcriptome and metabolome in response to photoperiod variation. Plant, Cell & Environment 33:1298–1313.
Hogg E.H., Brandt J.P., Kochtubajda B. (2002) Growth and dieback of aspen forests in northwestern Alberta, Canada, in relation to climate and insects. Canadian Journal of Forest Research 32:823–832.
Hogg E.H., Wein R.W. (2005) Impacts of drought on forest growth and regeneration following fire in southwestern Yukon, Canada. Canadian Journal of Forest Research 35:2141–2150.
Hsiao T.C., Xu L.-K. (2000) Sensitivity of growth of roots versus leaves to water stress: biophysical analysis and relation to water transport. Journal of Experimental Botany 51:1595–1616.
Hyodo H., Yamakawa S., Takeda Y., Tsuduki M., Yokota A., Nishitani K., Kohchi T. (2003) Active gene expression of a xyloglucan endotransglucosylase/hydrolase gene, XTH9, in inflorescence apices is related to cell elongation in Arabidopsis thaliana. Plant Molecular
161Biology 52:473–482.
Ingram J., Bartels D. (1996) The molecular basis of dehydration tolerance in plants. Annual Review of Plant Physiology and Plant Molecular Biology 47:377–403.
Intergovernmental Panel on Climate Change (2007) IPCC Fourth Assessment Report: Climate Change, 4th edn. Geneva, Switzerland.
Janz D., Behnke K., Schnitzler J.-P., Kanawati B., Schmitt-Kopplin P., Polle A. (2010) Pathway analysis of the transcriptome and metabolome of salt sensitive and tolerant poplar species reveals evolutionary adaption of stress tolerance mechanisms. BMC Plant Biology 10:150.
Jarvis P.G., Jarvis M.S. (1963) The water relations of tree seedlings: iv. some aspects of the tissue water relations and drought resistance. Physiologia Plantarum 16:501–516.
Joshi V., Jander G. (2009) Arabidopsis methionine-lyase is regulated according to isoleucine biosynthesis needs but plays a subordinate role to threonine deaminase. Plant Physiology 151:367–378.
Kanaoka M.M., Pillitteri L.J., Fujii H., Yoshida Y., Bogenschutz N.L., Takabayashi J., Zhu J.-K., Torii K.U. (2008) SCREAM/ICE1 and SCREAM2 Specify three cell-state transitional steps leading to Arabidopsis stomatal differentiation. The Plant Cell 20:1775–1785.
Kanehisa M., Goto S. (2000) KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Research 28:27–30.
Kanehisa M., Goto S., Sato Y., Furumichi M., Tanabe M. (2011) KEGG for integration and interpretation of large-scale molecular data sets. Nucleic Acids Research 40:D109–D114.
Klingler J.P., Batelli G., Zhu J.-K. (2010) ABA receptors: the START of a new paradigm in phytohormone signalling. Journal of Experimental Botany 61:3199–3210.
Keller S.R., Olson M.S., Silim S., Schroeder W.R., Tiffin P. (2010) Genomic diversity, population structure, and migration following rapid range expansion in the Balsam poplar, Populus
162balsamifera. Molecular Ecology 19:1212–1226.
Knight H., Knight M.R. (2001) Abiotic stress signalling pathways: specificity and cross-talk. Trends in Plant Science 6:262–267.
Kondo T., Kajita R., Miyazaki A., Hokoyama M., Nakamura-Miura T., Mizuno S., Masuda Y., Irie K., Tanaka Y., Takada S., Kakimoto T., Sakagami Y. (2010) Stomatal density is controlled by a mesophyll-derived signaling molecule. Plant and Cell Physiology 51:1–8.
Kopka J., Provart N.J., Müller-Rober B. (1997) Potato guard cells respond to drying soil by a complex change in the expression of genes related to carbon metabolism and turgor regulation. The Plant Journal 11: 871-882
Koussevitzky S., Suzuki N., Huntington S., Armijo L., Sha W., Cortes D., Shulaev V., Mittler R. (2008) Ascorbate peroxidase 1 plays a key role in the response of Arabidopsis thaliana to stress combination. 283: 34197-34203.
Kozlowski T., Pallardy S. (2002) Acclimation and adaptive responses of woody plants to environmental stresses. The Botanical Review 68:270–334.
Krasensky J., Jonak C. (2012) Drought, salt, and temperature stress-induced metabolic rearrangements and regulatory networks. Journal of Experimental Botany 63:1593–1608.
Kreps J.A., Wu Y., Chang H.-S., Zhu T., Wang X., Harper J.F. (2002) Transcriptome changes for Arabidopsis in response to salt, osmotic, and cold stress. Plant Physiology 130:2129–2141.
Lai Z., Gross B.L., Zou Y., Andrews J., Rieseberg L.H. (2006) Microarray analysis reveals differential gene expression in hybrid sunflower species. Molecular Ecology 15:1213–1227.
Lake J.A., Quick W.P., Beerling D.J., Woodward F.I. (2001) Plant development: signals from mature to new leaves. Nature 411:154–154.
Langfelder P., Horvath S. (2008) WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9:559.
163Larson P.R., Isebrands J.G. (1971) The plastochron index as applied to developmental studies of
cottonwood. Canadian Journal of Forest Research 1:1–11.
Lee J.M., Joung J.-G., McQuinn R., Chung M.-Y., Fei Z., Tieman D., Klee H., Giovannoni J. (2012) Combined transcriptome, genetic diversity and metabolite profiling in tomato fruit reveals that the ethylene response factor SlERF6 plays an important role in ripening and carotenoid accumulation. The Plant Journal 70:191–204.
Lei Y.B., Yin C.Y., Li C.Y. (2006) Differences in some morphological, physiological, and biochemical responses to drought stress in two contrasting populations of Populus przewalskii. Physiologia Plantarum 127:182–191.
Li X., Wu H.X., Dillon S.K., Southerton S.G. (2009) Generation and analysis of expressed sequence tags from six developing xylem libraries in Pinus radiata D. Don. BMC Genomics 10:41.
Liang Y.K., Dubos C., Dodd I.C., Holroyd G.H., Hetherington A.M., Campbell M.M. (2005) AtMYB61, an R2R3-MYB transcription factor controlling stomatal aperture in Arabidopsis thaliana. Current Biology 15:1201–1206.
Liu W., Fairbairn D.J., Reid R.J., Schachtman D.P. (2001) Characterization of two HKT1 homologues from Eucalyptus camaldulensis that display intrinsic osmosensing capability. Plant Physiology 127:283–294.
Lopes C.T., Franz M., Kazi F., Donaldson S.L., Morris Q., Bader G.D. (2010) Cytoscape Web: an interactive web-based network browser. Bioinformatics 26:2347–2348.
Lorenz W.W., Sun F., Liang C., Kolychev D., Wang H., Zhao X., Cordonnier-Pratt M.-M., Pratt L.H., Dean J.F.D. (2006) Water stress-responsive genes in loblolly pine (Pinus taeda) roots identified by analyses of expressed sequence tag libraries. Tree Physiology 26:1–16.
Lu Y., Duan B., Zhang X., Korpelainen H., Berninger F., Li C. (2009) Intraspecific variation in drought response of Populus cathayanagrown under ambient and enhanced UV-B radiation. Annals of Forest Science 66:613.
Luo Z.W., Potokina E., Druka A., Wise R.P., Waugh R., Kearsey M.J. (2007) SFP genotyping
164from affymetrix arrays is robust but largely detects cis-acting expression regulators. Genetics 176:789–800.
Ma Y., Szostkiewicz I., Korte A., Moes D., Yang Y., Christmann A., Grill E. (2009) Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science 324:1064 –1068.
MacAlister C.A., Bergmann D.C. (2011) Sequence and function of basic helix–loop–helix proteins required for stomatal development in Arabidopsis are deeply conserved in land plants. Evolution & Development 13:182–192.
MacAlister C.A., Ohashi-Ito K., Bergmann D.C. (2007) Transcription factor control of asymmetric cell divisions that establish the stomatal lineage. Nature 445:537–540.
Maclennan D.H., Beevers H., Harley J.L. (1963) “Compartmentation” of acids in plant tissues. Biochemical Journal 89:316–327.
Mahajan S., Tuteja N. (2005) Cold, salinity and drought stresses: An overview. Archives of Biochemistry and Biophysics 444:139–158.
Mahdieh M., Mostajeran A., Horie T., Katsuhara M. (2008) Drought stress alters water relations and expression of PIP-type aquaporin genes in Nicotiana tabacum plants.
Mardis E.R. (2008) Next-generation DNA sequencing methods. Genomics and Human Genetics 9:387–402.
Marron N., Delay D., Petit J.-M., Dreyer E., Kahlem G., Delmotte F.M., Brignolas F. (2002) Physiological traits of two Populus x euramericana clones, Luisa Avanzo and Dorskamp, during a water stress and re-watering cycle. Tree Physiology 22:849–858.
Masle J., Gilmore S.R., Farquhar G.D. (2005) The ERECTA gene regulates plant transpiration efficiency in Arabidopsis. Nature 436:866–870.
Masoudi-Nejad A., Goto S., Endo T.R., Kanehisa M. (2007) KEGG bioinformatics resource for plant genomics research. Methods in molecular biology (Clifton, NJ) 406:437–458.
165Maurel C., Verdoucq L., Luu D.-T., Santoni V. (2008) Plant Aquaporins: membrane channels with
multiple integrated functions. Annu Rev Plant Biol 59:595–624.
McCarthy D.J., Smyth G.K. (2009) Testing significance relative to a fold-change threshold is a TREAT. Bioinformatics 25:765–771.
McCourt P., Creelman R. (2008) The ABA receptors – we report you decide. Current Opinion in Plant Biology 11:474–478.
Micheli F. (2001) Pectin methylesterases: cell wall enzymes with important roles in plant physiology. Trends in Plant Science 6:414–419.
Miyazawa S.-I., Livingston N.J., Turpin D.H. (2006) Stomatal development in new leaves is related to the stomatal conductance of mature leaves in poplar (Populus trichocarpa x P. deltoides). Journal of Experimental Botany 57:373–380.
Monclus R., Dreyer E., Villar M., Delmotte F.M., Delay D., Petit J.-M., Barbaroux C., Le Thiec D., Brechet C., Brignolas F. (2006) Impact of drought on productivity and water use efficiency in 29 genotypes of Populus deltoides x Populus nigra. New Phytologist 169:765–777.
Munne-Bosch S., Alegre L. (2004) Die and let live: leaf senescence contributes to plant survival under drought stress. Functional Plant Biol 31:203–216.
Nadeau J.A., Sack F.D. (2002) Control of stomatal distribution on the Arabidopsis leaf surface. Science 296:1697–1700.
Nagaraj S.H., Gasser R.B., Ranganathan S. (2007) A hitchhiker’s guide to expressed sequence tag (EST) analysis.
Nambara E., Kawaide H., Kamiya Y., Naito S. (1998) Characterization of an Arabidopsis thaliana mutant that has a defect in ABA accumulation: ABA-dependent and ABA-independent accumulation of free amino acids during dehydration. Plant and Cell Physiology 39:853–858.
166Namroud M.-C., Beaulieu J., Juge N., Laroche J., Bousquet J. (2008) Scanning the genome for
gene single nucleotide polymorphisms involved in adaptive population differentiation in white spruce - NAMROUD - 2008 - Molecular Ecology - Wiley Online Library. Molecular Ecology 17:3599–3613.
Nguyen A., Lamant A. (1989) Variation in growth and osmotic regulation of roots of water-stressed maritime pine (Pinus pinaster Ait.) provenances. Tree Physiology 5:123–133.
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.
Ohashi-Ito K., Bergmann D.C. (2006) Arabidopsis FAMA Controls the final proliferation/differentiation switch during stomatal development. The Plant Cell Online 18:2493–2505.
Oleksiak M.F., Churchill G.A., Crawford D.L. (2002) Variation in gene expression within and among natural populations. Nature Genetics 32:261–266.
Ollas C., Hernando B., Arbona V., Gómez-Cadenas A. (2012) Jasmonic acid transient accumulation is needed for abscisic acid increase in citrus roots under drought stress conditions. Physiologia Plantarum doi:10.111/j.1399-3054.2012.01659.x.
Olson M.S., Amanda L.R., Takebayashi N., Salim S., Schroeder W.R., Tiffin P. (2010) Nucleotide diversity and linkage disequilibrium in balsam poplar (Populus balsamifera). New Phytologist 186:526–536.
Park S.Y., Fung P., Nishimura N., Jensen D.R., Fujii H., Zhao Y., Lumba S., Santiago J., Rodrigues A., Chow T.F.F., Alfred S.E., Bonetta D., Finkelstein R., Provart N.J., Desveaux D., Rodriguez P.L., McCourt P., Zhu J.-K., Schroeder J.I., Volkman B.F., Cutler S.R. (2009) Abscisic acid inhibits type 2c protein phosphatases via the PYR/PYL family of START proteins. Science 324: 1068–1071.
Pechanova O., Hsu C.-Y., Adams J.P., Pechan T., Vandervelde L., Drnevich J., Jawdy S., Adeli A., Suttle J.C., Lawrence A.M., Tschaplinski T.J., Séguin A., Yuceer C. (2010) Apoplast proteome reveals that extracellular matrix contributes to multistress response in poplar. BMC Genomics 11:674–674.
167Pena-Rojas K., Aranda X., Joffre R., Fleck I. (2005) Leaf morphology, photochemistry and water
status changes in resprouting Quercus ilex during drought. Functional Plant Biology 32:117–130.
Pfaffl M.W. (2001) A new mathematical model for relative quantification in real-time RT-qPCR. Nucleic Acids Research 29:e45.
Pillitteri L.J., Sloan D.B., Bogenschutz N.L., Torii K.U. (2007) Termination of asymmetric cell division and differentiation of stomata. Nature 445:501–505.
Plomion C., Lalanne C., Claverol S., Meddour H., Kohler A., Bogeat-Triboulot M.B., Barre A., Le Provost G., Dumazet H., Jacob D., Bastien C., Dreyer E., De Daruvar A., Guehl J.M., Schmitter J.M., Martin F., Bonneu M. (2006) Mapping the proteome of poplar and application to the discovery of drought-stress responsive proteins. Proteomics 6:6509–6527.
Polle A., Altman A., Jiang X. (2006) Towards genetic engineering for drought tolerance in trees. In: Fladung M, Ewald D (eds) Tree transgenesis. Springer Berlin Heidelberg, pp 275–297.
Popko J., Hänsch R., Mendel R.R., Polle A., Teichmann T. (2010) The role of abscisic acid and auxin in the response of poplar to abiotic stress. Plant Biology 12:242–258.
Quarrie S.A., Jones H.G. (1977) Effects of abscisic acid and water stress on development and morphology of wheat. Journal of Experimental Botany 28:192–203.
R Development Core Team (2009) R: A language and environment for statistical computing.
Rabbani M.A., Maruyama K., Abe H., Khan M.A., Katsura K., Ito Y., Yoshiwara K., Seki M., Shinozaki K., Yamaguchi-Shinozaki K. (2003) Monitoring expression profiles of rice genes under cold, drought, and high-salinity stresses and abscisic acid application using cdna microarray and rna gel-blot analyses. Plant Physiology 133:1755–1767.
Radoglou K.M., Jarvis P.G. (1990) Effects of CO2 enrichment on four poplar clones. I. growth and leaf anatomy. Annals of Botany 65:617–626.
168Raj S., Bräutigam K., Hamanishi E.T., Wilkins O., Thomas B.R., Schroeder W.R., Mansfield
S.D., Plant A.L., Campbell M.M. (2011) Clone history shapes Populus drought responses. Proceedings of the National Academy of Sciences of the United States of America 108:12521–12526.
Ramanjulu S., Bartels D. (2002) Drought- and desiccation-induced modulation of gene expression in plants. Plant, Cell & Environment 25:141–151.
Regier N., Streb S., Cocozza C., Schaub M., Cherubini P., Zeeman S.C., Frey B. (2009) Drought tolerance of two black poplar (Populus nigra L.) clones: contribution of carbohydrates and oxidative stress defence. Plant, Cell & Environment 32:1724–1736.
Rice K.J., Matzner S.L., Byer W., Brown J.R. (2004) Patterns of tree dieback in Queensland, Australia: the importance of drought stress and the role of resistance to cavitation. Oecologia 139:190–198.
Robinson, A.R., Gheneim, R., Kozak, R.A., Ellis, D.D., Mansfield, S.D. (2005) The potential of metabolite profiling as a selection tool for genotype discrimination in Populus. Journal of Experimental Botany 56: 2807–2819.
Rodríguez-Calcerrada J., Atkin O.K., Robson T.M., Zaragoza-Castells J., Gil L., Aranda I. (2010) Thermal acclimation of leaf dark respiration of beech seedlings experiencing summer drought in high and low light environments. Tree Physiology 30:214–224.
Ronnberg-Wästljung A.C., Glynn C., Weih M. (2005) QTL analyses of drought tolerance and growth for a Salix dasyclados x Salix viminalis hybrid in contrasting water regimes. Theoretical and Applied Genetics 110:537–549.
Saavedra X., Modrego A., Rodriguez D., Gonzalez-Garcia M.P., Sanz L., Nicolas G., Lorenzo O. (2010) The nuclear interactor PYL8/RCAR3 of Fagus sylvatica FsPP2C1 is a positive regulator of abscisic acid signaling in seeds and stress. Plant Physiology 152:133–150.
Saez A., Apostolova N., Gonzalez-Guzman M., Gonzalez-Garcia M.P., Nicolas C., Lorenzo O., Rodriguez P.L. (2004) Gain-of-function and loss-of-function phenotypes of the protein phosphatase 2C HAB1 reveal its role as a negative regulator of abscisic acid signalling. The
169Plant Journal 37:354–369.
Sakurai N., Akiyama M., Kuraishi S. (1986) Irreversible effects of water stress on growth and stomatal development in cotyledons of etiolated squash seedlings. The Plant Cell Physiol 27:1177–1185.
Santarius K.A. (1973) The protective effect of sugars on chloroplast membranes during temperature and water stress and its relationship to frost, desiccation and heat resistance. Planta 113:105–114.
Saravitz D.M., Pharr D.M., Carter T.E. (1987) Galactinol synthase activity and soluble sugars in developing seeds of four soybean genotypes.
Schauer N.N., Fernie A.R.A. (2006) Plant metabolomics: towards biological function and mechanism. Trends in Plant Science 11:508–516.
Schindler D.W., Donahue W.F. (2006) An impending water crisis in Canada’s western prairie provinces. Proceedings of the National Academy of Sciences of the United States of America 103:7210–7216.
Secchi F., Zwieniecki M.A. (2010) Patterns of PIP gene expression in Populus trichocarpa during recovery from xylem embolism suggest a major role for the PIP1 aquaporin subfamily as moderators of refilling process. Plant, Cell & Environment 33:1285–1297.
Seibt U., Rajabi A., Griffiths H., Berry J. (2008) Carbon isotopes and water use efficiency: sense and sensitivity. Oecologia 155:441–454.
Seki M., Narusaka M., Ishida J., Nanjo T., Fujita M., Oono Y., Kamiya A., Nakajima M., Enju A., Sakurai T., Satou M., Akiyama K., Taji T., Yamaguchi-Shinozaki K., Carninci P., Kawai J., Hayashizaki Y., Shinozaki K. (2002) Monitoring the expression profiles of 7000 Arabidopsis genes under drought, cold and high-salinity stresses using a full-length cDNA microarray. The Plant Journal 31:279–292.
Seki M., Umezawa T., Urano K., Shinozaki K. (2007) Regulatory metabolic networks in drought stress responses. Current Opinion in Plant Biology 10:296–302.
170Sharp R.E., Poroyko V., Hejlek L.G., Spollen W.G., Springer G.K., Bohnert H.J., Nguyen H.T.
(2004) Root growth maintenance during water deficits: physiology to functional genomics. Journal of Experimental Botany 55:2343–2351.
Shaw, D J., Steed, E B., DeBlander, T L. (2005) Forest Inventory and Analysis (FIA) annual inventory answers the question: What is happening to pinyon-juniper woodlands? Journal of Forestry 103:280–285.
Shen Q., Zhang P., Ho T.-H.D. (1996) Modular nature of abscisic acid (ABA) response complexes: composite promoter units that are necessary and sufficient for aba lnduction of gene expressionin barley. The Plant Cell 8:1107–1119.
Shinozaki K., Yamaguchi-Shinozaki K. (1996) Molecular responses to drought and cold stress. Current Opinion in Biotechnology 7:161.
Shinozaki K., Yamaguchi-Shinozaki K. (2000) Molecular responses to dehydration and low temperature: differences and cross-talk between two stress signaling pathways. Current Opinion in Plant Biology 3:217–223.
Shinozaki K., Yamaguchi-Shinozaki K. (2007) Gene networks involved in drought stress response and tolerance. Journal of Experimental Botany 58:221–227.
Shpak E.D., McAbee J.M., Pillitteri L.J., Torii K.U. (2005) Stomatal patterning and differentiation by synergistic interactions of receptor kinases. Science 309:290–293.
Shulaev V., Cortes D., Miller G., Mittler R. (2008) Metabolomics for plant stress response. Physiologia Plantarum 132:199–208.
Silim S., Nash R., Reynard D., White B., Schroeder W. (2009) Leaf gas exchange and water potential responses to drought in nine poplar (Populus spp.) clones with contrasting drought tolerance. Trees - Structure and Function 23:959–969.
Silim S., Ryan N., Kubien, D.S. (2010) Temperature responses of photosynthesis and respiration in Populus balsamifera L.: acclimation versus adaptation. Photosynthesis Research 104:19-30.
171Silva E.C., NogueiraI R.J.M.C., Vale F.H.A., de Araújo F.P., Pimenta M.A. (2009) Stomatal changes
induced by intermittent drought in four umbu tree genotypes. Brazilian Journal of Plant Physiology 121:33–42.
Sjodin A., Street N.R., Sandberg G., Gustafsson P., Jansson S. (2009) The Populus Genome Integrative Explorer (PopGenIE): a new resource for exploring the Populus genome. New Phytologist 182:1013–1025.
Slovik S., Daeter W., Hartung W. (1995) Compartmental redistribution and long-distancetransport of abscisic acid (ABA) in plants as influenced by environmental changes in the rhizosphere—a biomathematical model. Journal of Experimental Biology 46:881–894.
Smyth G.K. (2004) Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Statistical Applications in Genetics and Molecular Biology 3:Article 3.
Sobeih W.Y., Dodd I.C., Bacon M.A., Grierson D., Davies W.J. (2004) Long-distance signals regulating stomatal conductance and leaf growth in tomato (Lycopersicon esculentum) plants subjected to partial root-zone drying. Journal of Experimental Botany 55:2353–2363.
Steele K. (2009) Novel upland rice variety bred using marker-assisted selection and client-oriented breeding released in Jharkhand, India. Food and Agriculture Organization (FAO).
Sterky F., Bhalerao R.R., Unneberg P., Segerman B., Nilsson P., Brunner A.M., Charbonnel-Campaa L., Lindvall J.J., Tandre K., Strauss S.H. (2004) A Populus EST resource for plant functional genomics. Proceedings of the National Academy of Sciences of the United States of America 101:13951–13956.
Sterky F., Regan S., Karlsson J., Hertzberg M., Rohde A., Holmberg A., Amini B., Bhalerao R., Larson M., Villarroel R., van Montagu M., Sandberg G., Olsson O., Teeri T.T., Boerjan W., Gustafsson P., Uhlen M., Sundberg B., Lundeberg J. (1998) Gene discovery in the wood-forming tissues of poplar: Analysis of 5,692 expressed sequence tags. Proceedings of the National Academy of Sciences of the United States of America 95:13330–13335.
Strabala T.J. (2004) Expressed sequence tag databases from forestry tree species. In: Kumar
172S, Fladung M (eds) Molecular genetics and breeding of forest trees. Hawthorn Press, Binghamton, NY
Street N.R., Skogstrom O., Sjodin A., Tucker J., Rodriguez-Acosta M., Nilsson P., Jansson S., Taylor G. (2006) The genetics and genomics of the drought response in Populus. The Plant Journal 48:321–341.
Street N.R., Skogstrom O., Sjodin A., Tucker J., Rodriguez-Acosta M., Nilsson P., Jansson S., Taylor G. (2006) The genetics and genomics of the drought response in Populus. The Plant Journal 48:321–341.
Sugano S.S., Shimada T., Imai Y., Okawa K., Tamai A., Mori M., Hara-Nishimura I. (2010) Stomagen positively regulates stomatal density in Arabidopsis. Nature 463:241–244.
Suzuki R., Shimodaira H. (2006) Pvclust: an R package for assessing the uncertainty in hierarchical clustering. Bioinformatics 22:1540–1542.
Sweetlove L.J., Fernie A.R. (2005) Regulation of metabolic networks: understanding metabolic complexity in the systems biology era. New Phytologist 168:9–24.
Taji T., Ohsumi C., Iuchi 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 29:417–426.
Taji T., Seki M., Satou M., Sakurai T., Kobayashi M., Ishiyama K., Narusaka Y., Narusaka M., Zhu J.-K., Shinozaki K. (2004) Comparative genomics in salt tolerance between Arabidopsis and Arabidopsis-related halophyte salt cress using Arabidopsis microarray. Plant Physiology 135:1697–1709.
Tanaka Y., Nose T., Jikumaru Y., Kamiya Y. (2013) ABA inhibits entry into stomatal-lineage development in Arabidopsis leaves [published online ahead of print March 19 2013]. The Plant Journal. doi:10.1111/tpj.12136
Tardieu F., Simonneau T. (1998) Variability among species of stomatal control under fluctuating soil water status and evaporative demand: modelling isohydric and anisohydric behaviours.
173Journal of Experimental Botany 49:419–432.
Teskey R.O., Bongarten B.C., Cregg B.M., Dougherty P.M., Hennessey T.C. (1987) Physiology and genetics of tree growth response to moisture and temperature stress: an examination of the characteristics of loblolly pine (Pinus taeda L.). Tree Physiology 3:41–61.
Tisne S., Barbier F., Granier C. (2011) The ERECTA gene controls spatial and temporal patterns of epidermal cell number and size in successive developing leaves of Arabidopsis thaliana. Annals of Botany 108:159–168.
Tschaplinski T.J., Tuskan G.A., Gunderson C.A. (1994) Water-stress tolerance of black and eastern cottonwood clones and four hybrid progeny. 1. Growth, water relations and gas exchange. Canadian Journal of Forest Research 24:364–371.
Tschaplinski T.J., Tuskan G.A., Sewell M.M., Gebre G.M., Donald E.T.I., Pendleyi C. (2006) Phenotypic variation and quantitative trait locus identification for osmotic potential in an interspecific hybrid inbred F-2 poplar pedigree grown in contrasting environments. Tree Physiology 26:595–604.
Tsuchihira A., Hanba Y.T., Kato N., Doi T., Kawazu T., Maeshima M. (2010) Effect of overexpression of radish plasma membrane aquaporins on water-use efficiency, photosynthesis and growth of Eucalyptus trees. Tree Physiology 30:417–430.
Tuskan G.A., DiFazio S., Jansson S., Bohlmann J., Grigoriev I., Hellsten U., Putnam N., Ralph S., Rombauts S., Salamov A., Schein J., Sterck L., Aerts A., Bhalerao R.R., Bhalerao R.P., Blaudez D., Boerjan W., Brun A., Brunner A., Busov V., Campbell M.M., Carlson J., Chalot M., Chapman J., Chen G.L., Cooper D., Coutinho P.M., Couturier J., Covert S., Cronk Q., Cunningham R., Davis J., Degroeve S., Dejardin A., Depamphilis C., Detter J., Dirks B., Dubchak I., Duplessis S., Ehlting J., Ellis B., Gendler K., Goodstein D., Gribskov M., Grimwood J., Groover A., Gunter L., Hamberger B., Heinze B., Helariutta Y., Henrissat B., Holligan D., Holt R., Huang W., Islam-Faridi N., Jones S., Jones-Rhoades M.W., Jorgensen R., Joshi C., Kangasjarvi J., Karlsson J., Kelleher C., Kirkpatrick R., Kirst M., Kohler A., Kalluri U., Larimer F., Leebens-Mack J., Leple J.C., Locascio P., Lou Y., Lucas S., Martin F., Montanini B., Napoli C., Nelson D.R., Nelson C., Nieminen K., Nilsson O., Pereda V., Peter G., Philippe R., Pilate G., Poliakov A., Razumovskaya J., Richardson P., Rinaldi C., Ritland K., Rouze P., Ryaboy D., Schmutz J., Schrader J.,
174Segerman B., Shin H., Siddiqui A., Sterky F., Terry A., Tsai C.J., Uberbacher E., Unneberg P., Vahala J., Wall K., Wessler S., Yang G., Yin T., Douglas C., Marra M., Sandberg G., de Peer Y.V., Rokhsar D. (2006) The genome of black cottonwood, Populus trichocarpa (Torr. & Gray). Science 313:1596–1604.
Tuskan G.A., Gunter L.E., Yang Z.K., Yin T., Sewell M.M., DiFazio S.P. (2004) Characterization of microsatellites revealed by genomic sequencing of Populus trichocarpa. Canadian Journal of Forest Research 34:85–93.
Tyree M.T., Kolb K.J., Rood S.B., Patiño S. (1994) Vulnerability to drought-induced cavitation of riparian cottonwoods in Alberta: a possible factor in the decline of the ecosystem? Tree Physiology 14:455–466.
Ujino-Ihara T., Yoshimura K., Ugawa Y., Yoshimaru H., Nagasaka K., Tsumura Y. (2000) Expression analysis of ESTs derived from the inner bark of Cryptomeria japonica. Plant Molecular Biology 43:451–457.
Uno Y., Furihata T., Abe H., Yoshida R., Shinozaki K., Yamaguchi-Shinozaki K. (2000) Arabidopsis basic leucine zipper transcription factors involved in an abscisic acid-dependent signal transduction pathway under drought and high-salinity conditions. Proceedings of the National Academy of Sciences of the United States of America 97:11632–11637.
Urao T., Yakubov B., Satoh R., Yamaguchi-Shinozaki K., Seki M., Hirayama T., Shinozaki K. (1999) A transmembrane hybrid-type histidine kinase in Arabidopsis functions as an osmosensor. The Plant Cell 11:1743–1754.
Urbanczyk-Wochniak E., Baxter C., Kolbe A., Kopka J., Sweetlove L.J., Fernie A.R. (2005) Profiling of diurnal patterns of metabolite and transcript abundance in potato (Solanum tuberosum) leaves. Planta 221: 891–903.
USDA-NRCS (2009) Populus balsamifera L. The PLANTS Database [online] URL: http://plants.usda.gov (accessed May 2009).
van Leeuwen H., Kliebenstein D.J., West M.A.L., Kim K., van Poecke R., Katagiri F., Michelmore R.W., Doerge R.W., St Clair D.A. (2007) Natural variation among Arabidopsis thaliana
175accessions for transcriptome response to exogenous salicylic acid. The Plant Cell 19:2099–2110.
van Mantgem P.J., Stephenson N.L., Byrne J.C., Daniels L.D., Franklin J.F., Fule P.Z., Harmon M.E., Larson A.J., Smith J.M., Taylor A.H., Veblen T.T. (2009) Widespread increase of tree mortality rates in the Western United States. Science 323:521–524.
von Groll U., Berger D., Altmann T. (2002) The subtilisin-like serine protease SDD1 mediates cell-to-cell signaling during Arabidopsis Stomatal development. The Plant Cell 14:1527–1539.
Walton D.C., Harrison M.A., Cote P. (1976) The effects of water stress on abscisic-acid levels and metabolism in roots of Phaseolus vulgaris L. and Other Plants. Planta 131:1–4.
Wang C., Wang Y., Diao G., Jiang J., Yang C. (2010) Isolation and characterization of expressed sequence tags (ESTs) from cambium tissue of birch (Betula platyphylla Suk). Plant Molecular Biology 28:438–449.
Wang H., Ngwenyama N., Liu Y., Walker J.C., Zhang S. (2007) Stomatal Development and Patterning are regulated by environmentally responsive mitogen-activated protein kinases in Arabidopsis. The Plant Cell 19:63–73.
Wang T., Hamann A., Spittlehouse D.L., Aitken S.N. (2006) Development of scale-free climate data for Western Canada for use in resource management. International Journal of Climatology 26:383–397.
Watkinson J.I., Sioson A.A., Vasquez-Robinet C., Shukla M., Kumar D., Ellis M., Heath L.S., Ramakrishnan N., Chevone B., Watson L.T., van Zyl L., Egertsdotter U., Sederoff R.R., Grene R. (2003) Photosynthetic acclimation is reflected in specific patterns of gene expression in drought-stressed loblolly pine. Plant Physiology 133:1702–1716.
Weckwerth W. (2003) Metabolomics in systems biology. Annual Review of Plant Biology
Welin B.V., Olson Å., Nylander M., Palva E.T. (1994) Characterization and differential expression of dhn/lea/rab-like genes during cold acclimation and drought stress in Arabidopsis thaliana. Plant Molecular Biology 26:131–144.
176Whitehead A., Crawford D.L. (2006) Variation within and among species in gene expression: raw
material for evolution. Molecular Ecology 15:1197–1211.
Wilkins O., Nahal H., Foong J., Provart N.J., Campbell M.M. (2009a) Expansion and diversification of the Populus R2R3-MYB family of transcription factors. Plant Physiology 149:981–993.
Wilkins O., Waldron L., Nahal H., Provart N.J., Campbell M.M. (2009b) Genotype and time of day shape the Populus drought response. The Plant Journal 60:703–715.
Wilkinson S., Corlett J.E., Oger L., Davies W.J. (1998) Effects of Xylem pH on transpiration from wild-type and flacca tomato leaves. Plant Physiology 117:703–709.
Wilkinson S., Davies W.J. (2002) ABA-based chemical signalling: the co-ordination of responses to stress in plants. Plant, Cell & Environment 25:195–210.
Williams M.E., Foster R., Chua N.-H. (1992) Sequences Flanking the Hexameric G-Box Core CACGTG Affect the Specificity of Protein Binding. The Plant Cell 4:485–496.
Wise R.P., Caldo R.A., Hong L., Shen L., Cannon E., Dickerson J.A. (2007) BarleyBase/PLEXdb. In: Edwards D (ed) Plant Bioinformatics: Methods and Protocols. Humana Press, Totowa, New Jersey, USA, pp 347–363.
Wisniewski M., Close T.J., Artlip T., Arora R. (1996) Seasonal patterns of dehydrins and 70-kDa heat-shock proteins in bark tissues of eight species of woody plants. Physiologia Plantarum 96:496–505.
Woodward F.I. (1987) Stomatal numbers are sensitive to increases in CO2 from pre-industrial levels. Nature 327:617–618.
Wu Z., Irizarry R.A., Gentleman R., Martinez-Murillo F., Spencer F. (2004) A Model Based Background Adjustment for Oligonucleotide Expression Arrays. Johns Hopkins University, Dept. of Biostatistics Working Papers.
177Xu Z., Zhou G. (2008) Responses of leaf stomatal density to water status and its relationship with
photosynthesis in a grass. Journal of Experimental Botany 59:3317–3325.
Yan D.-H., Fenning T., Tang S., Xia X., Yin W. (2012) Genome-wide transcriptional response of Populus euphratica to long-term drought stress. Plant Science 195:24–35.
Yang F., Wang Y., Miao L.-F. (2010) Comparative physiological and proteomic responses to drought stress in two poplar species originating from different altitudes. Physiologia Plantarum 139:388–400.
Yin C., Duan B., Wang X., Li C. (2004) Morphological and physiological responses of two contrasting Poplar species to drought stress and exogenous abscisic acid application. Plant Science 167:1091–1097.
Yoo C.Y., Pence H.E., Jin J.B., Miura K., Gosney M.J., Hasegawa P.M., Mickelbart M.V. (2010) The Arabidopsis GTL1 transcription factor regulates water use efficiency and drought tolerance by modulating stomatal density via transrepression of SDD1. The Plant Cell Online 22:4128–4141.
Yordanov I., Velikova V., Tsonev T. (2000) Plant responses to drought, acclimation, and stress tolerance. Photosynthetica 38:171–186.
Yu H., Chen X., Hong Y.-Y., Wang Y., Xu P., Ke S.-D., Liu H.-Y., Zhu J.-K., Oliver D.J., Xiang C.-B. (2008) Activated expression of an Arabidopsis HD-START protein confers drought tolerance with improved root system and reduced stomatal density. The Plant Cell Online 20:1134–1151.
Zeevaart J.A.D., Creelman R.A. (1988) Metabolism and physiology of abscisic acid. Annual Review of Plant Physiology and Plant Molecular Biology 39:439–473.
Zhang S., Chen F., Peng S., Ma W., Korpelainen H., Li C. (2010) Comparative physiological, ultrastructural and proteomic analyses reveal sexual differences in the responses of Populus cathayana under drought stress. Proteomics 10:2661–2677.
Zhang X., Zang R., Li C. (2004) Population differences in physiological and morphological
178adaptations of Populus davidiana seedlings in response to progressive drought stress. Plant Science 166:791–797.
Zhao M., Running S.W. (2010) Drought-induced reduction in global terrestrial net primary production from 2000 through 2009. Science 329:940–943.
179
Appendix
180Appendix A.1 Metabolites that are significant for a treatment (T):time-of-day (D) interaction (P
<0.05, n=15)
1.52.02.53.03.54.0
M91T611_Benzoic_acid
Abun
danc
e
1.52.02.53.03.54.0
M91T611_Benzoic_acid
1.52.02.53.03.54.0
M147T645_Thymidine.5..monophoph
Abun
danc
e
1.52.02.53.03.54.0
M147T645_Thymidine.5..monophoph
−0.50.00.51.01.52.02.53.0
M256T666_L.Isoleucine
Abun
danc
e
−0.50.00.51.01.52.02.53.0
M256T666_L.Isoleucine
2
3
4
5
M147T704_Glycolic_acid
Abun
danc
e
2
3
4
5
M147T704_Glycolic_acid
0
1
2
3
M179T804_Salicyl_alcoholAb
unda
nce
0
1
2
3
M179T804_Salicyl_alcohol
3.03.54.04.55.0
M147T928_Threonic_acid
Abun
danc
e
2.53.03.54.04.55.0
M147T928_Threonic_acid
1.52.02.53.03.54.0
M217T1015_NI_5C_sugar
Abun
danc
e
1.52.02.53.03.54.0
M217T1015_NI_5C_sugar
0.51.01.52.02.53.0
M333T1095_NI
Abun
danc
e
0.51.01.52.02.53.0
M333T1095_NI
3.5
4.0
4.5
5.0
M204T1132_Shikimic_acidAb
unda
nce
3.5
4.0
4.5
5.0
M204T1132_Shikimic_acid
3.0
3.5
4.0
4.5
M364T1195_NI_6C_sugar
Abun
danc
e
3.0
3.5
4.0
4.5
M364T1195_NI_6C_sugar
4.5
5.0
5.5
M147T1210_Glucose_1
Abun
danc
e
4.0
4.5
5.0
5.5
M147T1210_Glucose_1
3.5
4.0
4.5
5.0
M147T1224_Glucose_2
Abun
danc
e
3.5
4.0
4.5
5.0
M147T1224_Glucose_2
2.02.53.03.54.0
M204T1305_Glucose_3
Abun
danc
e
2.02.53.03.54.0
M204T1305_Glucose_3
3.03.54.04.55.05.5
M361T1617_Salicin
Abun
danc
e
3.03.54.04.55.05.5
M361T1617_Salicin
2
3
4
5
M361T2065_Raffinose
Abun
danc
e
2
3
4
5
M361T2065_Raffinose
1.51.5
2.5
4.0
2.52.5
Pre-dawn Mid-day
Time of Day
WetDry
Treatment
181
2.02.53.03.54.04.5
M147T569_Malonic_acid
Abun
danc
e
2.02.53.03.54.04.5
M147T569_Malonic_acid
1.52.02.53.03.54.0
M147T645_Thymidine.5..monophoph
Abun
danc
e
1.52.02.53.03.54.0
M147T645_Thymidine.5..monophoph
0.00.51.01.52.02.53.03.5
M142T667_L.Proline_1
Abun
danc
e
0.00.51.01.52.02.53.03.5
M142T667_L.Proline_1
0.51.01.52.02.53.03.5
M174T677_Glycine
Abun
danc
e
0.00.51.01.52.02.53.03.5
M174T677_Glycine
3.03.54.04.55.0
M147T682_Succinic_acidAb
unda
nce
2.53.03.54.04.55.0
M147T682_Succinic_acid
2.0
2.5
3.0
3.5
M254T688_Catechol
Abun
danc
e
2.0
2.5
3.0
3.5
M254T688_Catechol
2
3
4
5
M147T704_Glycolic_acid
Abun
danc
e
2
3
4
5
M147T704_Glycolic_acid
3.03.54.04.5
M245T714_Fumaric_acid
Abun
danc
e
2.53.03.54.04.5
M245T714_Fumaric_acid
2.02.53.03.54.0
M188T733_L.AlanineAb
unda
nce
2.02.53.03.54.0
M188T733_L.Alanine
1234
M204T734_L.Serine
Abun
danc
e
1234
M204T734_L.Serine
4.5
5.0
5.5
M147T857_Malic_acid
Abun
danc
e
4.0
4.5
5.0
5.5
M147T857_Malic_acid
0.51.01.52.02.53.03.5
M233T886_L.Aspartic_acid
Abun
danc
e
0.51.01.52.02.53.03.5
M233T886_L.Aspartic_acid
2.53.03.54.04.5
M156T888_Pyroglutamic_acid
Abun
danc
e
2.53.03.54.04.5
M156T888_Pyroglutamic_acid
0.00.51.01.52.02.53.0
M174T892_Butyric_acid.4.amino.n
Abun
danc
e
0.00.51.01.52.02.53.0
M174T892_Butyric_acid.4.amino.n
2.53.03.54.04.55.0
M147T928_Threonic_acid
Abun
danc
e
2.53.03.54.04.55.0
M147T928_Threonic_acid
AP947AP1005AP1006AP2278AP2298AP2300
Genotype
WetDry
Treatment
Appendix A.2 Metabolites that are significant for a treatment (T):genotype (G) interaction (P
<0.05, n=41)
182
0.00.51.01.52.02.53.0
M306T949_NI
Abun
danc
e
0.00.51.01.52.02.53.0
M306T949_NI
1
2
3
M192T980_L.Phenylalanine
Abun
danc
e
1
2
3
M192T980_L.Phenylalanine
2.0
2.5
3.0
M221T995_NI
Abun
danc
e
2.0
2.5
3.0
M221T995_NI
1.52.02.53.03.54.0
M217T1015_NI_5C_sugar
Abun
danc
e
1.52.02.53.03.54.0
M217T1015_NI_5C_sugar
0.51.01.52.02.53.0
M200T1035_NI_5C_sugarAb
unda
nce
0.51.01.52.02.53.0
M200T1035_NI_5C_sugar
1.52.02.53.03.54.04.5
M273T1143_Citric_acid
Abun
danc
e
1.52.02.53.03.54.04.5
M273T1143_Citric_acid
4.85.05.25.45.65.86.06.2
M345T1180_Quinic_acid
Abun
danc
e
4.85.05.25.45.65.86.06.2
M345T1180_Quinic_acid
1.52.02.53.03.54.0
M201T1190_Fructose_1
Abun
danc
e
1.52.02.53.03.54.0
M201T1190_Fructose_1
3.0
3.5
4.0
4.5
M364T1195_NI_6C_sugarAb
unda
nce
3.0
3.5
4.0
4.5
M364T1195_NI_6C_sugar
3.63.84.04.24.44.64.85.0
M218T1196_Fructose_2
Abun
danc
e
3.63.84.04.24.44.64.85.0
M218T1196_Fructose_2
4.5
5.0
5.5
M147T1210_Glucose_1
Abun
danc
e
4.5
5.0
5.5
M147T1210_Glucose_1
3.5
4.0
4.5
5.0
M147T1224_Glucose_2
Abun
danc
e
3.5
4.0
4.5
5.0
M147T1224_Glucose_2
1.52.02.53.03.54.0
M217T1234_NI_Sugar_alcohol
Abun
danc
e
1.52.02.53.03.54.0
M217T1234_NI_Sugar_alcohol
2.02.53.03.54.0
M204T1305_Glucose_3
Abun
danc
e
2.02.53.03.54.0
M204T1305_Glucose_3
2.5
3.0
3.5
4.0
M73T1365_NI_6C_sugar
Abun
danc
e
2.5
3.0
3.5
4.0
M73T1365_NI_6C_sugar
1.51.5
2.5
2.5
4.0
2.52.5
AP947AP1005AP1006AP2278AP2298AP2300
Genotype
WetDry
Treatment
183
2.53.03.54.0
M204T1411_NI
Abun
danc
e
2.02.53.03.54.0
M204T1411_NI
1.01.52.02.53.03.5
M91T1539_NI
Abun
danc
e
1.01.52.02.53.03.5
M91T1539_NI
2.02.53.03.54.04.5
M204T1542_NI
Abun
danc
e
2.02.53.03.54.04.5
M204T1542_NI
2.53.03.54.04.55.05.5
M361T1617_Salicin
Abun
danc
e
2.53.03.54.04.55.05.5
M361T1617_Salicin
2.0
2.5
3.0
3.5
M219T1659_NI
Abun
danc
e
2.0
2.5
3.0
3.5
M219T1659_NI
1.01.52.02.53.03.5
M236T1674_Adenosine
Abun
danc
e1.01.52.02.53.03.5
M236T1674_Adenosine
2.02.53.03.54.04.5
M368T1807_Catechin
Abun
danc
e
2.02.53.03.54.04.5
M368T1807_Catechin
0.00.51.01.52.02.53.0
M461T1841_NI
Abun
danc
e
0.00.51.01.52.02.53.0
M461T1841_NI
−10123
M456T1869_NI
Abun
danc
e
−10123
M456T1869_NI
1.01.52.02.53.03.54.04.5
M204T1876_Galactinol
Abun
danc
e
1.01.52.02.53.03.54.04.5
M204T1876_Galactinol
2.53.03.54.04.5
M204T1938_Digalactosylglycerol
Abun
danc
e
2.02.53.03.54.04.5
M204T1938_Digalactosylglycerol
1.51.5
2.5
2.5
4.0
AP947AP1005AP1006AP2278AP2298AP2300
Genotype
WetDry
Treatment
184Appendix A.3 Metabolites that are significant for a genotype (G):time of day (D) interaction(P
<0.05, n=6)
1.52.02.53.03.54.0
M91T611_Benzoic_acid
Abun
danc
e
1.52.02.53.03.54.0
M91T611_Benzoic_acid
2.0
2.5
3.0
M172T844_NI_Amino_acid
Abun
danc
e
1.5
2.0
2.5
3.0
M172T844_NI_Amino_acid
1.52.02.53.03.54.0
M201T1190_Fructose_1
Abun
danc
e1.01.52.02.53.03.54.0
M201T1190_Fructose_1
3.5
4.0
4.5
5.0
M147T1224_Glucose_2
Abun
danc
e
3.5
4.0
4.5
5.0
M147T1224_Glucose_2
Abun
danc
e
0.51.01.52.02.53.03.5
M294T1683_NI
0.00.51.01.52.02.53.03.5
M294T1683_NI
1.51.5
AP947AP1005AP1006AP2278AP2298AP2300
Genotype
Mid-dayPre-dawn
Time of Day
M204T1876_Galactinol
01
23
4
185
Citrate
Acetyl CoA
cis-Aconitate
Isocitrate
2-ketoglutarateSuccinyl-CoA
Succinate
Fumarate
Malate
Pyruvate
Oxaloacetate
4.2.1.2
1.3.5.1
glycolysis IV
1.1.1.37
6.2.1.5
1.1.1.41
2.3.3.1
4.2.1.3
−2 −1 0 1 2Fold Ratio [Log2(Fold-change)]
Color K ey Populus balsamifera, genotype AP-947
Malic AcidCitric Acid Succinic Acid
EC:1
.3.5
.1
EC:1
.1.1
.37
EC:4
.2.1
.3
EC:2
.3.3
.1
EC:1
.1.1
.41
EC
:6.2
.1.5
EC:4
.2.1
.2
(a)
(b)
−1 −0.5 0 0.5 1Pearson Correlation Coefficient
Color Key
AP-947
Appendix A.4 Analysis of the citrate cycle (TCA; pop00020) pathway in genotype AP-947, AP-
1005, AP-2278, AP-2298 and AP-2300. (a) Correlation among select transcripts and metabolites
from the KEGG pathway pop00020 ‘Citrate cycle (TCA cycle)’ for genotype AP-1006. Colors
represent Pearson correlation value. Red indicates positive correlation and blue represents negative
correlation values. (b) Map displays selected steps from citrate cycle pathway. Colours indicate
fold-change in transcript or metabolite abundance between water-deficit and well-watered treated
samples; red indicates higher abundance in water-deficit treated samples and blue indicates lower
abundance in water-deficit treated samples. Enzymes are given as EC numbers. EC 1.1.1.37,
malate dehydrogenase; EC:1.1.1.41, isocitrate dehydrogenase (NAD+); EC:1.3.5.1, succinate
dehydrogenase; EC:2.3.3.1, citrate synthase; EC:5.2.1.2, fumarate hydratase, EC: 5.2.1.3, aconitate
hydratase, EC: 6.2.1.5, succinate-CoA ligase, beta subunit. Pearson correlation and pathway maps
for AP-1006 can be found in Figure 5.11.
186
Citrate
Acetyl CoA
cis-Aconitate
Isocitrate
2-ketoglutarateSuccinyl-CoA
Succinate
Fumarate
Malate
Pyruvate
Oxaloacetate
4.2.1.2
1.3.5.1
glycolysis IV
1.1.1.37
6.2.1.5
1.1.1.41
2.3.3.1
4.2.1.3
−2 −1 0 1 2Fold Ratio [Log2(Fold-change)]
Color K ey Populus balsamifera, genotype AP-2278
Malic AcidCitric Acid Succinic Acid
EC:1
.3.5
.1
EC:1
.1.1
.37
EC:4
.2.1
.3
EC:2
.3.3
.1
EC:1
.1.1
.41
EC
:6.2
.1.5
EC:4
.2.1
.2
(a)
(b)
−1 −0.5 0 0.5 1Pearson Correlation Coefficient
Color Key
AP-2278
AP-1005
Citrate
Acetyl CoA
cis-Aconitate
Isocitrate
2-ketoglutarateSuccinyl-CoA
Succinate
Fumarate
Malate
Pyruvate
Oxaloacetate
4.2.1.2
1.3.5.1
glycolysis IV
1.1.1.37
6.2.1.5
1.1.1.41
2.3.3.1
4.2.1.3
−2 −1 0 1 2Fold Ratio [Log2(Fold-change)]
Color K ey Populus balsamifera, genotype AP-1005
Malic AcidCitric Acid Succinic Acid
EC:1
.3.5
.1
EC:1
.1.1
.37
EC:4
.2.1
.3
EC:2
.3.3
.1
EC:1
.1.1
.41
EC
:6.2
.1.5
EC:4
.2.1
.2
(a)
(b)
−1 −0.5 0 0.5 1Pearson Correlation Coefficient
Color Key
187
AP-2298
Citrate
Acetyl CoA
cis-Aconitate
Isocitrate
2-ketoglutarateSuccinyl-CoA
Succinate
Fumarate
Malate
Pyruvate
Oxaloacetate
4.2.1.2
1.3.5.1
glycolysis IV
1.1.1.37
6.2.1.5
1.1.1.41
2.3.3.1
4.2.1.3
−2 −1 0 1 2Fold Ratio [Log2(Fold-change)]
Color K ey Populus balsamifera, genotype AP-2298
Malic AcidCitric Acid Succinic Acid
EC:1
.3.5
.1
EC:1
.1.1
.37
EC:4
.2.1
.3
EC:2
.3.3
.1
EC:1
.1.1
.41
EC
:6.2
.1.5
EC:4
.2.1
.2
(a)
(b)
−1 −0.5 0 0.5 1Pearson Correlation Coefficient
Color Key
AP-2300
Citrate
Acetyl CoA
cis-Aconitate
Isocitrate
2-ketoglutarateSuccinyl-CoA
Succinate
Fumarate
Malate
Pyruvate
Oxaloacetate
4.2.1.2
1.3.5.1
glycolysis IV
1.1.1.37
6.2.1.5
1.1.1.41
2.3.3.1
4.2.1.3
−2 −1 0 1 2Fold Ratio [Log2(Fold-change)]
Color K ey Populus balsamifera, genotype AP-2300
Malic AcidCitric Acid Succinic Acid
EC:1
.3.5
.1
EC:1
.1.1
.37
EC:4
.2.1
.3
EC:2
.3.3
.1
EC:1
.1.1
.41
EC
:6.2
.1.5
EC:4
.2.1
.2
(a)
(b)
−1 −0.5 0 0.5 1Pearson Correlation Coefficient
Color Key
188Appendix A.5 Summary statistics for all metabolites (n=87). Mean peak intensity values
represented for all samples, wet samples and dry samples ± standard deviation. Relative abundance
between water-deficit and well-watered samples represented as log2(fold-ratio); positive values
indicate increased abundance in water-deficit conditions, whereas negative values indicate decreased
abundance in water-deficit conditions relative to control conditions. P-values for treatment (T)
main effect as calculated per the factorial ANOVA; metabolites significant for treatment main
effect denoted with an * (FDR < 0.05). Metabolite classes are as follows: AA = Amino Acid; C =
Carbohydrate; P = Phenolic, SA = Sugar Alcohol, and NI = Not Identified.
189
Peak
IDM
etab
olite
Cla
ssM
ean
Peak
Inte
nsity
Mea
n Pe
ak In
tens
ity
(Wet
)M
ean
Peak
Inte
nsity
(D
ry)
Rel
ativ
e A
bund
ance
[lo
g 2(Dry
/W
et)]
AN
OVA
: T-
mai
n ef
fect
P ad
j-val
ue
M23
6T16
74Ad
enos
ine
N52
6 .56
±39
6 .21
530 .
53±
291 .
5752
2 .80
±47
6 .15
-0 .0
20 .
137
M23
3T88
6As
parti
c Ac
idAA
296 .
28±
294 .
0736
7 .67
±30
7 .16
228 .
74±
265 .
49↓
-0 .6
80 .
031*
M91
T611
Benz
oic
Acid
OA
1504
.38
±15
08 .1
093
3 .98
±47
0 .81
2043
.78
±19
06 .3
2↑
1 .13
<0 .0
01*
M17
4T89
2Bu
tyric
_aci
d-4-
amin
o-n-
OA
259 .
23±
257 .
5327
5 .43
±25
0 .97
243 .
59±
264 .
24-0
.18
0 .74
7M
368T
1807
Cat
echi
nP
3100
.35
±37
08 .6
422
67 .4
5±
1948
.34
3888
.47
±46
93 .4
90 .
780 .
099
M25
4T68
8C
atec
hol
P70
0 .87
±43
6 .49
590 .
17±
324 .
8780
4 .35
±49
9 .59
↑0 .
450 .
004*
M27
3T11
43C
itric
Aci
dO
A31
39 .6
8±
3234
.99
2001
.96
±19
82 .6
442
15 .5
7±
3787
.82
↑1 .
07<0
.001
*M
204T
1938
Dig
alac
tosy
l gly
cero
lC
1063
7 .32
±50
37 .5
697
21 .9
6±
4424
.63
1150
3 .47
±54
38 .5
10 .
240 .
054
M20
1T11
90Fr
ucto
se (1
) C
2337
.75
±14
39 .9
723
71 .9
2±
1296
.96
2305
.45
±15
69 .7
3-0
.04
0 .05
4M
218T
1196
Fruc
tose
(2)
C26
732 .
90±
1171
5 .37
2823
4 .79
±11
885 .
4525
311 .
75±
1143
4 .45
↓-0
.16
0 .01
7*M
245T
714
Fum
aric
Aci
dO
A56
25 .5
9±
5424
.24
8033
.48
±59
34 .2
532
71 .2
1±
3570
.98
↓-1
.30
<0 .0
01*
M20
4T18
76G
alac
tinol
SA48
02 .0
0±
4877
.83
3830
.27
±46
78 .5
854
84 .8
4±
4929
.92
↑0 .
520 .
001*
M14
7T12
10G
luco
se (1
) C
1343
86 .6
3±
8520
6 .15
1435
66 .5
6±
8926
1 .75
1257
00 .2
4±
8070
3 .20
-0 .1
90 .
054
M14
7T12
24G
luco
se (2
)C
3471
2 .43
±25
603 .
7937
184 .
52±
2697
4 .81
3234
7 .82
±24
130 .
43-0
.20
0 .07
2M
204T
1305
Glu
cose
(3)
C25
46 .2
0±
2145
.51
2701
.56
±22
02 .6
723
86 .9
1±
2087
.23
-0 .1
80 .
091
M24
9T95
7G
lyce
rol
C33
2 .42
±16
2 .50
396 .
90±
177 .
1226
5 .63
±11
2 .71
↓-0
.58
<0 .0
01*
M17
4T67
7G
lyci
neAA
292 .
85±
258 .
5925
9 .50
±19
3 .18
326 .
21±
308 .
750 .
330 .
160
M14
7T70
4G
lyco
lic A
cid
OA
1352
4 .11
±14
317 .
9215
219 .
75±
1677
1 .14
1162
7 .66
±10
743 .
03↓
-0 .3
9<0
.001
*M
560T
1902
Kaem
pfer
olP
196 .
31±
268 .
0719
8 .54
±24
4 .28
194 .
14±
290 .
83-0
.03
0 .06
4M
188T
733
L-Al
anin
eAA
2307
.59
±16
29 .8
023
09 .3
5±
1539
.03
2305
.93
±17
19 .6
3-0
.00
0 .97
1M
332T
1243
L-As
corb
ic A
cid
OA
1109
.39
±12
58 .7
612
70 .8
9±
1608
.13
958 .
55±
787 .
16-0
.41
0 .31
5M
246T
973
L-G
luta
mat
eAA
4966
.54
±45
26 .0
556
72 .6
6±
5243
.92
4283
.69
±36
01 .9
4-0
.41
0 .87
3M
256T
666
L-Is
oleu
cine
AA40
.75
±10
2 .79
7 .28
±7 .
6272
.68
±13
6 .36
↑3 .
32<0
.001
*M
192T
980
L-Ph
enyl
alan
ine
AA49
2 .51
±57
5 .02
397 .
35±
343 .
9257
0 .23
±70
3 .82
0 .52
0 .61
1M
142T
667
L-Pr
olin
e (1
)AA
237 .
69±
309 .
2522
7 .17
±29
3 .06
248 .
39±
327 .
210 .
130 .
777
190
M14
2T94
1L-
Prol
ine
(2)
AA81
.42
±10
5 .25
64 .0
4±
50 .5
397
.65
±13
6 .40
0 .61
0 .27
4M
204T
734
L-Se
rine
AA38
34 .3
9±
3110
.60
4211
.38
±34
79 .9
034
82 .2
0±
2696
.85
-0 .2
70 .
708
M10
1T76
1L-
Thre
onin
eAA
1600
.57
±12
58 .7
418
36 .0
6±
1331
.99
1396
.48
±11
62 .2
3↓
-0 .3
90 .
031*
M28
0T12
29L-
Tyro
sine
AA25
.64
±40
.20
28 .5
1±
42 .2
722
.80
±38
.11
-0 .3
20 .
473
M14
7T85
7M
alic
Aci
dO
A78
173 .
33±
3822
4 .18
8145
1 .00
±31
450 .
0675
071 .
88±
4362
6 .83
↓-0
.12
0 .00
4*M
147T
569
Mal
onic
Aci
dO
A25
69 .2
8±
2729
.46
3884
.80
±32
28 .1
512
98 .1
0±
1157
.31
↓-1
.58
<0 .0
01*
M26
1T15
75M
alto
seC
505 .
10±
306 .
5051
3 .81
±26
8 .38
496 .
38±
341 .
76-0
.05
0 .60
7M
204T
1529
Mel
ibio
seC
2821
8 .02
±16
597 .
2825
181 .
91±
1514
5 .70
3109
0 .89
±17
460 .
72↑
0 .30
<0 .0
01*
M30
5T13
48M
yo-in
osito
lSA
1592
17 .2
0±
3658
3 .20
1509
16 .8
8±
3295
2 .17
1670
71 .2
7±
3825
4 .21
↑0 .
150 .
009*
M15
6T88
8Py
rogl
utam
ic A
cid
AA54
67 .0
0±
3942
.39
5908
.21
±38
06 .3
250
49 .5
2±
4043
.02
-0 .2
30 .
105
M64
8T19
55Q
uerc
itin
P15
56 .9
8±
1857
.22
1693
.12
±18
30 .6
614
29 .6
1±
1882
.66
↓-0
.24
0 .00
5*M
345T
1180
Qui
nic
Acid
OA
3232
18 .6
6±
1225
27 .5
136
5912
.78
±10
8908
.25
2828
19 .9
1±
1215
05 .2
2↓
-0 .3
7<0
.001
*M
361T
2065
Raf
finos
eC
1753
6 .64
±20
279 .
0010
427 .
71±
9171
.20
2282
4 .99
±24
336 .
41↑
1 .13
<0 .0
01*
M36
1T16
17Sa
licin
P60
860 .
18±
5190
2 .98
4346
5 .31
±24
729 .
5477
319 .
84±
6425
8 .13
↑0 .
830 .
004*
M17
9T80
4Sa
licyl
_alc
ohol
P14
2 .47
±28
8 .50
72 .2
1±
68 .4
021
2 .73
±39
1 .13
↑1 .
560 .
039*
M20
4T11
32Sh
ikim
ic A
cid
OA
2538
3 .58
±14
920 .
9430
322 .
58±
1634
3 .72
2071
0 .12
±11
729 .
35↓
-0 .5
5<0
.001
*M
147T
682
Succ
inic
Aci
dO
A13
497 .
11±
9757
.16
1798
7 .78
±10
381 .
9191
54 .4
9±
6731
.39
↓-0
.97
<0 .0
01*
M36
1T16
93Su
cros
eC
1124
108 .
52±
3286
78 .8
910
1311
1 .69
±26
6108
.64
1229
137 .
78±
3485
13 .6
6↑
0 .28
<0 .0
01*
M14
7T92
8Th
reon
ic a
cid
OA
2239
4 .89
±11
990 .
8324
317 .
40±
1070
0 .24
2051
5 .57
±12
915 .
39↓
-0 .2
5<0
.001
*
M24
7T74
8
Thre
onic
aci
d
1,4-
lact
one
OA
685 .
46±
295 .
4582
2 .50
±29
7 .12
554 .
58±
227 .
77↓
-0 .5
7<0
.001
*
M14
7T64
5
Thym
idin
e-5’
-
mon
opho
phat
eN
3865
.36
±22
50 .4
041
28 .6
9±
2390
.89
3605
.86
±20
87 .7
3-0
.20
0 .97
1M
129T
1048
NI (
5C S
ugar
; 1)
C24
63 .6
2±
1017
.39
2303
.75
±83
3 .85
2607
.50
±11
43 .6
50 .
180 .
393
M21
7T10
15N
I (5C
Sug
ar; 2
)C
2109
.60
±18
80 .1
129
59 .1
4±
2237
.40
1305
.72
±91
5 .68
↓-1
.18
<0 .0
01*
M21
7T10
85N
I (5C
Sug
ar; 3
)C
4853
.83
±24
75 .7
245
22 .5
5±
2030
.95
5146
.13
±27
89 .9
10 .
190 .
930
M74
T101
9N
I (5C
Sug
ar; 4
)C
207 .
35±
302 .
3426
1 .53
±35
5 .67
155 .
63±
231 .
16-0
.75
0 .05
7M
200T
1035
NI (
5C S
ugar
; 5)
C63
.21
±15
1 .28
59 .3
8±
119 .
5266
.84
±17
6 .76
0 .17
0 .05
2
191
M36
4T11
95N
I (6C
Sug
ar; 1
)C
8461
.74
±39
49 .0
289
29 .4
2±
3897
.17
8019
.20
±39
67 .6
0↓
-0 .1
60 .
033*
M73
T136
5N
I (6C
Sug
ar; 2
)C
7730
.26
±28
67 .0
574
35 .1
4±
2482
.71
7992
.96
±31
60 .8
30 .
100 .
971
M11
6T62
7N
I (Am
ino
Acid
; 1)
AA10
04 .4
1±
1144
.97
974 .
87±
971 .
5710
31 .2
7±
1287
.54
0 .08
0 .93
0M
172T
844
NI (
Amin
o Ac
id; 2
)AA
459 .
23±
167 .
8650
1 .17
±15
7 .30
418 .
26±
168 .
60↓
-0 .2
60 .
001*
M21
8T58
2N
I (Am
ino
Acid
; 3)
AA33
0 .13
±32
4 .71
271 .
36±
246 .
9438
5 .71
±37
7 .08
0 .51
0 .09
5M
292T
913
NI (
Org
anic
Aci
d)O
A36
3 .84
±20
5 .57
363 .
97±
191 .
4836
3 .70
±22
0 .34
-0 .0
00 .
304
M21
7T12
34N
I (Su
gar A
lcoh
ol)
SA20
90 .8
9±
1519
.31
1982
.08
±12
81 .7
721
91 .5
1±
1710
.81
0 .14
0 .90
0M
105T
1602
NI(1
)N
I79
5 .07
±75
3 .31
788 .
33±
746 .
7180
1 .44
±76
3 .50
0 .02
0 .60
7M
173T
1167
NI (
2)N
I46
17 .4
9±
1773
.08
4841
.04
±15
54 .4
244
05 .9
5±
1942
.42
↓-0
.14
0 .00
5*M
191T
1139
NI (
3)N
I91
6 .25
±10
23 .1
110
55 .3
8±
1138
.90
783 .
90±
886 .
17↓
-0 .4
30 .
039*
M20
4T14
11N
I (4)
NI
3729
.61
±19
55 .5
032
47 .0
1±
1890
.32
4170
.25
±19
19 .2
1↑
0 .36
0 .00
1*M
204T
1442
NI (
5)N
I19
07 .5
1±
1338
.76
1447
.71
±95
7 .32
2335
.24
±14
97 .7
5↑
0 .69
0 .00
2*M
204T
1480
NI (
6)N
I74
9 .78
±46
1 .85
780 .
06±
476 .
2572
1 .13
±44
8 .50
-0 .1
10 .
057
M20
4T15
42N
I (7)
NI
1147
0 .15
±88
29 .5
710
852 .
32±
8244
.30
1201
7 .77
±93
30 .3
8↑
0 .15
0 .00
1*M
204T
1673
NI (
8)N
I18
69 .1
9±
1193
.35
1704
.32
±93
3 .29
2024
.00
±13
82 .2
00 .
250 .
132
M21
7T15
30N
i (9)
NI
4711
.25
±28
00 .2
043
06 .6
3±
2678
.62
5094
.12
±28
72 .5
4↑
0 .24
0 .02
3*M
217T
1771
NI (
10)
NI
1616
.81
±22
56 .7
686
8 .07
±64
3 .35
2224
.49
±28
48 .8
7↑
1 .36
0 .00
1*M
219T
1659
NI (
11)
NI
469 .
32±
281 .
9138
2 .76
±15
5 .86
551 .
22±
344 .
35↑
0 .53
<0 .0
01*
M22
1T99
5N
I (12
)N
I22
7 .48
±18
2 .53
221 .
02±
167 .
7823
3 .59
±19
6 .19
0 .08
0 .95
0M
253T
1990
NI (
13)
NI
2769
.39
±18
70 .8
525
86 .2
3±
1495
.32
2941
.65
±21
60 .6
10 .
190 .
304
M29
4T16
83N
I (14
)N
I32
8 .85
±29
6 .08
290 .
87±
255 .
1936
5 .97
±32
8 .44
0 .33
0 .97
1M
306T
949
NI (
15)
NI
160 .
21±
137 .
2610
7 .17
±65
.46
210 .
40±
166 .
04↑
0 .97
<0 .0
01*
M33
3T10
95N
I (16
)N
I28
9 .64
±13
6 .85
315 .
97±
139 .
3525
7 .17
±12
7 .48
-0 .3
00 .
172
M35
5T17
80N
I (17
)N
I11
96 .4
9±
902 .
5310
96 .6
8±
807 .
7212
90 .9
4±
978 .
870 .
240 .
194
M35
6T17
19N
I (18
)N
I71
1 .70
±56
9 .84
614 .
91±
515 .
9180
3 .28
±60
5 .12
↑0 .
390 .
003*
M36
1T19
97N
I (19
)N
I28
3548
.70
±10
8725
.29
2826
53 .0
1±
1116
95 .6
528
4396
.23
±10
6437
.97
0 .01
0 .87
0M
370T
1763
NI (
20)
NI
70 .6
7±
54 .4
057
.00
±38
.41
83 .4
6±
63 .5
4↑
0 .55
0 .01
7*M
373T
1980
NI (
21)
NI
6686
4 .97
±38
927 .
7865
389 .
43±
3511
7 .64
6826
1 .18
±42
362 .
490 .
060 .
524
192
M38
8T19
95N
I (22
)N
I13
772 .
62±
1538
6 .42
1372
4 .68
±13
532 .
3013
817 .
99±
1702
9 .85
0 .01
0 .97
1M
396T
1886
NI (
23)
NI
744 .
93±
664 .
7366
7 .89
±53
2 .68
817 .
01±
763 .
950 .
290 .
971
M45
6T18
69N
I (24
)N
I96
.57
±17
6 .14
89 .8
5±
203 .
3410
3 .05
±14
6 .02
0 .20
0 .17
2M
461T
1841
NI (
25)
NI
163 .
79±
175 .
3413
2 .77
±12
1 .86
191 .
64±
209 .
01↑
0 .53
0 .02
2*M
469T
2109
NI (
26)
NI
596 .
45±
582 .
4457
3 .39
±60
0 .45
618 .
21±
567 .
500 .
110 .
393
M47
6T20
51N
I (27
)N
I10
64 .4
7±
794 .
6392
9 .97
±59
7 .17
1191
.75
±92
9 .80
0 .36
0 .24
7M
523T
2058
NI (
28)
NI
147 .
51±
123 .
5514
1 .41
±10
5 .57
152 .
99±
138 .
100 .
110 .
971
M91
T153
9N
I (29
)N
I89
2 .93
±75
2 .00
790 .
84±
618 .
8698
1 .96
±84
4 .89
0 .31
0 .32
9
193
M147T569_Malonic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .361 0 .672 6 .309 2 .40E-5 ***Treatment 1 9 .325 9 .325 87 .528 2 .20E-16 ***TimePoint 1 0 .341 0 .341 3 .202 0 .076 .Genotype:Treatment 5 1 .881 0 .376 3 .531 0 .005 **Genotype:TimePoint 5 0 .578 0 .116 1 .085 0 .371Treatment:TimePoint 1 0 .300 0 .300 2 .814 0 .096 .Genotype:Treatment:TimePoint 5 0 .659 0 .132 1 .238 0 .294Residuals 151 16 .087 0 .107
M218T582_NI_Amino_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .069 0 .614 3 .400 0 .006 **Treatment 1 0 .678 0 .678 3 .755 0 .054 .TimePoint 1 0 .306 0 .306 1 .694 0 .195Genotype:Treatment 5 1 .250 0 .250 1 .385 0 .233Genotype:TimePoint 5 1 .891 0 .378 2 .095 0 .069 .Treatment:TimePoint 1 0 .344 0 .344 1 .907 0 .169Genotype:Treatment:TimePoint 5 0 .807 0 .161 0 .894 0 .487Residuals 155 27 .983 0 .181
M91T611_Benzoic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .686 0 .737 9 .203 1 .09E-7 ***Treatment 1 2 .290 2 .290 28 .594 3 .15E-7 ***TimePoint 1 0 .752 0 .752 9 .390 0 .003 **Genotype:Treatment 5 1 .016 0 .203 2 .536 0 .031 *Genotype:TimePoint 5 1 .593 0 .319 3 .978 0 .002 **Treatment:TimePoint 1 0 .892 0 .892 11 .132 0 .001 **Genotype:Treatment:TimePoint 5 0 .225 0 .045 0 .561 0 .730Residuals 155 12 .416 0 .080
M116T627_NI_Amino_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 27 .431 5 .486 10 .024 3 .01E-8 ***Treatment 1 0 .021 0 .022 0 .039 0 .843TimePoint 1 1 .439 1 .439 2 .630 0 .107Genotype:Treatment 5 5 .537 1 .107 2 .023 0 .079 .
Appendix A.6 ANOVA results: metabolite abundance. Significance: ***P < 0.001; **P < 0.01; * P
< 0.05; . P < 0.1
194
Genotype:TimePoint 5 2 .551 0 .510 0 .932 0 .462Treatment:TimePoint 1 0 .87 0 .870 1 .590 0 .209Genotype:Treatment:TimePoint 5 3 .772 0 .754 1 .378 0 .236Residuals 144 78 .815 0 .547
M147T645_Thymidine .5 . .monophophateDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .762 1 .152 8 .256 1 .12E-6 ***Treatment 1 0 .000 0 .000 0 .001 0 .971TimePoint 1 0 .617 0 .617 4 .424 0 .038 *Genotype:Treatment 5 2 .005 0 .401 2 .872 0 .018 *Genotype:TimePoint 5 1 .171 0 .234 1 .678 0 .146Treatment:TimePoint 1 1 .759 1 .759 12 .604 0 .001 ***Genotype:Treatment:TimePoint 5 0 .546 0 .109 0 .783 0 .564Residuals 113 15 .772 0 .140
M256T666_L .IsoleucineDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .491 1 .098 3 .803 0 .003 **Treatment 1 10 .425 10 .425 36 .094 1 .43E-8 ***TimePoint 1 5 .724 5 .724 19 .819 1 .68E-5 ***Genotype:Treatment 5 2 .504 0 .501 1 .734 0 .130Genotype:TimePoint 5 2 .738 0 .548 1 .896 0 .098 .Treatment:TimePoint 1 4 .456 4 .456 15 .429 0 .000 ***Genotype:Treatment:TimePoint 5 2 .745 0 .549 1 .901 0 .098 .Residuals 146 42 .168 0 .289
M142T667_L .Proline_1Df Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 4 .662 0 .932 2 .111 0 .071 .Treatment 1 0 .085 0 .085 0 .193 0 .661TimePoint 1 2 .912 2 .912 6 .593 0 .012 *Genotype:Treatment 5 7 .004 1 .401 3 .172 0 .011 *Genotype:TimePoint 5 5 .675 1 .135 2 .570 0 .032 *Treatment:TimePoint 1 0 .194 0 .194 0 .439 0 .510Genotype:Treatment:TimePoint 5 1 .375 0 .275 0 .623 0 .683Residuals 89 39 .309 0 .442
M174T677_GlycineDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .874 0 .775 3 .901 0 .003 **Treatment 1 0 .544 0 .544 2 .740 0 .101TimePoint 1 1 .193 1 .193 6 .004 0 .016 *
195
Genotype:Treatment 5 2 .990 0 .598 3 .010 0 .015 *Genotype:TimePoint 5 2 .842 0 .568 2 .862 0 .019 *Treatment:TimePoint 1 0 .417 0 .417 2 .102 0 .151Genotype:Treatment:TimePoint 5 1 .233 0 .247 1 .242 0 .296Residuals 92 18 .274 0 .199
M147T682_Succinic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .080 1 .016 22 .876 2 .20E-16 ***Treatment 1 6 .116 6 .116 137 .722 2 .20E-16 ***TimePoint 1 3 .403 3 .403 76 .616 3 .33E-15 ***Genotype:Treatment 5 1 .609 0 .322 7 .244 4 .02E-6 ***Genotype:TimePoint 5 0 .189 0 .038 0 .850 0 .517Treatment:TimePoint 1 0 .122 0 .122 2 .750 0 .099 .Genotype:Treatment:TimePoint 5 0 .442 0 .088 1 .990 0 .083 .Residuals 155 6 .884 0 .044
M254T688_CatecholDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .052 0 .210 4 .669 0 .001 ***Treatment 1 0 .494 0 .494 10 .952 0 .001 **TimePoint 1 0 .035 0 .035 0 .785 0 .377Genotype:Treatment 5 0 .804 0 .161 3 .568 0 .004 **Genotype:TimePoint 5 0 .416 0 .083 1 .846 0 .107Treatment:TimePoint 1 0 .129 0 .129 2 .851 0 .093 .Genotype:Treatment:TimePoint 5 0 .370 0 .074 1 .642 0 .152Residuals 154 6 .942 0 .045
M147T704_Glycolic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .199 1 .040 9 .964 3 .79E-8 ***Treatment 1 3 .703 3 .703 35 .486 2 .05E-8 ***TimePoint 1 0 .003 0 .003 0 .027 0 .870Genotype:Treatment 5 5 .291 1 .058 10 .141 2 .80E-8 ***Genotype:TimePoint 5 0 .187 0 .037 0 .358 0 .876Treatment:TimePoint 1 1 .014 1 .014 9 .713 0 .002 **Genotype:Treatment:TimePoint 5 0 .690 0 .138 1 .323 0 .258Residuals 137 14 .295 0 .104
M245T714_Fumaric_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 8 .223 1 .645 23 .924 2 .20E-16 ***Treatment 1 7 .873 7 .873 114 .540 2 .20E-16 ***
196
TimePoint 1 0 .248 0 .248 3 .609 0 .059 .Genotype:Treatment 5 2 .300 0 .46 6 .692 1 .14E-5 ***Genotype:TimePoint 5 0 .330 0 .066 0 .961 0 .444Treatment:TimePoint 1 0 .011 0 .011 0 .158 0 .692Genotype:Treatment:TimePoint 5 1 .163 0 .233 3 .385 0 .006 **Residuals 154 10 .586 0 .069
M188T733_L .AlanineDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 6 .420 1 .284 16 .919 2 .33E-13 ***Treatment 1 0 .000 0 .000 0 .001 0 .971TimePoint 1 0 .198 0 .198 2 .609 0 .108Genotype:Treatment 5 1 .597 0 .319 4 .207 0 .001 **Genotype:TimePoint 5 1 .164 0 .233 3 .066 0 .011 *Treatment:TimePoint 1 0 .012 0 .012 0 .153 0 .696Genotype:Treatment:TimePoint 5 0 .433 0 .087 1 .142 0 .341Residuals 157 11 .915 0 .076
M204T734_L .SerineDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 9 .801 1 .960 5 .077 0 .000 ***Treatment 1 0 .115 0 .115 0 .298 0 .586TimePoint 1 0 .004 0 .004 0 .011 0 .916Genotype:Treatment 5 5 .299 1 .060 2 .745 0 .022 *Genotype:TimePoint 5 4 .276 0 .855 2 .215 0 .057 .Treatment:TimePoint 1 0 .264 0 .264 0 .683 0 .410Genotype:Treatment:TimePoint 5 1 .748 0 .350 0 .905 0 .480Residuals 123 47 .494 0 .386
M247T748_Threonic acid 1,4-lactoneDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .248 0 .050 1 .682 0 .142Treatment 1 1 .595 1 .595 54 .058 1 .17E-11 ***TimePoint 1 0 .068 0 .067 2 .287 0 .133Genotype:Treatment 5 0 .253 0 .051 1 .715 0 .135Genotype:TimePoint 5 0 .203 0 .041 1 .376 0 .236Treatment:TimePoint 1 0 .003 0 .003 0 .113 0 .738Genotype:Treatment:TimePoint 5 0 .403 0 .081 2 .733 0 .022 *Residuals 150 4 .425 0 .030
M101T761_L .ThreonineDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 8 .979 1 .796 26 .238 2 .00E-16 ***
197
Treatment 1 0 .434 0 .434 6 .335 0 .013 *TimePoint 1 0 .064 0 .064 0 .940 0 .334Genotype:Treatment 5 0 .428 0 .086 1 .250 0 .291Genotype:TimePoint 5 0 .797 0 .159 2 .328 0 .047 *Treatment:TimePoint 1 0 .017 0 .017 0 .249 0 .619Genotype:Treatment:TimePoint 5 0 .383 0 .077 1 .119 0 .354Residuals 116 7 .939 0 .068
M179T804_Salicyl_alcoholDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 2 .576 0 .515 1 .556 0 .178Treatment 1 1 .915 1 .915 5 .785 0 .018 *TimePoint 1 0 .582 0 .582 1 .758 0 .187Genotype:Treatment 5 3 .543 0 .709 2 .140 0 .065 .Genotype:TimePoint 5 2 .547 0 .509 1 .539 0 .183Treatment:TimePoint 1 5 .18 5 .180 15 .645 0 .000 ***Genotype:Treatment:TimePoint 5 4 .208 0 .842 2 .542 0 .032 *Residuals 116 38 .405 0 .331
M172T844_NI_Amino_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .116 0 .223 8 .354 5 .58E-7 ***Treatment 1 0 .365 0 .365 13 .662 0 .000 ***TimePoint 1 0 .000 0 .000 0 .013 0 .908Genotype:Treatment 5 0 .045 0 .009 0 .337 0 .890Genotype:TimePoint 5 0 .523 0 .105 3 .913 0 .002 **Treatment:TimePoint 1 0 .006 0 .006 0 .219 0 .640Genotype:Treatment:TimePoint 5 0 .175 0 .035 1 .310 0 .263Residuals 148 3 .954 0 .027
M147T857_Malic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .807 0 .361 14 .965 5 .28E-12 ***Treatment 1 0 .264 0 .264 10 .936 0 .001 **TimePoint 1 0 .004 0 .004 0 .160 0 .690Genotype:Treatment 5 0 .892 0 .178 7 .389 3 .01E-6 ***Genotype:TimePoint 5 0 .073 0 .015 0 .607 0 .695Treatment:TimePoint 1 0 .005 0 .005 0 .199 0 .656Genotype:Treatment:TimePoint 5 0 .076 0 .015 0 .632 0 .675Residuals 157 3 .791 0 .024
M233T886_L .Aspartic_acidDf Sum Sq Mean Sq F-value Pr(>F)
198
Genotype 5 4 .149 0 .830 3 .361 0 .007 **Treatment 1 1 .561 1 .561 6 .322 0 .013 *TimePoint 1 0 .627 0 .627 2 .540 0 .113Genotype:Treatment 5 3 .549 0 .710 2 .875 0 .016 *Genotype:TimePoint 5 4 .067 0 .813 3 .294 0 .007 **Treatment:TimePoint 1 0 .118 0 .118 0 .479 0 .490Genotype:Treatment:TimePoint 5 1 .65 0 .330 1 .336 0 .252Residuals 157 38 .764 0 .247
M156T888_Pyroglutamic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 6 .202 1 .240 15 .247 3 .34E-12 ***Treatment 1 0 .286 0 .286 3 .511 0 .063 .TimePoint 1 0 .059 0 .059 0 .729 0 .395Genotype:Treatment 5 1 .255 0 .251 3 .086 0 .011 *Genotype:TimePoint 5 1 .195 0 .239 2 .938 0 .015 *Treatment:TimePoint 1 0 .020 0 .020 0 .244 0 .622Genotype:Treatment:TimePoint 5 0 .657 0 .131 1 .615 0 .159Residuals 157 12 .773 0 .081
M174T892_Butyric_acid-4-amino-n- Df Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 20 .773 4 .155 12 .753 2 .97E-10 ***Treatment 1 0 .078 0 .078 0 .238 0 .626TimePoint 1 0 .928 0 .928 2 .848 0 .094 .Genotype:Treatment 5 5 .095 1 .019 3 .128 0 .010 *Genotype:TimePoint 5 1 .045 0 .209 0 .642 0 .668Treatment:TimePoint 1 1 .044 1 .044 3 .204 0 .076 .Genotype:Treatment:TimePoint 5 1 .997 0 .399 1 .226 0 .300Residuals 143 46 .584 0 .326
M292T913_NI_Organic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .714 0 .143 0 .819 0 .538Treatment 1 0 .273 0 .273 1 .567 0 .213TimePoint 1 0 .333 0 .332 1 .906 0 .170Genotype:Treatment 5 0 .740 0 .148 0 .849 0 .518Genotype:TimePoint 5 1 .199 0 .240 1 .374 0 .238Treatment:TimePoint 1 0 .583 0 .583 3 .341 0 .070 .Genotype:Treatment:TimePoint 5 0 .874 0 .175 1 .003 0 .419Residuals 134 23 .371 0 .174
M147T928_Threonic_acid
199
Df Sum Sq Mean Sq F-value Pr(>F)Genotype 5 3 .504 0 .701 17 .129 2 .03E-13 ***Treatment 1 1 .255 1 .255 30 .668 1 .31E-7 ***TimePoint 1 0 .010 0 .010 0 .248 0 .619Genotype:Treatment 5 2 .017 0 .403 9 .859 3 .52E-8 ***Genotype:TimePoint 5 0 .077 0 .015 0 .375 0 .865Treatment:TimePoint 1 0 .361 0 .361 8 .825 0 .003 **Genotype:Treatment:TimePoint 5 0 .402 0 .080 1 .966 0 .087 .Residuals 152 6 .219 0 .041
M142T941_L .Proline_2Df Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .572 0 .314 1 .396 0 .229Treatment 1 0 .393 0 .393 1 .743 0 .189TimePoint 1 2 .806 2 .806 12 .462 0 .001 ***Genotype:Treatment 5 2 .245 0 .449 1 .994 0 .083 .Genotype:TimePoint 5 2 .001 0 .400 1 .778 0 .121Treatment:TimePoint 1 0 .398 0 .398 1 .767 0 .186Genotype:Treatment:TimePoint 5 0 .961 0 .192 0 .854 0 .514Residuals 152 34 .228 0 .225
M306T949_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .589 0 .718 5 .288 0 .000 ***Treatment 1 2 .336 2 .336 17 .203 5 .48E-5 ***TimePoint 1 0 .036 0 .036 0 .266 0 .607Genotype:Treatment 5 2 .596 0 .519 3 .824 0 .003 **Genotype:TimePoint 5 1 .107 0 .221 1 .630 0 .155Treatment:TimePoint 1 0 .015 0 .015 0 .112 0 .738Genotype:Treatment:TimePoint 5 1 .591 0 .318 2 .343 0 .044 *Residuals 157 21 .314 0 .136
M249T957_GlycerolDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .613 0 .123 4 .055 0 .002 **Treatment 1 1 .273 1 .273 42 .080 1 .26E-9 ***TimePoint 1 0 .007 0 .007 0 .225 0 .636Genotype:Treatment 5 0 .103 0 .021 0 .683 0 .637Genotype:TimePoint 5 0 .140 0 .028 0 .924 0 .467Treatment:TimePoint 1 0 .135 0 .135 4 .454 0 .037 *Genotype:Treatment:TimePoint 5 0 .167 0 .033 1 .106 0 .360Residuals 147 4 .447 0 .030
200
M246T973_L .GlutamateDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 17 .696 3 .539 7 .860 1 .28E-6 ***Treatment 1 0 .041 0 .041 0 .091 0 .763TimePoint 1 0 .1 0 .1 0 .222 0 .638Genotype:Treatment 5 5 .384 1 .077 2 .391 0 .040 *Genotype:TimePoint 5 1 .932 0 .386 0 .858 0 .511Treatment:TimePoint 1 0 .125 0 .125 0 .278 0 .599Genotype:Treatment:TimePoint 5 2 .009 0 .402 0 .892 0 .488Residuals 155 69 .791 0 .450
M192T980_L .PhenylalanineDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 17 .407 3 .481 7 .192 1 .18E-5 ***Treatment 1 0 .223 0 .223 0 .461 0 .499TimePoint 1 0 .903 0 .904 1 .867 0 .175Genotype:Treatment 5 7 .202 1 .441 2 .976 0 .016 *Genotype:TimePoint 5 3 .413 0 .683 1 .410 0 .229Treatment:TimePoint 1 0 .088 0 .088 0 .183 0 .670Genotype:Treatment:TimePoint 5 2 .737 0 .547 1 .131 0 .350Residuals 85 41 .145 0 .484
M221T995_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .572 0 .314 5 .177 0 .000 ***Treatment 1 0 .002 0 .002 0 .025 0 .874TimePoint 1 0 .004 0 .004 0 .070 0 .791Genotype:Treatment 5 1 .136 0 .227 3 .739 0 .003 **Genotype:TimePoint 5 0 .697 0 .139 2 .294 0 .048 *Treatment:TimePoint 1 0 .001 0 .001 0 .020 0 .888Genotype:Treatment:TimePoint 5 0 .250 0 .050 0 .822 0 .535Residuals 157 9 .537 0 .061
M217T1015_NI_5C_sugarDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .964 0 .793 11 .400 2 .14E-9 ***Treatment 1 6 .660 6 .660 95 .759 2 .20E-16 ***TimePoint 1 0 .154 0 .154 2 .211 0 .139Genotype:Treatment 5 3 .218 0 .644 9 .255 9 .68E-8 ***Genotype:TimePoint 5 0 .104 0 .021 0 .300 0 .912Treatment:TimePoint 1 0 .546 0 .546 7 .852 0 .006 **Genotype:Treatment:TimePoint 5 0 .525 0 .105 1 .510 0 .190Residuals 157 10 .919 0 .070
201
M74T1019_NI_5C_sugarDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 4 .764 0 .953 5 .591 9 .50E-5 ***Treatment 1 0 .823 0 .823 4 .828 0 .030 *TimePoint 1 0 .031 0 .030 0 .179 0 .673Genotype:Treatment 5 2 .052 0 .410 2 .408 0 .039 *Genotype:TimePoint 5 2 .058 0 .412 2 .415 0 .039 *Treatment:TimePoint 1 0 .073 0 .073 0 .426 0 .515Genotype:Treatment:TimePoint 5 1 .501 0 .300 1 .762 0 .124Residuals 148 25 .223 0 .170
M200T1035_NI_5C_sugarDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 15 .478 3 .096 17 .498 9 .42E-14 ***Treatment 1 0 .909 0 .909 5 .140 0 .025 *TimePoint 1 0 .010 0 .010 0 .058 0 .811Genotype:Treatment 5 4 .527 0 .905 5 .118 0 .000 ***Genotype:TimePoint 5 2 .504 0 .501 2 .831 0 .018 *Treatment:TimePoint 1 0 .01 0 .010 0 .057 0 .812Genotype:Treatment:TimePoint 5 1 .321 0 .264 1 .493 0 .195Residuals 157 27 .776 0 .177
M129T1048_NI_5C_sugarDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .332 0 .666 17 .287 1 .96E-13 ***Treatment 1 0 .042 0 .042 1 .092 0 .298TimePoint 1 0 .057 0 .057 1 .486 0 .225Genotype:Treatment 5 0 .217 0 .043 1 .126 0 .349Genotype:TimePoint 5 0 .148 0 .030 0 .77 0 .573Treatment:TimePoint 1 0 .004 0 .004 0 .112 0 .739Genotype:Treatment:TimePoint 5 0 .097 0 .019 0 .502 0 .774Residuals 147 5 .666 0 .039
M217T1085_NI_5C_sugarDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .819 0 .364 3 .763 0 .003 **Treatment 1 0 .004 0 .004 0 .039 0 .844TimePoint 1 0 .098 0 .098 1 .017 0 .315Genotype:Treatment 5 0 .409 0 .082 0 .846 0 .520Genotype:TimePoint 5 0 .794 0 .159 1 .642 0 .153Treatment:TimePoint 1 0 .039 0 .039 0 .406 0 .525Genotype:Treatment:TimePoint 5 0 .35 0 .07 0 .724 0 .607
202
Residuals 136 13 .151 0 .097
M333T1095_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 2 .515 0 .503 4 .844 0 .000 ***Treatment 1 0 .266 0 .266 2 .561 0 .112TimePoint 1 0 .045 0 .045 0 .433 0 .512Genotype:Treatment 5 1 .306 0 .261 2 .515 0 .034 *Genotype:TimePoint 5 0 .923 0 .185 1 .778 0 .123Treatment:TimePoint 1 0 .798 0 .798 7 .687 0 .007 **Genotype:Treatment:TimePoint 5 1 .132 0 .226 2 .18 0 .061 .Residuals 110 11 .423 0 .104
M204T1132_Shikimic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .412 0 .682 16 .238 6 .81E-13 ***Treatment 1 1 .283 1 .282 30 .517 1 .35E-7 ***TimePoint 1 0 .002 0 .002 0 .042 0 .839Genotype:Treatment 5 0 .436 0 .087 2 .075 0 .071 .Genotype:TimePoint 5 0 .396 0 .079 1 .883 0 .100Treatment:TimePoint 1 0 .325 0 .325 7 .738 0 .006 **Genotype:Treatment:TimePoint 5 0 .583 0 .117 2 .777 0 .020 *Residuals 157 6 .598 0 .042
M191T1139_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 4 .449 0 .890 4 .600 0 .001 ***Treatment 1 1 .113 1 .112 5 .750 0 .018 *TimePoint 1 0 .936 0 .936 4 .839 0 .030 *Genotype:Treatment 5 2 .066 0 .413 2 .135 0 .065 .Genotype:TimePoint 5 1 .247 0 .249 1 .289 0 .272Treatment:TimePoint 1 1 .234 1 .234 6 .380 0 .013 *Genotype:Treatment:TimePoint 5 1 .757 0 .351 1 .816 0 .114Residuals 136 26 .311 0 .193
M273T1143_Citric_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 8 .473 1 .695 15 .551 2 .18E-12 ***Treatment 1 3 .785 3 .785 34 .733 2 .28E-8 ***TimePoint 1 0 .738 0 .738 6 .774 0 .010 *Genotype:Treatment 5 2 .881 0 .576 5 .288 0 .000 ***Genotype:TimePoint 5 0 .663 0 .133 1 .216 0 .304Treatment:TimePoint 1 0 .262 0 .262 2 .406 0 .123
203
Genotype:Treatment:TimePoint 5 0 .835 0 .167 1 .532 0 .183Residuals 155 16 .891 0 .109
M173T1167_NI
Df Sum Sq Mean Sq F-value Pr(>F)Genotype 5 1 .663 0 .333 10 .856 5 .54E-9 ***Treatment 1 0 .313 0 .313 10 .231 0 .002 **TimePoint 1 0 .01 0 .01 0 .326 0 .569Genotype:Treatment 5 0 .189 0 .038 1 .232 0 .297Genotype:TimePoint 5 0 .136 0 .027 0 .889 0 .490Treatment:TimePoint 1 0 .019 0 .019 0 .635 0 .427Genotype:Treatment:TimePoint 5 0 .057 0 .011 0 .374 0 .866Residuals 157 4 .809 0 .031
M345T1180_Quinic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .635 0 .127 6 .902 7 .51E-6 ***Treatment 1 0 .739 0 .739 40 .149 2 .37E-9 ***TimePoint 1 0 .003 0 .003 0 .167 0 .683Genotype:Treatment 5 0 .345 0 .069 3 .751 0 .003 **Genotype:TimePoint 5 0 .045 0 .009 0 .488 0 .785Treatment:TimePoint 1 0 .025 0 .025 1 .347 0 .248Genotype:Treatment:TimePoint 5 0 .148 0 .030 1 .604 0 .162Residuals 157 2 .888 0 .018
M201T1190_Fructose_1Df Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .444 1 .089 13 .836 3 .77E-11 ***Treatment 1 0 .390 0 .390 4 .952 0 .028 *TimePoint 1 1 .523 1 .523 19 .358 2 .02E-5 ***Genotype:Treatment 5 2 .802 0 .560 7 .121 5 .14E-6 ***Genotype:TimePoint 5 1 .614 0 .323 4 .103 0 .002 **Treatment:TimePoint 1 0 .033 0 .033 0 .414 0 .521Genotype:Treatment:TimePoint 5 0 .705 0 .141 1 .792 0 .118Residuals 153 12 .040 0 .079
M364T1195_NI_6C_sugarDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 2 .965 0 .593 18 .198 3 .20E-14 ***Treatment 1 0 .200 0 .200 6 .129 0 .014 *TimePoint 1 0 .033 0 .033 1 .018 0 .314Genotype:Treatment 5 0 .591 0 .118 3 .626 0 .004 **
204
Genotype:TimePoint 5 0 .289 0 .058 1 .772 0 .122Treatment:TimePoint 1 0 .236 0 .236 7 .247 0 .008 **Genotype:Treatment:TimePoint 5 0 .102 0 .020 0 .629 0 .678Residuals 157 5 .115 0 .033
M218T1196_Fructose_2Df Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 2 .608 0 .522 24 .551 2 .20E-16 ***Treatment 1 0 .164 0 .164 7 .699 0 .006 **TimePoint 1 0 .092 0 .092 4 .309 0 .040 *Genotype:Treatment 5 0 .603 0 .121 5 .677 7 .67E-5 ***Genotype:TimePoint 5 0 .222 0 .044 2 .091 0 .069 .Treatment:TimePoint 1 0 .050 0 .050 2 .331 0 .129Genotype:Treatment:TimePoint 5 0 .166 0 .033 1 .563 0 .174Residuals 157 3 .335 0 .021
M147T1210_Glucose_1Df Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 2 .761 0 .552 10 .270 1 .57E-8 ***Treatment 1 0 .269 0 .269 4 .996 0 .027 *TimePoint 1 0 .644 0 .644 11 .982 0 .001 ***Genotype:Treatment 5 3 .890 0 .778 14 .467 1 .19E-11 ***Genotype:TimePoint 5 0 .923 0 .185 3 .435 0 .006 **Treatment:TimePoint 1 0 .407 0 .407 7 .574 0 .007 **Genotype:Treatment:TimePoint 5 0 .519 0 .104 1 .929 0 .092 .Residuals 157 8 .442 0 .054
M147T1224_Glucose_2Df Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .769 0 .754 11 .240 2 .88E-9 ***Treatment 1 0 .289 0 .289 4 .311 0 .040 *TimePoint 1 0 .947 0 .947 14 .116 0 .000 ***Genotype:Treatment 5 5 .274 1 .055 15 .729 1 .59E-12 ***Genotype:TimePoint 5 1 .486 0 .297 4 .430 0 .001 ***Treatment:TimePoint 1 0 .659 0 .659 9 .826 0 .002 **Genotype:Treatment:TimePoint 5 0 .835 0 .167 2 .489 0 .034 *Residuals 156 10 .462 0 .067
M280T1229_L .TyrosineDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 2 .12 0 .424 0 .939 0 .458Treatment 1 0 .374 0 .374 0 .829 0 .364TimePoint 1 0 .504 0 .504 1 .116 0 .293
205
Genotype:Treatment 5 2 .427 0 .485 1 .075 0 .377Genotype:TimePoint 5 1 .569 0 .314 0 .695 0 .628Treatment:TimePoint 1 1 .304 1 .304 2 .888 0 .092 .Genotype:Treatment:TimePoint 5 3 .506 0 .701 1 .553 0 .178Residuals 129 58 .244 0 .452
M217T1234_NI_Sugar_alcoholDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 7 .451 1 .490 23 .036 2 .20E-16 ***Treatment 1 0 .004 0 .004 0 .067 0 .797TimePoint 1 0 .101 0 .101 1 .558 0 .214Genotype:Treatment 5 1 .074 0 .215 3 .319 0 .007 **Genotype:TimePoint 5 0 .132 0 .026 0 .408 0 .843Treatment:TimePoint 1 0 .116 0 .116 1 .791 0 .183Genotype:Treatment:TimePoint 5 0 .267 0 .053 0 .826 0 .533Residuals 155 10 .027 0 .065
M332T1243_L .Ascorbic_acidDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .677 0 .335 2 .870 0 .017 *Treatment 1 0 .171 0 .171 1 .464 0 .228TimePoint 1 0 .052 0 .052 0 .443 0 .507Genotype:Treatment 5 0 .784 0 .157 1 .341 0 .250Genotype:TimePoint 5 0 .283 0 .057 0 .485 0 .787Treatment:TimePoint 1 0 .148 0 .148 1 .265 0 .263Genotype:Treatment:TimePoint 5 0 .734 0 .147 1 .256 0 .286Residuals 152 17 .766 0 .117
M204T1305_Glucose_3Df Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 6 .779 1 .356 9 .699 6 .12E-8 ***Treatment 1 0 .54 0 .540 3 .863 0 .051 .TimePoint 1 1 .658 1 .658 11 .861 0 .001 ***Genotype:Treatment 5 5 .242 1 .048 7 .500 3 .03E-6 ***Genotype:TimePoint 5 1 .565 0 .313 2 .238 0 .054 .Treatment:TimePoint 1 1 .066 1 .066 7 .625 0 .007 **Genotype:Treatment:TimePoint 5 1 .370 0 .274 1 .960 0 .089 .Residuals 136 19 .012 0 .140
M305T1348_Myo .inositolDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .124 0 .025 2 .618 0 .026 *Treatment 1 0 .084 0 .084 8 .873 0 .003 **
206
TimePoint 1 0 .008 0 .008 0 .806 0 .371Genotype:Treatment 5 0 .012 0 .002 0 .253 0 .938Genotype:TimePoint 5 0 .011 0 .002 0 .242 0 .943Treatment:TimePoint 1 0 .005 0 .005 0 .506 0 .478Genotype:Treatment:TimePoint 5 0 .035 0 .007 0 .737 0 .597Residuals 157 1 .483 0 .009
M73T1365_NI_6C_sugarDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .576 0 .115 3 .135 0 .010 *Treatment 1 0 .000 0 .000 0 .002 0 .967TimePoint 1 0 .001 0 .001 0 .035 0 .853Genotype:Treatment 5 0 .529 0 .106 2 .876 0 .017 *Genotype:TimePoint 5 0 .065 0 .013 0 .354 0 .879Treatment:TimePoint 1 0 .029 0 .029 0 .777 0 .380Genotype:Treatment:TimePoint 5 0 .046 0 .009 0 .249 0 .940Residuals 148 5 .442 0 .037
M204T1411_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .11 0 .222 3 .690 0 .004 **Treatment 1 0 .795 0 .795 13 .218 0 .000 ***TimePoint 1 0 .018 0 .018 0 .304 0 .582Genotype:Treatment 5 1 .033 0 .207 3 .434 0 .006 **Genotype:TimePoint 5 0 .571 0 .114 1 .900 0 .098 .Treatment:TimePoint 1 0 .134 0 .134 2 .220 0 .138Genotype:Treatment:TimePoint 5 0 .134 0 .027 0 .444 0 .817Residuals 152 9 .144 0 .060
M204T1442_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 9 .533 1 .907 16 .735 5 .60E-13 ***Treatment 1 1 .373 1 .373 12 .049 0 .001 ***TimePoint 1 0 .165 0 .165 1 .450 0 .230Genotype:Treatment 5 0 .625 0 .125 1 .098 0 .364Genotype:TimePoint 5 0 .644 0 .129 1 .130 0 .347Treatment:TimePoint 1 0 .015 0 .015 0 .129 0 .720Genotype:Treatment:TimePoint 5 0 .241 0 .048 0 .422 0 .833Residuals 142 16 .177 0 .114
M204T1480_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 6 .468 1 .294 54 .170 <2e-16 ***
207
Treatment 1 0 .114 0 .114 4 .779 0 .030 *TimePoint 1 0 .041 0 .041 1 .735 0 .190Genotype:Treatment 5 0 .156 0 .031 1 .302 0 .266Genotype:TimePoint 5 0 .043 0 .009 0 .357 0 .877Treatment:TimePoint 1 0 .008 0 .008 0 .319 0 .573Genotype:Treatment:TimePoint 5 0 .083 0 .017 0 .697 0 .627Residuals 157 3 .749 0 .024
M204T1529_MelibioseDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 8 .055 1 .611 61 .058 2 .20E-16 ***Treatment 1 0 .537 0 .537 20 .338 1 .26E-5 ***TimePoint 1 0 .015 0 .015 0 .583 0 .446Genotype:Treatment 5 0 .244 0 .049 1 .851 0 .106Genotype:TimePoint 5 0 .166 0 .033 1 .256 0 .286Treatment:TimePoint 1 0 .012 0 .012 0 .447 0 .505Genotype:Treatment:TimePoint 5 0 .018 0 .004 0 .134 0 .984Residuals 157 4 .142 0 .026
M217T1530_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .502 1 .100 13 .725 4 .07E-11 ***Treatment 1 0 .554 0 .554 6 .906 0 .009 **TimePoint 1 0 .139 0 .139 1 .729 0 .190Genotype:Treatment 5 0 .257 0 .051 0 .640 0 .669Genotype:TimePoint 5 0 .273 0 .055 0 .681 0 .639Treatment:TimePoint 1 0 .059 0 .059 0 .739 0 .391Genotype:Treatment:TimePoint 5 0 .217 0 .043 0 .542 0 .744Residuals 157 12 .588 0 .080
M91T1539_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 9 .971 1 .994 17 .160 3 .70E-13 ***Treatment 1 0 .160 0 .160 1 .38 0 .242TimePoint 1 0 .025 0 .025 0 .211 0 .647Genotype:Treatment 5 2 .252 0 .450 3 .876 0 .003 **Genotype:TimePoint 5 0 .579 0 .116 0 .997 0 .422Treatment:TimePoint 1 0 .002 0 .002 0 .016 0 .901Genotype:Treatment:TimePoint 5 0 .406 0 .081 0 .698 0 .626Residuals 137 15 .920 0 .116
M204T1542_NIDf Sum Sq Mean Sq F-value Pr(>F)
208
Genotype 5 35 .033 7 .007 119 .323 2 .20E-16 ***Treatment 1 0 .791 0 .791 13 .477 0 .000 ***TimePoint 1 0 .022 0 .022 0 .369 0 .544Genotype:Treatment 5 1 .329 0 .266 4 .527 0 .001 ***Genotype:TimePoint 5 0 .851 0 .170 2 .900 0 .016 *Treatment:TimePoint 1 0 .003 0 .003 0 .044 0 .835Genotype:Treatment:TimePoint 5 0 .103 0 .021 0 .352 0 .880Residuals 142 8 .338 0 .059
M261T1575_MaltoseDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 7 .245 1 .449 13 .609 5 .97E-11 ***Treatment 1 0 .053 0 .053 0 .497 0 .482TimePoint 1 0 .015 0 .015 0 .139 0 .710Genotype:Treatment 5 0 .486 0 .097 0 .913 0 .475Genotype:TimePoint 5 0 .625 0 .125 1 .174 0 .325Treatment:TimePoint 1 0 .001 0 .001 0 .011 0 .918Genotype:Treatment:TimePoint 5 0 .601 0 .120 1 .130 0 .347Residuals 150 15 .972 0 .106
M105T1602_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 30 .354 6 .071 61 .515 <2e-16 ***Treatment 1 0 .048 0 .048 0 .482 0 .488TimePoint 1 0 .013 0 .013 0 .135 0 .714Genotype:Treatment 5 0 .620 0 .124 1 .256 0 .286Genotype:TimePoint 5 0 .056 0 .011 0 .114 0 .989Treatment:TimePoint 1 0 .040 0 .040 0 .407 0 .524Genotype:Treatment:TimePoint 5 0 .775 0 .155 1 .571 0 .171Residuals 157 15 .494 0 .099
M361T1617_SalicinDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .668 1 .134 11 .340 2 .37E-9 ***Treatment 1 1 .061 1 .061 10 .616 0 .001 **TimePoint 1 0 .421 0 .421 4 .212 0 .042 *Genotype:Treatment 5 1 .695 0 .339 3 .391 0 .006 **Genotype:TimePoint 5 1 .771 0 .354 3 .544 0 .005 **Treatment:TimePoint 1 0 .712 0 .712 7 .123 0 .008 **Genotype:Treatment:TimePoint 5 0 .525 0 .105 1 .050 0 .391Residuals 157 15 .696 0 .100
M219T1659_NI
209
Df Sum Sq Mean Sq F-value Pr(>F)Genotype 5 1 .244 0 .249 6 .514 1 .56E-5 ***Treatment 1 0 .764 0 .764 20 .004 1 .48E-5 ***TimePoint 1 0 0 .000 0 .000 0 .988Genotype:Treatment 5 0 .611 0 .122 3 .201 0 .009 **Genotype:TimePoint 5 0 .330 0 .066 1 .729 0 .131Treatment:TimePoint 1 0 .037 0 .037 0 .964 0 .328Genotype:Treatment:TimePoint 5 0 .078 0 .016 0 .407 0 .843Residuals 157 5 .999 0 .038
M204T1673_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .533 1 .106 15 .604 4 .20E-12 ***Treatment 1 0 .220 0 .220 3 .107 0 .080 .TimePoint 1 0 .028 0 .028 0 .394 0 .531Genotype:Treatment 5 0 .153 0 .030 0 .43 0 .827Genotype:TimePoint 5 0 .322 0 .064 0 .908 0 .478Treatment:TimePoint 1 0 .032 0 .032 0 .445 0 .506Genotype:Treatment:TimePoint 5 0 .113 0 .023 0 .319 0 .901Residuals 135 9 .573 0 .071
M236T1674_AdenosineDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 15 .328 3 .066 41 .473 2 .20E-16 ***Treatment 1 0 .222 0 .222 3 .007 0 .085 .TimePoint 1 0 .095 0 .095 1 .289 0 .258Genotype:Treatment 5 1 .549 0 .310 4 .192 0 .001 **Genotype:TimePoint 5 0 .752 0 .150 2 .034 0 .077 .Treatment:TimePoint 1 0 .111 0 .111 1 .503 0 .222Genotype:Treatment:TimePoint 5 0 .385 0 .077 1 .043 0 .395Residuals 157 11 .605 0 .074
M294T1683_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 42 .363 8 .473 78 .231 2 .20E-16 ***Treatment 1 0 .001 0 .001 0 .011 0 .916TimePoint 1 0 .18 0 .180 1 .665 0 .199Genotype:Treatment 5 0 .409 0 .082 0 .755 0 .583Genotype:TimePoint 5 2 .315 0 .463 4 .274 0 .001 **Treatment:TimePoint 1 0 .02 0 .020 0 .184 0 .669Genotype:Treatment:TimePoint 5 0 .705 0 .141 1 .302 0 .266Residuals 150 16 .245 0 .108
210
M361T1693_SucroseDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .233 0 .047 3 .38 0 .006 **Treatment 1 0 .321 0 .321 23 .237 3 .36E-6 ***TimePoint 1 0 .128 0 .128 9 .266 0 .003 **Genotype:Treatment 5 0 .069 0 .014 1 .003 0 .418Genotype:TimePoint 5 0 .092 0 .018 1 .335 0 .252Treatment:TimePoint 1 0 .005 0 .005 0 .348 0 .556Genotype:Treatment:TimePoint 5 0 .082 0 .016 1 .187 0 .318Residuals 157 2 .166 0 .014
M356T1719_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 6 .154 1 .231 19 .088 8 .27E-15 ***Treatment 1 0 .751 0 .751 11 .646 0 .001 ***TimePoint 1 0 .159 0 .159 2 .458 0 .119Genotype:Treatment 5 0 .323 0 .065 1 .001 0 .419Genotype:TimePoint 5 0 .721 0 .144 2 .235 0 .053 .Treatment:TimePoint 1 0 .064 0 .064 0 .996 0 .320Genotype:Treatment:TimePoint 5 0 .288 0 .058 0 .893 0 .487Residuals 157 10 .124 0 .064
M370T1763_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .037 1 .007 7 .899 1 .18E-6 ***Treatment 1 0 .973 0 .973 7 .627 0 .006 **TimePoint 1 0 .005 0 .005 0 .041 0 .840Genotype:Treatment 5 0 .941 0 .188 1 .475 0 .201Genotype:TimePoint 5 1 .267 0 .253 1 .988 0 .083 .Treatment:TimePoint 1 0 .029 0 .029 0 .226 0 .635Genotype:Treatment:TimePoint 5 0 .178 0 .036 0 .279 0 .924Residuals 156 19 .895 0 .128
M217T1771_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .814 0 .763 6 .181 4 .87E-5 ***Treatment 1 1 .773 1 .773 14 .371 0 .000 ***TimePoint 1 0 .010 0 .010 0 .083 0 .774Genotype:Treatment 5 0 .995 0 .199 1 .613 0 .163Genotype:TimePoint 5 0 .428 0 .086 0 .694 0 .629Treatment:TimePoint 1 0 .144 0 .144 1 .168 0 .282Genotype:Treatment:TimePoint 5 0 .106 0 .021 0 .172 0 .972Residuals 101 12 .463 0 .123
211
M355T1780_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 4 .514 0 .903 15 .508 2 .19E-12 ***Treatment 1 0 .135 0 .135 2 .322 0 .130TimePoint 1 0 .009 0 .009 0 .162 0 .688Genotype:Treatment 5 0 .188 0 .038 0 .647 0 .664Genotype:TimePoint 5 0 .418 0 .084 1 .436 0 .214Treatment:TimePoint 1 0 .039 0 .039 0 .662 0 .417Genotype:Treatment:TimePoint 5 0 .049 0 .010 0 .168 0 .974Residuals 157 9 .139 0 .058
M368T1807_CatechinDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 8 .501 1 .700 16 .144 7 .91E-13 ***Treatment 1 0 .384 0 .383 3 .641 0 .058 .TimePoint 1 0 .013 0 .013 0 .122 0 .728Genotype:Treatment 5 4 .275 0 .855 8 .118 7 .75E-7 ***Genotype:TimePoint 5 1 .473 0 .295 2 .798 0 .019 *Treatment:TimePoint 1 0 .100 0 .100 0 .951 0 .331Genotype:Treatment:TimePoint 5 0 .209 0 .042 0 .397 0 .850Residuals 157 16 .535 0 .105
M461T1841_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 22 .781 4 .556 25 .334 2 .20E-16 ***Treatment 1 1 .270 1 .270 7 .059 0 .009 **TimePoint 1 0 .058 0 .058 0 .325 0 .570Genotype:Treatment 5 3 .845 0 .769 4 .276 0 .001 **Genotype:TimePoint 5 1 .441 0 .288 1 .603 0 .163Treatment:TimePoint 1 0 .064 0 .064 0 .355 0 .552Genotype:Treatment:TimePoint 5 0 .578 0 .116 0 .643 0 .667Residuals 143 25 .719 0 .180
M456T1869_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 46 .1 9 .220 21 .611 2 .81E-16 ***Treatment 1 1 .096 1 .096 2 .569 0 .111TimePoint 1 0 .276 0 .276 0 .646 0 .423Genotype:Treatment 5 19 .601 3 .920 9 .189 1 .19E-7 ***Genotype:TimePoint 5 3 .682 0 .736 1 .726 0 .132Treatment:TimePoint 1 0 .005 0 .005 0 .012 0 .914Genotype:Treatment:TimePoint 5 1 .094 0 .219 0 .513 0 .766
212
Residuals 151 64 .423 0 .427
M204T1876_GalactinolDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .019 0 .204 1 .650 0 .153Treatment 1 1 .966 1 .966 15 .922 0 .000 ***TimePoint 1 46 .155 46 .155 373 .821 2 .20E-16 ***Genotype:Treatment 5 1 .993 0 .399 3 .228 0 .010 **Genotype:TimePoint 5 6 .732 1 .346 10 .905 2 .02E-8 ***Treatment:TimePoint 1 0 .296 0 .296 2 .401 0 .124Genotype:Treatment:TimePoint 5 2 .763 0 .553 4 .475 0 .001 ***Residuals 102 12 .594 0 .123
M396T1886_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 6 .297 1 .259 8 .693 2 .73E-7 ***Treatment 1 0 .001 0 .001 0 .006 0 .941TimePoint 1 0 .033 0 .033 0 .228 0 .634Genotype:Treatment 5 0 .888 0 .178 1 .226 0 .300Genotype:TimePoint 5 1 .141 0 .228 1 .575 0 .170Treatment:TimePoint 1 0 .011 0 .011 0 .075 0 .785Genotype:Treatment:TimePoint 5 0 .601 0 .120 0 .830 0 .530Residuals 156 22 .600 0 .145
M560T1902_KaempferolDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 15 .743 3 .149 9 .183 1 .15E-7 ***Treatment 1 1 .558 1 .558 4 .545 0 .035 *TimePoint 1 0 .015 0 .015 0 .045 0 .833Genotype:Treatment 5 0 .875 0 .175 0 .510 0 .768Genotype:TimePoint 5 2 .682 0 .536 1 .565 0 .173Treatment:TimePoint 1 0 .149 0 .149 0 .435 0 .511Genotype:Treatment:TimePoint 5 1 .59 0 .318 0 .927 0 .465Residuals 154 52 .804 0 .343
M204T1938_DigalactosylglycerolDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .610 0 .122 2 .204 0 .057 .Treatment 1 0 .278 0 .278 5 .020 0 .026 *TimePoint 1 0 0 0 0 .999Genotype:Treatment 5 1 .047 0 .209 3 .784 0 .003 **Genotype:TimePoint 5 0 .336 0 .067 1 .212 0 .306Treatment:TimePoint 1 0 .001 0 .001 0 .013 0 .910
213
Genotype:Treatment:TimePoint 5 0 .137 0 .027 0 .495 0 .780Residuals 157 8 .688 0 .055
M648T1955_Quercitin
Df Sum Sq Mean Sq F-value Pr(>F)Genotype 5 11 .507 2 .301 6 .970 6 .66E-6 ***Treatment 1 3 .384 3 .384 10 .247 0 .002 **TimePoint 1 0 .004 0 .004 0 .013 0 .908Genotype:Treatment 5 1 .03 0 .206 0 .624 0 .682Genotype:TimePoint 5 2 .458 0 .492 1 .489 0 .196Treatment:TimePoint 1 0 .39 0 .39 1 .181 0 .279Genotype:Treatment:TimePoint 5 1 .132 0 .227 0 .686 0 .635Residuals 156 51 .511 0 .330
M373T1980_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 8 .983 1 .797 33 .203 2 .00E-16 ***Treatment 1 0 .037 0 .037 0 .683 0 .410TimePoint 1 0 .013 0 .013 0 .241 0 .624Genotype:Treatment 5 0 .092 0 .018 0 .340 0 .888Genotype:TimePoint 5 0 .513 0 .103 1 .897 0 .098 .Treatment:TimePoint 1 0 0 .000 0 .000 0 .989Genotype:Treatment:TimePoint 5 0 .225 0 .045 0 .833 0 .528Residuals 157 8 .495 0 .054
M253T1990_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 1 .749 0 .350 2 .898 0 .016 *Treatment 1 0 .186 0 .186 1 .542 0 .216TimePoint 1 0 .404 0 .404 3 .347 0 .069 .Genotype:Treatment 5 0 .279 0 .056 0 .463 0 .803Genotype:TimePoint 5 0 .782 0 .156 1 .296 0 .269Treatment:TimePoint 1 0 .227 0 .227 1 .878 0 .173Genotype:Treatment:TimePoint 5 0 .563 0 .113 0 .933 0 .462Residuals 139 16 .774 0 .121
M388T1995_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 22 .555 4 .511 90 .681 2 .00E-16 ***Treatment 1 0 .000 0 .000 0 .004 0 .949TimePoint 1 0 .004 0 .004 0 .088 0 .767Genotype:Treatment 5 0 .564 0 .113 2 .266 0 .051 .
214
Genotype:TimePoint 5 0 .311 0 .062 1 .251 0 .288Treatment:TimePoint 1 0 .038 0 .038 0 .758 0 .385Genotype:Treatment:TimePoint 5 0 .246 0 .049 0 .988 0 .427Residuals 157 7 .810 0 .050
M361T1997_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 0 .615 0 .123 4 .177 0 .001 **Treatment 1 0 .003 0 .003 0 .102 0 .750TimePoint 1 0 .021 0 .021 0 .715 0 .399Genotype:Treatment 5 0 .181 0 .036 1 .232 0 .297Genotype:TimePoint 5 0 .142 0 .028 0 .963 0 .442Treatment:TimePoint 1 0 .025 0 .025 0 .836 0 .362Genotype:Treatment:TimePoint 5 0 .118 0 .024 0 .800 0 .551Residuals 157 4 .621 0 .029
M476T2051_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 3 .576 0 .715 5 .915 4 .88E-5 ***Treatment 1 0 .233 0 .233 1 .925 0 .167TimePoint 1 0 .000 0 .000 0 .004 0 .952Genotype:Treatment 5 0 .425 0 .085 0 .702 0 .622Genotype:TimePoint 5 1 .112 0 .222 1 .839 0 .108Treatment:TimePoint 1 0 .009 0 .009 0 .073 0 .787Genotype:Treatment:TimePoint 5 0 .196 0 .039 0 .324 0 .898Residuals 157 18 .984 0 .121
M523T2058_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 7 .472 1 .494 5 .552 0 .000 ***Treatment 1 0 .003 0 .003 0 .011 0 .916TimePoint 1 0 .216 0 .216 0 .804 0 .371Genotype:Treatment 5 0 .578 0 .116 0 .429 0 .828Genotype:TimePoint 5 2 .724 0 .545 2 .024 0 .079 .Treatment:TimePoint 1 0 .18 0 .180 0 .667 0 .415Genotype:Treatment:TimePoint 5 1 .082 0 .216 0 .804 0 .548Residuals 145 39 .029 0 .269
M361T2065_RaffinoseDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 5 .731 1 .146 4 .762 0 .001 ***Treatment 1 9 .111 9 .111 37 .848 0 .000 ***TimePoint 1 22 .607 22 .607 93 .915 0 .0 ***
215
Genotype:Treatment 5 2 .272 0 .455 1 .888 0 .101Genotype:TimePoint 5 2 .213 0 .443 1 .839 0 .110Treatment:TimePoint 1 3 .076 3 .076 12 .780 0 .001 ***Genotype:Treatment:TimePoint 5 1 .445 0 .289 1 .200 0 .313Residuals 119 28 .646 0 .241
M469T2109_NIDf Sum Sq Mean Sq F-value Pr(>F)
Genotype 5 18 .746 3 .749 29 .105 <2e-16 ***Treatment 1 0 .141 0 .141 1 .095 0 .297TimePoint 1 0 .008 0 .008 0 .065 0 .800Genotype:Treatment 5 0 .265 0 .053 0 .411 0 .840Genotype:TimePoint 5 0 .419 0 .084 0 .651 0 .661Treatment:TimePoint 1 0 .133 0 .133 1 .030 0 .312Genotype:Treatment:TimePoint 5 0 .120 0 .024 0 .186 0 .968Residuals 147 18 .936 0 .129
216Appendix A.7 Pair-wise comparisons among genotypes for absolute magnitude log2(fold-change)
variation for the drought transcriptome. Bolded asterisks indicate significant differences according
to Bonferroni’s test (P < 0.05).
Padj (Bonferroni)
AP-947 AP-1005 <0.001 *AP-947 AP-1006 <0.001 *AP-947 AP-2278 <0.001 *AP-947 AP-2298 <0.001 *AP-947 AP-2300 0.024 *AP-1005 AP-1006 <0.001 *AP-1005 AP-2278 <0.001 *AP-1005 AP-2298 <0.001 *AP-1005 AP-2300 <0.001 *AP-1006 AP-2278 1 .000 AP-1006 AP-2298 <0.001 *AP-1006 AP-2300 <0.001 *AP-2278 AP-2298 <0.001 *AP-2278 AP-2300 <0.001 *AP-2298 AP-2300 1 .000
217
Padj (Bonferroni)
AP-947 AP-1005 1 .000AP-947 AP-1006 0 .510AP-947 AP-2278 1 .000AP-947 AP-2298 1 .000AP-947 AP-2300 1 .000AP-1005 AP-1006 0.007 *AP-1005 AP-2278 1 .000AP-1005 AP-2298 1 .000AP-1005 AP-2300 0.022 *AP-1006 AP-2278 0.024 *AP-1006 AP-2298 0 .778AP-1006 AP-2300 1 .000AP-2278 AP-2298 1 .000AP-2278 AP-2300 0 .069AP-2298 AP-2300 1 .000
Appendix A.8 Pair-wise comparisons among genotypes for absolute magnitude log2(fold-change)
variation for the drought metabolome. Bolded asterisks indicate significant differences according to
Bonferroni’s test (P < 0.05).
218Appendix A.9 GO term enrichment among transcripts that are significantly correlated with at least
one metabolite. (a) For transcripts with increased abundance in water-deficit treated samples and
(b) for transcripts with decreased abundance in water-deficit treated samples.
219
(a)
GO
Ter
mTy
peA
nnot
atio
n
Num
ber i
n R
efer
ence
Li
st
Num
ber i
n B
ackg
roun
d Li
stP-
valu
ePa
dj
GO
:001
6023
Ccy
topl
asm
ic m
embr
ane-
boun
ded
vesi
cle
1241
30 .
0001
30 .
0064
GO
:003
1988
Cm
embr
ane-
boun
ded
vesi
cle
1243
30 .
0002
10 .
0064
GO
:003
1410
Ccy
topl
asm
ic v
esic
le12
482
0 .00
053
0 .01
1G
O:0
0319
82C
vesi
cle
1250
70 .
0008
20 .
013
GO
:000
9536
Cpl
astid
4837
280 .
0012
0 .01
5G
O:0
0056
67C
trans
crip
tion
fact
or c
ompl
ex61
850 .
0035
0 .03
6G
O:0
0057
73C
vacu
ole
1059
70 .
021
0 .18
GO
:000
5739
Cm
itoch
ondr
ion
3025
750 .
025
0 .19
GO
:004
4444
Ccy
topl
asm
ic p
art
1081
1331
0 .03
60 .
25G
O:0
0352
50F
UD
P-ga
lact
osyl
trans
fera
se a
ctiv
ity33
30 .
0021
0 .17
GO
:000
4022
Fal
coho
l deh
ydro
gena
se (N
AD) a
ctiv
ity35
90 .
011
0 .33
GO
:000
3743
Ftra
nsla
tion
initi
atio
n fa
ctor
act
ivity
5176
0 .01
30 .
33G
O:0
0083
78F
gala
ctos
yltra
nsfe
rase
act
ivity
382
0 .02
60 .
51
GO
:001
6667
Fox
idor
educ
tase
act
ivity
, act
ing
on s
ulfu
r gr
oup
of d
onor
s41
940 .
066
0 .99
GO
:000
8194
FU
DP-
glyc
osyl
trans
fera
se a
ctiv
ity74
680 .
076
0 .99
GO
:000
6560
Ppr
olin
e m
etab
olic
pro
cess
526
1 .70
E-06
0 .00
038
GO
:000
6525
Par
gini
ne m
etab
olic
pro
cess
533
5 .70
E-06
0 .00
065
GO
:000
6012
Pga
lact
ose
met
abol
ic p
roce
ss43
20 .
0001
10 .
0082
GO
:000
9069
Pse
rine
fam
ily a
min
o ac
id m
etab
olic
pr
oces
s61
000 .
0001
40 .
0082
GO
:000
9821
Pal
kalo
id b
iosy
nthe
tic p
roce
ss31
50 .
0002
0 .00
91G
O:0
0064
46P
regu
latio
n of
tran
slat
iona
l ini
tiatio
n44
80 .
0005
40 .
021
GO
:000
5985
Psu
cros
e m
etab
olic
pro
cess
456
0 .00
097
0 .03
2G
O:0
0059
82P
star
ch m
etab
olic
pro
cess
479
0 .00
350 .
086
GO
:000
6573
Pva
line
met
abol
ic p
roce
ss34
10 .
004
0 .08
6G
O:0
0424
30P
indo
le a
nd d
eriv
ativ
e m
etab
olic
pro
cess
483
0 .00
410 .
086
220
GO
:004
5426
Pqu
inon
e co
fact
or b
iosy
nthe
tic p
roce
ss34
20 .
0043
0 .08
6G
O:0
0423
75P
quin
one
cofa
ctor
met
abol
ic p
roce
ss34
30 .
0046
0 .08
6
GO
:004
2401
Pce
llula
r bio
geni
c am
ine
bios
ynth
etic
pr
oces
s48
70 .
0049
0 .08
6
GO
:000
6576
Pce
llula
r bio
geni
c am
ine
met
abol
ic
proc
ess
5143
0 .00
550 .
089
GO
:000
9064
Pgl
utam
ine
fam
ily a
min
o ac
id m
etab
olic
pr
oces
s51
490 .
0065
0 .08
9G
O:0
0065
86P
indo
lalk
ylam
ine
met
abol
ic p
roce
ss34
90 .
0066
0 .08
9G
O:0
0065
68P
trypt
opha
n m
etab
olic
pro
cess
349
0 .00
660 .
089
GO
:000
5984
Pdi
sacc
harid
e m
etab
olic
pro
cess
4101
0 .00
820 .
1G
O:0
0193
18P
hexo
se m
etab
olic
pro
cess
8393
0 .01
30 .
16G
O:0
0442
62P
cellu
lar c
arbo
hydr
ate
met
abol
ic p
roce
ss16
1093
0 .01
40 .
16
GO
:003
4641
Pce
llula
r nitr
ogen
com
poun
d m
etab
olic
pr
oces
s15
1031
0 .01
80 .
2G
O:0
0060
90P
pyru
vate
met
abol
ic p
roce
ss37
70 .
022
0 .23
GO
:000
5996
Pm
onos
acch
arid
e m
etab
olic
pro
cess
9526
0 .02
40 .
24
GO
:000
9066
Pas
parta
te fa
mily
am
ino
acid
met
abol
ic
proc
ess
4145
0 .02
70 .
24G
O:0
0424
34P
indo
le d
eriv
ativ
e m
etab
olic
pro
cess
383
0 .02
70 .
24G
O:0
0098
20P
alka
loid
met
abol
ic p
roce
ss41
440 .
027
0 .24
GO
:000
9311
Pol
igos
acch
arid
e m
etab
olic
pro
cess
4151
0 .03
10 .
26
GO
:000
9081
Pbr
anch
ed c
hain
fam
ily a
min
o ac
id
met
abol
ic p
roce
ss38
90 .
032
0 .26
GO
:001
6137
Pgl
ycos
ide
met
abol
ic p
roce
ss41
590 .
036
0 .27
GO
:000
6417
Pre
gula
tion
of tr
ansl
atio
n41
590 .
036
0 .27
GO
:000
6413
Ptra
nsla
tiona
l ini
tiatio
n41
600 .
037
0 .27
GO
:000
9072
Par
omat
ic a
min
o ac
id fa
mily
met
abol
ic
proc
ess
3101
0 .04
40 .
32G
O:0
0421
80P
cellu
lar k
eton
e m
etab
olic
pro
cess
2218
860 .
046
0 .32
GO
:004
4106
Pce
llula
r am
ine
met
abol
ic p
roce
ss12
890
0 .05
0 .33
GO
:000
6520
Pce
llula
r am
ino
acid
met
abol
ic p
roce
ss12
890
0 .05
0 .33
221
GO
:000
6732
Pco
enzy
me
met
abol
ic p
roce
ss74
690 .
076
0 .48
GO
:000
9308
Pam
ine
met
abol
ic p
roce
ss14
1171
0 .08
0 .49
GO
:000
6073
Pce
llula
r glu
can
met
abol
ic p
roce
ss52
980 .
085
0 .51
222
GO
Ter
mTy
peA
nnot
atio
n
Num
ber i
n R
efer
ence
Li
st
Num
ber i
n B
ackg
roun
d Li
stP-
valu
eP ad
j
GO
:001
6023
Ccy
topl
asm
ic m
embr
ane-
boun
ded
vesi
cle
3041
32 .
.00E
-20
1 .50
E-18
GO
:003
1988
Cm
embr
ane-
boun
ded
vesi
cle
3043
37 .
60E-
202 .
80E-
18G
O:0
0314
10C
cyto
plas
mic
ves
icle
3048
21 .
50E-
183 .
60E-
17G
O:0
0319
82C
vesi
cle
3050
75 .
80E-
181 .
10E-
16G
O:0
0095
36C
plas
tid59
3728
9 .80
E-8
1 .40
E-6
GO
:004
4444
Ccy
topl
asm
ic p
art
113
1133
10 .
001
0 .00
9G
O:0
0319
78C
plas
tid th
ylak
oid
lum
en5
110
0 .00
10 .
009
GO
:003
1977
Cth
ylak
oid
lum
en5
110
0 .00
10 .
009
GO
:000
9543
Cch
loro
plas
t thy
lako
id lu
men
511
00 .
001
0 .00
9G
O:0
0095
34C
chlo
ropl
ast t
hyla
koid
1163
30 .
006
0 .04
3G
O:0
0319
76C
plas
tid th
ylak
oid
1165
50 .
008
0 .05
GO
:003
1984
Cor
gane
lle s
ubco
mpa
rtmen
t11
677
0 .00
90 .
057
GO
:000
9579
Cth
ylak
oid
1392
40 .
015
0 .08
1
GO
:000
8287
Cpr
otei
n se
rine/
thre
onin
e ph
osph
atas
e co
mpl
ex4
136
0 .01
60 .
081
GO
:004
4436
Cth
ylak
oid
part
1069
20 .
026
0 .13
GO
:000
9507
Cch
loro
plas
t33
3355
0 .04
0 .17
GO
:000
5737
Ccy
topl
asm
113
1329
30 .
040 .
17G
O:0
0056
67C
trans
crip
tion
fact
or c
ompl
ex4
185
0 .04
10 .
17G
O:0
0198
43F
rRN
A bi
ndin
g5
189
0 .01
10 .
65G
O:0
0161
68F
chlo
roph
yll b
indi
ng3
710 .
014
0 .65
GO
:001
6709
F
oxid
ored
ucta
se a
ctiv
ity, a
ctin
g on
pai
red
dono
rs, w
ith in
corp
orat
ion
or re
duct
ion
of m
olec
ular
oxy
gen,
NAD
H o
r NAD
PH
as o
ne d
onor
, and
inco
rpor
atio
n of
one
at
om o
f oxy
gen
311
20 .
044
1
GO
:000
3755
Fpe
ptid
yl-p
roly
l cis
-tran
s is
omer
ase
activ
ity3
114
0 .04
61
(b)
223
GO
:001
0876
Plip
id lo
caliz
atio
n5
322 .
90E-
60 .
001
GO
:000
9069
Pse
rine
fam
ily a
min
o ac
id m
etab
olic
pr
oces
s6
100
0 .01
60 .
008
GO
:000
6570
Pty
rosi
ne m
etab
olic
pro
cess
317
0 .00
00 .
014
GO
:000
9072
Par
omat
ic a
min
o ac
id fa
mily
met
abol
ic
proc
ess
510
10 .
001
0 .03
8G
O:0
0066
64P
glyc
olip
id m
etab
olic
pro
cess
335
0 .00
20 .
069
GO
:004
4106
Pce
llula
r am
ine
met
abol
ic p
roce
ss15
890
0 .00
20 .
069
GO
:003
4641
Pce
llula
r nitr
ogen
com
poun
d m
etab
olic
pr
oces
s16
1031
0 .00
30 .
082
GO
:000
6694
Pst
eroi
d bi
osyn
thet
ic p
roce
ss5
142
0 .00
30 .
082
GO
:000
6568
Ptry
ptop
han
met
abol
ic p
roce
ss3
490 .
005
0 .08
2G
O:0
0093
08P
amin
e m
etab
olic
pro
cess
1711
710 .
005
0 .08
2G
O:0
0065
86P
indo
lalk
ylam
ine
met
abol
ic p
roce
ss3
490 .
005
0 .08
2G
O:0
0065
20P
cellu
lar a
min
o ac
id m
etab
olic
pro
cess
1489
00 .
005
0 .08
2G
O:0
0161
25P
ster
ol m
etab
olic
pro
cess
499
0 .00
50 .
082
GO
:000
6470
Ppr
otei
n am
ino
acid
dep
hosp
hory
latio
n5
173
0 .00
80 .
11G
O:0
0422
54P
ribos
ome
biog
enes
is9
489
0 .00
90 .
11G
O:0
0068
69P
lipid
tran
spor
t5
188
0 .01
10 .
14G
O:0
0163
11P
deph
osph
oryl
atio
n5
194
0 .01
20 .
14G
O:0
0434
36P
oxoa
cid
met
abol
ic p
roce
ss22
1849
0 .01
40 .
14G
O:0
0060
82P
orga
nic
acid
met
abol
ic p
roce
ss22
1852
0 .01
40 .
14G
O:0
0197
52P
carb
oxyl
ic a
cid
met
abol
ic p
roce
ss22
1849
0 .01
40 .
14G
O:0
0464
17P
chor
ism
ate
met
abol
ic p
roce
ss3
740 .
015
0 .14
GO
:000
9073
Par
omat
ic a
min
o ac
id fa
mily
bio
synt
hetic
pr
oces
s3
730 .
015
0 .14
GO
:004
2180
Pce
llula
r ket
one
met
abol
ic p
roce
ss22
1886
0 .01
70 .
15
GO
:000
6576
Pce
llula
r bio
geni
c am
ine
met
abol
ic p
ro-
cess
414
30 .
018
0 .15
GO
:001
6126
Pst
erol
bio
synt
hetic
pro
cess
382
0 .02
0 .15
GO
:004
2434
Pin
dole
der
ivat
ive
met
abol
ic p
roce
ss3
830 .
021
0 .15
224
GO
:004
2430
Pin
dole
and
der
ivat
ive
met
abol
ic p
roce
ss3
830 .
021
0 .15
GO
:002
2613
Prib
onuc
leop
rote
in c
ompl
ex b
ioge
nesi
s9
583
0 .02
30 .
17G
O:0
0082
02P
ster
oid
met
abol
ic p
roce
ss5
233
0 .02
50 .
17G
O:0
0066
43P
mem
bran
e lip
id m
etab
olic
pro
cess
310
90 .
041
0 .28
GO
:000
8610
Plip
id b
iosy
nthe
tic p
roce
ss12
973
0 .04
50 .
29
225Appendix A.10 Pair-wise M:T Spearman correlation values in AP-1006.
226
Met
abol
iteTr
ansc
ript
Des
crip
tion
p-va
lue
Spea
rman
C
orre
latio
nM
361T
2065
_Raf
finos
ePt
p .14
60 .1
.S1_
a_at
inte
gral
mem
bran
e fa
mily
pro
tein
8 .32
E-5
-0 .8
95M
147T
682_
Succ
inic
_aci
dPt
p .19
69 .1
.A1_
atun
know
n pr
otei
n0 .
000
-0 .8
85M
361T
2065
_Raf
finos
ePt
p .20
97 .1
.S1_
s_at
CKI
1 (C
ASEI
N K
INAS
E I);
kin
ase
0 .00
0-0
.884
M20
4T18
76_G
alac
tinol
Ptp .
1398
.1 .S
1_at
TPI (
TRIO
SEPH
OSP
HAT
E IS
OM
ERAS
E); t
riose
-pho
spha
te
isom
eras
e0 .
000
-0 .8
81M
361T
2065
_Raf
finos
ePt
p .33
3 .1 .
S1_a
tzi
nc fi
nger
(AN
1-lik
e) fa
mily
pro
tein
0 .00
0-0
.876
M91
T611
_Ben
zoic
_aci
dPt
p .13
39 .1
.S1_
s_at
PRA1
.B4
(PR
ENYL
ATED
RAB
AC
CEP
TOR
1 .B
4)0 .
000
-0 .8
74M
204T
1876
_Gal
actin
olPt
p .24
9 .1 .
S1_a
tSE
C (s
ecre
t age
nt);
trans
fera
se, t
rans
ferri
ng g
lyco
syl g
roup
s0 .
000
-0 .8
73M
204T
1876
_Gal
actin
olPt
p .14
60 .1
.S1_
a_at
inte
gral
mem
bran
e fa
mily
pro
tein
0 .00
0-0
.873
M36
1T16
17_S
alic
inPt
p .13
39 .1
.S1_
s_at
PRA1
.B4
(PR
ENYL
ATED
RAB
AC
CEP
TOR
1 .B
4)0 .
000
-0 .8
72M
361T
2065
_Raf
finos
ePt
p .24
9 .1 .
S1_a
tSE
C (s
ecre
t age
nt);
trans
fera
se, t
rans
ferri
ng g
lyco
syl g
roup
s0 .
000
-0 .8
67
M91
T611
_Ben
zoic
_aci
dPt
p .31
63 .1
.A1_
atBC
CP2
(BIO
TIN
CAR
BOXY
L C
ARR
IER
PR
OTE
IN 2
); bi
otin
bi
ndin
g0 .
000
-0 .8
67M
256T
666_
L .Is
oleu
cine
Ptp .
1501
.1 .A
1_s_
atM
YB4;
DN
A bi
ndin
g 0 .
000
-0 .8
63M
204T
1876
_Gal
actin
olPt
p .26
33 .1
.S1_
atAT
GLX
1 (G
LYO
XALA
SE I
HO
MO
LOG
); la
ctoy
lglu
tath
ione
lyas
e0 .
000
-0 .8
54M
179T
804_
Salic
yl_a
lcoh
olPt
p .15
86 .1
.S1_
atLS
H10
(LIG
HT
SEN
SITI
VE H
YPO
CO
TYLS
10)
0 .00
0-0
.850
M20
4T18
76_G
alac
tinol
Ptp .
2097
.1 .S
1_s_
atC
KI1
(CAS
EIN
KIN
ASE
I); k
inas
e0 .
001
-0 .8
45M
204T
1876
_Gal
actin
olPt
p .33
03 .1
.S1_
atC
KB1;
pro
tein
kin
ase
regu
lato
r0 .
001
-0 .8
42M
361T
1617
_Sal
icin
Ptp .
1346
.1 .A
1_at
mev
alon
ate
diph
osph
ate
deca
rbox
ylas
e, p
utat
ive
0 .00
1-0
.842
M14
7T68
2_Su
ccin
ic_a
cid
Ptp .
2056
.1 .S
1_s_
atC
LASP
(CLI
P-AS
SOC
IATE
D P
RO
TEIN
); bi
ndin
g0 .
001
-0 .8
41M
147T
682_
Succ
inic
_aci
dPt
p .10
64 .1
.A1_
atpo
lyga
lact
uron
ase,
put
ativ
e / p
ectin
ase,
put
ativ
e0 .
001
-0 .8
39M
204T
1876
_Gal
actin
olPt
p .33
3 .1 .
S1_a
tzi
nc fi
nger
(AN
1-lik
e) fa
mily
pro
tein
0 .00
1-0
.839
M20
4T18
76_G
alac
tinol
Ptp .
1271
.3 .S
1_a_
atAT
SC35
; RN
A bi
ndin
g 0 .
001
-0 .8
37M
256T
666_
L .Is
oleu
cine
Ptp .
3510
.1 .S
1_at
DN
A bi
ndin
g0 .
001
-0 .8
36M
256T
666_
L .Is
oleu
cine
Ptp .
1339
.1 .S
1_s_
atPR
A1 .B
4 (P
REN
YLAT
ED R
AB A
CC
EPTO
R 1
.B4)
0 .00
1-0
.829
M17
9T80
4_Sa
licyl
_alc
ohol
Ptp .
1404
.1 .S
1_at
NTM
C2T
2 .1
0 .00
1-0
.828
M36
1T20
65_R
affin
ose
Ptp .
3337
.1 .S
1_s_
atun
know
n pr
otei
n0 .
001
-0 .8
28M
91T6
11_B
enzo
ic_a
cid
Ptp .
1346
.1 .A
1_at
mev
alon
ate
diph
osph
ate
deca
rbox
ylas
e, p
utat
ive
0 .00
1-0
.825
227
M20
4T18
76_G
alac
tinol
Ptp .
1489
.2 .S
1_s_
atAt
RAB
A5d
(Ara
bido
psis
Rab
GTP
ase
hom
olog
A5d
); G
TP
bind
ing
0 .00
1-0
.823
M25
6T66
6_L .
Isol
euci
nePt
p .15
88 .1
.S1_
s_at
PIP2
;5 (P
LASM
A M
EMBR
ANE
INTR
INSI
C P
RO
TEIN
2;5
); w
ater
ch
anne
l0 .
001
-0 .8
21M
306T
949_
NI
Ptp .
1501
.1 .A
1_s_
atM
YB4;
DN
A bi
ndin
g 0 .
001
-0 .8
21M
147T
682_
Succ
inic
_aci
dPt
p .11
40 .1
.A1_
atph
otos
yste
m II
reac
tion
cent
er P
sbP
fam
ily p
rote
in0 .
001
-0 .8
20M
91T6
11_B
enzo
ic_a
cid
Ptp .
1416
.1 .A
1_s_
atFK
BP15
-2; F
K506
bin
ding
0 .
001
-0 .8
20
M20
4T18
76_G
alac
tinol
Ptp .
2026
.1 .S
1_s_
atEL
F5A-
1 (E
UKA
RYO
TIC
ELO
NG
ATIO
N F
ACTO
R 5
A-1)
; tra
nsla
tion
initi
atio
n fa
ctor
0 .00
1-0
.818
M91
T611
_Ben
zoic
_aci
dPt
p .35
10 .1
.S1_
atD
NA
bind
ing
0 .00
1-0
.816
M14
7T70
4_G
lyco
lic_a
cid
Ptp .
3022
.1 .A
1_s_
atH
AT22
; tra
nscr
iptio
n fa
ctor
0 .00
1-0
.813
M36
1T20
65_R
affin
ose
Ptp .
2629
.1 .S
1_s_
atse
nesc
ence
-ass
ocia
ted
prot
ein-
rela
ted
0 .00
1-0
.812
M25
6T66
6_L .
Isol
euci
nePt
p .13
23 .1
.S1_
atD
EAD
box
RN
A he
licas
e, p
utat
ive
0 .00
1-0
.809
M20
4T18
76_G
alac
tinol
Ptp .
2716
.1 .S
1_at
glut
ared
oxin
, put
ativ
e0 .
001
-0 .8
09M
91T6
11_B
enzo
ic_a
cid
Ptp .
1510
.1 .S
1_s_
atpr
otea
se in
hibi
tor
0 .00
1-0
.808
M36
1T20
65_R
affin
ose
Ptp .
1549
.1 .S
1_at
UBP
7 (U
BIQ
UIT
IN-S
PEC
IFIC
PR
OTE
ASE
7); u
biqu
itin
thio
lest
eras
e0 .
001
-0 .8
08M
91T6
11_B
enzo
ic_a
cid
Ptp .
2462
.1 .S
1_at
ribos
omal
pro
tein
L5
fam
ily p
rote
in0 .
002
-0 .8
07
M20
4T18
76_G
alac
tinol
Ptp .
2536
.1 .A
1_at
NF-
YB11
(NU
CLE
AR F
ACTO
R Y
, SU
BUN
IT B
11);
trans
crip
tion
fact
or0 .
002
-0 .8
05M
361T
2065
_Raf
finos
ePt
p .12
71 .3
.S1_
a_at
ATSC
35; R
NA
bind
ing
0 .00
2-0
.805
M20
4T18
76_G
alac
tinol
Ptp .
321 .
1 .S1
_a_a
tAT
HVA
22E
0 .00
2-0
.804
M20
4T18
76_G
alac
tinol
Ptp .
1267
.1 .S
1_x_
atid
entic
al p
rote
in b
indi
ng
0 .00
2-0
.803
M20
4T18
76_G
alac
tinol
Ptp .
1886
.1 .S
1_at
CIB
1 (C
RYPT
OC
HR
OM
E-IN
TER
ACTI
NG
BAS
IC-H
ELIX
-LO
OP-
HEL
IX 1
); D
NA
bind
ing
0 .00
20 .
801
M24
9T95
7_G
lyce
rol
Ptp .
1075
.1 .A
1_a_
atun
know
n pr
otei
n0 .
002
0 .80
3M
247T
748_
Thre
onic
.ac
id .1
.4 .la
cton
e . .2
TMS .
. .tra
ns .
Ptp .
1323
.1 .S
1_at
DEA
D b
ox R
NA
helic
ase,
put
ativ
e0 .
002
0 .80
6M
247T
748_
Thre
onic
.ac
id .1
.4 .la
cton
e . .2
TMS .
. .tra
ns .
Ptp .
1075
.1 .A
1_a_
atun
know
n pr
otei
n0 .
001
0 .80
8
228
M14
7T70
4_G
lyco
lic_a
cid
Ptp .
3043
.1 .S
1_s_
at4C
L2 (4
-CO
UM
ARAT
E:C
OA
LIG
ASE
2); 4
-cou
mar
ate-
CoA
lig
ase
0 .00
10 .
809
M14
7T68
2_Su
ccin
ic_a
cid
Ptp .
3348
.2 .A
1_a_
atps
eudo
urid
ine
synt
hase
fam
ily p
rote
in0 .
001
0 .80
9M
345T
1180
_Qui
nic_
acid
Ptp .
1308
.1 .S
1_at
calm
odul
in-re
late
d pr
otei
n, p
utat
ive
0 .00
10 .
812
M14
7T70
4_G
lyco
lic_a
cid
Ptp .
2269
.1 .S
1_s_
atba
nd 7
fam
ily p
rote
in0 .
001
0 .81
4
M36
1T16
93_S
ucro
sePt
p .29
22 .1
.S1_
atAT
CAP
1 (A
RAB
IDO
PSIS
TH
ALIA
NA
CYC
LASE
ASS
OC
IATE
D
PRO
TEIN
1);
actin
bin
ding
0 .00
10 .
819
M14
7T68
2_Su
ccin
ic_a
cid
Ptp .
333 .
1 .S1
_at
zinc
fing
er (A
N1-
like)
fam
ily p
rote
in0 .
001
0 .82
3M
147T
704_
Gly
colic
_aci
dPt
p .13
08 .1
.S1_
atca
lmod
ulin
-rela
ted
prot
ein,
put
ativ
e0 .
001
0 .82
4M
147T
682_
Succ
inic
_aci
dPt
p .20
97 .1
.S1_
s_at
CKI
1 (C
ASEI
N K
INAS
E I);
kin
ase
0 .00
10 .
825
M24
7T74
8_Th
reon
ic .
acid
.1 .4
.lact
one .
.2TM
S . . .t
rans
.Pt
p .13
82 .1
.A1_
atrib
osom
al p
rote
in-re
late
d0 .
001
0 .82
6M
204T
1876
_Gal
actin
olPt
p .19
58 .1
.S1_
s_at
unkn
own
prot
ein
0 .00
10 .
829
M14
7T68
2_Su
ccin
ic_a
cid
Ptp .
2536
.1 .A
1_at
NF-
YB11
(NU
CLE
AR F
ACTO
R Y
, SU
BUN
IT B
11);
trans
crip
tion
fact
or0 .
001
0 .83
0M
247T
748_
Thre
onic
.ac
id .1
.4 .la
cton
e . .2
TMS .
. .tra
ns .
Ptp .
3653
.1 .S
1_at
OR
P1C
(OSB
P(O
XYST
ERO
L BI
ND
ING
PR
OTE
IN)-R
ELAT
ED
PRO
TEIN
1C
); ox
yste
rol b
indi
ng
0 .00
10 .
831
M14
7T68
2_Su
ccin
ic_a
cid
Ptp .
2271
.1 .S
1_s_
atun
know
n pr
otei
n0 .
001
0 .83
3M
204T
1876
_Gal
actin
olPt
p .11
40 .1
.A1_
atph
otos
yste
m II
reac
tion
cent
er P
sbP
fam
ily p
rote
in0 .
001
0 .83
5M
147T
682_
Succ
inic
_aci
dPt
p .14
60 .1
.S1_
a_at
inte
gral
mem
bran
e fa
mily
pro
tein
0 .00
10 .
838
M14
7T68
2_Su
ccin
ic_a
cid
Ptp .
321 .
1 .S1
_a_a
tAT
HVA
22E
0 .00
10 .
840
M14
7T68
2_Su
ccin
ic_a
cid
Ptp .
2716
.1 .S
1_at
glut
ared
oxin
, put
ativ
e0 .
000
0 .85
5M
204T
1876
_Gal
actin
olPt
p .21
36 .1
.S1_
atzi
nc fi
nger
(C3H
C4-
type
RIN
G fi
nger
) fam
ily p
rote
in0 .
000
0 .85
6M
361T
1617
_Sal
icin
Ptp .
341 .
1 .A1
_at
unkn
own
prot
ein
0 .00
00 .
867
M20
4T18
76_G
alac
tinol
Ptp .
1064
.1 .A
1_at
poly
gala
ctur
onas
e, p
utat
ive
/ pec
tinas
e, p
utat
ive
0 .00
00 .
869
M36
1T20
65_R
affin
ose
Ptp .
1140
.1 .A
1_at
phot
osys
tem
II re
actio
n ce
nter
Psb
P fa
mily
pro
tein
0 .00
00 .
871
M14
7T68
2_Su
ccin
ic_a
cid
Ptp .
249 .
1 .S1
_at
SEC
(sec
ret a
gent
); tra
nsfe
rase
, tra
nsfe
rring
gly
cosy
l gro
ups
0 .00
00 .
885
M24
9T95
7_G
lyce
rol
Ptp .
1779
.1 .S
1_at
glyc
osyl
hyd
rola
se fa
mily
17
prot
ein
8 .67
E-5
0 .89
4M
101T
761_
L .Th
reon
ine
Ptp .
3215
.2 .S
1_at
RN
A-bi
ndin
g pr
otei
n 45
(RBP
45),
puta
tive
0 .00
00 .
894
Top Related