This is to certify that the
dissertation entitled

Comparative and Functional Genomics Approaches to
Understand Environmental Adaptation in Woody
Perennial, Populus

presented by

Sunchung Park

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Plant Breeding and Genetics

 

 

/ ’ t‘ érofessor’s Signature

Majo
//z/>./o,¢

Date

MSU is an Affirmative Action/Equal Opportunity Institution

0-0-0-0-0-0-0—¢-O-I-I-n-o-u-Q-o-.-n—I-o-l-o-n-c-o-o-Q-O-.-o-g— .-.—. -

 

- .----...—.-.—.—._.—o—.—.—----o-.-o

 

 

LIBRARY
Michigan State
University

PLACE iN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p65~p.15

 

COMPARATIVE AND FUNCTIONAL GENOMICS APPROACHES TO
UNDERSTAND ENVIRONMENTAL ADAPTATION IN WOODY PERENNIAL,
POPULUS

By

Sunchung Park

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirement
for the degree of

DOCTOR OF PHILOSOPHY

Plant Breeding and Genetics - Forestry

2004

ABSTRACT
COMPARATIVE AND FUNCTIONAL GENOMICS APPROACHES TO
UNDERSTAND ENVIRONMENTAL ADAPTATION IN WOODY PERENNIAL,
POPULUS
By

Sunchung Park

In order to survive various environmental changes over their long life span, woody
perennials have evolved several different physiological properties compared with annual
herbaceous plants such as vegetative dormancy, seasonal nutrient reallocation, and
extensive frost hardiness. Despite these distinguishable properties, most of our
knowledge on the molecular basis underlying these processes comes from extrapolation
based on findings from annual plant Studies. Since poplar ESTS and genomic resources as
tree model plant have now been available, I have taken comparative and functional
genomic approaches to understand the genetic regulation of these properties. First I have
identified a set of 25,282 unigenes from 105,831 poplar ESTS Obtained from public
databases. I then estimated poplar multigene families. Several multigene families had a
higher copy number in poplar than in Arabidopsis, implying potential lineage-specific
proliferation of poplar protein families. The protein families with such expansion may
reflect the adaptation to extensive environmental stresses in perennial poplar. In addition,
comparison of poplar unigenes with the Arabidopsis transcriptomes revealed that genes
involved in transcriptional regulation are the most divergent while metabolism-related
genes are most conserved. Second, to understand genetic regulation underlying seasonal
growth, I carried out a series of global transcriptional analysis using stem samples from

poplar trees grown in field and under controlled environment conditions. The results

showed that extensive metabolic switch and alteration in cellular functions occurred
during transition from rest to quiescent. The genes involved in the pentose phosphate
pathway were elevated and defense-related genes were overrepresented during early
winter whereas the genes involved in fermentation and fatty acid B-Oxidation were
upregulated and signaling-related genes were overrepresented during late winter.
Furthermore, clustering of these genes into eight groups has revealed that plants regulate
seasonal growth by integrating environmental factors with developmental stages. For
example, Short daylength could modulate the cold signal to achieve maximum cold
hardiness. This mechanism may help plants to control cold hardiness more closely timed
with seasonality than unpredictable temperature alone does. Finally, in this study, an
auxin-repressed gene from black locust (Robim'a pseudoacacia) was investigated. The

gene is posttranscriptionally regulated and negatively associated with shoot elongation.

ACKNOWLEDGEMENTS
This study was supported by Department of Forestry, Plant Breeding&Genetics, Genome
Technology Support Facility, and Tree Research Centr in Michigan State University. I
thank all the staff for their kind assistance.

I want to thank several people who have contributed to this work in one way or
another. First of all, I would like to thank to my supervisor, Dr. Kyung—Hwan Han for
guiding and advising me during the course of the study. I also thank my guaidance
committee, Dr. Steve van Nocker, Dr. Rebecca Grumet and and Dr. Dean Dellapenna for
their helpful comments, encouragement, and positive criticism. Dr. Daniel E. Keathley is
acknowledged for his careful work in reviewing the manuscript and coauthoring one
paper. As well, Merilyn Ruthig are acknowledged for fruitful editing manuscripts and her
friendship. Especially, I want to thank Dr. Joanne H.Whallon for giving me the
opportunity to serve as teaching assistant and her support. I thank my colleagues at the
Lab for sharing every day life.

Above all, I wish to express my deepest gratitude to my wife, Sookyung for
rewarding collaboration, invaluable support and coauthoring one paper. Certainly, I

would like to thank my family for their interest and support to my work and life.

iv

TABLE OF CONTENTS

 

 

 

 

 

LIST OF FIGURES VII
INTRODUCTION
References ..................................................................................................................... 1 1
CHAPTER 1 15
Abstract ......................................................................................................................... 15
Introduction ................................................................................................................... 16
Materials and Methods .................................................................................................. 17
Results and Discussion .................................................................................................. 20
Conclusion ..................................................................................................................... 30
References ..................................................................................................................... 32
Web Site References ..................................................................................................... 36
CHAPTER 2 46
Abstract ................................................................................................................. . ........ 46
Introduction ................................................................................................................... 48
Materials and Methods .................................................................................................. 50
Results ........................................................................................................................... 54
Discussion ..................................................................................................................... 64
References ..................................................................................................................... 70
CHAPTER 3 95
Abstract ......................................................................................................................... 95
Introduction ................................................................................................................... 97
Materials and Methods .................................................................................................. 99
Results ......................................................................................................................... 102
Discussion ................................................................................................................... 108
References ................................................................................................................... 1 12
CONCLUSION 125

 

LIST OF TABLES

Table 1-1. Poplar ESTS and unigene set statistics. ........................................................... 44

Table 1-2. The 14 Highest copy number poplar unigene families determined by
tBLASTX with an E-Value threshold of <1.0E-20. ....................................... 45

Table 2-1. Eight groups of genes identified by K-means clustering as shown in Figure .2-

9 ....................................................................................................................... 87

vi

LIST OF FIGURES

Figure 1-1. Phylogenetic relationships for the nine plant Species used in the sequence
analyses. .......................................................................................................... 37
Figure 1-2. Distribution of sequence length (bp) of singletons and contigs that constitute
the poplar unigene set .................................................................................... 38
Figure 1-3. Functional categorization of the 14,291 poplar unigenes that had tBLASTX
matches to Arabidopsis transcriptome with E-value threshold of $1 .0E-20. . 39
Figure 1-4. Gene copy number distribution of putative poplar gene families estimated by
tBLASTX search against the poplar unigene data set itself. .......................... 40
Figure 1—5. Percentage of poplar unigenes belonging to each functional category with
different E-value thresholds. ................................................................... ....... 41
Figure 1-6. Functional categorization of the 7,760 poplar unigenes that had counterparts
in all three groups (dicots monocots and pine) of plant species. .................... 42
Figure 1-7. Phylogenic tree of DNA sequences of peptidylprolyl isomerase family,
constructed using the neighbor-joining method .............................................. 43
Figure 2-1. Microarray experimental design. ................................................................... 75
Figure 2-2. The temperature, daylength and tree growth change during the experiment
period (i.e., July 2002 to June 2003) ............................................................... 76
Figure 2-3. Hierarchical clustering (A) and Functional categorization (B) of genes based
on the seasonal variation of bark. ................................................................... 77
Figure 2-4. Hierarchical clustering (A) and Functional categorization (B) of genes based
on seasonal variation of xylem. ...................................................................... 78
Figure 2-5. Magnified view of of the bark cluster A genes as shown Figure 2-3. ........... 79
Figure 2-6. Magnified view of the genes of the bark cluster B as shown Figure 2-3. ...... 81
Figure 2-7. The whole plant picture and stem sections of the poplar trees treated under
controlled environments .................................................................................. 83
Figure 2-8. Dendrogram showing relationships between seasonal changes and controlled

environmental treatments in bark (A) and xylem tissue (B). .......................... 84

vii

Figure 2-9. K-means clusters based on expression profiles of both annual cycle and
controlled environments .................................................................................. 85
Figure 2-10. Validation of rnicroarray results by Northern blot analysis. ........................ 86
Figure 3-1. Comparison and evolutionary tree of the predicted amino acid sequence of
RpARP with other plant ARPs. .................................................................... 116
Figure 3-2. Analysis of RpARP mRN A expression in response to auxin and
cycloheximide. .............................................................................................. 1 17

Figure 3—3. Comparison of RpARP transcript levels under various hormonal and abiotic

conditions. ..................................................................................................... l 18
Figure 3-4. Fusicoccin and acidic pH can repress RpARP mRN A expression. ............. 119
Figure 3-5. Repression of RpARP mRN A is correlated with elongation Stimulating

phenotype. ..................................................................................................... 120
Figure 3-6. Effects Of sucrose and low temperature on auxin suppression on RpARP

mRN A expression. ........................................................................................ 121
Figure 3-7. The expression of RpARP gene is regulated posttranscriptionally .............. 122
Figure 3-8. Nucleotide sequence of RpARP promoter region. ....................................... 123

Figure 3-9. Expression pattern of RpARP mRN A is correlated negatively with the extent
of elongation of hypocotyl. ........................................................................... 124

viii

INTRODUCTION

Trees represent the majority of terrestrial biomass production, providing a range of
ecological services, such as carbon sequestration, bioremediation, nutrient cycling and
biofilteration. Trees are also economically important by producing a variety of wood-
based products including timber, pulp and paper. In addition, trends in developed
countries to reduce their use of non-renewable resources and to conserve the environment
are likely to increase the use of products generated from renewable resources such as
wood (Bhalerao et al., 2003; Wullschleger et al., 2002)

Despite the economical and ecological importance of trees, our understanding at
the molecular level of tree development and adaptation to environments is very limited,
compared to that in herbaceous plants. Most of what we know about the regulation of
these processes comes from research on Arabidopsis thaliana, a small herbaceous annual
plant. Until recently complex biological questions of these types could not be reasonably
addressed in woody plants. However, the development of Populus genomic resources
(e. g., its complete genome sequences and a large number of ESTS from various tissues)
and the recent advances in genomic and bioinfomatic tools allow us to take integrated

functional and comparative genomics approaches to address these questions (Brunner et

al., 2004).

Populus as a model woody plant
Trees are unique among plants since they can achieve extreme longevity and great size.
Bristlecone pine (Pinus longaeva), for example, can live to be more than 4700 years Old,

and coast redwood (Sequoia sempervirens) and giant sequoia (Sequoiadendron

giganteum can exceed 100 m in height and 25 m in circumference, respectively. Because
of their great potential size and longevity, trees differ developmentally in so many
respects from herbaceous plants, including patterns of apical dominance, perennial
growth habit, secondary xylem (wood) formation, mature and juvenile phase change,
dormancy cycles, seasonal nutrient reallocation and reproduction. Consequently, tree
development is unlikely to be understood based solely on information from herbaceous
systems.

Arabidopsis thaliana has been widely accepted as a model plant in plant biology
since it has a small genome, rapid life cycle and various value in experimental studies at
the physiological and biochemical levels (Goodman et al., 1995). However, it has often
been questioned whether arabidopsis can serve as the model for tree biology. For
example, secondary xylem formation in Arabidopsis could be used to study wood
formation in tree but it is essential that results be placed in the context provided by
parallel studies with trees (Chaffey et al., 2002; Oh et al., 2003). Another such example
could be the annual growth cycle, which is essential for perennial growth of woody
plants. The complex regulation patterns of the annual growth cycle involve a complex of
interactions between environmental Signals (including day length and temperature) and
plant signal transduction pathways. It is hard to imagine that Arabidopsis could be
adequate for the study of such processes (Taylor, 2002). Thus, to understand further
unique development and adaptation processes that occur in trees, a ‘model’ tree is
needed.

Populus has been accepted as a ‘model’ woody plant to complement the genetic

resources being developed in Arabidopsis. The genus Populus (poplars, cottonwoods and

aspens) belongs to the Salicaceae family and is comprised of about 30 species, all found
in the Northern hemisphere and exhibiting some of the fastest growth rates observed in
temperate trees. Poplar provides an attractive model system for several reasons. First, and
most important, are the recent completion of the Populus genome sequence and a large
number of expressed sequence tags (ESTS) have been produced and analyzed from
various tissues of Populus species (Hertzberg et al., 2001; Kohler et al., 2003; Sterky et
al., 1998). Other reasons include the relatively small genome size (ca. 550 Mbp, only 4x
larger than Arabidopsis but 40x smaller than pine), the large number of molecular genetic
maps and the relative ease of genetic transformation. Populus can also be propagated
vegetatively, making mapping populations immortal and facilitating the production of
large amounts of clonal materials for experimentation. In addition, the genus Populus has
genuine commercial value as a tree for timber, plywood, pulp and paper (Dinus et al.,

2001)

F unctionaLgenomics approach

Because of their longevity, trees possess various means of adapting to environmental
changes. While fundamental cellular and physiological aspects of gene function in plants
can be far more precisely analyzed in Arabidopsis, rice and other annual model plants, a
key challenge for poplar research is how to conduct affordable experiments that enable
gene-level inferences to be made about these whole organism-level attributes. Until
recently complex biological question of these types could not be reasonably addressed in
woody plants. Today, genomics is, however, providing tools that allow examination of

whole-plant responses at single—gene resolution, thus providing a bridge between

molecular biology and whole-plant physiology. This will provide an integrated view of
tree responses at the finest scale (Matsuyama et al., 2002; Seki et al., 2001). As a result,
genomics may provide a mechanism for translating the understanding of cellular
processes to system-level understanding of tree growth.

On the other hand, the study of woody plants presents many challenges to the use
of the conventional genetic methods such as the self-crossing, inbreeding lines and
recessive mutant screening due to the long generation time and long time-to-flower, even
though physiological and biochemical experiments are more readily conducted. AS a
result, there are no well-characterized mutant collections of forest trees. Therefore,
reverse genetics, in which a gene sequence is first obtained and then its function. learned
via directed alteration in transgenic, is the route taken in most transgenic experiments
with trees. However, the efficiency of reverse genetic method is highly depending on the
potential of the selected candidate genes, and the candidate genes should be carefully
selected for functional evaluation. Functional genomic approaches could facilitate the
selection process of the candidate genes by providing the single gene resolution in global

scale.

Comparative Genomic approach

The real power of genomics lies in the fact that all known organisms Share the same
genetic code. Sequences for many genes have been largely conserved through
evolutionary time, so that information gained about the functioning of genes in one
organism can be applied to distantly related organisms. For example, the functions of

more than half the genes in Arabidopsis were inferred from experimental data gathered in

animal, yeast and microbial model species (Somerville and Somerville 1999).
Furthermore, comparative analyses of the sequences of expressed genes have provided a
powerful means for gene discovery (e.g., the identification of species-specific genes and
multigene family) as well as the study of the molecular basis of plant diversity (Aravind
et al., 2000; Lespinet et al., 2002; Riechmann et al., 2000).

Recently, the complete genome sequence of Arabidopsis thaliana and rice (Oryza
sativa L.) has been available (Eckardt 2000) and efforts have been initiated to sequence
the genomes of other plants, including Medicago truncatula, a model legume (Frugoli
and Harris 2001). In addition, large numbers of ESTS are being generated from a wide
variety of plant species and are publicly available (http://www.tigr.orgitdb/tgi/-
plant.shtml).

Since most of genomic resouces are generated from herbaceous plants, the
physiological properties unique to perennial woody plants could be addressed through
comparative genomics. It can also be expected that, through comparative genomics, many
of the rapid advances in understanding structural and regulatory genes in other model
organisms can be directly translated to poplar and other tree Species. And the reverse will
also be true; research with Populus will aid in understanding the genetic regulation on

metabolic processes in other plants.

Annual growth cycle in woody perennials
Trees growing in temperate climates are subjected to seasonal environmental variations
such as chilling/frost and drought. Thus, temperate woody perennials have evolved

annual growth cycles that involve alternation between active growth and vegetative

dormancy (Arora et al 2003). The annual growth cycle Of woody perennials is closely
timed with local seasonal changes, which could be enabled by interaction between
endogenous (e.g., adaptable genetic factors) and environmental factors (e.g., photoperiod
and temperature) (Perry, 1971). The genetic adaptation to local climatic conditions is a
dominant determining factor for geographical distribution, growth capacity, and survival
Of tree species (Sarnish, 1954; Weiser, 1970). Thus, understanding of genetic regulation
on annual growth cycle is not only of scientific interest, but also of economic and
environmental significance.

The annual growth cycle has two major traits, dormancy and cold hardiness,
which are genetically and temporally correlated with each other (Chen et al., 2002).
Dormancy, in general, is defined as temporary suspension of visible growth of any plant
structure containing a meristem (Lang, 1987). Transition from active to dormancy state
occurs gradually (Perry, 1971), and termination of meristematic cell division is one of the
first visible indicators of the establishment of a dormant state. Physiologically, dormancy
can be distinguished into three states: paradormancy, ecodormancy and endodormancy
(Lang, 1987). The paradormancy describes dormancy when the initial reaction leading to
growth control involves a specific signal originating in or initially perceived in a different
structure from the one in which dormancy is manifested. The specific signal may be due
to a continuous production of inhibitory factors, as in apical dominance. The
ecodormancy describes dormancy when one or more factors in the basic growth
environment are unsuitable for overall growth metabolism. For example, water or
nutrient deficiencies or cold could cease plant growth, which is not limited to a specific

meristernic tissue. The endodormancy describes dormancy when the initial reaction

leading to growth control is a specific perception Of an environmental or endogenous
signal in the affected structure alone. The vascular cambium in woody perennials, as
shoot meristem does, exhibits different stages of dormancy (Little and Bonga, 1974), that
is, endodoramancy (i.e., rest) and ecodormancy (i.e, quiescence). At the end of growing
season, typically late summer, the camibium enter dormancy by responding to short
daylength (SD). At this stage, cambial dormancy is regulated by internal factors, which
has been denoted rest and is characterized by non-responsiveness to indole-3—acetic acid
(1AA) under favorable conditions for cambial activity (Little and Bonga, 1974; Little,
1981). Rest is characterized by a requirement for sustained exposure to low temperatures
before active growth can resume in the spring (i.e., chilling requirement). The cambium
gradually regains the ability to produce xylem in response to 1AA as exposure to natural
chilling is increased (Christersson, 1978). When fully responsive, the cambium is in the
quiescence stage of dormancy, which is imposed solely by adverse external factors. Upon
the return Of warmer temperatures in spring, plants are released from dormancy and also
fully dehardened, primarily by responding to warm temperature in whole shoots of
woody species (Fuchigami and Nee, 1987; Smitspinks, et al., 1985).

Cold acclimation is distinguished by typically two stages in woody plants. First
stage of cold acclimation together with the onset of dormancy is induced by short
daylength under warm temperatures (Weiser, 1970), at which plants can survive
temperatures slightly below freezing. The response to SD has an adaptive significance.
First, by responding to photoperiod cues, plants are able to begin acclimating to winter
conditions prior to the onset of low temperature. Second, photoperiodic cue is much more

reliable signal of the approaching winter compared to temperature, which can vary

dramatically from year to year. By responding to SD closely timed with local seasonal
time, plants are able to synchronize dormancy induction and cold acclimation with the
end of growing season and coming winter. However, higher capacity of cold acclimation,
i.e., frost hardiness could be achieved by subsequent cold after dormancy induction by
SD, suggesting the possible interrelation between dormancy and cold hardiness (Nissila
and Fuchigarni, 1978). The idea is supported further by the fact that dormant deciduous
genotypes always had higher extent of cold hardiness than non-dormant evergreen
genotypes in peach (Arora and Wisniewski, 1994). However, the molecular mechanism
underlying the interaction between SD and cold is poorly understood.

Recent studies showed that cold acclimation by low temperature (LT) . and SD
could be uncoupled in hybrid aspen, suggesting that SD and LT might utilize independent
pathways for the cold acclimation (Welling et al., 2002). Olsen et a1 (1997) reported that
cold acclimation and dormancy initiation by SDS were controlled by phytochorrne A-
involving pathway in Populus (Howe et al., 1996). Since the molecules involved in day
length perception Signaling (e.g., photoreceptors, CONSTANS, circadian proteins) have
been well studied and well known as high conservation across distant species, e.g.,
Arabidopsis and rice (Yanovsky and kay, 2003), it is plausible that trees must have
functionally homologous genes playing similar roles in their response to daylength
changes. Thus, it is interesting to question the mechanism by which the conserved
pathway could crosstalk with the cold acclimation pathways. Furthermore, most
dormancy/cold hardiness-related traits appear to be inherited in a quantitative manner,

displaying continuous distribution in progeny tests (Bradshaw and Stettler, 1995; Howe

et al., 2000). This is indicative of multigenic control. Thereby, these results also justify
my genomic approaches to the study Of annual growth cycle.

Despite these many observations, we still do not clearly understand the molecular
basis of seasonal development. Perhaps our limited understanding of the molecular
biology of seasonal development is, in part, due to superimposition Of various internal
and external events occurring during annual growth cycle. Dorrnancy development and
cold acclimation partially overlap and are interrelated, making it difficult to associate
physiological and molecular changes specifically with one or the other phenological
events. Thus, to separate dormancy processes from cold hardiness processes has been a
fundamental problem in dormancy research. It is important to identify which of the
changes is associated with Specific events of seasonal growth cycle. Although there were
some previous studies (Espinosa-Ruiz et al., 2004; Rowland and Arora, 1997; Welling et
al., 2004) that have identified the genes associated with different aspects of dormancy or

cold acclimation, they were limited to a few of genes or a short period of study time.

Auxin repressed gene
Since auxin, a plant growth regulator produced mainly in apical buds, plays a key role in
a wide variety of growth and developmental processes (Berleth, et al., 2000), the
characterization of auxin regulated genes will expand our knowledge on the various
biological implication of auxin.

In this thesis, an auxin repressed protein (ARP) from black locust has been
characterized. To date, three putative orthologs of ARP were isolated and characterized

from pea, strawberry, and tobacco. They were isolated from different tissue/organs (e.g.,

strawberry receptacles, Reddy and Poovaiah, 1990; pea dormant bud, Stafstrom et al.,
1998; tobacco pollen, Steiner et al., 2003), however, they all showed transcriptional
repression in response to auxin. The previous results suggest that this family of genes
may have conserved biological function across species although they were expressed in
different organs. Recently, we identified an ARP ortholog (named RpARP) from the
expressed sequence tags (ESTS) derived from the trunk wood of a 10—year old black
locust (R. pseudoacacia) (Yang et a1, 2003). The gene was expressed across the trunk
wood (bark/cambium, sapwood, and sapwood-heartwood transition zone). The deduced
amino acid sequence of RpARP shows high sequence homology with other ARP
orthologs. However, the underlying mechanism regulating repression by auxin, and
potential biological function that accounts for their expression in various organs or plant

species, is not known.

10

References

Aravind L, Watanabe H, Lipman DJ,Koonin EV (2000) Lineage-specific loss and
divergence of functionally linked genes in eukaryotes. Proc Natl Acad Sci U S A. 97:
11319-24

Arora R, Rowland LJ,Tanino K (2003) Induction and release of bud dormancy in woody
perennials: A science comes of age. Hortscience. 38: 911-921

Arora R,Wisniewski ME (1994) Cold-Acclimation in Genetically Related (Sibling)
Deciduous and Evergreen Peach (PrunuS-Persica L Batsch) .2. A 60-Kilodalton Bark
Protein in Cold-Acclimated Tissues of Peach IS Heat-Stable and Related to the
Dehydrin Family of Proteins. Plant Physiology. 105: 95-101

Berleth T, Mattsson J,Hardtke CS (2000) Vascular continuity and auxin signals. Trends
in Plant Science. 5: 387-393

Bhalerao R, Nilsson O,Sandberg G (2003) Out of the woods: forest biotechnology enters
the genomic era. Curr Opin Biotechnol. 14: 206-13

Bradshaw HD, Jr.,Stettler RF (1995) Molecular genetics of growth and development in
populus. IV. Mapping QTLS with large effects on growth, form, and phenology traits
in a forest tree. Genetics. 139: 963-73

Brunner AM, Busov VB,Strauss SH (2004) Poplar genome sequence: functional
genomics in an ecologically dominant plant species. Trends Plant Sci. 9: 49-56

Chaffey N, Cholewa E, Regan S,Sundberg B (2002) Secondary xylem development in
Arabidopsis: a model for wood formation. Physiol Plant. 114: 594—600

Chen THH, Howe GT,Bradshaw HD (2002) Molecular genetic analysis of dormancy-
related traits in poplars. Weed Science. 50: 232-240

Christersson L (1978) Influence of Photoperiod and Temperature on Development of
Frost Hardiness in Seedlings of Pinus-Silvestris and Picea-Abies. Physiologia
Plantarum. 44: 288-294

Dinus RJ, Payne P, Sewell NM, Chiang VL,Tuskan GA (2001) Genetic modification of
short rotation popular wood: Properties for ethanol fuel and fiber productions. Critical
Reviews in Plant Sciences. 20: 51-69

Eckardt NA (2000) Sequencing the rice genome. Plant Cell. 12: 2011-7

Espinosa-Ruiz A, Saxena S, Schmidt J, Mellerowicz E, Miskolczi P, Bako L,Bhalerao RP
(2004) Differential stage-specific regulation of cyclin-dependent kinases during
cambial dormancy in hybrid aspen. Plant J. 38: 603-615

Frugoli J,Harris J (2001) Medicago truncatula on the move! Plant Cell. 13: 458-63

11

Fuchigami LH,Nee CC (1987) Degree Growth Stage Model and Rest-Breaking
Mechanisms in Temperate Woody Perennials. Hortscience. 22: 836-845

Goodman HM, Ecker JR,Dean C (1995) The genome of Arabidopsis thaliana. Proc Natl
Acad Sci U S A. 92: 10831-5

Hertzberg M, Aspeborg H, Schrader J, Andersson A, Erlandsson R, Blomqvist K,
Bhalerao R, Uhlen M, Teeri 'IT, Lundeberg J, Sundberg B, Nilsson P,Sandberg G
(2001) A transcriptional roadmap to wood formation. Proc Natl Acad Sci U S A. 98:
14732-7

Howe GT, Gardner G, Hackett WP,Fumier GR (1996) Phytochrome control of short-day-
induced bud set in black cottonwood. Physiologia Plantarum. 97: 95-103

Howe GT, Saruul P, Davis J,Chen THH (2000) Quantitative genetics of bud phenology,
frost damage, and winter survival in an F-2 family of hybrid poplars. Theoretical and
Applied Genetics. 101: 632-642

Kohler A, Delaruelle C, Martin D, Encelot N,Martin F (2003) The poplar root
transcriptome: analysis of 7000 expressed sequence tags. FEBS Lett. 542: 37-41

Lang GA, Early JD, Martin GC,Darnell RL (1987) Endodormancy, Paradormancy, and
Ecodormancy - Physiological Terminology and Classification for Dormancy
Research. Hortscience. 22: 371-377

Lespinet 0, Wolf YI, Koonin EV,Aravind L (2002) The role of lineage-specific gene
family expansion in the evolution of eukaryotes. Genome Res. 12: 1048-59

Little CHA (1981) Effect of Cambial Dormancy State on the Transport of 1-C-l4 Indol-
3-Ylacetic Acid in Abies-Balsamea Shoots. Canadian Journal of Botany-Revue
Canadienne De Botanique. 59: 342-348

Little CHA,Bonga JM (1974) Rest in Cambium of Abies-Balsamea. Canadian Journal of
Botany—Revue Canadienne De Botanique. 52: 1723-1730

Matsuyama T, Tamaoki M, Nakajima N, Aono M, Kubo A, Moriya S, Ichihara T, Suzuki
O,Saji H (2002) CDNA rnicroarray assessment for ozone-stressed Arabidopsis
thaliana. Environ Pollut. 117: 191-4

Nissila PC,Fuchigami LH (1978) Relationship between Vegetative Maturity and lst
Stage of Cold-Acclimation. Journal of the American Society for Horticultural
Science. 103: 710-711

Oh S, Park S,Han KH (2003) Transcriptional regulation of secondary growth in
Arabidopsis thaliana. J Exp Bot. 54: 2709-22

Olsen JE, Junttila O, Nilsen J, Eriksson ME, Martinussen I, Olsson O, Sandberg G,Moritz
T (1997) Ectopic expression of oat phytochrome A in hybrid aspen changes critical
daylength for growth and prevents cold acclimatization. Plant Journal. 12: 1339-1350

12

Perry TO (1971) Dormancy of Trees in Winter. Science. 171: 29-&

Reddy ASN,Poovaiah BW (1990) Molecular-Cloning and Sequencing of a Cdna for an
Auxin-Repressed Messenger-Rna - Correlation between Fruit-Growth and Repression
of the Auxin-Regulated Gene. Plant Molecular Biology. 14: 127-136

Riechmann JL, Heard J, Martin G, Reuber L, Jiang C, Keddie J, Adam L, Pineda O,
Ratcliffe OJ, Samaha RR,al. e (2000) Arabidopsis transcription factors: genome-wide
comparative analysis among eukaryotes. Science. 290: 2105-10

Rowland LJ,Arora R (1997) Proteins related to endodormancy (rest) in woody perennials.
Plant Science. 126: 119-144

Samish RM (1954) Dormancy in woody plants. Annual Review of Plant Physiology and
Plant Molecular Biology. 5: 183-203

Seki M, Narusaka M, Abe H, Kasuga M, Yamaguchi-Shinozaki K, Caminci P,
Hayashizaki Y,Shinozaki K (2001) Monitoring the expression pattern of 1300
Arabidopsis genes under drought and cold stresses by using a full-length cDNA
rnicroarray. Plant Cell. 13: 61-72

Srnitspinks B, Swanson BT,Markhart AH (1985) The Effect of Photoperiod and
Thermoperiod on Cold-Acclimation and Growth of Pinus-Sylvestris. Canadian
Journal of Forest Research-Revue Canadienne De Recherche Forestiere. 15: 453-460

Somerville C,Somerville S (1999) Plant functional genomics. Science. 285: 380-3

Stafstrom JP, Ripley BD, Devitt ML,Drake B (1998) Dormancy-associated gene
expression in pea axillary buds. Planta. 205: 547-552

Steiner C, Bauer J, Amrhein N,Bucher M (2003) Two novel genes are differentially
expressed during early germination Of the male gametophyte of Nicotiana tabacum.
Biochim Biophys Acta. 1625: 123-33

Sterky F, Regan S, Karlsson J, Hertzberg M, Rohde A, Holmberg A, Amini B, Bhalerao
R, Larsson M, Villarroel R, Van Montagu M, Sandberg G, Olsson O, Teeri TI‘,
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.
Proc Natl Acad Sci U S A. 95: 13330-5

Taylor G (2002) Populus: arabidopsis for forestry. Do we need a model tree? Ann Bot
(Lond). 90: 681-9

Weiser CJ (1970) Cold resistance and injury in woody plants. Science. 169: 1269-78

Welling A, Rinne P, Vihera-Aarnio A, Kontunen-Soppela S, Heino P,Palva ET (2004)
Photoperiod and temperature differentially regulate the expression of two dehydrin
genes during overwintering of birch (Betula pubescens Ehrh.). Journal of
Experimental Botany. 55: 507-5 16

13

Wullschleger SD, Tuskan GA, DiFazio SP (2002) Genomics and the tree physiologist.
Tree Physiol. 22: 1273-6

Yang J, Park S, Kamdem DP, Keathley DE, Retzel E, Paule C, Kapur V,Han KH (2003)
Novel gene expression profiles define the metabolic and physiological processes
characteristic of wood and its extractive formation in a hardwood tree species,
Robinia pseudoacacia. Plant Mol Biol. 52: 935-56

Yanovsky MJ, Kay SA (2003) Living by the calendar: how plants know when to flower.
Nat Rev Mol Cell Biol. 4: 265-75

14

CHAPTER 1

Large-Scale Computational Analysis of Poplar ESTS Reveals the Repertoire and
Unique Features of Expressed Genes in the Poplar Genome

Abstract

Perennial woody plants differ from annual herbaceous plants in several ways and are
expected to have evolved to adopt a unique repertoire and expression profiles of
functional genes. Poplar, a model tree species for which a large number of ESTS are
publicly available, was used to carry out a large-scale comparative analysis with the
expressed sequences of eight plant species. First we have obtained 105,831 poplar ESTS
from public databases and identified a set of 25,282 unigenes (i.e., tentative non-
redundant sequences). The majority of the unigenes (56%) had significant matches to
Arabidopsis genes. We then estimated poplar multigene families by counting the
tBLASTX matches of each unigene against the poplar unigene dataset itself. Forty-seven
percent of the 25,282 unigenes were subsequently organized into 3,481 multigene
families 89% of which had less than five copy members. In poplar protein kinases
represent the largest family followed by GTP-binding proteins and Myb transcription
factors. Several multigene families had a higher copy number in poplar than in
Arabidopsis hinting potential lineage-specific proliferation of poplar protein families. The
protein families with such expansion might be related to the adaptation to high degree of
environmental stresses of perennial poplar growth. Comparison of poplar unigenes with
the Arabidopsis transcriptome revealed that genes involved in transcriptional regulation

are the most divergent while metabolism-related genes are most conserved.

15

Introduction

Tree growth is one of the most important biological processes on the Earth. Its product
wood is of primary importance to humans as timber for construction fuelwood and wood-
pulp for paper manufacturing. It is also the most environmentally cost-effective
renewable source of energy. In order to survive various environmental changes over long
periods of time trees have evolved to adopt unique physiological properties such as bud
dormancy to overcome unfavorable growth conditions, deciduous habit to avoid deep
frost, extended juvenile period, and secondary growth. We are interested in learning
about the genetic regulation of the tree-specific developmental processes using Populus
as a model.

The genus Populus belongs to the Salicaceae family and is comprised of about 30
species. Poplar has been widely accepted as a model system for tree biology by many
researchers for several reasons: a) a small genome (ca. 550 Mbp only 4x larger than
Arabidopsis but 40x smaller than pine); b) easy clonal propagation which allows for
replication of experiments; c) rapid growth; d) extensive genetic variation; e) high
expected synteny with Arabidopsis; 0 available high-throughput transgenic technology;
and g) high quality genetic maps. Furthermore its complete genome sequencing project
now has approximately 8x coverage of the genome (Tuskan et al., 2003). It is therefore
no surprise that a large number of expressed sequence tags (ESTS) have been produced
and analyzed from various tissues of the Populus species (Sterky et al., 1998; Hertzberg
et al., 2001; Kohler et al., 2003)

Single-pass sequencing of CDNA clones can reveal a substantial portion of the

expressed genes of a genome. As a result large numbers of ESTS are being generated

16

from a wide variety of plant species. Comparative analyses of the expressed sequence
tags (ESTS) provide a powerful means for gene discovery as well as the study of
molecular basis of plant diversity. First we retrieved a total of 105,831 poplar ESTS from
NCBI (http://www.ncbi.nih.gov) and organized them into clusters to generate unigene
sets using the StackPACK program. This program has been applied successfully to
identify different genes from redundant ESTS of other species such as barley (Michalek et
al., 2002) and fungus (Trail et al., 2003). With the identified unigene sets we conducted
large-scale computational comparisons against ca. 394,100 expressed sequences of eight
plant species that include four dicots, three monocots and one gymnosperrn. These
species have diverse phylogenetic relationships (Soltis et al., 1999) (Figure 1-1). Of these
eight species, two complete plant genome sequences (Arabidopsis and rice) were
included in the analyses. These analyses provided us with a unique opportunity to
examine the repertoire and organization of expressed genes in the poplar genome and to
observe the differences in expression profiles between poplar and other plant Species. The
unigene set generated in this study are accessible on a publicly searchable website

(httpzllwww.genomics.msu.edu/ ~han/poplar.html).

Materials and Methods

Assembly and clustering of poplar ES Ts

As of January 2003 a total of 105,831 poplar ESTS were available from four Populus
species and were retrieved using a custom Perl script from dbEST and the non-redundant
nucleotide database available at the NCBI (http://www.ncbi.nih.gov). The ESTS were

assembled separately by individual species using the StackPACKTM (version 2.2)

17

clustering system (http://www.egenetics.com) using default settings for EST assembly
(Miller et al., 1999). The process includes sequential steps of masking clustering
assembly alignment analysis and consensus partitioning. The masking step employed
CrossMatch (httpzllwww.genome.washington.edu/uwgc/analysistOOlS/phrap.htm) to mask
vector artifacts and repetitive sequences as simple repeat. The masked sequences were
clustered based on their relative similarin (the default is greater than 96 % identical over
a window of 150 bases) determined by d2 cluster (Hide et al., 1994) that is word-based
greedy clustering algorithm. The related but loose clusters were further aligned and
assembled using PHRAP (http://www.genome.washington.edu/uwgc/analysistoolsl-
phrap.htm) to improve alignment quality by generating particularly distinct sequences as
singletons and highly related sequences as sub-contigs. The aligned sub-contigs were
further analyzed using the CRAW alignment analysis tool (Chou and Burke 1999).
CRAW is used to analyze sub-contigs for error and alternative expression forms partition
the sub-contigs maximize consensus sequence length create final alignments and select
the best consensus sequence. For each species, the ESTS were organized into contiguous
overlapping consensus (‘contigs’) while the low-frequency ESTS were not incorporated
into contig assemblies and remain as singletons. The resulting singletons and consensus
sequences were used as tentative unigene sets in this study. Then the four unigene sets (a
total of 41,955 sequences) were assembled again using the same StackPACK clustering
system to produce a 29,818 poplar unigene set. In order to eliminate any transcripts that
potentially resulted from the same locus this poplar unigene set was re-assembled using
the BLASTCLUST program (ftp://ftp.ncbi.nlm.nih.gov/blast/documentationsl-

README.bcl) with a threshold of 95 % sequence identity covering a 90 % length region.

18

To avoid potential mosaic sequences of different species the sequence with the longest
sequence read in each contig was selected instead of the resulting consensus sequences
and used as a representative of the contig regardless of the species origin. The resulting
contigs and singletons were further screened to eliminate any sequences shorter than 250-
bp. The final 25,282 unigenes that may represent distinct genes were used for further

analysis.

Other plant sewnce data sets used in the analyses

Computational analyses were performed on several sets of plant gene sequences Obtained
from public databases. The Arabidopsis and rice gene sets as predicted from the entire
genomic sequence contain 27,290 and 80,916 sequences respectively (as of March 2003)
and are available through TIGR (http://www.tigr.org/tdb/). The minimally-redundant
EST sets of three dicots three monocots and one gymnosperrn species were obtained from
the Gene Indices of TIGR (http://www.tigr.org/tdb/tgi/plant.shtml) which were
constructed by first clustering and then assembling EST and annotated gene sequences
from GenBank to produce a set of unique high-fidelity virtual transcripts or tentative

consensus (TC) sequences (Quackenbush et al., 2000).

Similarity searches

Similarity searches were carried out with the Standalone BLAST programs (Altschul et
al., 1997) using executable copies obtained from the NCBI (v 2.2.2). The sequence data
sets used here were formatted first as a BLAST searchable database file using Forrnatdb

executable one of the BLAST tools available from NCBI (for a manual see

19

ftp://ftp.ncbi.nih.gov/blast/documents fREADME.formatdb). Searches were performed
through comparisons of protein sequences with translation of nucleotide query or
database sequences using gapped BLASTX or tBLASTX (Altschul et al., 1997).
Nucleotide queries were preprocessed with DUST to mask low-complexity regions and
protein query sequences (including six-frame translations of ESTS) were filtered with
SEG (Wootton and Federhen 1996). Custom Perl scripts (for scripts visit our laboratory
web site at http://foresg.msu.edulbiotech/Projects/Projects Poplar.htm) and relational
databases (Microsoft Access) were used to automate searches on large sets of query
sequences and to extract summary information (e.g., score and E-value of best hit).
Queries were considered to have a significant Similarity for E-value cutoff of $1.0E-20 a
potential similarity for E-value SIDE-10 and no significant similarity for E-value cutoff
of >1.0E-5. These user-defined E-values were based on the criteria used in other

comparative genomics studies (Rubin et al., 2000; Van der Hoeven etal., 2002).

Results and Discussion

Analysis of Poplar ES Ts and Establishment of a Unigene Set

We downloaded a total of 105,831 ESTS (as of January 2003) from public databases that
were derived from various tissues of four poplar species (Table 2-1). A hybrid aspen
species (Populus tremula x P. tremuloides) contributed the highest number of ESTS
(56,147) followed by P. balsamifera ssp. trichocarpa (24,050) P. tremula (14,091) and P.
tremuloides (11,543). The ESTS were organized into 25,282 of unigenes as described in
method, each of which may represent a potentially distinct gene (i.e., unigene). While
redundant transcripts are assembled into contigs, the low-frequency ESTS are not

incorporated into contig assemblies and remain as singletons. The resulting 25,282-

20

unigene set consists of 15,178 singletons and 10,104 contigs with an average length of
476-bp and 692-bp, respectively (Table 2-1). The 15,178 singletons consist of 9,005 from
aspen, 2,892 from P. trichocarpa, 1,966 from P. tremula and 1,317 from tremuloides.
The 25,282-unigenes account for about 72 % of an estimated 35,000 poplar genes
(Tuskan et al., 2003). A list of the entire poplar unigenes is provided on our project web
site at http://www.genorrrics.msu.edu/~han/mplar.html. However, we should
acknowledge that the number of unigenes might have been overestimated mostly due to
low quality and short-length attributes of ESTS that has been reported by other studies
(Zhu et al., 2003; Rudd, 2003). The attributes may be responsible for the singletons that
failed to merge into contigs. Indeed, the singletons of the poplar unigene set had shorter
sequence length than did the contigs (Figure 1-2). Fifty-five percent of singletons ranged
less than SOO-bp while twenty percent of contigs ranged less than 500-bp. The potential
drawbacks (i.e. short length and singleton) were considered during interpretation of
further analyses. The singletons could have been generated by orthologs less conserved
among the four species. To assess the likelihood of orthologs appearing as unigenes, we
compared the singletons of three poplar species with the entire unigenes from aspen using
tBLASTX algorithm. We defined as the potential orthologs the singletons that had higher
similarity with aspen unigenes than with genes of any other non-poplar species (i.e. the
eight species used in this study; Figure l-l) although they could be closely related
paralogs. Using the bits score as similarity criteria and less than 1E-20 as a threshold for
significance of the similarity, we identified 571 singletons (275 from P. trichocarpa, 195

from P. tremula and 97 from P. tremuloides) as the potential orthologs. This number is

21

relatively small compared to the 25,282 unigenes, suggesting that the less conserved
orthologs are unlikely to affect the overall conclusions at the analysis stringency we used.

A surrogate annotation approach was used to annotate the poplar unigene set. The
entire poplar unigene set was searched using the tBLASTX algorithm against the
Arabidopsis transcriptome predicted from the complete genomic sequence (TIGR
Genome Databases http://www.tigr.org/tdb/). Functional categories have been assigned
using the Munich Information Center for Protein Sequences (MIPS) Arabidopsis database
(MATDB httpzllmips.gsf.de/proj/thal/db/search/search_frame.html) search function.
Poplar unigenes with an E-value of SIDE-20 were assigned to the corresponding
Arabidopsis annotation. This approach is based on the assumption that functionality is
transferable based on sequence conservation. The majority of the unigenes (56 %) had
significant matches to Arabidopsis genes 75 % of which matched to genes of known
function but the remaining 25 % were found to be unclassified or of unknown function.
Of the assigned functions metabolism-related genes were the most numerous at 18 %
followed by cellular organization (9 %) and Signal transduction (8 %) (Figure 1-3). Other
significant functional categories include transcription regulation (7 %), protein
destination (7 %), protein synthesis (4 %), transport facilitation (5 %) and cell division (3
%). Kirst et al., (2003) recently reported that ~90 % of pine ESTS had counterparts in
Arabidopsis sequences (E value < 1.0E-10). The 56 % match in this study is substantially
lower than their estimate. The difference might be explained by the facts that many of the
poplar unigenes used in our analysis were not full-length (average length of 572-bp) and
different cut-off values (1.0E-10 in pine study and 1.0E-20 in the current study) were

used. Indeed, the unmatched unigenes had shorter sequence length with an average of

22

490-bp than did the matched ones with an average of 645-bp. In addition, this functional
classification analysis was performed solely based on ESTS, whose representation is
highly dependent on the source libraries and expression level. Genes with low level of
expression might not have been represented in this analysis. This notion is supported by
the fact that a large number of Arabidopsis ESTS (178,000 ESTS) from various sources
representing over 60 distinct tissues, biotic and abiotic challenges, and developmental
stages matched to only to 16,115 unigenes (reviewed in Rudd 2003). Likewise, in the
current classification analysis the presence of an EST is a reliable attribute while the
absence of an EST does not necessarily mean that the gene is not present in the genome
and not expressed. Completion of poplar genome sequencing will provide a better

opportunity to examine the differences between poplar and other plant species.

Poplar Multigene Families Estimated by tBLASTX Analysis of Unigenes

One of the characteristic features of eukaryotic genomes is that a significant fraction of
protein-coding genes belong to multigene families that are likely derived from gene
duplication. For example, up to 80 % of the Arabidopsis genes are members of multigene
families (Lespinet et al., 2002). The main contributing forces for organizational and
regulatory diversity in eukaryotes include the changes in domain architectures via domain
accretion and shuffling gene loss in a particular lineage and lineage-specific proliferation
of protein families (Aravind et al., 2000; Lander et al., 2001; Lespinet et al., 2002).
Recent advances in structural and comparative genomics have provided valuable

information about gene family organization in diverse plant species.

23

In order to describe the gene family organization in poplar, we first
computationally determined the number and size of poplar gene families by counting the
tBLASTX matches of each unigene against the poplar unigene set itself with an E-value
threshold of 51.0E-20. It is important to note that these gene families do not necessarily
indicate functional groupings. Of the 25,282 poplar unigenes, 11,909 (47 %) had
significant matches with other poplar unigenes at this threshold and were subsequently
organized into 3,481 gene families. To determine the potential bias caused by short
sequence length or singleton, we performed the same analysis using unigenes greater than
500-bp or excluding singletons (i.e., contigs). Of 15,088 poplar unigenes greater than
500-bp, 9,448 (63 %) were grouped into 2,033 gene families. There were 10,104 poplar
contigs (25,282 minus singletons), 5,401 (53 %) of which had significant matches with
other poplar unigenes, resulting in 1,995 gene families. The distribution of the gene
family sizes is shown in Figure 1-4. Overall, the majority (56-55 %) had two copy
members. Gene families with fewer than five copy members accounted for around 90 %
of the entire gene families estimated in this study. In order to assess how the copy
number of the gene families was conserved between the two species, we have compared
the copy numbers of the 14 largest poplar gene families with those of corresponding
Arabidopsis gene families. The Arabidopsis gene copy number was obtained by
subjecting the Arabidopsis gene set (TIGR http://www.tigr.org[tdb[) to the same analysis
as the poplar tBLASTX. In order to increase the stringency and therefore not to
overestimate the poplar gene copy number in the analysis the initially tBLASTX-
deterrnined copy numbers of the 14 largest poplar gene families were subjected to several

refinement steps. as described in Table 2-2. First, we removed the sequences that had no

24

matches to the known multigene family proteins in public databases. Then, the sequences
that had less than 90 % identity with any member of the multigene family with over 90 %
length region were removed. Finally, we removed the sequences that either did not share
the conserved region of the protein family due to its short length, or had long gap regions
probably resulting from chimera when they were aligned using ClustalW software
(Thompson et al., 1994). The resulting alignments are provided as supplemental materials
(httpzllforestrv.m_su.edu/biotech/Proj&cts/Proiects¥ Popl_ar.htm). Table 2-2 lists the 14
highest copy number poplar gene families along with their corresponding Arabidopsis
gene families. In poplar protein kinases represent the largest family followed by Ras-
related GTP-binding protein and Myb transcription factor. Several multigene families had
lower copy numbers in poplar than in Arabidopsis (Table 2-2). However caution should
be exercised in interpreting these results because two factors exist that may lead to a
substantial underestimate of cow numbers for each gene family. First, many Of the
25,282 poplar unigene sequences used in the analysis are partial sequences that generally
result in higher E-values (i.e. lower similarity) compared to full-length sequences. It is
quite possible that the tBLASTX search with a short coding region of partial EST
sequences might miss the consensus motif(s) characteristics of the gene family if they are
located in the middle of the protein. Second, the 25,282 poplar unigenes do not represent
the entire poplar transcriptome while Arabidopsis sequences represent whole
transcriptome on its genome. Possibly the genes with low expression levels were not
picked up during the EST sequencing resulting in a lower copy number estimation than
actual number. This may account for the lower copy number of some multigene families

in poplar when compared to Arabidopsis.

25

Five of the 14 largest poplar multigene families had a relatively higher copy
number in poplar than in Arabidopsis (Table 2-2). These families include ubiquitin-
domain protein, ubiquitin-conjugating enzyme, tubulin, peptidylprolyl isomerase and
ABA-inducible protein. The proteins of these families had predicted biochemical
characteristics that suggest their roles in protein degradation, energy, Stress response, and
cellular structure. The higher copy number of ubiquitin gene families in poplar may
represent an evolutionary adaptation important for high protein turnover throughout
different seasons. Interestingly functionally related two multigene families such as
ubiquitin-conjugating enzyme and peptidylprolyl isomerase which are involved in protein
degradation and folding respectively (Hershko and Ciechanover, 1998; Schiene and
Fischer 2000) have a slightly higher copy number in poplar than in Arabidopsis (Table 2-
2) and indeed these genes were more abundantly expressed in senescent autumn leaves
than in the young summer leaves of poplar (Bhalerao et al., 2003). Another example of
potential adaptational change in poplar is the tubulin gene family that has 19 members in
poplar and 17 in the Arabidopsis genome. Perennial nature of poplar growth and
development requires adaptive evolution for cold hardiness. The microtubules of higher
plants have been suggested to participate in low temperature stress response and
adaptation (Nick, 1998; Nyporko et al., 2003). In fact those factors known to increase
plant cold hardiness were also involved in the induction of elevated cold-stability in
microtubules (Pihakaskimaunsbach and Puhakainen 1995). Also it has been shown that
depolymerization of microtubules is related to freezing injury in cotton (Gossypium
hirsutum L.) (Rikin et al., 1980). The cold-resistance of microtubules appears to be

correlated with their cold stability (Detrich 1997; Gupta et al., 2001). Parker and Detrich

26

(1998) reported that an Antarctic fish (Notothenia coriiceps) had a substantial number
(~15) of a-tubulin subunits encoded by a multigene family suggesting that the expansion
of tubulin gene families might have occurred as a result of adaptive evolution for low
temperatures. Considering these observations in the literature it is not surprising that
poplar had an increase in copy number for this gene family. Another multigene family
with a higher copy number in poplar encodes ABA-inducible protein which is a
subfamily of late embryogenesis abundant (LEA) protein family (Wise, 2003). These
hydrophilic proteins have been suggested to be involved in the protection of cellular
structures from dehydration or hyper-osmotic stress by acting as a hydration buffer by
sequestration of ions by direct protection of other proteins or membranes or by re-
naturation of unfolded proteins (Bray et al., 1993; Soulages et al., 2002; Koag et al.,
2003).

Multigene families with higher copy numbers could imply greater functional
diversity providing resources for specific adaptations and evolution of new functional
systems. Considering that the poplar genome is approximately four times larger than the
Arabidopsis genome and only about 72 % of the transcriptome is represented in this
analysis it is reasonable to assume that the copy number of each poplar family should be
less than or at most equal to that of corresponding Arabidopsis families if there is no
poplar specific gene family expansion since divergence from the last common ancestor.
Nonetheless we noted several multigene families had relatively higher copy numbers in
poplar than in Arabidopsis. Potential physiological roles of the protein families (as
described above) suggest that this kind of multigene family expansion might be common

among the perennial temperate tree species. However it should be noted that the copy

27

numbers of multigene families could also be overestimated considering the high rate of
error in EST datasets (Ewing and Green 1998; Hillier et al., 1996), and highly expressed

gene families could be overrepresented during clustering analysis.

Comparison of Poplar with Other Plant Species

Arabidopsis is an herbaceous non-woody plant that does not undergo secondary growth
under normal conditions. However it has recently been shown that Arabidopsis can be
induced to express all of the major components of secondary growth (Lev-Yadun, 1994;
Zhao et al., 2000; Chaffey et al., 2002). It is therefore of great scientific interest to
examine the degree of sequence similarity in expressed gene sequences between
Arabidopsis an herbaceous plant and poplar a perennial woody species. In order to
determine whether certain functional classes of genes represent characteristic differences
between the perennial tree species (i.e. polar) and an annual herbaceous plant (i.e.
Arabidopsis) we carried out computational comparisons of the 25,282 poplar unigenes in
all translated frames (tBLASTX algorithm) with Arabidopsis transcriptome predicted
from the complete genomic sequence using the E-values of tBLASTX searches as an
estimate of sequence conservation. Figure 1-5 displays the distribution of E-value
matches with regard to functional categories for the poplar unigene set. Of the 25,282
poplar unigenes, 58 % show strong similarity (£1.0E-20) with their Arabidopsis
counterparts. Even at the strongest match threshold (_<_l.0E-100), eight percent (1,907) of
the 25,282 poplar unigenes matched with Arabidopsis sequences showing a very high
level of conservation. While the frequencies of genes in most functional categories were

similar among the five different match stringency (E-value) categories those of genes

28

involved in a few cellular functions changed significantly as the match stringency
increased. For example, the proportion of genes belonging to metabolism category
decreased from 29 % at the strongest match stringency (SIDE-100) to 12 % when the E-
value cutoff was reduced to 1.0E-20 ~ 1.1E-30 suggesting that metabolic functions are
more constant across plant species (i.e. more ancestral gene functions) (Figure 1-5). On
the other hand, transcription regulation-related genes showed opposite trends. Their
frequencies decreased from 9 to 3 % as the E-value stringency changed from 1.0E-20 to
1.0E-100. This along with the fact that Arabidopsis can be induced to undergo secondary
growth suggests that the major differences between poplar and Arabidopsis (e.g.
secondary growth perennial growth habit) may be in transcriptional control rather than in
structural genes. Genes of unknown function are among the ‘less-conserved.’ For
example 33 % of the poplar unigenes having relatively weak similarity (£1.0E-20) with
their Arabidopsis counterparts were categorized as ‘unclassified’ function while only 9 %
of those with the strongest similarity (SIDE-100) matched to Arabidopsis genes for
which no putative functions have been assigned. Genes involved in most other cellular
functions had similar frequencies among the five different match stringency categories
suggesting that they are relatively fast evolving. The differential distribution of the
functional categories based on the match stringencies was confirmed by contingency test
(P <1E-5).

In order to investigate whether the highly conserved or divergent genes between
poplar and Arabidopsis are also conserved or divergent in other vascular plants we
computationally compared the poplar unigenes with ca. 394,100 expressed sequences

including putative ORFS from the genome sequences of four dicots (Arabidopsis barrel

29

medic, soybean, tomato) three monocots (maize, rice, wheat) and one gymnosperrn (pine)
(Figure 1-1). For this analysis the 394,100 plant gene sequences were divided into three
groups (‘dicots’, ‘monocots’ and ‘pine’). All sequences from Glycine max, Medicago
truncatula, Arabidopsis thaliana, and Lycopersicon esculentum were pooled into the
‘dicots’ group which had a total of 151,832 sequences. Likewise the sequences from
Oryza sativa, Zea mays and Triticum aestivum went into the ‘monocots’ group making a
total of 220,859 sequences. The ‘pine’ group represents only one species (Pinus taeda)
with 21,409 ESTS available. A large number of poplar unigenes did not have any
detectable homologues to other plant species, ranging from 6,835 (27 %) (an E-value
threshold of £1.0E-10) to 9,398 (37 %) (SIDE-20). However, about 30 % (7,581) of the
25,282 poplar unigenes had significant matches (SIDE-20) in all of the three groups
suggesting their involvement with common biological functions in plant species. The
proportions in each functional category of the 7,581 unigenes that had counterparts in all
of the three groups were very similar to those of the entire poplar unigene set. The largest
functional category was ‘metabolism’ (20 %) followed by ‘unclassified proteins’ (18 %)
and ‘signal transduction’ (9 %) (Figure 1-6). A total of 7,760 poplar unigenes matched to
pine ESTS with E-value cutoff of S 1.0E—20 while 13,422 and 15,257 poplar unigenes had
significant matches to monocot and dicot sequences respectively ($1.0E-20). The number
Of poplar unigenes that matched to monocot sequences represents a substantially high

percentage (53 %) of the 25,282 poplar unigenes.

Conclusion
About 72 % of the 35,000 estimated poplar genes were identified in the unigene set

described in this report. High proportions of the unigenes had counterparts in the

30

genomes of Arabidopsis and other plant species supporting the hypothesis that many of
the functional genes are conserved among different plant species. We estimate that at
least 47 % of poplar genes belong to multigene families based on the observation that
11,909 of the 25,282 unigenes had significant tBLASTX matches (SIDE-20) to other
poplar unigenes. Comparison of poplar multigene families with those of Arabidopsis
suggests that the expansion of poplar multigene families occurred in support of the
adaptation to various stresses of poplar growth. Another striking feature of the difference
between poplar and Arabidopsis transcriptome is that the genes involved in
transcriptional regulation are the most divergent while metabolism-related genes are
highly conserved. While the findings described in this report increase our understanding
of the poplar transcriptome it should be noted that completion of the on-going poplar
genome sequencing would provide an invaluable opportunity to verify these findings and

advance our knowledge of tree growth.

Acknowledgement

We would like to thank the Genomics Technology Support Facility staff at Michigan
State University for their help and consultation with the bioinforrnatics analysis. This
work was supported by USDA Grants to the Eastern Hardwood Utilization Program at

Michigan State University (No. 98-34158-5995 00-34158-9236 and 01-34158-11222).

31

References

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997)
Gapped BLAST and PSI-BLAST: a new generation of protein database search
programs. Nucleic Acids Res. 25: 3389-402

Aravind L, Watanabe H, Lipman DJ, Koonin EV (2000) Lineage-specific loss and
divergence Of functionally linked genes in eukaryotes. Proc Natl Acad Sci U S A. 97:
11319-24

Bergelson J, Dwyer G, Emerson JJ (2001) Models and data on plant-enemy coevolution.
Annu Rev Genet. 35: 469-99

Bessler JB, Torredagger JZ, Zakian VA (2001) The Pif 1p subfamily of helicases: region-
specific DNA helicases? Trends Cell Biol. 11: 60-5

Bhalerao R, Keskitalo J, Sterky F, Erlandsson R, Bjorkbacka H, Birve SJ, Karlsson J,
Gardestrom P, Gustafsson P, Lundeberg J,Jansson S (2003) Gene expression in
autumn leaves. Plant Physiol. 131: 430-42

Bray EA (1993) Molecular Responses to Water Deficit. Plant Physiol. 103: 1035-1040

Chaffey N, Cholewa E, Regan S, Sundberg B (2002) Secondary xylem development in
Arabidopsis: a model for wood formation. Physiol Plant. 114: 594-600

Chou A,Burke J (1999) CRAinew: for viewing splicing variation, gene families, and
polymorphism in clusters of ESTS and full-length sequences. Bioinformatics. 15: 376-
81

Coleman GD, Banados MP, Chen T (1994) Poplar Bark Storage Protein and a Related
Wound-Induced Gene Are Differentially Induced by Nitrogen. Plant Physiol. 106:
211-215

Dean C, Pichersky E, Dunsmuir P (1989) Structure, Evolution, and Regulation of Rbcs
Genes in Higher- Plants. Annu. Rev. Plant Physiol. Plant Molec. Biol. 40: 415-439

Detrich HW, 3rd (1997) Microtubule assembly in cold-adapted organisms: functional
properties and structural adaptations of tubulins from antarctic fishes. Comp Biochem
Physiol A Physiol. 118: 501-13

Ewing B, Green P (1998) Base-calling of automated sequencer traces using phred. 11.
Error probabilities. Genome Res. 8: 186-94

Galili S, Avivi Y, Millet E, Feldman M (2000) RFLP-based analysis of three Rch
subfamilies in diploid and polyploid species of wheat. Mol Gen Genet. 263: 674-80

Gupta ML, Jr., Bode CJ, Dougherty CA, Marquez RT, Himes RH (2001) Mutagenesis of
beta-tubulin cysteine residues in Saccharomyces cerevisiae: mutation of cysteine 354
results in cold-stable microtubules. Cell Motil Cytoskeleton. 49: 67-77

32

Haruta M, Major IT, Christopher ME, Patton JJ, Constabel CP (2001) A Kunitz trypsin
inhibitor gene family from trembling aspen (Populus tremuloides Michx.): cloning,

functional expression, and induction by wounding and herbivory. Plant Mol Biol. 46:
347-59

Hershko A,Ciechanover A (1998) The ubiquitin system. Annu Rev Biochem. 67: 425-79

Hertzberg M, Aspeborg H, Schrader J, Andersson A, Erlandsson R, Blomqvist K,
Bhalerao R, Uhlen M, Teeri TT, Lundeberg J, Sundberg B, Nilsson P, Sandberg G
(2001) A transcriptional roadmap to wood formation. Proc Natl Acad Sci U S A. 98:
14732-7

Hide W, Burke J ,Davison DB (1994) Biological evaluation of d2, an algorithm for high-
perforrnance sequence comparison. J Comput Biol. 1: 199-215

Hillier LD, Lennon G, Becker M, Bonaldo MF, Chiapelli B, Chissoe S, Dietrich N,
DuBuque T, Favello A, Gish W, Hawkins M, Hultrnan M, Kucaba T, Lacy M, Le M,
Le N, Mardis E, Moore B, Morris M, Parsons J, Prange C, Rifldn L, Rohlfing T,
Schellenberg K, Marra M,et al. (1996) Generation and analysis of 280,000 human
expressed sequence tags. Genome Res. 6: 807-28

Keeling PJ,Palmer JD (2001) Lateral transfer at the gene and subgenic levels in the
evolution of eukaryotic enolase. Proc Natl Acad Sci U S A. 98: 10745-50

Koag MC, Fenton RD, Wilkens S, Close TJ (2003) The binding of Maize DHN 1 to Lipid
Vesicles. Gain of Structure and Lipid Specificity. Plant Physiol. 131: 309-16

Kohler A, Delaruelle C, Martin D, Encelot N,Martin F (2003) The poplar root
transcriptome: analysis of 7000 expressed sequence tags. FEBS Lett. 542: 37-41

Koiwa H, Bressan RA, Hasegawa PM (1997) Regulation of protease inhibitors and plant
defense. Trends Plant Sci. 2: 379-384

Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K,
Doyle M, FitzHugh W,al. e (2001) Initial sequencing and analysis of the human
genome. Nature. 409: 860-921

Lespinet 0, Wolf YI, Koonin EV, Aravind L (2002) The role of lineage-specific gene
family expansion in the evolution of eukaryotes. Genome Res. 12: 1048-59

Levyadun S (1994) Induction of Sclereid Differentiation in the Pith of Arabidopsis-
Thaliana (L) Heynh. J. Exp. Bot. 45: 1845-1849

Mandal S, Kundu P, Roy B, Mandal RK (2002) Precursor of the inactive 28 seed storage
protein from the Indian mustard Brassica juncea is a novel trypsin inhibitor.
Charaterization, post-translational processing studies, and transgenic expression to
develop insect-resistant plants. J Biol Chem. 277: 37161-8

Michalek W, Weschke W, Pleissner KP,Graner A (2002) EST analysis in barley defines a
unigene set comprising 4,000 genes. Theor Appl Genet. 104: 97-103

33

Miller RT, Christoffels AG, Gopalakrishnan C, Burke J, Ptitsyn AA, Broveak TR,Hide
WA (1999) A comprehensive approach to clustering of expressed human gene

sequence: the sequence tag alignment and consensus knowledge base. Genome Res.
9: 1143-55

Nick P: Signaling to the microtubular cytoskeleton in plants International Review of
Cytology - a Survey of Cell Biology, Vol 184, pp. 33-80 (1998).

Nyporko AY, Demchuk ON,Blume YB (2003) Cold adaptation of plant microtubules:
structural interpretation of primary sequence changes in a highly conserved region of
alpha-tubulin. Cell Biol Int. 27: 241-3

Oguri S, Kamoshida M, Nagata Y, Momonoki YS,Kamimura H (2003) Characterization
and sequence of tomato 28 seed albumin: a storage protein with sequence similarities
to the fruit lectin. Planta. 216: 976-84

Parker SK,Detrich HW, 3rd (1998) Evolution, organization, and expression of alpha-
tubulin genes in the antarctic fish Notothenia coriiceps. Adaptive expansion of a gene
family by recent gene duplication, inversion, and divergence. J Biol Chem. 273:
34358-69

Pihakaskimaunsbach K,Puhakainen T (1995) Effect of Cold-Exposure on Cortical
Microtubules of Rye (Secale-Cereale) as Observed by Immunocytochemistry.
Physiol. Plant. 93: 563-571

Quackenbush J, Liang F, Holt 1, Pertea G,Upton J (2000) The TIGR gene indices:
reconstruction and representation of expressed gene sequences. Nucleic Acids Res.
28: 141-5

Riechmann JL, Heard J, Martin G, Reuber L, Jiang C, Keddie J, Adam L, Pineda O,
Ratcliffe OJ, Samaha RR,al. e (2000) Arabidopsis transcription factors: genome-wide
comparative analysis among eukaryotes. Science. 290: 2105-10

Rikin A, Atsmon D,Gitler C (1980) Chilling Injury in Cotton (Gossypium-Hirsutum-L) -
Effects of Anti-Microtubular Drugs. Plant Cell Physiol. 21: 829-837

Rubin GM, Yandell MD, Wortman JR, Gabor Miklos GL, Nelson CR, Hariharan 1K,
Fortini ME, Li PW, preiler R, Fleischmann W, Cherry JM, Henikoff S, Skupski
MP, Misra S, Ashbumer M, Bimey E, Boguski MS, Brody T, Brokstein P, Celniker
SE, Chervitz SA, Coates D, Cravchik A, Gabrielian A, Galle RF, Gelbart WM,
George RA, Goldstein LSnB, Gong F, Guan P, Harris NL, Hay BA, Hoskins RA, Li
J, Li Z, Hynes RO, Jones SJnM, Kuehl PM, Lemaitre B, Littleton JT, Morrison DK,
Mungall C, O'Farrell PH, Pickeral OK, Shue C, Vosshall LB, Zhang J, Zhao Q,
Zheng XH, Zhong F, Zhong W, Gibbs R, Venter JC, Adams MD,Lewis S (2000)
Comparative Genomics of the Eukaryotes. Science. 287: 2204-2215

Rudd S (2003) Expressed sequence tags: alternative or complement to whole genome
sequences? Trends Plant Sci. 8: 321-329

34

Schiene C, Fischer G (2000) Enzymes that catalyse the restructuring of proteins. Curr
Opin Struct Biol. 10: 40-5

Soltis PS, Soltis DE,Chase MW (1999) Angiosperrn phylogeny inferred from multiple
genes as a tool for comparative biology. Nature. 402: 402-4

Soulages JL, Kim K, Walters C,Cushman JC (2002) Temperature-induced extended
helix/random coil transitions in a group 1 late embryogenesis-abundant protein from
soybean. Plant Physiol. 128: 822-32

Sterky F, Regan S, Karlsson J, Hertzberg M, Rohde A, Holmberg A, Amini B, Bhalerao
R, Larsson M, Villarroel R, Van Montagu M, Sandberg G, Olsson O, Teeri 'IT,
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.
Proc Natl Acad Sci U S A. 95: 13330-5

Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving the sensitivity
of progressive multiple sequence alignment through sequence weighting, position-
specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673-80

Trail F, Xu JR, San Miguel P, Halgren RG, Kistler HC (2003) Analysis of expressed
sequence tags from Gibberella zeae (anamorph Fusarium graminearum). Fungal
Genet Biol. 38: 187-97

Van der Hoeven R, Ronning C, Giovannoni J, Martin G, Tanksley S (2002) Deductions
about the Number, Organization, and Evolution of Genes in the Tomato Genome
Based on Analysis of a Large Expressed Sequence Tag Collection and Selective
Genomic Sequencing. Plant Cell. 14: 1441-1456

Wang WX, Pelah D, Alergand T, Shoseyov O,Altman A (2002) Characterization of SP1,
a stress-responsive, boiling-soluble, homo-oligomeric protein from aspen. Plant
Physiol. 130: 865-75

Wetzel S, Demmers C,Greenwood JS (1989) Seasonally Fluctuating Bark Proteins Are a
Potential Form of Nitrogen Storage in 3 Temperate Hardwoods. Planta. 178: 275-281

Wootton JC,Federhen S (1996) Analysis of compositionally biased regions in sequence
databases. Methods Enzymol. 266: 554-71

Yoon M, Putterill JJ, Ross GS, Laing WA (2001) Determination of the relative
expression levels of rubisco small subunit genes in Arabidopsis by rapid
amplification of cDN A ends. Anal Biochem. 291: 237-44

Zardoya R, Ding X, Kitagawa Y, Chrispeels MJ (2002) Origin of plant glycerol
transporters by horizontal gene transfer and functional recruitment. Proc Natl Acad
Sci U S A. 99: 14893-6

Zhao C, Johnson BJ, Kositsup B, Beers EP (2000) Exploiting secondary growth in
Arabidopsis. Construction of xylem and bark cDNA libraries and cloning of three
xylem endopeptidases. Plant Physiol. 123: 1185-96

35

Zhou J, Monson EK, Teng S, Schulz VP, Zakian VA (2000) Piflp helicase, a catalytic
inhibitor of telomerase in yeast. Science. 289: 771-4

Zhu B, Coleman GD (2001) The poplar bark storage protein gene (Bspa) promoter is
responsive to photoperiod and nitrogen in transgenic poplar and active in floral
tissues, immature seeds and germinating seeds of transgenic tobacco. Plant Mol Biol.
46: 383-94

Zhu W, Schlueter SD,Brendel V (2003) Refined annotation of the Arabidopsis genome
by complete expressed sequence tag mapping. Plant Physiology. 132: 469-484

Web Site References
http://www.tigr.org/tbd/plant.shtml; TIGR Gene Index Databases.

ftpz/lfypncbi.nih.gov/blast/executables; Stand-alone BLAST program.

ftp://ftp.ncbi.nih.gov/blast/documents/README.form_atdb; Documentation for the
Formatdb program.

ftp://ftp.ncbi.nih.gov/blast/documents/README.formatdb; Documentation for the
BLASTCLUST program.

ftp://ftp.ncbi.nih.gov/blast/db; NCBI Public Databases.
http://www.tigr.org/tdb/; The TIGR Genome Databases
http://mips.gsf.de/proj/thal/db/search/search_frame.htrnl; MATDB Search Page
http://www.egenetics.com; Electric Genetics Home Page.

httpzllforestry.msu.edu/biotech/Projects/Proiects Poplarhtm The Han Lab Project Web
Site

mzllwwwgenomics.msu.edu/~haLn/poplar.html Searchable web site for the data
presented.

36

Fabales Glycine max (55,990)
Medicago truncatula (36,235)

 

Malpighiales Populus spp. (105,831)

Malvales

 

Brassicales Arabidopsis thaliana (27,290)

 

[: Solanales L ycopersicon esculentum (32,317)

Asterales

Oryza sativa (132,485)
Monocot Zea mays (36,022)

 

 

Triticum aestivum (52,352)

 

Gymnosperm Pinus taeda (21,409)

 

Figure 1-1. Phylogenetic relationships for the nine plant species used in the sequence
analyses.

The number of sequences used in the analyses for each species is given in parenthesis.
The phylogenetic tree was adapted from Soltis et al. (1999) and abbreviated only to
reflect the relative branching order of each species involved.

37

4000

3500 ............ ................ El Singletons
I Contigs

 

 

b--- ------

_-——---

r—————--‘

hl--—--‘

 

Frequency
-I~ -l N N or
O 0| O 01 O
O O O O O
c O O O O

 

 

500

o ——!L- - .. -

300 400 500 600' 700 800 900 1000 1100
Length (bp)

 

 

 

 

 

 

 

Figure 1-2. Distribution of sequence length (bp) of singletons and contigs that
constitute the poplar unigene set.

38

Transport (5) Transcription (7)

Miscellaneous (6) ' Protein Synthesis (4)
Organization (9)

Transduction (8)

   
 
   
  

    

Protein Destination (7)

Unclassified
Proteins (25)

Metabolism (18)

Cell Division
DNA Synthesisq(3)

Defense, Cell Death (4) , .
Cellular BlogeneSIs (1 )Signal

Energy (2)
Development (1 )

Figure 1-3. Functional categorization of the 14,291 poplar unigenes that had
tBLASTX matches to Arabidopsis transcriptome with E-value threshold of $1.0E-20.
The 10,991 unigenes that did not have matches were excluded from the chart. Functional
categories are according to the Munich Information Center For Protein Sequence (MIPS
http://mips.gsf.de). Percentage of the unigenes in each functional category is given in
parenthesis.

39

 

60

 

50

 

I Based on whole unigenes
I Based on > SOO-bp unigenes
U Based on Contigs

 

4O

 

30

 

 

20

 

1O

 

 

 

 

Fraction of Poplar Gene Families (%)

o
5 6

2 3 4 .7 8 9 10 11-20 >20
Gene Copy Number Estimated by tBLASTX with an E-value of <1 .0E-20

Figure 1-4. Gene copy number distribution of putative poplar gene families
estimated by tBLASTX search against the poplar unigene data set itself.

A total of 3,481, 2,303 or 1,995 multigene families were identified based on different
stringency, and about 56 % of which had two copy members.

40

2365(17%) 4130(29%) 3115(22%)

1007.1:E

 

 

 

 

sol W
80‘ ~

 

70

so”

50’ ~
4014
30?

.l a
10s

l
t_

1.1E-30 4

 

 

EL

.1 E-50

. ‘ ‘

 

 

 

 

1.0E-201.0E-30 1.0E-50 1.0E-70

Figure 1-5. Percentage of poplar unigenes belonging to each functional category

with different E-value thresholds.

A total of 14,291 sequences had significant tBLASTX matches to the genes of known
function according to MIPS (http://mips.gsf.de) functional categorization. The actual
number and percentage of sequences matched with ranges of l.0E—20~1.1E-30 1.1E-
30~1.0E-50 1.1E-30~1.0E-50 1.1E-50~l.0E-70 l.1E—70~1.0E-100 and 5 1.0E-100
respectively were indicated on top of the graph. Asterisks indicated that the differed

 

 

4 .1 E70 4 .1 E-100 51.0E-1

2774(19°/.) 1907(13%)

Transport Facilitation
D Transcription

0 Protein Synthesis

0 Protein Destination
'3 Metabolism

I Energy

‘3 Development

I Cellular Transport

I Cellular Organization
I Signal Transduction

' 0 Cellular Biogenesis

U Defense
I DNA Synthesis

' I Unclassified

I Miscellaneous

00

distribution of the categories was supported by contingency test (p < lE-5).

41

Transport (5) Transcription (7)
Miscellaneous (6) Protein Synthesis (4)

   
 
  
  

Protein Destination (7)

Unclassified
Proteins (25)

. Metabolism (18)
Cell Divrsion (3)

Defense (4)

Cellular Biogenesis (1) " ’ ‘ Energy (2)

Development (1)

Si nal

Transductionw) Cellular Cellular Transport (1)
Organization (9)

Figure 1-6. Functional categorization of the 7,760 poplar unigenes that had
counterparts in all three groups (dicots monocots and pine) of plant species.
Percentage of the unigenes in each functional category is given in parenthesis. For this
analysis the plant gene sequences were pooled into three groups. All of the sequences
from Glycine max Medicago truncatula Arabidopsis thaliana Lycopersicon esculentum
were pooled into the ‘dicots’ group which had a total of 151,832 sequences. Likewise, the
sequences from Oryza sativa Zea mays Triticum aestivum into the ‘monocots’ group with
a total of 220,859 sequences.

42

cn1 5054Aspn
cn15055Aspn

79
I 8 cn687ch
L__
EEO
L‘_

 

cn15056Aspn

cn959ch

cn17060Aspn

cn13829Aspn
cn1 607ch

31 cn16853Aspn

cn13069Aspn

 

 

4 cn1313Tmo
24069786ch '
_93| cn16843Aspn
99 cn14348Aspn
cn15588Aspn
cn666ch
cn14982Aspn
cn14980Aspn '
anOOTma

 

 

 

 

 

 

 

 

Fail

 

 

 

 

 

Figure 1-7. Phylogenic tree of DNA sequences of peptidylprolyl isomerase family,
constructed using the neighbor-joining method.

To support the gene duplication within the families, phylogenic analysis using
peptidylprolyl isomerase protein family were performed as an example. Bootstrap
confidence values (%) based on 100 replications are shown for each branch.

43

Table 1-1. Poplar ESTS and unigene set statistics.

 

 

 

Species Number Number of Average
of ESTS Uniggres Length (bp)

P. tremula x P. tremuloides 56147 20864 511

P. trichocarpa 24050 9113 542

P. tremula 14091 7295 461

P. tremuloides 11543 4683 507

Total 105831 41955 ND "

Poplar unigenes identified by clustering

the combined 41955 sequences 41955 a 29818 ND

Filtering with BLASTCLUST 29644 " ND

Final Poplar Unigene Set 25282 c 572

 

" The sequences consists of four species unigene sets.

b BLASTCLUST algorithm was used to further remove any sequences with 95 % nucleic
acid identity covering 90 % length region. ’

c Any ESTS shorter than 250-bp sequence read were discarded in order to reduce
sequence length bias in subsequent analyses. The poplar unigene set consist of 15,178
singletons and 10,104 contigs

d ND not determined.

Table 1-2. The 14 Highest copy number poplar unigene families determined by
tBLASTX with an E-Value threshold of <1.0E-20.

 

 

Representative
. GenBank . . Gene of Arabidopsis
{GE 11:). 3:31:11? Accession Copy No. mmgefeiuncillgn 0f Corresponding Gene Copy
Number Arabidopsis Number
Family
cn726Aspn BU896558 53"(73b) Protein kinase At1g56140 501
cn8047Aspn 811894782 33 (53) $31M‘doma‘“ At4g02890 19
cn16304Aspn BU810510 41 (48) GTP-binding protein A13g46830 60
24067636ch BU876112 31 (44) Pg: transcmptlonAt4g18770 133
cn15003ASpn BU894972 33 (37) $3$M0nflganng At3g08690 30
cn933Aspn BU895108 25 (34) Peroxidase At2g18150 73
cn66l4Aspn BU831012 24 (33) Cytochrome P450 At4g37400 152
cn15387Aspn BU831392 24 (31) aggrrbosylatron At2g47170 25
cn11888Aspn 81128682 19 (30) Tubulin Atlg50010 17
cn14980Aspn BU831411 20 (29) Pept’dylprf’ly‘ At4g34870 20
lsomerase
24074413ch BU882889 21 (28) MADSib‘.“ At4g18960 41d
transcrlptlon factor
cn753ch BU882276 18 (28) Water channel protein At3g16240 24
cn51ch BU874797 7 (27) ABAilnduc‘b'e At5g38760 3
protein
23534478Tmo AY095297 18 (26) Cellulose synthase At4g39350 26

 

" The copy number estimation was further refined by excluding the genes that 1) had no
matches to the known multigene family proteins present in NCBI protein database, 2)
were shorter than 500-bp, 3) shared <90 % nucleic acid sequence identity with any
member of the multigene family over 90 % length region, 4) did not share the
conserved region mostly due to its short length, or 5) had long gap regions probably
resulting from chimera when they were aligned using ClustalW software (Thompson et

al. 1994).

b The copy number determined using tBLASTX with an E-value threshold of $1.0E-20.
c To support the gene duplication within the families, we performed phylogenic analysis
using peptidylprolyl isomerase protein family as an example, which was provided as

Figure 1-7.

d Actual number of Arabidopsis MADS-box genes that can be identified by MADS-box
domain search is ~80 (Riechmann et al., 2000). The number 41 reflects the outcome of
tBLASTX search using the entire cDNA sequence.

45

CHAPTER 2

Genetic regulation of annual growth cycle in woody perennial, Populus

Abstract

Temperate woody plants have evolved annual growth cycle as an adaptive mechanism for
winter survival, which is controlled by interactions between endogenous and
environmental factors (e.g., daylength, temperature). To understand the genetic regulation
underlying this process, we carried out a series of global transcriptional analysis using
stem samples from poplar trees grown in field and under controlled environment
conditions. The analysis showed that extensive metabolic switches and alterations in
cellular functions occurred during transition from rest to quiescent. The genes involved in
the pentose phosphate pathway were elevated and defense-related genes were over-
represented during early winter whereas the genes involved in fermentation and fatty acid
B-Oxidation were upregulated. Moreover, signaling-related genes were over-represented
during late winter. Furthermore, clustering of these genes into eight groups has revealed
that plants regulate seasonal growth by integrating environmental factors with
developmental stages. For example, short daylength (SD) could modulate the cold signal
in at least two different ways to achieve maximum cold hardiness. First, short daylength
could induce cold associated genes independent of actual cold temperatures, resulting in a
higher expression of the genes in both cold and SD than in cold alone. Second, the short
daylength itself did not induce some of the genes associated with cold acclimation, but

did enhance their expression in response to cold temperature. This mechanism may help

46

plants to control cold hardiness more closely timed with seasonality than unpredictable

temperature alone does.

47

Introduction

In the temperate zone, woody perennial plants exhibit an annual cycle alternating
between growth and dormancy. By integrating information from multiple environmental
cues, they can coordinate their development with seasonal changes, and ultimately,
control their growth capacity. During the warm summer season, woody plants grow
actively, and at the end of summer they enter dormancy and initiate cold acclimation. At
this stage, cambial dormancy is regulated by internal factors, which has been denoted rest
and characterized by non-responsiveness to indole-3-acetic acid (1AA) under favorable
conditions for cambial activity (Little and Bonga 1974, Little 1981). Short daylength
(SD) has been known to be one of the most important environmental factors controlling
the onset of dormancy and cold acclimation (Greet et al., 1989; Olsen et al., 1997). By
responding to SD concordant with local seasonal change, plants are able to synchronize
dormancy development and cold acclimation with the coming winter cold. Previous
studies have showed that cold acclimation facilitated by low temperature (LT) could be
uncoupled from dormancy, suggesting that SD and LT might utilize independent
pathways for the cold acclimation (Welling et al., 2002). However, LT and dormancy
development would be needed for plants to achieve the higher degree of cold hardiness,
i.e., second stage of cold hardiness (Anisko et al., 1994; Christersson, 1978). Indeed,
dormant deciduous genotypes always reach a higher extent of cold hardiness than the
non-dormant evergreen genotypes, as discovered in peach (Arora and Wisniewski, 1994).
In nature, dormant plants could induce a second stage of cold hardiness by responding to
the subsequent low or freezing temperatures that occur in late fall. The facts indicated the

possible interaction between dormancy development and cold acclimation. Cold is also

48

required to break dormancy, after which the cambium gradually regains the ability to
produce xylem in response to 1AA, referred to as a chilling requirement (Little and
Bonga, 1974). When fully responsive, the cambium is in the quiescence stage of
dormancy, which is imposed solely by adverse external factors, typically by low
temperature. Upon the return of warmer temperatures in the Spring, plants are gradually
dehardened, primarily by responding to the rising temperature, and the cambial cells are
released from dormancy (Mellerowicz et al., 1992). The periodicity regulated timely with
seasonal changes is a dominant factor in the distribution, growth, and survival of woody
plants (Samish, 1954; Weiser, 1970).

Despite these early Observations, the molecular basis of seasonal development
remains a mystery. Perhaps this lack of understanding of the molecular mechanism
underlying seasonal development has been, in part, due to superimposition of various
internal and external events occurring during annual growth cycle, which make it difficult
to associate molecular changes with specific events. In nature, dormancy development
and cold acclimation/deacclimation partially overlap and are interrelated. Thus, it is
important to identify which of the changes is associated with specific events in the
seasonal growth cycle. Although there were some previous Studies that have identified
the genes associated with different aspects of dormancy or cold acclimation, they were
limited to a few of genes or a short period of study time (Espinosa-Ruiz et al., 2004;
Rowland and Arora, 1997; Welling et al., 2004).

To understand the molecular basis of seasonal development, we sought to obtain
global transcriptional profiles of eastern cottonwood (Populus deltoides) on an annual

cycle as well as under controlled environmental conditions. We identified 714 and 743 of

49

genes that were differentially regulated during seasons in bark and xylem, respectively.
Clustering and functional analysis of the genes showed that various functional gene sets
were differentially expressed from season to season, which provided us with a global
characterization of seasonally regulated biological processes in woody plants.
Furthermore, we analyzed seasonal behaviors of the genes in context of the
responsiveness to environmental factors. These approaches showed that the cold-
associated genes are under both developmental and environmental control, and thus
provided new insights into the interrelation between cold acclimation and dormancy

development.

Materials and Methods

Plant Materials and Growth Conditions.

Populus deltoides (clone ill-129) was used throughout this work. First, stem cuttings were
planted and grown in a green house under long day (LDW, 18 h) or short day (SDW,
natural photoperiod) conditions for 13 weeks (Aug. 15 - Nov. 22, 2002). The temperature
was kept at 23°C and the plants were watered three times a week. Then half of the plants
from each photoperiod group were transferred to lower temperature (5 °C, LDC and
SDC) conditions for three weeks. For a drought treatment (LDD), we took some of the
plants that had been grown under LD and discontinued watering them for 40 days.
Second, for the plant samples grown under the natural environmental conditions, stem
cuttings were planted in field soil at the Tree Research Center on the campus of Michigan

State University (East Lansing, MI). Those plants were harvested in the indicated months

50

(i.e., July 2002 to June 2003). Bark and xylem tissue samples were collected separately

from the subject plants and stored at — 80 °C until used in the microarray experiments.

Experimental Design and Data Analysis.

The first set of experiments included a 12-month comparison between the bark and xylem
tissues. The second set of experiments compared the bark and xylem tissues of plants
grown under controlled environmental conditions, i.e., LDW, LDD, LDC, SDW, and
SDC (Figure 2-1). To avoid variability due to sample preparation, the bark and xylem
tissue samples used in all of the experiments were pooled from at least three independent
plants. To eliminate potential variability due to aRNA amplification, we duplicated the
amplification process. Each one of the duplicate amplifications was labeled with Cy5 or
Cy3 using a dye swap strategy. Therefore, four data sets were derived for each
experimental unit. A woven loop design was used for both sets of experiments. In the
first set of experiments, RNA from each treatment plant group was competitively
hybridized to the RNA in each of the other plant treatment groups. In the second set of
experiments, RNA from the plants at N-th month was competitively hybridized to RNA
from the plants at n-l‘h, n+1‘h, n+4th and n+8‘h month (Figure 2-1). For statistical analysis,
systemic errors were corrected using lowess method after a log2-transformation of raw
data. The normalized signal values were analyzed with ANOVA to evaluate the sources
of variability. This method was previously described in Yang et a1 (2002) and
implemented in MAANOVA package (http://www.jax.org/staff/churchill/labsitel-
software/anova/rmaanova) in the statistical language R. The MAANOVA package

provides three statistics tests (called F 1, F2 and F3). The F1 is based on the error variance

51

of a single gene, the F3 is based on common error variance across all genes, and the F2 is
based on a hybrid of the other two tests (Chui and Churchill, 2003). Probes for which
there were no significant treatment effects (p >0.001, >0.05, and >0.05 for F1, F2 and F3
for bark and p >0.0001, >0.005, and >0.005 for xylem) were removed from the dataset.
Finally, the signal values were further normalized by subtracting the variances caused by
factors other than the treatment effects, thus generating a population with the relative
expression values of each gene across the treatments. The relative expression values are
treatment-based normalized data in the sense that effects due to the array and dye have
been removed (Cui and Churchill 2003). Hierarchical and k-means clustering were
performed on the normalized data by using CLUSTER and dendograrns and k-lmeans

groups were visualized with TREEVIEW (Eisen et al., 1998).

Preparation of the cDNA Arrays

The six cDNA libraries were created using RNA prepared from either bark or xylem
tissues harvested on Sep-06-01, Oct-30-01 and Apr-09-02. Total RNAs were extracted
according to the modified Hu, et al (2002), and poly(A)+ enriched RNAs were prepared
using the Poly(A)PuristTM MAG (Ambion, Austin, TX). cDNAs were synthesized using a
SMART cDNA Library Construction Kit (Clontech, Palo Alto, CA), directionally ligated
to pBluescript which was modified to contain sfi I site, and transformed to
ElectroMAXTM DHlZSTM competent cells (Invitrogen, Carlsbad, CA) by electroporation.
The six cDNA libraries were subtracted by the randomly selected 1,920 sequences from
the September bark and xylem cDNA libraries following Scares et a1. (1994) with a few

of modifications. The driver RNAs were prepared using Megascript T3 system (Ambion)

52

and BrightStar Psoralen-Biotin Nonisotopic Labeling Kits (Ambion). Target phagemid
ssDNAs were prepared using M13K07 as a helper phage. The subtracted ssDNAs by the
driver antisense RNAs were repaired to dsDNA using Klenow DNA polymerase and
transformed to ElectroMAXTM DH12STM to make the subtracted cDNA libraries. For the
cDNA array, the 960 clones were randomly selected from each of the six subtracted
libraries, which made a total of 7,680 clones including 1,920 of the driver cDNA clones.
The cDNA inserts were amplified with T7 and T3 primer directly from the selected
colonies. The 7,680 cDNA inserts were spotted on SuperAmine glass slides (Telechem,
Sunnydale, CA) in a format containing 32 blocks of 240 inserts using an Omnigridder
robot (Gene Machines, San Carlos, CA). After spotting, the slides were UV light cross-
linked, washed and denatured as described by the manufacturer. Hybridized micrOarrays
were scanned using an Affymetrix 428 array scanner (Affymetrix, Santa Clara, CA) and
analyzed with GenePix Pro 4.0 software (Axon Instruments, Union City, CA). The 4,965
probes that showed significant signals when compared to the background were selected,
sequenced and then clustered by StackPack program (Miller et al., 1999) resulting in

1,953 contigs and singletons that were supposed to represent unique genes.

RNA amplification. Probe Preparation and Hybridilztion.

For hybridization, total RNAs were prepared according to the modified Hu, et al (2002).
The amplification of total RNA was performed based on a previously described method
(Wang, et al., 2000). The amplified aRNA was cleaned up using an RNeasy mini kit
(Qiagen, Valencia, CA). Two micrograms of aRNA were labeled with aminoallyl-dUTP

(Sigma, St. Louis, MO) during reverse transcription using random hexamers and

53

SuperscriptTM II (Invitrogen). The labeled probes were purified using a QIAquick PCR
purification kit (Qiagen). After purification, the probes were coupled with Cy3 or Cy5
monoreactive dye (Amersham, Piscataway, NJ). The labeled cDNA probes were then
hybridized to cDNA arrays at 42°C for 12-18 hours in 40111 of hybridization buffer (50 %
formarnide, 5X SSC, 0.1 % SDS and 5 ug yeast tRNA). Following hybridization, the
arrays were washed with 1X SSC, 0.1% SDS at 42°C for 5 min, 0.1X SSC, 0.1 % SDS at
room temperature for 5 min and deionized water for 5 min. Hybridized microarrays were
scanned using an Affymetrix 428 array scanner (Affymetrix) and analyzed with GenePix

Pro 4.0 software (Axon Instruments)

Northern Blotting

Seven micrograms of each RNA sample were subjected to electrophoresis through 1.1%
formaldehyde-agarose gels and transferred to Hybond-N nylon membranes (Amersham).
Hybridization was performed in UltraHybTM (Ambion) following the manufacturer’s
protocol. Probes were 32P-dATP-labeled with Strip-EZ DNA kit (Ambion) and Northern

blots were analyzed via QUANTITY ONE 4.1.1 software (Bio-Rad, Hercules, CA).

Results

Relationship between annual growth capacity and local environgntal change.

To provide an assessment of relationship between the annual growth capacity of poplars
and local seasonal changes, we measured the length and diameter growth of poplars in
monthly intervals throughout the experiment year, from July 2002 to June 2003, and

correlated the results to local seasonal changes, i.e., day length and temperature (Figure

54

2-2). Length and diameter growth were highly correlated indicating that shoot apical and
cambial growth are correlated. Measurable diameter growth was first detected between
April and May, correlating with bud break in mid-April, and reached its peak between
July and August. The growth stopped around mid-September with the setting of terminal
bud. As expected, plants entered dormancy in mid-September under relatively warm
temperatures, around an average of 18 °C, and about 12 hr Of daylength, implying that
poplar plants perceive SD to be less than 12 hr. Re-growth of poplars in the spring was
likely to be correlated with rising temperature rather than photoperiod because the plants
did not resume their growth despite the increase in daylength, to longer than 13 h, that
occurred in April. Based on these observations and previous studies (Chen et a1.,_2002;
Howe et al., 1995) we defined the growing season as the period from May to August,
September being the onset of dormancy, and the period from October to the following
April as the non-growing season during which the cambium could transition from rest to
quiescence along with the accumulation of chilling. This definition was further used to

interpret results from the microarray experiment.

Major alterations in gene expression during winter reflect the difierent stages of cambial
dormancy.

Seasonal growth of temperate woody plants is regulated by poorly understood
interactions between endogenous and external environmental factors, which are
orchestrated by multi-genetic factors (Chen et al., 2002). To understand the genetic
regulation of seasonal growth, we investigated the global transcript profiles for an annual

cycle using cDNA microarrays. The microarrays carrying 1,953 tentative unigenes were

55

used to determine the levels of mRNAs in bark and xylem tissues from field-grown
poplar plants in monthly intervals as shown in Figure 2-2. During the annual growth
cycle, 714 of the genes in bark and 743 from xylem were identified to show significant
differential expressions, which had been determined by ANOVA as described in the
method section. These selected genes were further clustered hierarchically, based on the
similarity of their monthly expressions. The genes in bark were grouped largely into three
clusters (Figure 2-3A) while the genes in xylem, grouped into two clusters (Figure 2-4A).
The distinctive expression profiles in bark within the non-growing season were consistent
with previous studies showing the existence of the two stages of cambial dormancy, i.e.,
rest and quiescence (Little and Bonga, 1974). Thus, we speculated that cluster A and
cluster B might represent rest and quiescence, respectively. This speculation was further
supported by the functional correlation of the genes in the clusters to the corresponding
dormancy states as discussed later. Furthermore, the fact that the temperatures in Jan
were similar to those in Feb, when cluster A/B genes were up- or down-regulated,
suggests that the genes might be regulated not only by external environmental factors but

also by internal factors, e. g., dormancy development.

Up-regulation of the genes associated with cold hardiness during early winter.

In bark, a major fraction of the genes (443) were identified as up-regulated in the growing
season (Figure 2-3, Cluster C), representing 22 % of the total genes on the array. That
group was followed by cluster A, identified as up-regulated in rest and representing 7 %
(138), as well as cluster B, identified as up-regulated in quiescence and representing 6.8

% (133). On the other hand, in xylem, 34 % (669) of genes in the growing season, and

56

9.5 % (186) of the genes in non-growing season were identified as being up-regulated.
The major biological activities represented by each cluster may be indicated by the over-
or under-representation of categories within the clusters. For functional categorization,
the genes of each cluster were sorted into ten functional groups according to MIPS
(http://mips.gsf.de) categorization (Figure 2-3B, 3-4B). The genes that could be assigned
to any group due to unknown function or had no significant matches in the current
database were assigned to Unclassified or No-Hit groups, respectively. In cluster A of
bark and xylem, the cell defense/rescue-associated genes and No-Hits were over-
represented, and signal transduction and protein synthesis-associated genes were under-
represented, indicating activation of cell protection mechanisms with reduced cellular
communication and activity. The cell defense/rescue category codes for detoxification-,
cold/drought- and pathogenesis-associated proteins (Figure 2-5). Many of them were
reportedly induced during cold acclimation in arabidopsis and winter annuals, suggesting
their roles in protection of cells from cold damages. For example, the antioxidant proteins
may be involved in the detoxification of reactive oxygens probably caused by excess
excitation of photosystem and imbalanced respiratory metabolism (Fowler and
Thomashow 2002). Additionally, pathogenesis-associated proteins have been known to
have antifreeze activity (Hon et al., 1995). However, unlike transiently up-regulation of
the antioxidant genes in arabidopsis, the genes in poplar were induced throughout winter,
suggesting that they could have more functions in woody plants. Indeed, the antioxidants
have been speculated to be associated with breaking dormant bud based on a correlation
of increased capacity to scavenge free radicals with chilling accumulation (Wang et a1

1991; Wang and Faust, 1994). Down-regulation of these genes under freezing

57

temperatures in Feb supported this notion for cambial dormancy as well. One intriguing
result is the up-regulation of the putative histone deacetylase gene, suggesting possible
involvement of epigenetic regulation in transition from rest to quiesecence. During
winter, temperate woody species have been known to considerably modify their
metabolic processes, including the synthesis of proteins and sugars and membrane lipids.
For example, the proportion of phospholipids and unsaturated fatty acids increases, while
free sterol decrease (Wang and Faust 1988), and the pentose phosphate (PP) pathway is
elevated (Sagisaka, 1974). The biochemical alterations of plasma membranes enable
plants to maintain permeability of plasma membrane by increasing fluidity of membranes
at low temperatures and the PP pathway could provide the reducing power needed for
desaturation (Thomashow, 1999). Consistent with previous studies, our results showed an
up-regulation of genes encoding enzymes associated with PP pathway, lipid desaturation,
and cell wall modification during rest (Figure 2-5). Additionally, transcripts encoding
galactinol synthases, the enzyme that catalyzes the first committed Step in the synthesis of
raffinose, and invertase were up-regulated, indicating the accumulation of soluble sugars,

which are related to cold hardiness (Wanner and Junttila, 1999).

Elevation of signaling-associated genes occurred diging quiescent dormanc.

For bark cluster B upregulated during late winter and early winter, signal transduction-
associated genes were over-represented while cellular organization- and protein
synthesis-associated genes were under-represented, indicating the activation of cellular
signaling with reduced cellular activity during quiescence (Figure 2-3B). These results

are consistent with the previous Observation in which cambial cells could restore their

58

capability of responsiveness to growth hormones without any visible growth after being
exposed to some period of chilling (Little and Bonga, 1974). The reactivation of the
cellular signaling was reflected by receptor kinases, intracellular signaling proteins and
the genes implicated in cytokinin and auxin signaling. Cytokinin as well as auxin has
been known to be involved in breaking bud dormancy (Cutting et a1 1991; Hewett and
Wareing, 1974), suggesting the potential role of this hormone in the cambial transition to
quiescence (Figure 2-6). Additionally, up-regulation of the putative cyclin gene further
supports restoration of growth-competency after fulfillment of chilling requirements.
Compared to rest dormancy, biological activities during quiescence dormancy have been
less studied. We found that there were considerable changes in metabolism during late
winter (i.e., quiescence). The transcripts encoding fermentation, (e.g., pyruvate
decarboxylases and alcohol dehydrogenase) were up-regulated, suggesting that plants
might be under oxygen deficiency stress, probably caused by ice encasement formed
during freeze-thaw cycles in late winter (Bertland et al., 2003). We also found that
transcripts encoding fatty acid oxidizing enzymes, e.g., acyl-COA dehydrogenase, fatty
acid CoA ligase, and fatty acid B-oxidation multifunctional protein were induced.
Interestingly, induction of fatty acid oxidation is consistent with previous histochemical
Observations showing a gradual decline of lipid bodies in bark according to proceeding
toward reactivation of cambium after transition to quiescence (Farrar and Evert, 1997;
Riding and Little, 1984; Arend and Formm, 2003). The hydrolysis of lipids might have an
effect on the freeing of water, which increased gradually with the end of bud
endodormancy (Erez et al., 1998) and could allow plants to preserve the cryoprotective

sugars during late winter. In addition, lignin synthesis-, phospholipid- and sterol-

59

associated genes were found to be up-regulated during quiescence. Although we could
not determine the roles of these genes in cold hardiness or quiescence, these findings may
reflect that, with dormancy break, there are changes in cell wall and membrane
composition allowing increased permeability of solutes and water to the cytoplasm.
Overall, the functional analysis suggests that plants have different biochemical
characteristics between rest and quiescence, and the difference is represented in

transcription levels.

The dissection of different stages of cold acclimation and dormancy under controlled
growth conditions.

To dissect the effects of temperature and photoperiod, plants were grown under five
different environmental conditions as described in methods and used for microarray
experiments. In our growth system, LDC-treated plants did not show any visible change
of the shoot but reduced vessel size in xylem, indicating reduced activity of cambial cells
(Figure 2-7). LDD- or SDW-treated plants induced terminal bud and showed compressed
layers of cambial zone, indicating that they were in dormancy. However, cambial cells in
SDW plants had thickened cell wall that is characteristic of the rest (Chaffey et al., 1998).
Based on these observations and previous studies, we speculated that LDC would induce
cold acclimation without dormancy; LDD, quiescent dormancy; SDW, both first stage of
cold acclimation and rest dormancy; SDC, second stage of cold acclimation and partly
quiescent dormancy. Thus, comparison between them could lead us to associate the genes
with various aspects of seasonal development. The speculation was further supported by

hierarchical clustering based on the controlled environment and monthly expression

60

profiles (Figure 2-8). In bark, SDW were the closest to September; SDC, the closest to
December; LDD and LDC are largely grouped to growing and non-growing months,
respectively. On the other hand, in xylem, SDC is close to Oct/Nov while SDW and
LDD, LDC is largely grouped to growing and non-growing months. The difference
between bark and xylem implies that they might be under different regulation

mechanism.

The modulation of cold signal by SD-induced dormang

For the next level of analysis, we performed k-means clustering based on seasonal and
controlled environmental expression profiles of the genes selected in Figure 2-8. These
genes were placed into eight groups (Figure 2-9). The genes of group G1, G2, G3 and G4
were largely expressed during the growing season, the genes of G5, G6, G7 and G8 were
expressed during the non-growing season. The differences in responsiveness to controlled
treatments was of interest since they may suggest what treatment(s) was likely to be a
main regulatory group factor(s) during the annual cycle. Furthermore, the variety of
responsiveness suggests that multiple pathways might be involved in regulation of the
genes. For example, all of G1 — 4 groups have responded negatively to SDC while they
have responded differently to LDD or SDW. G1 was up-regulated during the active
growing season and responded negatively to four treatments, suggesting that they were
involved mostly in active growth. Consistent with this view, G1 includes active growth-
related genes, i.e., photosynthesis-, cell division-, and auxin signaling-associated genes
(Table 3-1). G2 showed the positive responsiveness only to LDD, indicating that their

genes might be associated with drought stress. G3 responds positively to LDD and SDW

61

and G4 responds positively only in SDW, indicating the association with onset of
dormancy. The positive response to SDW appeared to be correlated with a relatively high
level of expression in September. Interestingly, G3 included starch synthesis- and amino
acid- as well as dehydration- related genes (Table 3-1), which have been reported to
occur before onset of dormancy (Perry, 1971). Moreover, the fact that they could respond
to both SD and drought might provide a molecular insight on how drought stress could
hasten dormancy (Rinne et al., 1994). On the other hand, genes in G4 were up-regulated
solely by SDW rather than LDC and had a higher expression in September when
compared to the genes in G3. This result suggested that G4 genes might be involved in
the onset of dormancy rather than cold acclimation. G4 includes many cell wall-
associated genes, suggesting that cell wall thickness induced during dormancy
development could be regulated primarily by SD signal. Finally, the negative
responsiveness to LDC and SDC may be responsible for down-regulation of G1-4 gene
groups during winter.

The groups up-regulated during late fall/winter, GS — 8 (Figure 2-9), are of the
most immediate interest for an integrative view of environmental and developmental
factors in dormancy/cold hardiness. It has been known that SD and cold could induce
cold acclimation independently (Welling et al., 2002). Consistent with the fact, many of
genes (G5 -7) were up-regulated by SD as well as cold, suggesting that these genes
should be under independent regulation of cold or SD. Woody perennials have been well
known to require both cold and SD to achieve a higher capacity of cold hardiness.
Consistently, G6 and 7 showed much a higher expression in SDC than LDC or SDW

alone. Furthermore, the different response to cold between, before and after dormancy

62

induction suggested that dormancy development might modulate cold signal.
Interestingly, a few of the genes up-regulated in early spring (i.e., bark cluster B) were
grouped into G8 that showed up—regulation in LDD and LDC. The result was consistent
with the fact that plants were imposed only by an adverse environment during
quiescence. Moreover, the fact that most of cluster C genes were excluded from this
analysis due to a non-significant difference between treatments suggested that the cluster
C genes might be regulated by the unknown internal factors (e.g., hormones) rather than

external environment (e.g., day length and temperature)

Validation of microarray results
Nothem blots were performed to confirm microarray results. Some examples are shown
in Figure 2-10, which represents genes from G6 or G7 (Figure 2-9). The selected gene

results followed the overall patterns obtained with microarrays.

63

Discussion

Annual growth cycle in woody perennials is a complex developmental process that is
controlled by interactions between environmental factors and internal factors (Arora et
al., 2003; Dennis, 1994; Lang, 1987). To understand the molecular basis of seasonal
growth processes, we designed microarray experiments by which we could associate gene
expressions with various environmental factors or growth stages. Using this approach, it
was possible to show that seasonal growth was regulated through interaction between

environmental and developmental signals at the molecular level.

Cell wall modification is associated with dormancy while the accumulation of energy
reserves is associated with both growth cessation and dormancn

Growth cessation has been known to be a prerequisite of cold acclimation since actively
growing woody plants could not acclimate to cold. Moreover, treatments that stop growth
(e.g., nutrient deficiency, water stress and chemicals) could increase hardiness to a
limited extent (Weiser, 1970; Perry, 1971). However, the effect of growth cessation on
cold hardiness is not clearly understood in woody plants. Since SD and drought could
induce growth cessation, G3 responding positively to both drought and SD could provide
insight on how the growth cessation promotes cold hardiness. Interestingly, G3 includes
starch synthesis-associated genes, i.e., granule-bound starch synthase and starch
branching enzymes (Table 3-1). The starch accumulation is necessary for cold hardiness
because woody plants severely depleted in photosynthetic reserves cannot acclimate
(Weiser, 1970). The coordinated up-regulation of plastidial forms of aldolase and
Glyceraldehyde 3 phosphate dehdrogenase and cytosolic forms of triosephosphate

isomerase implies that bark tissue may obtain carbon sources by its photosynthesis as

well as importing them from source organs, e.g., leaves. Thus, growth cessation by water
stress appears to trigger accumulation of photosynthates that could contribute to
promoting cold hardiness. In addition, the amino acid synthesis and ribosomal genes were
found in G3, which might be related to nitrogen and sulfur mobilization. The coordinated
expression of starch and amino acid synthesis associated genes suggests that growth
cessation triggered by the prolonged exposure to adverse environments could activate the
pathway to accumulate the energy reserves, e.g., starch and storage proteins, which could
be used in return of a favorable environment. On the other hand, based on microscopic
observations (Figure 2-7), SDW-treated plants were in rest dormancy while LDD-treated
plants were in growth cessation (i.e., quiescence). Thus, we speculated that G4
responding solely to SD might be associated with rest dormancy. G4 included many cell
wall and lignin/flavonoid associated genes. During cambial dormancy, it has been well
known that extensive cell wall thickening and modification occur (Ermel et al., 2000).
Thus, the cell wall related genes in G4 might reflect the difference between dormancy
and growth cessation. Assuming that the structural changes are associated with rest
dormancy while the metabolic changes are associated with growth cessation/quiescent
dormancy, it is interesting to hypothesize that the biochemical differences could account
in part for the different reversibility between them to active growth state in return of

favorable environmental conditions.

The transition from rest to quiescence

The transition from rest to quiescence in the cambium is gradual and is accomplished by

chilling (Little and Bonga 1974). However, the mechanism of chilling induced-transition

65

to quiescence is little understood. One Of the activities associated with satisfaction of
chilling requirement is a restoration of cell-to-cell signaling networks among individual
cells of the meristem that had been disrupted during dormancy induction. Resumption of
cell-to-cell communication, regulated by plasmodesmata (PD), may allow for symplastic
movement of small signaling molecules, hormones, or proteins responsible for dormancy
release. Rinne et al. (2001) found that the restoration of the symplastic organization of
birch apical meristem was correlated with removing of 1,3-B-glucan, which is digested by
1,3-B-glucanase. This may be the case on cambial dormancy. Putative 1,3-B—glucanase
highly increased by chilling and decreased in time supposed to be in quiescence (Figure
2-4A). Moreover, Schrader et al. (2004) reported up-regulation of 1,3-B—glucan synthase
in poplar dormant cambium, which can produces 1,3-B—glucan to block PD. In addition,
we found up-regulation of pectin methylesterase and pectin glucuronyltransferase during
early winter (Figure 2-4A). Recently, pectin glucuronyltransferase has been reported to
be essential for intercellular attachment of meristemic cells, of which the disruption could
disconnect cell signaling via the plasmodesmata (lwai et al., 2002). The up-regulation of
the genes during chilling led us to assume that these enzymes could be associated with
symplastic restoration of cambium. An intriguing finding was the upregulation Of a
popoar homolog of histone deacetylase during rest (Figure 2-4A). Many characteristics
are shared between vemalization and chilling requirement, e.g., the range of effective
temperature, the experienced period and the site of perception, and the role of histone
deacetylation on transition from vegetative to reproductive stage by prolonged exposure
to cold (i.e., vemalization) has been well known in arabidopsis (He et al., 2003). Thus,

the up-regulation of the gene provides the possibility that chilling requirement could be

66

regulated by transcriptional regulation through the modification of chromosomal
structure.

Woody plants to maintain cold hardiness differently depending on growth stages.
Woody plants are hardly dehardened during rest, whereas quiescent plants are rapidly
dehardened in response to rising temperature (Perry, 1971, Anisko et al., 1994).
However, there is little data on the nature of enzymes involved in this process. Our data
suggested that the difference on dehardening rate could be accounted by reactivation of
signaling-related genes (e.g., receptor kinases, calmodulin) during late winter. In
addition, the coordinated up-regulation of phospholipid (i.e., diacylglycerol kinase and
phosphoesterase) and sterol modification associated genes (i.e., cholesterol
acyltransferase, sterol reductase and sterol sulfotransferase) could imply that the
molecules might play a role as intra/intercellular signaling messengers. Recently, the
significant role of sterols on vascular patterning and differentiation were reported
(Carland et al., 2002; Schrick et al., 2004). Since dehardening process is correlated with
re-activation of cambium, thus, it is plausible that sterols could act on reactivation of the

cambium.

The modulation of cold signal by SD

It has been well known that both low temperature and SD signal are required for
maximum cold hardiness. However, how SD signal could modulate the response to cold
at the molecular level is less understood. Our expression profiles may provide a
molecular glimpse of these phenomena. According to our results, there were at least two

mechanisms. First, SD enhanced the expression of the cold-associated genes as well as

67

SD itself induced the genes independently of cold. As shown at G6 of k-means clustering
(Figure 2-9), cold associated genes had much a higher expression in SDC than LDC.
Chen and Li (1978) proposed that LT, SD and water stress trigger independent frost
hardening mechanisms, and total hardiness would be a sum of these. It might be reflected
by G6. SD and cold signal could induce the genes independently, resulting in the higher
expression of the genes than cold signal alone. Second, SD signal itself could not induce
the cold-associated genes but it enhanced the genes much higher than cold signal alone
when plants had cold signal, which was supported by G7 (Figure 2-9). Thus, the
responsiveness to cold signal appeared to depend on how the dormancy development
proceeded. Moreover, the fact that both groups of genes were up-regulated in first stage
of cold acclimation, (i.e., LDC) suggested that the higher level of cold hardiness (i.e.
SDC) could be achieved by a higher expression of the same battery of genes as with the
first stage of cold acclimation. In annual herbaceous plants, the expression level of cold-
associated genes is regulated primarily by responding to the degree of temperature, i.e.,
non-freezing temperature or freezing temperature (Zhu et al., 2000) while in woody
plants, they are induced differentially depending on different development stages as
shown here. This fact suggests that woody plants may evolve different mechanisms to
attain extreme freezing tolerance. The ultimate survival Of woody plants is dependent on
not only the maximal capacity of cold hardening, but also on the timing and rate of cold
acclimation and the stability of cold hardiness against unseasonably warm periods during
winter (Weiser, 1970). Hence, the successful performance of a woody species in a
particular locality implies synchronization of the annual development of cold hardiness

with the seasonal temperature changes. By controlling cold hardiness via developmental

68

stage, which is regulated by predictable photoperiod, plants are able to distinguish
between unseasonable and seasonable cold and keep cold hardiness in accordance with

local seasonality.

69

References

Anisko T, Lindstrom OM, Hoogenboom G (1994) Development of a Cold-Hardiness
Model for Deciduous Woody-Plants. Physiologia Plantarum. 91: 375-382

Arora R, Rowland LJ, Tanino K (2003) Induction and release of bud dormancy in woody
perennials: A science comes of age. Hortscience. 38: 911-921

Arora R, Wisniewski ME (1994) Cold-Acclimation in Genetically Related (Sibling)
Deciduous and Evergreen Peach (Prunus-Persica L Batsch) .2. A 60-Kilodalton Bark
Protein in Cold-Acclimated Tissues of Peach Is Heat-Stable and Related to the
Dehydrin Family of Proteins. Plant Physiology. 105: 95-101

Bertrand A, Castonguay Y, Nadeau P, Laberge S, Michaud R, Belanger G, Rochette P
(2003) Oxygen deficiency affects carbohydrate reserves in overwintering forage
crops. J Exp Bot. 54: 1721-30

Carland FM, Fujioka S, Takatsuto S, Yoshida 8, Nelson T (2002) The identification of
CVPl reveals a role for sterols in vascular patterning. Plant Cell. 14: 2045-58

Chaffey NJ, Barlow PW, Barnett JR (1998) A seasonal cycle of cell wall structure is
accompanied by a cyclical rearrangement of cortical microtubules in fusiforrn

cambial cells within taproots of Aesculus hippocastanum (Hippocastanaceae). New
Phytologist. 139: 623-635

Chen HH, Li PH (1978) Interactions of Low-Temperature, Water Stress, and Short Days

in Induction of Stem Frost Hardiness in Red Osier Dogwood. Plant Physiology. 62:
833-835

Chen THH, Howe GT, Bradshaw HD (2002) Molecular genetic analysis of dormancy-
related traits in poplars. Weed Science. 50: 232-240

Christersson L (1978) Influence of Photoperiod and Temperature on Development of
Frost Hardiness in Seedlings of Pinus-Silvestris and Picea-Abies. Physiologia
Plantarum. 44: 288-294

Cui X, Churchill GA (2003) Statistical tests for differential expression in cDNA
microarray experiments. Genome Biol. 4: 210

Cutting JGM, Strydom DK, Jacobs G, Bellstedt DU, Vanderrnerwe KJ, Weiler EW
(1991) Changes in Xylem Constituents in Response to Rest-Breaking Agents Applied
to Apple before Budbreak. Journal of the American Society for Horticultural Science.
116: 680-683

Dennis FG (1994) Dormancy - What We Know (and Dont Know). Hortscience. 29:
1249-1255

Eisen MB, Spellman PI‘, Brown PO, Botstein D (1998) Cluster analysis and display of
genome-wide expression patterns. Proc Natl Acad Sci U S A. 95: 14863-8

70

Erez A, Couvillon GA, Kays SJ (1980) Effect of Oxygen Concentration on the Release of
Peach Leaf Buds from Rest. Hortscience. 15: 39-41

Erez A, Faust M, Line MJ (1998) Changes in water status in peach buds on induction,
development and release from dormancy. Scientia Horticulturae. 73: 111-123

Ermel FF, Follet-Gueye ML, Cibert C, Vian B, Morvan C, Catesson AM, Goldberg R
(2000) Differential localization of arabinan and galactan side chains of
rhamnogalacturonan 1 in cambial derivatives. Planta. 210: 732-740

Espinosa-Ruiz A, Saxena S, Schmidt J, Mellerowicz E, Miskolczi P, Bako L, Bhalerao
RP (2004) Differential stage-specific regulation of cyclin-dependent kinases during
cambial dormancy in hybrid aspen. Plant J. 38: 603-615

Farrar JJ, Evert RF (1997) Seasonal changes in the ultrastructure of the vascular cambium
of Robinia pseudoacacia. Trees-Structure and Function. 11: 191-202

Fowler S, Thomashow MF (2002) Arabidopsis transcriptome profiling indicates that
multiple regulatory pathways are activated during cold acclimation in addition to the
CBF cold response pathway. Plant Cell. 14: 1675-90

Greer DH, Stanley CJ, Warrington U (1989) Photoperiod Control of the Initial Phase of
Frost Hardiness Development in Pinus-Radiata. Plant Cell and Environment. 12: 661-
668

He Y, Michaels SD, Amasino RM (2003) Regulation of flowering time by histone
acetylation in Arabidopsis. Science. 302: 1751-4

Hewett EW, Wareing PF (1973) Cytokinins in Populus by Robusta - Changes During
Chilling and Bud Burst. Physiologia Plantarum. 28: 393-399

Hon WC, Griffith M, Mlynarz A, Kwok YC, Yang DS (1995) Antifreeze proteins in
winter rye are similar to pathogenesis-related proteins. Plant Physiol. 109: 879-89

Howe GT, Hackett WP, Fumier GR, Klevom RE (1995) Photoperiodic Responses of a
Northern and Southern Ecotype of Black Cottonwood. Physiologia Plantarum. 93:
695-708

Hu C, Honda C, Kita M, Zhang Z, Tsuda T, Moriguchi T (2002) A Simple Protocol for
RNA isolation from Fruit Trees Containing High Levels of Polysaccharides and
Polyphenol Compounds. Plant Molecular Biology Reporter. 20: 69a-69g

lwai H, Masaoka N, Ishii T, Satoh S (2002) A pectin glucuronyltransferase gene is
essential for intercellular attachment in the plant meristem. Proc Natl Acad Sci U S A.
99: 16319-24

Jiang M, Zhang J (2002) Water stress-induced abscisic acid accumulation triggers the
increased generation of reactive oxygen species and up-regulates the activities of
antioxidant enzymes in maize leaves. J Exp Bot. 53: 2401-10

71

Kim HJ, Kim YK, Park JY, Kim J (2002) Light signalling mediated by phytochrome
plays an important role in cold-induced gene expression through the C-

repeat/dehydration responsive element (C/DRE) in Arabidopsis thaliana. Plant J. 29:
693-704

Lang GA, Early JD, Martin GC, Darnell RL (1987) Endodormancy, Paradormancy, and
Ecodormancy - Physiological Terminology and Classification for Dormancy
Research. Hortscience. 22: 371-377

Little CHA (1981) Effect of Cambial Dormancy State on the Transport of l-C-l4 Indol-
3-Ylacetic Acid in Abies-Balsamea Shoots. Canadian Journal of Botany-Revue
Canadienne De Botanique. 59: 342-348

Little CHA, Bonga JM (1974) Rest in Cambium of Abies-Balsamea. Canadian Journal of
Botany-Revue Canadienne De Botanique. 52: 1723-1730

Miller RT, Christoffels AG, Gopalakrishnan C, Burke J, Ptitsyn AA, Broveak TR, Hide
WA (1999) A comprehensive approach to clustering of expressed human gene

sequence: the sequence tag alignment and consensus knowledge base. Genome Res.
9: 1143-55

Nissila PC, Fuchigami LH (1978) Relationship between Vegetative Maturity and lst
Stage of Cold-Acclimation. Journal of the American Society for Horticultural
Science. 103: 710-711

Olsen JE, Junttila O, Nilsen J, Eriksson ME, Martinussen I, Olsson O, Sandberg G,
Moritz T (1997) Ectopic expression of oat phytochrome A in hybrid aspen changes

critical daylength for growth and prevents cold acclimatization. Plant Journal. 12:
1339-1350

Perry TO (1971) Dormancy of Trees in Winter. Science. 171: 29-&

Riding RT, Little CHA (1984) Anatomy and Histochemistry of Abies-Balsamea Cambial
Zone Cells During the Onset and Breaking of Dormancy. Canadian Journal of
Botany-Revue Canadienne De Botanique. 62: 2570-2579

Rinne P, Saarelainen A, Junttila O (1994) Growth Cessation and Bud Dormancy in
Relation to Aba Level in Seedlings and Coppice Shoots of Betula-Pubescens as
Affected by a Short Photoperiod, Water-Stress and Chilling. Physiologia Plantarum.
90: 451-458

Rinne PLH, Kaikuranta PM, van der Schoot C (2001) The Shoot apical meristem restores

its symplasmic organization during chilling-induced release from dormancy. Plant
.Joumal. 26: 249-264

Rowland LJ, Arora R (1997) Proteins related to endodormancy (rest) in woody
perennials. Plant Science. 126: 119-144

Sagisaka S (1974) Transition of metabolisms in living poplar bark from growing to
wintering stages and vice versa. Plant Physiol. 54: 544-549

72

Samish RM (1954) Dormancy in woody plants. Annual Review of Plant Physiology and
Plant Molecular Biology. 5: 183-203

Schrader J, Baba K, May ST, Palme K, Bennett M, Bhalerao RP, Sandberg G (2003)
Polar auxin transport in the wood-forming tissues of hybrid aspen is under
simultaneous control of developmental and environmental signals. Proceedings of the
National Academy of Sciences of the United States of America. 100: 10096-10101

Schrader J, Moyle R, Bhalerao R, Hertzberg M, Lundeberg J, Nilsson P, Bhalerao RP
(2004) Cambial meristem dormancy in trees involves extensive remodelling of the
transcriptome. Plant Journal. In press

Schrick K, Fujioka S, Takatsuto S, Stierhof YD, Stransky H, Yoshida S, Jurgens G
(2004) A link between sterol biosynthesis, the cell wall, and cellulose in Arabidopsis.
Plant J. 38: 227-43

Thomashow MF (1999) Plant cold acclimation: Freezing tolerance genes and regulatory
mechanisms. Annual Review of Plant Physiology and Plant Molecular Biology. 50:
571-599

Wang E, Miller LD, Ohnmacht GA, Liu ET, Marincola FM (2000) High-fidelity mRNA
amplification for gene profiling. Nat Biotechnol. 18: 457-9

Wang SY, Faust M (1988) Changes of fatty acids and sterols in apple buds during bud
break induced by a plant bioregulator, thidiazuron. Physiologia Plantarum. 72: 115-
120

Wang SY, Faust M (1994) Changes in the Antioxidant System Associated with Budbreak
in Anna Apple (Malus-Domestica Borkh) Buds. Journal of the American Society for
Horticultural Science. 119: 735-741

Wang SY, Jiao HJ, Faust M (1991) Changes in Ascorbate, Glutathione, and Related
Enzyme-Activities During Thidiazuron-Induced Bud Break of Apple. Physiologia
Plantarum. 82: 231-236

Wanner LA, Junttila O (1999) Cold-induced freezing tolerance in Arabidopsis. Plant
Physiol. 120: 391-400

Weiser CJ (1970) Cold resistance and injury in woody plants. Science. 169: 1269-78

Welling A, Moritz T, Palva ET, Junttila O (2002) Independent activation of cold
acclimation by low temperature and short photoperiod in hybrid aspen. Plant
Physiology. 129: 1633-1641

Welling A, Rinne P, Vihera-Aamio A, Kontunen-Soppela S, Heino P, Palva ET (2004)
Photoperiod and temperature differentially regulate the expression of two dehydrin
genes during overwintering of birch (Betula pubescens Ehrh.). Journal of
Experimental Botany. 55: 507-516

73

Yang YH, Dudoit S, Luu P, Lin DM, Peng V, Ngai J, Speed TP (2002) Normalization for
cDNA microarray data: a robust composite method addressing single and multiple
slide systematic variation. Nucleic Acids Res. 30: e15

74

 

 

 

 

Figure 2-1. Microarray experimental design.

For a particular target sample, the pooled material was divided into two batches; total
RNAs were amplified before labeling. These batches were split in two and labeled with
Cy5 or Cy3 dye. The labeled probes were randomly assigned to each of the four
hybridizations, respectively. Arrows represent hybridizations between the target samples.
The sample at the tail of the arrow is labeled with Cy3, and the sample at the head of the
arrow is labeled with Cy5. Four array experiments were performed based on each target
sample. LDW represent long day (16hr) in warm temperature (25 °C) condition; LDD,
long day in warm temperature (25 0C) with drought treatment; LDC, long day in cold
temperature (5 oC); SDW, short day (~9hr) in warm temperature; SDC, short day in cold
temperature.

75

A 46 444 4

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1"“?— ; - 18
30- ‘ + Daylength i i 3 -15
A i “ll — Max. Temp. A
9 20 j I — Min. Temp. i ‘ 14 E
E .. I" “2 8
* it. all
as 0 ll' lllllllll 1.1 ll: 8 2
c. . ; : "11‘ ‘1111 iii : a
E 40 g ' W“ A” ‘ 6 Q
0 5 i i . . . I
F l l l i l I I i‘ 4
'20 "l I i i I I l l i ‘_ 2
.30 i;-- it i :i i .- — 1"‘—i——1_ o
H 0) Qt JJ > U L". .0 H H >1 Ci
:3 :3 G) U 0 (D (U 0) 15 Or 15 :5
'3 ‘1: U) 0 Z Q 1') [u '2: 43 El '3
B 3m _,_ 1 g _1 _ i1 ___ 1 _ 25
A E
E
e E-
“ 2004 C
5 “5’
,E, e
E, +Diame1ter l 10 8
.. —B— ' z:
c 100 A Lengt 2
‘5, or
c 5 E
3 ' .92
D
o - — 444 —— ———-- o

 

Figure 2-2. The temperature, daylength and tree growth change during the
experiment period (i.e., July 2002 to June 2003).

(A) Daily minimum and maximum temperatures are illustrated throughout the experiment
period. Photoperiod is represented by the length of daylight on 15th day of each month at
experimental site in East Lansing, MI. (B) The length or volume growth of primary stems
is an average of 20 independent measurements, and the tree materials used for both the
measurements and the microarray experiments were from the same field. Error bar
represent :tS.E.

76

   

 

 

 

too-it .
”"1
m l
m .
ems .
m .
m f
1% ‘ A
o... i - . , ,.__,L. ,
Whole ClusterA Cluster B ClusterC
I Cell Growth E] Protein Fate
El Defense, Cell Death El Protein Synthesis
I Signal Transduction D Transcription
Cl Cellular Organization I Transport
I Energy I Unclassified
El Metabolism No Hit

 

 

Figure 2-3. Hierarchical clustering (A) and Functional categorization (B) of genes
based on the seasonal variation of bark.

The 714 genes from the bark samples were selected based on ANOVA. In hierarchical
clustering, months are indicated numerically by columns, and genes are indicated by
rows. In functional categorization, all 1,953 of the genes on the array were used for
comparison reference, and referred to as Whole in the graph. The asterisks indicate the
differed distribution of the functional groups between clusters, which was supported by a
contingency test (p <0.001).

77

AVEWMB

        

 

 

 

 

 

5 7810111212 43 100%
90% -
80% t
70% i
60% 4. — .
50% -
40% ~
30% ~-
20% ‘ - 1.
10% l
0% . 4
Whole ClusterA ClusterB
I Cell Growth El Protein Fate
Cl Defense, Cell Death Cl Protein Synthesis
I Signal Transduction Cl Transcription
Cl Cellular Organization
Energy I Unclassified
El Metabolism No Hit

up

Figure 2-4. Hierarchical clustering (A) and Functional categorization (B) of genes
based on seasonal variation of xylem.

The 743 genes were selected based on ANOVA in xylem. Hierarchical clustering and
functional categorization were analyzed as described in Figure. 2.

78

Cell rowth
[ch DNA processing

Defense

  

 

Abiotic stress
Abiotic stress
Abiotic stress
‘ Abiotic stress
Abiotic stress

Senescence

lntercellular
Intracellular
intracellular
lntercellular

Photosynthesis

 

JJASNDJFMAM

ref|NP_563817.1| histone deacetylase family protein (SE-7)

gb|AAM46894.1| early drought induced protein (9E-19)
sp|Q39644|Late embryogenesis abundant protein (3E-12)
gb|AA015460.1| cold-regulated LTCOR12 (SE-32)
gb|AAM65256.1| pollen coat-like protein (2E-8)
sp|P46519| Desiccation protectant protein (SE-52)
gb|AAM23263.1| DnaJ-like protein (BE-29)
gb|AAP03085.1| class IV chitinase (7E-49)
gb|AAD33696.1| PR1a precursor (5E-19)
gb|AAT01418.1 lstress-responsive protein (7E-77)
sp|Q41596|Endochitinase 1 (4E-32)

gb|AAM60932.1| disease resistance protein (1 E-83)
pir||T08154 pathogenesis-related protein PR1 (SE-19)
rel|NP_181620.1| cysteine protease inhibitor (1 E-20)
gb|AAM67134.1 | glutaredoxin-like protein (1 E-25)
releP_192897.2| glutathione peroxidase (8E-80)
gb|AABZ4588.1 | cupredoxin (2E-14)

pir||T10484 blue copper protein stellacyanin (7E-16)
ref|NP_179828.1| peroxidase 17 (PER17) (1 E105)
gb|AAT02523.1| metallothionein 1b (SE-26)
sp|P17598| Catalase isozyme 1 (1 E-144)
emblCAE18065.1| cytochrome P450 (1 E78)
gb|AAT02526.1| metallothionein 3a (SE-11)
gb|AAT02524.1| metallothionein 2a (7E-27)
gb|AAM65608.1|senescence-associated protein 12 (GE-44)

ref|NP_195785.1| macrophage migration inhibitory (SE-48)
ref|NP_175057.2| protein phosphatase 2C (2E-28)
ref|NP_568466.1| CBL-interacting protein kinase (2E-40)
ref|NP_196467.1| transmembrane CLPTM1 family (1 E126)

emb|CAA47812.1| ptxA (SE-20)

pir|lT14329 extracellular dermal glycoprotein (2E-31)
pir||T06482 cell wall protein (2E-26)
ref|NP_563679.1lextracellular dermal glycoprotein (2E-7)
ref|NP_194094.1|reticuion family protein (GE-82)
ref|NP_200309.1| integral membrane family protein (2E-46)
gb|AAFlO7596.1| fiber protein Fb34 (SE-34)
ref|NP_180436.1| cupin family (storage of Nitrogen) (4E-80)

ret|XP_331149.1| NADH-Ubiquinone Oxidoreductase (9E-86)
ref|NP_188923.1| chlorophyll A-B binding family (SE-14)
gblAAL32038.1| early light-induced protein (35-40)
gb|AA033591.1learly light induced protein (6546)
sp|064416|FluBisCO small subunit (1 E-48)

Figure 2-5. Magnified view of of the bark cluster A genes as shown Figure 2-3.
The genes that had a known function were shown. Assigned categories and sub-roles
were shown in the first and second columns, respectively. The column indicates months,

June to March

79

Protein Fate

Transcri tion

 

NAD salvage
Nucleic acid

Degradation
Degradation
Degradation
Modification
Modification

mRNA syn.
mFlNA syn.
mRNA syn.
mRNA syn.
RNA processing

Protein
Proton

Figure 2-5 (cont’d)

pir||T50664 methylenetetrahydrofolate dehydrogenase (1 E-1o7)
gb|AAM96867.1| galactinol synthase (1 E-1 18)
gb|AAM22411.1| cell-wall invertase (2E-79)
emblCA029733.1| pectin methylesterase (415-45)

pir||T16976 pectinesterase (25-99)

dbj|BAD15422.1 |pectin-glucuronyltransferase (SE-41 )
gb|AAP87281 .1 | beta-1 ,3-glucanase (1 E104)

pir|lT50680 beta-1,3 glucanase (1 E-141)

dbj|BAA02729.1| cytoplasmic aldolase (4E-16)

gi|17380672| phosphoglycerate dehydrogenase (0E0)
ref|NP_850759.1 | fructose-bisphosphate aldolase (1 E-124)
gi|22296820| pyruvate kinase (1 E103)

ref|NP_200316.1| GDSL-motif lipase/hydrolase family (6E-74)
gb|AAN87573.1| delta 12 oleic acid desaturase (1 E131)
gb|AAN74808.1|Nicotinate phosphoribosyltransferase (2E-79)
emb|CADS7200.1|guanine deaminase (7E-70)

pir||T07790 transaldolase (1 E102)

ref|NP_198892.1| glucose-S-P dehydrogenase (0E0)
ref|NP_176904.1| rubber elongation factor family (4E-31)
ref|NP_595138.1| oxosterol binding domain (SE-41)
emblCAD87012.1 |progesterone 5-beta-reductase (1 E136)
ref|NP_172904.2| oxygenase (1 E-1 16)

ref|NP_190468.1| AMP-dependent synthetase (BE-73)

ref|NP_850481.1| F-box family protein (1 E-82)
pir||S42882 cysteine proteinase (9E-64)
ref|NP_564919.1| F-box family protein (2E-90)
ref|NP_922744.1 |palmitoyI-protein thioesterase (1 E-86)
ref|NP_194414.1| ubiquitin-like protein (2E-40)

ref|NP_179571.1| zinc finger (2E-75)

gb|AAT37999.1|RNA polymerase sigma factor (SE-97)
ref|NP_194461.1| zinc finger (5E-13)

ref|NP_180326.1| zinc finger (AN1-like) family protein (7E-43)
emb|CAD48198.1|RNA-binding protein (4E-42)

ref|NP_849394.1| mitochondrial membrane translocase (SE-60)
pir||T14564 inorganic H+ diphosphatase (2E-23)
ref|NP_567178.1| plasma membrane intrinsic protein (1 E-129)
pir||J02288 nodulin-26 (SE-39)

ref|NP_565876.2| SOUL heme-binding family protein (2E-31)

80

   
  

Cell growth

  

Defense

Abiotic stress
Defense
Defense
Defense
Detoxification
Detoxification
Detoxification
Senescence

 

Intracellular
Intracellular
Ion Transport
lntercellular

it "1 Signalling
Cellular organization
1 Cell wall
;’ '. Cytoskeleton
Energy

. Cell cycle
DNA processing

ref|NP_191614.1| cyclin family protein (cyclin U) (4E-23)
ref|NP_172363.1| histone H2A, (SE-35)

gb|AAG44839.1| Hs1pro-1-like receptor (6E-58)
ref|NP_850009.1| disease resistance-responsive (2E-35)
gb|AA084216.1l Kunitz trypsin inhibitor 3 (SE-58)
ref|NP_565872.1| protease inhibitor (1 E-30)
gb|AAC18566.1|2.4-D inducible GTT S-transferase (1 E-75)
gb|AAT02526.1| metallothionein 3a (SE-27)
gb|AAT02527.1| metallothionein 3b (2E-6)
ref|NP_199977.1| leaf senescence protein (1 E-21)

dbj|BA882502.1| ciga (SE-25)

ref|NP_565316.1| auxin-responsive protein (3E-40)
emblCAA98167.1| Rab GTPases (2E-8)

ref|NP_190797.1| calmodulin-binding family protein (8E-12)
gb|AAGOl 180.1 | opsin (4E-42)

ref|NP_180875.1| leucine-rich transmembrane kinase (2E-8)
embiCA036200.1|TlR-NBS disease resistance protein (2E-83)
ref|NP_176532.2| LRR transmembrane kinase. (5E-70)
dbj|BAD15416.1| receptor protein kinase PERKl (SE-36)
ref|NP_171714.1| GRAM domain-containing protein (2E-10) ‘
ref|NP_176683.1| transducin family protein (2E-46)

gb|AAT37957.1| fasciclin-like AGP 14 (1E-76)
ref|NP_197478.1 |tubulin alpha-3/-5 chain (1 E-1 19)

Electron transport emblCAA04703.1] cytochome b5 (1E-17)
Electron transport ref|NP_200930.1|complex1ubiquinone oxidoreductase (1E-18)
Electron transport ref|NP_198005.1| plastocyanin-like domain (SE-10)

 

 
 

   

Protein Fate
' ; Degradation
Degradation
Degradation
Degradation
Modification
Protein Sorting

Protemynthesis

Ribosome
Ribosome
tFlNA

ref|NP_192212.2| kelch repeat-containing F-box family (1 E28)
ref|NP_850050.1| serine carboxypeptidase $28 (4E-97)
emblCAC43295.1| vacuolar processing enzyme (1 E-34)
ref|NP_566473.2| subtilase family protein (6E-17)
ref|NP_190104.1| ubiquitin family protein (4E-33)
gb|AAK25949.1| ER lumen protein retaining receptor (2E-60)

gb|AAP80667.1| ribosomal Pr 117 (SE-33)
ref|NP_850375.1| ribosomal protein L1 family protein (2E-12)
ref|NP_175439.2| aminoacyl-tRNA synthetase (SE-42)

Figure 2-6. Magnified view of the genes of the bark cluster B as shown Figure 2-3.

81

Metabolism

 

Transport

  
 

Amino acid
Carbohydrate
Cell wall

Cell wall
Detoxification

Figure 2-6 (cont’d)

ref|NP_197318.1| glutamate dehydrogenase 1 (6E-71)
ref|NP_180697.1| pka-type carbohydrate kinase family (7E—51)
ref|NP_180377.2lalpha-L-fucosidase. (4E-24)

ref|NP_1 9521 2.1 | beta-fructofuranosidase (1 E-102)
ref|NP_563973.1| lactoylglutathione lyase (glyoxalase I) (1 E12)
dbj|BAC83542.1| acyl-CoA dehydrogenase (SE-41)
pir||T07929Iong-chain-fatty-acid-CoA Iigase (2E-86)
ref|NP_194630.1| fatty acid oxidation multifunctional (1 E-23)
ref|NP_014051.1| NADPH alcohol dehydrogenase (2E-38)
sp|P93436|Alcohol dehydrogenase class III (1 E-86)
sp|P51850|Pyruvate decarboxylase isozyme 1 (1 E-120)
ref|NP_172203.1 | phosphoesterase family protein (2E-16)
ref|NP_200577.1| diacylglycerol kinase. (2E-16)
gb)AAP23033.1| sphingolipid delta-8 desaturase (BE-63)
ref|NP_179967.3| 5' nucleotidase family protein (1 E-34)
sp|P9371 1 |Caffeoyl-CoA O-methyltransferase (1 E-124)
gb|AAK58908.1| 4-coumarate:CoA ligase 3 (SE-38)
emblCA865335.1 | ferulate-s-hydroxylase (1 E-1 14)
gb|AAP47635.1| lecithine cholesterol acyltransferase (BE-84)
gb|AAF81279.1 | C-14 sterol reductase (BE-91)

dbj|BAD1 1 159.1 | steroid sulfotransferase (2E-95)

ref|NP_1 71 840.1 |oxoglutaratedependent dioxygenase (1 E-24)
ref|NP_18081 1 .1| dienelactone hydrolase family (BE-54)

ref|NP_564690.1| Kln17 DNA-binding protein (7E-57)
ref|NP_194573.1| DNA-binding protein (5640)
ref|NP_567546.1| DNA-binding family protein (1 E-38)
dbj|BAD19748.1| helix-loop-helix DNA-binding protein (1 E7)
ref|NP_920230.1| transcription initiation factor (1 E-5)
ref|NP_188169.1| no apical meristem (2E-76)
gb|AAM09320.2|Homeobox-containing protein (1 E-20)
ref|NP_196202.1| zinc finger (03HC4-type ) family (SE-42)
ref|NP_200445.1| zinc finger (C3HC4-type) (4E-7)
pir||T01735 homeobox protein NTH15 (2E-79)
ref|NP_565774.1| DNA-binding protein (7E-47)

sp|004011|Auxin-binding protein ABP20 (1 E47)
ref|NP_195616.2| kinesin-related protein (SE-27)
gb|AAK38196.1 | phosphate transporter 1 (1 E-107)
pir||T00435 mitochondrial carrier protein (SE-12)
gb|AA091200.1 | glutathione transporter (1 E-131)

82

.14-

. . IV 3 ’
use.

.‘7 i

wi a"

 

.-: .,- , i '
' fiifiifii
“irfiwiflh‘

-. 50. ,. ,, .,
81 ‘
'3; k

8‘ Ir.
, mums!

' 1’... -
, " ‘ “C

Figure 2-7. The whole plant picture and stem sections of the poplar trees treated

under controlled environments.

The tree materials used were the same as for the microarray experiments. Colored
pictures of stem sections (second and third rows) were taken on microscopy. Black and
white pictures at the fourth row were taken on confocal microscopy. Scale bar represents
50 um. Long daylength and warm (LDW); Long daylength and drought (LDD); Long
daylength and cold (LDC); Short daylength and warm (SDW); Short daylength and cold

(SDC)

83

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2-8. Dendrogram showing relationships between seasonal changes and
controlled environmental treatments in bark (A) and xylem tissue (B).

Months were indicated numerically. For dendrogram, sets of genes were selected based
on P < 0.001, 0.05, 0.05 for bark or P < 0.0001, 0.005, 0.005 (for xylem) of
MAANOVA at both seasonal changes and controlled treatments.

84

 

 

 

T _ a ‘ __.__ ,-______.__.
‘ .LDD lsow
BLDG Dsoc

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2- 9. K-means clusters based on expression profiles of both annual cycle and
controlled environments.

A total of 272 genes were individually clustered into eight groups. Each group is
represented by a pattern of seasonal variation (line graphs, left) and environment
responsiveness (bar graphs, right) for genes in the group. The number of genes in each
group is indicated at the bottom of the bar graph. For the seasonal variation graph, the x
axis represents twelve months (i. e., Jan to Dec) and the y axis represent the normalized
value ranging between 4 and —4 in log 2 space as described in method section. The thick
black line indicates the means value. In the environmental responsiveness graph, the
means fold changes ranging between 3 and —3 for each of the four treatments (i.e., LDD,
LDC, SDW, and SDC) over LDW were shown as the y axis in log 2 scale.

85

 

JFMAMJJASOND

*3“ \ _ .fi. m *1..- M‘. 'QgI‘i
, 1 '

 

 

G6
1

“I“ "‘

 

't. u hm: “ can

 

 

0030
on
0391mm

G6PDH
GPX
«Iv—Q LTCOR12
n . ~ 258

 

 

 

E

 

l3

 

G7

E

WUI‘

1.... h-zi' i. :r u... in» LJ '* J

 

 

 

 

 

  
 
 
 

a Lipase

Q EDG
B-glucanase
‘ ’ 258

Figure 2-10. Validation of microarray results by Northern blot analysis.

RNAs were prepared from the bark tissues of trees grown under the indicated
environment conditions or in out-field. 258 rRN A was used as a loading control. G6PDH,
glucose 6-phosphate dehydrogenase; GPX, glutathione peroxidase; LTCORIS, Cold-
regulated COR15; Lipase, GDSL-motif lipase; EDG, Extracellular dermal glycoprotein;

B-glucanase, endo- 1,3-B-glucanase.

86

Table 2-1. Eight groups of genes identified by K-means clustering as shown in
Figure 2-9.

 

 

CloneID Description (BlastX E-val) Category Sub-role
GROUP 1
L2P10d10 dbj|BAD00018.1|histone l (6E-19) CELL GROWTH DNA processing
B9SP10b05 gb|AAK01359.l|dehydration stress-induced CELL RESCUE, Abiotic stress
protein (BE-57) DEFENSE
B9P02g09 gb|AA858469.1|ultraviolet-B-repressible protein CELL RESCUE, Abiotic stress
(IE-28) DEFENSE
B9P07f 10 gb|AAR09168.l|alpha-expansin l (IE-112) CELLULAR Cell wall
ORGANIZATION
B9SP09a02 gb|AAF78903.1|proiine-rich protein (2E-34) CELLULAR Cell wall
ORGANIZATION
X9P08b06 emb|CAD58680. l Imicrotubuie associated protein CELLULAR Cytoskeleton
(8E-44) ORGANIZATION
L1P08fll gb|AAP3219l.lIalpha-tubulin (IE-126) CELLULAR Cytoskeleton
ORGANIZATION
X9P09d02 emb|CAB66336. 1|alpha-tubulin (IE-107) CELLULAR Cytoskeleton
ORGANIZATION
B9P07d01 gb|AAP03873.l|photosystem I reaction center ENERGY Photosynthesis
(3E-44) '
B9SP08d05 reflNP_l76940. l [photosystem 11 core ENERGY Photosynthesis
complex(5E-25)
X9SP07c12 reflNP_192772.l|chlorophyll A-B binding ENERGY Photosynthesis
protein (9E-78)
LlPlOclO pir||SO4125 chlorophyll a/b-binding protein type ENERGY Photosynthesis
III (IE-36)
X9P07e05 sp|P34105 IN ADP Malic enzyme (IE-158) METABOLISM Carbohydrate

B9SP09a05 reflNP_19823 6. 1 INAD-epimerase/dehydratase METABOLISM Carbohydrate
family(lE-103)

X98 P06f0 l gb|AAF 80591 .1 |xyioglucan METABOLISM Cell wall
endotransglycosylase (IE-22)

X9P10c09 ref|NP_973975.1|lipase class 3 family protein METABOLISM Lipid

(3E-55)

L3P02b09 sp|023787|Thiazole biosynthetic enzyme (lE- METABOLISM
1 l9)

X9SP04f03 gb|AAM51555.l|metacaspase 1(protease) (2E- PROTEIN FATE Degradation
50)

B9P03e01 gb|AAB97006.l|Gibbereilin regulated SIGNAL Hormone
protein(4E-23) TRANSDUCTION

X9SP05e05 dbj|BAA33810.1|phi-1(phosphate induced) ( IE- SIGNAL Intracellular
130) TRANSDUCTION

X9P05f04 reflNP_195130. l|zinc finger (C3HC4-type) (3E- SIGNAL Intracellular
56) TRANSDUCTION

B9SP06h06 pir||T07086 acid phosphatase (8E-9l) SIGNAL Intracellular

TRANSDUCTION

B9P10a12 gb|AAM29182.l|Aux/IAA protein (2E-64) TRANSCRIPTION mRNA synthesis
L3P04b02 emb|CAC84706.l|aux/IAA protein (IE-127) TRANSCRIPTION mRNA synthesis
B9P03fll ref|NP_l78396. l|zinc finger (C3HC4-WP€)(9E- TRANSCRIPTION mRNA synthesis

38)
X9P05d09 pir||T10807 annexin 2 (8E-87) TRANSPORT
X9P05g06 emb|CAC82712.1|major intrinsic protein 1(1E- TRANSPORT

l 15) I
PIN gi|10441744|pttPINl (OE—0) TRANSPORT Hormone

87

B9SP07h06
X9SP01h07
X9SP04f06

B9P06bl 1

GROUP 2

L4P02008
X9P10a05
B9P03e05
L1 P09a01
L2P04e07
L3 P09g1 l
B9P09g06

L2P09h10

X9SP06a05

L 1 P06e05

B9SP04h05

L4P07h02

X9P03f10

L1P02a09

L2P08h01

L3 P09e05

X9P08a10

B9SP07eO3

L3P05b05

L3P05b03

X9SP03eO6

L1P07b05
X9P02g09

X9SP05g09

B9SP07g09

X9P01h07

Table 2-1 (cont’d)

pir||T48886 aquaporin (lE-l 15)
reflNP_921277. 1|Transposase of Tn 10 (0E0)
reflNP_l99968.l|phosphate-responsive 1 family
( l E-l 12)

gb|AAM28295. l |PVR3-iike protein (6E-20)

TRANSPORT
UNCLASSIFIED
UNCLASSIFIED

UNCLASSIFIED

sp|P09189|dnaK-type molecular chaperone hsp70 CELL RESCUE,

(113-133)
emb|CAC35772.3|dhnl (413.42)

gb|AAD01605. l|copper/zinc-superoxide
dismutase (6E-83)
sp|O65768|Superoxide dismutase (2E-12)

emb|CAC33845. l|cytosoiic CuZn-superoxide
dismutase (SE-85)

ref|NP_194825. l |CBL-interacting protein kinase
(4E-67)

reflNP_1 8801 5. 1 lhydroxyproline-rich
glycoprotein (4E-71)
gb|AAG61120.1|RuBisCO activase 1 (lE-l l9)
sp|P32869|Photosystem I reaction center subunit
II (7E-88)

pir||JC5876 early light-inducible protein (2E-l l)
reflNP.192705. l |oxidoreductase family protein
(3E-58)

sp|Q9LW96|Inositoi-3-phosphate synthase (1E-
124)

sp|O78328| l -deoxy-D-xyiulose S-phosphate
synthase (IE-140)

reflNP_566550. l |UDP-glucoronosyl transferase
(213-56)

gb|AAM6473 l . l |nitriiase associated protein
(2E- 10)

gb|AAM64577. 1 |nitriiase-associated protein
(ZE-l l)

gb|AAM95643. l |WD-repeat anthocyanin
biosynthesis (lE-13l)

emb|CAA73220. 1|isoflavone reductase (IE-108)
gb|AAS93803. l |progesterone S-beta-reductase

( l E-78)

emb|CAC80137. l |progesterone 5-beta-reductase
(8E-98)

gb|AAM65243. 1 |carboxyl-terminal peptidase

( l E- 129)

DEFENSE

CELL RESCUE,
DEFENSE

CELL RESCUE,
DEFENSE

CELL RESCUE.
DEFENSE

CELL RESCUE.
DEFENSE
SIGNAL
TRANSDUCTION
CELLULAR
ORGANIZATION
ENERGY
ENERGY

ENERGY
METABOLISM

METABOLISM
METABOLISM
METABOLISM
METABOLISM
METABOLISM
METABOLISM

METABOLISM
METABOLISM

METABOLISM

PROTEIN FATE

reflNP_l79037.2|proline iminopeptidase (SE-91) PROTEIN FATE

gb|AA063777. l Icyclophilin (2E-83)
sp|O49884|ribosomal protein L30 (IE-53)

sp|P19951|4OS RIBOSOMAL PROTEIN 514
(213.49)

PROTEIN FATE
PROTEIN
SYNTHESIS
PROTEIN
SYNTHESIS

gb|AAR01683.l|ribosomal protein Ll3a (9E-78) PROTEIN

88

SYNTHESIS

Water

Abiotic stress
Abiotic stress
Detoxification
Detoxification
Detoxification
Intracellular
Cell wall

Photosynthesis
Photosynthesis

Photosynthesis

Carbohydrate
Carbohydrate
Cell wall
Nitrogen
Nitrogen
Secondary

Secondary
Sterol

Sterol
Degradation
Degradation
Folding
Ribosome
Ribosome

Ribosome

X9SP05e04
B9P04b03
B9P05d07
X9P10gl 1
X9P02h02
X9SP06g05
L4PO7f05
B9P02g05

X9P04d05
L4P07f09

B9P05b07

GROUP 3
X9P02e09

B9P07a09
X9SP06e04
B9SP05b06
B9SP05b04
B9P01b02
X9P10d02
L2P10bl 1
X9SP03b06
X9Pl 1e02
X9P07f07
X9P03hl l
X9SP01b03
X9P02cl 1
B9SP08h07

B9SP06b10
B9P03a03

L4P07g09

Table 2-1 (cont’d)

gb|AAQ08196.1|eukaryotic translation initiation
factor (3E-84)

sp|O48650|TRANSLATION FACTOR SUIl
(SE-50)

gb|AAQ08l91 . l leukaryotic translation initiation
factor (7E-86)

gb|AAC61786. 1 Iglycine-rich RNA-binding
protein (2E-38)

dbj|BAB84582.l|transcription factor LIM (IE-
1 17)

reflNP_567050.l|remorin family protein (6E—23) TRANSCRIPTION

gb|AAQ84334. l Izinc-finger protein (2E-6 l)
sp|Q43019|Nonspecific lipid-transfer protein
(IE-29)

emb|CAE53882. l |aquaporin (l‘E- 149)
sp|Q9ZRXO|Translationally controlled tumor
protein (SE-9)

sp|P3568 l |Translationally controlled tumor
protein (SE—65)

reflNP_l93537.2|dehydration-responsive protein
( 1E-87)
gb|AAQ56599. l Ichitinase-like protein (9E-5 1)

gb|AAL7 1857. l Idehydroascorbate reductase
(2E-96)
gb|AAT02522.l|metallothionein la (4E-27)

gb|AAQ74889. l |AI-induced protein (lE-l 13)
pir||JQ2252 peroxidase (4E-87)

gb|AAK84479. l lauxin growth promotor protein
( lE-70)

dbj|BAD19898.1|AB13-interacting protein 2
(2E-4 l )

emb|CAA72330. l|shaggy-like kinase (BE-32)

sp|Q9FVIl|Actin-depolymerizing factor 2 (1E-
66)
gb|AAP73462.1|actin (SE-68)

reflNP_566065. l Ireticulon family protein (7E-
86)

reflNP_l96738. l |NADH-ubiquinone
oxidoreductase (SE-83)

emb|CAC86995.1|ATP citrate lyase a-subunit

( lE- 141)

pdbl l QGNICystathionine Gamma-Synthase (1B-
1 2 l )

gb|AAF26735. l |methionine synthase (lE- 107)
ref|NP__ 193 129. l Iglycine/serine
hydroxymethyitransferase (SE-66)
gb|AAB61597.1|glutamine synthetase (lE- 100)

89

PROTEIN
SYNTHESIS
PROTEIN
SYNTHESIS
PROTEIN
SYNTHESIS
TRANSCRIPTION

TRANSCRIPTION

TRANSCRIPTION
TRANSPORT

TRANSPORT
UNCLASSIFIED

UNCLASSIFIED

CELL RESCUE.
DEFENSE

CELL RESCUE,
DEFENSE

CELL RESCUE.
DEFENSE

CELL RESCUE,
DEFENSE

CELL RESCUE,
DEFENSE

CELL RESCUE.
DEFENSE
SIGNAL
TRANSDUCTION
SIGNAL
TRANSDUCTION
SIGNAL
TRANSDUCTION
CELLULAR
ORGANIZATION
CELLULAR
ORGANIZATION
CELLULAR
ORGANIZATION
ENERGY

METABOLISM
METABOLISM

METABOLISM
METABOLISM

METABOLISM

Translation
Translation
Translation
mRNA
processing
mRNA synthesis
mRNA synthesis
mRNA synthesis
Lipid

Water

Abiotic stress
Defense
Detoxification
Detoxification
Detoxification
Detoxification
Hormone
Hormone
Intracellular
Cytoskeleton
Cytoskeleton
ER

Electron Transfer
Acetyi-CoA
Amino acid

Amino acid
Amino acid

Amino acid

Table 2-1 (cont’d)

L3P10f05 reflNP_ 1 78224. l|fructose-bisphosphate METABOLISM Carbohydrate
aldolase/Plastid (3E-91)

LlP10c05 sp|P12858|Glyceraldehyde 3-p METABOLISM Carbohydrate
dehydrogenase/Plastid ( lE- 103)

X9P05f02 sp|P21820|Triosephosphate isomerase (IE-89) METABOLISM Carbohydrate

B9P01a06 dbj|BAB85987.l|Acy1-CoA-binding protein METABOLISM fatty acid
(6E-32)

L1P09e05 sp|Q43784|Granule-bound starch synthase (lE- METABOLISM Starch
l 18)

X9P04a01 emb|CAA54308.l|l,4-alpha-giucan branching METABOLISM Starch
enzyme ( 1E-157)

B9P03h05 emb|CAA54308.1|1,4-alpha—glucan branching METABOLISM Starch
enzyme (6E-47)

X9P07f06 gb|AAF63239.1|alpha-amylase (IE-66) METABOLISM Starch

B9SP05a09 reflNP_175460.l|dehydrocholesterol reductase METABOLISM Sterol
(IE-74)

X9SP09h03 gb|AAR12194.l|molecular chaperone Hsp90—2 PROTEIN FATE Folding
(2E-47)

B9SP06c05 gb|AAM78180.1|polyubiquitin (6E-55) PROTEIN FATE Modification

X9SP04g10 dbj|BAA85750.1|polyubiquitin (7E-7) PROTEIN FATE Modification

X9Pl 1f02 emb|CAA54603.1|pentameric polyubiquitin (lE- PROTEIN FATE Modification
1 l l)

L2P08g09 gb|AAF34767.1|60S acidic ribosomal protein PROTEIN Ribosome
(IE-128) SYNTHESIS

B9SP07d03 sp|P17093|ribosomal protein 811 (2E-69) PROTEIN Ribosome

SYNTHESIS

X9P06a06 dbj|BAD07825.l|60S ribosomal protein L9 (4E- PROTEIN Ribosome
64) SYNTHESIS

X9P02g02 gb|AAP80650.l|eiongation factor (3E-31) PROTEIN Translation

SYNTHESIS

L1P10g02 dbj|BAA34348. l |elongation factor-l alpha (lE- PROTEIN Translation
162) SYNTHESIS

X9Pl 1h08 sp|O49169|ELONGATlON FACTOR l-ALPHA PROTEIN Translation
(IE-69) SYNTHESIS

B9P05d10 gb|AAD56020.l|elongation factor-l alpha 3 PROTEIN Translation
(4E-75) SYNTHESIS

L2P10c05 reflNP_056176.2|R3H domain (2E-l3) TRANSCRIPTION mRNA synthesis

X9SP06gll reflNP_l9l705.l|zinc finger (C3HC4-type RING TRANSCRIPTION mRNA synthesis
finger) (2E-1 l)

B9SP03f05 pir||T14332 homeotic protein (6E-52) TRANSCRIPTION mRNA synthesis

B9SP03g02 pir||T14563 inorganic diphosphatase ( 1E-178) TRANSPORT Proton

B9P04h05 pir||S54l72 inorganic diphosphatase (2E-9l) TRANSPORT Proton

X9SP03c01 reflNP_178443.1|nodulin-related (IE-25) TRANSPORT

L4P02f09 dbj|BAB09103.1|endosomal protein-like (lE- TRANSPORT
174)

GROUP 4

X9SP07d02 ref|NP_567575. l |dehydration stress protein CELL RESCUE, Abiotic stress
(ERD3) (IE-133) DEFENSE

X9P07d04 gb|AAC24588. l |iate embryogenesis-like protein CELL RESCUE, Abiotic stress
(6E-68) DEFENSE

X9SP04a12 dbj|BAC78212.1|thaumatin/PR5-like protein CELL RESCUE, Defense
( lE-83) DEFENSE

X9SP09c10 gb|AAD25355. l|pathogenesis-related protein CELL RESCUE, Defense
(1E-64) DEFENSE

90

B9P07b10
139901105
B9P04g09
B9P07f03
B9SP10c06
B9SP09d08
B9SPO8c05
X9P10c03

B9SP04d10
B9SP04bO8

B9SP06f12
B9SP06dl l

X9SP01bO7
X9SP09b09
B9SP01bO6
B9SP05g08

B9SP10cl l
B9SP05e02

X9P08c05
X9P01f08

X9P01c10
X9P09f01
X9P08c12

X9P07a12
B9SP04c06

B9SP10d05
X9SP05h08
B9SP07f06
B9P05a10

LlPlOdOS
X9SP05gO6

Table 2-1 (cont’d)

gb|AAF65767. 1|giutathione S-transferase (2E-
79)
gb|AAK52084. l |peroxidase (2E-96)

pir||T10790 peroxidase (IE-137)
reflNP_l77592. l |ABA-responsive protein (2E-9)

gb|AAM76226.1|germin E protein precursor
(2E-78)
gb|AAT37955.l|fasciclin-like AGP 12 (IE—89)

pir||T09546 extensin (3E-27)

sp|P479l6|S-adenosyimethionine synthetase
(IE-134)

dbj|BAA89049. 1 Isucrose synthase (0E0)
gb|AAS45400. 1 lendo- l ,4-beta-glucanase ( 1E-
108)

gb|AAK66161. l|pectate lyase (3E-14)
reflNP_564059.1|glycosyl hydrolase family
protein 17 (SE-20)

gb|AAD39086. l |xyioglucan endo-
transglycosylase (IE-102)

pir||T09870 endo-xyioglucan transferase (3E-92)
emb|CAE76634. l |pectin methylesterase (3E-76)
ref|NP_197666.1|glycosyl transferase family 2

( IE- 15)

pir||T10054 pyruvate kinase/Plastid (6E-75)
dbj|BAD09193.1|PrMC3 Esterases and lipases
(4E-58)

gb|AAG30576. l |nitrate reductase (2E-64)
gb|AAN5 2279. 1 |phenylalanine ammonia-lyase
(IE-107)

gb|AAQ74878. l |phenylaianine ammonia lyase
(SE-80)

gb|AAN5 2280. 1 |phenylaianine ammonia-lyase
(IE-93)

sp|P45730|Phenylalanine ammonia-lyase (1E-

l 14)

gb|AAF60951 . 1 |O-methyltransferase (0E0)
dbj|BAA21643. l |phenylaianine ammonia-lyase
(lE- 105)

gb|AAP31058. l|flavonoid 3',5'-hydroxy1ase (1E-
109)

gb|AAC49929. llflavanone 3beta-hydroxyiase
(IE-104)

sp|O65333|Chalcone--flavonone isomerase (1E-
88)

reflNP_l78637.2|glycosyl hydrolase family (2E-
12)

pir||S34285 polyubiquitin (IE-129)

gb|AAN03468.1|bZIP transcription factor ATBZ TRANSCRIPTION mRN A synthesis

(25-44)

91

CELL RESCUE,
DEFENSE
CELL RESCUE.
DEFENSE
CELL RESCUE.
DEFENSE
CELL RESCUE.
DEFENSE
CELLULAR

ORGANIZATION

CELLULAR

ORGANIZATION

CELLULAR

ORGANIZATION

METABOLISM

METABOLISM
METABOLISM

METABOLISM
METABOLISM

METABOLISM
METABOLISM
METABOLISM
METABOLISM

METABOLISM
METABOLISM

METABOLISM
METABOLISM

METABOLISM

METABOLISM

METABOLISM

METABOLISM
METABOLISM

METABOLISM

METABOLISM

METABOLISM

METABOLISM

PROTEIN FATE

Detoxification
Detoxification
Detoxification
Hormone
Cell wall
Cell wall
Cell wall
Amino acid

Carbohydrate
Cell wall

Cell wall
Cell wall ‘

Cell wall
Cell wall
Cell wall
Cell wall

Glycolysis
Lipid

Nitrogen

Secondary
Secondary

Secondary
Secondary

Secondary
Secondary

Secondary
Secondary

Secondary

Modification

B9P10g03
X9SP07dOl
X9P07d05

B9SP06c02
X9SP02a1 1

GROUP 5
L3P09e07

B9P05d04
B9P07f06
B9SP09d05

X9SP02f06
L2P10a02

L4P02609
B9SP07aO3

X9SPO3C 10
B9P08e05
L2P09c03

GROUP 6
L3P03e02

X9P08e03
B9SP02a02
L 1 P07c04
L2P08dl l
L4P05f06
L2P09aO6
L3P03d02
L1P10g09
L1P10f07

L l P06608

Table 2-1 (cont’d)

reflNP_568932.2|YT521-B-like family (6E-15)
gb|AAF28385.”lipid-transfer protein (3E-13)
ref|NP_200019. 1|BAG domain-containing

protein (6E-21)

reflNP_5667 12. l |protease inhibitor/storage/lipid

transfer (4E-24)

reflNP_l79803.3|C2 domain-containing protein

(2E-95)

gb|AAL6799l . l |dehydration-induced protein

RD22 (9E-29)

gb|AAL26909. 1|dehydration—responsive protein

RD22 (IE-20)

reflNP_l95785. l |macrophage migration

inhibitory family (SE-48)

emb|CAD66637.1|phytocyanin protein, PUP2

(4E- 1 9)

sp|O64416|RuBisCO small subunit (lE-48)
sp|P26301 |phosphoglycerate
dehydratase/cytosolic ( lE- 125)
gi|22296820|pyruvate kinase/cytosolic (0E0)
gi|17380672|phosphoglycerate dehydrogenase

(0E0)

ref|NP_568818. 1|eukaryotic translation initiation

factor (2E-34)

gb|AAM77753. l |translation initiation factor B04

(IE-41)

TRANSCRIPTION RNA processing

TRANSPORT
UNCLASSIFIED

UNCLASSIFIED

UNCLASSIFIED

CELL RESCUE.
DEFENSE
CELL RESCUE.
DEFENSE
SIGNAL

TRANSDUCTION

ENERGY

ENERGY
METABOLISM

METABOLISM
METABOLISM

PROTEIN
SYNTHESIS
PROTEIN
SYNTHESIS

dbj|BAD04010.l|tonoplast intrinsic protein (ZE- TRANSPORT

29)

gb|AAC15460. l|coid-regu1ated LTCOR12 (5E-

32)

gb|AAM23263. l |DnaJ ~1ike protein (SE-29)

gb|AAM46894.1|early drought induced protein

(9E-19)

sp|Q39644|Late embryogenesis abundant protein

(3E-12)

gb|AAD33696. l |PR1a precursor (5E- 19)

pir||T08154 pathogenesis-related protein PR1

(SE 1 9)

gb|AAT01418. l|stress-responsive protein (7E-

77)

sp|Q41596|Endochitinase 1 (4E-32)

ref|NP_l79828.1|peroxidase l7 (PER17) (1E-

105)

pir||T10484 blue copper protein steilacyanin (7E-

16)

gb|AABZ4588. l |cupredoxin (2E- 14)

92

CELL RESCUE,
DEFENSE
CELL RESCUE.
DEFENSE
CELL RESCUE,
DEFENSE
CELL RESCUE,
DEFENSE
CELL RESCUE.
DEFENSE
CELL RESCUE.
DEFENSE
CELL RESCUE.
DEFENSE
CELL RESCUE,
DEFENSE
CELL RESCUE,
DEFENSE
CELL RESCUE,
DEFENSE
CELL RESCUE.
DEFENSE

Lipid

Abiotic stress
Abiotic stress
lntercellular
Electron transport

Photosynthesis
Glycolysis

Glycolysis ‘
Glycolysis

Translation

Translation

Abiotic stress
Abiotic stress
Abiotic stress
Abiotic stress
Defense
Defense
Defense
Defense
Detoxification
Detoxification

Detoxification

Table 2-1 (cont’d)

B9SP03b11 reflNP_192897.2|g1utathione peroxidase (SE-80) CELL RESCUE, Detoxification

DEFENSE
L2P08c05 sp|P17598|Catalase isozyme 1 (IE-144) CELL RESCUE, Detoxification
DEFENSE
L1P01h03 reflNP_l75057.2|protein phosphatase 2C (2E—28) SIGNAL Intracellular
TRAN SDUCTION
L1P07a08 pir||T06482 probable cell wall protein (2E-26) CELLULAR Cell wall
ORGANIZATION
L2P03f02 reflNP_l94094.l|reticulon family protein (6E- CELLULAR ER
82) ORGANIZATION
L3P03f10 gb|AAO33591. l|ear1y light induced protein (6E- ENERGY Photosynthesis
46)
L2P04h07 gb|AAL32038.l|early light-induced protein (3E- ENERGY Photosynthesis
40)
L1P02f10 ref|NP_188923.1|chlorophyll A-B binding family ENERGY Photosynthesis
(6E- 14)
L3P04g06 pir||T16976 pectinesterase (2E-99) METABOLISM Cell wall
L1P04cll emb|CAD29733.l|pectin methylesterase (4E-45) METABOLISM Cell wall
L4P04f05 pir||T50680 beta-1,3 glucanase (IE-141) METABOLISM Cell wall
L3P03e07 reflNP_850759. l |fructose-bisphosphate METABOLISM Glycolysis

aldolase/cytosolic (IE-124)
L1P05d01 gb|AAN87573.l|delta 12 oleic acid desaturase METABOLISM Lipid

( [E l 3 l)

L1P04b06 ref|NP_198892. l|glucose-6-P dehydrogenase METABOLISM PPP
(0E0)

X9P08e06 reflNP_l76904.l|rubber elongation factor family METABOLISM Rubber syn.
(4E-31)

L2P10b07 emb|CAD87012.1|progesterone 5-beta—reductase METABOLISM Sterol
(IE-136)

L4P01 g09 gb|AAN74808. 1 [Nicotinate METABOLISM
phosphoribosyltransferase (2E-79)

L2P01b01 pir||S42882 cysteine proteinase (9E-64) PROTEIN FATE Degradation

L1P06c05 reflNP_194414.llubiquitin-like protein (2E-40) PROTEIN FATE Modification

L3P08d02 ref|NP_194461.l|zinc finger (SE-l3) TRANSCRIPTION mRNA synthesis

L1P06302 emb|CAD48i98.1|RNA-binding protein (4E42) TRANSCRIPTION RNA processing
X9Pl lb09 gb|AAM66l l4.l|nifU-like protein (213-53) TRANSPORT Nitrogen
L3P08e06 pirIIJ Q2288 nodulin-26 (3E-39) TRANSPORT
B9P05c12 ref|NP_849394.l|mitochondria1 membrane TRANSPORT
translocase (3E-60)
LlP08e03 reflNP_567l78. l |piasma membrane intrinsic TRANSPORT
protein (lE—129)
L3P03d05 sp|Q07469|BARK STORAGE PROTEIN (3E- UNCLASSIFIED Storage
1 l)
L2P10e07 reflNP_568160. l |protease inhibitor/storage/lipid UNCLASSIFIED
transfer(7E-38)

L1P10g05 gb|AASSS470.l|iittle protein 1 (IE-16) UNCLASSIFIED

L4P07a02 ref|NP_194476.2|CBS domain-containing protein UNCLASSIFIED
(9E-39)

GROUP 7

L4P04f03 gb|AAM60932.1|disease resistance protein (lE- CELL RESCUE, Defense
83) DEFENSE

L1P01e05 gb|AAP03085.1|c1ass 1V chitinase (7E-49) CELL RESCUE, Defense

DEFENSE

93

L1P03d10
X9P03c07
L 1 P04d01
L2P05h10
L4P04c02
B9P04b01
B9P04f06
L1P06d04
L1P04e02
L1P09h10
GROUP 8
L2P10h02
L2P01a05
Ll P06a05
X9SP07d12
L1P06f10
X9P06g05

L2P05fl 1
X9P05b12

Table 2-1 (cont’d)

pir||T14329 extracellular dermal glycoprotein
(2E-3 l)
gb|AAR07596.l|fiber protein Fb34 (3E-34)

ref|NP_200316. l IGDSL-motif lipase/hydrolase
family (6E-74)

gb|AAM96867.1|galactinol synthase (IE-l l8)
gb|AAP87281.l|beta-l,3-glucanase (IE-104)
dbj|BAA02729.l|cytoplasmic aldolase (4E-16)
emb|CAD37200. l|guanine deaminase (7E-70)
reflNP_l90468. l|AMP-binding synthetase (CoA
ligase) (SE-73)

releP_850481.l|F-box family protein (IE-82)
reflNP_l79571.1|zinc finger (2E-75)

gb|AAT02527. ”metallothionein 3b (2E-6)
gb|AAT02526. llmetallothionein 3a (3E-27)

dbj|BA882502. l |cytokinine inducible gene3
(3E-25)

sp|P51850|Pyruvate decarboxylase isozyme 1

( l E- l 20)

reflNP_l72203.1|phosphoesterase family protein
(2E— 16)

releP_566473.2|subtilase family protein (6E-l7) PROTEIN FATE

ref|NP_188169.1|no apical meristem (2E-76)

CELLULAR

ORGANIZATION

CELLULAR

ORGANIZATION

METABOLISM

METABOLISM
METABOLISM
METABOLISM
METABOLISM
METABOLISM

PROTEIN FATE
TRANSCRIPTION

CELL RESCUE,
DEFENSE
CELL RESCUE.
DEFENSE
SIGNAL

TRANSDUCTION

METABOLISM

METABOLISM

Cell wall

Lipid
Carbohydrate
Cell wall

Glycolysis
Nucleic acid

Degradation
mRN A synthesis
Detoxification
Detoxification
Hormone .
fermentation
Lipid

Degradation

TRANSCRIPTION mRNA synthesis
gblAAQ91200.1[glutathione transporter (IE-131) TRANSPORT

 

94

CHAPTER 3

An auxin-repressed gene (RpARP) from black locust (Robinia pseudoacacia) is
posttranscriptionally regulated and negatively associated with shoot elongation.
Abstract

The plant hormone auxin regulates various plant growth and developmental processes by
controlling the expression of the auxin-response genes. While many genes up-regulated
by auxin have been characterized, less is known about the genes that are down regulated
by auxin. Here, we isolated and characterized an auxin-repressed gene (RpARP) from the
tree legume, Robinia pseudoacacia L. A sequence similarity search in public databases
showed that RpARP gene has homologues in various higher plants including monocots
and dicots. The deduced amino acid sequences are highly conserved among them (up to
85% identity). Northern blot analysis showed that auxin can repress RpARP gene
expression and the repression was dependent on metabolizable sugar and protein
synthesis. In addition, cold treatment abolished the auxin-mediated repression of RpARP
gene expression. Results from the transgenic plant analyses suggest that the expression of
RpARP gene is post-transcriptionally regulated by auxin and the untranslated regions
(UTRs) are required for the regulation. Sequence analysis of the promoter region (-70bp
and -500bp upstream of putative transcription initiation site) of RpARP gene identified
four sucrose-repressible-response elements (TATCCAT—motifs, Huang et al. 1990),
suggesting that the cis-elements responsible for the regulation by sucrose are located in

the promoter region. In fact, the expression of transgenic RpARP gene was not affected

95

by sucrose when driven by CaMV35 promoter. We also present evidence that RpARP

gene expression is negatively associated with hypocotyl elongation.

96

Introduction

Auxin, a plant growth regulator produced mainly in apical buds, plays a key role in a
wide variety of growth and developmental processes such as lateral root formation, apical
dominance, tropism, and differentiation of vascular tissue (Berleth et al. 2000, Hamann
2001, Muday 2001). There have been two major experimental approaches to understand
the molecular basis of auxin action in such a variety of developmental processes. One is
to identify and analyze mutants that lack normal auxin responses. Two related families of
proteins, the AUX/1AA proteins and auxin response factors (ARFs), were identified as
key regulators of auxin-modulated gene expression (Ulmasov et al. 1997, Rouse et a1.
1998). Those protein families function as transcriptional regulators and are thought to
activate or repress the auxin-induced gene expression through interaction between them.
Recently, the protein degradation by ubiquitin-mediated process was found to be
important in auxin response (Ruegger et al. 1998). The other approach involves the use of
molecular biology tools to identify and characterize the genes regulated by auxin signals.
This molecular approach has led to the identification of several classes of early auxin
response genes and auxin responsive cis-elements within the promoters of these genes
(Guilfoyle et al.1998). The expression level of the early (or primary) auxin response
genes increases within minutes of auxin application, independent of de novo protein
synthesis (Abel and Theologis 1996, Walker and Estelle 1998). Although many mutants
and cDNA clones that are associated with auxin signaling have been isolated and
characterized, our knowledge of the mechanisms by which auxin regulates various

biological functions is still limited. This lack of information may be partially due to the

97

fact that most of the molecular studies were focused on the ‘primary response genes’ that
are up-regulated by auxin signal (Hagen and Guilfoyle 2002).

Auxin—represssed protein (ARP) genes and their role for plant growth and
development are relatively understudied. So far, three orthologs of ARP were isolated
and characterized from pea, strawberry, and tobacco. SAR5 was first isolated by
differential screening from auxin-deprived strawberry receptacles (Reddy and Poovaiah
1990). They showed that SAR5 gene was repressed by auxin and the repression of the
gene was positively correlated with strawberry fruit maturation. Another ortholog
PsDRMl has been reported as bud dormancy-related genes from pea. The abundance of
PsDRMl transcripts in axillary buds declines quickly, within six hours of the removal of
the apical buds. Then, they re-accumulate when the axillary buds become latent again and
the newly established apical bud becomes active (Stafstrom et al. 1998). Therefore, the
gene was named “dormancy-related protein (DRP).” However, it is not known how the
gene is regulated at the molecular level, or how the two different plant developmental
processes are related with ARP expression. Steiner et al. (2003) reported a new member
of auxin-repressed protein family from tobacco, whose transcripts were high during
pollen maturation but rapidly declined to minimal levels in germinating pollen.

Recently, we identified an ARP ortholog (named RpARP) from the expressed
sequence tags (ESTS) derived from the trunk wood of a 10-year old black locust (R.
pseudoacacia) (Yang et al., 2003). The gene was expressed across the trunk wood
(bark/cambium, sapwood, and sapwood-heartwood transition zone). The deduced amino
acid sequence of RpARP shows high sequence homology with other ARP orthologs.

Repression of RpARP gene by auxin was dependent on sucrose and post-transcriptionally

98

regulated. The expression of this auxin-repressed gene is negatively associated with shoot

elongation.

Materials and Methods

Plant materials

Black locust seeds were germinated as previously described (Han et al. 1993). Seven-day
old hypocotyls (~2.5 cm long) from the seedlings were used. Arabidopsis thaliana (L.)
Heynh. plants, used in this study, were grown under sterile conditions, the seeds of
ecotype Columbia were surface sterilized with 1.5 % (v/v) sodium hypochlorite solution
containing 0.02 % (v/v) triton-XIOO for 5 min with vigorous shaking, and then washed
several times with sterile water. Seeds were chilled at 4°C for 2 to 4 days in water and
then were plated onto MS medium (Murashige and Skoog 1962) containing 0.8% agar
and 1% (w/v) sucrose (pH 5.7-5.8). The plants were grown at 25°C under a 16h/day

photoperiod.

Isolation dim-length RpARP gene and its promoter region

The RpARP gene (Genbank AY009094) was isolated through the black locust EST
sequencing project (Yang et al., 2003). Total RNAs were purified from both the cambial
and heartwood region of 10-year old black locust trees and used to construct directional
cDNA libraries in the pTripleEx plasmid vector (Clontech, Palo Alto, CA). The resulting
libraries were applied to EST sequencing. Using a computational contig search,
approximately 15 RpARP clones were detected out of the 4000 sequenced clones.

Presumed whole transcript of RpARP was determined through multiple alignments. The

99

RpARP gene promoter was cloned using Genome Walker Kit (Clontech, Palo Alto, CA)
following a manufacturer’s instruction with 2.5 pg of genomic DNA isolated from a

black locust seedling.

Construction of binary vectors and plant transformation

The 700bp fragment of RpARP cDNA including 5’UTR and 3’UTR was prepared by
restriction digestion with SmaI and SacI from pTripleARP. The fragment was inserted
into the corresponding sites of the binary vector pB1121 under the control of CaMV 358
promoter to produce the resulting recombinant vector, pBIARP (Figure l-7A). For the
construction of pBIARPGUS, a PCR reaction with primers containing linker sites, Xbal
(5’) and SmaI (3’) was performed on pTripleARP to yield 400bp of RpARP coding
region. The resulting PCR product was cloned to pGEM-T (Promega, Madison, WI) to
produce plasmid pGEMARP (Figure l-7A). The SmaI-Xbal fragment of pGEMARP was
cloned to the corresponding sites of pBllOl (Genbank U12639.1) to yield pBIARPGUS.
Arabidopsis ecotype Columbia was transformed using recombinant A. tumefaciens by the
vacuum infiltration method (Bechhold et al. 1998). Transformants were selected on MS
supplemented with 0.7 % (w/v) agar medium containing 1 % (w/v) sucrose and 50 mg/L

kanamycin.

Elongation measurement
Black locust seedlings that had uniform hypocotyl lengths (~ 25 mm) were selected and
used for elongation measurements. Hypocotyls were cut into S-mm segments and

immersed in MS+2 % (w/v) sucrose medium (Figure 3-9) or in liquid MS medium in

100

absence or presence of 2 % sucrose, 10 M NAA, 20 1.1M Fusicoccin, 2 % mannose, or
low temperature (4°C) (Figure 3-5). Incubation was carried out at 25°C with a l6h/day
photoperiod on a shaker at 30 cycles per min. Elongation ratios were determined by

measuring the increase in segment length after 24 hours of incubation.

Hormone and chemical treatment

Hypocotyl segments (5 mm in length) were cut from below the apical meristem of 7-day
old seedlings and incubated in liquid MS medium supplemented with indicated hormones
or chemicals with gentle agitation on a shaker at 30 cycles per min at 25°C. All hormones
and chemicals used were purchased from Sigma-Aldrich. Working and stock
concentrations of hormones and chemicals were used as follows unless otherwise noted:
10 11M Fusicoccin (FC, diluted from 10mM stock solution in DMSO); 50 11M
cycloheximide (CHX, from 200 mM in DMSO); 20 11M abscisic acid (ABA, from 10mM
in EtOH); 20 11M 2-chloroethyl phosphonic acid (EP, from 100mM in EtOH); 20 M
gibberellic acid (GA, from 50mM in H20); 20 11M jasmonic acid (JA, from 200mM in
EtOH); 10 11M (it-Naphthalene acetic acid (NAA, from lmM in H20). After the
treatments, the tissues were frozen in liquid nitrogen and stored at -80°C until use for

RNA analysis.

Northern blot analysis
For northern-blot analysis, total RNA was isolated from black locust hypocotyl segments
or from whole tissues of transgenic Arabidopsis plant using Trizol reagent (Invitrogen,

San Diego). Six micrograms of total RNA were separated on formaldehyde-agarose gels,

101

transferred to Hybond-N membranes. Hybridization of northern blots was performed
using UltraHyb solution (Amersham-Pharmacia, Piscataway, NJ) at 42°C according to the
manufacturer’s instruction. RpARP or GUS DNA labeled with 32P-dCTP in a random
primed reaction was used as probe. DNAs used for the probes were obtained by PCR
using gene—specific primers. Hybridization signals were detected by autoradiography. For
quantification of northern signal, the autoradiography films were scanned and analyzed

using ImageQuant program (Amersham-Pharmacia)

Results

The ARP gene family is higher plant-specific.

The BLAST search against NCBI dbEST and The Institute For Genomic Research
(TIGR) Plant Gene Indices (httpzllwww.tigr.org/tdb/tgi) identified homologous sequences
of RpARP, all of which are from higher plant species (four monocots and eight dicots).
No significant hits were obtained from other organisms such as animal and fungal species
using a 1.0E-5 E-value cut-off value. Multiple alignments of the deduced amino acid
sequences of the orthologs revealed the presence of highly conserved amino acid domains
at both N-terminal and C-terminal (Figure 3-1A). The deduced amino acid sequence of
RpARP is highly conserved among the plant genes (up to 85% identity at amino acid
level). In order to examine the evolutionary relatedness of the 12 putative ARP proteins,
we generated an unrooted phylogenic tree with the deduced amino acid sequences (Figure
3-1B). The tree showed two distinct clades of plant ARPs that are separated into dicot
and monocot. The two clades are consistent with the evolutionary distance of monocot

and dicot (Soltis et al. 1999), suggesting that the ARP gene was attained before monocot

102

and dicot diverged from the last common ancestor. Furthermore, sequences from all of
the legume species (black locust, pea, and soybean) were branched together, indicating
that RpARP (from a legume tree) is structurally closer to that of other legume species
(pea and soybean) than to that of a tree species poplar. The fact that ARP genes are
present only in higher plants and down-regulated by auxin suggests that the gene
products play a role in plant growth and development processes specific to higher plants.
While the RpARP was repressed rapidly by exogenous auxin as in the previous studies
with pea and strawberry ARP genes, it is not known whether these genes are functionally

conserved.

S_upDression oprARP gene expression by auxin.

The expression of the RpARP gene in response to exogenous auxin was analyzed using
northern blot analysis (Figure 3-2). Total RNA was extracted from hypocotyl fragments
of black locust seedlings after treatment with 10 M NAA or pre-treatment with 10 11M
NAA for 24 hours, followed by incubation in auxin-deprived MS medium. The transcript
level was found to decline within 3 hours of the application of exogenous auxin, reaching
level undetectable by northern blot after 6 hours. However, the RpARP transcripts re-
accumulated rapidly when the tissues were transferred to auxin-free MS media (Figure 3-
2A). This auxin-mediated repression of the gene was effectively achieved using as low as
0.1 M NAA (data not shown). To determine if the auxin-mediated reduction of RpARP
transcripts was dependent on de novo protein synthesis, we carried out the same
experiment with a protein synthesis inhibitor, cycloheximide. The cycloheximide

treatment retarded the reduction of RpARP transcripts by auxin and also blocked the re-

103

accumulation of the transcripts in the auxin-free medium (Figure 3-2B), indicating that
the protein synthesis was required for the suppression of the gene expression by auxin.
Our next question was whether other hormones or abiotic stress signaling was involved in
the regulation of the ARP gene expression. To answer that question, we examined the
effects of various growth regulators such as ABA, BA, ethylene and GA, and abiotic
stresses such as salt stress and cold treatment with or without exogenous auxin (Figure 3-
3). Cold treatment seemed to increase RpARP gene expression with or without auxin
treatment, while salt slightly depressed the repression by auxin. On the other hand, GA
treatment resulted in loss of the transcript even in the absence of auxin. In the presence of
sucrose, auxin effectively repressed RpARP gene expression regardless of the treatments

with the exception of cold stress.

RpARP gene is woressed in response to fitsicoccin andjacidic pH.

According to the acid growth hypothesis (Kutschera 1994), auxin begins by stimulating
proton pumping, which results in a hyperpolarization of the plasma membrane, thereby
inducing cell elongation through acid-dependent cell wall loosening (Senn and Goldsmith
I988, Rayle and Cleland 1992). To determine the relationship between the physiological
action of auxin and RpARP gene suppression, we treated the samples with fusicoccin
(PC), which is known to induce rapid cell elongation through the activation of the
transport protein H+ATPase in the plasma membrane (Blatt 1988), causing strong
acidification of the cell walls (Marre 1979, Kutschera and Schopfer 1985). Figure 3-4
shows that RpARP gene expression was suppressed by both fusicoccin and acidic pH

treatments. These results suggest that auxin may suppress the_RpARP gene through the

104

acidification of the cells. Interestingly, under FC treatment, RpARP was suppressed in
one hour. Furthermore, FC and low pH treatments appear to be less effective in
depressing the RpARP transcript than is auxin. The RpARP transcript was not detectable
after 2 hours of auxin treatment (Figure 3-2), while it is clearly detectable even after 6
hours of treatments with PC or low pH (Figure 3-4). In addition, the transcript level was
increasing again after 9 hours of FC treatment, suggesting that PC effect might be

transient.

Effect of sucrose and low temperature on auxin suppression of RpARP mRNA expression.
Our next question was whether the factors associated with the auxin-mediated growth can
also affect the expression of RpARP gene. We examined the effects of sucrose starvation
and cold treatment (4°C). Sucrose has been known to affect the auxin-mediated
elongation and growth (Stevenson and Cleland 1981, Seyedin et al. 1982, Gray et a1.
1998) and low temperature has known as a general inhibitor of auxin transport and
metabolism (George et al. 1967). In fact, sucrose-deprivation and low temperature
conditions inhibit the auxin-modulated elongation of hypocotyl segments (Figure 3-5A).
Sucrose was required for auxin-mediated RpARP gene suppression. In addition, sucrose
deprivation increased the RpARP expression and under low temperature treatment, the
expression of RpARP gene was not affected by exogenous auxin (Figure 3-6A).

In order to gain some insights on the physiological role of sucrose in auxin-
mediated suppression of the gene, we tested the effects of sugar analogs, mannose and
mannitol on the expression of RpARP gene (Figure 3-6B). Mannitol is known for its

impermeability of plasma membrane and mannose as a non-metabolizable sugar by most

105

plant systems (Gibson 2000). We analyzed the time course expression (0.5, l, 3, 6, 9, 24
h) of the gene in MS medium containing 2% sucrose, 2% mannose, or 2% mannitol.
Mannitol, the impermeable form of sugar, had no effect on the expression of the gene. On
the other hand, mannose, permeable but non-metabolizable sugar in most plant systems
(Gibson 2000), had an intermediate effect between those of control and sucrose. These
results indicate that sucrose is needed for both intracellular osmotic regulation and
metabolic energy for the regulation of RpARP gene expression by auxin. This is
consistent with previous findings. Auxin can induce an initial growth (Senn and
Goldsmith 1988, Rayle and Cleland 1992), but after 4 to 6 hrs the initial growth rate
begins to decrease without supply of exogenous sucrose, which is needed for osmotic
regulator and new cell wall compounds (Cheung and Cleland 1991, Inouhe and

Yamamoto 1991).

Regulation of RpARP expression through mRNA stability

In order to investigate the regulation mechanism and biological function of RpARP, we
generated transgenic Arabidopsis plants harboring T-DNA containing a full-length
RpARP cDNA, with or without the untranslated regions (UTRs), driven by the CaMV
358 promoter (Figure 3-7A). The transgenic plants showed a very similar phenotype to
that of the wild type. The expression of RpARP (with UTRs) was repressed in response
to exogenous auxin in pBIARP transgenic plants, even though the gene is under the
control of constitutive CaMV promoter (Figure 3-7B). On the other hand, the level of
RpARP transcipts was not affected by exogeous auxin in the transgenic plants harboring

the 35S::RpARPorf:GUS construct (pBIARPGUS, Figure 3-7A) which contains only 350

106

bp of RpARP coding region but no UTRs. These results suggest that auxin suppression of
the recombinant RpARP transcript is through post-transcriptional regulation and that the
elements responsible for the auxin suppression are located on the untranslated regions of
RpARP. However, the expression of the transgenic RpARP was not affected by sucrose
in transgenic plants transformed with either pBIARP or pBIARPGUS constructs. This
suggests that the cis-element(s) for sucrose signaling is likely to be located in the
promoter region. We sequenced the 2 kb genomic region upstream of the RpARP coding
region and searched for putative cis-elements. Four sucrose repressible response element
(Huang et al. 1990) were found between 70bp and 500bp upstream from the putative

transcription start site (Figure 3-8).

RpARP gene expression is ngativelv associated with hypocotyls elongation

Previous studies indicated that the expression level of ARP genes were low in actively
growing tissue (Reddy and Poovaiah 1990, Stafstrom et al. 1998). We examined the
expression levels of RpARP gene in the segments of elongating hypocotyls that have
different elongation rates. Hypocotyls (25 mm in length) were cut into 5-mm segments
and incubated on MS media for 24 hours. Then, percent elongation and transcript levels
of RpARP were measured at different regions. The hypocotyl segments showing higher
percent elongation had a lower level of RpARP expression and vice versa (Figure 3-9).
To further confirm that the expression of RpARP gene was negatively associated with
hypocotyl elongation, we examined the effects of various treatments that affect shoot
elongation. In general, relatively lower levels of the gene expression were observed in the

treatments that increased shoot elongation, with exception of SU+FC+ and Mn+N+

107

treatments (Figure 3-5). This may be due to the transient nature of FC and Mn effects on
the repression of the gene. Those results suggest that RpARP gene products might play a

role in the cellular processes that are characteristic to non-actively growing tissues.

Discussion

It has long been appreciated that the plant growth regulator, auxin, is an important
regulator of such developmental processes in higher plants as tropism, apical dominance,
and rhizoid initiation. Auxin can regulate the developmental processes through
modification of the expression patterns of the genes involved in the processes (Theologis
1989, Worley et al. 2000). This auxin-mediated regulation mechanism is often highly
conserved. For instance, despite the differences in the chemical composition of cell walls,
the physiological response of all seed plants to auxin and acid is very similar and the key
protein mediating the rapid elongation process, expansin, is highly conserved structurally
and functionally among seed plants (Shcherban et al. 1995, Cooke et al. 2002). The ARP
gene represents another such example. Here, we isolated one ARP ortholog (RpARP)
from a tree legume, black locust (Robinia pseudoacacia). The fact that the RpARP gene
was also repressed rapidly by exogenous auxin, is consistent with previous studies,
supporting the idea that ARP genes might have common patterns of expression and
hormonal regulation. A sequence homology search with dbESTs indicates that ARP gene
is present in many plant species and its amino acid sequences are highly conserved. The
high level of structural conservation between monocotyledonous and dicotyledonous
plants suggests that l) the ARP gene family might have formed before the evolutionary

divergence of monocots and dicots, 2) the ARP protein has strict functional constraints

108

that limit structural modification while maintaining function, and 3) the function of the
protein is important to normal development or physiology of the higher plants.
Furthermore, it is notable that ARP homologous sequences are found only in higher
plants, suggesting its role in higher plant-specific processes.

The RpARP gene is regulated by sucrose as well as auxin. However, these two
signals have different regulation mechanisms. For example, the transcript level of
RpARP was effectively reduced by exogenous auxin in transgenic plants, transformed
with 35S::RpARP construct that contains RpARP structural gene with UTRs, but not in
transgenics carrying 35S::RpARPorf:GUS construct that has the coding region but no
UTRs. On the other hand, sucrose had no effect on either construct. These results suggest
that auxin regulates the ARP gene posttanscriptioally and the responsible cis-elements are
likely to be located in the UTR regions. A number of mRNA instability sequences have
been identified in UTR regions. For example, AU-rich elements (ARES) known as the
archetypal mRNA instability determinant were found in 3’ UTR of such mammalian
genes as lymphokine and cytokine (Chen and Shyu 1995). Down stream (DST) element
which is responsible for mRNA instability in plant, was found in the 3’ UTR of the small
auxin-up RNA (SAUR) transcripts (Gil and Green 1996). However, despite the fact that
the UTRs are involved in auxin-mediated regulation of RpARP gene, no known
regulatory element was found in the UTRs of the gene. Therefore, the auxin-mediated
regulation of RpARP’s transcript stability might represent a novel mechanism for post-
transcriptional regulation of gene expression. Fusicoccin, known as H+-ATPase activator,
also repress RpARP gene expression (Figure 3-4). Whether H+-ATPase activation is

involved in this auxin-mediated repression of RpARP is not known.

109

Cis-elements responsible for sucrose signal might be located at genomic regions
upstream of RpARP gene. Four copies of conserved TATCCA cis-elements were present
in the promoter region of RpARP. This sucrose repressible response element was
reported to be responsible for sugar repression of alpha-amylase gene in rice (Lu et al.
1998, Toyofuku et al. 1998). Lu et al. (1998) identified nuclear proteins binding to the
TATCCA in a sequence-specific and sugar dependent manner. Sucrose is needed both
osmotically and metabolically for auxin- mediated RpARP repression. We tested various
sugar analogs for their impacts on RpARP expression. Mannitol (non-permerable through
plasma membrane) mimicked the effect of the no sugar control treatment, while mannose
(permeable but non—metabolizable sugar; Gibson 2000) showed an intermediate effect
between the no sugar control and sucrose treatment. Those results indicate that sucrose is
needed for both intracellular osmotic regulation and metabolic energy. Sugars function as
important regulators of various processes associated with plant growth and development
as well as metabolic resources and structural constituents of cells. In recent studies, sugar
was found to favor the expression of growth-related genes while repressing the
expression of the stress-related gene (Ho et al. 2001). Both auxin and sucrose are required
to completely repress the expression of RpARP gene. This is consistent with the auxin-
induced elongation process. According to prevailing acid growth hypothesis, auxin can
induce an initial growth through acidification (Senn and Goldsmith 1988, Rayle and
Cleland 1992), but after 4 to 6 h the rate begins to decrease without supply of exogenous
sucrose in which sucrose is needed for osmotic regulator and new cell wall compound

(Cheung and Cleland 1991, Inouhe and Yamamoto 1991). These results led us to a

110

hypothesis that RpARP gene repression is associated with the elongation process. In
other words, RpARP gene needs to be down-regulated for auxin-mediated elongation.

Here, we present two pieces of evidence in support of the hypothesis. First,
RpARP mRNA expression was lower at upper positions of hypocotyl which has more
enhanced elongation ability than at lower positions. Second, the physiological conditions
stimulating elongation can repress the RpARP mRNA expression effectively. Indeed,
RpARP gene is expressed at various developmental stages (e. g. from seedling to mature
wood) or in various tissue (i.e. leaves, root, hypocotyl and stem), and most abundantly in
non-growing tissues. It is quite feasible that RpARP gene needs be repressed for cell
growth probably to loose cell wall, modification of cell wall, or increase elasticity of
cytoskeleton. This seems to corroborate with previous results with two ARP genes that
were isolated and characterized from pea (psDRMl; Stafstrom et al. 1998) and
strawberry (SAR5; Reddy and Poovaiah 1990). The repression of SAR5 was associated
with strawberry fruit maturation, while the expression level of PsDRMl gene was high in
dormant axillary buds. These two different plant developmental processes are related
with cell elongation or expansion process regulated by auxin.

When considering all the data we gathered, RpARP gene is likely to play a role in
biological processes that are characteristic for non-growing (i.e., dormant) conditions or
stress conditions such as cold and sucrose-starvation. Subcellular localization of the
protein, and its interaction with other proteins, and analysis of transgenic plants over-
expressing or suppressing the gene will help understand the biological function of the

gene. We are in the process of generating transgenic plants for such studies.

lll

References

Abel S,Theologis A (1996) Early genes and auxin action. Plant Physiol. 111: 9-17

Berleth T, Mattsson J,Hardtke CS (2000) Vascular continuity and auxin signals. Trends
Plant Sci. 5: 387-393

Blatt MR (1988) Mechanisms of Fusicoccin Action - a Dominant Role for Secondary
Transport in a Higher-Plant Cell. Planta. 174: 187-200

Castle LA, Errampalli D, Atherton TL, Franzmann LH, Yoon ES,Meinke DW (1993)
Genetic and Molecular Characterization of Embryonic Mutants Identified Following
Seed Transformation in Arabidopsis. Mol Gen Genet. 241: 504-514

Chen CY,Shyu AB (1995) AU-rich elements: characterization and importance in mRNA
degradation. Trends Biochem Sci. 20: 465-70

Cheung SP,Cleland RE (1991) Galactose Inhibits Auxin-Induced Growth of Avena
Coleoptiles by 2 Mechanisms. Plant Cell Physiol. 32: 1015-1019

Cooke TJ, Poli D, Sztein AE,Cohen JD (2002) Evolutionary patterns in auxin action.
Plant Mol Biol. 49: 319-38

Gibson SI (2000) Plant sugar-response pathways. Part of a complex regulatory web. Plant
Physiol. 124: 1532-1539

Gil P,Green PJ (1996) Multiple regions of the Arabidopsis SAUR-ACl gene control
transcript abundance: the 3' untranslated region functions as an mRNA instability
determinant. Embo J. 15: 1678-86

Gray WM, Ostin A, Sandberg G, Romano CP,Estelle M (1998) High temperature
promotes auxin-mediated hypocotyl elongation in Arabidopsis. P Natl Acad Sci USA.
95: 7197-7202

Grunwald C, Deutsch F, Eckstein D,Fladung M (2000) Wood formation in rolC
transgenic aspen trees. Trees-Struct Funct. 14: 297-304

Guilfoyle T, Hagen G, Ulmasov T,Murfett J (1998) How does auxin turn on genes? Plant
Physiol. 118: 341-347

Hagen G,Guilfoyle T (2002) Auxin-responsive gene expression: genes, promoters and
regulatory factors. Plant Mol Biol. 49: 373-385

Hamann T (2001) The role of auxin in apical-basal pattern formation during Arabidopsis
embryogenesis. J Plant Growth Regul. 20: 292-299

Han KH, Keathley DE, Davis JM,Gordon MP (1993) Regeneration of a Transgenic
Woody Legume (Robinia-Pseudoacacia L, Black Locust) and Morphological
Alterations Induced by Agrobacterium-Rhizogenes-Mediated Transformation. Plant
Sci. 88: 149-157

112

Ho S, Chao Y, Tong W,Yu S (2001) Sugar Coordinately and Differentially Regulates
Growth- and Stress-Related Gene Expression via a Complex Signal Transduction
Network and Multiple Control Mechanisms. Plant Physiol. 125: 877-90

Huang N, Sutliff TD, Litts JC,Rodriguez RL (1990) Classification and characterization of
the rice alpha-amylase multigene family. Plant Mol Biol. 14: 655-68

Inouhe M,Yamamoto R (1991) Effects of 2-Deoxygalactose on Auxin-Induced Growth
and Levels of Udp-Sugars in Higher-Plants. Plant Cell Physiol. 32: 433-438

Kotake T, Nakagawa N, Takeda K,Sakurai N (2000) Auxin-induced elongation growth
and expressions of cell wall-bound exo- and endo-beta-glucanases in barley
coleoptiles. Plant Cell Physiol. 41: 1272-1278

Kurata T,Yamamoto KT (1998) petitl, a conditional growth mutant of Arabidopsis
defective in sucrose-dependent elongation growth. Plant Physiol. 118: 793-801

Kutschera U (1994) The Current Status of the Acid-Growth Hypothesis. New Phytol.
126: 549-569

Kutschera U,Schopfer P (1985) Evidence for the Acid-Growth Theory of Fusicoccin
Action. Planta. 163: 494-499

Liscum E,Reed JW (2002) Genetics of Aux/1AA and ARF action in plant growth and
development. Plant Mol Biol. 49: 387-400

Lu CA, Lim EK,Yu SM (1998) Sugar response sequence in the promoter of a rice alpha-
amylase gene serves as a transcriptional enhancer. J Biol Chem. 273: 10120-31

Marre E (1979) Fusicoccin - Tool in Plant Physiology. Annu Rev Plant Phys. 30: 273-
288

Muday GK (2001) Auxins and tropisms. J Plant Growth Regul. 20: 226-243

Muday GK,DeLong A (2001) Polar auxin transport: controlling where and how much.
Trends Plant Sci. 6: 535-542

Reddy ASN,Poovaiah BW (1990) Molecular-Cloning and Sequencing of a Cdna for an
Auxin-Repressed Messenger-Rna - Correlation between Fruit-Growth and Repression
of the Auxin-Regulated Gene. Plant Mol Biol. 14: 127-136

Rouse D, Mackay P, Stimberg P, Estelle M,Leyser O (1998) Changes in auxin response
from mutations in an AUX/1AA gene. Science. 279: 1371-1373

Ruegger M, Dewey B, Gray WM, Hobbie L, Turner J,Estelle M (1998) The TIRl protein
of Arabidopsis functions in auxin response and is related to human SKP2 and yeast
Grrlp. Gene Dev. 12: 198-207

Senn AP,Goldsmith MHM (1988) Regulation of Electrogenic Proton Pumping by Auxin
and Fusicoccin as Related to the Growth of Avena Coleoptiles. Plant Physiol. 88:
131-138

113

Seyedin N, Burris JS, Lamotte CE,Anderson IC (1982) Temperature-Dependent
Inhibition of Hypocotyl Elongation in Some Soybean Cultivars .1. Localization of
Ethylene Evolution and Role of Cotyledons. Plant Cell Physiol. 23: 427-431

Shcherban TY, Shi J, Durachko DM, Guiltinan MJ, McQueen-Mason SJ, Shieh
M,Cosgrove DJ (1995) Molecular cloning and sequence analysis of expansins--a

highly conserved, multigene family of proteins that mediate cell wall extension in
plants. Proc Natl Acad Sci U S A. 92: 9245-9

Soltis PS, Soltis DE,Chase MW (1999) Angiosperm phylogeny inferred from multiple
genes as a tool for comparative biology. Nature. 402: 402-4

Stafstrom JP, Ripley BD, Devitt ML,Drake B (1998) Dormancy-associated gene
expression in pea axillary buds. Planta. 205: 547-552

Stevenson T'T,Cleland RE (1981) Osmoregulation in the Avena-Coleoptile in Relation to
Auxin and Growth. Plant Physiol. 67: 749-753

Takahashi Y, Ishida S,Nagata T (1995) Auxin-Regulated Genes. Plant Cell Physiol. 36:
383-390 '

Theologis A (1989) Auxin-regulated gene expression in plants. Biotechnology. 12: 229-
43

Toyofuku K, Umemura T,Yamaguchi J (1998) Promoter elements required for sugar-
repression of the RAmy3D gene for alpha-amylase in rice. FEBS Lett. 428: 275-80

Uggla C, Mellerowicz EJ,Sundberg B (1998) Indole-3-acetic acid controls cambial
growth in Scots pine by positional signaling. Plant Physiol. 117: 113-121

Uggla C, Moritz T, Sandberg G,Sundberg B (1996) Auxin as a positional signal in
pattern formation in plants. P Natl Acad Sci USA. 93: 9282-9286

Ulmasov T, Hagen G,Guilfoyle TJ (1997) ARE 1, a transcription factor that binds to auxin
response elements. Science. 276: 1865-1868

Walker L,Estelle M (1998) Molecular mechanisms of auxin action. Curr Opin Plant Biol.
1: 434-439

Walker L,Estelle M (1998) Molecular mechanisms of auxin action. Curr Opin Plant Biol.
1: 434-439

Worley CK, Zenser N, Ramos J, Rouse D, Leyser O, Theologis A,Callis J (2000)
Degradation of Aux/1AA proteins is essential for normal auxin signalling. Plant J. 21:
553-62

114

   

2 QM>HmmZQOimeh 020.8Mmm memmm>mm3>2.MB¢mH .......
Z QM>PmHZmOimMQh>m¢UH MH<QZmUmmm>mm3>2mm4m>m 11111

(H ........ .Bkmh>mhm
mUH.mH<AZmUmBm>mmB>Z.M m8 ....... Bmkhmh
m88k<988mmmmmmh>mhk
mmB ....... mmflmh>mhk
mmfifi ..... mmmmmh>mhfi
mUmEm>mm3>2.>Nm<mm ...... mkhmh>nhk

              

a99m_92mm0muz>mmm0mo<oan
em emz---mam4mmowxxmmemnx
<><4>B ...... x---mmmeme>ex
em><¢mzaamm=z>>xnmmemeHam
ma.aemo<--mx>ueoa<ax.--.z
seam m09--mx><45n<meavox=
zeae mam--oxH<me<<e ..... z
zeae mom--ax=<mo¢<a ..... z

1111111111 HHMmmmOmQMAZAmOm

  

mHBmHmb 11111 mEqum----4UQQ¥q-HH Axmfifl
afihm mAZIImABMOdem-iitwomQHBXJH AxmAMwaflmnOm04>>QDSQMmfld>2
mHHm mmziimqizqdwm0----mwmmxbiZHmAMMAMUQUHHQEQU¢>>QDSAMHflabz

%Mnumdz dummmomofl HDD3AMQAQ>Z

 

ae.ammaflmao

    

 

H

HHHHHHHHHHH

E3385
__aaoa
cozoo

«£88.82
E80282
unis.

ooE

$9.;
From
coogom
5:03 .m

3.332%
anion.
cotoo

«£88.82
E8832,“

5:03 .m

Econiabw
anion.
cozoo

«£88.82
53323.2
cums.

3.:

use—=5

>o=am
coon>om
$300.. .m

115

B 3'"me Cotton Wheat

Barley
Pea
Nee
Soybean
B.locust "3'29

Poplar

Arabidopsis 1

Arabidopsis 2

Figure 3-1. Comparison and evolutionary tree of the predicted amino acid sequence
of RpARP with other plant ARPs.

(A) Sequences of Arabidopsis] (At1g28330)a, ArabidopsisZ (At2g38330)a, Barley
(TC26321)b, Cotton (TC9854)b, Maize (TC l38763)b, Rice (TC82951)b, Soybean
(TC132184)b, Wheat (TC40040)b, Pea (AF029242)C, and Strawberry (L44142)C, Poplar
(BIl36544)C are compared. Reverse type indicate amino acid identities, and shaded gray
indicates similar residues. The sequence alignment and sequence identity/similarity was
displayed using the Clustal W and BOXSHADE programs of the General Computer
Group Package. (B) An unrooted phylogenic tree of the putative ARP proteins shown in
(A). The unrooted tree was generated using CTREE program based on the Neighbor-
Joining algorithm at A National Laboratory for Compuational Science and Engineering at
the University of California San Diego (http://biowb.sdsc.edu). aAGI Number, bTIGR
Number, CGenbank Number.

116

 

NAA+ NAA-
13 6 912 24252730333648h
- m g

 

NAA+CHX+ NAA-CHX+
13691224252730333648h

 

Figure 3-2. Analysis of RpARP mRNA expression in response to auxin and
cycloheximide.

(A) The hypocotyl segments of black locust seedling were treated with MS + sucrose
(2%,w/v) medium containing 10 Li M NAA (NAA+) for indicated times, or with the
NAA free MS-sucrose medium (NAA-) after 24 h pre-treatment with 10 u M NAA. (B)
The hypocotyl segments were treated with the same condition as (A) except for adding 40
U M cycloheximide (CHX). Total RNA was extracted in indicated times of each blot
and loaded 60 g of total RNA per lane. The blots were probed with a random-primer-
labeled RpARP cDNA. Ethidium bromide stained ribosomal RNA is shown in the bottom
of each panel as a loading control.

117

ABA BA Cold EP GA MeJA Salt Con
NAA+-+-+-+-+-+-+-+-

 

 

Figure 3-3. Comparison of RpARP transcript levels under various hormonal and
abiotic conditions.

Total RNA was extracted from black locust hypocotyl segments immersed for 24 h in
liquid MS + sucrose (w/v, 2%) medium containing indicated hormones, abscisic acid
(ABA), benzyladenine (BA), ethephon (EP), gibberellin (GA), and methyl jasmonate
(MeJA) (each of 20 u M) or indicated abiotic conditions, 40C (Cold) and 300 mM NaCl
(Salt) in the absence or presence of 10 u M NAA as indicated (+/-). As control, the
samples were incubated in only MS + sucrose medium (Con). 60 g RNA was loaded per

lane and hybridized with RpARP cDNA. Ethidium bromide stained ribosomal RNA is
shown in the bottom as a loading control.

118

 

FC a
pH3-5 i ! 35:-1‘
C m H uni-r M

Figure 3-4. Fusicoccin and acidic pH can repress RpARP mRNA expression.

Excised hypocotyl segments of black locust seedlings (7 day old) were treated with MS +
2% sucrose (w/v) medium containing 10 Li M Fusicoccin (FC), buffered with 5 mM MES
(pH 3.5), or with no fusicoccin (C) as a control. Total RNA (6 u g) was isolated after
treatment for the indicated times (0.5, 1, 3, 6, 9, 12, 24 h), and then subjected to RNA
blot hybridization with probes specific for RpARP. The 258 rRNA bands of the
corresponding ethidum bromide-stained gels are shown in the right.

119

100

 

%Elongation
'8 8 8 8 8 5‘ 8 8
Hit
was.
ESE—1

 

 

 

1

Relative amount of RpARP mRNA.
(A)

 

 

o
'9‘“ 9’28" s“ at“ s§§§° as“
Figure 3-5. Repression of RpARP mRNA is correlated with elongation stimulating
phenotype.
The hypocotyl (25 mm length) of black locust seedlings were excised into 0.5 mm
segments. The segments were incubated in liquid MS medium in absence or presence of
2 % sucrose (SU), 10 u M NAA (N), 20 u M Fusicoccin (FC), 2 % mannose (Mn), or

low temperature (Cold, 40C) for 24 h. (A) For measurement of percent elongation, the
length of the segments were measured after 24 h treatment. Data are means + SD. (%).
(B) For measurement of RpARP mRNA expression, total RNA (6 Li g) were isolated
from the segments after length measurement, and the autoradiogram signal of the RNA
gel blot was quantified using ImageQuant program (Amersham). The values are
expressed as a relative value, where the amount of the lowest RpARP mRNA is defined
as 1.

120

013 6 9 12 24 h
SU+NAA+Cold+ "' *1" * ... «~— ~ m
811+NAA+ .- m
E

0 1 3 6 9 2 4 h
%NAA- {in A” «iii an rm «and

m
su-NAA- w n-Qi.§fl m

SJ-NAA-i- "" '.

 

B
0.5 1 3 6 9 24 h
91 " “W ”"" m
MN W .wquIL-i_--cmn
MT - “C‘nm
c esctm

Figure 3-6. Effects of sucrose and low temperature on auxin suppression on RpARP
mRNA expression.

(A) The hypocotyl segments were treated with liquid MS medium containing 10 u M
NAA and 2 % sucrose (SU+NAA+), 10 LI M NAA only (SU-NAA+), 2 % sucrose (w/v)
only (SU+NAA-), MS medium only (SU-NAA-) or 10 u M NAA and 2 % sucrose with
cold treatments, 40C (SU+NAA+Cold+). (B) Analysis of RpARP mRNA expression
under various sucrose isomers. Hypocotyl segments were incubated for 0.5, 1, 3, 6, 9, 24
h in MS medium containing 2% sucrose (SU), 2% mannose (MN), or 2% mannitol (MT).
As a control, MS medium only (C) was used. Total RNA (6 u g) was isolated after
treatment for the indicated times, and then subjected to RNA blot hybridization with
probes specific for RpARP. Ethidium bromide stained ribosomal RNA is shown in the

right of each panel as a loading control. (SU+NAA+ is a duplication of the left half of
Figure 3A

121

 

1313:) s'urn r RpAFPOFF Himflos_ pBIAH’
RpARPOl-‘F H GUS HNosh pBIARPGtS

 

 

B an su-
NAA + - + -
3123123123121"

“0. 'L-

 

 

pBIARPGUS

 

 

Figure 3-7. The expression of RpARP gene is regulated posttranscriptionally.

(A) Schematic representation of the plasmids used in Arabidopsis transformation and
nothem blot. 358 represents CaMV3SS;UTR, untranslation region; ORF, open reading
frame; NOS, nos terminator. Maps are not to scale. (B) Total RNA was isolated from
whole tissues of RpARP overexpressing (pBIARP) transgenic Arabidopsis or RpARP-
GUS expressing (pBIARPGUS) after treatment with liquid MS medium in absence (-) or
presence (+) of 2 % sucrose (SU) or 10 u M NAA (NAA) for 3 and 12 b. Each lane was
loaded with 6 u g of total RNA. For hybridization signals, RpARP (Top panel) and GUS
(Bottom panel) were used as probe.

122

61
121
181
241
301
361
421
481
541
601
661
721
781
841
901
961

1021
1081
1141
1201
1261

tgggctgtgt
agagggggta
aaggctggga
ttgtggagat
acccatttgc
tgttgttatt
tagcacatgc
agtccaaaat
agatattaca
caaacatcaa
caaaggaaat
tctttttttt
ttatagcatt
gtatcatctt
cctttgagga
tgactaatct
aaaaaaggag

attgctctga

ggggtgaccc

ggttgtatgg
tgatggtgca
agagagcggt
cgccgggaga
tgcatattcc
gttgttctct
tacttctctc
agagcgtacg
aaacaagtta
caaaatatta
aaaataaaaa
atatgtgtca
ttagaatttc
tcccatctca
attacatgat
ttacattaaa
tcgggtccac
agaataggac

acgacctact

aagatagtat gtcgcgggcc

cttctcactc

cacgcaaata

—>

tttccaccct

agttacactt

acgaatgatc
catgaaatgg
taagggggga
gtttacgagt
tattttcaac
tatttagaaa
gaattgggtg
catgtgacta
tttatcatta
aaagaaatat
taagacattt
tacagtacaa
tcaataaata
tcaaggaaga
taatctagac
agaaagcata
gataattttc
aaaaggctag
actactacta
ctacctacgc
tctttttttc

ggaagcaaga

cacttcggga
gaagtggggg
acactttggc
ggagagaaca
ctcacttgcg
atttaataca
agcataacaa
cggttaattc
aaaaatatat
tatttttgtt
ttgttctttc
tgaatatagt
cagtattttg
accgtaactt
Ctagtttata

gcatcaagga

caggtccaga
tgtaaggaga
tggataacag
ctgagtacat
ataaagcaga
acacaaacta
catattgttg
gaacttgttt
ttataaccaa
ccacttcatt
aaaatccaac
gaaaaaaaaa
ggtatctaat
ctctccaaat

tcataatcta

gtggggtcca

 

agtccakatc
ctaattgtgt
caaaactctc
atttaccaag
actatatcca

aaacactcca

tgacaaggta
caagggataa
gtgattgata
gagtaatatt
ggattattat
tttaacgagc
ctacttaacc
tttttaccaa
ttataatagt
atgtgacaca
acatattcat
tatcttactc
Meg
ctagatcctt
caccattaca

tgatgctgac

caqflatccgfl agagctccac

attttcatct
tctccccatc
tcttcatgct
ctcactacca

tcccatttta

Figure 3-8. Nucleotide sequence of RpARP promoter region.

gtacacttgt
atgctcacca
..
cgtcaaaacg

gttagagata

The arrow indicates putative transcriptional start site. The putative TATA box was
underlined. The sucrose repressible response element (TATCCAT-motif) is boxed
(Huang et al. 1990). The analysis of cis-elements has done using the database of
PIantCARE (http://oberon.rug.ac.be/PlantCARE)

123

 

 

.,
" . H.10'olfi'i-‘Jl0fi "1 ‘11

 

 

 

B i
i
‘5
s
s
s
E
C 40
35,
301'
E25:
'gzol
I:
8151
e\"

10 --
0 a L, 1£,.,L, ,1
1 2 4 5

3

Figure 3-9. Expression pattern of RpARP mRNA is correlated negatively with the
extent of elongation of hypocotyl.

(A) The hypocotyl (25 mm length) of black locust seedlings were excised into 0.5 mm
segments in order of from top and bottom (l,2,3,4,5). (B) For measurement of RpARP
mRNA expression in each segment position, total RNA (6 u g) were isolated from the
segments, and the autoradiogram signal of the RNA gel blot was quantified using
ImageQuant program (Amersham). The values is expressed as a relative value, where the
amount of the top segments RpARP mRNA is defined as l. (C) For measurement of
percent elongation, the segments were incubated in MS, 2 % Sucrose medium for 24 h
and then, the length of the segments were measured. Data are means + SD. (%)

124

CONCLUSION

Woody perennials differ in many developmental and physiological respects from
herbaceous plants. Recent development of Populus genomic resources (e.g., its complete
genome sequence and a large number of ESTS) and the advances in genomic and
bioinfomatic tools allow me to take an integrated functional and comparative genomics
approaches to study the tree specific biological properties.

The large-scale computational analysis of poplar ESTS as part of my thesis works,
reveals unique features of expressed genes in the poplar genome. Several multigene
families had a higher copy number in poplar than in Arabidopsis suggesting potential
lineage-specific proliferation of poplar protein families. Another different feature
between Populus and Arabidopsis transcriptome is that the genes involved in
transcriptional regulation are the most divergent while metabolism-related genes are
highly conserved.

Another part of my research is about the genetic regulation of seasonal growth
cycle in Populus. The functional genomic analyses showed that extensive metabolic
switch and alteration in cellular functions occurred during transition of rest to quiescent.
The genes involved in the pentose phosphate pathway were elevated and defense-related
genes were overrepresented during early winter while the genes involved in fermentation
and fatty acid B-oxidation were upregulated and signaling-related genes were
overrepresented during late winter. Furthermore, clustering of these genes into eight
groups based on expression profiles has revealed that plants regulate seasonal growth by
integrating environmental factors with developmental stages. The cold related genes were

regulated by developmental stages of dormancy. This mechanism could help plants to

125

control cold hardiness more closely timed with seasonality than unpredictable
temperature alone does.

Finally in this thesis work, the auxin repressed gene from black locust (RpARP)
has been characterized. As a result, the RpARP gene was repressed by auxin, and the
repression was dependent on metabolizable sugar and temperature. In addition, RpARP
gene expression is negatively associated with shoot elongation. Interestingly, ARP genes
of various plant species characterized from until recently have showed the auxin
repression and the close correlation with maturation of organ/tissues. The results may
suggest that the ARP gene family perform a similar biological function in various organs.

In conclusion, this study provided the first comprehensive view on genetic
regulation of the annual growth cycle and evolutionary adaptation to environment in a
woody perennial, Populus. This study also showed that the functional and comparative
genomic tools recently developed could be used to overcome or compensate for the
limitation of the conventional biochemical methods in the study of tree molecular
biology.

Ultimately, I believe that the knowledge gained through this study could be
leveraged to uncover the evolutionary mechanisms that may account for genetic
architechtures underlying the developmental and physiological differences between
woody perennials and herbaceous annuals. Within this framework, my thesis study could
be further extended to (1) cis- and trans- transcriptional regulatory networks for the
associated signal pathways, (2) species difference and conservation from evolutionary

perspective, (3) application of the identified mechanisms to heterologous systems.

126