View of the Four Members of the GT43 Family Identified in Arabidopsis

IRX14 and IRX14-LIKE: Two Glycosyl Transferases involved in Glucuronoxylan

Biosynthesis in Arabidopsis

A thesis presented to

the faculty of

the College of Arts and Sciences of Ohio University

In partial fulfillment

of the requirements for the degree

Master of Science

Brian D. Keppler

March 2010

This thesis titled

IRX14 and IRX14-LIKE: Two Glycosyl Transferases involved in Glucuronoxylan

Biosynthesis in Arabidopsis

BRIAN D. KEPPLER

has been approved for

the Department of Environmental and Plant Biology

and the College of Arts and Sciences by

Allan M. Showalter

Professor of Environmental and Plant Biology

Benjamin M. Ogles

Dean, College of Arts and Sciences 3

ABSTRACT

KEPPLER, BRIAN D., M.S., March 2010, Plant Biology

IRX14 and IRX14-LIKE: Two Glycosyl Transferases involved in Glucuronoxylan

Biosynthesis in Arabidopsis (96 pp.)

Director of Thesis: Allan M. Showalter

IRX14 and IRX14-LIKE are two closely related glycosyl transferases in the glycosyl transferase 43 (GT43) family of Arabidopsis. IRX14 was initially identified

during a coexpression search to identify glycosyl transferases involved in cell wall

biosynthesis. A T-DNA insertion mutant for IRX14 results in comparatively minor

changes, such as irregular xylem, while a mutation for IRX14L results in no changes.

However, an irx14 irx14L double mutant severely affects growth and development with

the dwarf plants failing to produce an inflorescence stem. Plants which are homozygous

for the mutant version of IRX14 but heterozygous for IRX14L [irx14 irx14L(±)] exhibit

an intermediate phenotype, including noticeably smaller leaves, stems, and underdeveloped siliques which contain only a few seeds. Additionally, the T-DNA

insertion mutant for IRX14 was found to result in a drought tolerant phenotype.

Carbohydrate analysis of total cell wall extracts revealed a reduction in xylose for the

irx14 and irx14 irx14L(±) mutants, consistent with a defect in glucuronoxylan

biosynthesis. Glucuronoxylan, along with cellulose and lignin, are the major components

of secondary cell walls in Arabidopsis. Immunolocalization of xylan with the LM10

antibody revealed a loss of xylan in irx14 mutants and a further reduction in the irx14 4 irx14L(±) mutants. IRX14L likely functions redundantly with IRX14 in glucuronoxylan biosynthesis, with IRX14 having the more important role in the process.

Approved: ______

Allan M. Showalter

Professor of Environmental and Plant Biology 5

ACKNOWLEDGMENTS

First and foremost, I want to express my gratitude for my advisor, Dr. Allan

Showalter, for his constant help, support, and guidance. I have learned so much and

developed a much greater interest in research since I first began working in the lab as an

undergraduate, largely due to the influence of Dr. Showalter. I also thank Dr. Sarah

Wyatt, Dr. Ahmed Faik, and Dr. Marcia Kieliszewski for serving on my advisory

committee and for their helpful advice and guidance.

I am very grateful to the other members of the Showalter lab who have made my time here such an enjoyable experience including Dr. Harjinder Sardar, Dr. Jie Yang, Dr.

Yizhu Zhang, Yan Liang, Dr. Wenliang Xu, Dr. Wei Tang, and Debarati Basu. I am also thankful for the undergraduates who have worked in the lab and assisted me with my

work, Shawna Callaghan, Celeste Taylor, and Rebecca Vondrell, particularly for their

help in screening mutant plants and phenotypic analysis. I thank Dr. Chris Havran for

teaching me the histological techniques and Jeff Thuma for instruction in using the

confocal microscope. I am grateful to Daniel Mullendore at Washington State University

for capturing images with scanning electron microscopy and the Complex Carbohydrate

Research Center in Athens, GA for performing the monosaccharide analysis.

TABLE OF CONTENTS

Page

Abstract ...... 3 Acknowledgments...... 5 List of Tables ...... 8 List of Figures ...... 9 List of Abbreviations ...... 11 Chapter 1: Introduction to the Plant Cell Wall ...... 13 Importance of the Cell Wall ...... 13 Cellulose ...... 13 Hemicelluloses ...... 14 Pectin ...... 16 Glycoproteins ...... 17 Chapter 2: Identification of Glycosyl Transferases by a Coexpression Approach ...... 18 Introduction ...... 18 Materials and Methods ...... 19 The Arabidopsis thaliana Co-response Database ...... 19 Carbohydrate-Active Enzymes (CAZy) ...... 19 Results ...... 20 Discussion ...... 33 Effectiveness of a Coexpression Approach ...... 33 Selection of a Glycosyl Transferase for Further Analysis ...... 35 Chapter 3: IRX14 and IRX14-Like, Two Glycosyl Transfereases involved in Glucuronoxylan Biosynthesis in Arabidopsis ...... 37 Introduction ...... 37 Glucuronoxylan Biosynthesis ...... 37 Materials and Methods ...... 39 Plant Growth Conditions ...... 39 Identification of T-DNA Insertion Lines ...... 40 Bioinformatics ...... 40 RNA Extraction and RT-PCR ...... 41 7

Histology ...... 42 Drought Stress ...... 42 NaCl Treatment ...... 43 Monosaccharide Analysis ...... 43 Xylan Immunolocalization ...... 44 Results ...... 44 The GT43 Family ...... 44 Expression Pattern ...... 47 T-DNA Insertion Mutants ...... 48 Comprehensive Phenotypic Analysis ...... 52 irx14 irx14L Double Mutant ...... 54 RT-PCR Analysis ...... 56 Irregular Xylem ...... 57 Drought Tolerance ...... 61 Salt Tolerance ...... 64 Carbohydrate Analysis ...... 65 Xylan Immunolocalization ...... 66 Allelic T-DNA Mutants ...... 67 Discussion ...... 69 Understanding the Drought Tolerance Phenotype ...... 69 IRX14 and IRX14L Function Redundantly in Glucuronoxylan Biosynthesis ...... 71 Chapter 4: Discussion, Conclusions, and Future Work ...... 74 Complexity of GX Biosynthesis ...... 74 Future Work ...... 75 Appendix A: Genevestigator Expression for the GT43 Family ...... 86 Appendix B: Scanning Electron Microscopy of irx14 and irx14 irx14L Mutants ...... 90

LIST OF TABLES

Page

Table 2.1: Glycosyl transferases coexpressed with AGPs ...... 20

Table 2.2: Summary of the GT Families coexpressed with AGPs ...... 32

Table 3.1: Primers used in the confirmation of T-DNA mutants ...... 41

Table 3.2: Overview of the four members of the GT43 family identified in Arabidopsis ...... 45

Table 3.3: T-DNA insertion mutants currently available for IRX14 and IRX14L ...... 49

Table 3.4: Plate based phenotypic analysis of the irx14 and irx14L mutants ...... 53 9

LIST OF FIGURES

Page

Figure 1.1: Structure of Type I and Type II cell walls ...... 16

Figure 3.1: Structure of Glucuronoxylan in dicots ...... 37

Figure 3.2: Phylogenic tree of the GT43 family ...... 45

Figure 3.3: Amino acid alignments of IRX14 and IRX14L ...... 46

Figure 3.4: Amino acid alignments of IRX9 and IRX9L ...... 46

Figure 3.5: Selected expression data for IRX14 and IRX14L from Genevestigator ...... 47

Figure 3.6: T-DNA insertion mutants for IRX14 and IRX14L ...... 50

Figure 3.7: Identification of irx14 homozygous mutants by PCR ...... 51

Figure 3.8: Identification of irx14L homozygous mutants by PCR...... 51

Figure 3.9: WT, irx14, and irx14L plants grown under normal conditions ...... 52

Figure 3.10: Root length of WT, irx14, and irx14L at day 14 ...... 53

Figure 3.11: Comparison of the WT and irx14 irx14L double mutant rosettes...... 55

Figure 3.12: WT, irx14(±) irx14L, irx14 irx14L(±), and irx14 irx14L plants grown under normal conditions...... 55

Figure 3.13: Phenotypic differences of the irx14 irx14L(±) plants ...... 56

Figure 3.14: RT-PCR analysis of the irx14 and irx14L mutants ...... 57

Figure 3.15: Sections of stems and roots from WT, irx14, and irx14L plants ...... 59

Figure 3.16: Sections of stems and roots from WT, irx14(±) irx14L, and irx14 irx14L(±) plants ...... 60

Figure 3.17: The drought tolerant phenotype of the irx14 mutants ...... 62

Figure 3.18: Water loss from leaves excised from WT, irx14, and irx14L plants ...... 63

Figure 3.19: Expression of SDR1 in WT and irx14 plants ...... 63

Figure 3.20: WT and irx14 mutants grown on various concentrations of NaCl ...... 64

Figure 3.21: Non-cellulosic sugar composition of the cell walls from WT, irx14, irx14L, irx14(±) irx14L, and irx14 irx14L(±) stems ...... 65

Figure 3.22: Xylan immunolocalization in stem sections of WT, irx14, and irx14 irx14L(±) plants ...... 66

Figure 3.23: Location of the GABI_0112C01 T-DNA insert ...... 68

Figure 3.24: Structure of the xylem in the stems and roots of GABI_0112C01 mutants...... 68

Figure 3.25: Characterization of the SALK_117947 mutant line ...... 69

Figure 4.1: Phylogenetic analysis of GT43 family members in Arabidopsis thaliana, Oryza Sativa, Triticum aestivum, and Zea Mays ...... 78

Figure A.1: Expression profile of At1g27600 (IRX9L) ...... 86

Figure A.2: Expression profile of At2g37090 (IRX9) ...... 87

Figure A.3: Expression profile of At4g36890 (IRX14) ...... 88

Figure A.4: Expression profile of At5g67230 (IRX14L) ...... 89

Figure B.1: Cross section of a WT stem ...... 91

Figure B.2: Normal xylem vessel elements of a WT stem...... 92

Figure B.3: Cross section of an irx14 stem ...... 93

Figure B.4: Irregular xylem vessel elements of an irx14 stem ...... 94

Figure B.5: Cross section of an irx14 irx14L double mutant stem ...... 95

Figure B.6: Irregular and collapsed cells of the irx14 irx14L stem ...... 96

LIST OF ABBREVIATIONS

AGP: Arabinogalactan-protein

Ara: Arabinose

CAZy: Carbohydrate-Active Enzymes

CES: Cellulose Synthase

CSL: Cellulose Synthase like

EXT: Extensin

FRA: Fragile Fiber

Fuc: Fucose

Gal: Galactose

GalUA: Galacturonic Acid

GAX: Glucuronoarabinoxylan

Glc: Glucose

GlcUA: Glucuronic Acid

GSL: Glucan Synthase like

GX: Glucuronoxylan

GT: Glycosyl Transferase

HRGP: Hydroxyproline-rich Glycoprotein

Hyp: Hydroxyproline

IRX: Irregular Xylem

Man: Mannose

MS: Murashige and Skoog 12

PCR: Polymerase Chain Reaction

PRP: Proline-rich Protein

Rha: Rhamnose

RT-PCR: Reverse Transcriptase-Polymerase Chain Reaction

T-DNA: Transfer DNA

UTR: Untranslated Region

WT: Wild Type

Xyl: Xylose

CHAPTER 1: INTRODUCTION TO THE PLANT CELL WALL

Importance of the Cell Wall

The cell wall is one of the primary characteristics that distinguishes plant cells from animal cells. It is immensely important for the plant, providing both shape and support. The cell wall can be divided into the primary and secondary cell wall, as well as the middle lamella that exists between the walls of adjacent cells (Carpita and McCann,

2000). The primary cell wall occurs in all cells and continues to grow, while the secondary cell wall is deposited only after cell growth has ceased. The secondary cell wall is often found in specialized cells, such as the xylem tracheary elements and ﬁbers

(Zhong and Ye, 2009).

The plant cell wall is important for humans in terms of wood, paper, food products, and potentially biofuels (Boudet et al., 2003; Harris and Smith, 2006; Gray et al., 2006). It is primarily composed of polysaccharides, particularly cellulose, hemicelluloses, and pectin (Carpita and McCann, 2000). Although the biosynthesis of these components is beginning to be understood, much more needs to be elucidated. In order to manipulate the cell wall for various purposes, a complete understanding of its biosynthesis is necessary.

Cellulose

Cellulose is common to both the primary and secondary plant cell walls and is the major component of each. Cellulose is composed of β-(1,4)-linked D-glucose chains, but 14

in the cell wall it exists in the form of microfibrils. These microfibrils consist of

approximately 36 of the cellulose chains hydrogen bonded together (Herth, 1983).

Cellulose microfibrils are synthesized at the plasma membrane by an enzyme complex known as the rosette (Delmer, 1999). These rosettes consist of six large subunits, which each contain six of the cellulose synthase (CesA) catalytic subunits that synthesize the individual glucose chains (Somerville, 2006). The result is 36 individual chains produced simultaneously that combine to form the microfibrils. A single cellulose microfibril can consist of 15,000 or more individual glucose molecules (Brett, 2000).

Hemicelluloses

Hemicelluloses (also known as cross-linking glycans) are an important aspect of both primary and secondary cell walls. These hemicelluloses serve the purpose of cross linking with the cellulose microfibrils. The particular hemicellulose that predominates depends on the particular species. In dicots, such as Arabidopsis, xyloglucan predominates in the primary cell wall (Fry, 1988). Xyloglucan is composed of β-(1,4)-D- glucan chains similar to cellulose, but this glucan backbone is further substituted with xylose, which in turn can be substituted with galactose and fucose (Carpita and McCann,

2000). Glucuronoxylan (GX) is another hemicellulose in dicots, but it is largely found in the secondary cell wall. GX consists of a β-(1,4)-D-xylan backbone substituted with α-D- glucuronic acid (GlcA), and 4-O-methyl-α-D-glucuronic acid (MeGlcA) (Ebringerová and Heinze, 2000). In dicots, the xylan is substituted with GlcA and MeGlcA at approximately once every eight xylose residues (Brown et al., 2007). In addition, GX in 15

Arabidopsis contains a reducing end structure (known as Sequence 1), consisting of 4-β-

D-Xyl-(1,4)-β-D-Xyl-(1,3)-α-L-Rha-(1,2)-α-D-GalA-(1,4)-D-Xyl, which was previously identified in birch and spruce (Johansson and Samuelson, 1977; Andersson et al., 1983;

Pena et al., 2007).

In some monocots, notably the grasses, the amount of xyloglucan present is quite low, accounting for only about 4% of the dry weight (Fry, 1988). Instead, xylan is the major hemicelluloses found in the primary cell wall of grasses. In grasses, xylan is in the form of glucuronoarabinoxylan (GAX). GAX is similar in structure to the GX found in the secondary cell walls of dicots, but with the addition of α-L-arabinose substitutions

(Ebringerová and Heinze, 2000). Based on whether xyloglucan or GAX is the primary hemicellulose present, plant cell walls can be classified into two types (Figure 1.1)

(Carpita and Gibeaut, 1993; Yokoyama and Nishitani, 2004). In addition, β-(1,3),(1,4)-D- glucans, also known as mixed-linkage β-glucans or simply β-glucans, are particularly prevalent in the primary cell wall of grasses (Fry, 1988). The β-glucans are unbranched polysaccharides consisting of a mixture of β-(1,4)-D-glucose and β-(1,3)-D-glucose

(Carpita and Gibeaut, 1993).

Figure 1.1: Structure of Type I and Type II cell walls. Xyloglucan is the predominate hemicellulose in Type I walls, while GAX predominates in Type II walls. (Yokoyama, et al., Genomic Basis for Cell-Wall Diversity in Plants. A Comparative Approach to Gene Families in Rice and Arabidopsis, Plant and Cell Physiology, 2004, 9, 45, 1111-1121, by permission of Oxford University Press.)

Pectin

Pectin is another group of polysaccharides prevalent in the cell wall, which are responsible for determining the wall porosity (Baron-Epel et al., 1988). In the primary cell wall of dicots, pectin is a major component and can account for up to 35% of the cell wall (Fry, 1988). Two of the more common pectin molecules in the cell wall are homogalacturonan (HGA) and rhamnogalacturonan I (RG-I). HGA consists of α-(1,4)- linked D-galacturonic acid chains, while in RG-I every other the galacturonic acid residues is replaced with rhamnose in the form of an α-(1,4)-D-GalA-α-(1,2)-L-Rha 17

repeating unit (Carpita and McCann, 2000). HGA and RG-I can be further structurally

modified to generate other less common pectins, such as xylogalacturonan and

rhamnogalacturonan II (RG-II) (Carpita and McCann, 2000).

Glycoproteins

In addition to the polysaccharides, glycoproteins represent another component of

the cell wall. The hydroproline-rich glycoproteins (HRGPs) are one such example.

HRGPs can be subdivided into three categories: Arabinogalactan-proteins (AGPs),

extensins (EXTs), and proline-rich proteins (PRPs) (Showalter, 1993). These categories

are based on their level of glycosylation with AGPs being highly glycosylated, EXTs

moderately glycosylated, and PRPs only lightly glycosylated.

AGPs are highly glycosylated cell surface proteins which are unique and ubiquitous in the plant kingdom. AGPs consist of approximately 90% carbohydrate with just 10% protein (Nothnagel, 1997). The protein backbone is rich in proline (P), alanine

(A), serine (S), and threonine (T) and often consists of AP, SP, and TP repeats.

Somewhere between 80-90% of these proline residues are converted to hydroxyproline

(Hyp, O) by enzymes known as prolyl-4-hydroxylases (P4Hs) (Hieta and Myllyharju,

2002). According to the Hyp contiguity hypothesis, either arabinosylation or galactosylation occurs at the Hyp residues depending upon whether they are contiguous

(SOOO) or non-contiguous (AO, SO, TO) (Kieliszewski, 2001). Galactosylation predominates in AGPs and forms the arabinogalactan (AG) polysaccharides, whereas arabinosylation is more common in EXTs where SPPP and SPPPP repeats are abundant. 18

CHAPTER 2: IDENTIFICATION OF GLYCOSYL TRANSFERASES BY A

COEXPRESSION APPROACH

Introduction

In Arabidopsis thaliana, over 450 glycosyl transferases divided into 41 different families are identified in the Carbohydrate-Active enzymes (CAZy) database (Cantarel et al., 2009). Many of these enzymes currently have no known function and few are biochemically characterized. The CAZy database includes not only enzymes that create

glycosidic bonds, but also enzymes involved in their modification and degradation. In

terms of glycosyl transferases, the database currently contains a total of 92 families when

all organisms are considered, but new families are frequently added. New families are

created only when at least one member with an available sequence is biochemically

characterized (Cantarel et al., 2009). Other members are then added to the families based

on sequence similarity.

The complete structure of an AG polysaccharide from AGPs was determined by

nuclear magnetic resonance (NMR) techniques (Tan et al., 2004). Based on this structure,

there are likely to be numerous glycosyl transferases involved in the glycosylation of

AGPs, including a β-galactosyltransferase, β-1,3-galactosyltransferae, β-1,6-

galactosyltransferae, α-1,3-arabinosyltransferase, α-1,5-arabinosyltransferase, α-1,4-

rhamnosyltransferase, α-1,6-glucuronosyltransferase, and α-1,2-fucosyltransferase.

However, none of the enzymes, or genes that encode them, specific for AGPs are definitively identified. In an attempt to identify the enzymes/genes involved in the 19

glycosylation of AGPs, a coexpression approach was utilized. Those glycosyl

transferases coexpressed with a greater number of AGPs were predicted to be involved in

the glycosylation of AGPs.

Materials and Methods

The Arabidopsis thaliana Co-response Database

Coexpression information was obtained from [email protected] - The A.

thaliana Co-Response Database (http://csbdb.mpimp-golm.mpg.de/csbdb/dbcor/ath.html)

(Steinhauser et al., 2004). Each known AGP (including Classical AGPs, Lysine-rich

AGPs, AG-Peptides, and FLAs) was entered into each of the four available matrices

(nasc0271, atge0100, atge0200, and atge0250). These results were compiled and a list of

all genes coexpressed with each of the known AGPs was generated.

Carbohydrate-Active Enzymes (CAZy)

The lists of genes coexpressed with each AGP were then cross referenced against the list of all known glycosyl transferases found in the CAZy database. The final result was a list of each known glycosyl transferase gene and the number of AGPs coexpressed with each. In addition, the particular GT family, number of members in a particular family, and the mechanism of action (inverting or retaining) was also recorded.

Results

Among the over 450 glycosyl transferases in the CAZy database, 148 were found to be coexpressed with at least one AGP (Table 2.1). A total of 30 different GT families were represented (Table 2.2).

Table 2.1: Glycosyl transferases coexpressed with AGPs.

Locus ID Family Members Mechanism # of AGP AGP Spearman (Name) in Coexpressed Locus ID Name Coefficient Family AGPs At5g22940 GT47 39 Inverting 14 At5g10430 AGP4 0.7930 (F8H) GT47 39 Inverting 14 At2g47930 AGP26 0.7876 At5g22940 GT47 39 Inverting 14 At1g03870 FLA9 0.7771 At5g22940 GT47 39 Inverting 14 At4g26320 AGP13 0.7154 At5g22940 GT47 39 Inverting 14 At4g12730 FLA2 0.7152 At5g22940 GT47 39 Inverting 14 At5g53250 AGP22 0.7109 At5g22940 GT47 39 Inverting 14 At5g55730 FLA1 0.6739 At5g22940 GT47 39 Inverting 14 At2g04780 FLA7 0.6221 At5g22940 GT47 39 Inverting 14 At5g56540 AGP14 0.6167 At5g22940 GT47 39 Inverting 14 At3g11700 FLA18 0.6066 At5g22940 GT47 39 Inverting 14 At2g45470 FLA8 0.5923 At5g22940 GT47 39 Inverting 14 At1g28290 AGP31 0.5734 At5g22940 GT47 39 Inverting 14 At4g37450 AGP18 0.5652 At5g22940 GT47 39 Inverting 14 At5g60490 FLA12 0.5618 At4g38040 GT47 39 Inverting 13 At2g47930 AGP26 0.6925 At4g38040 GT47 39 Inverting 13 At1g55330 AGP21 0.6661 At4g38040 GT47 39 Inverting 13 At5g53250 AGP22 0.6563 At4g38040 GT47 39 Inverting 13 At3g11700 FLA18 0.6546 At4g38040 GT47 39 Inverting 13 At5g10430 AGP4 0.6349 At4g38040 GT47 39 Inverting 13 At1g03870 FLA9 0.6232 At4g38040 GT47 39 Inverting 13 At5g65390 AGP7 0.6056 At4g38040 GT47 39 Inverting 13 At3g13520 AGP12 0.6029 At4g38040 GT47 39 Inverting 13 At5g11740 AGP15 0.5990 At4g38040 GT47 39 Inverting 13 At5g56540 AGP14 0.5771 At4g38040 GT47 39 Inverting 13 At4g37450 AGP18 0.5721 At4g38040 GT47 39 Inverting 13 At2g45470 FLA8 0.5704 At4g38040 GT47 39 Inverting 13 At2g04780 FLA7 0.5610 At1g24170 GT8 42 Retaining 12 At1g03870 FLA9 0.8555 At1g24170 GT8 42 Retaining 12 At5g44130 FLA13 0.7805 At1g24170 GT8 42 Retaining 12 At4g12730 FLA2 0.7725 At1g24170 GT8 42 Retaining 12 At4g37450 AGP18 0.7057 At1g24170 GT8 42 Retaining 12 At3g13520 AGP12 0.6835 At1g24170 GT8 42 Retaining 12 At3g11700 FLA18 0.6719 21

At1g24170 GT8 42 Retaining 12 At2g45470 FLA8 0.6649 At1g24170 GT8 42 Retaining 12 At2g04780 FLA7 0.6473 At1g24170 GT8 42 Retaining 12 At5g10430 AGP4 0.6449 At1g24170 GT8 42 Retaining 12 At1g55330 AGP21 0.6297 At1g24170 GT8 42 Retaining 12 At5g11740 AGP15 0.6214 At1g24170 GT8 42 Retaining 12 At2g33790 AGP30 0.5654 At4g39350 GT2 42 Inverting 11 At2g35860 FLA16 0.8399 (CESA2) GT2 42 Inverting 11 At3g52370 FLA15 0.7682 At4g39350 GT2 42 Inverting 11 At4g12730 FLA2 0.6949 At4g39350 GT2 42 Inverting 11 At2g04780 FLA7 0.6195 At4g39350 GT2 42 Inverting 11 At2g45470 FLA8 0.6092 At4g39350 GT2 42 Inverting 11 At5g60490 FLA12 0.6046 At4g39350 GT2 42 Inverting 11 At1g03870 FLA9 0.5950 At4g39350 GT2 42 Inverting 11 At5g10430 AGP4 0.5781 At4g39350 GT2 42 Inverting 11 At5g65390 AGP7 0.5639 At4g39350 GT2 42 Inverting 11 At4g26320 AGP13 0.5618 At4g39350 GT2 42 Inverting 11 At2g47930 AGP26 0.5586 At1g08660 GT29 3 Inverting 10 At1g28290 AGP31 0.7147 At1g08660 GT29 3 Inverting 10 At2g04780 FLA7 0.6670 At1g08660 GT29 3 Inverting 10 At4g12730 FLA2 0.6638 At1g08660 GT29 3 Inverting 10 At1g03870 FLA9 0.6452 At1g08660 GT29 3 Inverting 10 At5g10430 AGP4 0.6400 At1g08660 GT29 3 Inverting 10 At4g37450 AGP18 0.6342 At1g08660 GT29 3 Inverting 10 At5g18690 AGP25 0.6172 At1g08660 GT29 3 Inverting 10 At5g60490 FLA12 0.5844 At1g08660 GT29 3 Inverting 10 At4g26320 AGP13 0.5777 At1g08660 GT29 3 Inverting 10 At5g55730 FLA1 0.5775 At4g02500 GT34 8 Retaining 10 At2g35860 FLA16 0.7929 (XXT2) GT34 8 Retaining 10 At5g55730 FLA1 0.7309 At4g02500 GT34 8 Retaining 10 At4g12730 FLA2 0.6970 At4g02500 GT34 8 Retaining 10 At4g26320 AGP13 0.6907 At4g02500 GT34 8 Retaining 10 At2g04780 FLA7 0.6779 At4g02500 GT34 8 Retaining 10 At5g60490 FLA12 0.6413 At4g02500 GT34 8 Retaining 10 At2g45470 FLA8 0.6318 At4g02500 GT34 8 Retaining 10 At3g60900 FLA10 0.6154 At4g02500 GT34 8 Retaining 10 At3g52370 FLA15 0.5955 At4g02500 GT34 8 Retaining 10 At5g56540 AGP14 0.5786 At5g05170 GT2 42 Inverting 10 At4g12730 FLA2 0.7891 (CESA3) GT2 42 Inverting 10 At2g04780 FLA7 0.7014 At5g05170 GT2 42 Inverting 10 At1g03870 FLA9 0.6765 At5g05170 GT2 42 Inverting 10 At5g10430 AGP4 0.6707 At5g05170 GT2 42 Inverting 10 At2g35860 FLA16 0.6671 At5g05170 GT2 42 Inverting 10 At5g44130 FLA13 0.6574 At5g05170 GT2 42 Inverting 10 At2g45470 FLA8 0.6533 At5g05170 GT2 42 Inverting 10 At4g12730 FLA2 0.6331 At5g05170 GT2 42 Inverting 10 At3g11700 FLA18 0.6131 At5g05170 GT2 42 Inverting 10 At5g60490 FLA12 0.5678 At5g47780 GT8 42 Retaining 10 At2g47930 AGP26 0.7787 22

At5g47780 GT8 42 Retaining 10 At3g11700 FLA18 0.7260 At5g47780 GT8 42 Retaining 10 At1g03870 FLA9 0.7174 At5g47780 GT8 42 Retaining 10 At4g12730 FLA2 0.6774 At5g47780 GT8 42 Retaining 10 At5g65390 AGP7 0.6474 At5g47780 GT8 42 Retaining 10 At5g44130 FLA13 0.6304 At5g47780 GT8 42 Retaining 10 At2g45470 FLA8 0.6057 At5g47780 GT8 42 Retaining 10 At1g55330 AGP21 0.5863 At5g47780 GT8 42 Retaining 10 At3g61640 AGP20 0.5814 At5g47780 GT8 42 Retaining 10 At4g37450 AGP18 0.5750 At1g70090 GT8 42 Retaining 9 At5g65390 AGP7 0.7001 At1g70090 GT8 42 Retaining 9 At5g11740 AGP15 0.6883 At1g70090 GT8 42 Retaining 9 At1g55330 AGP21 0.6538 At1g70090 GT8 42 Retaining 9 At3g11700 FLA18 0.6317 At1g70090 GT8 42 Retaining 9 At5g64310 AGP1 0.6253 At1g70090 GT8 42 Retaining 9 At5g44130 FLA13 0.5895 At1g70090 GT8 42 Retaining 9 At3g61640 AGP20 0.5671 At1g70090 GT8 42 Retaining 9 At3g13520 AGP12 0.5668 At1g70090 GT8 42 Retaining 9 At1g03870 FLA9 0.5663 At1g74380 GT34 8 Retaining 9 At1g03870 FLA9 0.7675 (XXT5) GT34 8 Retaining 9 At4g12730 FLA2 0.7083 At1g74380 GT34 8 Retaining 9 At3g11700 FLA18 0.6663 At1g74380 GT34 8 Retaining 9 At2g45470 FLA8 0.6563 At1g74380 GT34 8 Retaining 9 At5g44130 FLA13 0.6489 At1g74380 GT34 8 Retaining 9 At5g11740 AGP15 0.6381 At1g74380 GT34 8 Retaining 9 At2g04780 FLA7 0.6212 At1g74380 GT34 8 Retaining 9 At5g65390 AGP7 0.5875 At1g74380 GT34 8 Retaining 9 At4g37450 AGP18 0.5622 At2g32620 GT2 42 Inverting 9 At4g12730 FLA2 0.6721 (CSLB2) GT2 42 Inverting 9 At1g03870 FLA9 0.6602 At2g32620 GT2 42 Inverting 9 At5g10430 AGP4 0.6423 At2g32620 GT2 42 Inverting 9 At2g45470 FLA8 0.6332 At2g32620 GT2 42 Inverting 9 At4g26320 AGP13 0.6201 At2g32620 GT2 42 Inverting 9 At5g56540 AGP14 0.6002 At2g32620 GT2 42 Inverting 9 At5g53250 AGP22 0.5966 At2g32620 GT2 42 Inverting 9 At2g33790 AGP30 0.5837 At2g32620 GT2 42 Inverting 9 At5g60490 FLA12 0.5750 At3g18170 GT61 7 Inverting 9 At4g26320 AGP13 0.7236 At3g18170 GT61 7 Inverting 9 At5g60490 FLA12 0.7011 At3g18170 GT61 7 Inverting 9 At5g56540 AGP14 0.6899 At3g18170 GT61 7 Inverting 9 At4g12730 FLA2 0.6619 At3g18170 GT61 7 Inverting 9 At5g53250 AGP22 0.6482 At3g18170 GT61 7 Inverting 9 At1g03870 FLA9 0.6354 At3g18170 GT61 7 Inverting 9 At3g52370 FLA15 0.6278 At3g18170 GT61 7 Inverting 9 At2g33790 AGP30 0.5950 At3g18170 GT61 7 Inverting 9 At2g35860 FLA16 0.5939 At5g03760 GT2 42 Inverting 9 At2g35860 FLA16 0.8053 (CSLA9) GT2 42 Inverting 9 At2g45470 FLA8 0.7345 At5g03760 GT2 42 Inverting 9 At3g52370 FLA15 0.7121 23

At5g03760 GT2 42 Inverting 9 At1g55330 AGP21 0.6988 At5g03760 GT2 42 Inverting 9 At4g12730 FLA2 0.6599 At5g03760 GT2 42 Inverting 9 At4g37450 AGP18 0.6545 At5g03760 GT2 42 Inverting 9 At2g04780 FLA7 0.6281 At5g03760 GT2 42 Inverting 9 At5g55730 FLA1 0.6144 At5g03760 GT2 42 Inverting 9 At2g47930 AGP26 0.5633 At1g23480 GT2 42 Inverting 8 At1g03870 FLA9 0.8729 (CSLA3) GT2 42 Inverting 8 At3g11700 FLA18 0.7706 At1g23480 GT2 42 Inverting 8 At2g47930 AGP26 0.7640 At1g23480 GT2 42 Inverting 8 At5g44130 FLA13 0.7241 At1g23480 GT2 42 Inverting 8 At4g12730 FLA2 0.6928 At1g23480 GT2 42 Inverting 8 At4g37450 AGP18 0.6672 At1g23480 GT2 42 Inverting 8 At2g45470 FLA8 0.6218 At1g23480 GT2 42 Inverting 8 At5g40730 AGP24 0.6117 At3g24040 GT14 11 Inverting 8 At1g28290 AGP31 0.7426 At3g24040 GT14 11 Inverting 8 At5g60490 FLA12 0.6611 At3g24040 GT14 11 Inverting 8 At5g18690 AGP25 0.6341 At3g24040 GT14 11 Inverting 8 At2g04780 FLA7 0.6005 At3g24040 GT14 11 Inverting 8 At2g35860 FLA16 0.5880 At3g24040 GT14 11 Inverting 8 At4g26320 AGP13 0.5848 At3g24040 GT14 11 Inverting 8 At1g03870 FLA9 0.5738 At3g24040 GT14 11 Inverting 8 At5g55730 FLA1 0.5614 At3g28180 GT2 42 Inverting 8 At1g03870 FLA9 0.8152 (CSLC4) GT2 42 Inverting 8 At5g44130 FLA13 0.7591 At3g28180 GT2 42 Inverting 8 At4g12730 FLA2 0.7523 At3g28180 GT2 42 Inverting 8 At1g55330 AGP21 0.7104 At3g28180 GT2 42 Inverting 8 At2g22470 AGP2 0.6814 At3g28180 GT2 42 Inverting 8 At5g64310 AGP1 0.6776 At3g28180 GT2 42 Inverting 8 At3g11700 FLA18 0.6527 At3g28180 GT2 42 Inverting 8 At4g09030 AGP10 0.5796 At3g62720 GT34 8 Retaining 8 At5g64310 AGP1 0.8008 (XXT1) GT34 8 Retaining 8 At3g61640 AGP20 0.7297 At3g62720 GT34 8 Retaining 8 At2g22470 AGP2 0.7166 At3g62720 GT34 8 Retaining 8 At3g11700 FLA18 0.7041 At3g62720 GT34 8 Retaining 8 At4g09030 AGP10 0.6580 At3g62720 GT34 8 Retaining 8 At5g11740 AGP15 0.6401 At3g62720 GT34 8 Retaining 8 At1g03870 FLA9 0.6185 At3g62720 GT34 8 Retaining 8 At3g13520 AGP12 0.6098 At4g15290 GT2 42 Inverting 8 At4g26320 AGP13 0.7140 (CSLB5) GT2 42 Inverting 8 At5g53250 AGP22 0.7080 At4g15290 GT2 42 Inverting 8 At5g56540 AGP14 0.7036 At4g15290 GT2 42 Inverting 8 At2g33790 AGP30 0.6974 At4g15290 GT2 42 Inverting 8 At5g60490 FLA12 0.6511 At4g15290 GT2 42 Inverting 8 At5g40730 AGP24 0.6079 At4g15290 GT2 42 Inverting 8 At1g03870 FLA9 0.5965 At4g15290 GT2 42 Inverting 8 At2g45470 FLA8 0.5894 At1g19360 GT77 19 Retaining 7 At4g09030 AGP10 0.6586 At1g19360 GT77 19 Retaining 7 At5g18690 AGP25 0.6314 24

At1g19360 GT77 19 Retaining 7 At4g12730 FLA2 0.6226 At1g19360 GT77 19 Retaining 7 At5g40730 AGP24 0.6206 At1g19360 GT77 19 Retaining 7 At1g28290 AGP31 0.6189 At1g19360 GT77 19 Retaining 7 At1g03870 FLA9 0.5741 At1g19360 GT77 19 Retaining 7 At5g10430 AGP4 0.5658 At2g03220 GT37 10 Inverting 7 At1g28290 AGP31 0.7126 (FUT1) GT37 10 Inverting 7 At5g18690 AGP25 0.7001 At2g03220 GT37 10 Inverting 7 At5g60490 FLA12 0.6438 At2g03220 GT37 10 Inverting 7 At3g52370 FLA15 0.6370 At2g03220 GT37 10 Inverting 7 At2g04780 FLA7 0.6041 At2g03220 GT37 10 Inverting 7 At4g09030 AGP10 0.6040 At2g03220 GT37 10 Inverting 7 At2g35860 FLA16 0.5648 At2g15370 GT37 10 Inverting 7 At2g04780 FLA7 0.7445 (FUT5) GT37 10 Inverting 7 At1g03870 FLA9 0.6837 At2g15370 GT37 10 Inverting 7 At5g60490 FLA12 0.6681 At2g15370 GT37 10 Inverting 7 At3g52370 FLA15 0.6182 At2g15370 GT37 10 Inverting 7 At5g56540 AGP14 0.5904 At2g15370 GT37 10 Inverting 7 At4g26320 AGP13 0.5660 At2g15370 GT37 10 Inverting 7 At2g45470 FLA8 0.5640 At4g00300 GT31 33 Inverting 7 At5g44130 FLA13 0.7445 At4g00300 GT31 33 Inverting 7 At1g03870 FLA9 0.7369 At4g00300 GT31 33 Inverting 7 At3g11700 FLA18 0.6508 At4g00300 GT31 33 Inverting 7 At1g55330 AGP21 0.6400 At4g00300 GT31 33 Inverting 7 At3g61640 AGP20 0.6221 At4g00300 GT31 33 Inverting 7 At4g12730 FLA2 0.5578 At4g00300 GT31 33 Inverting 7 At2g47930 AGP26 0.5503 At5g64740 GT2 42 Inverting 7 At4g12730 FLA2 0.8210 (CESA6) GT2 42 Inverting 7 At3g11700 FLA18 0.7612 At5g64740 GT2 42 Inverting 7 At1g03870 FLA9 0.7125 At5g64740 GT2 42 Inverting 7 At5g60490 FLA12 0.6513 At5g64740 GT2 42 Inverting 7 At2g35860 FLA16 0.6505 At5g64740 GT2 42 Inverting 7 At5g44130 FLA13 0.6120 At5g64740 GT2 42 Inverting 7 At5g65390 AGP7 0.5513 At1g27440 GT47 39 Inverting 6 At5g03170 FLA11 0.8429 (IRX10) GT47 39 Inverting 6 At5g60490 FLA12 0.6165 At1g27440 GT47 39 Inverting 6 At2g33790 AGP30 0.6153 At1g27440 GT47 39 Inverting 6 At2g35860 FLA16 0.6075 At1g27440 GT47 39 Inverting 6 At5g40730 AGP24 0.5913 At1g27440 GT47 39 Inverting 6 At2g04780 FLA7 0.5850 At2g31750 GT1 121 Inverting 6 At5g44130 FLA13 0.7069 At2g31750 GT1 121 Inverting 6 At3g52370 FLA15 0.6572 At2g31750 GT1 121 Inverting 6 At1g03870 FLA9 0.6201 At2g31750 GT1 121 Inverting 6 At4g37450 AGP18 0.6132 At2g31750 GT1 121 Inverting 6 At2g04780 FLA7 0.5811 At2g31750 GT1 121 Inverting 6 At2g45470 FLA8 0.5740 At3g03050 GT2 42 Inverting 6 At4g09030 AGP10 0.6630 (CSLD3) GT2 42 Inverting 6 At5g10430 AGP4 0.6410 At3g03050 GT2 42 Inverting 6 At3g11700 FLA18 0.6220 25

At3g03050 GT2 42 Inverting 6 At1g03870 FLA9 0.5933 At3g03050 GT2 42 Inverting 6 At3g61640 AGP20 0.5849 At3g03050 GT2 42 Inverting 6 At5g11740 AGP15 0.5723 At3g25140 GT8 42 Retaining 6 At5g55730 FLA1 0.6591 At3g25140 GT8 42 Retaining 6 At2g35860 FLA16 0.6364 At3g25140 GT8 42 Retaining 6 At2g04780 FLA7 0.5853 At3g25140 GT8 42 Retaining 6 At5g03170 FLA11 0.5851 At3g25140 GT8 42 Retaining 6 At1g28290 AGP31 0.5601 At3g25140 GT8 42 Retaining 6 At3g61640 AGP20 0.5473 At3g56000 GT2 42 Inverting 6 At1g03870 FLA9 0.6487 (CSLA14) GT2 42 Inverting 6 At5g53250 AGP22 0.5821 At3g56000 GT2 42 Inverting 6 At5g44130 FLA13 0.5786 At3g56000 GT2 42 Inverting 6 At5g40730 AGP24 0.5777 At3g56000 GT2 42 Inverting 6 At3g11700 FLA18 0.5686 At3g56000 GT2 42 Inverting 6 At4g37450 AGP18 0.5639 At4g02130 GT8 42 Retaining 6 At4g37450 AGP18 0.7607 At4g02130 GT8 42 Retaining 6 At5g44130 FLA13 0.6998 At4g02130 GT8 42 Retaining 6 At2g45470 FLA8 0.6568 At4g02130 GT8 42 Retaining 6 At3g11700 FLA18 0.6261 At4g02130 GT8 42 Retaining 6 At5g65390 AGP7 0.5775 At4g02130 GT8 42 Retaining 6 At1g03870 FLA9 0.5578 At5g09870 GT2 42 Inverting 6 At3g11700 FLA18 0.7428 (CESA5) GT2 42 Inverting 6 At5g44130 FLA13 0.7113 At5g09870 GT2 42 Inverting 6 At3g52370 FLA15 0.6850 At5g09870 GT2 42 Inverting 6 At2g35860 FLA16 0.6523 At5g09870 GT2 42 Inverting 6 At2g47930 AGP26 0.6081 At5g09870 GT2 42 Inverting 6 At5g11740 AGP15 0.5970 At5g17420 GT2 42 Inverting 6 At5g03170 FLA11 0.8724 (CESA7) GT2 42 Inverting 6 At5g60490 FLA12 0.7619 At5g17420 GT2 42 Inverting 6 At1g28290 AGP31 0.7289 At5g17420 GT2 42 Inverting 6 At5g40730 AGP24 0.6634 At5g17420 GT2 42 Inverting 6 At2g04780 FLA7 0.6088 At5g17420 GT2 42 Inverting 6 At5g10430 AGP4 0.5671 At5g22740 GT2 42 Inverting 6 At1g28290 AGP31 0.7608 (CSLA2) GT2 42 Inverting 6 At5g55730 FLA1 0.6956 At5g22740 GT2 42 Inverting 6 At2g35860 FLA16 0.6895 At5g22740 GT2 42 Inverting 6 At5g10430 AGP4 0.5992 At5g22740 GT2 42 Inverting 6 At2g04780 FLA7 0.5926 At5g22740 GT2 42 Inverting 6 At5g18690 AGP25 0.5607 At5g50420 GT68 3 Inverting 6 At2g04780 FLA7 0.7046 At5g50420 GT68 3 Inverting 6 At5g18690 AGP25 0.7045 At5g50420 GT68 3 Inverting 6 At3g52370 FLA15 0.6122 At5g50420 GT68 3 Inverting 6 At5g60490 FLA12 0.6114 At5g50420 GT68 3 Inverting 6 At1g55330 AGP21 0.5794 At5g50420 GT68 3 Inverting 6 At1g03870 FLA9 0.5755 At5g61840 GT47 39 Inverting 6 At5g60490 FLA12 0.6615 (IRX10L) GT47 39 Inverting 6 At1g28290 AGP31 0.6607 At5g61840 GT47 39 Inverting 6 At5g65390 AGP7 0.5837 26

At5g61840 GT47 39 Inverting 6 At2g35860 FLA16 0.5785 At5g61840 GT47 39 Inverting 6 At2g04780 FLA7 0.5725 At5g61840 GT47 39 Inverting 6 At5g53250 AGP22 0.5635 At1g02730 GT2 42 Inverting 5 At1g28290 AGP31 0.8137 (CSLD5) GT2 42 Inverting 5 At5g55730 FLA1 0.7562 At1g02730 GT2 42 Inverting 5 At4g37450 AGP18 0.7389 At1g02730 GT2 42 Inverting 5 At3g60900 FLA10 0.6751 At1g02730 GT2 42 Inverting 5 At2g04780 FLA7 0.6007 At1g13250 GT8 42 Retaining 5 At5g55730 FLA1 0.7647 At1g13250 GT8 42 Retaining 5 At2g45470 FLA8 0.6576 At1g13250 GT8 42 Retaining 5 At2g35860 FLA16 0.6229 At1g13250 GT8 42 Retaining 5 At3g52370 FLA15 0.5934 At1g13250 GT8 42 Retaining 5 At4g37450 AGP18 0.5770 At1g34270 GT47 39 Inverting 5 At3g61640 AGP20 0.7197 At1g34270 GT47 39 Inverting 5 At1g28290 AGP31 0.6445 At1g34270 GT47 39 Inverting 5 At5g55730 FLA1 0.6104 At1g34270 GT47 39 Inverting 5 At2g35860 FLA16 0.5773 At1g34270 GT47 39 Inverting 5 At2g04780 FLA7 0.5690 At3g02350 GT8 42 Retaining 5 At4g12730 FLA2 0.6804 At3g02350 GT8 42 Retaining 5 At2g35860 FLA16 0.6640 At3g02350 GT8 42 Retaining 5 At1g03870 FLA9 0.6344 At3g02350 GT8 42 Retaining 5 At2g04780 FLA7 0.5933 At3g02350 GT8 42 Retaining 5 At2g45470 FLA8 0.5827 At3g05320 GT65 1 Inverting 5 At1g03870 FLA9 0.7018 At3g05320 GT65 1 Inverting 5 At5g44130 FLA13 0.6954 At3g05320 GT65 1 Inverting 5 At3g11700 FLA18 0.6657 At3g05320 GT65 1 Inverting 5 At3g13520 AGP12 0.6138 At3g05320 GT65 1 Inverting 5 At3g61640 AGP20 0.6049 At4g32410 GT2 42 Inverting 5 At4g12730 FLA2 0.7341 (CESA1) GT2 42 Inverting 5 At2g35860 FLA16 0.6990 At4g32410 GT2 42 Inverting 5 At1g03870 FLA9 0.6870 At4g32410 GT2 42 Inverting 5 At5g44130 FLA13 0.6111 At4g32410 GT2 42 Inverting 5 At3g11700 FLA18 0.5829 At5g11110 GT4 24 Retaining 5 At5g64310 AGP1 0.7312 At5g11110 GT4 24 Retaining 5 At2g22470 AGP2 0.7211 At5g11110 GT4 24 Retaining 5 At4g09030 AGP10 0.7056 At5g11110 GT4 24 Retaining 5 At3g13520 AGP12 0.6508 At5g11110 GT4 24 Retaining 5 At5g11740 AGP15 0.6474 At5g15650 GT75 5 Inverting 5 At1g28290 AGP31 0.7019 At5g15650 GT75 5 Inverting 5 At2g04780 FLA7 0.6056 At5g15650 GT75 5 Inverting 5 At2g35860 FLA16 0.5622 At5g15650 GT75 5 Inverting 5 At5g64310 AGP1 0.5588 At5g15650 GT75 5 Inverting 5 At5g18690 AGP25 0.5560 At5g16190 GT2 42 Inverting 5 At5g55730 FLA1 0.6506 (CSLA11) GT2 42 Inverting 5 At4g37450 AGP18 0.5932 At5g16190 GT2 42 Inverting 5 At4g26320 AGP13 0.5797 At5g16190 GT2 42 Inverting 5 At1g03870 FLA9 0.5708 At5g16190 GT2 42 Inverting 5 At4g12730 FLA2 0.5632 27

At5g44030 GT2 42 Inverting 5 At5g03170 FLA11 0.9057 (CESA4) GT2 42 Inverting 5 At5g60490 FLA12 0.7720 At5g44030 GT2 42 Inverting 5 At5g40730 AGP24 0.7112 At5g44030 GT2 42 Inverting 5 At2g04780 FLA7 0.6387 At5g44030 GT2 42 Inverting 5 At5g10430 AGP4 0.5887 At1g08280 GT29 3 Inverting 4 At2g04780 FLA7 0.6117 At1g08280 GT29 3 Inverting 4 At1g28290 AGP31 0.6077 At1g08280 GT29 3 Inverting 4 At5g10430 AGP4 0.5776 At1g08280 GT29 3 Inverting 4 At2g35860 FLA16 0.5507 At1g19710 GT4 24 Retaining 4 At1g28290 AGP31 0.6822 At1g19710 GT4 24 Retaining 4 At2g04780 FLA7 0.6741 At1g19710 GT4 24 Retaining 4 At5g18690 AGP25 0.6244 At1g19710 GT4 24 Retaining 4 At5g60490 FLA12 0.5471 At1g34130 GT66 2 Inverting 4 At1g28290 AGP31 0.8278 At1g34130 GT66 2 Inverting 4 At3g60900 FLA10 0.7218 At1g34130 GT66 2 Inverting 4 At2g04780 FLA7 0.6692 At1g34130 GT66 2 Inverting 4 At5g55730 FLA1 0.6317 At1g71070 GT14 11 Inverting 4 At1g28290 AGP31 0.7329 At1g71070 GT14 11 Inverting 4 At2g04780 FLA7 0.6508 At1g71070 GT14 11 Inverting 4 At3g52370 FLA15 0.5957 At1g71070 GT14 11 Inverting 4 At2g35860 FLA16 0.5595 At2g22900 GT34 8 Retaining 4 At1g28290 AGP31 0.7978 At2g22900 GT34 8 Retaining 4 At5g55730 FLA1 0.6736 At2g22900 GT34 8 Retaining 4 At2g35860 FLA16 0.6508 At2g22900 GT34 8 Retaining 4 At2g04780 FLA7 0.6006 At2g37585 GT14 11 Inverting 4 At1g55330 AGP21 0.7156 At2g37585 GT14 11 Inverting 4 At2g04780 FLA7 0.7148 At2g37585 GT14 11 Inverting 4 At1g28290 AGP31 0.6449 At2g37585 GT14 11 Inverting 4 At5g55730 FLA1 0.6025 At3g14570 GT48 12 Inverting 4 At5g56540 AGP14 0.7530 (Gsl4) GT48 12 Inverting 4 At5g60490 FLA12 0.7041 At3g14570 GT48 12 Inverting 4 At4g09030 AGP10 0.6338 At3g14570 GT48 12 Inverting 4 At5g40730 AGP24 0.6283 At3g59100 GT48 12 Inverting 4 At5g60490 FLA12 0.6386 (Gsl11) GT48 12 Inverting 4 At1g28290 AGP31 0.6129 At3g59100 GT48 12 Inverting 4 At5g18690 AGP25 0.6051 At3g59100 GT48 12 Inverting 4 At2g04780 FLA7 0.5917 At3g61130 GT8 42 Retaining 4 At4g12730 FLA2 0.6536 At3g61130 GT8 42 Retaining 4 At4g37450 AGP18 0.6220 At3g61130 GT8 42 Retaining 4 At2g45470 FLA8 0.5866 At3g61130 GT8 42 Retaining 4 At5g55730 FLA1 0.5472 At3g62660 GT8 42 Retaining 4 At4g12730 FLA2 0.6843 At3g62660 GT8 42 Retaining 4 At3g11700 FLA18 0.6786 At3g62660 GT8 42 Retaining 4 At3g13520 AGP12 0.6664 At3g62660 GT8 42 Retaining 4 At2g35860 FLA16 0.6283 At4g01220 GT77 19 Retaining 4 At3g52370 FLA15 0.6471 At4g01220 GT77 19 Retaining 4 At4g37450 AGP18 0.5814 At4g01220 GT77 19 Retaining 4 At2g04780 FLA7 0.5677 28

At4g01220 GT77 19 Retaining 4 At1g03870 FLA9 0.5640 At4g11350 GT31 33 Inverting 4 At4g09030 AGP10 0.7576 At4g11350 GT31 33 Inverting 4 At5g64310 AGP1 0.7276 At4g11350 GT31 33 Inverting 4 At2g22470 AGP2 0.6018 At4g11350 GT31 33 Inverting 4 At5g11740 AGP15 0.5684 At4g36890 GT43 4 Inverting 4 At5g03170 FLA11 0.6653 (IRX14) GT43 4 Inverting 4 At5g60490 FLA12 0.6515 At4g36890 GT43 4 Inverting 4 At2g35860 FLA16 0.5713 At4g36890 GT43 4 Inverting 4 At5g40730 AGP24 0.5556 At5g05860 GT1 121 Inverting 4 At2g22470 AGP2 0.6354 At5g05860 GT1 121 Inverting 4 At4g09030 AGP10 0.6213 At5g05860 GT1 121 Inverting 4 At5g18690 AGP25 0.6168 At5g05860 GT1 121 Inverting 4 At1g03870 FLA9 0.5913 At5g19690 GT66 2 Inverting 4 At1g28290 AGP31 0.8285 At5g19690 GT66 2 Inverting 4 At2g04780 FLA7 0.6859 At5g19690 GT66 2 Inverting 4 At3g60900 FLA10 0.6788 At5g19690 GT66 2 Inverting 4 At5g55730 FLA1 0.6493 At5g39990 GT14 11 Inverting 4 At1g28290 AGP31 0.7538 At5g39990 GT14 11 Inverting 4 At2g04780 FLA7 0.6604 At5g39990 GT14 11 Inverting 4 At5g60490 FLA12 0.6173 At5g39990 GT14 11 Inverting 4 At5g55730 FLA1 0.5509 At5g54690 GT8 42 Retaining 4 At5g03170 FLA11 0.7996 (IRX8) GT8 42 Retaining 4 At5g40730 AGP24 0.7103 At5g54690 GT8 42 Retaining 4 At5g60490 FLA12 0.6534 At5g54690 GT8 42 Retaining 4 At2g35860 FLA16 0.5520 At5g55500 GT61 7 Inverting 4 At5g18690 AGP25 0.7262 GT61 7 Inverting 4 At4g09030 AGP10 0.7112 At5g55500 GT61 7 Inverting 4 At5g60490 FLA12 0.6608 At5g55500 GT61 7 Inverting 4 At3g52370 FLA15 0.5764 At5g62220 GT47 39 Inverting 4 At5g64310 AGP1 0.6333 At5g62220 GT47 39 Inverting 4 At5g11740 AGP15 0.6108 At5g62220 GT47 39 Inverting 4 At2g22470 AGP2 0.5979 At5g62220 GT47 39 Inverting 4 At3g11700 FLA18 0.5711 At1g06780 GT8 42 Retaining 3 At4g09030 AGP10 0.6979 At1g06780 GT8 42 Retaining 3 At5g56540 AGP14 0.6755 At1g06780 GT8 42 Retaining 3 At5g40730 AGP24 0.6383 At1g16570 GT33 1 Inverting 3 At1g28290 AGP31 0.6756 At1g16570 GT33 1 Inverting 3 At5g18690 AGP25 0.6306 At1g16570 GT33 1 Inverting 3 At2g04780 FLA7 0.5626 At1g16900 GT22 3 Inverting 3 At5g55730 FLA1 0.7303 At1g16900 GT22 3 Inverting 3 At1g28290 AGP31 0.6487 At1g16900 GT22 3 Inverting 3 At5g64310 AGP1 0.5963 At1g67880 GT17 6 Inverting 3 At5g64310 AGP1 0.6671 At1g67880 GT17 6 Inverting 3 At2g22470 AGP2 0.6004 At1g67880 GT17 6 Inverting 3 At4g09030 AGP10 0.5959 At1g74800 GT31 33 Inverting 3 At1g28290 AGP31 0.6857 At1g74800 GT31 33 Inverting 3 At2g04780 FLA7 0.6295 At1g74800 GT31 33 Inverting 3 At5g55730 FLA1 0.5756 29

At2g35610 GT77 19 Retaining 3 At2g04780 FLA7 0.6334 (XEG113) GT77 19 Retaining 3 At4g09030 AGP10 0.6291 At2g35610 GT77 19 Retaining 3 At5g10430 AGP4 0.6012 At2g35650 GT2 42 Inverting 3 At2g45470 FLA8 0.6911 (CSLA7) GT2 42 Inverting 3 At4g37450 AGP18 0.6369 At2g35650 GT2 42 Inverting 3 At2g47930 AGP26 0.5476 At2g38650 GT8 42 Retaining 3 At1g28290 AGP31 0.6873 At2g38650 GT8 42 Retaining 3 At2g04780 FLA7 0.6634 At2g38650 GT8 42 Retaining 3 At5g03170 FLA11 0.5628 At2g44660 GT57 2 Inverting 3 At5g55730 FLA1 0.5949 At2g44660 GT57 2 Inverting 3 At5g18690 AGP25 0.5535 At2g44660 GT57 2 Inverting 3 At1g28290 AGP31 0.5475 At3g11420 GT31 33 Inverting 3 At2g22470 AGP2 0.6844 At3g11420 GT31 33 Inverting 3 At5g64310 AGP1 0.6253 At3g11420 GT31 33 Inverting 3 At4g09030 AGP10 0.6087 At3g15350 GT14 11 Inverting 3 At4g09030 AGP10 0.7275 At3g15350 GT14 11 Inverting 3 At5g64310 AGP1 0.6476 At3g15350 GT14 11 Inverting 3 At2g22470 AGP2 0.5971 At3g29320 GT35 2 Retaining 3 At2g45470 FLA8 0.7544 At3g29320 GT35 2 Retaining 3 At4g37450 AGP18 0.5889 At3g29320 GT35 2 Retaining 3 At5g44130 FLA13 0.5737 At4g23490 GT31 33 Inverting 3 At5g55730 FLA1 0.6433 At4g23490 GT31 33 Inverting 3 At1g28290 AGP31 0.6130 At4g23490 GT31 33 Inverting 3 At3g52370 FLA15 0.5920 At4g31590 GT2 42 Inverting 3 At4g12730 FLA2 0.7256 (CSLC5) GT2 42 Inverting 3 At2g35860 FLA16 0.6607 At4g31590 GT2 42 Inverting 3 At2g45470 FLA8 0.6537 At4g31780 GT28 4 Inverting 3 At5g44130 FLA13 0.7007 At4g31780 GT28 4 Inverting 3 At2g45470 FLA8 0.6063 At4g31780 GT28 4 Inverting 3 At4g37450 AGP18 0.5867 At5g02410 GT59 1 Inverting 3 At1g28290 AGP31 0.6235 At5g02410 GT59 1 Inverting 3 At2g04780 FLA7 0.5665 At5g02410 GT59 1 Inverting 3 At5g18690 AGP25 0.5494 At5g05890 GT1 121 Inverting 3 At5g10430 AGP4 0.7073 At5g05890 GT1 121 Inverting 3 At1g03870 FLA9 0.6943 At5g05890 GT1 121 Inverting 3 At5g53250 AGP22 0.5767 At5g15050 GT14 11 Inverting 3 At5g55730 FLA1 0.6137 At5g15050 GT14 11 Inverting 3 At2g45470 FLA8 0.5933 At5g15050 GT14 11 Inverting 3 At3g60900 FLA10 0.5912 At5g24300 GT5 6 Retaining 3 At3g11700 FLA18 0.6442 At5g24300 GT5 6 Retaining 3 At5g44130 FLA13 0.6148 At5g24300 GT5 6 Retaining 3 At2g45470 FLA8 0.5615 At5g37180 GT4 24 Retaining 3 At5g56540 AGP14 0.6673 At5g37180 GT4 24 Retaining 3 At3g52370 FLA15 0.5736 At5g37180 GT4 24 Retaining 3 At5g65390 AGP7 0.5510 At5g53340 GT31 33 Inverting 3 At5g18690 AGP25 0.6150 At5g53340 GT31 33 Inverting 3 At1g28290 AGP31 0.5915 30

At5g53340 GT31 33 Inverting 3 At2g04780 FLA7 0.5839 At5g62620 GT31 33 Inverting 3 At5g64310 AGP1 0.6048 At5g62620 GT31 33 Inverting 3 At4g09030 AGP10 0.5947 At5g62620 GT31 33 Inverting 3 At2g22470 AGP2 0.5672 At1g05570 GT48 12 Inverting 2 At5g18690 AGP25 0.5856 (Gsl6) GT48 12 Inverting 2 At5g65390 AGP7 0.5639 At1g07240 GT1 121 Inverting 2 At1g03870 FLA9 0.7161 At1g07240 GT1 121 Inverting 2 At4g12730 FLA2 0.5494 At1g21480 GT47 39 Inverting 2 At5g56540 AGP14 0.5897 At1g21480 GT47 39 Inverting 2 At5g18690 AGP25 0.5633 At1g30530 GT1 121 Inverting 2 At5g55730 FLA1 0.7979 At1g30530 GT1 121 Inverting 2 At3g52370 FLA15 0.6795 At1g52420 GT4 24 Retaining 2 At1g28290 AGP31 0.5775 At1g52420 GT4 24 Retaining 2 At5g18690 AGP25 0.5621 At1g53290 GT31 33 Inverting 2 At2g04780 FLA7 0.6791 At1g53290 GT31 33 Inverting 2 At5g60490 FLA12 0.5633 At1g60470 GT8 42 Retaining 2 At4g09030 AGP10 0.6972 At1g60470 GT8 42 Retaining 2 At5g64310 AGP1 0.5693 At1g68470 GT47 39 Inverting 2 At3g52370 FLA15 0.8260 At1g68470 GT47 39 Inverting 2 At2g35860 FLA16 0.5682 At1g73160 GT4 24 Retaining 2 At1g03870 FLA9 0.5849 At1g73160 GT4 24 Retaining 2 At3g52370 FLA15 0.5646 At2g20370 GT47 39 Inverting 2 At3g52370 FLA15 0.6817 (MUR3) GT47 39 Inverting 2 At2g35860 FLA16 0.6304 At2g28080 GT1 121 Inverting 2 At5g18690 AGP25 0.6373 At2g28080 GT1 121 Inverting 2 At3g61640 AGP20 0.5697 At2g29750 GT1 121 Inverting 2 At4g37450 AGP18 0.6568 At2g29750 GT1 121 Inverting 2 At4g12730 FLA2 0.5766 At2g41640 GT61 7 Inverting 2 At5g64310 AGP1 0.6198 At2g41640 GT61 7 Inverting 2 At2g22470 AGP2 0.5973 At3g27540 GT17 6 Inverting 2 At5g64310 AGP1 0.6548 At3g27540 GT17 6 Inverting 2 At2g22470 AGP2 0.5860 At3g50740 GT1 121 Inverting 2 At1g03870 FLA9 0.7527 At3g50740 GT1 121 Inverting 2 At5g10430 AGP4 0.6108 At3g50760 GT8 42 Retaining 2 At5g64310 AGP1 0.6832 At3g50760 GT8 42 Retaining 2 At4g09030 AGP10 0.5667 At4g17770 GT20 11 Retaining 2 At2g35860 FLA16 0.6263 At4g17770 GT20 11 Retaining 2 At1g28290 AGP31 0.5447 At4g32120 GT31 33 Inverting 2 At1g28290 AGP31 0.7214 At4g32120 GT31 33 Inverting 2 At2g04780 FLA7 0.5566 At5g01220 GT4 24 Retaining 2 At5g44130 FLA13 0.6249 At5g01220 GT4 24 Retaining 2 At1g03870 FLA9 0.5529 At5g14850 GT22 3 Inverting 2 At5g55730 FLA1 0.6511 At5g14850 GT22 3 Inverting 2 At1g28290 AGP31 0.5661 At5g16510 GT75 5 Inverting 2 At1g28290 AGP31 0.5994 At5g16510 GT75 5 Inverting 2 At2g35860 FLA16 0.5459 At5g41460 GT31 33 Inverting 2 At5g55730 FLA1 0.8334 31

At5g41460 GT31 33 Inverting 2 At2g45470 FLA8 0.5582 At1g06000 GT1 121 Inverting 1 At5g55730 FLA1 0.6262 At1g10400 GT1 121 Inverting 1 At4g37450 AGP18 0.6387 At1g12990 GT17 6 Inverting 1 At3g61640 AGP20 0.6205 At1g14080 GT37 10 Inverting 1 At1g03870 FLA9 0.5772 (FUT6) At1g18580 GT8 42 Retaining 1 At5g44130 FLA13 0.5613 At1g20575 GT2 42 Inverting 1 At5g56540 AGP14 0.5667 At1g24100 GT1 121 Inverting 1 At2g35860 FLA16 0.6326 At1g27120 GT31 33 Inverting 1 At5g55730 FLA1 0.8558 At1g28710 GT77 19 Retaining 1 At2g45470 FLA8 0.5833 At1g60140 GT20 11 Retaining 1 At5g10430 AGP4 0.5909 At1g68020 GT20 11 Retaining 1 At2g22470 AGP2 0.5822 At1g71220 GT24 1 Retaining 1 At1g28290 AGP31 0.7680 At1g73370 GT4 24 Retaining 1 At5g56540 AGP14 0.6549

At1g78800 GT4 24 Retaining 1 At5g18690 AGP25 0.5643 At2g19880 GT21 1 Inverting 1 At4g09030 AGP10 0.5700 At2g24630 GT2 42 Inverting 1 At5g55730 FLA1 0.7547 (CslC8) At2g25300 GT31 33 Inverting 1 At1g28290 AGP31 0.5845 At2g32430 GT31 33 Inverting 1 At5g40730 AGP24 0.6101 At2g37090 GT43 4 Inverting 1 At5g40730 AGP24 0.6673 (IRX9) At3g07330 GT2 42 Inverting 1 At5g18690 AGP25 0.5522 (CslC6) At3g15940 GT4 24 Retaining 1 At1g28290 AGP31 0.5990 At3g21750 GT1 121 Inverting 1 At4g09030 AGP10 0.5937 At3g46970 GT35 2 Retaining 1 At5g44130 FLA13 0.6209 At3g58790 GT8 42 Retaining 1 At5g11740 AGP15 0.5681 At4g01750 GT77 19 Retaining 1 At4g12730 FLA2 0.6232 At4g04970 GT48 12 Inverting 1 At2g45470 FLA8 0.5600 (Gsl1) At4g07960 GT2 42 Inverting 1 At3g57690 AGP23 0.5857 (CslC12) At4g09500 GT1 121 Inverting 1 At5g55730 FLA1 0.6963 At4g18230 GT1 121 Inverting 1 At5g18690 AGP25 0.5740 At4g18240 GT5 6 Retaining 1 At5g44130 FLA13 0.5640 At4g18780 GT2 42 Inverting 1 At5g40730 AGP24 0.7099 (CesA8) At4g21060 GT31 33 Inverting 1 At1g28290 AGP31 0.5578 At4g24000 GT2 42 Inverting 1 At3g57690 AGP23 0.7639 (CslG2) At4g26940 GT31 33 Inverting 1 At2g35860 FLA16 0.6081 At4g38240 GT13 1 Inverting 1 At4g12730 FLA2 0.5677 At5g07720 GT34 8 Retaining 1 At1g03870 FLA9 0.5944 At5g16910 GT2 42 Inverting 1 At4g09030 AGP10 0.6024 (CslD2) At5g20410 GT28 4 Inverting 1 At3g57690 AGP23 0.6164 32

At5g38460 GT57 2 Inverting 1 At5g55730 FLA1 0.5939 At5g44820 GT77 19 Retaining 1 At5g44130 FLA13 0.5874 At5g66690 GT1 121 Inverting 1 At5g65390 AGP7 0.5860

Table 2.2: Summary of the GT Families coexpressed with AGPs.

# of Members # of % of # of Coexpressed Members Members Coexpressed Family with AGPs in Family Coexpressed AGPs Notable Members GT1 15 121 12.4% 16 GT2 24 42 57.1% 30 CESA, CSL GT4 9 24 37.5% 14 GT5 2 6 33.3% 3 GT8 16 42 38.1% 25 IRX8, GAUT1 GT13 1 1 100.0% 1 GT14 6 11 54.5% 15 GT17 3 6 50.0% 4 GT20 3 11 27.3% 4 GT21 1 1 100.0% 1 GT22 2 3 66.7% 3 GT24 1 1 100.0% 1 GT28 2 4 50.0% 3 GT29 2 3 66.7% 11 GT31 15 33 45.5% 19 GT33 1 1 100.0% 3 GT34 5 8 62.5% 22 FUT1 GT35 2 2 100.0% 3 GT37 3 10 30.0% 11 XXT1, XXT2, XXT5 GT43 2 4 50.0% 4 IRX9, IRX14 GT47 9 39 23.1% 27 F8H, MUR3, IRX10, IRX10L GT48 4 12 33.3% 9 GSL GT57 2 2 100.0% 3 GT59 1 1 100.0% 3 GT61 3 7 42.9% 13 GT65 1 1 100.0% 5 GT66 2 2 100.0% 4 GT68 1 3 33.3% 6 GT75 2 5 40.0% 5 GT77 6 19 31.6% 12 XEG113

Discussion

Effectiveness of a Coexpression Approach

Because many of the glycosyl transferases identified in this study have known

functions, assessing the effectiveness of the approach is feasible. For instance,

At5g22940, a member of the GT47 family, was found to be coexpressed with 14 different

AGPs, more than any other glycosyl transferase. However, At5g22940, also known as

FRA8 Homolog (F8H), is now thought to function in xylan biosynthesis, specifically

involved in the synthesis of the reducing end structure along with FRA8 (Lee et al.,

2009). Other members of the GT47 family coexpressed with AGPs include IRX10 and

IRX10L, both of which are also involved in xylan biosynthesis (Brown et al., 2009; Wu

et al., 2009). At4g39350, a member of the GT2 family, was found to be coexpressed with

11 AGPs. However, At4g39350 is a cellulose synthesis subunit (CESA2). In fact, eight of

the 10 CESA subunits were found to be coexpressed with at least one AGP. The fact that so many of the CESA genes are coexpressed with AGPs is significant and could suggest some sort of interaction between these components. Research suggests that AGPs and cellulose synthase may have a connection, perhaps indirectly, through microtubules

(Driouich and Baskin, 2008). Disruption of AGPs by Yariv reagent results in

disorganized microtubules (Sardar et al., 2006), while microtubules are proposed to act as

a pathway for the cellulose synthase rosette (Paredez et al., 2006). At1g74380 and

At4g02500, members of the GT34 family, were coexpressed with nine and 10 AGPs,

respectively. At4g02500 and At1g74380 are better known as XXT2 and XXT5 and are

known to function as a xylosyltransferases in xyloglucan biosynthesis (Cavalier et al., 34

2008; Zabotina et al., 2008). Although four of the 12 members of the GT48 family are coexpressed with AGPs, these are putative callose synthase genes. Even At4g36890

(IRX14), which was selected for further analysis (see Chapter 3) and coexpressed with four AGPs, was found to function in xylan biosynthesis rather than the glycosylation of

AGPs (Brown et al., 2007).

However, many of the other glycosyl transferases coexpressed with large numbers of AGPs have yet to be characterized and could still be involved in the glycosylation of

AGPs. Interestingly, the glycosyl transferase coexpressed with the second most AGPs is also a member of the GT47 family, but currently has no known function and could potentially be a good candidate for further analysis. The GT8 family is also particularly prevalent in the results with 16 of the 42 family members being coexpressed with at least one AGP. Some members of the GT8 family have been identified as putative galacturonosyltransferases (GAUT) (Caffall et al., 2009). GAUT1 is involved in the synthesis of the pectin homogalacturonan (Sterling et al., 2006), while IRX8 is hypothesized to be involved in the synthesis of the xylan reducing end structure (Pena et al., 2007; Persson et al., 2007). Notably, the GT8 family member (At1g24170) coexpressed with the most AGPs (12) is not among these and currently has no known function. Given that a member of the GT77 family (XEG113) was identified as a putative arabinosyltransferase involved in the arabinosylation of EXTs (Gille et al., 2009), the possibility that the GT77 family also contains arabinosyltransferases specific for AGPs should be considered. One member of the GT77 family, At1g19360, was coexpressed with seven different AGPs. Because some AGPs (e.g. AGP9) contain contiguous 35

hydroxyprolines (SOOO), they may also undergo arabinosylation. In this case, XEG113, which was coexpressed with three AGPs, may also be involved in the arabinosylation of

AGPs. Research suggests that a member of GT31 family may contain

galactosyltransferases specific for AGPs (Strasser et al., 2007; Qu et al., 2008). Among

the 33 members of the GT31 family, 15 were coexpressed with at least one AGP.

At4g00300, for instance, was found to be coexpressed with seven different AGPs.

While not yet shown to be successful at identifying the glycosyl transferases

involved in the glycosylation of AGPs, at the very least it was successful at identifying

those involved in cell wall biosynthesis in general. Given the complexity of the

interactions taking place in the cell wall and the relative lack of knowledge on exactly

what role each AGP is playing, perhaps it is not surprising that AGPs and the glycosyl

transferases involved in the synthesis of other cell wall components would be

coexpressed.

Selection of a Glycosyl Transferase for Further Analysis

The coexpression results were used as a starting point for the selection of a

glycosyl transferase for further analysis by a reverse genetics approach. In addition to the

number of AGPs coexpressed with a particular glycosyl transferase, a few other factors

were also considered. The number of members in a particular family was also a factor in

determining which glycosyl transferase would be selected for further analysis. For

instance, At1g24170, a member of the glycosyl transferase family 8 (GT8), was found to

be coexpressed with the 12 AGPs. However, At1g24170 could potentially prove difficult 36 to study in a reverse genetics approach due to the fact that the GT8 family contains 42 members. In a large family with 42 members, there is less likely to be an apparent phenotype when only one gene is mutated since some of the other 41 genes may be able to compensate for the loss.

One of the primary factors to consider was the availability of multiple T-DNA insertion mutants for a particular gene. For example, At1g08660, a member of the GT29 family, is coexpressed with 10 different AGPs. As such, it may seem like an ideal candidate for study because the GT29 family contains only three members, but unfortunately only one T-DNA insertion mutant (GABI_202E09) is currently available for the At1g08660 gene.

A member of the GT43 family, At4g36890, was ultimately selected for further analysis. At4g36890, also known as IRX14, was only shown to be coexpressed with four different AGPs, but At4g36890 is particularly well suited to study since the GT43 family contains only four members and because many T-DNA insertion mutants are available

(See Chapter 3). At4g36890 was thus hypothesized to function in the glycosylation of

AGPs. More specifically, because many members of the GT43 family are known glucuronosyltransferases, At4g36890 was hypothesized to function as a β-1,6- glucuronosyltransferase in the glycosylation of AGPs. However, a subsequent publication has shown evidence that At4g36890 is instead involved in glucuronoxylan biosynthesis

(Brown et al., 2007). 37

CHAPTER 3: IRX14 AND IRX14-LIKE, TWO GLYCOSYL TRANSFEREASES

INVOLVED IN GLUCURONOXYLAN BIOSYNTHESIS IN ARABIDOPSIS

Introduction

Glucuronoxylan Biosynthesis

Glucuronoxylan (GX), along with cellulose and lignin, are the major components of secondary cell walls in Arabidopsis. While the biosynthesis of cellulose and lignin are

now relatively well understood, only recently have the enzymes involved in GX

biosynthesis begun to be identified. Identifying all the enzymes involved is critical in

order to have a complete understanding of GX biosynthesis and to manipulate plants for

superior wood or biofuel production.

Several glycosyl transferases, in three different families, are believed to be

responsible for synthesis of the xylan backbone, addition of the side chains, and the

synthesis of the reducing end structure (Figure 3.1).

Figure 3.1: Structure of glucuronoxylan in dicots. (A) GX consists of a repeating β-(1,4)- xylose backbone substituted with glucuronic acid (GlcUA) and 4-O-methyl-glucuronic acid (Me-GlcUA) at approximately every eight residues (Brown et al., 2007). (B) The unique reducing end structure of GX (Pena et al., 2007).

In terms of synthesis of the xylan backbone, two GT families are implicated.

IRX9 and IRX14, both members of the GT43 family, along with IRX10 and IRX10L,

members of GT47, are hypothesized to be involved with synthesis of the backbone

(Brown et al., 2005; Brown et al., 2007; Lee et al., 2007a; Pena et al., 2007; Brown et al.,

2009; Wu et al., 2009). A significant amount of research is reported on the analysis and characterization of one member of the GT43 family, At2g37090 (IRX9). As compared to

WT plants, irx9 mutants are much smaller and cross sections of irx9 stems revealed a

severe irregular xylem phenotype (Brown et al., 2005). Carbohydrate analysis of the noncellulosic cell wall material from stems revealed a significant decrease in the amount of xylose present (Brown et al., 2005). A homolog of IRX9 in Poplar, PoGT43B, is expressed in cells undergoing secondary cell wall thickening and overexpression of

PoGT43B in irx9 plants completely restores the plants to normal size and restores the xylose content to a certain extent (Zhou et al., 2007). The glucuronoxylan chain length is decreased in irx9 plants (Pena et al., 2007) and a biochemical assay demonstrated that irx9 was deficient in xylosyl transferase activity (Lee et al., 2007a). These results have lead to the hypothesis that IRX9 is involved in the transfer of xylose in the synthesis of the glucuronoxylan backbone. During the course of this thesis research, a mutant of

IRX14 was identified to exhibit an irregular xylem phenotype in stems and a reduction of xylose in the cell wall (Brown et al., 2007). The irx14 mutant was also deficient in the xylosyl transferase activity observed in WT, suggesting it is involved with synthesis of the xylan backbone along with IRX9 (Brown et al., 2007; Brown et al., 2009). 39

Genes hypothesized to affect the reducing end structure of GX include members

of the GT47 and GT8 families. In the GT8 family, both IRX8 and PARVUS are

hypothesized to function in the synthesis of the reducing end structure, although their

precise roles are currently unknown (Lee et al., 2007b; Pena et al., 2007). FRA8 and

FRA8 Homolog (F8H), two closely related members of the GT47 family, are also

proposed to function in the synthesis of the reducing end structure (Zhong et al., 2005;

Lee et al., 2009).

In this study, the role of IRX14 in GX biosynthesis is further investigated and a novel drought tolerant phenotype for irx14 mutants is identified. In addition, a closely related gene in the GT43 family named IRX14-LIKE (IRX14L) is studied for the first time. While mutation of IRX14L results in no phenotypic differences, a double mutant of

IRX14 and IRX14L is severely affected in plant growth and development and has a further reduction in glucuronoxylan as compared to a mutation in IRX14 alone. This leads to the

conclusion that IRX14 and IRX14L are functionally redundant and that IRX14L also functions in GX biosynthesis, although to a lesser degree than IRX14.

Materials and Methods

Plant Growth Conditions

Arabidopsis plants were grown in autoclaved Fafard 52 Mix soil (Agawam, MA)

in growth chambers under long day conditions (16 hr light/8 hr dark) at 22°C and 70%

relative humidity. In some instances, plates were instead grown on Murashige and Skoog

(MS) (4.33 g/L) plates with 1% sucrose. 40

Identification of T-DNA Insertion Lines

T-DNA insertion lines were identified on the Salk Institute Genomics Analysis

Laboratory (SIgNal) website (http://signal.salk.edu/cgi-bin/tdnaexpress) for IRX14

(At4g36890) and IRX14L (At5g67230) and ordered from The Arabidopsis Biological

Resource Center (ABRC). Homozygous plants were identified by PCR with gene specific

primers (Table 3.1). Either LBa1 (5’-TGGTTCACGTAGTGGGCCATCG-3’) or LBb1.3

(5’-ATTTTGCCGATTTCGGAAC-3’) was used as the left border primer of the T-DNA

insertion.

Bioinformatics

Protein sequences for IRX14 (At4g36890), IRX14L (At5g67230), IRX9

(At2g37090), and IRX9L (At1g27600) were obtained from The Arabidopsis Information

Resource (TAIR, http://www.arabidopsis.org/). The four protein sequences were aligned

using ClustalW (http://www.ebi.ac.uk/Tools/clustalw2/index.html), and MEGA 4.0

(http://www.megasoftware.net/) was utilized to generate phylogenetic trees. Expression

profiles were obtained from Genevestigator (https://www.genevestigator.ethz.ch/).

Additionally, the protein sequences were examined for a signal peptide (SignalP,

http://www.cbs.dtu.dk/services/SignalP/, transmembrane domain (TMHMM,

http://www.cbs.dtu.dk/services/TMHMM/, and golgi localization (Golgi Predictor, http://ccb.imb.uq.edu.au/golgi/).

Table 3.1: Primers used in PCR confirmation of T-DNA mutants.

IRX14 SALK_028143 LP: AATGGAAAGCCCATGTTTCTC RP: GAGGAATGTAGACGAATTTGGG SALK_038212 LP: AACGACACGTGTACCTCCTTG RP: AACATCACAATCCCATCAAGC SALK_038215 LP: AACGACACGTGTACCTCCTTG RP: AACATCACAATCCCATCAAGC SALK_055193 LP: ATAAGTAAAATTGAGGGGCGG RP: GCTGTGGAAGACAAGTTCTGC GABI_011C12 LP: TTGCATCATGTTGGATGTGTC RP: GCTGTGGAAGACAAGTTCTGC

IRX14L SALK_080872 LP: TTCGACTCTCTTTGCAGCTTC RP: CAACGGACATATTGGAAATCG SALK_117947 LP: TGGAGCAACGATATACTTGGG RP: CGAGAGTTCTTCTTCACCGTG SALK_066961 LP: CTTGCTCTTCGACACTCTTGG RP: ATCGATGTACGGTGTGAGGAG SALK_067044 LP: CTTGCTCTTCGACACTCTTGG RP: ATCGATGTACGGTGTGAGGAG

RNA Extraction and RT-PCR

Total RNA was extracted from WT and transgenic plants using the RNeasy Plant

Mini Kit (QIAGEN, Valencia, CA). RT-PCR was performed with the OneStep RT-PCR

Kit (QIAGEN) according to the manufacturer’s instructions. The primers for IRX14 were

5’-ATGCGGAAGAGATGGTTTTG-3’ and 5’-TTCGGATTTTCGTTGGATGT-3’. The

primers for IRX14L were 5’-GCCAATCCAAGGTCCTAGTTG-3’ and 5’-

ATGATCCAGCCAGGTGGGAA-3’. ACTIN was used as an internal control with the

following primers: 5’-GTCCTCGACTCTGGAGATGGTGTG-3’ and 5’- 42

GTCGTACTACCGGTATTGTGCTCG-3’. Primers were designed to either span an intron/exon boundary to eliminate amplification of DNA or flank an intron to detect amplification of DNA.

Histology

Stems and roots were obtained from five week old WT, irx14, irx14L, irx14 irx14L(±), and irx14(±) irx14L plants. All tissues were immediately placed in SafeFix II

All Purpose Fixative (Fisher Scientific, Pittsburgh, PA). Tissues were then dehydrated through an increasing series of EtOH (30%, 50%, 60%, 70%, 80%, 90%, 95%, 100%) for two hours each. Tissues were cleared with xylene and embedded in paraffin. Ten micrometer thick sections were obtained with a rotary microtome. Sections were dried overnight and dewaxed with xylene. All sections were rehydrated and stained with toluidine blue and viewed on a Motic BA400 microscope (Motic, Xiamen, China).

Images were captured with a Moticam 2300 3.0M pixel camera using the Motic Images

Plus 2.0 software and processed with Adobe Photoshop (Adobe Systems, San Jose, CA).

Drought Stress

In order to achieve drought conditions, plants were watered normally for three weeks before water was withheld. After water was withheld for approximately two weeks, plants were again watered and photos were taken. For the leaf excision test, leaves were excised from three week old plants and weighed immediately. After 60 minutes the 43

leaves were weighed again. By comparing the initial and final weights, the amount of

water lost from the leaves over the 60 minutes was determined.

NaCl Treatment

WT and irx14 plants were placed on MS plates and grown for five days. The

seedlings were then transferred to MS plates supplemented with either 0 mM, 100 mM,

200 mM, or 300 mM NaCl. After five days of growth on the salt containing plates,

photos were taken and the effect of the NaCl on the WT and irx14 plants was recorded.

Monosaccharide Analysis

Stems were obtained from five week old WT, irx14, irx14L, irx14 irx14L(±), and

irx14(±) irx14L plants grown in soil. Stems were ground to a fine powder in liquid

nitrogen and extracted for 1 hour twice with 70% ethanol at 70°C followed by a wash in

100% acetone. Glycosyl composition analysis of the non-cellulosic cell wall fraction was

performed at the Complex Carbohydrate Research Center (CCRC) in Athens, GA.

Glycosyl composition analysis was performed by combined gas chromatography/mass

spectrometry (GC/MS) of the per-O-trimethylsilyl (TMS) derivatives of the

monosaccharide methyl glycosides produced from the sample by acidic methanolysis.

Methyl glycosides were first prepared from a dry sample by methanolysis in 1 M

HCl in methanol at 80°C (18-22 hours), followed by re-N-acetylation with pyridine and

acetic anhydride in methanol (for detection of amino sugars). The samples were then

per-O-trimethylsilylated by treatment with Tri-Sil (Pierce, Rockford, IL) at 80°C (0.5 44

hours) (York et al., 1985; Merkle and Poppe, 1994). GC/MS analysis of the TMS methyl glycosides was performed on an HP 6890 GC interfaced to a 5975b MSD, using a All

Tech EC-1 fused silica capillary column (30m x 0.25 mm ID).

Xylan Immunolocalization

Stem sections were incubated with the LM10 antibody (PlantProbes, Leeds, UK)

(1/20 dilution) for 1 hour (McCartney et al., 2005). The samples were washed in PBS and then incubated for 1 hour with an FITC conjugated anti-rat secondary antibody (1/100 dilution) (Jackson ImmunoResearch, West Grove, PA). The samples were visualized with a Zeiss LSM510 laser scanning confocal system (Carl Zeiss, Oberkochen, Germany).

Results

The GT43 Family

The GT43 family consists of four members in Arabidopsis (Table 3.2). While two of the members, IRX9 and IRX14, were previously implicated with the synthesis of glucuronoxylan, the remaining two members were not previously examined. Comparative sequence analysis of the four proteins identified that At5g67230 is a homolog of IRX14

(Figure 3.2). The other family member, At1g27600, is more similar to IRX9. An amino acid alignment of IRX14 and At5g67230 revealed that the proteins are similar in terms of amino acid sequences with an identity of 75% and a similarity of 83% (Figure 3.3).

Because of this similarity, At5g67230 was named IRX14-LIKE (IRX14L). By comparison, IRX9 and At1g27600 have an identity and similarity of only 39% and 56%, respectively (Figure 3.4). 45

Table 3.2: Overview of the four members of the GT43 family identified in Arabidopsis.

Locus ID Name Length Exons Signal Golgi (AA)a Anchor/Peptideb Localizedc At1g27600 IRX9L 394 4 No Yes At2g37090 IRX9 351 3 No Yes At4g36890 IRX14 525 3 Anchor Post-Golgi At5g67230 IRX14L 492 3 Anchor Yes a TAIR: http://www.arabidopsis.org/ b SignalP: http://www.cbs.dtu.dk/services/SignalP/ c Golgi Predictor: http://ccb.imb.uq.edu.au/golgi/

Figure 3.2: Phylogenic tree of the GT43 family. The GT43 family is divided into two distinct groups with IRX14 and IRX14L being particularly similar. Phylogeny constructed using the Maximum Parsimony method in MEGA 4.0. Bootstrap values provided as a percentage of 1000 replicates.

Figure 3.3: Amino acid alignment of IRX14 and IRX14L. Identical and similar amino acids are highlighted in black and grey, respectively. Sequences aligned using ClustalW2.

Figure 3.4: Amino acid alignment of IRX9 and IRX9L. Identical and similar amino acids are highlighted in black and grey, respectively. Sequences aligned using ClustalW2.

Expression Pattern

According to publically available microarray data from Genevestigator

(https://www.genevestigator.ethz.ch/), the IRX14 and IRX14L genes are expressed in similar organs and tissues. Specifically, IRX14 and IRX14L are expressed in virtually the same organs and tissues, but in each case IRX14L is expressed at significantly lower levels (Figure 3.5). Both IRX14 and IRX14L are expressed highest in stems and nodes with slightly lower levels of expression in the roots, hypocotyls, and siliques. Complete expression profiles of the four GT43 family members are available in Appendix A.

5000 4500 4000 3500 3000 2500 2000 1500 1000 Relative Expression Level Relative Expression 500 0

Figure 3.5: Selected expression data for IRX14 (blue) and IRX14L (red) from Genevestigator. IRX14 and IRX14L exhibit similar expression patterns, but IRX14 shows higher levels of expression.

T-DNA Insertion Mutants

T-DNA insertion mutants were selected for both IRX14 and IRX14L. Numerous

T-DNA insertion mutants are available in both genes (Table 3.3). Highest priority was given towards T-DNA insertions located in an exon as these are most likely to result in knockout mutants. Initially, the SALK_038212 line was obtained for IRX14 and

SALK_066961 for IRX14L (Figure 3.6).

Table 3.3: T-DNA insertion mutants currently available for IRX14 and IRX14L. SALK_038212 (irx14) and SALK_066961 (irx14L) were the focus of this work and are highlighted in grey.

Gene Name Mutant Line Location At4g36890 FLAG_594B12 FLAG FST 1000-Promotor (IRX14) SALK_028143 SALK T-DNA 1000-Promotor GABI_742C07 GABI-Kat FST 1000-Promotor SAIL_81_C03 SAIL FST 1000-Promotor SAIL_1151_C06 SAIL FST 300-UTR5 SALK_038215 SALK T-DNA Exon SALK_038212 SALK T-DNA Exon SALK_055284 SALK T-DNA Intron SALK_055193 SALK T-DNA Intron GABI_011C12 GABI-Kat Confirmed Exon SALK_032012 SALK T-DNA 300-UTR3 At5g67230 GABI_166B04 GABI-Kat FST 1000-Promotor (IRX14L) GABI_345H04 GABI-Kat FST 1000-Promotor GABI_322A11 GABI-Kat FST 1000-Promotor SAIL_156_D12 SAIL FST 1000-Promotor SALK_080872 SALK T-DNA HM 1000-Promotor SAIL_156_D12 SAIL FST 1000-Promotor GABI_909E10 GABI-Kat FST 1000-Promotor DX575973 Community FST 1000-Promotor SAIL_883_C02 SAIL FST 1000-Promotor SALK_117947 SALK T-DNA 300-UTR5 SAIL_579_H02 SAIL FST 300-UTR5 SAIL_579_H03 SAIL FST 300-UTR5 SAIL_598_D01 SAIL FST 300-UTR5 SAIL_598_F01 SAIL FST 300-UTR5 GABI_322A11 GABI-Kat FST Exon SK21633 SK FST Exon SALK_066961 SALK T-DNA Exon SALK_067044 SALK T-DNA Exon SK17829 SK FST 300-UTR3

Figure 3.6: T-DNA insertion mutants for IRX14 and IRX14L. The IRX14 and IRX14L genes are similar in structure consisting of three exons and two introns. The exons are represented by the thick blue lines, while introns are represented by the thinner black lines. Light blue regions indicate the 3’UTR and 5’UTR. Locations of the SALK_038212 and SALK_066961 T-DNA insertions and the left (LP), right (RP), and T-DNA (LBa1) primers used for PCR confirmation are indicated.

Plants homozygous for the T-DNA insertions were obtained for each mutant line

and confirmed by PCR (Figures 3.7 and 3.8). Using a combination of two gene specific

primers (LP and RP) and a T-DNA primer (LBa1 or LBb1.3), heterozygous and

homozygous mutants can be distinguished from WT plants. Interestingly, the irx14

mutant exhibits two bands rather than one in homozygous mutants. Further investigation by sequencing revealed that this results from the presence of (at least) two T-DNA inserts

adjacent to each other, but in the opposite orientation. As such, both the combination of

the RP and the T-DNA primer (819 bp) and LP and the T-DNA primer (678 bp) generate

a band (Figure 3.7). 51

In terms of overall growth and development, neither the irx14 nor irx14L mutants

show any gross phenotypic differences compared to WT when grown under normal

conditions (Figure 3.9).

Figure 3.7: Identification of irx14 homozygous mutants by PCR. DNA was extracted and amplified from 10 unknown plants and compared with a known WT (left). Both heterozygous (HZ) and homozygous mutant (MU) plants were indentified. For the SALK_038212 line, both the combination of LP and LBb1.3 and RP and LBb1.3 generates a band (right).

Figure 3.8: Identification of irx14L homozygous mutants by PCR. DNA was extracted and amplified from seven unknown plants and compared with a known WT. Both heterozygous (HZ) and homozygous mutant (MU) plants were indentified. 52

Figure 3.9: WT, irx14, and irx14L plants grown under normal conditions. No obvious phenotypic differences are apparent.

Comprehensive Phenotypic Analysis

Comprehensive soil and plate based phenotypic analysis was performed on both

irx14 and irx14L to identify any differences from WT (Boyes et al., 2001). In terms of the plate based analysis, root length and number of lateral roots were examined over the course of the first two weeks of growth. For the plate based analysis, the day at which various developmental milestones occur was recorded, including the day the plants have four rosette leaves, 10 rosette leaves, and first flowering. Additionally, the height of the inflorescence stem and size of rosette perimeter were recorded at day 21. At day 32, the

number of branches off the main stem and the number of stems from base were recorded.

Based on the soil based analysis, the irx14 and irx14L mutants grow slightly slower,

reaching certain milestones significantly later than WT. (Table 3.4). For the plate based 53

analysis, one difference was observed with respect to root length. By day 14, irx14 roots

were shorter than WT roots (Figure 3.10).

Table 3.4: Plate based phenotypic analysis of the irx14 and irx14L mutants. Significant differences from wild type plants: * P < 0.05, ** P < 0.01.

4 Rosette Leaves (d) 10 Rosette Leaves (d) Flowering (d) WT 8.07 ± 1.03 17.40 ± 0.83 26.07 ± 2.71 irx14 9.23 ± 0.83** 18.54 ± 0.88** 27.85 ± 3.29 irx14L 9.15 ± 0.38** 18.54 ± 0.88** 30.38 ± 3.07**

Day 21 Day 34 Inflorescence Rosette Branches off Stems from Height (mm) Perimeter (mm) Inflorescence Base WT 4.80 ± 2.51 154.47 ± 25.85 2.93 ± 0.59 3.73 ± 1.22 irx14 4.00 ± 3.11 127.62 ± 24.71** 2.62 ± 0.51 2.31 ± 1.11** irx14L 3.38 ± 3.48 132.69 ± 12.17* 2.92 ± 0.49 3.46 ± 1.13

60.0

50.0 * 40.0

30.0

20.0 Root Length (mm) 10.0

0.0 WT irx14 irx14L

Figure 3.10. Root length of WT, irx14, and irx14L at Day 14. Bars indicate standard deviation (n=20). Significant differences from wild type plants: * P < 0.05.

irx14 irx14L Double Mutant

Because of the similarities in the sequences and expression patterns of IRX14 and

IRX14L, it was hypothesized that both genes function in the same process. As such, the irx14 and irx14L homozygous mutants were crossed in order to obtain an irx14 irx14L double mutant. Although both the irx14 and irx14L single mutants appeared identical to

WT, the irx14 irx14L double mutant had severely delayed growth and failed to produce an inflorescence stem (Figure 3.11). Over the course of the first week of growth, the irx14 irx14L mutant was indistinguishable from WT. However, growth of irx14 irx14L essentially ceased thereafter. In addition, plants homozygous for the mutation of IRX14 and heterozygous for IRX14L [irx14 irx14L(±)] and plants heterozygous for IRX14 and homozygous for IRX14L [irx14(±) irx14L] were examined (Figure 3.12). The irx14(±) irx14L plants also appeared identical to WT plants, whereas the irx14 irx14L(±) plants had an intermediate phenotype. The irx14 irx14L(±) plants were smaller overall with noticeably smaller leaves, stems, and siliques (Figure 3.13). The siliques of the irx14 irx14L(±) plants never fully developed, were reduced in size, and rarely contained any viable seeds.

Figure 3.11: Comparison of the WT and irx14 irx14L double mutant rosettes. The irx14 irx14L mutants never grew any larger and were significantly smaller than their WT counterparts.

Figure 3.12: WT, irx14(±) irx14L, irx14 irx14L(±), and irx14 irx14L plants grown under normal conditions. While irx14(±) irx14L were identical to WT, irx14 irx14L(±) plants exhibited an intermediate phenotype.

Figure 3.13: Phenotypic differences of the irx14 irx14L(±) plants. The mutants had an intermediate phenotype with notably smaller leaves (top) and siliques (bottom).

RT-PCR Analysis

Reverse transcriptase-PCR (RT-PCR) revealed a reduction of the IRX14 and

IRX14L transcripts in the mutants, but the transcript is not completely knocked out in either (Figure 3.14). However, because the RNA transcript contains the T-DNA insert, the RNA and any subsequent protein that may or may not be produced should be nonfunctional.

RT-PCR can also demonstrate whether or not compensation occurs between the

IRX14 and IRX14L genes. In other words, is the expression of IRX14 increased in the irx14L mutant or is the expression of IRX14L increased in the irx14 mutant?

Compensation was predicted to occur between the two genes due to the sequence similarity, but compensation is not evident from the RT-PCR results (Figure 3.14). 57

However, quantitative PCR (Q-PCR) would be necessary to conclusively demonstrate

whether or not compensation between the two genes is occurring.

Figure 3.14: RT-PCR analysis of the irx14, irx14L, irx14 irx14L(±), and irx14(±) irx14L mutants. RT-PCR provides confirmation that the levels of IRX14 and IRX14L transcripts are reduced in the mutants. Actin was used as a control.

Irregular Xylem

The irx14 mutant was previously reported to have an irregular xylem phenotype in stems (Brown et al., 2007). Stem sections from irx14 plants were obtained to confirm this phenotype (Figure 3.15). In addition to stems, sections of roots were also obtained as these were not previously examined. Both irx14 stems and roots contain an obvious irregularity of the xylem and the cell wall was noticeably thinner compared to WT. The 58

irx14L mutants were also examined for the irregular xylem phenotype, but the xylem in

this mutant was indistinguishable from wild type plants. Because of the lack of the irx14

irx14L stems available for sectioning, the xylem in the stems of the irx14 irx14L(±) and irx14(±) irx14L were examined (Figure 3.16). While the xylem of irx14(±) irx14L was

comparable to wild type, the xylem of the irx14 irx14L(±) mutants is considerably more

irregular than the irx14 single mutant, particularly in the roots. Many of the xylem vessel

elements in the roots of the irx14 irx14L(±) mutant were nearly completely collapsed and

were almost unrecognizable compared to wild type. 59

Figure 3.15: Sections of stems (A-C) and roots (D-F) from WT (A, D), irx14 (B, E), irx14L (C, F) plants. Only the irx14 mutants exhibit an irregular xylem phenotype in stems and roots. Representative irregular xylem vessel elements are indicated by arrows. Scale bars equal 50μm.

Figure 3.16: Sections of stems (A-C) and roots (D-F) from WT (A, D), irx14(±) irx14L (B, E), irx14 irx14L(±) (C, F) plants. Only the irx14 irx14L(±) mutants exhibit an irregular xylem phenotype in stems and roots. Representative irregular xylem vessel elements are indicated by arrows. Scale bars equal 50μm.

Drought Tolerance

The most interesting and potentially beneficial phenotypic difference observed with irx14 plants was drought tolerance. Water was withheld from three week old irx14 and WT plants for two weeks. In this initial experiment, all the WT plants died while all irx14 plants survived (Figure 3.17). A leaf excision test was also performed on both irx14 and irx14L mutants in order to further quantify the drought tolerance. Leaves were removed from each mutant, weighed immediately, and weighed again after 60 minutes.

Any weight lost was considered to be due to water lost from the leaf. In the leaf excision test, leaves from irx14 plants lost approximately 15% less water than WT over the course of 60 minutes. Water loss from irx14L plants was consistent with WT, indicating that irx14L plants did not share the drought tolerant phenotype (Figure 3.18).

A mutation of IRX1 (lew2) was also previously reported to exhibit both irregular xylem and drought tolerant phenotypes (Chen et al., 2005). Expression of various stress related genes were upregulated in lew2 mutants as compared to wild type. Expression of one such gene, SDR1 (Short-Chain Dehydrogenase/Reductase), was examined in irx14 mutants, but no difference was observed in expression compared to wild type (Figure

3.19)

Figure 3.17: The drought tolerant phenotype of the irx14 mutants. Three week old plants were subjected to drought conditions for two weeks. Representative WT (left) and irx14 (right) plants are shown. 63

Figure 3.18: Water loss from leaves excised from WT, irx14, and irx14L plants. Leaves measured over the course of 60 minutes. Bars indicate standard error (n=5). Significant differences from wild type plants: * P < 0.01

Figure 3.19: Expression of SDR1 in WT and irx14 plants. No difference in expression was observed. ACTIN was used as a control.

Salt Tolerance

Because of the observed differences in response to drought conditions, irx14 plants were also examined for salt tolerance. Plants were grown on MS plates supplemented with varying concentrations of NaCl (Figure 3.20). However, no differences between the wild type and irx14 mutants were observed. Both plants survived at 100mM NaCl while all died at 300mM. Only at 200 mM were varying survival rates observed, but these appeared similar for both WT and irx14.

Figure 3.20: WT and irx14 mutants grown on various concentrations of NaCl. No differences between the WT and irx14 mutants were observed. 65

Carbohydrate Analysis

Carbohydrate analysis was performed on the noncellulosic portion of total cell

wall extracted from WT, irx14, irx14L, irx14 irx14L(±), and irx14(±) irx14L stems

(Figure 3.21). In the irx14 mutant, the amount of xylose decreased by 24% from wild

type plants, while the irx14L mutants showed no reduction. The irx14 irx14L(±) plants showed a further reduction in xylose, a total reduction of 42% from wild type plants. The irx14(±) irx14L mutants had a monosaccharide composition virtually identical to that of wild type. Unfortunately, no stems were available from irx14 irx14L plants for a comparable analysis. Although the amount of GalUA appeared to increased in the irx14 and irx14 irx14L(±) plants, the amounts are based on the relative percentage of each. In this case, as the percentage of xylose decreased, the percentage of GalUA increased even though the absolute amount of GalUA in the cell wall was not necessarily increased.

Figure 3.21: Non-cellulosic sugar composition of the cell walls from WT, irx14, irx14L, irx14(±) irx14L, and irx14 irx14L(±) stems. The irx14 and irx14 irx14L(±) stems exhibited a substantial reduction in xylose, while no changes were apparent in the cell wall fractions of irx14L or irx14(±) irx14L stems. 66

Xylan Immunolocalization

In order to provide further support for the role of IRX14 and IRX14L in GX

biosynthesis, the LM10 monoclonal antibody to β-(1,4)-D-Xylan was utilized for the

immunolocalization of xylan in stem sections (Figure 3.22). The LM10 antibody can

recognize unsubstituted and relatively low-substituted xylans in several species, but has

no cross-reactivity with wheat arabinoxylan (McCartney et al., 2005). The LM11

antibody was also available; this antibody can accommodate more extensive substitution

of a xylan backbone and binds strongly to wheat arabinoxylan. However, because

previous reports indicate similar or identical results for LM10 and LM11, only LM10 was

utilized here (Wu et al., 2009). While irx14 mutants showed only a slight reduction in the

intensity of fluorescence, the irx14 irx14L(±) mutant showed a further reduction in

fluorescence indicating a substantial loss of GX in this mutant. No differences from WT

were apparent for either the irx14L or irx14(±) irx14L mutants (not shown).

Figure 3.22: Xylan immunolocalization in stem sections of WT, irx14, and irx14 irx14L(±) plants. While a slight reduction in fluorescence intensity was visible for irx14, a further reduction was visible for irx14 irx14L(±). Scale bars equal 50 µm.

Allelic T-DNA Mutants

Because T-DNA mutants can contain inserts in genes other than the one of interest, it becomes necessary to perform additional experiments to ensure that any

phenotypic differences are actually due to the disruption of the intended gene. The mutant

can be complemented with the gene to ensure that any phenotypic differences are

restored to wild type. Otherwise, multiple T-DNA mutants can be obtained in the same

gene. All of the mutants exhibiting similar phenotypic differences provide strong

evidence that the particular gene of interest is actually responsible. In order to confirm

the phenotypic differences observed in the irx14 (SALK_038212) and irx14L

(SALK_066961) mutants, additional T-DNA mutants were also obtained.

For a second mutant in IRX14, the GABI_0112C01 line was obtained. The

GABI_0112C01 line contains a T-DNA insert in the third and final exon (Figure 3.23).

Although the presence of the insert was confirmed by PCR, sectioning of the stems

revealed that the irregular xylem phenotype is absent (Figure 3.24). As such, this mutant

was not pursued further. Other lines examined in IRX14 include SALK_028143,

SALK_038215, and SALK_055193. Both SALK_038215 and SALK_055193 were

examined by PCR, but strangely no heterozygous or homozygous mutants were found. A

homozygous mutant was obtained for SALK_028143. As this insert is located in the

promoter, RT-PCR was performed to confirm the gene is knocked out. Unfortunately,

levels of the IRX14 transcript in the mutant were identical to WT. Although no other T-

DNA mutants were identified, previous complementation of the irx14 mutant restored the

irregular xylem phenotype to WT (Brown et al. 2007). 68

Figure 3.23: Location of the GABI_0112C01 T-DNA insert. The insert is located in the third and final exon, but did not result in a phenotype consistent with irx14.

Figure 3.24: Structure of the xylem in the stems (A) and roots (B) of GABI_0112C01 mutants. Unfortunately, this mutant did not exhibit the irregular xylem phenotype.

Similar problems occurred in the identification of a second T-DNA mutant for

IRX14L. Screening of SALK_067044 revealed no heterozygous or homozygous mutants, while SALK_080872 is an insert in the promoter which resulted in no reduction or loss of the IRX14L transcript. However, plants homozygous for the SALK_117947 insert in the

5’UTR were identified (Figure 3.25A). Although the insert did not result in a knockout mutant, the levels of the IRX14L transcript were reduced as compared to WT (Figure

3.25B). 69

Figure 3.25: Characterization of the SALK_117947 mutant line. (A) Homozygous mutants (MU) were identified by PCR. (B) RT-PCR analysis revealed a reduction in the level of the IRX14L transcript for the mutant.

Discussion

Understanding the Drought Tolerance Phenotype

Although the drought phenotype seemed promising at first, clear understanding of

this phenotype has proven difficult. Further examination revealed the presence of the

phenotype to be quite variable, with it appearing in some instances and being completely

absent in others. Various factors seem to have an effect on the phenotype including the

age of the plants, relative humidity, and how much the plants were watered prior to the

drought treatment. The leaf excision test is more consistent, but in a few instances no difference in water loss was found between the irx14 mutant and wild type plants.

One possibility to consider would be a role of stomata and guard cells in the

drought tolerance phenotype. F8H, an enzyme also involved in GX biosynthesis, is 70

expressed in guard cells based on GUS staining (Lee et al., 2009). This indicates that GX

is likely to be important for cell wall synthesis in guard cells. Treatment of guard cells in

Commelina communis with xylanase has previously been shown to disrupt guard cells, ultimately resulting in cell lysis (Jones et al., 2003). However, with C. communis being a monocot, xylan likely makes up a much larger percentage of the primary cell wall.

Certainly, a reduction of GX in the cell wall of the guard cell caused by a mutation in

IRX14 could affect stomatal opening and closing and result in drought tolerance.

However, leaf peels of both WT and irx14 mutants had no obvious differences in the

stomata (data not shown).

A similar drought tolerant phenotype as was observed for irx14 was previously reported in a mutation of the CesA8/IRX1 (lew2), a glycosyl transferase involved in cellulose synthesis (Chen et al., 2005). The drought tolerance of this mutant was attributed to an increased accumulation in osmolytes, such as sugar, proline, and ABA, in the mutant due to the irregular xylem even when the plant is grown under normal (not stressed) conditions (Chen et al., 2005). In essence, the lew2 mutants start preparing for drought conditions even when they were watered normally. Our findings that irx14 is also drought tolerant prompts the question of whether all irregular xylem mutants are drought tolerant or if this phenotype is unique to lew2 and irx14. For the lew2 mutant, expression of various stress related genes were upregulated as compared to wild type, but this is not the case with irx14. Expression of SDR1 was examined in irx14 mutants, but no difference in expression was observed. Additionally, lew2 mutants were not only shown to be resistant to drought stress, but also to osmotic stresses, such as NaCl, mannitol, and 71

glycerol (Chen et al., 2005). The irx14 mutants were grown on varying concentrations of

NaCl, but no enhanced salt tolerance compared to WT was observed. These results

suggest that the drought tolerance seen in irx14 may be due to a different mechanism than

observed in the lew2 mutants. Another possibility is that the irregular xylem phenotype in

irx14 does result in increased stress, but perhaps only under certain conditions. For

instance, expression of the SDR1 gene was only checked in plants that were consistently

well watered. However, the possibility exists that irx14 mutants do become more stressed

than WT, but only under slightly more adverse conditions. If this is the case, the

phenotype may be dependent on the plants’ experiencing a slight drought over an extended period of time rather than normal watering followed by a single drought event.

If the phenotype is dependent on how well the plants were taken care of prior to the test, this would also explain why the leaf excision test showed a difference in water loss between irx14 and WT appeared in some instances, but not others.

IRX14 and IRX14L Function Redundantly in Glucuronoxylan Biosynthesis

Although the irx14L mutant shows no phenotype on its own, a double mutant of irx14 and irx14L is severely delayed in growth and never produces an inflorescence stem under normal growth conditions. The intermediate phenotype of the irx14 irx14L(±) mutant indicates that IRX14L is also likely to be involved in glucuronoxylan biosynthesis and functionally redundant with IRX14. In every instance examined, the irx14 irx14L(±) mutant exhibits a more extreme phenotype as compared to the irx14 single mutant. The irx14 irx14L(±) plants exhibit an increased irregularity of the xylem and a further 72 decrease of xylose in stems. These results provide additional support that both IRX14 and IRX14L function in GX biosynthesis.

Because an irregular xylem phenotype is visible with irx14, but not irx14L, it appears that IRX14 plays a more important role in GX biosynthesis than IRX14L. This is consistent with the higher levels of expression in every tissue for IRX14 compared to

IRX14L. Nonetheless, IRX14L is obviously playing some role or at least is capable of compensating for a loss of irx14 considering the dramatic phenotype seen in an irx14 irx14L double mutant. Particularly fascinating are the irx14(±) irx14L mutants. Despite the fact that IRX14L is completely knocked out and IRX14 is partially knocked out, these plants are still identical to WT in every aspect examined.

The phenotype of irx14 and irx14L is similar to the previously reported relationship between IRX10 and IRX10L (Brown et al., 2009; Wu et al., 2009). In both cases, mutation of one gene yields a comparatively mild irregular xylem phenotype, but a mutation in the other results in no difference from wild type. However, a double mutant of the two genes is severely affected with respect to growth and development. Similar to

IRX14, IRX10 and IRX10L are proposed to function in the synthesis of the glucuronoxylan backbone (Brown et al., 2009; Wu et al., 2009). More recently, a similar relationship was shown for FRA8 and its homolog named F8H (Lee et al., 2009). A fra8 mutant results in an irregular xylem phenotype, while no differences are observed in the f8h mutant; however, a fra8 f8h double mutant exhibits a much more severe phenotype.

Because FRA8/F8H and IRX10/IRX10L appear to share a similar relationship with

IRX14/IRX14L, these genes may be expressed similarly in terms of compensation. 73

Expression of F8H in the fra8 mutant, IRX10 in the irx10L mutant, and IRX10L in the irx10 mutant were all examined by RT-PCR, but there is no mention of compensation occurring in any of these cases (Brown et al., 2009; Lee et al., 2009; Wu et al., 2009).

CHAPTER 4: DISCUSSION, CONCLUSIONS, AND FUTURE WORK

Complexity of GX Biosynthesis

The small size of the irx14 irx14L double mutant demonstrates the importance of

GX in plant growth and development. GX is obviously vital to maintain the strength of the secondary cell wall, which is particularly important for the xylem vessel elements due to the pressure of transpiration. As with other mutants with irregular xylem phenotypes, the small size of these plants can likely be attributed to impeding water flow through the plant, preventing the plant from reaching full size.

Currently, IRX9, IRX14, IRX14L, IRX10, and IRX10L are all suggested to be involved the synthesis of the β-(1,4)-xylan backbone. It is currently unclear why so many enzymes are required to synthesize a relatively simple structure and whether these different enzymes are interacting with one another in a complex. The IRX9 protein was expressed in yeast and examined for xylosyltransferases activity, but none was detected

(Pena et al., 2007). However, microsomes isolated from irx9, irx14, and irx10 irx10L mutants were shown to be deficient in xylosyltransferase activity compared to WT microsomes (Brown et al., 2007; Lee et al., 2007a). An attempt was also made to express

FRA8, but again no activity was detected (Zhong et al., 2005). As such, none of the enzymes currently thought to be involved in GX biosynthesis are confirmed biochemically. The lack of a biochemical activity for these enzymes has lead to the suggestion that multiple enzymes in a complex may be required in order to function properly (Brown et al., 2007; Pena et al., 2007; Brown et al., 2009). The large number of 75

enzymes now identified for synthesis of the backbone alone tends to support this

hypothesis. Not only must one consider the enzymes thought to be involved in the synthesis of the backbone, but because the pathway for GX biosynthesis is currently

unknown, the enzymes thought to be involved in the addition of side chains and in the

synthesis of the reducing end structure must all be considered together. One possibility would be to utilize a coexpression approach to identify the proteins that are potentially in a complex with IRX14 and IRX14L. Searching the coexpression database for IRX14 reveals that both IRX10 and IRX10L are coexpressed (data not shown). However, IRX9,

IRX9L, FRA8, F8H, and PARVUS were not found to be coexpressed with IRX14. In addition, many other glycosyl transferases are also identified that are not likely to be in the complex, such as CESA2, CESA4, and GSL06. Clearly more research is needed to determine exactly how all of the enzymes function and how they may interact with one another.

Future Work

Although this work lays the foundation of understanding IRX14 and IRX14L, much more work could be conducted to give a more complete understanding on their role in GX biosynthesis. The role of IRX9L has yet to be investigated, and whether it shares a similar relationship with IRX9 as was observed with IRX14 and IRX14L is unknown. If this is the case, irx9L likely exhibits no phenotype, but an irx9 irx9L double mutant would have a severe phenotype as seen with irx14 irx14L. Perhaps more interesting would be an irx9 irx14 double mutant as these genes seem to be more critical than 76

IRX14L and IRX9L. In order to understand the synthesis of the xylan backbone, crosses

between mutants of the GT43 and GT47 families will likely be required. For instance,

both an irx14 irx10 and irx9 irx10 double mutant would provide interesting insights into

the mechanism of synthesis of the xylan backbone.

Another open question is whether or not these enzymes exist in a complex.

Although it seems possible that IRX14 and IRX14L exist in a complex, further

experiments are certainly required for confirmation. Bimolecular fluorescence

complementation (BiFC) will be a particularly useful tool to investigate this question.

Particularly interesting will be to investigate if IRX10 and IRX10L are part of a complex with IRX14, IRX9, IRX14L, and IRX9L, despite the fact that they are members of an

entirely different family. Even more interesting will be to investigate if those enzymes

thought to be involved in xylan backbone synthesis interact with those thought to be

involved in the synthesis of the reducing end structure. For instance, is there an

interaction between IRX9 and FRA8?

Genes homologous to IRX14 are identified in important crop plants such as rice,

wheat, corn, and tomato (Figure 4.1). Interestingly, genes homologous to IRX14 and

IRX14L are found in both dicots and monocots, which is consistent with their role in the

synthesis of the xylan backbone because of its importance in both. Future work could be

conducted to see if a mutation in any of these homologous genes would also result in

irregular xylem or a drought tolerant phenotype in particular.

Although, no biochemical activity is reported for any of the glycosyl transferases

currently thought to be involved in GX biosynthesis, observing such activity will 77 ultimately be required to definitively identify the function of a particular enzyme. More than likely, numerous glycosyl transferases are required to work together to function properly and a single glycosyl transferase may have no activity individually. In any case, this work helps to lay a foundation for these future studies by identifying additional glycosyl transferases, IRX14 and IRX14L, and their roles in GX biosynthesis and plant physiology.

Figure 4.1: Phylogenetic analysis of GT43 family members in Arabidopsis thaliana (At), Oryza Sativa (Os), Triticum aestivum (Ta), and Zea mays (Zm). Genes homologous to IRX14/IRX14L are highlighted in yellow. Phylogeny constructed using the Maximum Parsimony method in MEGA 4.0. Bootstrap values provided as a percentage of 1000 replicates. The names of the genes in Triticum aestivum and Zea mays are based only on the order in which they appear in the CAZy database; GenBank accession numbers are provided in parentheses. Only full length sequences were included.

REFERENCES

Andersson S., Samuelson O., Ishihara M., and Shimizu K. (1983). Structure of the

reducing end-groups in spruce xylan. Carbohydr. Res. 111, 283-288.

Baron-Epel O., Gharyal P.K., and Schindler M. (1988). Pectins as mediators of wall porosity in soybean cells. Planta 175, 389-395.

Boudet A.M., Kajita S., Grima-Pettenati J., and Goffner D. (2003). Lignins and lignocellulosics: a better control of synthesis for new and improved uses. Trends Plant

Sci. 8, 576-581.

Boyes D.C., Zayed A.M., Ascenzi R., McCaskill A.J., Hoffman N.E., Davis K.R., and

Gorlach J. (2001). Growth stage-based phenotypic analysis of Arabidopsis: A model for high throughput functional genomics in plants. Plant Cell 13, 1499-1510.

Brett C.T. (2000). Cellulose microfibrils in plants: biosynthesis, deposition, and integration into the cell wall. Int. Rev. Cytol. 199, 161-199.

Brown D.M., Zeef L.A.H., Ellis J., Goodacre R., and Turner S.R. (2005).

Identification of novel genes in Arabidopsis involved in secondary cell wall formation using expression profiling and reverse genetics. Plant Cell 17, 2281-2295.

Brown D.M., Goubet F., Vicky W.W.A., Goodacre R., Stephens E., Dupree P., and

Turner S.R. (2007). Comparison of five xylan synthesis mutants reveals new insight into the mechanisms of xylan synthesis. Plant J. 52, 1154-1168. 80

Brown D.M., Zhang Z., Stephens E., Dupree P., and Turner S.R. (2009).

Characterization of IRX10 and IRX10-like reveals an essential role in glucuronoxylan

biosynthesis in Arabidopsis. Plant J. 57, 732-746.

Caffall K.H., Pattathil S., Phillips S.E., Hahn M.G., and Mohnen D. (2009).

Arabidopsis thaliana T-DNA mutants implicate GAUT genes in the biosynthesis of pectin and xylan in cell walls and seed testa. Mol Plant 2, 1000-1014.

Carpita N.C. and Gibeaut D.M. (1993). Structural models of primary cell walls in flowering plants: consistency of molecular structure with the physical properties of the

walls during growth. 3, 1-30.

Carpita N.C. and McCann M. (2000). The Cell Wall In Biochemistry & Molecular

Biology of Plants, 1st ed., B. B. Buchanan, W. Gruissem and R. L. Jones, eds (Rockville,

MD: American Society of Plant Physiologists) pp. 52-108.

Cavalier D.M., Lerouxel O., Neumetzler L., Yamauchi K., Reinecke A., Freshour

G., Zabotina O.A., Hahn M.G., Burgert I., Pauly M., Raikhel N.V., and Keegstra K.

(2008). Disrupting two Arabidopsis thaliana xylosyltransferase genes results in plants

deficient in xyloglucan, a major primary cell wall component. Plant Cell 20, 1519-1537.

Chen Z.Z., Hong X.H., Zhang H.R., Wang Y.Q., Li X., Zhu J.K., and Gong Z.Z.

(2005). Disruption of the cellulose synthase gene, AtCesA8/IRX1, enhances drought and

osmotic stress tolerance in Arabidopsis. Plant J. 43, 273-283. 81

Delmer D.P. (1999). Cellulose Biosyntehsis: Exciting times for a difficult field of study.

Annu. Rev. Plant Physiol. Plant Mol. Biol. 50, 245-276.

Ebringerová A. and Heinze T. (2000). Xylan and xylan derivatives - biopolymers with valuable properties, 1. Naturally occurring xylans structures, isolation procedures and properties. Macromolecular Rapid Communications 21, 542-556.

Fry S.C. (1988). The Growing Plant Cell Wall: Chemical and Metabolic Analysis, 1st ed.

(Longman Scientific & Technical) pp. 333.

Gille S., Hänsel U., Ziemann M., and Pauly M. (2009). Identification of plant cell wall mutants by means of a forward chemical genetic approach using hydrolases. Proc. Natl.

Acad. Sci. USA 106, 14699-14704.

Gray K.A., Zhao L., and Emptage M. (2006). Bioethanol. Curr. Opin. Chem. Biol. 10,

141-146.

Harris P.J. and Smith B.G. (2006). Plant cell walls and cell-wall polysaccharides: structures, properties and uses in food products. 41, 129-143.

Herth W. (1983). Arrays of plasma membrane “rosettes” in cellulose microfibril formation of Spirogyra. Planta 159, 347-356.

Hieta R. and Myllyharju J. (2002). Cloning and characterization of a low molecular weight prolyl 4-hydroxylase from Arabidopsis thaliana. J. Biol. Chem. 277, 23965-

23971. 82

Johansson M.H. and Samuelson O. (1977). Reducing end groups in brich xylan and their alkaline degradation. Wood Sci. Technol. 11, 251-263.

Jones L., Milne J.L., Ashford D., and McQueen-Mason S.J. (2003). Cell wall arabinan

is essential for guard cell function. Proc. Natl. Acad. Sci. USA 100, 11783-11788.

Kieliszewski M.J. (2001). The latest hype on Hyp-O-glycosylation codes.

Phytochemistry 57, 319-323.

Lee C., O'Neill M.A., Tsumuraya Y., Darvill A.G., and Ye Z. (2007a). The irregular

xylem9 mutant is deficient in xylan xylosyltransferase activity. Plant Cell Physiol. 48,

1624-1634.

Lee C., Zhong R., Richardson E.A., Himmelsbach D.S., McPhail B.T., and Ye Z.

(2007b). The PARVUS gene is expressed in cells undergoing secondary wall thickening

and is essential for glucuronoxylan biosynthesis. Plant Cell Physiol. 48, 1659-1672.

Lee C., Teng Q., Huang W., Zhong R., and Ye Z. (2009). The F8H glycosyltransferase is a functional paralog of FRA8 involved in glucuronoxylan biosynthesis in Arabidopsis.

Plant Cell Physiol. 50, 812-827.

McCartney L., Marcus S.E., and Knox J.P. (2005). Monoclonal antibodies to plant cell

wall xylans and arabinoxylans. J. Histochem. Cytochem. 53, 543-546.

Merkle R. and Poppe I. (1994). Carbohydrate composition analysis of glycoconjugates

by gas-liquid chromatography/mass spectrometry. Meth. Enzymol. 230, 1-15. 83

Nothnagel E.A. (1997). Proteoglycans and related components in plant cells. Int. Rev.

Cytol. 174, 195-291.

Pena M.J., Zhong R., Zhou G., Richardson E.A., O'Neill M.A., Darvill A.G., York

W.S., and Ye Z. (2007). Arabidopsis irregular xylem8 and irregular xylem9:

Implications for the complexity of glucuronoxylan biosynthesis. Plant Cell 19, 549-563.

Persson S., Caffall K.H., Freshour G., Hilley M.T., Bauer S., Poindexter P., Hahn

M.G., Mohnen D., and Somerville C. (2007). The Arabidopsis irregular xylem8 Mutant

Is Deficient in Glucuronoxylan and Homogalacturonan, Which Are Essential for

Secondary Cell Wall Integrity. Plant Cell 19, 237-255.

Qu Y., Egelund J., Gilson P.R., Houghton F., Gleeson P.A., Schultz C.J., and Bacic

A. (2008). Identification of a novel group of putative Arabidopsis thaliana beta-(1,3)- galactosyltransferases. Plant Mol. Biol. 68, 43-59.

Showalter A.M. (1993). Structure and function of plant cell wall proteins. Plant Cell 5,

9-23.

Somerville C. (2006). Cellulose Synthesis in Higher Plants. Annu. Rev. Cell Dev. Biol.

22, 53-78.

Steinhauser D., Usadel B., Luedemann A., Thimm O., and Kopka J. (2004).

CSB.DB: a comprehensive systems-biology database. Bioinformatics 20, 3647-3651. 84

Sterling J.D., Atmodjo M.A., Inwood S.E., Kumar Kolli V.S., Quigley H.F., Hahn

M.G., and Mohnen D. (2006). Functional identification of an Arabidopsis pectin biosynthetic homogalacturonan galacturonosyltransferase. Proc. Natl. Acad. Sci. USA

103, 5236-5241.

Strasser R., Bondili J.S., Vavra U., Schoberer J., Svoboda B., Glossl J., Leonard R.,

Stadlmann J., Altmann F., Steinkellner H., and Mach L. (2007). A unique β-1,3- galactosyltransferase is indispensable for the biosynthesis of N-glycans containing lewis a structures in Arabidopsis thaliana. Plant Cell 19, 2278-2292.

Tan L., Qiu F., Lamport D.T.A., and Kieliszewski M.J. (2004). Structure of a hydroxyproline (Hyp)-arabinogalactan polysaccharide from repetitive Ala-Hyp expressed in transgenic Nicotiana tabacum. J. Biol. Chem. 279, 13156-13165.

Wu A., Rihouey C., Seveno M., Hörnblad E., Singh S.K., Matsunaga T., Ishii T.,

Lerouge P., and Marchant A. (2009). The Arabidopsis IRX10 and IRX10-LIKE glycosyltransferases are critical for glucuronoxylan biosynthesis during secondary cell wall formation. Plant J. 57, 718-731.

Yokoyama R. and Nishitani K. (2004). Genomic basis for cell-wall diversity in plants.

A comparative approach to gene families in rice and Arabidopsis. Plant Cell Physiol. 45,

1111-1121. 85

York W.S., Darvill A.G., McNeil M., Stevenson T.T., and Albersheim P. (1985).

Isolation and characterization of plant cell walls and cell wall components. Meth.

Enzymol. 118, 1-54.

Zabotina O.A., van de Ven W.T.G., Freshour G., Drakakaki G., Cavalier D.,

Mouille G., Hahn M.G., Keegstra K., and Raikhel N.V. (2008). Arabidopsis XXT5 gene encodes a putative α-1,6-xylosyltransferase that is involved in xyloglucan

biosynthesis. 56, 101-115.

Zhong R. and Ye Z.H. (2009) Secondary cell walls. Encyclopedia of Life Sciences,

DOI:10.1002/9780470015902.a0021256.

Zhong R., Pena M.J., Zhou G., Nairn C.J., Wood-Jones A., Richardson E.A.,

Morrison W.H.,III, Darvill A.G., York W.S., and Ye Z. (2005). Arabidopsis fragile

fiber8, which encodes a putative glucuronyltransferase, is essential for normal secondary

wall synthesis. Plant Cell 17, 3390-3408.

Zhou G., Zhong R., Himmelsbach D.S., McPhail B.T., and Ye Z. (2007). Molecular

characterization of PoGT8D and PoGT43B, two secondary wall-associated

glycosyltransferases in poplar. Plant Cell Physiol. 48, 689-699.

APPENDIX A: GENEVESTIGATOR EXPRESSION FOR THE GT43 FAMILY

Figure A.1: Expression profile of At1g27600 (IRX9L) 87

Figure A.2: Expression profile of At2g37090 (IRX9)

Figure A.3: Expression profile of At4g36890 (IRX14) 89

Figure A.4: Expression profile of At5g67230 (IRX14L) 90

APPENDIX B: SCANNING ELECTRON MICROSCOPY OF IRX14 AND

IRX14 IRX14L MUTANTS

Although irx14 irx14L double mutants are extremely small and never produce an inflorescence stem under normal conditions, the irx14 irx14L plants do grow slightly larger and produce a small inflorescence stem when grown at 100% humidity. Plants were grown on MS plates for two weeks before being transferred to soil. Once in soil, the plants were covered in plastic wrap to maintain a 100% humidity environment. These stems, along with WT and irx14 stems of plants grown under normal conditions were examined by scanning electron microscopy.

Entire stems were cut from WT, irx14, and irx14 irx14L plants and frozen in liquid nitrogen. Stems were placed in super chilled 100% EtOH in 50ml Falcon tubes and stored in a -20°C freezer overnight. The stems in ethanol were then shipped overnight on dry ice to Daniel Mullendore at Washington State University for further processing.

The ethanol-submerged stems were allowed to warm to room temperature and the ethanol was replaced with distilled water and allowed to rest for at least two hours. Ten, one mm serial sections were made above and below the original wound with a new region of a double-edged razor blade for each slice. The sections were lyophilized for 12 hours using a Virtis Lyophilizer (The Virtis Co., Inc, Gardiner, NY). Stems were then mounted on to aluminum specimen mounts (TED Pella Inc. Redding, CA) with carbon

Pelco tabs (TED Pella Inc., Redding, CA) and approximately 200Å of Au was applied 91 using a Technics Hummer V (Anatech) sputter coater (San Jose, CA). Sections were imaged with high vacuum mode at 30KV accelerating voltage.

Figure B.1: Cross section of a WT stem.

Figure B.2: Normal xylem vessel elements of a WT stem.

Figure B.3: Cross section of an irx14 stem. 94

Figure B.4: Irregular xylem vessel elements of an irx14 stem.

Figure B.5: Cross section of an irx14 irx14L double mutant stem.

Figure B.6: Irregular and collapsed cells of the irx14 irx14L stem.