Identification and Characterization of Five Arabidopsis

Galactosyltransferases and Their Functional Roles in -

Glycosylation, Growth, Development, and Cellular Signaling

A dissertation presented to

the faculty of

the College of Arts and Sciences of Ohio University

In partial fulfillment

of the requirements for the degree

Doctor of Philosophy

Debarati Basu

August 2015

© 2015 Debarati Basu. All Rights Reserved.

2

This dissertation titled

Identification and Characterization of Five Arabidopsis Hydroxyproline

Galactosyltransferases and Their Functional Roles in Arabinogalactan-Protein

Glycosylation, Growth, Development, and Cellular Signaling

by

DEBARATI BASU

has been approved for

the Department of Environmental and Biology

and the College of Arts and Sciences by

Allan M. Showalter

Professor of Environmental and Plant Biology

Robert Frank

Dean, College of Arts and Sciences

3

ABSTRACT

BASU, DEBARATI, Ph.D., August 2015, Molecular and Cellular Biology

Identification and Characterization of Five Arabidopsis Hydroxyproline

Galactosyltransferases and Their Functional Roles in Arabinogalactan-Protein

Glycosylation, Growth, Development, and Cellular Signaling

Director of Dissertation: Allan M. Showalter

Arabinogalactan- (AGPs) are a class of highly glycosylated ubiquitous plant , implicated in diverse biological roles. Although contain substantial amounts of AGPs, enzymes responsible for glycosylation of AGPs are largely unknown. Given that sugar side chains account for about 90% of the molecular mass of AGPs, sugars likely define the interactive surface of the molecule and hence its function. Bioinformatic analysis indicated that hydroxyproline (Hyp)-O- galactosyltransferases (GALTs) for AGPs might belong to the -Active enZYme (CAZY) 31-family, which includes β-GALTs involved in the synthesis of glycoproteins. Here, five GT 31 Hyp-O-GALTs, namely AtGALT2

(At4g21060), AtGALT3 (At3g06440), AtGALT4 (At1g27120), AtGALT5 (At1g74800) and AtGALT6 (At5g62620), were characterized. Detergent permeabilized microsomes obtained from Pichia pastoris expressing AtGALT2 and AtGALT5 and from tobacco epidermal cells expressing AtGALT2, AtGALT3, AtGALT4, AtGALT5 and AtGALT6 specifically catalyzed the transfer of [14C]Gal from UDP-[14C]Gal to Hyp in the chemically synthesized AGP peptide substrate acceptor, [AO]7. Furthermore, these Hyp-

O-GALTs exhibited similar biochemical properties. Confocal microscopic analysis of

4 fluorescently tagged AtGALT2-6 indicated AtGALT2 was localized in the ER and

Golgi, while the other four proteins were localized exclusively in Golgi vesicles.

Additional support that these five GALTs encode AGP-specific Hyp-O-GALTs was provided by analysis of allelic knockout mutants of the five GALT genes. These mutants demonstrated significantly lower Hyp-O-GALT activities, reduced β-Yariv-precipitated

AGPs and pleiotropic growth and developmental phenotypes compared to wild type plants with increasing severity in galt2galt5 double mutants. To varying degrees, all five

Hyp-O-GALTs were essential for tip growth and involved in development in response to salt stress. Interestingly, the galt2galt5 double mutant phenocopied the root swelling phenotypes as well as the coat and cellulose-deficient phenotypes of previously characterized mutants, namely sos5 (a GPI-anchored fasciclin- like AGP), fei1fei2 (a pair of leucine rich cell wall receptor-like kinases), and sos5fei1fei2. These findings indicated that the arabinogalactan (AG) polysaccharides of

SOS5 are likely critical for cellular signaling and in stimulating cellulose biosynthesis.

In summary, this dissertation contributes to an understanding of AGP biosynthesis, particularly with respect to Hyp-galactosylation, and to the functional roles which AG polysaccharides play in plants.

5

ACKNOWLEDGMENTS

First and foremost, I am indebted to my supervisor, Dr. Allan Showalter, for granting me the opportunity to pursue my PhD, for his continued support, encouragement, financial assistance and guidance throughout my PhD study. I am very grateful for his patience, availability to discuss my coursework and thesis research, and to help answer the many questions that arose throughout my graduate studies. Most importantly he always encouraged me to develop independent scientific thinking and research skills. I am honored to have had the opportunity to learn from him and to grow as a scientist under his guidance.

I am grateful to all my committee members for their invaluable suggestions and criticism relevant to my thesis. They were instrumental in developing my professional and research career with regular inputs and critical suggestions. I am thankful to them for allowing me to use their instruments and reagents for my research.

I am very grateful to the past and present members of the Showalter lab who have made my time here such an enjoyable experience including Brian Keppler, Yan Liang,

Wuda Wang, and Lu Tian.

I am thankful to all the brilliant, enthusiastic, dedicated, diligent undergraduates for assisting me in the research including Shauni Bobbs, Hannah Fritts, Tayler DeBrosse,

Emily Poirier, Eric Sokup, Kirk Emch, Megan Moore, Siyi Ma, Hayley Herock, Andrew

Travers and Kiara Dillard.

I sincerely appreciate the help of Connie Pollard and Jamie Dewey for their incessant co-ordination and unconditional help.

6

I am thankful to Vijay Nadella for his valuable suggestions on QPCR analysis. I also thank Dr. Micheal Held and Jeffrey Thuma for their insightful suggestions on confocal microscopy, especially for image acquisition and analysis.

I am thankful all my friends including Claudia Lechler, Bailey Hunter, Hayley

Shurr, Svetlana Bondareva, Conny Bartholmes, Proma Basu, Mohor Chatterjee, Sutapa

Ghosh, Enakshi Ghosh, Laura Cristea, Aditi Vyas, Nilesh Khade, Aditya Kulkarni, and friends from the International Student Union for making my stay at Athens an enjoyable and memorable experience.

I am also thankful to the Department of Environmental and Plant Biology, the

Molecular and Cellular Biology Program, and the Ohio University Graduate College for providing me with financial assistance. I am also thankful to the Graduate Student

Senate for providing me with funds to attend national scientific meetings.

I also express my gratitude towards all my instructors in the University of Calcutta and Thakurpukur College, especially Prof. Sumita Jha, Moumita Banerjee Prof. Anita

Mukherjee, Dr. Asim Bhadra, Dr. Silanjan Bhattacharya, Dr. Meenakshi Mukherjee, Dr.

Kuntal Narayan Chaudhury and Dr. Sutapa Kumar who inspired me to embark upon my journey for a Ph.D.

Lastly, none of this would have been possible without the unending love, encouragement, and support of my family.

7

TABLE OF CONTENTS

Page

Abstract ...... 3

Acknowledgments ...... 5

List of Tables ...... 17

List of Figures ...... 20

List of Abbreviations ...... 26

Chapter 1: Introduction ...... 28

1.1. Plant Cell Walls ...... 28

1.1.1. Overview ...... 28

1.1.2. Cell wall composition ...... 29

1.1.3. Biosynthesis of cell wall polysaccharides ...... 30

1.1.4. Plant cell wall proteins ...... 41

1.1.5. Structure of AGPs ...... 46

1.1.6. Biosynthesis of AGPs ...... 48

1.1.7. Function of AGPs ...... 54

1.2. Objectives ...... 61

1.3. Organization of the Dissertation ...... 62

Chapter 2: Functional Identification of a Hydroxyproline-O-Galactosyltransferase

Specific for Arabinogalactan-Protein Biosynthesis in Arabidopsis ...... 64

2.1. Abstract ...... 64

2.2. Introduction ...... 65

8

2.3. Results ...... 68

2.3.1. Identification of putative AGP GALTs in by in-silico

analysis ...... 68

2.3.2. Heterologous expression of putative Hyp-GALT genes in Pichia cells ...... 71

2.3.3. Heterologously expressed AtGALT2 demonstrates Hyp-GALT activity ...... 71

2.3.4. [AO]7 and d[AO]51 are substrate acceptors for AtGALT2 ...... 72

2.3.5. Product characterization by acid and base hydrolysis shows that AtGALT2

transfers Gal to Hyp residues ...... 75

2.3.6. AtGALT2 is specific for AGPs ...... 77

2.3.7. Biochemical Characteristics of the AtGALT2 Enzyme ...... 78

2.3.8. AtGALT2 mutants have lower GALT activity and reduced β-Yariv

precipitable AGPs ...... 80

2.3.9. AtGALT2 is likely localized to the endomembrane system ...... 82

2.3.10. Computational modeling of AtGALT2 predicts UDP-sugar binding ...... 86

2.4. Discussion ...... 88

2.4.1. An unrooted phylogenetic analysis of animal and plant GT31 members

revealed three distinct cluster ...... 92

2.5. Materials and Methods ...... 96

2.5.1. Identification of putative GALTs involved in AGP biosynthesis ...... 96

2.5.2. Cloning and expression of AtGALTs in Pichia pastoris ...... 97

2.5.3. Preparation of Pichia microsomes and immunoblot analysis ...... 99

9

2.5.4. Galactosyltranferase Assay with microsomal preparations from Pichia

expressing AtGALT2 ...... 100

2.5.5. Purification of Hyp-GALT2 reaction products by Reverse-Phase HPLC ... 101

2.5.6. Analysis of the Hyp-[14C]galactoside profile by gel Permeation chromato-

graphy and HPAEC ...... 101

2.5.7. Monosaccharide composition analysis of GALT reaction products by High-

Performance Anion-Exchange Chromatography (HPAEC) ...... 102

2.5.8. Determination of substrate specificity of the AtGALT2 enzyme activity ... 102

2.5.9. Biochemical characterization of AtGALT2 enzyme activity ...... 103

2.5.10. AtGALT2 mutant analysis ...... 104

2.5.11. Transient expression and subcellular localization of AtGALT2 in Nicotiana

tabacum ...... 106

Chapter 3: Two Hydroxyproline Galactosyltransferases, GALT5 and GALT2, Function in Arabinogalactan-Protein Glycosylation, Growth and Development in Arabidopsis .. 108

3.1. Abstract ...... 108

3.2. Introduction ...... 109

3.3. Results ...... 112

3.3.1. At1g74800 (GALT5) encodes a putative galactosyltransferase ...... 112

3.3.2. Heterologous expression of GALT5 in Pichia cells ...... 115

3.3.3. Heterologously expressed GALT5 demonstrates Hyp-GALT activity ...... 117

3.3.4. Characterization of the GALT5 assay products by reverse-phase HPLC

analysis ...... 117

10

3.3.5. Product characterization by acid and base hydrolysis indicates GALT5

transfers Gal to Hyp residues ...... 120

3.3.6. GALT5 is specific for AGPs ...... 124

3.3.7. Biochemical characteristics of the GALT5 enzyme ...... 125

3.3.8. GALT5 is localized to the Golgi ...... 127

3.3.9. Isolation of T-DNA insertion alleles for the GALT2 and GALT5 genes ...... 128

3.3.10. GALT2 and GALT5 have overlapping but distinct expression patterns ..... 131

3.3.11. Compensatory mechanism of GALT2 and GALT5 ...... 133

3.3.12. Biochemical phenotypes of the mutants: GALT activity, β-Yariv-

precipitable AGPs and immunolabeling with AGP specific monoclonal

antibodies ...... 136

3.3.13. Pleiotropic growth and development phenotypes of the mutants ...... 141

3.3.14. Mutants demonstrate reduced inhibition of tube growth and root

growth in the presence of β- ...... 148

3.3.15 Mutant seed germination and root growth are hypersensitive to NaCl ...... 149

3.3.16. Double mutant (galt2galt5) displays less seed coat mucilage ...... 155

3.4. Discussion ...... 157

3.4.1. GALT5 is an AGP Hyp-GALT and other AGP GTs ...... 157

3.4.1. GALT5 is localized to Golgi vesicles ...... 163

3.4.2. AGP glycosylation required for normal growth and development: GALT and

AGP glycosyltransferase mutant phenotypes ...... 163

3.4.3. AGP glycosylation required for root and tip growth ...... 165

11

3.4.4 GALTs and cellular signaling ...... 167

3.5 Concluding remarks ...... 168

3.6. Materials and Methods ...... 169

3.6.1. In silico analysis of GALT5 and GALT2 ...... 169

3.6.2. Heterologous expression of GALT5 in Pichia pastoris ...... 169

3.6.3. Preparation of Pichia microsomes expressing GALT5 and immunoblot

analysis ...... 170

3.6.4. Galactosyltranferase assay with microsomal preparations from Pichia

expressing GALT5 ...... 171

3.6.5. Purification of Hyp-GALT5 reaction products by reverse-phase HPLC ..... 171

3.6.6. Analysis of the Hyp-[14C]galactoside profile by gel permeation chromato-

graphy and high performance anion-exchange chromatography (HPAEC) ...... 171

3.6.7. Monosaccharide composition analysis of GALT reaction products by high

performance anion-exchange chromatography ...... 172

3.6.8. Determination of substrate specificity of the GALT5 enzyme activity ...... 172

3.6.9. Biochemical characterization of GALT5 enzyme activity ...... 173

3.6.10. Transient expression and subcellular localization of GALT5 in Nicotiana

tabacum leaves ...... 174

3.6.11. Plant material and genetic analysis ...... 177

3.6.12. Isolation of Golgi-enriched plant microsomal membranes ...... 178

3.6.13. Extraction of AGPs and AGP profiling by HPLC ...... 179

3.6.14. In vitro pollen germination assay ...... 180

12

3.6.15. Germination assays ...... 180

3.6.16. Root growth measurements ...... 181

3.6.17. Aberrant root hair morphology ...... 181

3.6.18. Seed staining and visualization ...... 182

3.6.19. AGP specific monoclonal antibodies ...... 182

3.6.20. Immunofluorescence detection of AGPs in root hairs, pollen tubes

and ...... 183

Chapter 4: A Small Multigene Hydroxyproline-O-Galactosyltransferase family

Functions in Arabinogalactan-Protein Glycosylation, Growth and Development in Arabidopsis ...... 185

4.1. Abstract ...... 185

4.2. Introduction ...... 186

4.3. Results ...... 188

4.3.1. In silico analysis of GALT1, GALT3, GALT4 and GALT6 ...... 188

4.3.2. Transiently expressed six Hyp-O-GALTs in has AGP

specific Hyp-O-GALT activity ...... 197

4.3.3 Substrate specificities of GALT2, GALT3, GALT4, GALT5 and GALT6 .. 199

4.3.4. Expression profile of the nine Hyp-O-GALTs ...... 201

4.3.5. GALT3, GALT4 and GALT6 are targeted to the Golgi vesicles ...... 206

4.3.6. Loss-of-function mutations in GALT3, GALT4 and GALT6 showed AGP

specific biochemical defects ...... 208

4.3.7. Disruption in GALT3, GALT4 and GALT6 exhibits root hair defects ...... 213

13

4.3.8. Disruption in GALT4 and GALT6 leads to reduction in seed number ...... 215

4.3.9. Knockout mutation of GALT3 and GALT6 results in reduced staining of

adherent mucilage ...... 219

4.3.10 Disruption of GALT6 led to early senescence ...... 224

4.3.11. GALT3, GALT4, and GALT6 mutants exhibit and root growth

which is less sensitive to β-Yariv reagent ...... 225

4.3.12. Conditional salt hypersensitive phenotype in galt mutants ...... 229

4.4. Discussion ...... 235

4.4.1. GALT2-6 encodes Hyp-GALTs for AGPs and are widely expressed in

Arabidopsis ...... 235

4.4.2. GALT3, GALT4, and GALT6 are localized to Golgi vesicles ...... 238

4.4.3. GALT mutant phenotypes reveal functional roles of AGP glycosylation in

normal growth and development ...... 238

4.4.4. Conditional phenotypes indicate GALT3, GALT4, and GALT6 function in tip

growth ...... 242

4.5. Materials and Methods ...... 246

4.5.1. In silico analysis of six Hyp-O-GALTs ...... 246

4.5.2. Plant lines and plant growth conditions ...... 247

4.5.3 Mutant confirmation by PCR and RT-PCR ...... 247

4.5.4. Heterologous expression of GALT1, GALT3, GALT4 and GALT6 and assay

for AGP galactosyltransferase activity ...... 252

4.5.5. Fluorescent protein fusion and subcellular localization ...... 253

14

4.5.6. Galactosyltranferase assay with microsomal preparations from transiently

expressed GALT1, GALT3, GALT4, GALT6 in tobacco epidermal cells ...... 254

4.5.7. Purification of Hyp-GALT5 reaction products by reverse-phase HPLC ..... 254

4.5.8. Determination of substrate specificity of the GALT1, GALT3, GALT4 and

GALT6 enzyme activity ...... 254

4.5.9. Characterization of GALT2-GALT6 [AO]7:GALT activity ...... 254

4.5.10. Isolation of Golgi-enriched plant microsomal membranes ...... 255

4.5.11. Extraction of AGPs ...... 255

4.5.12. Evaluation of seed set ...... 256

4.5.13. Root growth measurements ...... 256

4.5.14. In Vitro pollen germination assay ...... 257

4.5.15. Aberrant root hair morphology ...... 257

4.5.16. Cytochemical staining of seeds and determination of adherent mucilage size

and mass ...... 257

4.5.17. Measurement of chlorophyll content ...... 258

Chapter 5: Glycosylated SALT-OVERLY SENSITIVE5 Mediates Root Growth and

Seed Coat Mucilage Adherence via FEI/FEI2 kinase pathway ...... 259

5.1. Abstract ...... 259

5.2 Introduction ...... 260

5.3. Results ...... 263

5.3.1. Expression patterns of GALT2, GALT5, SOS5 and FEI1, FEI2 ...... 263

5.3.2. Quintuple mutants display cell expansion defects ...... 270

15

5.3.3. The quintuple mutant is defective in cellulose biosynthesis ...... 278

5.3.4. Role of ACC in AGP-FEI1/FEI2 mediated cell expansion ...... 283

5.3.5. Altered seed coat pectin mucilage organization in fei2, sos5, galt2galt5 and

quintuple mutant ...... 286

5.3.6. Quintuple mutants display reduced cellulosic rays in the adherent layer of

seed mucilage ...... 291

5.4. Discussion ...... 296

5.4.1. GALT2, GALT5, SOS5, FEI1, FEI2 act in a single genetic signaling

pathway ...... 296

5.4.2. SOS5-FEI1/FEI2 pathway is required for anisotropic growth in the root ... 298

5.4.2. Role of ethylene and ABA in GALT2/GALT5-SOS5/FEIs Pathway ...... 300

5.4.3. Effect of glycosylated SOS5-/FEIs pathway on cellulose biosynthesis ...... 301

5.5. Materials and Methods ...... 302

5.5.1. Plant material ...... 302

5.5.2. Plant growth conditions ...... 303

5.5.3. qRT-PCR analysis ...... 303

5.5.4. RT-PCR ...... 303

5.5.5. Cellulose synthesis assays ...... 306

5.5.6. Germination assay ...... 307

5.5.7. Crystalline cellulose determination ...... 307

5.5.8. Cell wall preparation ...... 308

5.5.9. Seed staining and visualization ...... 308

16

5.5.10. Phenotypic analysis of the mutants in response to salt ...... 308

5.5.11. Phloroglucinol Staining ...... 308

Chapter 6: General Conclusions ...... 310

References ...... 326

17

LIST OF TABLES

Page

Table 1.1 Comparison of polymer contents (%w/w) between Type I and type II primary cell walls ...... 30

Table 1.2 GTs involved in cellulose biosynthesis and their corresponding mutant ...... 33

Table 1.3 GTs involved in hemicellulose biosynthesis and their corresponding mutants 35

Table 1.4 GTs involved in pectin biosynthesis and their corresponding mutants ...... 38

Table 1.5 Information on the enzymes, genes, and mutants for extensin glycosylation ... 44

Table 1.6 Information on the prolyl-4-hydroxylase (P4H) genes and their mutants ...... 50

Table 1.7 Information on the enzymes, genes, and mutants for AGP glycosylation ...... 53

Table 1.8 List of well characterized AGP mutants and their respective phenotypes ...... 60

Table 2.1 GALT activity and the amount of β-Yariv-precipitated AGPs in wild type and galt2 mutant plants ...... 82

Table 2.2 Overview of the sequence properties of AtGALT1-6, the six putative

Arabidopsis galactosyltransferases ...... 95

Table 2.3 Primers used for cloning AtGALT1, 3, 4, 5 and 6 in Pichia ...... 99

Table 2.4 Primers used for genotyping and RT-PCR analysis of AtGALT2 ...... 105

Table 3.1 GALT activity and amount of β-Yariv precipitated AGPs in WT, galt2, galt5, and galt2galt5 mutants ...... 137

Table 3.2 Comparisons of various developmental phenotypes displayed by WT, galt2, galt5, and galt2galt5 mutant plants ...... 142

18

Table 3.3 Information on the known enzymes, genes, and mutants for AGP glycosylation ...... 161

Table 4.1 Nomenclature, locus identifiers, site of T-DNA insertion, length of open reading frames, predicted number of amino acids ...... 191

Table 4.2 Amino acid identity/similarity between the predicted amino acid sequences of the six AGP-specific GALT candidates involved in initial galactosylation of AGPs belonging to GT31 family by MATCHER (http://mobyle.pasteur.fr/cgi- bin/portal.py?#jobs::matcher) ...... 192

Table 4.3 Coexpression analysis of candidate genes involved in AGP biosynthesis in

Arabidopsis using the Gene CAT coexpression tool ...... 193

Table 4.4 Subcellular distribution of AGP specific GALT activity obtained from GALT1-

GALT6 transiently expressed in N. tabacum ...... 198

Table 4.5 GALT activity and amount of β-Yariv precipitated AGPs in WT and galt mutants ...... 211

Table 4.6 Weight, length, and seed number from WT and galt siliques ...... 218

Table 4.7 Determination of adherent mucilage mass and size in WT and galt ...... 222

Table 4.8 Quantification of total sugars from WT and galt mucilage sequentially extracted using ammonium oxalate, 0.2 N NaOH and 2 N NaOH ...... 223

Table 4.9 List of AGP specific GTs and well characterized AGPs ...... 244

Table 4.10 List of primers used in this study ...... 248

Table 5.1 List of candidate genes coexpressed with FEI1 as query genes using the Gene

CAT coexpression tool ...... 265

19

Table 5.2 Quantification of total sugars from WT, galt2-1, galt5-1, galt2galt5, sos5, sos5fei1fei2, and quintuple mucilage sequentially extracted using ammonium oxalate, 0.2

N NaOH, and 2 N NaOH ...... 291

Table 5.3 Primers used in this study ...... 304

20

LIST OF FIGURES

Page

Figure 1.1 Schematic representation of the different classes of AGPs deduced from their corresponding DNA sequence...... 48

Figure 2.1 Bioinformatics analysis of putative Arabidopsis β-galactosyltransferases

(GALTs) from CAZy GT family GT31...... 70

Figure 2.2 Screening for AtGALT2 expressed in Pichia cell lines. [AO]7-dependent

GALT activity tests of the 10 transgenic Pichia cell lines using Triton X-100 permeablized microsomal membranes...... 72

Figure 2.3 RP-HPLC fractionation of the [AO]7:AtGALT2 and d[AO]51:AtGALT2 reaction products on a PRP-1 reverse-phase column...... 74

Figure 2.4 Monosaccharide analysis of the RP-HPLC purified [AO]7:GALT reaction product following acid hydrolysis...... 75

Figure 2.5 Bio-gel P2 fractionation of the RP-HPLC purified [AO]7:AtGALT2 reaction product and High-Performance Anion-Exchange Chromatography (HPAEC) of the resulting base hydrolysis product...... 76

Figure 2.6 Effect of various peptide and polysaccharide acceptor substrates on incorporation of [14C]radiolabeled ...... 78

Figure 2.7 Biochemical characterization of the [AO]7:AtGALT2 activity...... 80

Figure 2.8 Characterization of the galt2 mutants...... 81

Figure 2.9 Subcellular localization of AtGALT2 in tobacco epidermal cells observed after 5 days of infiltration...... 84

21

Figure 2.10 A time course of the subcellular localization of AtGALT2 in tobacco leaf epidermal cells observed after 2, 3 and 4 days post infiltration...... 85

Figure 2.11 Predicted structural models of AtGALT2 generated by I-TASSER and Phyre

2...... 87

Figure 2.12 A 3D structural model of AtGALT2 as predicted by the COFACTOR server available in the I-TASSER program...... 88

Figure 3.1 Domain organization and sequence alignment of the GALT5 and GALT2 proteins...... 114

Figure 3.2 Screening for the presence of 6x His-tagged GALT5 and GALT activity .... 116

Figure 3.3 RP-HPLC fractionation of the [AO]7:GALT5 reaction products on a PRP1 reverse-phase column...... 119

Figure 3.4 Monosaccharide analysis of the RP-HPLC purified [AO]7:GALT5 reaction product following acid hydrolysis...... 121

Figure 3.5 Bio-gel P2 fractionation of the RP-HPLC purified [AO]7:GALT5 reaction product and High-Performance Anion-Exchange Chromatography (HPAEC) of the resulting base hydrolysis product...... 123

Figure 3.6 Effect of various peptide and polysaccharide acceptor substrates on incorporation of [14C]radiolabeled galactose...... 125

Figure 3.7 Biochemical characterization of the [AO]7:GALT5 activity...... 126

Figure 3.8 Subcellular localization of GALT5 in tobacco leaf epidermal cells observed 5 days after infiltration...... 128

Figure 3.9 Molecular characterization of galt single and double mutants...... 130

22

Figure 3.10 Expression profiles of GALT2 and GALT5 in publicly available databases.133

Figure 3.11 Organ-specific expression of GALT2 and GALT5 and gene compensation in galt2, galt5 and galt2galt5 mutants observed by qRT-PCR...... 135

Figure 3.12 Profiles of AGPs extracted from WT, galt2, galt5 and galt2galt5 mutants and separated by RP-HPLC...... 138

Figure 3.13 Immunofluorescent labeling of galt2galt5 and wild-type root hairs, pollen tubes and seeds using AGP specific monoclonal antibodies JIM4, JIM8, JIM13 and

MAC207...... 140

Figure 3.14 Morphological phenotypes of WT, galt2, galt5 and galt2galt5 mutants. .... 143

Figure 3.15 Root hair length and density reduced in the galt2galt5 double mutant...... 144

Figure 3.16 Disruption of tip growth in pollen tubes of galt2, galt5 and galt2galt5 mutants...... 145

Figure 3.17 The galt single and double mutants demonstrate reduced inhibition of pollen tube growth in response to β-Yariv reagent...... 146

Figure 3.18 Reduced inhibition of primary root growth of galt2, galt5 and galt2galt5 mutants in the presence of β-Yariv reagent...... 147

Figure 3.19 Delayed germination of galt2, galt5 and galt2galt5 seeds in the presence of

NaCl by using radicle length as an indication of delayed germination...... 151

Figure 3.20 Sensitivity of galt2 and galt5 seedlings to various salt and osmotic stresses as measured by root growth...... 152

Figure 3.21 Salt induced inhibition of primary root elongation in galt2, galt5 and galt2galt5 mutants...... 153

23

Figure 3.22 Root-Bending assay of wild type, galt, sos5, and fei mutant seedlings...... 154

Figure 3.23 Conditional root anisotropic growth defects of galt, sos5, and fei mutants. 155

Figure 3.24 Staining of seed coat mucilage for cellulose and pectin in wild type, galt, sos5, and fei mutant seeds...... 156

Figure 3.25 Sites of action of known GTs acting on AGPs are depicted within a representative AGP glycomodule sequence found within an AGP molecule...... 160

Figure 4.1 Sequence analysis of nine GALT proteins...... 195

Figure 4.2 HCA analysis of eight GALTs showing the DxD motif within a pocket of hydrophobic amino acids...... 196

Figure 4.3 Hyp-O-galactosyltransferase Activity of GALT1, GALT3, GALT4 and

GALT6 Transiently Expressed in N. tabacum...... 199

Figure 4.4 Substrate specificity of GALT 3 and 6 was monitored using potential substrates for GALT assay...... 201

Figure 4.5 Expression pattern of the six candidate GALT genes...... 203

Figure 4.6 Gene expression profile of nine GALTs in different organs/tissues...... 204

Figure 4.7 Transcript levels of the eight GALTs in the developing seed coat depicted by http://seedgenenetwork.net/ (by Le et al., 2010)...... 205

Figure 4.8 Subcellular localization of transiently expressed GALT3, GALT4 and

GALT6-YFP in N. benthamiana...... 207

Figure 4.9 Schematic gene models and transcript analysis of GALT1, GALT3,

GALT4 and GALT6 and T-DNA mutants...... 210

Figure 4.10 HPLC profile of AGPs obtained from wild type and single galt mutants. .. 212

24

Figure 4.11 Root hair length and density reduced in the galt3, galt4 and galt6 mutant.214

Figure 4.12 Morphology of the siliques from plants after crossing galt4 × wild type, galt6

× wild type compared with wild type and homozygous mutant galt4 and galt6...... 216

Figure 4.13 Determination of pollen viability using Alexander's staining solution...... 217

Figure 4.14 Staining of seed coat mucilage for pectin in wild type and the single and double galt mutants...... 221

Figure 4.15 Age-dependent leaf senescence phenotype of WT and galt6-1 and galt6-2 plants...... 225

Figure 4.16 The galt single mutants demonstrate reduced inhibition of pollen tube growth in response to β-Yariv reagent...... 227

Figure 4.17 Reduced inhibition of primary root growth of galt3, galt4 and galt6 mutants in the presence of β-Yariv reagent...... 228

Figure 4.18 Salt induced inhibition of primary root elongation in galt3, galt4 and galt6 mutants...... 231

Figure 4.19 Root elongation in response to 100 mM NaCl ...... 232

Figure 4.20 Root-bending assay of wild type, galt1, galt3, galt4 and galt6 mutant seedlings...... 233

Figure 4.21 Conditional root anisotropic growth defects of single and double galt mutants compared to the wild type plants...... 234

Figure 5.1 Transcript profiling of GALT2, GALT5, FEI1, and FEI2 throughout the different developmental stages in Arabidopsis as depicted by PlaNet...... 267

Figure 5.2 Expression analysis of indicated genes during the course of seed development.268

25

Figure 5.3 Transcript profiling of FEI1, FEI2, SOS5, GALT2 and GALT5 genes from different organs and developmental stages and in response to salinity stress...... 269

Figure 5.4 RT-PCR analysis of the quintuple mutants to confirm null status...... 272

Figure 5.5 Quintuple mutants exhibits reduced root elongation in response to elevated

NaCl...... 273

Figure 5.6 Quintuple mutant displays conditional root anisotropic growth defect...... 274

Figure 5.7 Hypocotyl phenotype of the quintuple mutants...... 276

Figure 5.8 Salt hypersensitivity assessed by root bending assay...... 277

Figure 5.9 Mutants display hypersensitivity towards isoxaben...... 279

Figure 5.10 Phloroglucinol staining to detect lignin...... 280

Figure 5.11 The quintuple mutant displays cellulose deficiency under elevated salt stress.282

Figure 5.12 Role of ethylene and ABA on conditional root phenotype...... 285

Figure 5.13 Staining of seeds with ruthenium red with no shaking, demonstrating both adherent and nonadherent layers of seed mucilage...... 288

Figure 5.14 Aberrant adherent mucilage structure as depicted by ruthenium red staining.289

Figure 5.15 Effect of cationic chelator on seed mucilage phenotype...... 290

Figure 5.16 Reduced cellulosic rays observed in mutant seed coat mucilage...... 293

Figure 5.17 Impaired cellulosic rays in the mutants...... 294

Figure 5.18 Seedling establishment percentage of the mutants belonging to the indicated genotypes on media containing increasing concentration of polyethylene glycol (PEG).295

Figure 5.19 Proposed model linking GALT2 and GALT5 with SOS5/FEI1/FEI2 in cellular signaling of root growth...... 297

26

LIST OF ABBREVIATIONS

[AO]7: [-hydroxyproline]7

AG: arabinogalactan

AGP: arabinogalactan-protein

Ala: alanine

Arabidopsis: Arabidopsis thaliana

CAZy: the Carbohydrate Active enZymes (CAZy) database d[AO]51: deglycosylated [alanine-hydroxyproline]51

DP: degree of polymerization

EDTA: ethylenediaminetetraacetic acid

ER:

EXT: extensin

ExtP: extensin peptide

FITC: fluorescein isothiocyanate

FLA: fasciclin-like

AGPFUT: fucosyltransferase

Gal: galactose

Galectin domain: galactose-binding lectin domain

GalNAcT: acetylgalactosaminyltransferases

GAUT: galacturonosyltransferase

GAX: glucuronoarabinoxylan

GFP: green fluorescent protein

27

GlcA: glucuronyltransferase

GPI: glycosylphosphatidylinositol

GT: glycosyltransferase

HF: hydrogen fluoride

HG: homogalacturonan

His: histidine

HPAEC: High pH Anion Exchange Chromatography

HRGP: Hydroxyproline rich-glycoproteins

Hyp: hydroxyproline mAb:

N-glycan: N-linked oligosaccharide

NMR: magnetic resonance

P4H: 4-hydroxylase

Pfam: protein family database

PMSF: phenylmethylsulphonyl fluoride

PRP: proline rich protein

RG-I: rhamnogalacturonan I

RG-II: rhamnogalacturonan I

Rha: rhamnoseRP-HPLC: reverse phase-high performance liquid chromatography

ST: sialyltransferase

TFA: trifluoroacetic acid

β-Yariv reagent: (β-D-galactosyl)3 or (β-D-glucosyl)3 Yariv phenylglycosides

28

CHAPTER 1: INTRODUCTION

1.1. Plant Cell Walls

1.1.1. Overview

One of most striking features that distinguish a plant cell from an animal cell is the cell wall. The cell wall is a highly heterogeneous and complex structure with remarkable properties, which combine extreme tensile strength with extensibility (Wolf et al., 2012).

Plant walls are important not only because they are integral to plant growth and development, but also because they determine the quality of plant-based products. The plant cell wall is of enormous economic importance to humans. Walls serve as a source of natural textile fibers, paper and wood products, and components of fine chemicals and medicinal products. Recently, deconstruction of the the plant cell wall is a major focus in research, as walls, contain lignocellulosic material which can be used for the production of sustainable bioethanol (Sarkar et al., 2009). Thus, understanding how plants synthesize and remodel their cell walls is an important and expanding area of research, particularly in the light of renewable energy.

It is clear that cell walls are involved in diverse functions in plants, so the biosynthesis and differentiation of these walls has to be tightly regulated spatially, temporally and developmentally. As much as 10% of the Arabidopsis genome is estimated to be involved in wall biosynthesis, modification and degradation (Carpita,

2001). Clearly, a better understanding of the structure and function of walls depends on the availability of detailed knowledge regarding their biosynthesis.

29

Based on their morphology and chemistry, plant cell walls can be classified into two broad types: primary cell walls, which are formed by growing cells; and secondary cell walls, which are formed in certain cells after cell expansion has slowed down and are often lignified. Generally, the role of the primary cell wall is to sustain cell expansion, driven by turgor pressure, by resisting tensile forces. About 20-30% of the primary cell wall dry weight is composed of cellulose. In contrast, secondary cell walls are synthesized adjacent to the primary cell walls and create thickened specialized structures in different cell types (i.e., vessels or fibers), due to deposition of lignin for resisting compressive forces (Lee et al., 2011). Secondary cell walls are known to contain less pectin than primary cell walls (Harholt et al., 2010). The detailed mechanism involving in the synthesis of secondary cell wall is not yet clearly elucidated (Taylor et al., 1999).

However, recent progress has been made in understanding transcriptional network that regulates secondary cell wall biosynthesis (Zhong et al., 2008; Dubos et al., 2010; Zhao et al., 2010). The most prominent families of transcription factors involved in this network appear to be the NAC (NAM/ATAF/CUC) and R2R3-type MYB

(MYELOBLASTOSIS) family proteins (Zhong and Ye, 2015).

1.1.2. Cell wall composition

The primary cell wall is a robust structure consisting of partially redundant but interdependent polysaccharide and protein networks (Wolf et al., 2012). Carpita and

Gibeaut (1993) coined the terms “Type I” and “Type II” primary cell wall based on the differences between grass species and all other angiosperms with respect to polymer structure, architecture, and biosynthesis (Table 1.1). Type I walls, which are found in

30 dicots and non-commelinoid monocots, contain about equal amounts of cellulose and cross-linking xyloglucans (XyGs), with various minor amounts of arabinoxylans, glucomannans, and galacto-glucomannans. On the other hand, graminaceous monocots have xylans and mixed linked glucans as their major hemicelluloses with only 1-5% of the cell wall dry weight as XyG, and gluronoaraniboxylans. By contrast, dicots and other monocots have XyGas their major hemicellulose, which constitutes 10-20% of the cell wall dry weight (Fry, 1989).

Table 1.1 Comparison of polymer contents (%w/w) between Type I and type II primary cell walls Polymer Type I wall Type II wall Cellulose 30-40 30-40 Pectin ~30 2-10 Proteins 10 2-10 Xyloglucan 20 Minor Xylan 5 20-30 Mannan 3 5 (According to Carpita and Gibeaut, 1993; Carpita, 1996; Carpita and McCann, 2000).

1.1.3. Biosynthesis of cell wall polysaccharides

Synthesis of cell wall polysaccharides requires sophisticated, well-coordinated biosynthetic machinery involving specialized enzymes called GTs. Each glycosidic linkage in a polysaccharide requires a specific GT. Most of the GTs are integral membrane proteins, accounting for 1~2% of eukaryotic genomes (Lairson et al., 2008).

31

As of May 2015, eukaryotic and prokaryotic GTs are categorized into 97 distinct families according to the Carbohydrate Active Enzymes (CAZy) database

(http://www.cazy.org/, Campbell et al., 1997; Cantarel et al., 2009; Lombard et al.,

2014). The CAZy database has classified GTs based on amino acid sequence linking the sequence to the specificity and 3D structure of the GTs that assemble, modify and breakdown oligo- and polysaccharides (Lombard et al., 2014).However, there are reports of wrongly assigned GT revealed upon extensive biochemical approaches (Busk and

Lene, 2015).

1.1.3.1. Cellulose

Cellulose is the most abundant component of plant cell wall and is common to both primary and secondary cell walls. It is the major load-bearing network that is responsible for regulation of cell expansion during growth and maintenance of cell shape

(Wrightman and Turner, 2010). It exists as microfibrils and is composed of β -(1,4)- linked D- - chains. Unlike most cell wall polysaccharides, which are synthesized in the , cellulose is synthesized at the plasma membrane by cellulose synthase complexes (CSCs), large membrane-bound complexes with six-fold symmetry.

Three different proteins encoded by the cellulose synthase gene family A (CESAs) are responsible for forming a functional CSC. CESA1 and CESA3, together with CESA2,

CESA5 and CESA6 are involved in primary cell wall biosynthesis, whereas CESA4,

CESA7, and CESA8 are required for secondary cell wall synthesis (Persson et al., 2007).

Excluding CESA10, nine other genes were identified, and an extensive characterization of the mutants was undertaken (Table 1.2). Although CESA9 has sequence similarity

32 with CESA6, a recent study by Stork et al. (2010) revealed that it might be involved in secondary cell wall synthesis. There are 10 CesA genes in the Arabidopsis genome, 10 in rice, 20 in maize, 10 in sorghum, 18 in poplar, and at least 7 in barley (Richmond and

Somerville 2000; Holland et al., 2004; Burton et al., 2004; Djerbi et al., 2005; Penning et al., 2009). Functionally, cellulose microfibrils are highly oriented, creating mechanical anisotropy in the wall, which in turn determines the growth direction of the cell and provide tensile strength to the cell wall.

33

Table 1.2 GTs involved in cellulose biosynthesis and their corresponding mutant Identifier Gene Mutant alleles CAZy Reference GT family AtCESA1 At4g32410 rsw1 GT2 Arioli et al., 1998 AtCESA2 At4g39350 cesa2-1; GT2 Mendu et al., cesa2-2 2011 AtCESA3 At5g05170 ixr; cev1; GT2 Cano-Delgado et eli1; rsw5 al., 2003 AtCESA4 At5g44030 irxr GT2 Taylor et al., 2003 AtCESA5 At5g09870 mum3; cesa5- GT2 Mendu et al., 1; ces5-2 2011 AtCesA6 At5g64740 prc1; irx2 GT2 Fagard et al., 2000; Desprez et al., 2002 AtCESA7 At5g17420 irx3; fra5; GT2 Taylor et al., mur10 2007 AtCESA8 At4g18780 irx; fra6; lew2 GT2 Taylor et al., 2000: Chen et al., 2005 AtCESA9 At2g21770 ces9-1; ces9-2 GT2 Persson et al., 2007 AtCESAA10 At2g25540 nr nr nr nr denotes not reported

34

1.1.3.2. Hemicellulose

Hemicellulose is a heterogenous group of wall polysaccharides also known as cross- linking glycans. They play an important role in strengthening cell wall structure by cross-linking cellulose microfibrils. Golgi localized GTs are involved in the biosynthesis of hemicelluloses. The different types of hemicelluloses include XyG, xylans, mannans glucoronoarabinoxylans (GAX), glucomannans, and β-(1→3, 1→4)-glucans. Depending on the plant species, a particular type of hemicellulose will be predominant in its wall.

For example, in dicots and non commelinoid monocots, XyGis in their primary cell walls and glucuronoxylan (GX) is in their secondary cell walls. In contrast, monocots, like grasses have GAX as the major hemicellulose present in their primary cell walls. A number of GTshave been identified that are responsible for biosynthesis of hemicelluloses (Table 1.3).

Structurally, XyGconsists of a β -(1,4)-D glucan chain backbone substituted by xylose that in turn can be substituted with galactose (Gal) and (Fuc) (Carpita and

McCann, 2000). GX and GAX both consists of a β-(1,4)- D-xylan backbones substituted with α -D-glucuronic acid (GlcA), and 4-O-methyl-α-D-glucuronic acid (MeGlcA), but

GAX also has α-L- substitutions (Ebringerová and Heinze, 2000). In addition to GAX, grasses accumulate mixed-linked β glucans. These are unbranched polysaccharides consisting of a mixture of β-(1,4)-D-glucose and β-(1,3)-D-glucose.

35

Table 1.3 GTs involved in hemicellulose biosynthesis and their corresponding mutants Identifier and Activity CAZy Reference Gene name GT Family Xyloglucan At3g28180 (CSLC4) β-(1,4)-glucan synthase GT2 Cocuron et al., 2007 At3g62720 (XXT1) α-(1,6)-XylT GT34 Faik et al., 2002 At4g02500 (XXT2) α-(1,6)-XylT GT34 Cavalier et al., 2006 At2g20370 (MUR3) β-(1,2)-GalT GT47 Madson et al., 2003 At2g03220 α-(1,6)-FucT GT37 Vanzin et al., (MUR2/FUT1) 2002 Xylan At2g37090 (IRX9) Putative β-(1,4)-xylan GT43 Brown et al., synthase 2007 At4g36890 (IRX14) Putative β-(1,4)-xylan GT43 Levy et al., synthase 1991 At1g27440 Putative β-(1,4)-xylan GT47 Perrin et al., (IRX10/GUT2) synthase 1999 At5g61840 (IRX10- - GT47 Brown et al., LIKE/GUT1) 2009 At2g28110 Rhamnosyl β-(1,3)- GT47 Pena et al., (IRX7/FRA8) Putative XylT 2007

36

Table 1.3 continued At5g22940 (F8H) - GT47 Lee et al., 2009 At5g54690 Xylosyl α-(1,4)-Putative GT8 York et al., (IRX8/GAUT12) GalAT 2008 At1g19300 Synthesis of reducing GT8 Lee et al., (PARVUS/GLZ1) end oligosaccharide. 2007 putative α-XylT Mannan At5g22740 (CSLA2), β-(1,4)-mannan GT2 Dhugga et synthase al., 2004; Liepman et al. 2005; At1g23480 (CSLA3), β-(1,4)-mannan GT2 Wang et al., synthase 2013 At2g35650 (CSLA7), β-(1,4)-mannan GT2 Goubet et synthase al., 2003 At5g03760 (CSLA9) β-(1,4)-mannan GT2 Wang et al., synthase 2013 Mixed linked Glucan CSL-F β-(1,3)- and β-(1,4) GT2 Burton et linkages al.,2006; Doblin et al., 2009 This table is modified from Scheller and Ulvskov, (2010).

1.1.3.3. Pectins

Pectins are the most soluble cell wall polysaccharide, rich in D -galacturonic acid.

Pectin forms the matrix in which cellulose fibrils and hemicellulose are embedded.

37

Pectins not only determine cell wall porosity, but also act as an external barrier against pathogens. Pectins can be grouped into the following categories based on their composition: homogalacturonan (HG), xylogalacturonan (XGA), apiogalacturonan, rhamnogalacturonan I (RGI), and rhamnogalacturonan II (RGII). Homogalacturonan is a homopolymer of α -1,4-linked D-galacturonic acid (GalUA), which can get substituted with other sugar moieties like Xyl in XGA or with apiofuranose (apiogalacturonan). In contrast to XGA, RGI is a branched structure with repeating units of the disaccharide (α-

1,4-D-GalUA- α-1, 2-L-Rha) as the backbone, which in turn can be substituted with either β-1,4-, branched arabinan, and/or AGside chains. RGII on the other hand has the same backbone structure like that of HG, a homopolymer of α-1,4 D-GalUA, but it has elaborate side chains at the O-2 and O-3 positions (Harholt et al., 2010). RGII can occur as a dimer via a borate ester linkage, which contributes to the tensile strength of the wall (O'Neill et al., 2001). The negatively charged HG domains interact with each other in the presence of Ca2+ ions, forming the gel structure of pectin. Bridging between

HGs through Ca2+ is blocked by HG methylesterification, which can be reversed by pectin methylesterases in the cell wall. Numerous GTs involved in pectin biosynthesis have been characterized as listed in Table 1.4 (Atmodjo et al., 2011; Sterling et al.,

2006; Held et al., 2011).

Boughanmi et al. (2010) suggested that unesterified pectin may associate with cellulose, hemicellulose and AGPs, which may in turn strengthen cell wall cohesion.

Interactions between AGPs and pectic polymers were often suggested based on co- purification of the two polymers as well as in an in vitro binding experiment (Carpita,

38

1989; Immerzeel et al., 2006; Iraki et al., 1989 Baldwin et al., 1993). In another study,

AGPs are also speculated to act as calcium capacitors and potentially bind to calcium via the GlcA residuesin the oligosaccharide side-chains (Lamport and Varnai, 2013;

Pickard, 2013). A recent report demonstrated that, a classical AGP, namely, APAP1

(At3g45230), forms cross-links with both pectin and xylan (Tan et al., 2013). Thus elucidating the interaction between pectin and AGP may unravel some of the mysteries of cell wall integrity as physiological role.

Table 1.4 GTs involved in pectin biosynthesis and their corresponding mutants Identifier and gene CAZy Function Reference name GT Family Homogalacturonan (HG)

At3g61130 GT8 Proven HG:GalAT Atmodjo et al., (GAUT1) 2011 At2g38650 GT8 Proven Golgi Sterling et al., (GAUT7) membrane anchor 2006 of GAUT1 At3g25140 GT8 Putative HG:GalAT Bouton et al., (GAUT8/QUA1) 2002 At1g78240 Putative HG:MT Mouille et al., (QUA2/TSD2) 2011 At4g00740 (QUA3) Putative HG:MT Miao et al., 2011 At5g65810 (CGR3) Putative HG:MT Held et al., 2011 Rhamnogalacturonan I (RG-I) At2g35100 (ARAD1) GT47-B Putative Harholt et RG-I:α-1,5-AraT al., 2012

39

Table 1.4 continued At5g44930 (ARAD2) GT47-B Putative Harholt et RG-I:α-1,5-AraT al., 2012 At2g33570 (GALS1) GT92 Proven β-1,4- Liwanag et GalT al., 2012 At5g44670 (GALS2) GT92 Proven β-1,4- Liwanag et GalT al., 2012 At4g20170 (GALS3) GT92 Proven β-1,4- Liwanag et GalT al., 2012 Rhamnogalacturonan II (RG-II) At4g01770 (RGXT1) GT77 Proven RG-II:α- Egelund et 1,3-XylT al., 2006 At4g01750 (RGXT2) GT77 Proven RG-II:α- Egelund et 1,3-XylT al., 2006 At1g56550 (RGXT3) GT77 Proven RG-II:α- Egelund et 1,3-XylT al., 2008 At4g01220 GT77 Proven RG-II:α- Liu et al., (RGXT4/MGP4) 1,3-XylT 2011 Xylogalacturonan (XGA) At5g33290 (XGD1) GT47-C Proven XGA:β- Jensen et al., 1,3-XylT 2008 This table is modified from Atmodjo et al. (2013)

1.1.3.4. Lignin

Lignin is a complex aromatic heteropolymer that imparts strength, rigidity and hydrophobicity to plant secondary cell walls (Boerjan et al., 2003). It is the second most

40 abundant biopolymer after cellulose. It is mainly deposited in terminally differentiated cells of supportive and conducting tissues to withstand the force of gravity, mechanical stress, and the negative pressure generated by transpiration (Bonawitz and Chapple

2010). In addition, to lignification during the course of normal tissue development, various biotic and abiotic stresses, like pathogen infection and wounding can also trigger the production and deposition of lignin at specific sites. Lignin is synthesized mainly from three hydroxycinnamyl alcohols, differing in their degree of methoxylation: a) p- coumaryl, b) coniferyl, and c) sinapyl alcohols, which in turn gives rise to a different types of lignin units called p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S) units, respectively (Rencoret et al., 2011). Common phenotypes of lignin-deficient plants include dwarfing, sterility, collapsed , the inability to stand erect, and in the most extreme cases, developmental arrest shortly after germination (Schilmiller et al., 2009).

Understanding lignin biosynthesis is particularly important for potential bioengineering strategies for mitigating the negative impact of lignin on biofuel production, pulping, or ruminants. To date, ten different Arabidopsis genes involved in phenylpropanoid and monolignol biosynthetic pathways were characterized (Vanholme et al., 2010). This includes phenylalanine ammonia lyase 1 (PAL1), PAL2, cinnamate 4-hydroxylase

(C4H), 4-coumarate: CoA ligase 1 (4CL1), 4CL2, caffeoyl-CoA O-methyltransferase 1

(CCOAOMT1), cinnamoyl-CoA reductase 1 (CCR1), ferulate 5-hydroxylase 1 (F5H1),

COMT, and cinnamyl alcohol dehydrogenase 6 (CAD6). Decreasing lignin leads to unavoidable compromises in plant growth, thus recent approaches are directed towards altering lignin composition or introducing novel monomers or linkages that maintain the

41 function of the polymer, while at the same time, increasing its degradability in industrial processes, thereby facilitating biofuel production.

1.1.4. Plant cell wall proteins

Besides polysaccharides and lignin, cell wall proteins represent another component of plant cell walls. Cell wall proteins are involved in wall structure, support, signaling and interactions with other wall components and with the plasma membrane (Jamet et al., 2006). They are ubiquitous and relatively abundant in all land plants and are also present in . Broadly, cell wall proteins can be grouped into functional and structural proteins, and most of these proteins undergo extensive post translational modifications. Functional proteins include all extracellular enzymes that are required for cell wall construction and remodeling, whereas the structural proteins include hydroxyproline-rich proteins (HRGPs) and glycine-rich proteins (GRPs). Apart from these, there are other wall proteins like cystein-rich thionins and solanaceous lectins.

Structural proteins are mostly rich in one or more specific amino acids, have a repetitive sequence motif, and are often glycosylated.

1.1.4.1. HRGPs

HRGPs are a super family of cell wall proteins represented by a spectrum of molecules. This spectrum ranges from the water soluble, highly glycosylated - AGPs to the water insoluble and moderately glycosylated extensins (EXTs) and lastly to the least glycosylated proteins, the proline-rich proteins (PRPs) (Showalter, 1993). Glycosylation of cell wall proteins plays an important role in dictating the interaction between wall proteins and other wall components and also in defining cell type specificity. Unraveling

42 the biosynthesis of these glycoproteins remains a daunting scientific challenge, as isolating and characterizing the enzymes involved in glycosylation is difficult. The most critical aspect of this challenge is to isolate and solubilize these enzymes, which are integral membrane proteins, while retaining enzyme activity. The lack of a robust and reproducible enzyme assay and suitable substrate acceptors to validate their function may also hinder their characterization. A recent bioinformatics analysis on the

Arabidopsis thaliana genome/proteome identified 166 HRGPs classified as 85 AGPs, 59

EXTs, 18 PRPs, and 4 AGP/EXT hybrid proteins, highlighting the question why plants need so many HRGPs (Showalter et al., 2010).

1.1.4.1.1. Proline-rich proteins (PRPs)

Proline-rich proteins (PRPs) are the least glycosylated HRGPs and are slightly basic in nature. They often contain the repetitive motif of Val-Tyr-Lys-Pro, and are involved in rapid defense responses triggered by either fungal elicitors or wounding (Bradley et al., 1992). Two broad subclasses of PRPs exist: 1) PRPs present as a component of cell wall and 2) PRPs as plant nodulins (Showalter, 1993). Recently, Zhan et al. (2012) demonstrated that a proline-rich protein, named SICKLE (SIC), is critical for development and abiotic stress tolerance in Arabidopsis. In addition, Boron et al. (2015) reported on the identification and characterization of a PRP-like protein (PRPL1), which plays a critical role in root hair elongation. Arabidopsis polysaccharide synthases and

GTs are usually membrane proteins and are present in low amounts in plant cells, making their purification notoriously difficult. However, substantial progress has been made in recent years in the identification of genes for the biosynthesis of cell wall

43 polymers. A variety of strategies have been developed and are usually applied in combination, including homolog identification based on the polysaccharide synthases or the GTs known in other organisms; conventional biochemical purification methods coupled with proteomic methods; forward genetic and reverse genetic methods; and gene mining based on expression data of expressed sequence tags (ESTs) or microarray data. Table 1.5 summarizes the polysaccharide synthases and GTs identified unambiguously for the synthesis of specific wall polymers to date, with the activity tests and mutant phenotypes included if such information is available.

1.1.4.1.2. Extensins (EXTs)

Extensins are moderately glycosylated proteins, highly basic, and widely distributed in the plant kingdom. Structurally, they are characterized by a pentapetide repetitive motif, Ser-Hyp4, and by a protein backbone rich in not only Hyp and Ser residues, but also Val, Tyr, Lys, and His residues to a lesser extent (Kieliszewski and Lamport, 1987).

Hyp residues in the Ser-(Hyp)4 repeats are each modified by 1 to 4 arabinosyl residues, whereas a single galactose residue modifies several of the Ser residues. The gene responsible for galactosylation (i.e. addition of galactose onto Ser residues and addition of short arabino-oligosaccharides onto Hyp residues were recently characterized by heterologous expression, in vitro assays and mutant analysis (Saito et al., 2014; Ogawa-

Ohnishi et al., 2013; Velasquez et al., 2011; Gille et al., 2009; Table 1.5). EXTs are involved in (Cannon et al., 2008) and are induced upon wounding

(Wycoff et al., 1995) and pathogen infection (Garcia-Muniz et al., 1998).

44

Table 1.5 Information on the enzymes, genes, and mutants for extensin glycosylation

Enzyme GT Gene Mutants Mutant References Family identifier phenotype -O- - At3g01720 sgt1 -1 longer Saito et al., α− (SGT1) (SALK_ and 2014 galactosyl 05987) larger transferase sgt1-2 leaves SALK_ 054682)

Hydroxy GT At5g25265 hpat1 pleiotropic Ogawa- proline-O- 8 (HPAT1), (GABI_ phenotypes Ohnishi et β-arabinosyl like At2g25260 298B0), affecting al., 2013; transferase (HPAT2), hpat2 growth and Velasquez At5g13500 (SAIL_ developme et al., 2011 (HPAT3) 178_H), n hpat3 (SALK_ 04766) β-1,2- GT At1g75120 rra1 Reduced Egelund et arabinosyl 77 (RRA1), (SAIL_590_ root hair al. , 2007; transferase At1g75110 Grra2 growth Velasquez (RRA2) (SAIL et al., 2011; _70_D0) At1g19360 rra3 Reduced ; (RRA3) (GABI_ root hair 233B0) growth

45

Table 1.5 continued β-1,2 GT77 At2g35610 xeg113–1 Reduced Gille et al., arabinosyl (XEG113) (SALK_ roothair 2009; transferase 0669) growth Velasquez xeg113–3 et al., 2011 (SALK_ 05809)

1.1.4.1.3. Arabinogalactan proteins (AGPs)

Arabinogalactan proteins are highly glycosylated extracellular matrix proteins that

are implicated in diverse biological roles that span through multiple

stages of plant growth and development (Ellis et al., 2010). They are found attached to

the plasma membrane, in the wall, in the apoplastic space, and in plant secretions (e.g.,

surfaces and wound exudates). They are also found in detergent-resistant

membranes in Arabidopsis, indicting their presence in lipid rafts (Borner et al., 2005).

AGPs occur ubiquitously in all plant cells and in all plant species, from algae to

angiosperms. There are reports that AGPs are capable of cross-linking either to one

another or to other cell wall molecules, namely, pectin and arabinoxylan (Kjellbom et al.

1997; Tan et al. 2013). Periplasmic AGPs reportedly act as calcium capacitors, which is

significant because calcium ion gradients are important for cell expansion (Lamport and

Varnai, 2012). All AGPs share three common features: 1) their ability to bind to β-Yariv

reagent (Yariv et al., 1967), 2) the presence of non-contiguous, clustered Hyp residues in

their protein backbone, and 3) the presence of highly heterogenous O-linked AG

polysaccharides attached to Hyp residues.

46

A major challenge in AGP research is to establish the function of a single AGP. So far, specific biological function cannot be ascribed to a single AGP gene because of the inherent drawbacks of the methods adopted in inferring its function. Most studies either use anti-AGP antibodies or β-Yariv reagent, both of which can bind to the carbohydrate epitopes of more than one AGP (Knox, 1997; Ellis et al., 2010). Alternatively, a reverse genetic approach can circumvent this problem by using mutants for specific AGP genes.

However, mutant analysis has its own limitations; genes involved in biosynthesis of

AGPs protein backbones belong to large gene families and could be functionally redundant. Significant phenotypic changes may remain undetected unless double or triple mutants are produced. This has led to shifting the focus to genes involved in the glycosylation of AGPs, as a possible approach to identify AGP functions and the contribution of the AG polysaccharides to such functions.

1.1.5. Structure of AGPs

Arabinogalactan-proteins (AGPs) are undoubtedly one of the most complex families of macromolecules found in nature. AGPs display a high degree of heterogeneity which arises not only from the incredible diversity of glycans decorating the protein backbone, the array of peripheral sugars decorating the large AG polysaccharide chains, and the varying extent of glycosylation, but also due to the micro-heterogeneity of protein backbone (Ellis et al., 2010). AGPs are defined by the abundance of specific amino acids, Hyp, Ala, Ser, Thr, and Gly, and the presence of type IIAG, which generally comprises about 90% or more of the molecular weight. The AG polysaccharides (Type

II AGs) on AGPs are covalently attached to Hyp residues in the protein backbone, which

47 frequently contains dipeptide repeats of hypalternating with Ala, Thr and Ser, and are mostly composed of galactose and arabinose with varying amounts of , fucose, and glucuronic acid (Kieliszewski and Shpak, 2001; Tan et al., 2010; Tryfona et al.,

2010).

The protein backbones of AGPs are synthesized by members of a large multigene family, consisting of 85 members in Arabidopsis thaliana (Schultz et al., 2002; Johnson et al., 2003; Showalter et al., 2010) and 69 members in rice () (Ma and

Zhao, 2010). The glycan composition of AGPs can vary greatly between species, between organs within the same species, and may even be developmentally regulated within the same organ in different cell-types (Tsumuraya et al., 1988; Pennell et al.,

1991). Based on structural domains of AGP protein backbones, there are several classes of AGPs: 1) classical AGPs, 2) AG peptides, 3) lysine-rich classical AGPs, 4) fasciclin like AGPs (FLAs), and 5) other chimeric AGPs (Figure1.1).

Fasciclin-like arabinogalactan-proteins (FLAs) contain AGP-like glycosylated regions as well as putative cell adhesion domains known as fasciclin domains (FAS).

Twenty-one fasciclin genes have been identified in the Arabidopsis genome (Johnson et al., 2003). FLAs have been classified into four general groups based on the number of

FAS domains (one or two), location and number of AG domains (one or two), and whether or not they contain GPI anchor addition sites (Johnson et al., 2003; MacMillan et al., 2011). MacMillan et al (2011) also suggested that a subset of single FAS containing FLAs contributes to strength by affecting cellulose deposition.

48

Figure 1.1 Schematic representation of the different classes of AGPs deduced from their corresponding DNA sequence. The left-hand panel depicts predicted protein structures of native AGPs, whereas the right-hand panel denotes the AGP structures after processing and post-translational modification. Not drawn to scale. Processing involves removal of the predicted N- terminal sequence, removal of the predicted C-terminal GPI anchor addition signal sequence (if present), followed by the attachment of the GPI-anchor to the C-terminus. Pro residues are hydroxylated to Hyp and O-linked glycans (indicated by feathers) are added to most of the Hyp residues. This image is modified from Gaspar et al. (2001).

1.1.6. Biosynthesis of AGPs

AGPs are one of the most extensively post-translationally modified proteins in nature. Biosynthesis of AGPs consists of several steps: 1) cleavage of the N-terminal

49 signal peptide, 2) GPI anchoring of the C-terminus of the AGP core protein, 3) posttranslational modification of Pro residues to form Hyp, 4) O-glycosylation of

Hyp residues, which likely begins in the Endoplasmic reticulum (ER) but is more extensive in the Golgi, and 5) glycosyl hydrolase-mediated hydrolysis of AGPs.

1.1.6.1. GPI anchor addition in AGPs

In , several integral membrane proteins remain attached to the plasma membrane by glycosylphosphatidylinositol (GPI) molecules. GPI anchors provide an alternative to transmembrane domains for anchoring proteins to cell surfaces (Shultz et al., 2000). Oxley and Bacic (1999) demonstrated membrane bound GPI-anchored AGPs

(PcAGP1) from pear suspension culture cells, and they also showed the presence of an extracellular form of PcAGP1, that lacked the GPI anchor. This finding indicated that, for secretion of AGPs, GPI anchors of AGPs are likely cleaved by the action of a specific phospholipase. So far, 12 classical AGPs, five AG peptides, and 17 fasciclin- like AGPs have been predicted to be GPI-anchored in Arabidopsis (Gaspar et al., 2001).

1.1.6.2. of prolyl residues by prolyl hydroxylase

Prolyl 4-hydroxylase (P4H) is the enzyme responsible for conversion of proline

(Pro) into 4-hydroxyproline (4-Hyp) in the ER. Earlier studies by Cohen et al. (1983) showed that P4H, a membrane-associated enzyme catalyzes this reaction in presence of oxygen, α-ketoglutarate, ferrous ion, and ascorbate. The Arabidopsis genome contains

13 P4H gene family, of which only two enzymes, AtP4H1 and AtP4H2, are characterized using an in vitro assay (Vlad et al., 2007; Tiainen et al., 2005). Recent reports on the genetic analysis of P4H2, P4H5, and P4H13 loss of function mutants

50 indicated that they play a pivotal role in root hair tip growth (Velasquez et al 2011;

Velasquez et al., 2015; Table 1.6). Velasquez et al. (2015) also provided evidence of an enzyme complex formed by P4H5 with P4H2 and P4H13.

Table 1.6 Information on the prolyl-4-hydroxylase (P4H) genes and their mutants

Gene Mutants Mutant phenotypes References Identifier At3g06300 p4h2.1 Reduced root hair Velasquez et al., (P4H2.1) (SALK_042128) growth 2011 p4h2.2 (SALK_118623) At2g17720 p4h5 Reduced root hair Velasquez et al., (P4H5) (SALK_152869) growth 2011 At2g23096 p4h13 (SAIL Reduced root hair Velasquez et al., (P4H13) 425_H02) growth 2011

At2g43080 - - Hieta and (P4H1 ) Myllyharju, 2002; Vlad et al., 2007

1.1.6.3. Glycosylation of AGPs

Despite intense research to unravel AGP function, their molecular mechanism(s) of action remain elusive. To address the function and regulation of AGP biosynthesis, it is important to understand the mechanism underlying AGP glycosylation. Identification of cell wall biosynthetic enzymes through a biochemical purification route is often

51 hampered by the low abundance of the enzymes, unavailability of substrate acceptors or difficulties in maintaining enzymatic activities during the solubilization/purification process.

1.1.6.3.1. Hyp-galactosyltransferases (Hyp-O-GALTs)

In mammals, galactosyltransferases are thoroughly studied, and their activity and biological functions are well characterized. AGPs are analogous to animal and mucins (Nothangel, 1997). Thus, it is hypothesized that the GTs responsible for adding the first sugar to the protein core of mammalian proteoglycans could be similar to the GT responsible for adding the first sugar to the AGP protein backbone in plants.

Galactosaminyltransferases (GalNAcTs) involved in biosynthesis of mammalian O- glycosylated proteins, namely mucins, are extensively studied. They are known to transfer N-acetylgalactosamine (GalNAc) to the hydroxyl group of Ser and Thr residues of the mucin core protein by an α -1 linkage (Julenius et al., 2004). Characterizing

GALTs in plants is extremely challenging as they are integral membrane proteins, and thus are difficult to solubilize without affecting their activity. Consequently, a bioinformatics approach adopted for identifying Arabidopisis homologs to mammalian

β-(1,3) GALTs identified 20 putative Arabidopsis β-(1,3) GALTs in the CAZy GT-31 family, one of which (GALT1) was previously identified as a β-(1,3) GALT involved in the biosynthesis of a protein-bound N-linked oligosaccharide (N-glycan) (Strasser et al.,

2007).

Since 90% or more of a typical AGP molecule is composed of carbohydrate, it can be assumed that AG polysaccharides play integral roles in AGP function. These

52 polysaccharides likely define the interactive surface of AGP molecules and hence their functions. Liang et al. (2010) proposed that there could be as many as 15 different

GTsinvolved in the biosynthesis of AGPs (Table 1.7). Considerable progress in recent years has led to the identification of two fucosyltransferases genes (FUT4 and FUT6), which are members of CAZy GT -37 family, five β-GALT genes (GALT31A, and three

HPGT1, HPGT2 and HPGT3) which are members of the GT- -31 family, and three β- glucoronosyltransferases genes (AtGlcAT14A, AtGlcAT14B and AtGlcAT14C) belonging to the GT- -14 family that encode GT specific for the biosynthesis of AGPs (Wu et al.,

2010; Knoch et al., 2013; Geshi et al., 2013; Table 1.7). Of all the GTs involved in the

O-glycosylation of AGPs, the galactosyltransferases that add the first galactose on to the protein backbone are crucial as they produces the acceptor for subsequent glycosylation events.

53

Table 1.7 Information on the enzymes, genes, and mutants for AGP glycosylation Enzyme GT Gene Mutants Mutant References Family Identifier phenotypes Hydroxy- GT31 At5g53340 SALK_ Longer Ogawa- proline-O- (HPGT1) 007547, lateral roots, Ohnishi et β- At4g32120 (hpgt1-1) longer root al., 2015 galactosyl (HPGT2) SALK_ hairs, radial transferase At2g25300 070368, expansion of (HPGT3) (hpgt2-1) the cells in SALK_ the root tip, 009405, small leaves, (hpgt3-1) shorter stems, reduced fertility and shorter siliques. β-1,6- GT29 At1g08280 Not reported - Dilokpimol galactosyl- (GALT29) et al., 2014 transferase β-1,3- GT31 At1g77810 Not reported - Qu et al., galactosyl- 2008 transferase β-1,6- GT31 At1g32930 galt31A Geshi et galactosyl- (GALT31) (FLAG_379B0) lethal mutant al., 2013 transferase

54

Table 1.7 continued β-1,6- GT14 At5g39990 glcat14a-1 Enhanced Knoch et glucuronosyl (GlcAT14) (SALK_ cell al., 2013; transferase At5g15050 06433) elongation Dilokpimol (GlcAT14) glcat14a-2 in and Geshi, At2g37585 (SALK seedlings 2014 (GlcAT14) _04390) α-1,2-fucosyl GT37 At2g15390 fut4-1 Reduced Wu et al., transferase (SAIL_ root 2010; 284_B) growth Liang et fut4-2 under salt al., 2014, At1g14080 (SALK_ stress Tryfona et 12530 al., 2014 fut6-1 (SALK_0 783) fut6-2 (SALK_ 09950))

1.1.7. Function of AGPs

Mounting evidence has ascribed several putative functional roles to various AGPs

(Seifert and Roberts, 2007; Ellis et al., 2010; Tan et al., 2012; Tan et al., 2013),

including cell expansion, xylem cell differentiation, and cell wall development. They are

also involved in a large number of other biological processes, including

(Langan and Nothnagel, 1997), (Gao and Showalter, 1999; Guan

and Nothnagel, 2004), cell differentiation (Majewska-Sawka and Nothngel, 2000; dos

55

Santos et al., 2006), cell expansion (Darley et al., 2001; Lu et al., 2001), host/microbe interactions (van Buuren et al., 1999; Johnson et al., 2003; Nguema-Ona et al., 2013), and abiotic stress responses (Shi et al., 2003; Xu et al., 2008; Table 1.7). Furthermore, a variety of functions were conferred to AGP specific-GT mutants, ranging from embryo lethality to conditional effects on root growth, cell elongation, and pollen tube growth, indicating that post-translational modification of AGPs can control the function (Ogawa-Ohnishi et al., 2015; Geshi et al., 2013; Knoch et al., 2013). Most of this evidence came from genetic mutant analysis of either a specific AGP coding gene or an AGP specific GT gene or from the use of β-Yariv reagent or AGP-specific antibodies

(Table 1.8). Such work highlights the importance of these glycoproteins in plant survival and development. Glycome profiling of cell wall fractions obtained from root and leaves of fut4fut6 compared to wild type plants revealed distinct patterns, indicating that fucosyl residues on AGPs may regulate intermolecular interactions between AGPs and other wall components (Ling et al., 2013). AGPs also possess valuable adhesive and emulsification properties that are utilized for commercial purposes. There is also some interest in studying the structure of AGPs as they apparently act as modulators of immune responses (Pettolino et al., 2006).

1.1.7.1. Role of AGPs in sexual reproduction

AGPs are implicated in pollen development and in development of the female gametophyte. Arabidopsis thaliana agp6agp11 double mutants showed reduced pollen biogenesis, which included both altered germination and pollen tube elongation

(Coimbra et al., 2010). Similarly, FLA3 is reported to be involved in microspore

56 development and may affect pollen intine formation, possibly by impacting cellulose deposition (Li et al., 2010). Furthermore, a recent study demonstrated that AGP6,

AGP11, AGP23 and AGP40 are integral to the formation of nexine, an inner exinic layer in pollen walls in Arabidopsis (Jia et al., 2014). Of the four pollen-specific AGPs, AGP6 and AGP11 are classical AGPs, whereas AGP23 and AGP40 are AG peptides

(Showalter et al., 2010). In another study, Arabidopsis RNAi-AtAGP18 lines demonstrated impaired female gametogenesis and differentiation of viable megaspores

(Demesa-Arevalo and Vielle-Calzada, 2013). Taken together, these studies clearly highlight the importance of AGPs in sexual reproduction in plants.

1.1.7.2. Role of AGPs in cell division and expansion

Indirect evidence indicates that AGPs are involved in cell expansion, cellulose deposition and cell wall thickening (Shi et al., 2003; Xu et al., 2008). Initial studies showed that root epidermal cell elongation was severely reduced in response to β-Yariv reagent, which specifically binds to glycosylated AGPs (Willats and Knox, 1996; Ding and Zhu, 1997). Similar findings were reported in treatment of carrot suspension cultures with β-Yariv reagent. Genetic mutants of FLA4 (denoted as fla4 or sos5 for salt overly sensitive 5) show abnormal root phenotypes, including reduced root growth and swelling of the root tip when compared to normal Arabidopsis plants grown on media containing elevated salt or sucrose (Shi et al., 2003; Xu et al., 2008). Given that AGPs are rich in carbohydrate moieties, they may be involved in cellular signaling, either directly or indirectly, which may in turn influence cellulose biosynthesis. AGPs were

57 also shown to be required for apical cell extension in the moss

(Lee et al., 2005).

Overexpression of AtAGP18 in Arabidopsis thaliana resulted in plants that displayed smaller rosettes, shorter stems and roots, more branches and fewer viable seeds (Zhang et al., 2011). In a separate study, mutant analysis of a lysine rich AGP mutant, atagp19, displayed delayed growth, shorter hypocotyls and inflorescence stems, and reduced seed production, indicating roles in cell division, cell expansion and leaf development (Yang et al., 2007). Further, the fla11fla12 T-DNA knockout double mutant had altered cellulose composition and microfibril angles, as well as compromised stem strength and stiffness (MacMillan et al., 2010).

1.1.7.3. Role of AGPs in tip growth

Polarized growth is essential for specialized cells like pollen tubes, root hairs and trichomes, and such polarized growth is called tip growth. Treatment of pollen tubes with β-GlcY that selectively binds to AGPs reversibly abolishes tip growth, which can be rescued by removal of the reagent (Roy et al., 1998; Mollet et al., 2002); most of the

AGP epitopes have been localized on the tube surface (Jauh and Lord, 1996; Chen et al.,

2007; Speranza et al., 2009). Coimbra et al. (2010) have demonstrated that pollen- specific AGPs, namely AGP6 and AGP11 are functionally redundant and integral to pollen biogenesis in Arabidopsis. Disruption of fucosylated AGPs in Arabidopsis mur1 mutants results in excessive bulging of root hair bases (van Hengel and Roberts, 2002).

Ding and Zhu (1997) reported similar findings in trichoblasts of the Arabidopsis reb1 mutant, which was depleted of AGPs in their cell walls. Dramatic bulging of trichoblast

58 cells and disruption root hair elongation was also observed in Arabidopsis upon treatment with β -Yariv, which selectively binds to glycosylated AGPs (Willats and

Knox, 1996). Recently, Marzec et al. (2015) suggested that of AGPs play an important role for root hair development in barley. All this evidence indicates that AGPs play a critical role in polarized tip growth in plants.

1.1.7.4. Role of AGPs in programmed cell death (PCD)

Tracheary elements (TEs) undergo autolysis followed by PCD as they differentiate and mature. Majewska-Sawka and Nothnagel, (2000) postulated that AGPs present in both tracheary elements and cotton fibers mark cells destined for PCD. In Zinnia mesophyll cells, AGPs are thought to act as inducers of tracheary element differentiation, a process culminating in PCD (Chaves et al. 2002). In maize coleoptiles,

AGPs are localized to plasma membranes and cell walls of cells destined to differentiate into tracheids and schlerenchyma; both of these processes require PCD (Buckner et al.,

2000). Furthermore, xylogen, an AGP-lipid transfer protein hybrid, localized to cell walls undergoing tracheary element differentiation, is crucial for xylem differentiation in

Zinnia mesophyll cells (Motose et al., 2004). Yariv reagent treatment also results in the induction of programmed cell death (PCD) in tobacco BY-2 cells and Arabidopsis suspension cultured cells (Chaves et al., 2002; Gao and Showalter, 1999). Consistent with such findings, LeAGP-1 was localized to differentiating xylem and metaxylem in stems and roots of plants from initiation to completion (Gao and Showalter,

2000). In addition, overexpression of LeAGP-1 in tomato produced dwarf plants, which was possibly related to interference of secondary cell wall synthesis. Interestingly, the

59 leaves of these plants showed delayed senescence (Sun et al., 2004). In poplar, classical

AGPs were enriched during fiber cell death (Moreau et al., 2005). Collectively, these findings point to AGPs playing a role in programmed cell death (Schindler et al., 1995), although their precise mechanism of action is unknown in this process.

1.1.7.5. Role of AGPs in cellular signaling

AGPs are implicated as cellular signaling molecules. Sardar et al. (2006) suggested that =GPI-anchored AGPs function to link the plasma membrane to the cytoskeleton; thus, they are ideally positioned to transmit signals between the cell wall and the cytoplasm. Driouich and Baskin (2008) proposed that AGPs are involved in rearranging scaffold proteins and serve to disrupt microtubule organization, thereby indirectly influencing cellulose deposition. AGPs could also direct cellulose deposition by serving as positional cues and adhesive cell wall components (Kohorn et al., 2000).

Furthermore, Zhang et al. (2011b) suggested that AtAGP18 may act as a co-receptor for signal transduction during plant development and growth, including binding cytokinins, interacting with receptor kinases, or ion channels. In addition, several lines of evidence indicate that AGPs are critical for pollen-pistil interaction (Cheung et al., 1995; Wu et al., 1995; Wu et al., 2000; Zhang et al., 2014). Pereira et al. (2014) suggested that two pairs of pistil AGPs are important for sexual reproduction, AGP4/AGP7 and

AGP25/AGP27. Xylogen, is a highly glycosylated AGP hybrid, which includes a GPI- anchor as well as a nonspecific lipid transfer , that directly acts as an extracellular developmental signal (Motose et al., 2004). Although these various lines of evidence are largely circumstantial, they implicate AGPs as likely signaling molecules.

60

Table 1.8 List of well characterized AGP mutants and their respective phenotypes Identifier and Type of Mutant Function Reference Gene name AGP alleles At3g46550 Fasciclin fla4 High-salinity stress Shi et al., (FLA4) like AGPs 2003 At2g24450 Fasciclin fla3 Development of male Li et al., (FLA3) like AGPs reproductive organs and 2010 pollen grains At5g55730 Fasciclin fla1 Shoot regeneration Johnson et (FLA1) like AGPs al., 2011 At5g03170 Fasciclin fla11 Synthesis of MacMillan (FLA11) like AGPs secondary cell wall et al., 2010 At5g60490 Fasciclin fla12 synthesis of MacMillan (FLA12) like AGPs secondary cell wall et al., 2010 At3g57690 AG peptide agp23 Formation of pollen wall Jia et al., (AGP23) 2014 At3g20865 AG peptide agp40 Formation of Jia et al., (AGP40) pollen wall 2014 At2g23130 Lys-rich agp17 influence Agrobacterium Gaspar et (AGP17) classical binding al., 2004; Yang et al., 2007

61

Table 1.8 continued At4g37450 Lys-rich agp18 Mediated megaspore Yang et al., (AGP18) classical selection 2007. Demesa- Arévalo et al., 2013 At1g68725 Lys-rich agp19 Shorter rosette Yang et al., (AGP19) classical leaves, delayed 2007 growth, fewer seeds At2g33790 Chimeric agp30 Root regeneration Van Hengel (AGP30) AGP and seed germination and Roberts, 2003 At1g28290 Chimeric agp31 Insensitive to ABA Liu and (AGP31) AGP induced inhibition of Mehdy, 2007 seed germination

1.2. Objectives

The identification of AGP specific Hyp-O-GALTs is a major step forward in our understanding of the biosynthesis of AGP and provides biochemical and molecular tools for further research especially probing the role of glycosylated AGP in plant growth and development. To accomplish these goals, the following specific objectives will be investigated

Identification of the candidate genes involved in initial steps of glycosylation of

AGPs-The main objective is to first identify candidate enzymes more specifically GALT involved in glycosylation of AGPs using in silico predictions in conjunction with biochemical, molecular and cellular biology. The identified candidate genes will be

62 heterologously expressed in Pichia pastoris as recombinant fusion proteins followed by testing their activity utilizing a robust in vitro AGP-GALT assay. Subcellular localization of these candidate proteins will be also studied using a transient expression system in tobacco leaves to support the functional characterization work.

Biochemical and physiological characterization of the corresponding genetic mutants-The identified putative GALTs will be functionally characterized using genetic mutants and analyzing the phenotypes of the mutants in comparison to the wild type plants.

Finally, probing the role of AGPs in cellular signaling by examining the response of the galt mutants towards plant growth regulators and specific chemical inhibitors.

1.3. Organization of the Dissertation

This dissertation has been organized into six chapters. Chapter 1 contains a literature review covering description on different components of plant cell wall and their respective roles with emphasis on their structure, biosynthesis and function of

AGP. In addition to this, it provides literature review of the past and current knowledge of (AGP) structure function has been discussed.Chapter 2 deals with identification of a six membered GT31 family and six candidate GALTs and characterization of these putative GALTs using an in vitro AGP specific GALT assay. Furthermore it provides an extensive biochemical characterization of the reaction product. Chapter 3 deals with additional characterization of another Hyp-O-GALT, AtGALT5 with complimentary in depth phenotypic characterization of single and double galt mutants. Chapter 4 mostly focused on summarization all the five GALT genes and subtle unique phenotypic

63 features. This chapter is intended to be published in BMC Plant Biology journal.

Chapter 5 delves into the cellular signaling aspect of AGP by taking advantage of a fasciclin like GPI anchored AGP, SOS5 and two well characterized cell wall receptor kinases, namely FEI1/FEI2 and two AGP specific Hyp-O-GALTs, GALT2 and GALT5.

This chapter is intended to be published in Plos One journal. Chapter 6 consists of a summary of above chapters and ends with future directions.

64

CHAPTER 2: FUNCTIONAL IDENTIFICATION OF A HYDROXYPROLINE-

O-GALACTOSYLTRANSFERASE SPECIFIC FOR ARABINOGALACTAN-

PROTEIN BIOSYNTHESIS IN ARABIDOPSIS

This work has been published in the following manuscript.

Basu D, Liang Y, Liu X, Himmeldirk K, Faik A, Kieliszewski M, Held M,

Showalter AM (2013) Functional identification of a hydroxyproline-O- galactosyltransferase specific for biosynthesis in Arabidopsis. J.

Biol. Chem. 288: 10132–10143.

2.1. Abstract

Although plants contain substantial amounts of AGPs, the enzymes responsible for

AGP glycosylation are largely unknown. Bioinformatics indicated AGP galactosyltransferases (GALTs) are members of the CAZy GT31involved in N- and O- glycosylation. Six Arabidopsis GT31 members were expressed in Pichia pastoris and tested for enzyme activity. The At4g21060 gene (named AtGALT2) was found to encode activity for adding galactose (Gal) to Hyp in AGP protein backbones. AtGALT2 specifically catalyzed incorporation of [14C]Gal from UDP-[14C]Gal to Hyp of model substrate acceptors having AGP peptide sequences, consisting of non-contiguous Hyp residues, such as [Ala-Hyp] repetitive units exemplified by chemically synthesized

[AO]7 and HF-deglycosylated d[AO]51. Microsomal preparations from Pichia cells

14 expressing AtGALT2 incorporated [ C]Gal to [AO]7, and the resulting product co-

14 eluted with [AO]7 by reverse phase HPLC. Acid hydrolysis of the [ C]Gal-[AO]7

65

14 14 product released [ C]radiolabel as Gal only. Base hydrolysis of the [ C]Gal-[AO]7 product released a [14C]radiolabeled fragment that co-eluted with a Hyp-Gal standard after HPAEC fractionation. AtGALT2 is specific for AGPs as substrates lacking AGP peptide sequences did not act as acceptors. Moreover, AtGALT2 uses only UDP-Gal as the substrate donor and requires Mg2+ or Mn2+ for high activity. Additional support that

AtGALT2 encodes an AGP GALT was provided by two allelic AtGALT2 knockout mutants, which demonstrated lower GALT activities and reductions in β -Yariv- precipitated AGPs compared to wild type plants. Confocal microscopic analysis of fluorescently tagged AtGALT2 in tobacco epidermal cells indicated AtGALT2 is likely localized in the endomembrane system consistent with its function.

2.2. Introduction

Plant cell walls are complex, dynamic structures composed of polysaccharides and glycosylated proteins (Cosgrove et al., 2005; Somerville et al., 2004). Proteins are important components in plant cell walls because of their contribution to cell wall architecture and function. HRGPs are one such structural cell wall . They are represented by a spectrum of molecules ranging from highly glycosylated - AGPs to the moderately glycosylated extensins (EXTs) and finally to the lightly glycosylated proline-rich proteins (PRPs) (3). Bioinformatic analysis has revealed the presence of 166

HRGPs from Arabidopsis, including 85 AGPs, 59 EXTs, 18 PRPs, and 4 AGP/EXT hybrid proteins (Showalter et al., 2010).

AGPs are the most structurally complex HRGPs and are implicated to function in plant growth, development, signaling and plant-pathogen interactions (Ellis et al., 2010;

66

Liu et al., 2007). They are found in plasma membranes, cell walls and plant exudates

(Seifert and Roberts, 2007). AGPs are defined by three criteria: the presence of type II arabino-3, 6-galactan chains, a hyp-rich protein backbone and the ability of most AGPs to bind to a class of synthetic phenylazo dyes, the β-Yariv reagents (Clarke et al., 1979).

AGP protein backbones are typically rich in Hyp alternating with Ala, Thr and Ser, whereas their carbohydrate moieties are mostly composed of galactose and arabinose with varying amounts of rhamnose, fucose and glucuronic acid (Qu et al., 2008). The polysaccharide chains of AGPs are comprised of β -(1,3)-galactan chains interrupted with β -(1,6)-galactopyranose (Gal) side chains and terminated mostly with arabinofuranose (Ara) residues (Bacic et al., 1987; Tan et al., 2004; Tan et al., 2010).

Recently, Tryfona et al. (2012) reported that Arabidopsis AGPs are similarly decorated with a linear β-(1,3)-galactan backbone with β-(1,6)-D-galactan side chains.

Although substantial progress has been made in elucidating GTs responsible for biosynthesis of many cell wall polysaccharides, little is known about the mechanisms and enzymes involved in the biosynthesis of AGPs. Unraveling the biosynthesis of these glycoproteins remains a daunting scientific challenge given that isolating and characterizing the enzymes involved in glycosylation is difficult. One critical aspect of this challenge is to isolate these enzymes, which are most likely integral membrane proteins, in their active form. In other cases, the lack of a robust and reproducible enzyme assay to validate their function presents another challenge. Liang et al. (2010) proposed that as many as 15 different GTs may be involved in the biosynthesis of AGPs.

Of which, only two GTs, specifically two fucosyltransferases (FUTs) have been

67 successfully characterized and shown to add terminal fucose (Fuc) residues on AGPs

(Wu et al., 2010). Another candidate gene, encoding a putative transferase has been recently demonstrated to transfer Gal to an O-methylated Gal-β-(1,3)-Gal disaccharide acceptor, an analogue of the β-(1,3)-galactan chains found in AGPs (Qu et al., 2008).

Of all the GTs involved in O-glycosylation of AGPs, the hyp-O- galactosyl- transferase (Hyp-O-GALT) that adds the first Gal onto the protein backbone is crucial as it produces the acceptor for further glycosylation events. In mammals, GALTs are extensively studied, and their activity and biological functions are well characterized

(Wandall et al., 2007; Pedersen et al., 2007; Bennett et al., 2011). AGPs are analogous to animal proteoglycans and mucins (Nothnagel, 1997). Hassan et al. (2000) suggested that lectin domain containing GTs are a large family of N-acetyl galactosaminyltransferase

(GalNAc-Ts) that add N-acetyl galactosamine (GalNAc) to mammalian mucins and other protein backbones initiating O-glycosylation on either the nascent polypeptide or on a glycopeptide acceptor. Thus, it is hypothesized that the GTs responsible for adding the first sugar to the protein core of mammalian proteoglycans should be similar to the

GTs responsible for adding the first sugar to the AGP protein backbone. Recently, Liang et al. (2010) and Oka et al. (2010) reported on novel in vitro assays using synthetic AGP peptides for detecting Hyp-O-GALT activities in Arabidopsis microsomal membranes.

Using the protocol published by Liang et al. (2010), we report here on the identification and functional characterization of an Arabidopsis At4g21060 gene (named AtGALT2) that encodes a β-GALT involved in the biosynthesis of the glycan chain of AGPs.

68

2.3. Results

2.3.1. Identification of putative AGP GALTs in Arabidopsis thaliana by in-silico analysis

A bioinformatics approach was adopted for identifying putative Hyp-GALT genes

(Hyp-GALTs) involved in adding the first Gal to Hyp residues in AGPs. First, a phylogenetic tree was generated by submitting homologous animal and plant sequences encoding a galactosyltransferase-(GALT) catalytic domain (i.e., pfam 06712) (Figure

2.1). Only 20, of the 33 members of Arabidopsis GALTs in the GT31 family were used in this phylogenetic analysis as the remaining 13 accessions do not contain a GALT domain but instead have a Domain of Unknown Function (DUF604). Two of these 20 family members have been characterized; At1g26810 (GALT1) was identified as a β -

(1,3)-GALT involved in biosynthesis of a Lewis a on N-linked glycans (Strasser et al., 2007), and At1g77810 was reported to be a β-(1,3)-GALT that catalyzes transfer of Gal to a O-methylated Gal-β-(1,3)-Gal disaccharide, which mimics a partial structure of AGP side chains (Qu et al., 2008). Interestingly, only six Arabidopsis proteins

(At1g26810-GALT1, At4g21060-GALT2, At3g06440-GALT3, At1g27120-GALT4,

At1g74800-GALT5 and At5g62620-GALT6) contain a GAL-LECTIN (GALECTIN) binding domain (pfam 00337) in addition to the GALT domain (pfam 06712). This finding was consistent with that reported by Qu et al. 2008. Moreover, this GALECTIN domain is absent in all mammalian β-1, 3-GALTs in the GT31 family and in all other plant GTsin the CAZy database. Interestingly, previous studies found that a lectin domain is present in polypeptide α -N-acetylgalactosaminyltransferases (GalNAc-Ts).

69

These enzymes belong to GT27 and are involved in catalyzing the first step of O- glycosylation of mucins (Wandall et al., 2007; Pedersen et al., 2011; Bennett et al.,

2012). Consequently, it was hypothesized that plant GALTs contain analogous lectin domains and function in initiating O-glycosylation of AGPs (Egelund et al., 2011).

Thus, bioinformatics analysis indicated AtGALT1-6 represent six promising candidates for having Hyp- GALT enzymatic activity.

70

Figure 2.1 Bioinformatics analysis of putative Arabidopsis β-galactosyltransferases (GALTs) from CAZy GT family GT31. (A) Phylogenetic analysis of putative full length Arabidopsis GALTs as inferred by phylogeny.fr (http://www.phylogeny.fr/version2_cgi/index.cgi). Deduced amino acids sequences of pfam 01762-containing proteins from Arabidopsis (20), rice (20), Medicago truncatula (1), Sorghum bicolor (1), Vitis vinifera (8), Poplus (7), Brachypodium (9), Zea mays (11), mammals (17), Drosophila (1) and C. elegans (1) were aligned using maximum likelihood. Numbers on the nodes indicate the confidence values using 100 replications. Bootstrap values of > 50% were presented. The star indicates position of AtGALT (1-6) in clade III.

71

2.3.2. Heterologous expression of putative Hyp-GALT genes in Pichia cells

Six recombinant proteins (AtGALT1, AtGALT2, AtGALT3, AtGALT4, AtGALT5 and AtGALT6) fused with a 6xHis-tag were expressed in Pichia. Microsomal proteins from these recombinant lines were examined by immunoblotting with antibodies against the 6xHis tag and demonstrated the presence of recombinant fusion proteins of the predicted sizes. For example, AtGALT2 recombinant lines had the expected 78 kD protein band reacting with the 6xHis antibody (data not shown). For the AtGALT2 transformants, as well as the other recombinant lines, an additional smaller protein band

(~50kDa) was often detected that may be attributed to protein degradation by endogenous proteases in Pichia. Pichia cells transformed with the empty expression vector served as a negative control (NC) and lacked the recombinant protein band.

2.3.3. Heterologously expressed AtGALT2 demonstrates Hyp-GALT activity

An in vitro GALT assay developed by Liang et al. (2010) was used to test for activity of the recombinant AtGALTs expressed in Pichia cells. The components of the

GALT assay were detergent-permeablized microsomal membranes from the transformed

Pichia cell lines expressing one of the six AtGALT proteins as the enzyme source,

14 UDP-[ C]Gal as the sugar donor and two AGP peptide analogs (d[AO]51 and [AO]7) as the substrate acceptors. Only AtGALT2 showed activity to date; consequently, further product characterization and biochemical analysis has focused on AtGALT2. The amount of GALT activity varied in the ten independent cell lines (C1 to C10) of Pichia

14 cells expressing AtGALT2 based on the rate of [ C]Gal incorporation using the [AO]7

72 substrate acceptor (Figure 2.2). The C2 clone demonstrated the highest enzyme activity

(Figure 2.1).

Figure 2.2 Screening for AtGALT2 expressed in Pichia cell lines. [AO]7-dependent GALT activity tests of the 10 transgenic Pichia cell lines using Triton X-100 permeablized microsomal membranes. For each line, 250 µg of total microsomal membrane protein was used for the assay. [14C]Gal radiolabel incorporation is expressed as pmol/h/mg protein and reflects the difference between total incorporation obtained in reaction products in the presence versus absence of [AO]7 acceptor substrate. Reactions were done in triplicate and mean values are presented. All cell lines tested had AtGALT2 activity but varied in the rate of incorporation. Student’s t -test was performed using Graphpad Quickcalcs (http://www.graphpad.com/quickcalcs/) and significant differences in GALT activity were detected with respect to NC. (* and ** denote p< 0.05 and p< 0.01 respectively).

2.3.4. [AO]7 and d[AO]51 are substrate acceptors for AtGALT2

Total microsomal membranes fromPichia transformants expressing AtGALT2 were analyzed for Hyp-GALT activity using two substrate acceptors: [AO]7, a synthetic peptide and d[AO]51, a transgenically expressed and chemically deglycosylated protein.

Incorporation of [14C]Gal from UDP-[14C]Gal onto the two substrate acceptors was

73 observed by HPLC fractionation (Figure 2.3C and 2.3F) and by comparison to the non- radioactive [AO]7 and d[AO]51 substrate acceptor peaks (Figure 2.3A and 2.3D). Two

[14C]-radioactive peaks were detected, of which peak II has the same retention times as their respective substrate acceptors ([AO]7 and d[AO]51). The identity of peak I is not known; it may represent free [14C]Gal released by an endogenous galactosidase (Kato et al., 2003) or be composed of oligosaccharides with [14C]Gal incorporated into endogenous sugar acceptors as suggested previously (Liang et al., 2010). Microsomal preparations from a Pichia cell line transformed with the empty expression vector were used as negative controls (NC) (Figure 2.3B and 2.3E). Thus, HPLC fractionation provided evidence for incorporation of the [14C]radiolabel from UDP-[14C]Gal onto the

[AO]7 and d[AO]51 acceptors with relatively higher AtGALT2 enzyme activity demonstrated with the [AO]7 substrate acceptor compared to d[AO]51. Consequently, the

[AO]7:AtGALT2 reaction product was subjected to further characterization.

74

Figure 2.3 RP-HPLC fractionation of the [AO]7:AtGALT2 and d[AO]51:AtGALT2 reaction products on a PRP-1 reverse-phase column. Acceptor substrate alone (Panels A and D), GALT reaction with microsomal membranes from the NC Pichia line transformed with the empty expression vector (Panels B and E) and the GALT reaction with microsomal membranes from the transgenic Pichia C2 line (Panels C and F) were fractionated by RP-HPLC using identical elution conditions. Radioactive Peak II coeluted with the [AO]7 and d[AO]51 acceptor substrates in the GALT2 reaction and was used for subsequent product analysis.

75

2.3.5. Product characterization by acid and base hydrolysis shows that AtGALT2 transfers Gal to Hyp residues

To confirm that the [14C]radiolabel remained associated with Gal, RP-HPLC

14 fractions containing the [ C]radiolabeled [AO]7:AtGALT2 reaction products were pooled and subjected to total acid hydrolysis. The resulting acid hydrolyzed

[14C]radiolabeled monosaccharide was fractionated by HPAEC and showed that

[14C]label co-eluted with Gal, thereby confirming incorporation of [14C]Gal onto the

[AO]7 peptide (Figure 2.4).

Figure 2.4 Monosaccharide analysis of the RP-HPLC purified [AO]7:GALT reaction product following acid hydrolysis. Permeablized microsomal membranes from the transgenic Pichia C2 line expressing 14 6xHis-AtGALT2 served as the enzyme source in the [AO]7:GALT reaction. The [ C]- labeled monosaccharides were analyzed by High-Performance Anion-Exchange Chromatography (HPAEC) on a CarboPac PA-10 column. Elution times of monosaccharide standards are as indicated with arrows at the top.

76

In another set of experiments, base hydrolysis was used to confirm that the [14C]Gal residues are added to Hyp residues and to examine the extent of galactosylation of the

[AO]7 peptide acceptor. Base hydrolysis degrades peptide bonds, but keeps Hyp-

14 glycosidic bonds intact (Shpak et al., 2001). The intact [ C]radiolabeled [AO]7 peptide product eluted in the void volume (V0) on the P2 column, whereas the base hydrolysate of this product eluted at DP4 (Figure 2.4A).

Figure 2.5 Bio-gel P2 fractionation of the RP-HPLC purified [AO]7:AtGALT2 reaction product and High-Performance Anion-Exchange Chromatography (HPAEC) of the resulting base hydrolysis product. (A) Bio-gel P2 fractionation of the RP-HPLC purified [AO]7:AtGALT2 reaction product before and after base hydroylysis. Permeablized microsomal membranes from the Pichia C2 line expressing 6xHis-GALT2 served as the enzyme source in the [AO]7:AtGALT2 reaction. Elution profiles of the reaction product before and after base hydrolysis are shown. The column was calibrated with high-Mr dextran (V0), galactose (Vt), xylo- oligosaccharides with degree of polymerization (DP) 2 to 5 and XyG-oligosaccharides (DP6-9); their elution positions are indicated with arrows at the top of the figure. The

77 elution position of free Hyp amino acid (corresponding to DP3) is shown with an arrow in the panel. Base hydrolysis produces a radioactive peak eluting at DP4, which corresponds to Hyp-Gal. (B) HPAEC profile of a chemically synthesized Hyp-Gal standard detected as a PAD response. (C) The radioactive peak eluting at DP4 coelutes with the chemically synthesized Hyp-Gal standard following HPAEC. Both the Hyp-Gal standard and the radioactive peak eluting at DP4 were fractionated in 20 mM NaOH elution buffer on a CarboPac PA-20 column.

2.3.6. AtGALT2 is specific for AGPs

Various substrates that might act as potential substrate acceptors for a β -(1,3)-

GALTs were tested to investigate AtGALT2 enzyme specificity. Namely, [AO]7,

[AO]14, and d[AO]51, consisting of non-contiguous peptidyl Hyp residues, were used to examine AGP peptide sequences of various lengths. [AP]7, consisting of alternating Ala and Pro residues, was used to test the requirement of peptidyl Hyp for galactosylation.

ExtP, a chemically synthesized extensin peptide consisting of contiguous peptidyl Hyp residues, was used to test whether contiguous peptidyl Hyp residues act as potential acceptors. Two pectic polysaccharides, RGI from potato and RG from soybean, were also used as potential substrates acceptors. All the non–AGP substrate acceptors,

14 including [AP]7, failed to incorporate [ C]Gal, indicating the AtGALT2 activity was specific for AGP sequences containing non-contiguous peptidyl Hyp. It was also observed that the incorporation of [14C]radiolabel decreased with increasing lengths of these [AO] acceptor substrates (Figure 2.6).

78

Figure 2.6 Effect of various peptide and polysaccharide acceptor substrates on incorporation of [14C]radiolabeled galactose. A Permeablized microsomal membranes from the NC Pichia line transformed with the empty expression vector and the C2 Pichia line expressing 6xHis-AtGALT2 served as the enzyme source in the GALT reactions. [AO]7, [AO]14 and d[AO]51 contain seven, 14 and 51 [AO] units, respectively. A chemically synthesized extensin peptide (ExtP) contains repetitive SO4 units. [AP]7 contains seven [AP] units. Rhamnogalactan I (RGI) from potato and RG from soybean represent pectin polymer substrates. Enzyme reactions were done in triplicate and mean values are presented.

2.3.7. Biochemical Characteristics of the AtGALT2 Enzyme

To determine the preference of nucleotide sugar donors, the standard GALT assay was performed with other potential sugar nucleotides including UDP-[14C]Glc, UDP-

14 14 [ C]Xyl, and GDP-[ C]Fuc in the presence and absence of the [AO]7 peptide acceptor.

Hyp-GALT activity was only detected with UDP-[14C]Gal as the sugar donor (Figure

2.7 A).

The effects of pH and divalent cations as well as the concentrations of enzyme and substrate acceptor on the enzyme reaction were determined. With a total of 250 µg of

79 microsomal proteins in the assay system, [AO]7:AtGALT2 activity approached saturation when 20 µg of [AO]7 was included in the reaction mixture (Figure 2.7 B).

14 With 20 µg [AO]7 in the GALT assay, incorporation of [ C]Gal increased proportionally with respect to the amount of microsomal protein up to 250 µg using an incubation time of 2 h (Figure 6C). The [AO]7:AtGALT2 activity had a pH optimum of

6.5 with a HEPES-KOH buffer (Figure 2.7 D). The recombinant AtGALT2 was relatively stable, as [14C]Gal incorporation into product increased over the first 6 h before decreasing significantly (Figure 2.7 E). Finally, a divalent cation requirement for optimal enzyme activity was also observed. Mg2+ followed by Mn2+ significantly enhanced AtGALT2 activity, whereas the presence of Ca2+, Cu2+, Zn2+, and Ni2+ had inhibitory effects to different extents (Figure 2.7 F).

80

Figure 2.7 Biochemical characterization of the [AO]7:AtGALT2 activity. Data presented are an average of duplicate assays. (A) Specificity of AtGALT2 enzyme for nucleotide sugar donors was analyzed by monitoring the incorporation of 14 14 [ C]radiolabeled galactose onto [AO]7 substrate acceptor in presence of UDP-[ C]Glc, 14 14 UDP-[ C]Gal, UDP-[ C]Xyl, and GDP-[14C]Fuc. (B) Relationship between [AO]7 14 concentration and incorporation of [ C]Gal into [AO]7. (C) Relationship between 14 microsomal protein concentration and incorporation of [ C]Gal into [AO]7. (D) Effect of pH on enzyme activity. (E) Effect of increasing the reaction time on enzyme activity. (F) Effect of different divalent ions (5 mM) on enzyme activity.

2.3.8. AtGALT2 mutants have lower GALT activity and reduced β-Yariv precipitable AGPs

To provide additional in vivo evidence that AtGALT2 encodes an AGP GALT, two allelic AtGALT2 homozygous knockout mutants, galt2-1 (SALK_117233) and galt2-2

81

(SALK_141126), were obtained (Figure 2.8 A). qRT-PCR confirmed that the AtGALT2 transcripts were absent in these mutants (Figure 2.8 B). The mutants appeared to be phenotypically indistinguishable from wild type under normal growth conditions.

Biochemical analysis of the mutants, however, revealed that galt2-1 and galt2-2 respectively contained 21% and 13% less GALT activity compared to wild type plants

(Table 2.1). In addition, both mutants contained approximately 33% less β -Yariv- precipitable AGPs compared to wild type plants (Table 2.1).

Figure 2.8 Characterization of the galt2 mutants. (A) AtGALT2 gene structure and insertion sites of the T-DNAs in the galt2-1 and galt2- 2 mutants. The intron-exon structure of AtGALT2 is indicated (introns are drawn as lines and exons as rectangles with white rectangles as coding sequence and black rectangles as UTRs), and the positions of the T-DNA insertions in galt2-1 and galt2-2 are marked (triangles) as well as the position of the primers (arrows) used for screening by PCR. (B) RT-PCR analysis of transcript from the leaves of wild type (Col-0) and the two homozygous galt2 mutant lines. In the mutants no GALT2 transcript was detected in the analyzed tissues. ACTIN2-specific primers were used as controls.

82

Table 2.1 GALT activity and the amount of β-Yariv-precipitated AGPs in wild type and galt2 mutant plants Genetic Line GALT Activity β-Yariv-Precipitated (pmol/hr/mg) AGPs (µg/g) Wild type 6.7 + 0.8 14.1 + 3.5 galt2-1 5.3 + 1.2 9.6 + 3.5 galt2-2 5.8 + 1.0 9.2 + 2.8 Detergent solubilized microsomal fractions were used for performing a standard GALT assay, and AGPs were extracted, precipitated by β-Yariv reagent and quantified from 14 day old plants. The values are the averages of at least two experiments from two biological replicates. The standard deviations are indicated.

2.3.9. AtGALT2 is likely localized to the endomembrane system

To establish the subcellular localization of AtGALT2, live-cell confocal imaging of fluorescently tagged AtGALT2 protein was performed. An AtGALT2-vYFP fusion was constructed and transiently co-expressed with either a Golgi marker protein, ST-mGFP5, or an ER marker, HDEL-mGFP5, in tobacco leaves. Upon co-infiltration with the Golgi marker, AtGALT2-vYFP was not only observed as discrete punctate structures typical of a Golgi-localized staining pattern but also observed in a reticulate ER localization pattern (Figure 2.9). Co-infiltration of AtGALT2-vYFP with the ER marker revealed characteristic reticulate structures typical of an ER localization but also showed punctate

Golgi localization (Figure 2.9). There is a concern regarding transient expression experiments about over burdening the secretory system as well as specifically identifying ER versus Golgi subcellular localization, as these two membrane systems are highly connected in plants (Boevink et al., 1998). To address this inherent problem, a time course of colocalization was performed, where co-infiltrated leaf sections were

83 observed consecutively over 4 days starting from the 2nd day of infiltration. The hypothesis is that if the localization is an outcome of over burdening the endomembrane system, then over time expression the transient AtGALT2 may decrease considerably from ER and accumulate in the Golgi. However, no significant difference in co- localization between ER and Golgi over time was observed here, indicating AtGALT2 is probably present in both ER and Golgi compartments (Figure 2.10 A-E). Additionally, control images of tobacco cells expressing only ST-mGFP5, only HDEL-mGFP5, and only AtGALT2-vYFP at day 2 post infiltrations were observed to exclude spectral overlaps between YFP and GFP channels (Figure 2.10 G-I). AtGALT2 was also examined using multiple subcellular localization prediction programs (TargetP, http://www.cbs.dtu.dk/services/TargetP/) and Golgi Predictor http://ccb.imb.uq.edu.au/golgi/) and the TMHMM server

(http://www.cbs.dtu.dk/services/TMHMM/) (25) for the prediction of transmembrane domains (TMD). Based on these analyses and consistent with the live cell imaging data,

AtGALT2 is targeted to the secretory pathway and has a single N-terminal TMD.

84

Figure 2.9 Subcellular localization of AtGALT2 in tobacco leaf epidermal cells observed after 5 days of infiltration. Transiently expressed At GALT2-vYFP co-localized with sialic acid transferase (ST)- mGFP5 fusion protein (a Golgi marker) as well as with HDEL-mGFP5 fusion protein (an ER marker). These constructs were examined by laser-scanning confocal microscopy under fluorescent and white light, and the fluorescent images were merged to observe co-localization.

85

Figure 2.10 A time course of the subcellular localization of AtGALT2 in tobacco leaf epidermal cells observed after 2, 3 and 4 days post infiltration. Transiently expressed AtGALT2-vYFP co-localized with the ST-mGFP5 fusion protein (a Golgi marker) as well as with HDEL-mGFP5 fusion protein (an ER marker) at all- time points. These constructs were examined by laser-scanning confocal microscopy under fluorescent and white light, and the fluorescent images were merged to observe co-localization. Upper panels (A, B and C) show localization of AtGALT2 in the Golgi, whereas the lower panels (D, E and F) show localization of AtGALT2 in the ER. Control images of epidermal tobacco cells expressing only ST-mGFP5 (G), only HDEL- mGFP5 (H), and only AtGALT2-vYFP (I) 2 days post infiltration.

86

2.3.10. Computational modeling of AtGALT2 predicts UDP-sugar binding

A 3D structural model of AtGALT2 was created using I-TASSER and corroborated by Phyre 2 (Figure 2.11) (Roy et al., 2010; Kelley et al., 2009). I-TASSER and Phyre2 identified mouse manic fringe protein (2j0aA) as the closest structural homolog; this protein is a β1,3-N-acetylglucosaminyltransferase. COFACTOR was then used to identify putative molecular functions of AtGALT2 based on the predicted 3D structure by I-TASSER (Roy et al., 2010). COFACTOR analysis revealed that three aspartic acid residues at positions 80, 81 and 82 of AtGALT2 are involved in the binding and catalysis of a UDP-sugar donor substrate (Figure 2.12).

87

Figure 2.11 Predicted structural models of AtGALT2 generated by I-TASSER and Phyre 2. (A) Amino acid residues corresponding to the GALT domain (450-639) of AtGALT2 were used to construct these models. The following five proteins were used as templates: A. 4fixA (Mycobacterial galactofuranosyltransferase-GlfT2), (B) 2bo4A1 (Rhodothermus marinus mannosylglycerate synthase), (C) 2j0aA (mouse manic fringe in2 complexed with UDP and manganese structure), (D) 2yqcA2 (Candida albicans uridine-diphospho-N-acetylglucosamine pyrophosphorylase), E. 2d7iA (human UDP- GalNAc: polypeptide alpha-N-acetylgalactosaminyltransferase -pp-GalNAc-T10). Query structure (the GALT domain of AtGALT2) is shown in cartoon, while the structural analogs are displayed using backbone trace. (F) The homology modeling was done by Protein Homology/analogY Recognition Engine V 2.0 –PHYRE (http://www.sbg.bio.ic.ac.uk/phyre/) using the structure of the catalytic domain of the mouse manic fringe in2 complexed with UDP and manganese (PDB ID: c2j0bA) as the template. PHYRE analysis returned an estimated structural homology precision of 99% and 77% coverage.

88

Figure 2.12 A 3D structural model of AtGALT2 as predicted by the COFACTOR server available in the I-TASSER program. This 3D-model shows the predicted ligand and its binding site with a confidence score (CscoreLB) of 0.19. The highlighted residues correspond to the ‘DXD’ motif with the numbers denoting the positions of the three aspartic acid residues at positions 80, 81 and 82.

2.4. Discussion

In contrast to the considerable knowledge about the biosynthesis of cell wall polysaccharides and lignin, relatively little is known about the mechanisms involved in biosynthesis of AGPs (Tan et al., 2012; Boerjan et al., 2003). A bioinformatics approach was used to identify six putative AGP-GALTs (named GALT1 to GALT6) that act directly on the AGP protein backbone based on the finding that these are the only

Arabidopsis proteins that contain both a GALECTIN domain and a GALT domain, similar to certain mammalian GTs which O-glycosylate mucins and are composed of

89 analogous lectin and GT domains. These six candidate genes were heterologously expressed in Pichia cells and tested for Hyp-GALT activity using an in vitro Hyp-GALT assay previous developed in our laboratory (Liang et al., 2010). Only AtGALT2 has shown activity in this assay to date and thus became the focus of this investigation.

Microsomal preparations obtained from Pichia cells expressing recombinant

AtGALT2 exhibited Hyp-GALT activity catalyzing the transfer of [14C]Gal from UDP-

14 [ C]Gal onto a chemically synthesized peptide [AO]7 substrate acceptor (Figures 2.3 and 2.4). Further product characterization revealed that a single Gal residue is transferred to Hyp residues; there was no evidence for additional Galunits being added to the substrate acceptor (Figures 2.4 and 2.5). This observation is consistent with the hypothesis that O-glycosylation in plants occurs by the stepwise addition of sugar residues, as opposed to en block transfer. The recent identification and characterization of two AGP fucosyltransferases that have the ability to fucosylate AGPs lacking terminal Fuc residues is also consistent with sequential sugar addition in plants (Wu et al., 2010). These observations are consistent with O-glycosylation in animals, which is viewed to occur by sequential addition of single sugar residues to the polypeptide, as exemplified by two well-defined processes, O-mannosylation (Goto, 2007) and mucin type O-glycosylation (Hanisch, 2001). In addition, glycoengineered mammalian mucin- type O-glycosylation in transgenic plants demonstrates stepwise sugar addition (Yang et al., 2012; Castilho et al., 2012).

Furthermore, the observation that a single Galresidue is transferred to Hyp is also consistent with the hypothesis that AtGALT2 is specific for peptidyl Hyp and lacks the

90 ability to transfer additional Gal units onto peptidyl Hyp-Gal units; however, given the relatively small amount of product produced, the possibility that insufficient amounts of peptidyl Hyp-Gal substrate are available for further enzyme action cannot be excluded.

It should be noted that the Hyp-GALT activity observed here for heterologously expressed AtGALT2 in Pichia was considerably lower than that observed using plant microsomes (Liang et al., 2010). One possible explanation for this could be that multiple

Hyp-GALT enzymes, multi-enzyme complexes, and/or plant-specific cofactors are involved in the biosynthesis of AGP glycans, which are absent in Pichia cells.

AtGALT2 is specific for AGP sequences and not for other related protein sequences, including extensin with its characteristic Ser-(Hyp)4 repeat units and a non-hydroxylated

AGP-like sequence containing Pro in place of Hyp (Figure 2.6). Pectic polysaccharides contain Gal residues (Geshi et al., 2002; Peugnet et al., 2001), but pectic substrate acceptors also failed to serve as substrate acceptors for AtGALT2. In addition, Gal-

(1,3)-β-Gal-O-Me, which mimics the β-(1,3) Gal sugar backbone of AGPs, also failed to serve as a substrate acceptor for AtGALT2 (data not shown). These findings are consistent with the Hyp contiguity hypothesis, which states that non-contiguous Hyp residues are sites of AG polysaccharide addition, whereas contiguous Hyp residues are sites for the addition of Ara oligosaccharides (Egelund et al., 2011; Kieliszewski and

Shpak, 2001). Interestingly, shorter AGP peptides served as more effective substrate acceptors. While this observation lacks an explanation, it is consistent with previous findings with plant microsomes (Liang et al., 2010). It should also be noted that Strasser

91 et al. (2007) tested GALT2, as well as GALT 1, 3, 4, 5 and 6, for N-glycosylation activity, and only GALT1 was found to have such activity.

Heterologously expressed AtGALT2 in Pichia microsomes has similar biochemical properties to the GALT(s) present in Arabidopsis microsomal membranes (Figure 2.7)

(Liang et al., 2010; Oka et al., 2010). AtGALT2 is specific for UDP-Gal as the sugar donor, has a pH optimum of 6.5 (in contrast to 7 for plant microsomes), and has a requirement for Mg2+ and Mn2+ (in contrast to Mn2+ for plant microsomes) for high activity. These differences are likely a reflection of studying the properties of a single

GALT enzyme in yeast microsomes in contrast to the more complex GALT enzyme mixture in Arabidopsis microsomes that includes plant-specific factors. The observed divalent cation requirement agrees with the structural conformation of all CAZy GT31 members, which share a catalytic domain containing a DXD motif in the GT-A superfamily. In addition, the 3D protein structure of AtGALT2 predicted by I-TASSER and Phyre had as its closest match the catalytic domain of the mouse manic fringe in2 complexed with UDP and manganese (Figure 2.11 C).

Biochemical analysis of the AtGALT2 mutants provided additional in vivo evidence that AtGALT2 is indeed an AGP GALT. The absence of a mutant phenotype under normal growth conditions and the reduced GALT activity and lower β -Yariv- precipitable AGPs are consistent with gene redundancy (Figure 2.8). Other Hyp-GALTs likely compensate for the loss of AtGALT2. Apparently, the reduced GALT activity and the reduced β-Yariv-precipitable AGPs in these mutants are not sufficient to bring about a phenotypic change under normal growth conditions. Examination of these mutants

92 under non-physiological conditions or the production of multi-gene mutants within this gene family may reveal novel phenotypes in the future.

The Golgi apparatus is not only a central sorting point within the secretory pathway but also plays a central biosynthetic role in processing of complex carbohydrate structures. Subcellular localization of AtGALT2 to the ER and Golgi in tobacco leaf epidermal cells is consistent with the localization of Hyp-GALT enzyme activity to the endomembrane system in tobacco and Arabidopsis cell cultures (Figure 2.9) (Liang et al., 2010; Oka et al., 2010). In addition, bioinformatics analysis predicts that AtGALT2 is a type II membrane protein localized to the Golgi (Table 2.2). Based on these data,

AtGALT2 may initiate Hyp galactosylation of AGP protein backbone in the ER following the action of prolyl hydroxylase and continue to act in the Golgi to insure that

Hyp galactosylation is complete so as to allow for subsequent glycosylation and elongation of the AG polysaccharide.

2.4.1. An unrooted phylogenetic analysis of animal and plant GT31 members revealed three distinct cluster

Clade I comprised of plant specific GALTs devoid of a GALECTIN domain; clade

II consisted of animal GALTs involved in core β-(1,3) O-glycosylation and β-(1,3) N- acetyl glucosamine or α N-acetyl galactos-aminyltransferase (GlcNAc-T;GalNAc-T); and clade III consisted of plant specific GALTs with a Gal-binding lectin domain

(Figure 2.1). The association of a GALECTIN lectin domain with a GALT domain was conserved across several plant GT31 members, including both dicots (Arabidopsis,

Medicago truncatula, Poplus and Vitis vinifera) and monocots (Brachypodium, Zea

93 mays, rice, and Sorghum bicolor). The GALECTIN domain is defined by the presence of a conserved carbohydrate recognition domain (CRD) that specifically binds to β - galactosides, although they can display a wide range of substrate specificities due to structural heterogeneity in the CRD (Dodd and Drickamer, 2001). Although lectin domains are common in mammalian GT27 members, it is absent in mammalian GT31 members. By analogy to the mammalian GTs containing a lectin domain, the

GALECTIN domain in plants may modulate the GALT activity (Bennett et al., 2011).

Future experiments can be designed to test this hypothesis.

Two independent homology modeling methods were used to generate a predicted structure for AtGALT2 (Figure 2.11 and 2.12). First, AtGALT2 was submitted to the protein fold recognition server PHYRE (Protein Homology/analogY Recognition Engine;

Kelley and Sternberg, 2009), which were used to generate a predicted structural model

(Figure 2.11 F). In the second approach, the automated homology-modeling server I-

TASSER was utilized to generate five predicted structures for AtGALT2 (Figure. 2.11

A-E). Both PHYRE and I-TASSER generated similar homology model predictions for

AtGALT2. The outputs of these predictions were then used as a template to guide further structure-function analyses. The resulting 3D structure revealed the interaction of

AtGALT2 with a UDP-nucleotide sugar in a hydrophobic pocket containing a ‘DXD’ motif (Figure 2.12).

In summary, this study indicates that AtGALT2 (At4g21060) catalyzes galactosylation of Hyp residues in AGP protein backbones and thus represents the initial step in the biosynthesis of the polysaccharide side chains that decorate AGPs. Moreover,

94 transient expression of fluorescently tagged AtGALT2 and bioinformatics analysis indicates that this enzyme is a membrane-bound protein localized in the endomembrane system, consistent with its established biochemical function. Future studies will now focus on examining galt2 knock-out mutants in Arabidopsis, and testing for the existence of AtGALT2-containing enzyme complexes involved in AGP biosynthesis and expression of AtGALT2 and other putative Hyp-GALTs in other host systems to test for

Hyp-GALT and additional GALT enzymatic activities.

95

Table 2.2 Overview of the sequence properties of AtGALT1-6, the six putative Arabidopsis galactosyltransferases found in the CAZy GT 31 family NCBI Genea Coding Exonsa aaa TMb DXDc GALTd gleactin SignalPe Golgi regiona predictore At1g26810 3327 1932 8 644 N/A Y 406-589 172-362 Peptide Post-Golgi GALT1 At4g21060 4131 2054 9 684 70- Y 450-693 248-461 Anchor Post-Golgi GALT2 92 At3g06440 3133 1860 8 620 N/A Y 385-568 166-342 Anchor Golgi GALT3 At1g27120 2893 2022 7 674 13- Y 440-620 185-394 N/A Golgi GALT4 32 At1g74800 3145 2019 7 673 27- Y 438-618 191-389 Anchor Golgi GALT5 49 At5g62620 3196 2046 7 682 26- Y 447-627 187-401 Anchor Golgi GALT6 48 aGene, coding region exons, and number of amino acids (aa) were as annotated by The Arabidopsis Information Resource database (TAIR), with the exception of the GALT2 sequence information which was retrieved from NCBI. bThe transmembrane domain (TMD) was predicted using the TMHMM server (http://www.cbs.dtu.dk/services/TMHMM/). c DxD motifs in hydrophobic surroundings were identified using hydrophobic cluster analysis; Y-yes. d The galactosyltransferase (GALT) domain (Pfam domain Pf01762) and the GALECTIN (Pfam domain Pf00337) were identified using the Pfam server (http://pfam.sanger.ac.uk/). eSubcellular localization was predicted by SignalP (http://www.cbs.dtu.dk/services/SignalP/), TargetP 98

2.5. Materials and Methods

2.5.1. Identification of putative GALTs involved in AGP biosynthesis

A DELTA-BLAST search was performed using human β -(1,3)-GALT1 (Hs-

B3GALT2; O43825), human β -(1,3)-N-acetylgalactosaminyltransferase (O75752), human UDP-Gal:glycoprotein-N-acetylgalactosamine β -(1,3)-GALT5 (Hs-B3GALT5;

Q9Y2C3), mouse β -(1,3)-GALT1 (Mm-B3GALT1; O54904), and Caenorhabditis elegans GALT (Ce-T09F5.1; O62375) for identification of 20 Arabidopsis, 20 rice, one

Medicago truncatula, one Sorghum bicolor, eight Vitis vinifera, seven Poplus, nine

Brachypodium, and 11 Zea mays proteins (Boratyn et al., 2012). All of these proteins contain the structural motif pfam 01762 which represents the GALT domain and all of these proteins except those from Poplus, Brachypodium and Zea mays, which are yet to be included in the CAZy database, belong to the GT31 family as defined by Henrissat and Davies (22). The Poplus, Brachypodium and Zea mays proteins were instead retrieved from ARAMEMNON (http://aramemnon.botanik.uni-koeln.de/) (Schwacke et al., 2003).

Phylogenetic analysis was performed with 68 sequences from the GT31 family using the online Web service Phylogeny.fr at www.phylogeny.fr (Dereeper et al., 2008).

Multiple sequence alignments were performed by MUSCLE and PhylML for tree building whereas TreeDyn was used for tree rendering. Accession numbers presented in this study are available through the CAZy database GT31 (http://afmb.cnrs- mrs.fr/CAZY/index.html), the National Center for Biotechnology Information (NCBI, http://www.ncbi.nlm.nih.gov/) or the ARAMEMNON website. For prediction of

97 transmembrane domains (TMDs), sequences were submitted to the TMHMM 2.0 server

(Krogh et al., 2001). The protein structure was depicted using Prosite Mydomain Image creator (http://prosite.expasy.org/mydomains/). GALECTIN and GALT domains were predicted from Pfam (http://www.sanger.ac.uk/Software/Pfam/). In order to characterize the catalytic motif (DXD) of AtGALT2, hydrophobic cluster analysis (HCA) was performed using the drawhca server (http://smi.snv.jussieu.fr/hca/hca-form.html).

Homology modeling of AtGALT2 was done by the Protein Homology/analogY

Recognition Engine V 2.0 –PHYRE (http://www.sbg.bio.ic.ac.uk/phyre/) (Kelley et al.,

2009) and also by the I-TASSER server (http://zhanglab.ccmb.med.umich.edu/I-

TASSER/) (Zhang et al., 2008). First, the full length sequence of the AtGALT2 protein was analyzed, and then to test for a sugar nucleotide binding site, only the sequence corresponding to the GALT domain (450–639 amino acid residues) was analyzed. These protein modeling tools used the structure of the catalytic domain of mouse manic fringe in2 complexed with UDP and manganese (PDB ID: c2j0bA) as the template.

2.5.2. Cloning and expression of AtGALTs in Pichia pastoris

The cDNA of the coding region of four candidate AtGALTs (1, 3, 4 and 5) were obtained from the RIKEN Bioresource center. The cDNA for AtGALT6 was obtained from CNRGV - The French Plant Genomic Resource Center. A cDNA of the coding region for At4g21060 (AtGALT2) was graciously provided by Dr. Richard Strasser. The open reading frame of AtGALT2 was amplified with primers with a 5’ restriction site for SacII followed by a 6xHis tag and a 3’ restriction site for ApaI (F-GCCGCGG

ATGCATCATCATCATCATCACATGAAAAGAGTAAAAAGCGAATCTTTTA and

98

R-TCATCTGAAATTGCAACATTGTGGGGCCC. The bold letters denote the restriction sites, the shaded region denotes the 6xHis tag and the underlined region denotes the translational start site. Amplified products were sequenced, cloned in the shuttle vector pPICZ A by a traditional ‘cut and paste’ strategy, and transformed into E. coli (DH5α) for zeocin resistance. Transformed plasmids were electroporated into competent Pichia pastoris X-33 cells following manufacturer’s instructions (Invitrogen).

Twenty individual Pichia clones were selected, and the presence of the gene was confirmed by PCR using genomic DNA isolated from transformants and gene-specific primers. Genomic DNA was isolated from Pichia cells as described previously

(Hoffman and Winston Winston 1987). A similar strategy was adopted for cloning and expressing other AtGALTs in Pichia. Primers for the respective AtGALTs are listed in

Table 2.3. Ten of the 20 independent transformants were screened for expression of the recombinant AtGALT2 protein as follows: 25 mL of BMGM (buffered minimal glycerol media) supplemented with 100 mg/L of zeocin in a 250-mL flask was inoculated with a single colony and grown at 28°C in a shaking incubator at 260 rpm for ~24 h to obtain an OD600 reading of ~2. Cells were harvested by centrifugation at 2,500 x g for 5 min and resuspended in ~75 mL of BMMM (buffered minimal methanol media) to obtain an

OD600 of ~1. Protein expression was induced by adding 0.5%(v/v) methanol (final concentration) every 24 h, and 2 mL of cell cultures were harvested every 24 h for 5 d.

Cells were pelleted by centrifugation at 2,500 x g for 5 min at 4 °C and stored at -80°C until analysis.

99

Table 2.3 Primers used for cloning AtGALT1, 3, 4, 5 and 6 in Pichia Gene name Forward primer Reverse primer At1g26810 CGCCGCGGATGCATCATCATC TGCTGCCGCGAATGG GALT1 ATCATCACATGAAGAGATTTTA TAAGGGCCC TGGAGGGCTTC At3g06440 CGCCGCGGATGCATCATCATC GAACCAATCTATTTG GALT3 ATCATCACATGAAGCAATTCAT CTGCGAATAAGGGCC GTCAGTGGT C At1g27120 CGCCGCGGATGCATCATCATC CAATGCTGCAACATG GALT4 ATCATCACATGAAGAAGTCTAA AGATGAGGGCCC ACTCGATAATTC At1g74800 CGCCGCGGATGCATCATCATC GAGTGTTGTAACATG GALT5 ATCATCACATGAAAAAACCCAA AGATGATCTAGA ATTGTCG At5g62620 CGCCGCGGATGCATCATCATC CAGTGCTGCAACATG GALT6 ATCATCACATGAGGAAGCCCAA AGATGATCTAGA GTTGTCA The grey shaded area denotes the 6xHis-tag. The underlined nucleotides in the forward primer column denotes a SacII restriction site whereas the underlined nucleotides in the reverse primer column denotes an ApaI restriction site for AtGALT1, 3 and 4 and a XbaI restriction site for AtGALT5 and 6, respectively.

2.5.3. Preparation of Pichia microsomes and immunoblot analysis

Transformed Pichia cells from a 75 mL culture grown in an Erlenmeyer flask in the presence of methanol for 5 days were centrifuged at 2,500 x g for 5 min at 4°C and resuspended in 10 mL homogenization buffer (0.1 M HEPES-KOH, pH 7, 0.4 M

Sucrose, 1 mM dithiothreitol, 5 mM MgCl2, 5 mM MnCl2, 1 mM phenylmethylsulfonyl fluoride, and one tablet of Roche EDTA-free complete protease inhibitor cocktail and

100 µL RPI protease inhibitor IV). Cells were disrupted by vortexing eight times for 1

100 min each, with 2 min on ice between each vortexing, in the presence of acid-washed

425-600 µm glass beads (Sigma-Aldrich). The supernatant was centrifuged at 2,500 x g for 5 min at 4°C to remove the beads and then at 150,000 x g for 60 min at 4°C to obtain the membrane fraction (Molhoj et al., 2004). This microsomal pellet was resuspended in

50 µL of homogenization buffer. For immunoblot analysis, 5 µg of microsomal protein from Pichia transformants were denatured, subjected to 10% SDS/PAGE, and electroblotted onto PVDF Immobilon membranes (Millipore) using the Mini Protean3 system according to manufacturer’s recommendations. Blots were probed with an anti-

His primary antibody (Clontech) at a 1:10,000 dilution and a secondary goat anti-mouse

IgG antibody conjugated to horseradish peroxidase (HRP) (Clontech) at a 1:20,000 dilution. West Femto Maximum Sensitivity Substrate (Thermo Scientific) was used for

HRP detection. Pichia cell lines transformed with the empty expression vector were used as the negative control (NC). Protein quantification was done using the Bradford reagent (Sigma). Blots were stained with Commassie Brilliant Blue R 250 following

HRP detection to ensure equal loading.

2.5.4. Galactosyltranferase Assay with microsomal preparations from Pichia expressing AtGALT2

The standard GALT reaction (100 µL) consisted of detergent-permeabilized microsomal membranes (250 µg total protein), acceptor substrate peptide (20 µg), and approximately 3 µM UDP-[14C]Gal (90,000 cpm, 465 cpm/pmol; MP biomedical sciences). Permeabilization was achieved in two steps, 50 µL of microsomal protein were first treated with 0.3% Triton X-100 (15 min, 4°C), followed by ultracentrifugation

101 at 100,000 x g for 45 min. The pellet obtained was resuspended in 50 µL of extraction buffer and subjected to a second permeabilization step with 1% Triton X-100 for 15 min

° at 4 C, followed by ultracentrifugation at 100,000 x g for 45 min. [AO]7 and d[AO]51 were the two substrates acceptor used in the standard GALT assay. The reaction mixture was incubated for 2 h at room temperature and was terminated by mixing with 400 µL anion-exchange resin (DOWEX 1X8-100 resin; Sigma-Aldrich; 1:1 v/v in double distilled water). The resin mixture was loaded on a Zeba spin column (Pierce) and centrifuged at 15,000 x g for 1 min to remove unreacted UDP-[14C]Gal retained by the ion-exchange resin. The flow-through contained the incorporated [14C]radiolabeled product and was analyzed with an LS6500 multipurpose scintillation counter

(Beckman). Two reactions were included as controls, one with no substrate acceptor and one with permeablized microsomal membranes from the Pichia line (X33) transformed with the empty expression vector (pPICZ A) to serve as a negative control (NC).

2.5.5. Purification of Hyp-GALT2 reaction products by Reverse-Phase HPLC

The GALT reactionproduct was purified by RP-HPLC as described by Liang et al.

(2010).

2.5.6. Analysis of the Hyp-[14C]galactoside profile by gel Permeation chromato- graphy and HPAEC

Twenty-five standard GALT reactions were fractionated by RP-HPLC and combined to generate enough [14C]radiolabeled-product for base hydrolysis and separation on a

Biogel P2 column (Liang et al., 2010). The radioactive peak eluting at DP4 on a Biogel

P2 column was analyzed along with a chemically synthesized Hyp-Gal standard by

102

High-Performance Anion-Exchange Chroma-tography (HPAEC) on a CarboPac PA-20 column using 20 mM NaOH as the elution buffer, to provide additional confirmation of this DP4 peak as Hyp-Gal. Trans-4-(β-D-galactopyranosyloxy)-L-proline (i.e., the Hyp-

Gal standard) was chemically synthesized from commercially available galactopyranosyl bromide and hydroxyproline methyl ester as described with minor modifications (Strahm et al., 1981).

2.5.7. Monosaccharide composition analysis of GALT reaction products by High-

Performance Anion-Exchange Chromatography (HPAEC)

Fifteen standard GALT assays were pooled to generate sufficient 14C-products for acid hydrolysis and monosaccharide composition analysis as described by Liang et al.

(2010).

2.5.8. Determination of substrate specificity of the AtGALT2 enzyme activity

A standard GALT assay was performed using 20 µg of various peptide substrate acceptors [AO]7 , [AO]14 and d[AO]51 (containing seven, 14 and 51 [AO] repeating dipeptide units, respectively), an extensin peptide (ExtP) containing repetitive SO4 units and a [AP]7 peptide containing seven [AP] units as described by Liang et al. (2010).

Rhamnogalactan I (RGI) from potato and RG from soybean (100 µg each) were used as potential pectin substrates. Permeablized microsomal membranes (250 µg) from the NC

Pichia line and the C2 Pichia line expressing 6xHis-AtGALT2 served as the enzyme source in the GALT reactions. For all the peptide substrate acceptors, the standard

GALT assay was performed, and the reaction products were fractionated by RP-HPLC before monitoring incorporation of radiolabeled [14C] in a liquid scintillation counter

103

(Beckman Coulter LS 6500). For the pectin substrate acceptors, reactions were incubated at room temperature for 2h, terminated by adding 1 mL of cold 70% ethanol, and precipitated overnight at -20°C. Reaction products were collected by centrifugation at 10,000 x g for 10 min, and pellets were washed five times with 1 mL of cold 70% ethanol to remove excess UDP-[14C]Gal. The [14C]radiolabel incorporation was estimated by resuspending the pellets in 300 µL of water before counting in a liquid scintillation counter.

2.5.9. Biochemical characterization of AtGALT2 enzyme activity

The standard GALT assay was modified for AtGALT2 characterization using

[AO]7 peptide as the acceptor substrate. Assay products from each reaction were fractionated by RP-HPLC to measure incorporated [14C]radiolabel into acceptor substrates.

The optimum pH for AtGALT2 activity was determined using permeablized microsomal membranes (250 µg) from the C2 Pichia line expressing 6xHis-GALT2 dissolved in test buffers at a final concentration of 100 mM. Test buffers included MES-

KOH buffer at pH 4, 5, 6 and 7; HEPES-KOH buffer at pH 6, 6.5, 7, 7.5 and 8; Tris-HCl buffer at pH 8, 9 and 10; and CAPS-KOH buffer at pH 10.

To examine the effect of divalent cations on AtGALT2 activity, microsomal membranes were extracted with homogenizing buffer lacking divalent ions. MnCl2,

MgCl2, CaCl2, CuCl2, NiCl2, or ZnSO4 were added to the GALT assay (at a final concentration of 5 mM) when tested. Two controls were added, one with no ions in the buffer used for resuspending the detergent permealized membrane fraction and the other

104 with EDTA (5 mM) to chelate any residual divalent cations trapped in the membranes.

An equal volume of deionized distilled water was added instead of divalent ions in the control reaction.

To analyze the enzyme specificity for nucleotide sugar donors, the standard activity

14 assay was performed with [AO]7 as the acceptor substrate and various [ C]-radiolabeled nucleotide sugar donors (90,000 cpm). The nucleotide sugars tested included UDP-

[14C]Glc (MP Biomedicals), UDP-[14C]Xyl (PerkinElmer Life Sciences), and GDP-

[14C]Fuc (PerkinElmer Life Sciences). Four separate GALT reactions with no substrate acceptors were performed as controls.

2.5.10. AtGALT2 mutant analysis

Two T-DNA insertional lines for At4g21060-AtGALT2 (galt2-1 [SALK_117233] and galt2-2 [SALK_141126]) were selected using the SIGnaL database

(http://signal.salk.edu/) and were obtained from the ABRC (Arabidopsis Biological

Research Centre) (http://abrc.osu.edu/). The wild type plants were Columbia (Col-0), and galt2 mutants were in the Columbia (Col-0) genetic background. Homozygous mutants were identified by PCR analysis using primer sequences obtained with the T-

DNA Primer Design Tool provided by the Salk Institute Genomics Analysis Laboratory

(http://signal.salk.edu/tdnaprimers.2.html) (Table 2.4). To confirm homozygous plants at the transcript level, RNA was extracted and analyzed by RT-PCR. RNA was isolated using a Qiagen RNeasy plant mini kit followed by DNase I digestion using Qiagen

RNase free DNase to remove traces of DNA. The Qiagen One-Step RT-PCR kit was

105 used for first-strand synthesis and subsequent PCR steps (primers are listed in Table

2.4).

Table 2.4 Primers used for genotyping and RT-PCR analysis of AtGALT2 Gene Forward primer Reverse primer

galt2-2 ATCCTCCTTATGCCAAT GTGCAGGTCAATAG (SALK_141126.3) GGAC CAGAAGC galt2-1 TCACTTGGTCATTCCCT CAAATCGATGGAGT (SALK_117233) TTTG CTCTCCA ACTIN 2 (At3g18780) GGCTTAAAAAGCTGGG TCGTTTTGCGTTTT GTTT AGTCCC RT-PCR for GALT2a TCTTTGTTGCACTTAAT ACACAGCTGGAAA CCAAGAAAGG TTTTGCC aThe forward RT-PCR primer spans an exon-intron junction. The shaded region corresponds to one exon, and the unshaded region corresponds to the adjacent exon.

Plants were germinated after 4 days of stratification in darkness at 4°C, and grown on soil at 22°C and 60% relative humidity. Plants were grown under long-day conditions

(16 h photoperiod and 8 h dark, 120 µmol m -2 s -1 of fluorescent light).

Plant microsomal membranes were prepared and assayed according to Liang et al.

(2010). Specifically, 8 g of leaf tissue from 14 day old wild type and galt2 mutant plants were used to perform GALT reactions with [AO]7 as the peptide substrate acceptor and

UDP-[14C]Gal as the sugar donor.

AGPs were extracted from WT, galt2-1, and galt2-2 plants as described by Schultz et al. (2000). Specifically, 5 g of aerial tissue from 14 day old plants were used for each

106 line to obtain β -Yariv-precipitable AGPs, which were quantified by a spectrophotometrically as described by Gao et al. (1999).

2.5.11. Transient expression and subcellular localization of AtGALT2 in Nicotiana tabacum leaves

The AtGALT2 coding region was subcloned into the pVKH18En6-vYFP plasmid to generate the AtGALT2:vYFP construct by a traditional cut and paste strategy using

XbaI and SalI (F- CAGGACTCTAGAATGAAAAGAGTAAAAAGCGAATCTTT

TAGAGGAG and R- CATGACGTCGACTCTGAAATTGCAACATTGTGATCG

ACCTTTC) respectively. The shaded regions denote restriction sites and the underlined region denotes the translational start site. Agrobacterium-mediated transient expression was performed in the leaves of three to four week-old tobacco plants (Nicotiana tabacum cv. Petit Havana) grown at 22°C to 24°C using a bacterial optical density

(OD600) of 0.05 for single infiltrations and 0.025 each for co-infiltrations (Saint-Jore et al., 2002). The AtGALT2-vYFP construct was co-expressed with either the ER marker mGFP5-HDEL (Saint-Jore et al., 2002) or the Golgi marker ST-mGFP5 (Batoko et al.,

2000) to ascribe subcellular localization. Transformed plants were incubated under normal growth conditions and sampled daily for 2-7 days post-infiltration.

Leaf epidermal sections were imaged using an upright Zeiss LSM 510 META laser scanning microscope (Jena, Germany), using a 40 X oil immersion lens and an argon laser. For imaging the expression of vYFP constructs, the excitation line was 514 nm, and emission data were collected at 535–590 nm, whereas for mGFP5 constructs, the excitation line was 458 nm and the emission data were collected at 505-530 nm. Singly

107 infiltrated controls were analyzed to optimize gain and pinhole settings for each channel and to exclude any bleed through fluorescence between channels. Post-acquisition image processing was done using the LSM Image Browser 4 (Zeiss).

108

CHAPTER 3: TWO HYDROXYPROLINE GALACTOSYLTRANSFERASES,

GALT5 AND GALT2, FUNCTION IN ARABINOGALACTAN-PROTEIN

GLYCOSYLATION, GROWTH AND DEVELOPMENT IN ARABIDOPSIS

This work has been published in the following manuscript.

Basu D, Wang W, Ma S, DeBrosse T, Poirier E, Emch K, Soukup E, Tian L,

Showalter AM. (2015) Two Hydroxyproline Galactosyltransferases, GALT5 and

GALT2, Function in Arabinogalactan-Protein Glycosylation, Growth and Development in Arabidopsis. PLoS ONE 10(5): e0125624.

3.1. Abstract

Hyp-O-galactosyltransferase (GALT) initiates O-glycosylation of AGPs. We

previously characterized GALT2 (At4g21060), and now report on functional

characterization of GALT5 (At1g74800). GALT5 was identified using heterologous

expression in Pichia and an in vitro GALT assay. Product characterization showed

GALT5 specifically adds galactose to hypin AGP protein backbones. Functions of

GALT2 and GALT5 were elucidated by phenotypic analysis of single and double

mutant plants. Allelic galt5 and galt2 mutants, and particularly galt2galt5 double

mutants, demonstrated lower GALT activities and reductions in β -Yariv-precipitated

AGPs compared to wild type. Mutant plants showed pleiotropic growth and

development phenotypes (defects in root hair growth, root elongation, pollen tube

growth, flowering time, leaf development, silique length, and inflorescence growth),

which were most severe in the double mutants. Conditional mutant phenotypes were

109 also observed, including salt-hypersensitive root growth and root tip swelling as well as reduced inhibition of pollen tube growth and root growth in response to β-Yariv reagent.

These mutants also phenocopy mutants for an AGP, SOS5, and two cell wall receptor- like kinases, FEI1 and FEI2, which exist in a genetic signaling pathway. In summary,

GALT5 and GALT2 function as redundant GALTs that control AGP O-glycosylation, which is essential for normal growth and development.

3.2. Introduction

The fundamental processes that underpin plant growth and development depend crucially on the action and assembly of gene products designed to form the cell wall

(Carpita and Gibeaut, 1993). Cell walls are composed of cellulose, hemicellulose, and pectin, along with protein and lignin (Albersheim et al., 2011). Wall proteins have emerged as essential components because of their contribution to wall architecture and function (Albenne et al., 2013). Among the cell wall proteins, the HRGP superfamily constitutes the most abundant and diverse group of cell wall glycoproteins (Showalter,

1993). The HRGP superfamily is composed of a spectrum of molecules, ranging from lightly glycosylated proline-rich proteins to highly glycosylated AGPs with the moderately glycosylated extensins in between these two extremes (Hijazi et al., 2014).

AGPs are ubiquitous in the plant kingdom and are expressed in virtually all cells, either at the cell surface as part of the plasma membrane, the cell wall, or as extracellular secretions (Fincher et al., 1983; Nothnagel, 1997; Showalter, 2001; Seifert and Roberts,

2007). AGPs are distinguished by their abundance of Hyp, alanine, serine and residues in the protein backbone, the occurrence of Ala-Hyp, Ser-Hyp, and/or Thr-Hyp

110 dipeptide repeats, the presence of type II AG polysaccharide sides chains covalently attached to the Hyp residues, and their ability to interact with a reddish-brown chemical dye called β-Yariv reagent. The polysaccharide side chains attached to peptidyl Hyp are composed of β-1,3-galactan backbones decorated with β-1,6-galactose side chains that are further decorated with α-arabinose as well as other sugars, such as β-

(methyl)glucuronic acid, α-rhamnose, and α-fucose, which are present in lesser amounts

(Ellis et al., 2010; Kitazawa et al., 2013). An alternate model of AGP polysaccharide structure supported by NMR data indicates the galactan backbone consists of repeating trigalactosyl blocks, containing two β-1,3- attached to β-1,6-galactose, with these blocks being decorated with side chains containing β-galactose and α-arabinose along with other sugars (Tan et al., 2004; Tan et al., 2010). One or more GTsare thought to be responsible for adding each sugar in the AG polysaccharide.

Considerable progress in recent years has led to the identification of several, but by no means all, of the enzymes and their corresponding genes responsible for AGP glycosylation (Tan et al., 2012; Knoch et al., 2014). In particular, the following enzymes were identified and cloned: two α-1,2-fucosyltransferases (FUT4 and FUT6), one hyp-

O-galactosyltransferase (GALT2), one β-1,3-galactosyltransferase (At1g77810), one β-

1,6-galactosyltransferase with elongation activity (GALT31A), one β-1,6- galactosyltransferase with branch initiation and branch elongating activities

(GALT29A), and three β-1,6-gluronosyltransferases (GlcAT14A, GlcAT14B,

GlcAT14C) (Wu et al., 2010; Basu et al., 2013; Qu et al., 2008; Geshi et al., 2013;

Dilokpimol et al., 2011; Knoch et al., 2013; Dilokpimol et al., 2014). FUT4 and FUT6

111 are members of the CAZy GT-37 family; GALT2, At1g77810, and GALT31A are members of the GT-31 family; GALT29A is a member of the GT-29 family; and

GlcAT14A, GlcAT14B, GlcAT14C are members of the GT-14 family. In addition, Gille et al. (2013) identified a putative AGP β-arabinosyltransferase (RAY1) that is a member of the GT-77 family. This finding, however, is puzzling given that arabinose reportedly only exists as α-linked sugars in AGPs.

AGPs are proposed to play essential roles in a variety of plant growth and development processes, including cell expansion, cell division, reproductive development, , xylem differentiation, abiotic stress responses, and hormone signaling pathways (Seifert and Roberts, 2007; Ellis et al., 2010; Shi et al.,

2003; MacMillan et al., 2010; Griffiths et al., 2014). Most of these functions were deduced from analyzing mutants of various AGP genes or by using antibodies or β-

Yariv reagent to bind to AGPs and disrupt their function. One complication frequently encountered with AGP gene mutants is that no abnormal phenotype is observed, presumably because of gene redundancy/compensation within the AGP gene family

(e.g., there are 85 predicted AGP genes in Arabibopsis) (Tan et al., 2012; Showalter et al., 2010). Given that polysaccharides account for approximately 90% of an AGP and largely dictate the molecular surface of an AGP, it is likely that this carbohydrate moiety plays a critical role in AGP function. Thus, mutants in the genes encoding the enzymes for AGP glycosylation may provide a more informative approach to elucidating AGP function.

112

The hydroxyproline-O-galactosyltransferase (Hyp-GALT) that adds the first galactose onto the peptidyl Hyp residues in the AGP core protein is the first committed step in AG polysaccharide addition and represents an ideal control point to investigate the contribution of AG polysaccharides to AGP function. Previously, GALT2

(At4g21060) was demonstrated to encode a Hyp-GALT (Basu et al., 2013). Here, another member of GT-31, GALT5 (At1g74800) is shown to encode this same activity.

In addition, extensive phenotypic characterization of allelic galt2 and galt5 single mutants and galt2galt5 double mutants at the biochemical and physiological levels are presented which corroborate the roles of these two enzymes in AG biosynthesis and elucidate the contributions of the AG polysaccharides to AGP function.

3.3. Results

3.3.1. At1g74800 (GALT5) encodes a putative galactosyltransferase

GALT5 and GALT2 proteins are members of the diverse GT-31 family in the CAZy database (Cantarel et al., 2009; Egelund et al., 2011). In plants, GT-31 includes three clades; one with proteins having only a catalytic GALT domain, another with proteins containing both a galactosyltransferase (GALT) domain and a GALECTIN domain, and the third with proteins having a domain of unknown function (Egelund et al., 2011). In

Arabidopsis, 14 proteins have only the GALT domain, 6 proteins contain both domains, and 13 proteins have a domain of unknown function (Qu et al., 2008). Four of these

Arabidopsis GT-31 family members have been characterized. At1g26810 (GALT1) was identified as a β-(1,3)-GALT involved in biosynthesis of a Lewis a epitope on N-linked glycans (Egelund et al., 2011; Srtrasser et al., 2007). At1g77810 was reported to be a β-

113

(1,3)-GALT that catalyzes transfer of galactose (Gal) to an O-methylated Gal-β-(1,3)-

Gal disaccharide, which mimics a partial structure of AGP side chains (Qu et al., 2008).

At4g21060 (GALT2) was identified as a Hyp-GALT specific for AGPs (Basu et al.,

2013). Finally, At1g32390 (GALT31A) was shown to elongate β –1,6-galactan side chains on AGPs (Geshi et al., 2013). Given that GTs containing a lectin domain are involved in catalyzing the first step of O-glycosylation of animal glycoprotein mucins, it was hypothesized that plant GALTs containing analogous lectin domains may also function in initiating O-glycosylation of AGPs. Thus, we focused on functional characterization of such GALT genes containing a GALECTIN domain and here present our findings on GALT5. The GALT5 open reading frame is 2019 bp and corresponds to a protein of 672 amino acids, with a calculated molecular mass of 77.3 kD. The predicted protein structures and alignment of GALT2 and GALT5 are depicted in

Figure 3.1. Both proteins are predicted to be type II membrane proteins with N-terminal transmembrane domains. Thus, we hypothesized that GALT5 protein functions as an

AGP-specific Hyp-GALT.

114

Figure 3.1 Domain organization and sequence alignment of the GALT5 and GALT2 proteins. (A) Schematic diagram of GALT5 and GALT2 with their conserved domains as determined by Prosite (http://prosite.expasy.org/); the image was prepared using Prosite MyDomains (http://prosite.expasy.org/cgi-bin/prosite/mydomains/). The N-terminal

115 transmembrane, GALECTIN (Pfam PF00337) and GALT (Pfam PF01762) domains are denoted by green, yellow and red structures. The solid blue triangles denote DXD motifs. B. Protein sequence alignment of GALT5 and GALT2. Sequences were aligned using the PRALINE multiple sequence alignment program at http://www.ibi.vu.nl as described by Pirovano et al. (2001). Both the proteins have a single membrane spanning domain (GALT5, residues 27-49 and GALT2, residues 70-92), a GALECTIN domain (GALT5, residues 194-392 and GALT2, residues 248-461), and a GALT domain (GALT5, residues 465-672 and GALT2, residues 508-693). GALT5 and GALT2 showed 57% identity and 79% similarity in their protein sequences.

3.3.2. Heterologous expression of GALT5 in Pichia cells

Microsomal proteins from five independent recombinant Pichia lines expressing His tagged GALT5 were examined by immunoblotting with antibodies directed against the

6x His tag. All five GALT5 recombinant lines had the expected 77 kD protein band that reacted with the 6x His antibody (Figure 3.2 A). A non-specific, smaller protein band

(50 kD) was also detected in these recombinant lines. Transformed Pichia cells with the empty expression vector served as a negative control (NC) and lacked the recombinant

77 kD protein band, but contained the 50 kD protein band.

116

Figure 3.2 Screening for the presence of 6x His-tagged GALT5 and GALT activity in transformed Pichia cell lines. (A) Western blot analysis of the 6x His-tagged GALT5 protein in Pichia microsomal membrane preparations using the His antibody. C1 to C5 designate five independent cloned lines of transgenic Pichia cells transformed with the 6x His-GALT5 gene construct. A Pichia cell line transformed with the empty expression vector was used as the negative control (NC). (B) [AO]7-dependent GALT activity tests of the five transgenic Pichia cell lines using Triton X-100 permeablized microsomal membranes. For each line, 250 µg of total microsomal membrane protein was used for the assay. [14C]Gal radiolabel incorporation is expressed as pmol/h/mg protein and reflects the difference between total incorporation obtained in reaction products in the presence versus absence of [AO]7 acceptor substrate. Reactions were done in triplicate and mean values +SE are presented. All cell lines tested had GALT5 activity but varied in the rate of incorporation. Student’s t -test was performed using Graphpad Quickcalc (http://www.graphpad.com/quickcalcs/) and significant differences in GALT activity were detected with respect to NC (*, P <0.05 and **, P <0.01).

117

3.3.3. Heterologously expressed GALT5 demonstrates Hyp-GALT activity

An in vitro GALT assay developed by Liang et al. [29] was used to test for activity of the recombinant GALT5 expressed in Pichia cells. GALT assay components included detergent-permeablized microsomal membranes from the transformed Pichia cell lines expressing GALT5 protein as the enzyme source, UDP-[14C]Gal as the sugar donor and one of two AGP peptide analogs (d[AO]51 and [AO]7) as the substrate acceptor. The amount of GALT activity varied in the five recombinant cloned cell lines (C1 to C5) of

14 Pichia based on the rate of [ C]Gal incorporation using the [AO]7 substrate acceptor, but all were significantly higher than Pichia cells transformed with the vector alone, which served as a negative control (NC) (Figure 3.2 B).

3.3.4. Characterization of the GALT5 assay products by reverse-phase HPLC analysis

Pichia transformants expressing GALT5 were further analyzed for Hyp-GALT activity using two substrate acceptors: [AO]7, a synthetic AGP peptide and d[AO]51, a transgenically expressed and chemically deglycosylated AGP analog. Incorporation of

[14C]Gal from UDP-[14C]Gal onto the two substrate acceptors was observed by HPLC fractionation (Figure 3.3 C and 3.3 F) and by comparison to the non-radioactive [AO]7

14 and d[AO]51 substrate acceptor peaks (Figure 3.3 A and 3.3 D). Two [ C]-radioactive peaks were detected, of which peak II has the same retention time as their respective substrate acceptors ([AO]7 and d[AO]51) (Figure 3.3 C and 3.3 F). The identity of peak

I is not known; it may represent free [14C]Gal released by an endogenous galactosidase

(Liang et al., 2010) or be composed of oligosaccharides with [14C]Gal incorporated into

118 endogenous sugar acceptors as suggested previously (Kato et al., 2003; Shpak et al.,

2001). Peak I was also present in previous studies with plant (Arabidopsis and tobacco

BY2) microsomes (Liang et al., 2010). Microsomal preparations from a Pichia cell line transformed with the empty expression vector were used as negative controls (NC)

(Figure 3.3 B and 3.3 E). In summary, HPLC fractionation provided evidence for

14 14 incorporation of the [ C]radiolabel from UDP-[ C]Gal onto the [AO]7 and d[AO]51 acceptors, and the [AO]7:GALT5 reaction product was subjected to further biochemical characterization.

119

Figure 3.3 RP-HPLC fractionation of the [AO]7:GALT5 reaction products on a PRP1 reverse-phase column. Acceptor substrate alone (A and D), GALT reaction with microsomal membranes from the NC Pichia line transformed with the empty expression vector (B and E) and the GALT reaction with microsomal membranes from the transgenic Pichia C5 line (C and F) were fractionated by RP-HPLC using identical elution conditions. Radioactive Peak II coeluted with the [AO]7 and d[AO]51 acceptor substrates in the GALT5 reaction and was used for subsequent product analysis.

120

3.3.5. Product characterization by acid and base hydrolysis indicates GALT5 transfers Gal to Hyp residues

To confirm that the [14C]radiolabel remained associated with Gal, RP-HPLC

14 fractions containing the [ C]radiolabeled [AO]7:GALT5 reaction products were pooled and subjected to total acid hydrolysis. The resulting acid hydrolyzed [14C]radiolabeled monosaccharide was fractionated by HPAEC and showed that the [14C]label co-eluted

14 with Gal, thereby confirming incorporation of [ C]Gal onto the [AO]7 peptide (Figure

3.4).

121

Figure 3.4 Monosaccharide analysis of the RP-HPLC purified [AO]7:GALT5 reaction product following acid hydrolysis. Permeablized microsomal membranes from the transgenic Pichia C5 line expressing 6x 14 His-GALT5 served as the enzyme source in the [AO]7:GALT reaction. The [ C]- labeled monosaccharides were analyzed by High-Performance Anion-Exchange Chromatography (HPAEC) on a CarboPac PA-20 column. Elution times of monosaccharide standards are as indicated with arrows at the top.

In another set of experiments, base hydrolysis was used to confirm that the [14C]Gal residues were added to Hyp residues and to examine the extent of galactosylation of the

[AO]7 peptide acceptor. Base hydrolysis degrades peptide bonds, but keeps Hyp-

14 glycosidic bonds intact (Shpak et al., 2001). The intact [ C]radiolabeled [AO]7 peptide product eluted in the void volume (V0) on the P2 column, whereas the base hydrolysate of this product eluted at DP4 (Figure 2A). Given that Hyp residues alone elute as a DP3

122 sugar on a P2 column, it was concluded that GALT5 catalyzes the addition of one Gal onto the [AO]7 peptide, consistent with our previous work (Basu et al., 2013). Further confirmation of this conclusion was provided by fractionation of the base hydrolysate on a CarboPac PA-20 column and observing that the [14C]radiolabel co-eluted with an authentic Hyp-Gal standard (Figure 3.5 and 3.5).

123

Figure 3.5 Bio-gel P2 fractionation of the RP-HPLC purified [AO]7:GALT5 reaction product and High-Performance Anion-Exchange Chromatography (HPAEC) of the resulting base hydrolysis product. (A) Bio-gel P2 fractionation of the RP-HPLC purified [AO]7:GALT5 reaction product before and after base hydroylysis. Permeablized microsomal membranes from the Pichia C5 line expressing 6x His-GALT5 served as the enzyme source in the [AO]7:GALT5 reaction. Elution profiles of the reaction product before and after base hydrolysis are shown. The column was calibrated with high-Mr dextran (V0), galactose (Vt), xylo- oligosaccharides with degree of polymerization (DP) 2 to 5 and XyG-oligosaccharides (DP6-9); their elution positions are indicated with arrows at the top of the figure. The elution position of free Hyp amino acid (corresponding to DP3) is shown with an arrow in the panel. Base hydrolysis produces a radioactive peak eluting at DP4, which corresponds to Hyp-Gal. (B) HPAEC profile of a chemically synthesized Hyp-Gal standard detected as a PAD response. C. The radioactive peak eluting at DP4 coelutes with the chemically synthesized Hyp-Gal standard following HPAEC. Both the Hyp-Gal standard and the radioactive peak eluting at DP4 were fractionated in 5 mM NaOH elution buffer on a CarboPac PA-20 column.

124

3.3.6. GALT5 is specific for AGPs

Various substrates that might act as potential substrate acceptors for a GALT were tested to investigate GALT5 enzyme specificity. Namely, [AO]7, [AO]14, and d[AO]51, consisting of non-contiguous peptidyl Hyp residues, were used to examine AGP peptide sequences of various lengths. [AP]7, consisting of alternating Ala and Pro residues, was used to test the requirement of peptidyl Hyp for galactosylation. ExtP, a chemically synthesized extensin peptide consisting of contiguous peptidyl Hyp residues, was used to test whether contiguous peptidyl Hyp residues act as potential acceptors. Two pectic polysaccharides, RGI from potato and RG from soybean fiber, were also used as potential substrates acceptors. All the non–AGP substrate acceptors, including [AP]7, failed to incorporate [14C]Gal, indicating the GALT5 activity was specific for AGP sequences containing non-contiguous peptidyl Hyp. It was also observed that the incorporation of the [14C]radiolabel decreased with increasing lengths of the [AO] acceptor substrates (Figure 3.6).

125

Figure 3.6 Effect of various peptide and polysaccharide acceptor substrates on incorporation of [14C]radiolabeled galactose. Permeablized microsomal membranes from the NC Pichia line transformed with the empty expression vector and the C5 Pichia line expressing 6x His-GALT5 served as the enzyme source in the GALT reactions. [AO]7, [AO]14, and d[AO]51 contain 7, 14, and 51 [AO] units, respectively. A chemically synthesized extensin peptide (ExtP) contains repetitive SO4 units. [AP]7 contains 7 [AP] units. Rhamnogalactan I (RGI) from potato and RG from soybean represent pectin polymer substrates. Enzyme reactions using UDP-[14C]Gal as the sugar donor were done in triplicate and mean values +SE are presented.

3.3.7. Biochemical characteristics of the GALT5 enzyme

To determine the preference of nucleotide sugar donors, the standard GALT assay was performed with other potential sugar nucleotides including UDP-[14C]Glc, UDP-

14 14 [ C]Xyl, and GDP-[ C]Fuc in the presence and absence of the [AO]7 peptide acceptor.

Hyp-GALT activity was only detected with UDP-[14C]Gal as the sugar donor (Figure

3.7 A). The effects of pH and divalent cations on the GALT assay catalyzed by GALT5

126 were also determined. The [AO]7:GALT5 activity had a pH optimum of 6.5 with a

HEPES-KOH buffer, which is consistent with the lumen of Golgi vesicles where the enzyme is predicted to be localized (Figure 3.7 B). Mg2+ followed by Mn2+ significantly enhanced GALT5 activity, whereas the presence of Ca2+, Cu2+, Zn2+, and Ni2+ had inhibitory effects to different extents (Figure 3.7 C).

Figure 3.7 Biochemical characterization of the [AO]7:GALT5 activity. Data presented are an average of duplicate assays. A. Specificity of the GALT5 enzyme for nucleotide sugar donors was analyzed by monitoring incorporation of 14 14 [ C]radiolabeled galactose onto [AO]7 substrate acceptor in presence of UDP-[ C]Glc, UDP-[14C]Gal, UDP-[14C]Xyl, and GDP-[14C]Fuc. (B) Effect of pH on enzyme activity. (C) Effect of different divalent ions (5 mM) on enzyme activity.

127

3.3.8. GALT5 is localized to the Golgi

To establish the subcellular localization of GALT5, live-cell confocal imaging of fluorescently tagged GALT5 protein was performed. A GALT5-YFP fusion was constructed and transiently co-expressed with a Golgi marker protein, sialyltransferase

(ST)-GFP, or an ER marker, HDEL-GFP, in tobacco leaves. Upon co-infiltration with the Golgi marker, GALT5-YFP was observed to co-localize with the Golgi marker as discrete punctate structures typical of a Golgi-localized staining pattern (Figure 3.8).

Whereas upon co-infiltration with the ER marker, GALT5-YFP was not observed in the characteristic reticulate structures typical of ER localization (Figure 3.8). GALT5 was also examined using multiple subcellular localization prediction programs (TargetP, http://www.cbs.dtu.dk/services/TargetP/) and Golgi Predictor http://ccb.imb.uq.edu.au/golgi/) and the TMHMM server

(http://www.cbs.dtu.dk/services/TMHMM/) (Krogh et al., 2001) for the prediction of transmembrane domains (TMD).

128

Figure 3.8 Subcellular localization of GALT5 in tobacco leaf epidermal cells observed 5 days after infiltration. Transiently expressed GALT5-YFP co-localized with sialyltransferase (ST)-GFP fusion protein (a Golgi marker) as well as with HDEL-GFP fusion protein (an ER marker). These constructs were examined by laser-scanning confocal microscopy under fluorescent and white light, and the fluorescent images were merged to observe co- localization. Bar= 10µm.

3.3.9. Isolation of T-DNA insertion alleles for the GALT2 and GALT5 genes

To elucidate the in vivo functions of GALT2 and GALT5 in Arabidopsis, a reverse genetic approach was adopted. Two independent mutant alleles were isolated for each of the genes, galt2-1 and galt2-2 for GALT2 and galt5-1 and galt5-2 for GALT5.

Homozygous lines were identified by PCR analysis and T-DNA insertion sites were confirmed by sequencing (Figure 3.9 A). Testing for genetic redundancy was addressed by crossing galt2 and galt5 single mutants and using PCR to screen for galt2galt5 double mutants in the resulting F2 generation. RT-PCR analysis showed that the GALT2 transcript was absent in both galt2 allelic mutants as well as in the double mutant and

129 that the GALT5 transcript was absent in both galt5 allelic mutants as well as in the double mutant (Figure 3.9 B and 3.9 C). The qRT-PCR analysis corroborated these findings and confirmed the identification of allelic knock-out galt2, galt5 single mutants as well as galt2galt5 double mutants (Figure 3.9 D and 3.9 E).

130

A

Figure 3.9 Molecular characterization of galt single and double mutants. (A) GALT2 and GALT5 gene structure and T-DNA insertion sites in galt2-1, galt2-2, galt5-1 and galt5-2 mutants. The intron-exon structures of GALT2 and AGALT5 are indicated (introns are drawn as lines and exons as rectangles, with white rectangles representing coding sequences and black rectangles representing UTRs). Sites of T- DNA insertions in galt2 and galt5 are marked (triangles) as are the locations of primer sequences (arrows) used for PCR screening. (B) RT-PCR analysis of transcripts from rosette leaves of 14-d-old wild type (Col-0), the allelic homozygous galt2 and galt5 mutant lines. Arrows indicate the position of primers used for RT-PCR analysis of transcript levels. UBQ10 primers were used as internal controls. (C) RT-PCR analysis of transcripts from rosette leaves of 14-d-old wild type (Col-0), the homozygous galt2 and galt5 mutant lines used for producing the double mutant, and the galt2galt5 double mutant. (D) and (E) Quantitative RT-PCR analysis to detect GALT2 and GALT5 transcript abundance in the galt mutants. Total RNA was isolated from rosette leaves of 14-d-old wild type, galt2-1, galt2-2, galt5-1, galt5-2, and galt2galt5 plants. UBQ10 primers were used as controls. Data were normalized to the level of wild type GALT2 expression in panel D and wild type GALT5 expression in panel E, which was set to 1 arbitrary unit (a.u.) in each case. Means ± SE of three biological replicates (n = 3) are shown.

131

3.3.10. GALT2 and GALT5 have overlapping but distinct expression patterns

To analyze the spatial and developmental expression of GALT2 and GALT5, RNA was isolated from different organs and tissues and analyzed by qRT-PCR (Figure 3.11

A). Both GALT genes were ubiquitously expressed in Arabidopsis in an overlapping, but distinct pattern. GALT2 was highly expressed in root and stem, whereas GALT5 was highly expressed in stem, root and leaf. These findings are consistent with data from publicly available expression databases (Figure 3.10). Transcriptomics analysis using

GeneCAT (http://genecat.mpg.de) (Mutwil et al., 2008), Genevestigator

(http://www.genevestigator.com/gv/) (Zimmermann et al., 2004 and the Arabidopsis eFP browser (http://bar.utoronto.ca/efp/cgi-bin/efpWeb.cgi) (Winter et al., 2007) indicate both genes are widely expressed with GALT5 having a higher overall expression compared to GALT2. It is also noteworthy that these databases indicate that GALT5 is highly expressed in mature pollen (Figure 3.11 A -C).

132

133

Figure 3.10 Expression profiles of GALT2 and GALT5 in publicly available databases. Expression profiles of GALT2 and GALT5 as depicted by (A) GeneCAT and (B) Genevestigator and (C) eFP browser. Both genes display expression in root and mature pollen as denoted by red arrows.

3.3.11. Compensatory mechanism of GALT2 and GALT5

To investigate whether transcriptional compensation occurs between GALT2 and

GALT5 or with the other four members of the Arabidopsis GT-31 family encoding both

GALT and GALECTIN domains, qRT-PCR analysis was conducted using the galt2,

134 galt5 and galt2galt5 mutants (Figure 3.11 B and C). While GALT2 and GALT5 transcripts were absent in their respective mutant lines, significant increases in the abundance of GALT4, GALT5, and GALT6 transcripts in the galt2 mutants and GALT2,

GALT3, GALT4, and GALT6 transcripts in the galt5 mutants were observed. The compensation mechanism was also observed in the galt2galt5 double mutants, with increases in the abundance of GALT6, GALT4, and GALT3 transcripts over that seen in the single mutants. These results indicate transcriptional compensation within this GT-

31 clade with the notable exception of GALT1, whose expression is unchanged in the mutant backgrounds. Interestingly, GALT2 and GALT5 demonstrated the most pronounced increases in transcriptional compensation.

135

Figure 3.11 Organ-specific expression of GALT2 and GALT5 and gene compensation in galt2, galt5 and galt2galt5 mutants observed by qRT-PCR. (A) Organ-specific relative expression of GALT2 and GALT5 genes. qRT-PCR was performed with total RNA samples from roots, stem, inflorescence, silique, seedling, cell culture, cauline leaves, and juvenile rosette leaves. The averages of three biological replicates are shown. The y axis shows x-fold expression with respect to the lowest encountered value of GALT2 expression in seedling equal to one arbitrary unit (a.u.). (B) Functional compensation of galt2 and galt2galt5 as revealed by qRT-PCR analysis. Data were normalized to the level of GALT2 expression in seedlings of wild type, which was set to 1 arbitrary unit (a.u.). RNA was isolated from 14-d-old seedlings. (C) Functional compensation of galt5 and galt2galt5 as revealed by qRT-PCR analysis. The expression values were normalized to the level of GALT5 expression in seedlings of wild type, which was set to 1 arbitrary unit (a.u.). UBQ10 was used as an internal control for all the

136 qRT-PCR experiments. The asterisks indicate significant differences in expression of transcripts of the six GALTs tested compared with wild-type according to a Student's t test (*, P < 0.05; **, P < 0.01, ***, P<0.001).

3.3.12. Biochemical phenotypes of the mutants: GALT activity, β -Yariv- precipitable AGPs and immunolabeling with AGP specific monoclonal antibodies

To provide in vivo evidence that GALT5 encodes an AGP GALT and examine potential functional redundancy with GALT2, the two allelic galt5 mutants were analyzed along with the galt2 mutants and the galt2galt5 double mutant with respect to

GALT activity and content of β -Yariv-precipitable AGPs (Table 3.1). GALT activity was reduced by 22% and 28% in the two galt5 mutants compared to wild type plants.

This was similar to the 21% and 14% reductions for the galt2 mutants. Double mutants, however, demonstrated a 34% reduction in activity. In addition, the galt5 mutants had

40% and 43% less β-Yariv precipitable AGPs compared to wild type, while the galt2 mutants had reductions of 32% and 35%. The double mutants, however, contained 56% less β-Yariv precipitable AGPs. The profiles of these β-Yariv precipitable AGPs were also examined by HPLC, and revealed that virtually all these AGPs, as opposed to a single or subset of these AGPs, were affected in the single and double mutants, with the double mutant being more severely affected (Figure 3.12). It was observed that the AGP peaks in the mutants eluted later and had less protein than the wild type AGP peaks, corresponding to reduced glycosylation.

137

Table 3.1 GALT activity and amount of β-Yariv precipitated AGPs in WT, galt2, galt5, and galt2galt5 mutants Genotype GALT activity β-Gal Yariv precipitated (pmol/hr/mg) AGP (µg/g) WT 6.70 + 0.79 13.92 + 3.75 galt2-1 5.30 + 1.20a 9.91 + 2.80a galt2-2 5.80 + 1.01a 9.78 + 3.50a galt5-1 5.25 + 2.20a 7.90 + 6.10b galt5-2 4.80 + 3.50b 8.10 + 3.20b galt2galt5 4.44 + 0.44b 5.83 + 0.59b Detergent-solubilized microsomal fractions were used for performing a standard GALT 14 assay using [AO]7 as the peptide substrate acceptor and UDP-[ C]Gal as the sugar donor, and AGPs were extracted, precipitated by β-Yariv reagent, and quantified from 14-d-old plants. The values are averages of at least two independent experiments from two biological replicates. Student’s t tests were performed to determine statistical significance (a P < 0.05, b P <0.01).

138

Figure 3.12 Profiles of AGPs extracted from WT, galt2, galt5 and galt2galt5 mutants and separated by RP-HPLC. HPLC chromatograms of (A) [AO]7 peptide (B) AGPs from WT, (C) AGPs from galt2- 1, (D) AGPs from galt5-1,and (E) AGPs from galt2galt5. Arrows indicate the most prominent AGP peaks in the chromatographs.

In addition, immunofluorescence staining of AGP epitopes was performed to confirm that reduced levels of glycosylated AGPs were present in the galt2galt5 mutants

139

compared to wild type. The galt2galt5 double mutant displayed reduced labeling

intensity using four AGP specific monoclonal antibodies namely, JIM4, JIM8, JIM13

and MAC207 in root hairs, pollen tubes and seeds compared to the strong signals

displayed by the corresponding wild type samples (Figure 3.13). It should be noted that

Arabidopsis pollen tubes do not react with JIM13, as previously reported by Dardelle et

Aal. (2010).

140

Figure 3.13 Immunofluorescent labeling of galt2galt5 and wild-type root hairs, pollen tubes and seeds using AGP specific monoclonal antibodies JIM4, JIM8, JIM13 and MAC207. (A) Confocal microscopy images of WT and galt2galt5 root hairs (B) pollen tubes and (C) seeds. Significant reduction in signal intensity was observed in the galt2galt5 samples compared to the WT. All immunohistochemical experiments were repeated twice with ten seedlings for root hairs, ten flowers with five pollen tubes and 50 seeds for each genotype respectively.

141

3.3.13. Pleiotropic growth and development phenotypes of the mutants

While the galt2 and galt5 single mutants were largely indistinguishable from wild type, the double mutants displayed several altered phenotypes related to growth and development under normal growth conditions (Table 3.2, Figure 3.14, Figure 3.15, and

Figure 3.16,). Such alterations were reflected in the larger number of rosette leaves, increased flowering time, reduced silique length, and reduced plant height (Table 3.1, and Figure 3.14). Root hair length and density were reduced in some single mutants and in the double mutant (Figure 3.15). Specifically, galt2-1 and galt5-1 showed a reduction in root hair length, as did the double mutant, while root hair density was reduced in galt2-1, galt2-2, and galt5-1 along with the double mutant. In addition, pollen tube frequently associated with disruption of pollen tube tip growth (Figure 3.16) and was slightly inhibited in the double mutant (Figure 3.17) and Root growth was also inhibited in some single mutants (galt5-1 and galt5-2) and in the double mutant (Figure 3.18).

Several conditional phenotypes were also examined and revealed marked differences in the mutants as described in the next section.

142

Table 3.2 Comparisons of various developmental phenotypes displayed by WT, galt2, galt5, and galt2galt5 mutant plants Genotype Rosette Cauline Flowering Silique Plant height leaves (#) leaves (#) time (days) length (cm) (mm) WT 9.6 + 0.36 3.6 + 0.09 22.4 + 0.49 13.3 +0.25 40.20 + 0.82 galt2-1 10.8 + 0.41 3.4 + 0.36 23.4 + 0.45 12.7 +0.15 40.42 + 0.63 galt2-2 11.1 + 0.59 3.8 + 0.36 23.4 + 0.32 12.9 +0.55 40.35 + 0.51 galt5-1 10.7 + 0.59 3.4 + 0.34 22.8 + 1.44 12.4 +0.30 40.25 + 0.46 galt5-2 11.0 + 0.15 3.9 + 0.60 23.3 + 0.25 12.9 +0.30 40.16 + 0.32 galt2galt5 18.9 +0.83a 4.0 + 0.05 26.9 + 0.52a 10.5+0.18a 36.42+0.68a All measurements are means +SE of 15 to 20 plants per genotype. Statistically significant differences were determined by performing Student’s t tests (P < 0.05) using Graphpad Quickcalcs (http://www.graphpad.com/quickcalcs/). Plants were grown under long day conditions. Plant height and silique length were measured from 40-d-old plants.

143

Figure 3.14 Morphological phenotypes of WT, galt2, galt5 and galt2galt5 mutants. (A) Phenotypes of the indicated seedlings grown in vitro on MS media (B) and adult plants grown in soil during long days. Key developmental stages were monitored according to Boyes et al. 2001. Data presented are combined from three experimental replicates. Bars represent averages of 25 or more plants ± SE (Student's t test, *, P <0.05 and **, P <0.01). C. Plants at 28 d after germination.

144

Figure 3.15 Root hair length and density reduced in the galt2galt5 double mutant. (A) Wild type, galt2-1, galt5-1, and galt2galt5 plants were grown on MS agar plates for 10 d with 1 % sucrose or (B) with 4.5 % sucrose. Bars = 1mm. (C) Quantification of root hair length and (D) density of the galt mutants. The asterisks indicate significantly reduced root hair length and density compared with wild-type controls according to a Student's t test (*, P < 0.05; **, P < 0.01; n > 300).

145

Figure 3.16 Disruption of tip growth in pollen tubes of galt2, galt5 and galt2galt5 mutants. Disruption of tip growth in pollen tubes as observed after 8 h of in vitro germination of pollen in the germination media. WT pollen tubes did not display tube disruption and were used as a control. Bar = 50 µm.

146

Figure 3.17 The galt single and double mutants demonstrate reduced inhibition of pollen tube growth in response to β-Yariv reagent. (A) Representative images of pollen tubes from wild-type, galt2, galt5, and galt2galt5 mutants after 16 h in pollen germination medium, and (B) in pollen germination medium supplemented with 30 µM β-Yariv and (C) in pollen germination medium supplemented with 30 µM α -Yariv reagent. Bar = 30 µm. (D) Pollen tube lengths (from wild-type, galt2, galt5, and galt2galt5 plants) were measured over time in the pollen germination medium E in pollen germination medium supplemented with 30 µM α-Yariv reagent and F. in pollen germination medium supplemented with 30 µM β -Yariv reagent. Twenty flowers from each genotype and 25 pollen tubes from each flower were measured using Image J. The experiment was done in triplicate and the values were subjected to statistical analysis by ANOVA, followed by the Tukey's honestly significant difference test. In response to β -Yariv reagent, WT pollen tubes were significantly shorter than pollen tubes from single mutants (P <0.05) and galt2galt5 double mutants (P <0.01).

147

Figure 3.18 Reduced inhibition of primary root growth of galt2, galt5 and galt2galt5 mutants in the presence of β-Yariv reagent. (A) Root lengths of WT, galt2, galt5, and galt2galt5 plants were measured 7, 14 and 21 d after germination and seedling establishment for 5 d on MS plates, on MS plates supplemented with 50 µM α-Yariv reagent, and on MS plates supplemented with 50 µM β-Yariv reagent. Statistical differences were determined by one way ANOVA, followed by the Tukey's honestly significant difference test. Asterisks represent the statistical significance between genotypes (*, P < 0.05; **, P < 0.01; ***, P <0.001) within a treatment group. Vertical bars represent mean ± SE of the experimental means from at least three independent experiments (n =5), where experimental means were obtained from 10 to 15 seedlings per experiment. (B) Representative images of WT, galt2, galt5, and galt2galt5 plants after 14 d of growth on MS plates supplemented with 50 µM β- Yariv reagent. (C) Representative images of WT, galt2, galt5, and galt2galt5 plants after 14 d of growth on MS plates supplemented with 50 µM α-Yariv reagent. Size bar = 1 cm.

148

3.3.14. Mutants demonstrate reduced inhibition of pollen tube growth and root growth in the presence of β-Yariv reagent

β-Yariv reagent is known to inhibit pollen tube growth by disrupting AGPs (Cheung et al., 2002; Mollet et al., 2002; Pereira et al., 2006). Here, pollen from galt2 and galt5 single mutants as well as the double mutant were germinated and grown in β -Yariv reagent (Figure 3.18). Wild type pollen was used as a control and showed reduced pollen tube growth in the presence of β-Yariv reagent as expected. In contrast, pollen tube growth was less inhibited in the single mutants, and was even less inhibited in the double mutant. In other words, the mutants showed less sensitivity to β-Yariv-induced pollen tube growth inhibition. α-Yariv reagent, which does not bind to AGPs but is similar in structure to β-Yariv reagent, was used as another control treatment in these experiments and produced results identical to the unsupplemented control treatment. It should be noted that this experiment also revealed other non-conditional mutant phenotypes in the control treatment, namely pollen tube growth was slightly inhibited in the double mutant (Figure 3.17) and frequently associated with disruption of pollen tube tip growth (Figure 3.16), as mentioned in the previous section.

β-Yariv reagent is equally well known to inhibit root growth by disrupting AGPs

(van Hengel and Roberts, 2002; Ding and Zhu, 1997; Willats and Knox, 1996). Here, galt2 and galt5 single mutant seedlings as well as double mutant seedlings were grown in the presence of β-Yariv reagent (Figure 3.18). Wild type seedlings were used as a control and showed reduced root growth in the presence of β-Yariv reagent as expected.

In contrast, the single mutants showed a β-Yariv insensitive root growth phenotype, and

149 the double mutant displayed even greater β -Yariv insensitivity with respect to root growth. α-Yariv reagent was used as another control treatment in these experiments and produced results identical to the unsupplemented control treatment. This experiment also revealed another non-conditional mutant phenotype in the control treatments at 21 d, namely root growth was inhibited in some single mutants (galt5-1 and galt5-2) and in the double mutant (Figure 3.18), as mentioned in the previous section.

3.3.15 Mutant seed germination and root growth are hypersensitive to NaCl

Seed germination and root growth are known to be impaired in response to salt (Zhu et al., 1998). Here, the double mutants showed a significant reduction in seed germination compared to wild type, while the single mutants showed some reduction in seed germination, which was not statistically significant (Figure 3.19). Hypersensitive root growth was observed in the presence of varying concentrations of NaCl, KCl, and

LiCl, but not in the presence of CsCl and mannitol (Figure 3.20). Root growth was also inhibited in the single mutants, and even more inhibited in the double mutant over a 21 d time course, in response to 100 and 150 mM NaCl (Figure 3.21).

Single and double mutants were also subjected to a root-bending assay, which is routinely used to screen salt-hypersensitive mutants or transgenic plants (Zhu et al.,

1998). WT plants readily reoriented root growth, whereas the single mutants showed delayed root bending with the double mutants showing a greater delay (Figure 3. 22).

Other known salt-hypersensitive mutants, including sos5, which encodes a fasciclin-like

AGP called FLA4, and fei1 and fei2, which encode cell wall receptor like kinases which interact with sos5, were also tested in this root bending assay as they may be related to

150 galt2 and galt5 and showed various degrees of delayed root bending (Shi et al., 2003;

Xu et al., 2008) (Figure 3.22 D). The angle of root curvature in the single mutants and to a larger extent in the double mutants was observed to be greater than that of the WT due to their delayed response towards salt stress (Figure 3. 22 E).

Root tip swelling in response to 100 mM NaCl was observed in the galt single mutants; this swelling was even more pronounced in the double mutant (Figure 3.23).

Other mutants, including sos5, fei1, fei2, fei1fei2, and sos5fei1fei2, were also examined and demonstrated root tip swelling in response to salt as reported previously (Xu et al.,

2008).

151

Figure 3.19 Delayed germination of galt2, galt5 and galt2galt5 seeds in the presence of NaCl by using radicle length as an indication of delayed germination. WT, galt2, galt5 and galt2galt5 seeds were germinated on MS medium supplemented with either 100 mM NaCl (A) or 150 mM NaCl (B). Radicle lengths were measured over time and standard errors were calculated from at least three independent experiments, each of which contained 50 seeds of WT and 50 seeds of each mutant. Emergence of the radicle was considered the indication of germination. Radicle length was measured by Motic Image version 3.2.

152

Figure 3.20 Sensitivity of galt2 and galt5 seedlings to various salt and osmotic stresses as measured by root growth. WT, galt2-1, galt2-2, galt5-1, and galt5-2 seeds were grown on MS medium with 1% sucrose for 5 d. Root growth was measured 7 d after transferring the seedlings to MS plates supplemented with various concentrations of (A) NaCl, (B) KCl, (C) LiCl, (D) CsCl, and (E) mannitol. Error bars indicate SEs (n = 25). The experiments were repeated at least three times with similar results

153

Figure 3.21 Salt induced inhibition of primary root elongation in galt2, galt5 and galt2galt5 mutants. Five-day-old wild-type, galt2, galt5 and galt2galt5 seedlings germinated on MS medium were transferred onto media containing (A) 100 mM NaCl or (B) 150 mM NaCl and grown vertically. Root elongation (i.e., increase in length after transfer) was measured after 7, 14 and 21 d of growth. Data are the means ± SE of measurements from five independent experiments (total n = 100). Statistical differences were determined by one way ANOVA, followed by the Tukey's honestly significant difference test (*, P <0.05 and **, P <0.01).

154

Figure 3.22 Root-Bending assay of wild type, galt, sos5, and fei mutant seedlings. Five-day-old seedlings grown on MS plates were transferred to MS plates with 100 mM NaCl and reoriented at an angle of 180° (upside down). The photographs were taken 3 d (A), 5 d (B) and 10 d (C and D) after seedling transfer. Bar = 10 mm. (E) Analysis of root curvature in WT, galt, fei1, fei2 and sos5 mutant plants. Statistical differences were determined by one way ANOVA and ‘a’ denotes a significant difference of root curvature (P<0.05) between WT and single galt mutants, ‘b’ denotes a significant difference of root curvature (P<0.01) between galt single mutants and galt2galt5, fei1fei2, sos5 and sos5fei1fei2 mutants, and ‘c’ denotes a significant difference of root curvature (P<0.001) between WT and galt2galt5, fei1fei2, sos5 and sos5fei1fei2 mutants. Vertical bars represent mean ± SE of the experimental means from at least two

155 independent experiments (n =5), where experimental means were obtained from 15 seedlings per experiment.

Figure 3.23 Conditional root anisotropic growth defects of galt, sos5, and fei mutants. Light microscopic images of root tips of plant seedlings from indicated genotypes grown for 10d in MS plates with 100 mM NaCl. Seeds were germinated in MS plates and grown for 3d before transferring to the MS plates with 100 mM NaCl. Bar =1mm.

3.3.16. Double mutant (galt2galt5) displays less seed coat mucilage

Calcoflour white, which stains cellulose, and ruthenium red, which stains pectin, were used to stain galt single and double mutant seeds along with sos5 and sos5fei1fei2 mutant seeds to examine seed coat mucilage (Figure 3.24). The double mutant displayed reduced cellulose ray staining and reduced pectin staining in the mucilage adhering to the seeds compared to the wild type and galt2 and galt5 single mutants. This double mutant seed phenotype was similar to that displayed by sos5 and sos5fei1fei2 (Griffiths et al., 2014; Harpaz-Saad et al., 2011).

156

Figure 3.24 Staining of seed coat mucilage for cellulose and pectin in wild type, galt, sos5, and fei mutant seeds. Seeds of the indicated genotypes were prehydrated with water and stained with Calcofluor white and ruthenium red to visualize cellulose and pectin with a Zeiss LSM 510 META laser scanning confocal microscope.

157

3.4. Discussion

3.4.1. GALT5 is an AGP Hyp-GALT and other AGP GTs

Biochemical and genetic evidence are presented here indicating that GALT5, similar to GALT2, functions as an AGP-Hyp-O-GT (Basu et al., 2013). Detergent permealized microsomal preparations from Pichia cells expressing GALT5 exhibit Hyp-GALT activity, catalyzing transfer of [14C]Gal from UDP-[14C]Gal onto both a chemically synthesized peptide [AO]7 and endogenously produced HF-deglycosylated d[AO]51 substrate acceptors (Figure 3.3). Product characterization revealed that a single Gal residue is transferred to Hyp residues, as was the case for GALT2 (Figure 3.4 and 3.5).

This observation is consistent with the hypothesis that O-glycosylation in plants occurs by the stepwise addition of sugar residues, as opposed to en block transfer that is characteristic of N-glycosylation.

Hyp-GALT activity observed here using heterologously expressed GALT5 in Pichia is considerably lower than observed using plant microsomes, but is consistent with our previous findings with GALT2 (Basu et al., 2013; Liang et al., 2010). One possible explanation for this could be that multiple Hyp-GALT enzymes, multi-enzyme complexes, and/or plant-specific cofactors are involved in the biosynthesis of AGP glycans, which are absent in Pichia cells.

Substrate specificity of GALT5 was investigated using various potential acceptor substrates and demonstrated that GALT5 is specific for AGP sequences (Figure 3.6).

These findings are consistent with the Hyp contiguity hypothesis, which states that clustered, non-contiguous Hyp residues are sites of AGpolysaccharide addition, whereas

158 contiguous Hyp residues are sites for the addition of Ara oligosaccharides (Kieliszewski et al., 1995; Kieliszewski and Shpak, 2001). Heterologously expressed GALT5 in Pichia microsomes has similar biochemical properties to the GALT(s) present in Arabidopsis microsomal membranes and GALT2 (Figure 3.7) (Basu et al 2013; Liang et al., 2010).

GALT5 specifically requires UDP-Gal as the sugar donor, has a pH optimum of 6.5 (in contrast to 7 for plant microsomes), and has a requirement for Mg2+ and Mn2+ (in contrast to Mn2+ for plant microsomes) for its optimal activity. These differences are likely a reflection of studying the properties of a single GALT enzyme in Pichia microsomes in contrast to the more complex GALT enzyme mixture in Arabidopsis microsomes that includes plant-specific factors and is consistent with the biochemical characteristics of GALT2 (Basu et al., 2013).

Genetic mutant analysis provides additional in vivo evidence that GALT5 functions as an AGP-Hyp-GALT, which is functionally redundant to GALT2 (Table 3.1). Two allelic galt5 knock-out mutants have reduced Hyp-GALT activity and contain considerably less glycosylated (i.e., β-Yariv precipitiable) AGPs. Allelic galt2 knock- out mutants demonstrate similar biochemical phenotypes, while galt2galt5 double mutants possess even less enzyme activity and glycosylated AGPs compared to the single mutants. In addition, HPLC AGP profiling of the galt2, galt5, and galt2galt5 mutants extends these finding and indicates that GALT2 and GALT5 activity is not limited to a particular AGP or a small subset of AGPs, but instead broadly act on coexpressed AGPs (Figure 3.12). Immunofluorescent labeling of AGPs in root hairs, pollen tubes and seeds demonstrates that disruption of GALT2 and GALT5 led to the

159 biosynthesis of AGPs with reduced glycosylation in different organs (Figure 3.13). This finding is consistent with the reduced GALT activity and amounts of β-Yariv precipitiable AGPs in the gat2galt5 mutants as well as the HPLC profiling of the AGPs obtained from the galt2galt5 mutants. In this context, it should be noted that GALT2 and

GALT5 have overlapping patterns of gene expression and demonstrate transcriptional compensation when either one or both genes are knocked out (Figure 3.10 and 3.11).

Identification of GALT5 as an AGP Hyp-GALT adds to the growing list of the enzymes responsible for AGP glycosylation (Figure 3.25 and Table 3.3). Currently, there is biochemical and/or genetic evidence for eleven AGP GTs residing in multiple

GT families, including two Hyp-O-GALTs in GT31 (GALT2 and GALT5), one β-1,3-

GALT in GT31 (At1g77810), one β-1,6-GALT in GT31 (GALT31A), one β-1,6-GALT in GT29 (GALT29A), three β -1,6-GlcATs in GT14 (GlcAT14A, GlcAT14B,

GlcAT14C), two α-1,2-FUTs in GT37 (FUT4 and FUT6), and one β -AraT in GT77

(RAY1). Several AGPGTs, however, remain to be cloned and identified, including α-

AraTs, α-rhamonsyltransferases, and α-xylosyltranferases, as well as additional enzymes related to those listed above. It will be particularly interesting to know the number of enzymes responsible for synthesizing the β -1,3 and β -1,6 galactose chains and their detailed substrate specificities, particularly with respect to processivity. It is unknown whether these enzymes exist in a large biosynthetic complex, although a recent study reports that GALT31A and GALT29A interact with one another (Dilokpimol et al., 2014).

160

Figure 3.25 Sites of action of known GTs acting on AGPs are depicted within a representative AGP glycomodule sequence found within an AGP molecule. This glycomodule structure is based on information presented by Tryfona et al. 2014. Additional details on each of the known GTs are listed in Table 3.

Table 3.3 Information on the known enzymes, genes, and mutants for AGP glycosylation Gene Enzyme GT Gene Mutants Localization Mutant References name Family Identifier phenotypes GALT2 hydroxyproline-O-β- GT31 At4g21060 galt2-1 Golgi and ER Reduced root Basu et al., galactosyltransferase (SALK_117233) growth, radial 2013 galt2-2 swell-ing of (SALK_14112) Golgi root tips in GALT5 At1g74800 galt5-1 salt, reduced Basu et al., (SALK_10540) seed mucilage 2015 galt5-2 adherence (SALK_11574) AT1G7780 β-1,3-galactosyl GT31 At1g77810 - Golgi - Qu et al., 2008 transferase GALT31A β-1,6 galactosyl GT31 At1g32930 galt31A Golgi Embryo lethal Geshi et al., transferase (FLAG_379B06 mutant 2013 ) GALT29A β-1,6 galactosyl GT29 At1g08280 - Unique - Dilokpimol et transferase subcellular al., 2014 compartments

162

Table 3 continued GlcAT14A β-1,6- GT14 At5g39990 glcat14a-1 Golgi and Enhanced cell Knoch et al., glucuronosyl (SALK_06433) also unique elongation in 2013; transferase glcat14a-2 subcellular seedlings Dilokpimol et (SALK_04390) compartments - al., 2014 GlcAT14B At5g15050 - - - GlcAT14C At2g37585 - - - FUT4 α-1,2-fucosyl GT37 At2g15390 fut4-1 Reduced root Wu et al., 2010; transferase (SAIL_284_B) growth under Liang et al., fut4-2 salt stress 2014, Tryfona (SALK_12530) et al., 2014 FUT6 At1g14080 fut6-1 (SALK_0783) fut6-2 (SALK_09950) RAY1 β-arabinofuranosyl GT77 At1g7063 ray1-1 - Reduced root Gille et al., transferase (SALK_053158) growth and 2013 ray1-2 reduced rosette (GABI_001C09) size and inflorescence

3.4.1. GALT5 is localized to Golgi vesicles

GALT5 was localized to Golgi vesicles, consistent with bioinformatics predictions using Signal P and Golgi predictor, biochemical pulse-chase experiments of HRGP biosynthesis (Gardiner and Chrispeels, 1975; Robinson and Glas 1982; Nikolovski et al.,

2012) proteomics technique for localization of organelle proteins by isotope tagging, and localization studies performed with other AGP GTs, including GALT2, AT1G77810,

GALT31A, GALT29A, GlcAT14A, and FUT6 (Figure 3.8 and Table 3.1).

Interestingly, GALT2 as well as Hyp-GALT activity was identified in the ER as well as the Golgi, indicating AGPs likely initiate Hyp galactosylation in the ER and continue to be Hyp galactosylated and further glycosylated in the Golgi (Liang et al., 2010; Basu et al., 2013) A recent study has also localized GALT31A, GALT29A, and GlcAT14A to unique subcellular compartments, which are not part of the trans-Golgi network, cis-

Golgi network or endosomes (Oka et al., 2010; Poulsen et al., 2014).

3.4.2. AGP glycosylation required for normal growth and development: GALT and

AGP glycosyltransferase mutant phenotypes

While single galt2 or galt5 mutants are largely indistinguishable from wild type with respect to their non-biochemical phenotypes, galt2galt5 double mutants are clearly compromised with respect to normal growth and development (Table 3.1, Figure 3.14,

Figure 3.15, Figure 3.16, Figure 3.18, and Figure 3.21). Given that the single mutants have biochemical phenotypes corresponding to a reduction in AGP glycosylation that is exacerbated in the double mutant, it is reasonable to conclude that critical threshold levels of gycosylated AGPs are required for normal growth and development. In other

164 words, single mutants have sufficient levels of glycosylated AGPs to appear normal, while double mutants do not and display pleiotropic phenotypes affecting the growth and development of roots, leaves, , flowers, pollen, and seeds.

Specifically, galt2galt5 double mutants display shorter roots (Figure 3.21), shorter and less dense root hairs (Figure 3.15), more rosette leaves (Table 3.2), shorter inflorescences (Table 3.2), delayed flowering (Table 3.2), shorter pollen tubes (Figure

3.17), disruption of pollen tips (Figure 3.16), and reduced seed coat mucilage (Figure

3.24). These observations also lend further support to the notion that GALT5 and

GALT2 are functionally redundant. More severe growth and development consequences, including lethality, are likely as additional Hyp-GALT genes are identified and cumulatively knocked-out.

In contrast, other mutant phenotypes were observed for some of the known AGP

GTs (Table 3.3). Specifically, mutants for the β-1,6-GALT (GT31A) were embryo lethal (Geshi et al., 2013), mutants for one of the β -1,6-GlcATs (GT14A) displayed longer roots Knoch et al , and mutants for the β -AraT in GT77 (RAY1) had longer hypocotyls (Gille et al., 2013). Without knowing the precise site(s) of enzyme action and the biochemical extent to which each of these mutants alters AGP structure, it is difficult to interpret the functional implications of these mutant data. One possibility is that specific glycomodules within the AG polysaccharide are responsible for specific functions.

Roots and root hairs are particularly sensitive to the loss of glycosylated AGPs since some of the single galt mutants display more subtle versions of the phenotypes observed

165 in galt2galt5 double mutants. Such root hair sensitivity was previously observed in mutants for proline hydroxylation, extensins, and extensin arabinosylation, which display impaired growth (Baumberger et al., 2003; Velasquez et al., 2011).

3.4.3. AGP glycosylation required for root and tip growth

Several conditional phenotypes also characterize the galt single and double mutants; the most interesting of which involve alterations in root growth, pollen tubes, and root hairs in response to β-Yariv or NaCl treatments. In roots, β-Yariv binds AGPs, specifically to their β -1,3-galactan chains (Kitazawa et al., 2013), and inhibit root elongation (Figure 3.18) (van Hengel and Roberts, 2002; Ding and Zhu, 1997; Willats and Knox,1996). This inhibition is alleviated in the single and double mutants, which have reduced AGP glycosylation. This conditional phenotype indicates a role for AG polysaccharides in root elongation; this role is largely, but not completely masked in the mutants under normal growth conditions due to gene redundancy. In particular, normal root growth is inhibited in the double mutant, which corroborates a function for AGP glycosylation in root elongation. Pollen tube growth is affected in a similar manner. In pollen, β-Yariv binds to AGPs and inhibits pollen tube growth (Figure 3.17; Cheung et al., 2002; Mollet et al., 2002; Pereira et al., 2006). This inhibition is alleviated in single and double mutants, which likely have reduced AGP glycosylation. This conditional phenotype indicates a role for AG polysaccharides in pollen tube growth; as in the roots, this role is largely, but not completely masked in the mutants under normal growth conditions due to gene redundancy. Notably, under normal growth conditions, the double mutants show reduced pollen tube growth, as well as reduced root hair growth.

166

These results are consistent with observations that knock-out mutants of pollen-specific

AGP genes (AGP6, AGP11 and AGP40) lead to impaired pollen tube elongation

(Levitin et al., 2008) and that knock-out mutants of prolyl 4-hydroxylase genes (P4H2,

P4H5, P4H13) display shorter root hairs (Velasquez et al., 2011). Taken together, these studies indicate the important functional contribution of the carbohydrate moiety of

AGPs to root and pollen tube growth and development, particularly with respect to polarized tip growth.

Salt treatment was also used to reveal conditional phenotypes in the single and double mutants. In roots, NaCl treatment results in reduced root growth (Kang et al.,

2008; Galvan-Ampudia et al., 2011; Zagorchev et al., 2014) and the root bending assay can be used to screen for salt sensitivity (Zhu et al., 1998). Here, the single and particularly the double mutants are salt hypersensitive, demonstrating significantly reduced root growth and delayed root bending, further corroborating the functional contribution of the carbohydrate moiety of AGPs to root growth (Figure 3.21 and 3.22).

Moreover, root elongation in the single and double galt mutants was hypersensitive to

NaCl, LiCl, and KCl, but not to CsCl, as was previously observed for the sos5 mutants

(Figure 3.20; Shi et al., 2003). Single and double mutants for FUT4 and FUT6 also demonstrate reduced root growth in response to salt treatment, illustrating more specifically the importance of fucose in the AG side chains in root growth (Levitin et al.,

2008; Liang et al., 2013; Tryfona et al., 2014). In addition, NaCl treatment can result in root tip swelling in salt hypersensitive mutants. The single and double galt mutants also display this phenotype. Interestingly, mutants for SOS5, a fasciclin-like AGP, as well as

167 two cell wall receptor like kinases (FEI1 and FEI2) which are in the same genetic pathway as SOS5, phenocopy the galt2, galt5, and galt2galt5 mutants with respect to

NaCl–induced root tip swelling and salt hypersensitivity in the root bending assay.

3.4.4 GALTs and cellular signaling

AGPs are implicated as cellular signaling molecules (Shi et al., 2003; Tryfona et al.,

2014; Schultz et al., 1998; Lamport et al., 2006). Such implications stem from the observation that many AGPs have the capacity to be GPI-anchored to the plasma membrane and contain information-rich AG polysaccharides which could serve as signals themselves or bind to signaling molecules, such as calcium, and/or signal transduction molecules, such as plasma membrane/cell wall kinases. One of the more compelling indications that AGPs function in cellular signaling centers on SOS5/FLA4, a GPI-anchored fasciclin-like AGP; sos5 mutants are known to be salt hypersensitive and exhibit root tip swelling (Shi et al.,

2003). This root tip swelling phenotype is phenocopied by fei1fei2 mutants, which lack a pair of cell wall receptor-like kinases, FEI1/FEI2, and are defective in cellulose biosynthesis

(van Hengel and Roberts, 2002; Ding and Zhu, 1997; Steinwand et al., 2014). Moreover, genetic evidence indicates SOS5 and FEI1/FEI2 act in the same pathway. More recent work in seeds, has indicated SOS5 and FEI2 act in a pathway to synthesize seed coat mucilage, including pectin and cellulose, but the mechanistic details remain to be elucidated (Griffiths et al., 2014; Harpaz-Saad et al., 2011). The observations here that the galt2galt5 double mutant phenocopies the root swelling phenotypes (Figure 3.23) as well as the seed coat mucilage and cellulose-deficient phenotypes (Figure 3.24) of sos5, fei1fei2, and sos5fei1fei2 likely indicates that the carbohydrate moiety of SOS5/FLA4 is important for

168 cellular signaling. Experiments are in progress in our lab and elsewhere to determine whether the AG polysaccharides of SOS5 are responsible for directly or indirectly interacting with the extracellular domains of FEI1 and FEI2 and thereby activating the intracellular kinase domains. Such interactions may be similar to that already discovered between pectin and wall-associated kinases (Steinwand et al., 2014; Kohorn et al., 2012).

3.5 Concluding remarks

AGPs are long known to be expressed throughout the plant kingdom and implicated to function in various aspects of plant growth and development. This work identifies two key enzymes, GALT5 and GALT2, responsible for O-glycosylation of AGPs and elucidates the importance of the carbohydrate moiety in carrying out such functions. The detailed mode of action by which the AG polysaccharides trigger such growth and developmental events, however, remains to be determined. It is unknown whether the multitude of defects can be traced to a common mode of action or multiple modes of actions. For example, alterations to the AG polysaccharides could prevent intermolecular interactions within the wall and lead to a loss of cell wall integrity. This brings to mind the recent finding that AGPs can serve a structural role in crosslinking pectin and hemicellulose as well as earlier suggestions that AGPs assist with the packaging and delivery of cell wall cargo in Golgi-derived secretory vesicles (Tan et al.,

2013; Gibeaut and Carpita, 1991) or act as water laden, plasma membrane-cell wall shock absorbers (Nothnagel, 1997). Alternatively, such AG polysaccharide alterations may disrupt cellular signaling by preventing interactions with signaling molecules or signal transduction molecules. For example, AG polysaccharides can act in cellular

169 signaling by their ability to bind calcium, to serve as cell wall integrity sensors, and/or to act as co-receptors which interact with signaling molecules such as kinases, particularly those located at the cell surface (Seifert and Roberts, 2007; Ellis et al., 2010; Zhang et al., 2011; Lamport and Várnai, 2013; Wolfe et al., 2012) Future research designed to elucidate the molecular interactions between the AG polysaccharides and other molecules at the cell surface promises to hold the key to understanding the mode of action for these intriguing molecules.

3.6. Materials and Methods

3.6.1. In silico analysis of GALT5 and GALT2

Arabidopsis GALT2 and GALT5 predicted protein structure was depicted using

Prosite Mydomain Image creator (http://prosite.expasy.org/mydomains/).

Transmembrane domain, GALECTIN and GALT domains were predicted using

TMHMM server (http://www.cbs.dtu.dk/services/TMHMM/), and Pfam

(http://www.sanger.ac.uk/Software/Pfam/) respectively.

3.6.2. Heterologous expression of GALT5 in Pichia pastoris

The coding region of GALT5 was obtained from the RIKEN Bioresource center. The open reading frame of AtGALT5 was amplified with primers with a 5′ restriction site for

SacII followed by a 6x His-tag and a 3′ restriction site for XbaI (forward-

CGCCGCGGATGCATCATCATCATCATCACATGAAAAAACCCAA

ATTGTCG and reverse GAGTGTTGTAACATGAGATGATCTAGA). The boldface letters denote the restriction sites, the italic type denotes the His6 tag, and the underlined region denotes the translational start site. Amplified products were sequenced, cloned in

170 the shuttle vector pPICZ B as described in Basu et al. (2013). Five individual

Pichia clones expressing AtGALT5 were selected, and the presence of the gene was confirmed by PCR using genomic DNA isolated from transformants and gene-specific primers. Genomic DNA was isolated from Pichia cells as described previously

(Hoffman and Winston, 1987). After confirmation of the clones of Pichia harboring

GALT5, induction for expression of clones expressing GALT5 was performed as described in Basu et al. (2013).

3.6.3. Preparation of Pichia microsomes expressing GALT5 and immunoblot analysis

Microsomal proteins were isolated from the clone five transformed Pichia cells as described in Basu et al. (2010). For immunoblot analysis, 5 µg of microsomal protein from Pichia transformants was denatured, subjected to 10% SDS-PAGE, and electroblotted onto PVDF Immobilon membranes (Millipore) using the Mini Protean3 system according to manufacturer's recommendations. Blots were probed with an anti-

His primary antibody (Clontech) at a 1:10,000 dilution and a secondary goat anti-mouse

IgG antibody conjugated to horseradish peroxidase (HRP) (Clontech) at a 1:20,000 dilution. West Femto Maximum Sensitivity Substrate (Thermo Scientific) was used for

HRP detection. Pichia cell lines transformed with the empty expression vector (pPCIZ

B) were used as the negative control (NC). Protein quantification was done using the

Bradford reagent (Sigma). Blots were stained with Coomassie Brilliant Blue R-250

(Sigma) following HRP detection to ensure equal loading.

171

3.6.4. Galactosyltranferase assay with microsomal preparations from Pichia expressing GALT5

The standard GALT reaction was performed as described in Basu et al. (2010) using detergent permealized microsomes from Pichia clone C5 expressing AtGALT5. Two reactions were included as controls, one with no substrate acceptor and one with permeabilized microsomal membranes from the Pichia line (X-33) transformed with the empty expression vector (pPICZ B) as NC.

3.6.5. Purification of Hyp-GALT5 reaction products by reverse-phase HPLC

The GALT reaction products were purified by RP-HPLC as described by Liang et al.

(2010).

3.6.6. Analysis of the Hyp-[14C]galactoside profile by gel permeation chromato- graphy and high performance anion-exchange chromatography (HPAEC)

Thirty standard GALT reactions were fractionated by RP-HPLC and combined to generate enough 14C-radiolabeled product for base hydrolysis and separation on a Biogel

P2 column (Liang et al., 2010). The radioactive peak eluting at degree of polymerization

4 (DP4) on a Biogel P2 column was analyzed along with a chemically synthesized Hyp-

Gal standard by HPAEC on a CarboPac PA-20 column using 5 mM NaOH as the elution buffer to provide additional confirmation of this DP4 peak as Hyp-Gal. Trans-4-(β-D-

Galactopyranosyloxy)-L-proline (i.e. the Hyp-Gal standard) was chemically synthesized from commercially available galactopyranosyl bromide and hyp methyl ester as previously described (Basu et al., 2013).

172

3.6.7. Monosaccharide composition analysis of GALT reaction products by high performance anion-exchange chromatography

Twenty-five standard GALT assays were pooled to generate sufficient 14C-products for acid hydrolysis and monosaccharide composition analysis as described by Liang et al. (2010) and Basu et al. (2013) with minor modifications. The product from total acid hydrolysis was dissolved in deionized water and analyzed on a CarboPac PA20 column

(4 × 250 mm; Dionex) in a BioLC system using pulsed amperometric detection (ED50 electrochemical detector; Dionex). The column was equilibrated at a flow rate of 0.5 mL/min with 200 mM NaOH for 10 min, double distilled water for 10 min, and 1mM

NaOH for 15 min. The sample was eluted with 1 mM NaOH at a flow rate of 0.5 mL/min.

3.6.8. Determination of substrate specificity of the GALT5 enzyme activity

A standard GALT assay was performed using 20 µg of various peptide substrate acceptors, (AO)7, (AO)14, and (containing 7, and 14 repeating dipeptide units, respectively), an extensin peptide (ExtP) containing repetitive SO4units, and a

(AP)7 peptide containing seven AP units as described by Liang et al. [29].

Rhamnogalactan I from potato and rhamnogalactan from soybean (100 µg each) were used as potential pectin substrates. Permeabilized microsomal membranes (250 µg) from the NC Pichia line and Pichia line expressing His6-GALT5 served as the enzyme source in the GALT reactions. For all of the peptide substrate acceptors, the standard GALT assay was performed, and the reaction products were fractionated by RP-HPLC before monitoring incorporation of radiolabeled 14C in a liquid scintillation counter (Beckman

173

Coulter LS 6500). For the pectin substrate acceptors, RG (soybean fiber; Megazyme) and RGI (potato; Megazyme), reactions were incubated at room temperature for 2 h, terminated by adding 1 ml of cold 70% ethanol, and precipitated overnight at −20 °C.

Reaction products were collected by centrifugation at 10,000 × g for 10 min, and pellets were resuspended by vortexing followed by ten washes with 1 ml of cold 70% ethanol to remove excess UDP-[14C]Gal. The 14C-radiolabel incorporation was estimated by resuspending the pellets in 300 µl of water before counting in a liquid scintillation counter.

3.6.9. Biochemical characterization of GALT5 enzyme activity

The standard GALT assay was modified for GALT5 characterization using (AO)7 peptide as the acceptor substrate. Assay products from each reaction were fractionated by RP-HPLC to measure incorporated 14C-radiolabel into acceptor substrates. The optimum pH for GALT5 activity was determined using permeabilized microsomal membranes (250 µg) from the C5 Pichia line expressing His6-GALT5 dissolved in test buffers at a final concentration of 100 mM. Test buffers included MES-KOH buffer at pH 4, 5, 6, and 7; HEPES-KOH buffer at pH 6, 6.5, 7, 7.5, and 8; Tris-HCl buffer at pH

8, 9, and 10; and CAPS-KOH buffer at pH 9 and 10.

To examine the effect of divalent cations on GALT5 activity, microsomal membranes were extracted with homogenizing buffer lacking divalent ions. MnCl2,

MgCl2, CaCl2, CuCl2, NiCl2, or ZnSO4 was added to the GALT assay (at a final concentration of 5 mM) when tested. Two controls were added, one with no ions in the buffer used for resuspending the detergent permealized membrane fraction and the other

174 with EDTA (5 mM) to chelate any residual divalent cations trapped in the membranes.

An equal volume of deionized distilled water was added instead of divalent ions in the control reaction.

To analyze the enzyme specificity for nucleotide sugar donors, the standard activity

14 assay was performed with (AO)7 as the acceptor substrate and various C-radiolabeled nucleotide sugar donors (90,000 cpm). The nucleotide sugars tested included UDP-

[14C]Glc (MP Biomedicals), UDP-[14C]Xyl (PerkinElmer Life Sciences), and GDP-

[14C]Fuc (PerkinElmer Life Sciences). Four separate GALT reactions with no substrate acceptors were performed as controls.

3.6.10. Transient expression and subcellular localization of GALT5 in Nicotiana tabacum leaves

The GALT5 coding region was subcloned into pEarleyGate 101 plasmid to generate the GAL5:YFP construct by a gateway cloning strategy. The primers used in cloning are listed in Table 3.4. Agrobacterium-mediated transient expression was performed in the leaves of three to four week-old tobacco plants (Nicotiana tabacum cv. Petit Havana) grown at 22-24°C using a bacterial optical density (OD 600) of 0.05 for single infiltrations and 0.025 each for co-infiltrations (Saint-Jore et al., 2002). The GALT5-

YFP construct was co-expressed with either the ER marker GFP-HDEL or the Golgi marker ST-GFP (Batoko et al., 2000) to ascribe subcellular localization. The ER and

Golgi markers are cloned into pVKH18-EN6 plasmid vector. Transformed plants were incubated under normal growth conditions and sampled daily for 2-5 days post- infiltration. Leaf epidermal sections were imaged using an upright Zeiss LSM 510

175

META laser scanning microscope (Jena, Germany), using a 40 X oil immersion lens and an argon laser. For imaging the expression of YFP constructs, the excitation line was

514 nm, and emission data were collected at 535–590 nm, whereas for GFP constructs, the excitation line was 458 nm and the emission data were collected at 505-530 nm.

Singly infiltrated controls were analyzed to optimize gain and pinhole settings for each channel and to exclude any bleed through fluorescence between channels. Post- acquisition image processing was done using the ZEN lite 2012 image analysis software

(Blue Edition; Carl Zeiss).

176

Table 3.4 List of primers used in this study Purpose Forward Reverse Cloning GALT5 CGCCGCGGATGCATCATCAT GAGTGTTGTAACATG CATCATCACATGAAAAAACC AGATGATCTAGA CAAATTGTCG Subcellular localization CACCATGAAAAAACCCAAA TCTCATGTTACAACA TTGTCGAA CTCAGGCTTG Screening for T-DNA galt5-1 GALT5-1RP- GALT5-1LP- TTTCCACTTTCGACAATTTG CTAATTACATGGTTTT G GCGGG galt5-2 GALT5-2RP- GALT5-2LP- TGGGGACATTGTACTTGTTC TGGTACGCTTGCAAA C ATTTTC LBa1.3 ATTTTGCCGATTTCGGAAC RT-PCR GALT2 RTF- RTR- TCTTTGTTGCACTTAATCCA CATAAGCTTAGGCTA AGAAG TTCAAGATGG GALT5 RTF- RTR- TATGTGAACACGGAGCTCT ACATAAATTACGGCT TGCATTC GTTCAAGATGGA

177

Table 3.4 continued UBQ 10 GTCGACCCTTCACTTGGTGT ATCCTCAAGCTGCT TTCCAG GALT2 QPCRF- QPCRR- TCTTAGACATCGTCCTCTTA ACACAGCTGGAAA GA TTTTGCC GALT5 QPCRF- QPCRR- ACATAAATTACGGCTGTTCA CTTATGGGATAAGC AGATGGA TCTTAA The grey shaded area denotes the 6x His-tag, the underlined nucleotides in the forward primer column denotes a SacII restriction site, whereas the underlined nucleotides in the reverse primer column denotes an XbaI restriction site for GALT5.

3.6.11. Plant material and genetic analysis

The Columbia (Col-0) ecotype of Arabidopsis thaliana was used in this study. Two

T-DNA insertional lines for At1g74800-GALT5 (galt5-1 SALK_105404 and galt5-2

SALK_115741) and At4g21060-GALT2 (galt2-1 SALK_117233 and galt2-2

SALK_141126) were selected using the SIGnaL database (http://signal.salk.edu/) and were obtained from the ABRC (Arabidopsis Biological Research Centre)

(http://abrc.osu.edu/). Other mutants including fei1 (SALK_080073), fei2-1, and sos5-2

(SALK_125874) were provided by Dr. Joseph Keiber. Arabidopsis plants used in this study were germinated after 4 days of stratification in the dark at 4°C, and grown on soil at 22°C with 60% relative humidity. Plants were grown under long-day conditions (16 h photoperiod and 8 h dark, 120 µmol m -2 s -1 of fluorescent light).

Genomic DNA was isolated from galt5-1, galt5-2, galt2-1, galt2-2 and galt2galt5 mutant leaves and subsequent PCR analysis was carried out using Extract-N-Amp™

178

Plant Kits (Sigma-Aldrich). The primer sequences used in PCR analysis were obtained from the T-DNA Primer Design Tool provided by the Salk Institute Genomics Analysis

Laboratory (http://signal.salk.edu/tdnaprimers.2.html) in conjunction with the gene specific left and right primers (Table 3.3). PCR products were purified by gel extraction with QIAquick Gel Extraction Kit and sequenced by the Ohio University Genomics

Facility. To confirm homozygous plants at the transcript level, RNA was extracted, reverse transcribed, and analyzed by PCR using RT primers. RNA was isolated using a

Qiagen RNeasy plant mini kit followed by DNase I digestion using Qiagen RNase free

DNase set to remove traces of DNA. Qiagen One-Step RT-PCR kit was used for first- strand synthesis and subsequent PCR steps (primers are listed in Table 3.3).

For qRT-PCR, the cDNAs were amplified using Brilliant II SYBR Green QRT-PCR

Master Mix with ROX (Agilent Technologies, La Jolla, CA, USA) in an MX3000P real- time PCR instrument (Agilent Technologies). PCR was optimized and reactions were performed in triplicate. The transcript level was standardized based on cDNA amplification of Ubiquitin 10 (At4g05320) RNA as a reference.

3.6.12. Isolation of Golgi-enriched plant microsomal membranes

Plant microsomal membranes were extracted according to Liang et al. (2010) with minor modifications. Eight grams of leaf tissue from 14 d old wild type, galt2-1, galt2-2, galt5-1, galt5-2 and galt2galt5 mutant plants were ground in liquid nitrogen followed by resuspension in 8 ml extraction buffer (0.1 M HEPES-KOH, pH 7, 0.4 M Sucrose, 1 mM dithiothreitol, 5 mM MgCl2, 5 mM MnCl2, 1 mM phenylmethylsulfonyl fluoride, and one tablet of Roche EDTA-free complete protease inhibitor cocktail and 100 µL RPI

179 plant protease inhibitor VI). The homogenate was filtered through two layers of miracloth, and the filtrate was centrifuged at 3,000 x g for 20 min. The resulting supernatant was layered over a 1.8 M Sucrose cushion buffer and centrifuged at 100,000 x g for 60 min. The uppermost layer was discarded without disturbing the membrane containing interphase layer. A discontinuous sucrose gradient was implemented by sequentially layering 1.1 and 0.25 M sucrose solutions onto the interphase layer and centrifuging at 10,000 x g for 60 min. The microsomal membranes enriched at the

0.25/1.1 M sucrose interphase were collected and pelleted by another centrifugation at

100,000 x g for 30 min. The pellet was resuspended in 50 µL extraction buffer and stored at −80°C until use. A 1% Triton-X 100 permealized membrane fraction was used to perform GALT reactions using [AO]7 as the peptide substrate acceptor and UDP-

[14C]Gal as the sugar donor.

3.6.13. Extraction of AGPs and AGP profiling by HPLC

AGPs were extracted from the WT, galt2-1, galt2-2, galt5-1, galt5-2, and galt2galt5 mutant plants as described in Schultz et al. (2000). Five grams of plant material was used for each of the lines. Quantification of AGPs was done following the method of

Gao et al. (1999), and β-Yariv reagent was prepared as described in Yariv et al. (1967).

AGP profiling was conducted as described by Youl et al. (1998) with modifications.

AGPs were obtained from eight grams of plant material, precipitated by β-Yariv reagent and dissolved in 1 mL of deionized water before applying 100 µl onto a polymeric reverse-phase column (PRP-1, 5 µm, 4.1 × 150 mm; Hamilton) equilibrated with buffer

A (0.1% trifluoroacetic acid). Fifty µg of [AO]7 was used as a control to monitor the

180 retention time of a pure AGP peptide. Samples were eluted from the column following a linear gradient with solvent B (0.1% trifluoroacetic acid in 80% acetonitrile): 0 to 30% solvent B in 30 min, then 30 to 100% in 30 min at a flow rate of 0.5 mL/min.

Chromatography was monitored by absorption at 215 and 280 nm.

3.6.14. In vitro pollen germination assay

Flowers collected from WT, galt2-1, galt2-2, galt5-1, galt5-2 plants 1 to 2 weeks after bolting were used for the examination of pollen tube phenotypes. Individual open flowers were germinated in vitro as described by Boavida and McCormick, (2007) on solid germination medium (0.01% H3BO3, 1 mM MgSO4, 5 mM KCl, 5 mM CaCl2,

10% sucrose, and 1.5% low-melting agarose, pH 7.5 and 30 µM β-Yariv reagent or 30

µM α-Yariv reagent) at 22°C and 100% humidity in the dark. Pollen tube germination rates were computed by dividing the total number of germinated tubes by the number of grains. Images and measurements of pollen tubes were done at either at 40X or 20 X magnifications in a Nikon microscope coupled with a SPOT RT color CCD camera and

SPOT analysis software.

3.6.15. Germination assays

Seeds of wild-type, galt2-1, galt2-2, galt5-1 and galt5-2 were surface-sterilized by washing in a 95% ethanol solution for 5 min followed by a 5 min wash in a 30% bleach with 0.1% Tween 20 solution and then rinsed seven times with sterile water. The seeds were sown on 1X MS nutrient medium containing 1 % sucrose and 0.6% agar. For stratification treatment, seeds were stratified at 4°C in the dark for 3 d. The germination rate was scored by counting the number of germinated seeds after 5 d. Experiments were

181 done in triplicate with 50 seeds for each experiment and genotype. Only seed batches that had been harvested and stored at the same time and under the same conditions were used. For each experiment, samples from four genotypes (WT, two allelic single mutants and the double mutant) were placed side by side on the same plate. Radicle length was measured by Motic Image version 3.2. Three replicate plates were used for each treatment to ensure reproducibility of data.

3.6.16. Root growth measurements

For monitoring root growth in response to Yariv reagent, wild-type, galt2-1, galt2-2, galt5-1, galt5-2, and galt2galt5 were grown on MS plates for 7 d before they were transferred to MS plates supplemented with 50 µM α-Yariv reagent or 50 µM β-Yariv reagent. For seedling growth in salt, 7-d-old seedlings of wild-type, galt2-1, galt2-2, galt5-1, galt5-2, and galt2galt5 plants were transferred to MS medium containing 1% agar and 100 mM or 150 mM NaCl. Root length was determined on low-magnification

(×10) digital images captured using a CCD camera and image analysis freeware

(ImageJ; http://rsb.info.nih.gov/ij/). For analysis of salt hypersensitivity of the mutant plants, root growth was monitored using a root bending assay (Zhu et al., 1998) and images were taken under Nikon SMZ1500 stereomicroscope coupled with a CCD

Infinity 2 camera and analysis software.

3.6.17. Aberrant root hair morphology

Root hair length from 8-d-old plants grown on agar plates was determined on low- magnification (×10) digital images captured using a CCD camera and image analysis freeware (ImageJ; http://rsb.info.nih.gov/ij/). To ensure comparable results, the area 3 to

182

5 mm behind the root tip was analyzed. Plants grown on agar plates were carefully removed in ∼100 µL of half-strength MS medium on microscope slides for analysis.

Quantification data are the means of 50 to 75 values representing 15 root hairs each of

20 to 35 individual plants measured.

3.6.18. Seed staining and visualization

Seeds of all the indicated genotypes were prehydrated in water and stained either with 0.01% ruthenium red or calcofluor white (25 µg/ml of fluorescent brightener). In both cases, staining was performed as described by Willats et al. (2001) and Harpaz-

Saad et al. (2011). Imaging was done using a Zeiss LSM 510 confocal microscope.

3.6.19. AGP specific monoclonal antibodies

Four well characterized AGP specific monoclonal antibodies namely, JIM4, JIM8,

JIM13 and MAC207 (CCRC; http://www.ccrc.uga.edu/) were used as primary antibodies for detection of AGP epitopes. Goat anti-rat secondary antibody conjugated to fluorescein isothiocyanate (FITC) was used as secondary antibody. Root hairs, pollen tubes and seeds treated with secondary antibody only were used as negative controls.

Images were examined with a Zeiss LSM 510 laser scanning confocal microscope equipped with an argon-ion laser, using single wavelength excitation at 488 nm and detection of FITC signals between 505 and 530 nm. Confocal parameters for each antibody treatment were preserved across genotypes. Z-stack sections of the images were taken, and three-dimensional projections from these stacks were used for the final images using LSM Software ZEN 2011.

183

3.6.20. Immunofluorescence detection of AGPs epitopes in root hairs, pollen tubes and seeds

Ten day old WT and galt2galt5 seedlings grown on MS plates were used for immuno-staining of AGP epitopes according to the method described by Sauer et al.

(2006). Briefly, seedlings were harvested in 1X MS liquid media followed by fixation in stabilizing buffer (SB) prepared in 1X MS media containing 50 mM PIPES buffer, 5 mM MgSO4, 5 mM EGTA pH 7.0 with 4% paraformaldehyde at 4°C overnight. After extensively washing the seedlings with SB without 4% paraformaldehyde, they were incubated for 60 min at room temperature in 1X MS media containing 3% IGEPAL followed by incubation for 1h with 3% BSA with 0.02% sodium azide in 1X MS.

Seedlings were incubated with the primary antibody (1:25 dilution) overnight at 4°C in the dark, followed by extensive rinsing and incubation with secondary antibody at a 1:50 dilution with 1X MS media for 5 h at room temperature. Finally, seedlings were washed in 1X MS media and mounted in 25% glycerol in 1X MS media.

Immunolocalization of AGPs in WT and galt2galt5 pollen tube was performed according to the method described by Dardelle et al. (2010). Briefly, pollen tubes were germinated in germination media (GM) containing 5 mM CaCl2 2H2O, 0.01% (w/v)

. H3BO3, 5 mM KCl, 1 mM MgSO4 7H2O, and 10% (w/v) Suc, pH 7.5 for 16 h at room temperature. Upon germination, they were mixed (v/v) with a fixation medium

. containing 100 mM PIPES buffer, pH 6.9, 4 mM MgSO4 7H2O, 4mM EGTA, 10%

(w/v) Suc, and 5% (w/v) formaldehyde and incubated for 90 min at room temperature.

Pollen tubes were rinsed three times by centrifugation at 3,200 g for 6 min with 50 mM

184

. PIPES buffer, pH 6.9, 2 mM MgSO4 7H2O, and 2 mM EGTA and three times with phosphate-buffered saline (100 mM potassium phosphate, 138 mM NaCl, and 2.7 mM

KCl, pH 7.4). Pollen tubes were incubated overnight at 4°C in the dark with primary antibodies diluted 1:10 with phosphate-buffered saline, rinsed, and incubated with secondary antibody diluted 1:50 for 3 h at room temperature.

Whole-seed immunolabeling was conducted according to the method described by

Harpaz-Saad et al. (2011), using seeds shaken in water before immunolabeling.

185

CHAPTER 4: A SMALL MULTIGENE HYDROXYPROLINE-O-

GALACTOSYLTRANSFERASE FAMILY FUNCTIONS IN

ARABINOGALACTAN-PROTEIN GLYCOSYLATION, GROWTH AND

DEVELOPMENT IN ARABIDOPSIS

This work will be submitted to the journal of BMC Plant Biology for review.

Basu D, Tian L, Wang W, Bobbs S, Herock H, Travers A and Showalter AM.

4.1. Abstract

Arabinogalactan-proteins (AGPs) are ubiquitous components of cell walls throughout the plant kingdom and are extensively post translationally modified by conversion of proline to hydroxyproline (Hyp) and by addition of arabinogalactan polysaccharides (AG) to Hyp residues. Hyp glycosylation is initiated by the action of a set of Hyp-O-galactosyltransferase (Hyp-O-GALT) enzymes. AGPs are implicated to function in various aspects of plant growth and development, but the functional contributions of AGP glycans remain to be elucidated. Three members of the GT31 family (GALT3-At3g06440, GALT4-At1g27120, and GALT6-At5g62620) were identified as Hyp-O-GALT genes by heterologous expression in tobacco leaf epidermal cells and examined along with two previously characterized Hyp-O-GALT genes, GALT2 and

GALT5. Transcript profiling by real-time PCR of these five Hyp-O-GALTs revealed overlapping but distinct expression patterns. Transiently expressed GALT3, GALT4, and GALT6 fluorescent protein fusions were exclusively localized within Golgi vesicles. Biochemical analysis of knock-out mutants for the five Hyp-O-GALT genes

186 revealed significant reductions in both AGP-specific Hyp-O-GALT activity and β-Yariv perceptible AGPs. Further phenotypic analysis of these mutants demonstrated reduced root hair growth, reduced seed coat mucilage, reduced seed set, and accelerated leaf senescence.The mutants also displayed several conditional phenotypes, including impaired root growth, and defective anisotropic growth of root tips under salt stress, as well as less sensitivity to the growth inhibitory effects of β-Yariv reagent in roots and pollen tubes. This study provides evidence that all five Hyp-O-GALT genes encode enzymes that catalyze the initial steps of AGP galactosylation and that AGP glycans play essential roles in both vegetative and reproductive plant growth.

4.2. Introduction

Arabinogalactan-proteins (AGPs) are members of the hydroxyproline (Hyp)-rich cell wall glycoproteins superfamily and are hyperglycosylated by O-linked AG polysaccharides. AGPs are found in cell walls, plasma membranes, and extracellular secretions of virtually all plant cells, tissues and organ types (Majewska-Sawka and

Nothnagel, 2000). Moderately sized gene families encode a variety of AGP protein backbones throughout the plant kingdom. For example, based on bioinformatics studies,

Arabidopsis contains 85 AGP genes, while rice contains 69 AGP genes (Showalter et al.,

2010; Ma and Zhao, 2010). Moreover, these genes are spatially and temporally expressed in a variety of patterns, which likely relates to their multiple functions.

AGPs are implicated to function in various aspects of plant growth and development, including root elongation, somatic embryogenesis, hormone responses, xylem differentiation, pollen tube growth and guidance, programmed cell death, cell expansion,

187 salt tolerance, host-pathogen interactions, and cellular signaling (Seifert and Roberts,

2007; Ellis et al., 2010; Tan et al., 2012; Tan et al., 2013; Pereira et al., 2015; Nguema-

Ona et al., 2012; Seifert et al., 2014). However, there remains a lack of understanding of the biophysical and biochemical modes of action of any individual AGP. This lack of understanding regarding function also extends to the carbohydrate moieties or AG polysaccharides, which extensively decorate AGP core proteins and largely define their interactive surfaces.

Given the importance of understanding plant cell wall biosynthesis to biofuel production, much of the recent work on AGPs has focused on their biosynthesis. Such efforts have identified several of the biosynthetic glycosyltransferase (GT) genes/enzymes responsible for AG polysaccharide production (Tan et al., 2013, Knochet al., 2014). In particular, the following enzymes were identified and cloned: two α-1,2- fucosyltransferases (FUT4 and FUT6) which are members of the CAZy GT-37 family

(Wu et al., 2010; Liang et al., 2013; Tryfona et al., 2013), two hydroxyproline-O- galactosyltransferases (GALT2 and GALT5) which are members of GT-31 and contain a galectin domain (Basu et al., 2013; Basu et al., 2015), three other hydroxyproline-O- galactosyltransferases (HPGT1-HPGT3) which are members of GT-31 but lack a galectin domain (Ogawa-Ohnishi and Matsubayashi, 2015), one β-1,3- galactosyltransferase (At1g77810) which is a member of GT-31 (Qu et al., 2008), one β-

1,6-galactosyltransferase with elongation activity which is a member of GT-31

(GALT31A) (Geshi et al., 2013), one β-1,6-galactosyltransferase with branch initiation and branch elongating activities which is a member of GT-29 (GALT29A) (Dilokpimol

188 et al., 2014), three β-1,6-gluronosyltransferases which are members of GT-14

(GlcAT14A, GlcAT14B, GlcAT14C) (Knoch et al., 2013; Dilokpimol and Geshi et al.,

2014), and a putative AGP β-arabinosyltransferase (RAY1) which is a member of the

GT-77 family (Gille et al., 2013).

The hydroxyproline-O-galactosyltransferases (Hyp-O-GALT) that add galactose onto the peptidyl Hyp residues in AGP core proteins represent the first committed step in AG polysaccharide addition and represent an ideal control point to investigate the contribution of AG polysaccharides to AGP function. Previously, we demonstrated that

GALT2 (At4g21060) and GALT5 (At1g74800) are members of a small multigene family and encode Hyp-GALTs (Basu et al., 2013; Basu et al., 2015). In addition, extensive phenotypic characterization of allelic galt2 and galt5 single mutants and galt2galt5 double mutants at the biochemical and physiological levels was performed which corroborated the roles of these two enzymes in AG biosynthesis and elucidated the contributions of AG polysaccharides to AGP function. Here, we extend that work by characterizing the remaining GALTs members (i.e., GALT1, GALT3, GALT4, and

GALT6) of this small six-membered gene family, which are distinguished by encoding a

GALT domain as well as a GALECTIN domain.

4.3. Results

4.3.1. In silico analysis of GALT1, GALT3, GALT4 and GALT6

This study focused on the six-member gene/protein family in Arabidopsis, which is found within the CAZy GT31 family and distinguished by the presence of both a GALT

(pfam 01762) and a GALECTIN (pfam 00337) domain. Recently, two of these six

189 members, GALT2 (At4g21060) and GALT5 (At1g74800) were demonstrated to catalyze the addition of galactose onto Hyp residues of AGP backbones (Basu et al.,

2013; Basu et al., 2015). Another member of this family, GALT1, encoded by

At1g26810, was previously characterized and identified as a β–1,3-galactosyltransferase involved in the formation of the Lewis a epitope on N–linked glycans (Strasser et al.,

2007). The open reading frames of the remaining members, At3g06440 (GALT3),

At1g27120 (GALT4) and At5g62620 (GALT6) correspond to 1860 bp, 2022 bp and 2046 bp and specify proteins with 619 (70 kDa), 673 (77.0 kDa), and 681 (77.7 kDa) amino acids, respectively (Table 4.1). The six proteins share amino acid identities ranging from 35%-70% (Table 4.2). In addition, comparisons of these six members were performed with the three recently identified AGP-specific Hyp-O-GALTs (HPGT1,

HPGT2, and HPGT3), which are also within the GT-31 family and contain a GALT domain but lack a GALECTIN domain (Ogawa-Ohnishi and Matsubayashi, 2015). All nine proteins were predicted to be type II Golgi localized integral membrane proteins by several subcellular localization prediction programs (TargetP, http://www.cbs.dtu.dk/services/TargetP/ and Golgi Predator, http://ccb.imb.uq.edu.au/golgi/) (Emanuelsson et al., 2007; Table 4.1). These nine

GALTs were also run through the TMHMM server

(http://www.cbs.dtu.dk/services/TMHMM/) (Krogh et al., 2001) for prediction of transmembrane domains (TMDs), a typical type II membrane topology commonly found in GTs (Figure 4.1). All were predicted to have a single TMD except for GALT3,

HPGT2, and HPGT3, which instead contained hydrophobic regions that may serve as an

190 anchor to the Golgi membrane. Hydrophobic cluster analysis (HCA) was performed by submitting the protein sequences to the drawhca server

(http://smi.snv.jussieu.fr/hca/hca-form.html) and used to identify the hydrophobic pockets containing the “DXD” motifs of the six GALTs; this analysis also included two previously characterized AGP-related GT31 members, GALT31A and At1g77810, which are involved with the elongation of β-1,6-galactan side chains and the β -1,3 backbone of AG polysaccharides, respectively (Figure 4.2) (Qu et al., 2008; Geshi et al., 2013; Callebaut et al., 1997; Stolz et al., 2002). HCA analysis revealed conserved

DDD motifs in all the proteins contained within various hydrophobic pockets. The DXD motif is implicated in the binding the divalent metal ion that assists in anchoring the pyrophosphoryl group of the UDP-sugar donor in the enzyme’s active site (Qu et al.,

2008). Co-expression analysis was performed using GENEMANIA

(http://www.genemania.org/) and revealed that GALT3, GALT4 and GALT6 expression is tightly correlated with well-characterized AGP-specific GT31 members as well as with a number of AGPs (Table 4.3) (Basu et al., 2013; Geshi et al., 2013; Basu et al.,

2015; Mostafavi et al., 2008).

191

Table 4.1 Nomenclature, locus identifiers, site of T-DNA insertion, length of open reading frames, predicted number of amino acids Gene Mutant Location Genomic Coding Amino Localization lines DNA (bp) Region acids (bp) GALT1 Salk_ Promoter 3327 1932 643 Golgi At1g26810 006871 Intron Sail_ 170A08 GALT3 Salk_ Promoter 3133 1860 619 Golgi At3g06440 085633 Promoter Salk_ 005178 GALT4 Salk_ Exon 2839 2022 673 Golgi At1g27120 136251 Exon Salk_ 131723 GALT6 Sail_ Promoter 3196 2046 681 Golgi At5g62620 59_D08 Promoter Sail_ 70_B02

192

Table 4.2 Amino acid identity/similarity between the predicted amino acid sequences of the six AGP-specific GALT candidates involved in initial galactosylation of AGPs belonging to GT31 family by MATCHER (http://mobyle.pasteur.fr/cgi-bin/portal.py?#jobs::matcher)

GALT1 GALT2 GALT3 GALT4 GALT5 GALT6

GALT1 - 33.7 45.4 37.1 35.7 39.3

GALT2 51.5 - 38.9 56.4 57.1 57.9

GALT3 61.3 60.2 - 39.5 39.5 41.9

GALT4 56.1 73.2 61.4 - 68.0 60.2

GALT5 53.1 74.9 61.7 82.0 - 73.1

GALT6 56.1 74.1 64.0 73.6 82.5 -

Identity is displayed in the upper-half and similarity in the lower half of the table shaded in grey. Pair-wise analysis was done using Clustal W with a gap opening penalty of 14 and a gap extension penalty of 4 using BLOSUM62 matrix.

193

Table 4.3 Coexpression analysis of candidate genes involved in AGP biosynthesis in Arabidopsis using the Gene CAT coexpression tool

Pearson Correlation Gene ID PredictedFunction Coefficient 0.59084 At4g21060 galactosyltransferase family protein, (GALT2) contains Pfam profile: PF01762 galactosyltransferase 0.59083 At1g74800 galactosyltransferase family protein, (GALT5) contains Pfam profile: PF01762 galactosyltransferase 0.43006 At4g32120 galactosyltransferase family protein, (HPGT1) contains Pfam profile: PF01762 galactosyltransferase 0.40435 At1g32930 galactosyltransferase family protein, (GALT31A) contains Pfam profile: PF01762 galactosyltransferase 0.32234 At5g53340 galactosyltransferase family protein, (HPGT2) contains Pfam profile: PF01762 galactosyltransferase 0.32234 At5g53340 galactosyltransferase family protein, (HPGT3) contains Pfam profile: PF01762 galactosyltransferase 0.32374 At2g35620 leucine-rich repeat transmembrane 0.27977 At5g39990 glycosyltransferase family 14 protein / (GlcAT14a) contains Pfam profile: PF02485

194

Table 4.3 continued 0.26253 At2g37585 glycosyltransferase family 14 protein (GlcAT14b) contains Pfam profile: PF02485 enzyme 0.24682 At1g27120 galactosyltransferase family protein, (GALT4) contains Pfam profile: PF01762 galactosyltransferase 0.23551 At3g61640 AGP20 0.20469 At2g14890 AGP9 0.17328 At2g25300 galactosyltransferase family protein, contains Pfam profile: PF01762 galactosyltransferase 0.17184 At5g15050 glycosyltransferase family 14 protein / (GlcAT14c) contains Pfam profile: PF02485 0.16708 At3g57690 AGP23 0.13205 At3g20865 AGP40 0.13140 At3g01700 AGP11 0.10589 At1g70630 fucosyltransferase, putative (FUT4) 0.09032 At1g24520 AGP1 0.08353 At3g06440 galactosyltransferase family protein, (GALT3) contains Pfam profile: PF01762 galactosyltransferase 0.07137 At4g26320 AGP13 0.06124 At5g40730 AGP24 0.05883 At1g14080 fucosyltransferase, putative (FUT6) 0.04499 At5g10430 AGP4 0.03263 At5g56540 AGP14

195

Figure 4.1 Sequence analysis of nine GALT proteins. Sequence analysis using the TMHMM2.0 program for prediction of transmembrane helices of five Hyp-O-GALT proteins three namely HPGT1, HPGT2, HPGT-3, GALT2, GALT5 and three putative Hyp-O-GALTs namely, GALT3, GALT4 and GALT6. GALT1 is reported to be involved in the addition galactose for formation of the Lewis an epitope on N–linked glycans (Strasser et al., 2007).

196

Figure 4.2 HCA analysis of eight GALTs showing the DxD motif within a pocket of hydrophobic amino acids. The protein sequences are represented on a duplicated a-helical net, and the clusters of contiguous hydrophobic residues (V, I, L, F, M, W and Y) are boxed. Gly, Pro, Ser and Thr, are represented by symbols: Gly (filled diamond), Pro (red star), Ser (square with a black dot) and Thr (empty square). Vertical lines delineate the hydrophobic pocket in which the DxD motif (highlighted in red circle) can be found. Six GT-31 family members containing both GALT and GALECTIN were used along with two previously reported GALTs were used in this analysis.

197

4.3.2. Transiently expressed six Hyp-O-GALTs in Nicotiana tabacum has AGP specific Hyp-O-GALT activity

For biochemical characterization, full-length GALT1, GALT2, GALT3, GALT4,

GALT5 and GALT6 gene constructions, each harboring an N-terminal 6XHis tag, were transiently expressed in the leaves of Nicotiana tabacum. Leaves infiltrated with desired constructs were initially separated into three fractions: supernatant, total microsomal membranes and Golgi-enriched microsomal membranes. The highest GALT activity was observed in Golgi-enriched detergent permeablized microsomal membranes (Table 4.4), and thus this fraction was subsequently used as the enzyme source in transient assays

(Figure 4.3). Here, five of the six GALTs (i.e., GALT2-6) displayed Hyp-O-GALT activity, when compared to controls tobacco WT leaves alone or infiltrated with either an empty vector or an unrelated glycosyltransferase gene, sialyl transferase (ST).

Previously characterized GALT2 and GALT5 were used as positive controls for this assay, while GALT1 effectively served as a negative control, given its involvement with

N-glycan biosynthesis (Basu et al., 2013, Basu et al., 2015; Strasser et al., 2007; Figure

4.3).

198

Table 4.4 Subcellular distribution of AGP specific GALT activity obtained from GALT1-GALT6 transiently expressed in N. tabacum

Genotype Supernatant Total microsome Golgi microsome

WT 0.82+0.02 8.14+ 0.07 9.20+ 0.50

GALT1 0.75+0.15 8.25+ 0.02 8.95+ 0.45

GALT2 0.69+0.07 9.94+ 0.05a 14.49+ 0.61b

GALT3 0.86+ 0.07 9.85+ 0.02a 14.46+ 0.42b

GALT4 0.68+ 0.01 9.20+ 0.01a 13.24+ 0.56b

GALT5 0.73+ 0.02 10.39+ 0.05a 15.27+ 0.65b

GALT6 0.78+ 0.01 10.10+ 0.07a 14.96+ 0.71b

Experiments were performed using duplicated samples, and the data represent mean ± SE from two independent experiments. Letter ‘a’ represents significant differences (one way ANOVA, p<0.05 and ‘b’ represents p<0.01) compared to the WT control.

199

Figure 4.3 Hyp-O-galactosyltransferase Activity of GALT1, GALT3, GALT4 and GALT6 Transiently Expressed in N. tabacum. GALTs (1, 3, 4 and 6) were expressed in epidermal cells of tobacco leaves by Agrobacterium-mediated transient expression, prepared as microsomes and detergent permealized microsomes were used for glycosyltransferase assays. Synthetic peptide [AO]7 was used as substrate acceptor. WT tobacco leaves infiltrated with Agrobacterium GV3101 strain and infiltrated with ST fused with GFP were used as controls. The dark bars represent the GALT assay whereas the grey bars denote no substrate control. Experiments were performed using duplicated samples, and the data represent mean ± SD from two independent experiments (*, P < 0.05; **, P < 0.01, ** P < 0.001).

4.3.3 Substrate specificities of GALT2, GALT3, GALT4, GALT5 and GALT6

Various potential substrate acceptors were tested to investigate enzyme specificity of

GALT3, GALT4, and GALT6. Namely, [AO]7, [AO]14, and d[AO]51, consisting of non- contiguous peptidyl Hyp residues, were used to examine the effect of AGP peptide sequences of various lengths on GALT activity. [AP]7, consisting of alternating Ala and

Pro residues, was tested for the requirement of peptidyl Hyp for galactosylation. ExtP, a chemically synthesized extensin peptide consisting of contiguous peptidyl Hyp residues, tested whether contiguous peptidyl Hyp residues act as potential acceptors. Two

200 commercially available pectic polysaccharides, Rhamnogalactan I from potato and

Rhamnogalactan (a mixture of RGI and RGII) from soybean, were also tested as potential substrates acceptors. All the non–AGP substrate acceptors, including [AP]7, failed to incorporate [14C]Gal, indicating the GALT activity was specific for AGP sequences containing non-contiguous peptidyl Hyp (Figure 4.4). It is interesting to note that GALT2 and GALT5 expressed in tobacco expressed displayed higher activity than when expressed in Pichia, even after taking into account the relatively high background activity in tobacco. This indicates that there are plant-specific factors or accessory proteins critical for Hyp-O-GALT activity (Basu et al., 2013; Basu et al., 2015).

201

Figure 4.4 Substrate specificity of GALT 3 and 6 was monitored using potential substrates for GALT assay. Permeablized microsomal membranes obtained from the transiently expressed GALT 3 and GALT6 fused with 6x His-tags was served as the enzyme source in the GALT reactions. [AO]7, [AO]14, and d[AO]51 contain 7, 14, and 51 [AO] units, respectively. Microsome obtained from tobacco leaves infiltrated with an empty pMDC32 was used as negative control. A chemically synthesized extensin peptide (ExtP) contains repetitive SO4 units. [AP]7 contains 7 [AP] units. Rhamnogalactan I (RGI) from potato and RG from soybean represent pectin polymer substrates. Enzyme reactions were done in triplicate and mean values +SE are presented (*, P < 0.05; **, P < 0.01).

4.3.4. Expression profile of the nine Hyp-O-GALTs

qRT-PCR and data mining of public databases were used to analyze expression profiles of the Hyp-O-GALT genes. qRT-PCR analysis indicated that GALT1-6 are broadly expressed and have overlapping but distinct expression patterns (Figure 4.5).

These qRT-PCR data were in good agreement with public expression data available from GENEVESTIGATOR and the eFP browser (Zimmermann et al., 2004; Winter et al., 2007) as well as from the previous study by Strasser et al. (2007) (Figure 4.6). Data

202 from large-scale transcriptomic databases were used to provide insight into GALT expression and provide clues as to where to focus phenotypic analysis of GALT knockout mutant plants. Notable patterns of expression were as follows: highest expression of GALT6 was observed in senescent leaves followed by seed, seed coat, root hairs, flowers and siliques, whereas GALT4 was predominantly expressed in young flowers, mature flowers with siliques and mature siliques. GALT3 was abundant in roots, mature pollen, and hypocotyl (Figure 4.6).

Numerous studies indicate that pollen tubes undergo dramatic transformations while growing in the pistil, where they rapidly grow, perceive and respond to navigational cues secreted by the pistil, with AGPs playing a critical role in such interactions

(Tetsuya, 2010; Cheung et al., 2008; Suárez et al., 2013). Nonetheless, genes expressed by pollen tubes in response to growth in the pistil are poorly characterized. Qin et al.

(2009) utilized the novel combination of semi in vitro pollination followed by microarray analysis to identify genes specifically involved in pollen-pistil interaction, including the Hyp-O-GALTs. GALT5 had the highest expression followed by GALT2 and

GALT4, whereas HPGT3 was only expressed in latter stages of pollen elongation.

Furthermore, it is interesting to note that there was a temporal difference in the expression patterns of these Hyp-O-GALTs during pollen elongation (Figure 4.6).

In addition, transcriptome analyses using RNA extracted from laser-capture dissected seed coat tissue (http://seedgenenetwork.net/arabidopsis) indicated that all five

Hyp-O-GALT transcript levels displayed unique expression patterns in the seed coat during embryogenesis (Figure 4.7) (Le et al., 2010). Notably, GALT6 was expressed

203 throughout seed development, while expression of GALT2 and GALT5 transcripts was higher in the early stages compared to the latter stages of seed development. In contrast,

GALT4 was only observed at later stages of seed development, while GALT3 showed the least expression in seeds.

Figure 4.5 Expression pattern of the six candidate GALT genes. qRT-PCR analysis of GALT1, GALT3, GALT4 and GALT6 expression in major organs during growth and development. The level of expression is calculated relative to the UBQ10 gene (mean ± SD of three biological replicates).

204

Figure 4.6 Gene expression profile of nine GALTs in different organs/tissues. (A) Genevestigator developmental expression plot of the indicated genes (Zimmermann et al., 2004). (B) Transcript profiling of the indicated genes using microarray data of semi-in vitro germination of pollen tubes (Qin et al., 2009). (C) Analysis of developmental expression using eFP browser (Winter et al., 2007).

205

Figure 4.7 Transcript levels of the eight GALTs in the developing seed coat depicted by http://seedgenenetwork.net/ (by Le et al., 2010). Five stages of seed development have been monitored for investigating gene networks in seed, namely pre-globular, globular, heart stage, linear cotyledon and mature green seed stage. PEN- Peripheral , GSC-General seed coat, EP- Embryo proper, S- Suspensor, ME-Micropylar endosperm, CE- Chalazal endosperm and CSC- Chalazal seed coat.

206

4.3.5. GALT3, GALT4 and GALT6 are targeted to the Golgi vesicles

Transient expression of C-terminal YFP fusions to GALT3, GALT4, and GALT6 were infiltrated in N. tobaccum epidermal leaf cells to examine the subcellular localization of these enzymes (Figure 4.8). Overlays of GALT3-YFP, GALT4-YFP and

GALT6-YFP individually co-expressed with the Golgi marker protein, sialyl transferase, fused to GFP (ST-GFP) indicated that all three GALTs were localized to the Golgi apparatus. Furthermore, the possibility that they were localized in the ER was excluded, as the GALT-YFP fusion constructions were not co-localized with the ER marker,

HDEL fused with GFP (HDEL-GFP).

207

Figure 4.8 Subcellular localization of transiently expressed GALT3, GALT4 and GALT6-YFP in N. benthamiana. GALT3-YFP, GALT4-YFP and GALT6-YFP fusion protein were expressed under the control of the CaMV 35S promoter. Transiently expressed GALT3-YFP, and GALT6- YFP co-localized with sialyl transferase (ST)-GFP fusion protein (a Golgi marker) as well as with HDEL-GFP fusion protein (an ER marker). These constructs were examined by laser-scanning confocal microscopy under fluorescent and white light, and the fluorescent images were merged to observe co-localization.

208

4.3.6. Loss-of-function mutations in GALT3, GALT4 and GALT6 showed AGP specific biochemical defects

Two independent allelic mutant lines with T-DNA insertions were identified for each of the six GALTs genes in order to examine the biochemical roles of the Hyp-O-

GALTs in vivo. Homozygous mutants were generated, identified by PCR, and confirmed by sequencing (Figure 4.9). The qRT-PCR analysis showed that no transcript could be detected in the mutants (Figure 4.9). Significant reductions in GALT activity as well as

β-Yariv precipitable AGPs obtained from 14-d old seedlings were observed in knock-out mutants of GALT3 (galt3-1 and galt3-2), GALT4 (galt4-1 and galt4-2), and GALT6

(galt6-1 and galt6-2) compared to WT (Table 4.5). Such reductions were previously reported for knock-out mutants of GALT2 (galt2-1 and galt2-2), GALT5 (galt5-1 and galt5-2) and a galt2galt5 double mutant and were used here as positive controls (Basu et al, 2015). Consistent with the findings that GALT1 synthesizes Lewis a structures and lacks Hyp-O-GALT activity (Figure 4.3), knock-out mutants of GALT1 (galt1-1 and galt1-2) demonstrated no such reductions and were indistinguishable from WT (Table

4.5) (Strasser et al., 2007).

Given the differential expression of these Hyp-O-GALTs and the broad expression of AGPs, AGPs were also quantified from other organs in the mutants. Similar patterns of reductions in β-Yariv precipitable AGPs were observed in these other organs for these mutants. However, consistent with expression profile data, β-Gal Yariv precipitable

AGP content was especially reduced in senescent leaves (~19-23%) and stems (~24-

25%) of galt6 mutants, stems of galt2 and galt5 single and double mutants (~29-34%),

209 and siliques of galt4 mutants (~34-39%) (Table 4.6). Profiles of these β-Yariv precipitable AGPs produced by RP-HPLC were also examined for various galt mutants and revealed that virtually all these AGPs, as opposed to a single or subset of these

AGPs, were affected when compared to WT or galt1 control profiles (Figure 4.10).

Furthermore, the AGP peaks in the galt3, galt4, and galt6 mutants eluted later and thus had less glycosylated protein compared to the WT or galt1 control AGP peaks, consistent with reduced Hyp-galactosylation.

210

Figure 4.9 Schematic gene models and transcript analysis of GALT1, GALT3, GALT4 and GALT6 and T-DNA mutants. (A) GALT1, GALT3. GALT4 and GALT6 gene structure and T-DNA insertion sites in galt1-1, galt1-2, galt3-1, galt3-2, galt4-1, galt4-2, galt6-1 and galt6-2 mutants. The intron-exon structures of GALT1, GALT3, GALT4 and GALT6 are indicated (introns are drawn as lines and exons as rectangles, with white rectangles representing coding sequences and black rectangles representing UTRs). Sites of T-DNA insertions are marked (triangles) as are the locations of primer sequences (arrows) used for PCR

211 screening. (B) RT-PCR analysis of transcripts from rosette leaves of 14-d-old wild type (Col-0), the allelic homozygous galt1, galt3, galt4 and galt6 mutant lines. Arrows indicate the position of primers (denoted as RTF and RTR) used for RT-PCR analysis of transcript levels. UBQ10 primers were used as internal controls. (C) qRT-PCR analysis was performed to quantify and compare the transcript levels of the indicated genes with that of the wild type.

Table 4.5 GALT activity and amount of β-Yariv precipitated AGPs in WT and galt mutants Genotype GALT activity β-Yariv precipitated AGPs

(pmol/hr/mg) (µg/g) WT 7.10 + 0.90 13.60 + 0.75

galt1-1 6.80 + 0.37 13.30 + 0.95 galt1-2 7.20 + 0.65 13.30 + 0.95

b b galt2-1 5.53 + 1.20 9.91 + 2.80 galt2-2 5.90 + 1.01b 9.28 + 1.50b

a galt3-1 6.08 + 1.20 13.30 + 0.95 galt3-2 5.51 + 1.01b 12.30 + 0.80a

a a galt4-1 6.04 + 2.20 12.00 + 1.10 galt4-2 5.83 + 1.50b 11.90 + 1.20a

b b galt5-1 5.45 + 1.10 7.90 + 2.10 galt5-2 4.90 + 1.50b 8.10 + 1.20b

b a galt6-1 5.30 + 0.44 10.90 + 0.59 galt6-2 5.00 + 1.71b 10.30 + 1.54a

b b galt2galt5 4.64 + 0.54 5.63 + 0.39

Detergent-solubilized microsomal fractions were used for performing a standard GALT assay, and AGPs were extracted, precipitated by β-Yariv reagent, and quantified from 14-d-old plants. The values are averages of at least two independent experiments from two biological replicates. Student’s t tests were performed to determine statistical significance (a P < 0.05, b P <0.01).

212

Figure 4.10 HPLC profile of AGPs obtained from wild type and single galt mutants. RP-HPLC profiles of AGPs prepared from 14 d old seedlings of WT and galt mutant lines by precipitation with βYariv. The x axis is retention time in minutes. The y axis is absorbance at 215 nm.

213

4.3.7. Disruption in GALT3, GALT4 and GALT6 exhibits root hair defects

To investigate the physiological function of these six GALTs in vivo, mutants were grown on MS plates and compared to WT. No significant phenotypic differences in primary root growth were observed with the exception of the root hairs. Single mutant knock-out lines for GALT3, GALT4, GALT6, as well as for GALT2 and GALT5 and the galt2galt5 double mutant, displayed shorter and less dense root hairs compared to WT and knock-out lines for GALT1 (Figure 4.11).

214

Figure 4.11 Root hair length and density reduced in the galt3, galt4 and galt6 mutant. (A) Wild type, galt1, galt3, galt4 and galt6 plants were grown on MS agar plates for 10 d. Bars = 1mm. (B) Quantification of root hair length and (C) density of the galt mutants. The asterisks indicate significantly reduced root hair length and density compared with wild-type controls according to a Student's t test (*, P < 0.05; **, P < 0.01; n > 300).

215

4.3.8. Disruption in GALT4 and GALT6 leads to reduction in seed number

The galt4 and galt6 mutants displayed a 16% and 13% reduction in seed set, respectively (Figure 4.12 and Table 4.6). Reciprocal crosses of the galt4 and galt6 mutants to wild type plants were performed to determine whether this defect was conferred by the male or female gametophyte. Such crosses indicated that the male gametophyte of these mutants exclusively conferred reduced seed set (Figure 4.12 and

Table 4.6). Pollen were consequently examined with Alexander’s stain which indicated that pollen were viable (Figure 4.13). Furthermore, in vitro pollen germination did not exhibit altered germination frequency in galt4 and galt6 mutants compared to WT

(Figure 4.13).

216

Figure 4.12 Morphology of the siliques from plants after crossing galt4 × wild type, galt6 × wild type compared with wild type and homozygous mutant galt4 and galt6. The siliques were treated with ethanol (thus turned brown) for easy observation of the seeds. Absence of ovules was indicated with an asterisk.

217

Figure 4.13 Determination of pollen viability using Alexander's staining solution. Bar = 200 µm.

218

Table 4.6 Weight, length, and seed number from WT and galt siliques Genotype Silique Length Seeds/Silique Seed weight (mm) (mg) WT x WT 12.90 + 0.87 56.10 + 3.80 4.50 + 0.34 galt1-2 x galt1-2 13.04 + 0.84 54.87 + 2.70 4.65 + 0.28 galt2-1 x galt2-1 12.51 + 0.22 52.45 + 3.52a 4.20 + 0.91 galt2-2 x galt2-2 13.21 + 0.34 53.65 + 2.93a 4.65 + 0.44 galt3-1 x galt3-1 13.06 + 0.68 52.12 + 3.29a 4.63 + 0.34 galt3-2 x galt3-2 12.80 + 0.77 53.37 + 2.66a 4.50 + 0.37 galt4-1 x galt4-1 13.06 + 0.56 47.37 + 2.28b 3.26 + 0.40a galt4-2 x galt4-2 12.85 + 0.59 47.50 + 2.44b 3.41 + 0.32a galt5-1 x galt5-1 13.32 + 0.34 53.67 + 3.4a 4.23 + 0.54 galt5-2 x galt5-2 13.65 + 0.89 55.28 + 2.7 4.67 + 0.89 galt6-1 x galt6-1 13.10 + 0.57 49.11 +4.24b 3.41 + 0.18a galt6-2 x galt6-2 13.60 + 0.56 50.56 +2.79b 3.72 + 0.27a WT x galt4-1 13.10 + 0.73 45.10 +6.40b 3.70 + 0.56a WT x galt4-2 12.91 + 0.45 43.45 +4.90b 3.54 + 0.38a galt4-1 x WT 13.06 + 0.56 53.37 +4.28a 4.56 + 0.40 galt4-2 x WT 12.85 + 0.59 52.10 +1.40a 4.34 + 0.62 WT x galt6-1 13.00 + 0.54 49.70 +7.40b 3.50 + 0.56a WT x galt6-2 12.88 + 0.47 50.60 +4.40b 3.50 + 0.56a galt6-1 x WT 13.06 + 0.71 53.37 +4.28a 4.56 + 0.40 galt6-2 x WT 13.13 + 0.96 53.37 +4.28a 4.56 + 0.40 Siliques were obtained from 6-week-old plants (n=20). Letter ‘a’ and ‘b’ denotes a significant difference from the wild type (P <0.05, P<0.01) respectively.

219

4.3.9. Knockout mutation of GALT3 and GALT6 results in reduced staining of adherent mucilage

Prior evidence for the involvement of AGPs (SOS5) and GALT2/GALT5 in seed coat mucilage prompted an examination of the potential functions of GALT3, GALT4 and GALT6 in modifying seed coat mucilage (Basu et al., 2015; Harpaz-Saad et al.,

2011; Griffiths et al., 2014). The effect of disruption of the six GALT gene family members on adherent seed mucilage was investigated by staining hydrated seeds with ruthenium red, which stains negatively charged biopolymers such as pectin (Koornneef,

1981). The galt3-1, galt3-2, galt6-1, galt6-2 and galt2galt5 mutant seeds showed a thin staining pattern of the adherent mucilage, whereas WT, galt2, galt4, galt5, and galt1 seeds showed an intense, regular, spherical staining pattern (Figure 4.14). In addition, adherent mucilage mass and volume were measured to confirm the reduction of adherent mucilage thickness. No difference was observed in adherent mucilage mass between WT and galt single and double mutant seeds, whereas the adherent mucilage size of galt3, galt6 and galt2galt5 was substantially reduced (20~30%) compared with WT (Table

4.7). In contrasts, galt1, galt4, galt2 and galt5 mutants were less dramatically altered

(5~13%) compared to WT.

In order to confirm and quantify the changes in non-adherent (soluble) and adherent mucilage, WT and galt mutant seeds were analyzed. Sequential extraction of seeds with ammonium oxalate, 0.2 N NaOH, and 2 N NaOH was performed to assess changes in the soluble and adherent mucilage (Table 4.8). Both galt6 and galt3 seeds had a significant increase in the total sugar present in the ammonium oxalate and 0.2 N NaOH

220 extracts (soluble and weakly attached pectins) compared to wild type seeds.

Consistently, all these galt mutants also displayed a decrease in total sugars present in the 2 N NaOH extracts, which represent the majority of the adherent mucilage and contain strongly linked pectins and cross-linking glycans/hemicelluloses (Arsovski et al., 2010; Huang et al., 2011). Such distinct differences were not observed in galt2, galt4, and galt5 single mutants nor in galt1 mutants; however, was a similar phenotype observed in galt2galt5 mutants.

221

Figure 4.14 Staining of seed coat mucilage for pectin in wild type and the single and double galt mutants. Seeds of the indicated genotypes were prehydrated with water for 90 min and stained with ruthenium red to visualize cellulose and pectin using Zeiss LSM 510 META laser scanning confocal microscope. Bar= 100 µm.

222

Table 4.7 Determination of adherent mucilage mass and size in WT and galt mutants Genotype Mass (µg) Size (mm3) WT 1.82 + 0.05 0.47 + 0.13 galt1-2 1.76 + 0.06 0.45 + 0.03 galt2-1 1.80 + 0.09 0.44 + 0.04a galt2-2 1.77 + 0.06 0.45 + 0.07a galt3-1 1.80 + 0.09 0.37 + 0.08b galt3-2 1.75 + 0.08 0.35 + 0.05b galt4-1 1.79 + 0.05 0.42 + 0.10a galt4-2 1.81 + 0.05 0.39 + 0.20a galt5-1 1.80 + 0.05 0.41 + 0.30a galt5-2 1.81 + 0.05 0.39 + 0.25b galt6-1 1.83 + 0.04 0.35 + 0.04b galt6-2 1.77 + 0.07 0.37 + 0.05b galt2galt5 1.75 + 0.09 0.30 + 0.05c The mass and size values are the average mass and size of adherent mucilage of 100 seeds of triplicate assays + SD. Letters ‘a’ and ‘b’ denotes a significantly difference from the wild type (P <0.05; P<0.01 respectively).

223

Table 4.8 Quantification of total sugars from WT and galt mucilage sequentially extracted using ammonium oxalate, 0.2 N NaOH and 2 N NaOH Extracta Genotype Ammonium 0.2N NaOH 2N NaOH oxalate WT 0.85 + 0.05 1.04 + 0.03 0.81 + 0.04 galt1-2 0.83 + 0.07 1.05 + 0.05 0.84 + 0.03 galt2-1 0.95 + 0.03b 1.12 + 0.04b 0.61 + 0.05b galt2-2 0.97 + 0.70b 1.19 + 0.03b 0.54 + 0.06b galt3-1 1.30 + 0.09 c 1.29 + 0.05 c 0.53 + 0.05b galt3-2 1.28 + 0.05 c 1.27 + 0.07 c 0.52 + 0.07b galt4-1 0.88 + 0.20 1.09 + 0.02 0.79 + 0.01 galt4-2 0.90 + 0.60b 1.01 + 0.06 0.73 + 0.03 galt5-1 0.95 + 0.10b 1.17 + 0.05b 0.63 + 0.05b galt5-2 0.90 + 0.08b 1.20 + 0.07c 0.60 + 0.04b galt6-1 1.25 + 0.09 c 1.30 + 0.07c 0.55 + 0.05b galt6-2 1.31 + 0.05 c 1.29 + 0.04c 0.61 + 0.08b galt2galt 1.40 + 0.09d 1.36 + 0.08d 0.47 + 0.05d 5 aIntact seeds were extracted sequentially with 0.2% ammonium oxalate, 0.2 N NaOH and 2 N NaOH, neutralized, and assayed by the phenol-sulfuric acid method against glucose standards. The results are shown as µg/mg of seeds. Analyses were performed in triplicate and results are given as µg/mg seed + SE. All genotypes were grown, harvested and stored together. Letters ‘a’ ‘b’ and ‘c’ denote a significantly difference from the wild type (P <0.05; P<0.01; P<0.001 respectively).

224

4.3.10 Disruption of GALT6 led to early senescence

Only the GALT6 mutants (galt6-1 and galt6-2) displayed early onset of senescence compared to WT and the other galt mutants. This was visualized by premature yellowing of leaves and was correlated with a slightly greater reduction in chlorophyll content and protein content in GALT6 mutants compared to WT (Figure 4.15). These observations were consistent with the abundance of GALT6 transcripts in senescent leaves as well as with the markedly greater reduction of β-Yariv precipitable AGPs in galt6 senescent leaves (Table 4.6).

225

Figure 4.15 Age-dependent leaf senescence phenotype of WT and galt6-1 and galt6- 2 plants. (A) and (B) depict enlarged images of 32 and 34 d old senescent leaves. (C) Progression of leaf senescence. (E) total protein content and F Chlorophyll content.

4.3.11. GALT3, GALT4, and GALT6 mutants exhibit pollen tube and root growth which is less sensitive to β-Yariv reagent

GALT3, GALT4 and GALT6 mutants displayed reduced inhibition of pollen tube and root growth elongation in response to β -Yariv reagent compared to WT or α -Yariv

226 reagent control treatments (Figures 4.16 and 4.17). Moreover, as expected, GALT1 mutants did not exhibit any difference in either pollen tube or root growth elongation compared to WT. Moreover, no significant difference in pollen tube or primary root growth elongation was observed in unsupplemented germination media, indicating the conditional nature of this phenotype (Figures 4.16 and 4.17).

227

Figure 4.16 The galt single mutants demonstrate reduced inhibition of pollen tube growth in response to β-Yariv reagent. (A) Representative images of pollen tubes from wild-type, galt1, galt3, galt4 and galt6 mutants after 16 h in pollen germination medium, and (B) in pollen germination medium supplemented with 30 µM β-Yariv and (C) in pollen germination medium supplemented with 30 µM α -Yariv reagent. Bar = 50 µm. (D) Pollen tube lengths (from wild-type, galt2, galt5, and galt2galt5 plants) were measured over 16 h in the pollen germination medium supplemented with 30 µM β -Yariv reagent. Twenty flowers from each genotype and 25 pollen tubes from each flower were measured using Image J. The experiment was done in triplicate and the values were subjected to statistical analysis by ANOVA, followed by the Tukey's honestly significant difference test (**, P <0.01 and *, P <0.05).

228

Figure 4.17 Reduced inhibition of primary root growth of galt3, galt4 and galt6 mutants in the presence of β-Yariv reagent. (A) galt1, (B) galt3, (C) galt4 and (D) galt6 and WT root lengths were measured for 7, 14 and 21 d after germination and seedling establishment for 5 d on MS plates, on MS plates supplemented with 50 µM α-Yariv reagent, and on MS plates supplemented with 50 µM β- -Yariv reagent. Statistical differences were determined by ANOVA, followed by the Tukey's honestly significant difference test. Asterisks represent the statistical significance between genotypes (*, P < 0.05; **, P < 0.01) within a treatment group. Vertical bars represent mean ± SE of the experimental means from at least three independent experiments (n =5), where experimental means were obtained from 10 to 15 seedlings per experiment.

229

4.3.12. Conditional salt hypersensitive phenotype in galt mutants

GALT3 (galt3-1 and galt3-2) and GALT6 (galt6-1 and galt6-2) mutants, and to a lesser extent the GALT4 (galt4-1 and galt4-2) mutants, exhibited significant reductions in root elongation compared to WT when grown in the presence of 100 and 150 mM

NaCl (Figure 4.18 and Figure 4.19). Such reductions in root elongation were previously reported for knock-out mutants of GALT2 (galt2-1 and galt2-2), GALT5 (galt5-1 and galt5-2) and the galt2galt5 double mutant (Basu et al., 2015). As expected, GALT1 mutants did not show salt hypersensitive growth and were indistinguishable from WT in this assay.

A root bending assay was used as another means to evaluate salt hypersensitivity of the GALT mutants (Figure 4.20). This assay is commonly used by plant researchers to evaluate salt sensitivity/tolerance and involves monitoring root growth reorientation after a 180 degree reorientation of the seedling to gravity. Results of this experiment indicated that GALT3 (galt3-1 and galt3-2) and GALT6 (galt6-1 and galt6-2) mutants, and to a lesser extent the GALT4 (galt4-1 and galt4-2) mutants were slow to reorient their root growth compared to WT when grown in the presence of 100 mM NaCl. Such delayed reorientation was previously reported for knock-out mutants of GALT2 (galt2-1 and galt2-2), GALT5 (galt5-1 and galt5-2), and the galt2galt5 double mutant; these mutants were used here as positive controls (Basu et al., 2015). Moreover, as expected,

GALT1 mutants (galt1-1 and galt1-2) reoriented quickly and were indistinguishable from WT in this assay.

230

Microscopic examination of the GALT3, GALT6, and to a lesser extent the GALT4 mutants also revealed defective anisotropic root tip growth (i.e., root tip swelling) in the presence of 100 mM NaCl, which was not observed in WT and GALT1 mutants (Figure

4.21). Such salt hypersensitive root tip swelling responses were previously reported in galt2, galt5, and galt2galt5 mutants and were included here as positive controls (Basu et al., 2015).

Lamport et al. (2006) have demonstrated up-regulation of AGPs in response to salt stress and Shi et al. (2008) identified a fasciclin domain containing AGP, namely SOS5 to be involved in salt stress responses. This has prompted us to investigate the effect of salt stress on the galt mutants. It was observed that the galt3 and galt6 and to lesser extent galt4 but not galt1 roots exhibited significant reduction in root elongation in response to 100 and 150 mM NaCl compared to the WT, as well as displayed delayed root bending phenotype and defective anisotropic root tip growth in response to 100 mM salt stress (Figure 4.18, 4.19, 4.20 and 4.21).

231

Figure 4.18 Salt induced inhibition of primary root elongation in galt3, galt4 and galt6 mutants. Five-day-old wild-type, (A) galt1, (B) galt3, (C) galt4 and (D) galt6 seedlings germinated on MS medium were transferred onto media containing 100 mM NaCl and grown vertically. Root elongation (i.e., increase in length after transfer) was measured after 7, 14 and 21 d of growth in the non-permissive media. Data are the means ± SE of measurements from five independent experiments (total n = 100). Statistical differences were determined by ANOVA, followed by the Tukey's honestly significant difference test (*, P <0.05 and **, P <0.01).

232

. Figure 4.19 Root elongation in response to 100 mM NaCl Representative images of WT, (A) galt1, (B) galt3, (C) galt4 and (D) galt6 plants after 14 d of growth on MS plates supplemented with 100 mM NaCl. Size bar = 1 cm.

233

Figure 4.20 Root-bending assay of wild type, galt1, galt3, galt4 and galt6 mutant seedlings. Five-day-old seedlings grown on MS plates were transferred to MS plates with 100 mM NaCl and reoriented at an angle of 180° (upside down). The photographs were taken for knock-out mutants of (A) GALT1, (B) GALT3, (C) GALT4 and (D) GALT6 seedling 7 d post transfer in the MS plates supplemented with 100 mM NaCl. Bar = 10 mm. (E) Analysis of root curvature in WT, galt1, galt3, galt4, galt5 and galt6 mutant plants. Statistical differences were determined by one way ANOVA and ‘a’ denotes a significant difference of root curvature (P<0.05) between WT and single galt mutants,

234

‘b’ denotes a significant difference of root curvature (P<0.01) between galt single mutants and galt2galt5 mutants. Vertical bars represent mean ± SE of the experimental means from at least two independent experiments (n =5), where experimental means were obtained from 15 seedlings per experiment

Figure 4.21 Conditional root anisotropic growth defects of single and double galt mutants compared to the wild type plants. (A) Light microscopic images of root tips of plant seedlings from indicated genotypes grown for 7d in MS plates with 100 mM NaCl. Seeds were germinated in MS plates and grown for 3d before transferring to the MS plates supplemented with 100 mM NaCl. Bar

235

=1mm. (B) Analysis of root curvature in WT and galt mutants. Statistical differences were determined by one way ANOVA (*P<0.05, ** P<0.01, ***P<0.001) between WT and single or WT and double mutants.

4.4. Discussion

4.4.1. GALT2-6 encodes Hyp-GALTs for AGPs and are widely expressed in

Arabidopsis

Bioinformatic approaches were previously used to identify a small, six-membered gene family within the GT 31 family of the CAZY database as potential candidates for encoding Hyp-O-GALTs for AGPs (Basu et al., 2013; Qu et al., 2008; Egelund et al.,

2011). Protein members of this family were designated GALT1-6 and distinguished by the presence of a GALT domain as well as a GALECTIN domain. Previously, GALT1 was shown to catalyze galactose addition for formation of the Lewis a epitope on N– linked glycans (Strasser et al., 2007), while GALT2 and GALT5 were shown to act as

Hyp-GALTs specific for AGPs (Basu et al., 2013; Basu et al., 2016). In this study, biochemical and genetic evidence indicates that GALT3, GALT4, and GALT6 also act as Hyp-GALTs for AGPs.

Heterologous transient expression of GALT1-6 in tobacco epidermal cells demonstrated a significant increase in Hyp-GALT activity compared with various tobacco control plants, with the notable exception of GALT1 (Figure 4.3 and Table 4.4).

The absence of Hyp-O-GALT activity over background levels in case of GALT1 is consistent with its reported, non-AGP related function and provided a useful control for the studies reported here (Strasser et al., 2007). Moreover, this transient expression study in tobacco corroborates previous findings that GALT2 and GALT5, which were

236 expressed in Pichia pastoris, act as AGP-specific Hyp-GALTs (Basu et al., 2013; Basu et al 2015). It should be noted that the amount of activity detected in these two heterologous expression systems varied; the tobacco system showed much higher levels of activity than the Pichia system, even when taking into account the higher level of endogenous activity associated with the tobacco system. This observation likely reflects the need for other plant-based proteins or factors to enhance or optimize Hyp-O-GALT activity.

Substrate specificity of GALT2-6 was investigated using various potential acceptor substrates and demonstrated that GALT2-6 is specific for AGP sequences (Figure 4.4).

These findings are consistent with the Hyp contiguity hypothesis, which states that clustered, non-contiguous Hyp residues are sites of AG addition, whereas contiguous

Hyp residues are sites for the addition of Ara oligosaccharides (Kieliszewski et al.,

1995; Kieliszewski and Shpak, 2001). Despite the higher enzyme activity observed in the transient tobacco expression system compared to Pichia, the same size-dependent preference for AGP substrates was observed for GALT2 and GALT5 in both systems, where [AO]7 was the preferred substrate (Basu et al., 2015). GALT3, GALT4, and

GALT6 also acted in a similar manner within the tobacco system.

Heterologous transiently expressed GALT2-6 in tobacco microsomes have similar biochemical properties to the GALT(s) present in Arabidopsis microsomal membranes and to GALT2 and GALT5 expressed in Pichia, specifically all require UDP-Gal as the sugar donor (Basu et al., 2013; Basu et al., 2015; Liang et al., 2010; Oka et al., 2010).

237

Genetic mutant analysis provides additional in vivo evidence that GALT3, GALT4, and GALT6 function as Hyp-GALTs, similar to GALT2 and GALT5 (Figure 4.9,

Tables 4.1 and 4. 2). Allelic galt knock-out mutants for all these genes exhibit reduced

(i.e., 15-35% less) Hyp-GALT activity and contain considerably less (i.e., 10-60% less) glycosylated (i.e., β-Yariv precipitiable) AGPs. In addition, HPLC AGP profiling of the galt3, galt4, and galt6 mutants extends these findings and indicates that their activity is not limited to a particular AGP or a small subset of AGPs, but instead broadly acts on coexpressed AGPs, similar to that previously reported for galt2 and galt5 mutants

(Figure 4.10).

The qRT-PCR analysis of GALT1-6 was performed to examine their expression patterns and provide information relevant to phenotypic analysis of their corresponding allelic mutants (Figure 4.5). All six genes were widely expressed, and in the cases of

GALT2-6 are consistent with the widespread distribution of AGPs and the multiple functions associated with them. These patterns were corroborated by searching public expression databases, which revealed even broader organ and tissue expression patterns

(Figure 4.5, 4.6 and 4.7). HPGT1-3, three recently identified Hyp-O-GALTs for AGPs that lack GALECTIN domains, were also included in this analysis and showed equally broad patterns of gene expression. Nonetheless, within a given organ or tissue, the Hyp-

O-GALT genes exhibit both temporal and spatial differences in their expression patterns.

Transcriptome analyses using RNA extracted from laser-capture dissected seed coat tissue (Le et al., 2010) provides a particularly striking illustration of the diverse tissue- specific expression patterns of GALT2-6 and HPGT1-3 (Figure 4.7).

238

4.4.2. GALT3, GALT4, and GALT6 are localized to Golgi vesicles

Various approaches have indicated that AGP glycosylation occurs in Golgi vesicles.

These approaches include bioinformatics predictions using Signal P and Golgi predictor, biochemical experiments on hydroxyproline-rich glycoprotein (HRGP) biosynthesis

(Emanuelsson et al., 2007;Gardiner and Chrispeels, 1975; Robinson and Glas, 1982;

Nikolovski et al., 2012) a proteomics technique for localization of organelle proteins by isotope tagging (Nikolovski et al., 2012), and localization studies performed with other

AGP GTs, including GALT2, GALT5, HPGT1-3, AT1G77810, GALT31A, GALT29A,

GlcAT14A, and FUT6, Basu et al., 2015; Basu et al., 2016; Geshi et al., 2013; Qu et al.,

2008; Ogawa-Ohnishi and Matsubayashi, 2015; Knoch et al., 2013). Given their similarity to GALT2 and GALT5 and their demonstrated Hyp-GALT activity, GALT3,

GALT4, and GALT6 were expected to reside in the Golgi vesicles, and this was confirmed by heterologous expression of fluorescently tagged protein fusions in tobacco leaves (Figure 4.8). Interestingly, only GALT2 is found in both the ER and Golgi, indicating that Hyp-galactosylation may be initiated in the ER, but completed in the

Golgi where the bulk of the Hyp-O-GALTs are located (Basu et al., 2015).

4.4.3. GALT mutant phenotypes reveal functional roles of AGP glycosylation in normal growth and development

Genetic mutant analysis was used to investigate and compare the in vivo functional contributions of AGP glycosylation by GALT3, GALT4 and GALT6 with that of GALT2 and GALT5 (Figure 4.9 and Table 4.1). To date, a variety of functions are attributed to certain GTs acting on AGPs based on mutant analysis; these mutants show embryo

239 lethality, conditional defects of primary root growth, cell elongation, and pollen tube growth (Table 4.9) (Liang et al., 2014; Tryfona et al., 2014; Basu et al., 2015; Ogawa-

Ohnishi and Matsubayashi, 2015; Geshi et al., 2013; Gille et al., 2013)]. Like galt2 and galt5, galt3, galt4 and galt6 single mutant lines showed subtle or no detectable growth phenotypes under normal soil-based growth conditions, which is likely attributed to the functional redundancy within the GALT2-6 gene family (Basu et al., 2015). The subtle phenotypes that were displayed by the single mutants here included reduced root hair length and density for galt3, galt4, and galt6 (Figure 4.11), reduced seed set for galt4 and galt6 (Figure 4.12 and Table 4.6), reduced seed mucilage for galt3 and galt6

(Figure 4.14, Table 4.7 and 4.8), and accelerated leaf senescence for galt6 (Figure

4.15). Although these phenotypes are consistent with the expression profiles of these genes, it would be difficult to predict such phenotypes from expression data alone. It is anticipated that double and multiple galt mutants will show more pronounced mutant phenotypes, as was the case when galt2galt5 double mutants were produced and characterized (Basu et al., 2015).

With respect to root hair length and density, knock-out mutants of GALT3 and

GALT6, and to a lesser extent GALT4, display shorter and less dense root hairs, indicating glycosylated AGPs play a role in tip growth of root hairs in Arabidopsis

(Figure 4.11). The galt2, galt5, and galt2galt5 double mutants also demonstrate this response and were used here as positive controls (Basu et al., 2015). Other studies, using

Yariv reagent and prolyl hydroxylase genetic mutants, have also indicated that HRGPs

240

(i.e., AGPs and extensins) are involved with root hair growth (Baumberger et al., 2003;

Marzec et al., 2015; Velasquez et al., 2011; Velasquez et al., 2015).

With respect to seed mucilage, galt3 and galt6 mutants display an adherent seed mucilage layer of decreased width upon staining with ruthenium red when compared to

WT seeds, which indicates that glycosylated AGPs are involved in maintaining the adherent mucilage layer in seeds (Figure 4.14, Table 4.7 and Table 4.8). The galt2galt5 double mutants also demonstrate this response and were used here as positive controls

(Basu et al., 2015). Harpaz-Saad et al. 2011 reported that FEI2, a cell wall leucine–rich receptor-like kinase, is critical for the synthesis and proper deposition of cellulosic rays in seed coat mucilage, which coincides with an increase in solubility of the pectinaceous component of seed coat mucilage. The GALT6 expression profile during seed coat differentiation is identical to that of FEI2. Moreover, GALT6 is expressed at a higher level during late embryogenesis compared to the early embryogenesis stages, consistent with its involvement with seed coat development and in seed coat mucilage adherence.

Taken together, GALT6, GALT3, GALT2 and GALT5 likely glycosylate AGPs that are essential for maintaining mucilage adherence in seeds during hydration.

With respect to seed set, galt4 and galt6 mutants phenocopied knock-out mutants of two pollen-specific AGPs genes, AGP6 and AGP11, in terms of reduction in number of seeds per silique, but not in abnormal pollen structures, (Figure 4.12, Table 4.6 and

Figure 4.13). Several studies have indicated that AGPs play key roles in pollen biogenesis, pollen tube growth and development of pollen tube guidance and pollen– pistil interaction during post pollination events (Jia et al., 2015; Coimbra et al., 2010;

241

Cheung et al.1995; Levitin et al., 2008; Coimbra et al., 2009; Pereira et al., 2014). .

Indeed, a number of AGP genes are reported to show pollen-specific expression in

Arabidopsis, including AGP6, AGP11, AGP23 and AGP40. Reciprocal crosses of the mutants with WT confirmed that genetic transmission of this phenotype is contributed by the male gametophyte (Figure 4.12 and Table 4.6). In contrast, it is interesting to note that AGP18 is reported to be essential for female gametogenesis, given that functional megaspores in RNAi plants fail to enlarge and divide, resulting in ovule abortion and reduced seed set (Acosta-Garcia et al., 2004).

With respect to leaf senescence, galt6 mutants displayed age-dependent early onset of leaf senescence, indicating AGP glycosylation is related to plant aging (Figure 4.15).

Several lines of evidence implicate the involvement of AGPs in regulating program cell death and senescence (Moreau et al., 2005; Schindler et al., 1995; Gao et al., 2000;

Chaves et al., 2002; Sun et al., 2004).

For example, β-Yariv treatment is known to promote programmed plant cell death, while overexpressing AGPs results in tomato plants with enhanced lifespansWillats and

Knox, 1996; Mollet et al, 2002; Nguema-Ona et al., 2007; Nguema-Ona E et al., 2012)

Any one defect leading to reduced AG glycosylation is likely to impair the function of multiple AGPs, leading to pleiotropic effects as observed here. At least some of the genes involved in AGP glycosylation, however, exist in small redundant or partially redundant genes families and may compensate for one another when a given gene in the family is knocked out (Basu et al., 2013; Basu et al., 2015). In some cases, aberrant

242 phenotypes may not be discernable under normal conditions, but may be revealed under suboptimal growth conditions.

4.4.4. Conditional phenotypes indicate GALT3, GALT4, and GALT6 function in tip growth

The galt3, galt4, and galt6 mutants display several conditional pollen and root phenotypes in response to β-Yariv treatment or salt treatment. In both pollen tubes and roots, β-Yariv treatment is known to bind AGPs, specifically to their β -1,3-galactan chains, and inhibit pollen tube and root elongation (Willats and Knox , Mollet et al.,

2002indicating AGPs are important for such growth. This inhibition is alleviated in single mutants, which have reduced AGP glycosylation due to the lack of respective

GALTs (Figures 4.16 and 4.17). Under normal conditions (i.e., without β-Yariv treatment), no inhibition is observed in the single mutants, most likely due to gene redundancy, a conclusion supported by the observation that galt2galt5 double mutants show inhibition under normal conditions (Basu et al., 2015). Thus, GALT3, GALT4, and GALT6, like GALT2 and GALT5, are important for pollen tube and root growth and indicate that the AG polysaccharides are required for these growth functions.

In roots, salt treatment results in reduced root growth, which can be measured directly or by the root bending assay, which is commonly used to screen plants or mutants for salt sensitivity. Here, galt3, galt4, and galt6 mutants display salt hypersensitive root growth in both of these assays (Figure 4.18, 4.19 and 4.20). Thus,

GALT3, 4, and 6, like GALT2 and 5, function in root growth and indicate the importance of AG polysaccharides in this process. GALT3, 4, and 6, like GALT2 and 5,

243 also functions to prevent root tip swelling in response to salt stress (Figure 4.21). Since

SOS5, a fasciclin-like AGP, and FEI1 and FEI2, two cell wall receptor-like kinases, also prevent such root tip swelling, and are in the same genetic pathway as GALT2 and

GALT5, it appears that AGP glycosylation via any of the five GALTs (GALT2-5) likely generates a carbohydrate signal on SOS5, which is detected and transduced by the

FEI1/FEI2 kinases to promote cell wall integrity (Shi et al., 2003; Basu et al., 2015; Xu et al., 2008).

In conclusion, biochemical and genetic evidence presented here indicates that

GALT3, GALT4, and GALT6, like GALT2 and GALT5, function as AGP-specific

Hyp-GALTs. The largely, but not completely, overlapping pleiotropic effects observed in genetic null mutants for each of the genes with respect to multiple aspects of plant growth, development, and reproduction indicate the importance of AG polysaccharides to the functions of AGPs. Thus, these Hyp-GALT genes function in a largely redundant manner, and it is anticipated that more severe biochemical and physiological phenotypes will occur when multiple genetic mutants are studied, revealing additional AGP functions. Indeed, future work aimed at examining such multiple mutants, the role of the

GALECTIN domain, the potential for enzyme complex formation between and among

Hyp-GALTs and other GTs involved with AGP biosynthesis, Hyp-GALT/AGP trafficking, and the signaling roles of will provide deeper insight to the evolution and biology of this small enzyme family and the AGP family members that serve as their substrates.

244

Table 4.9 List of AGP specific GTs and well characterized AGPs Gene Function Mutants Mutant References Identifier Phenotype At4g21060 hydroxyproline- galt2-1 Reduced root Basu et al., (GALT2) O-β-galactosyl (SALK_ growth under 2013 transferase 11723) salt, radial galt2-2 swelling of (SALK_ root tips, At1g74800 14112) reduced seed (GALT5) galt5-1 mucilage (SALK_ adherence 10540) galt5-2 (SALK_ 11574) At5g53340 hpgt1-1 Longer lateral Ogawa- (HPGT1) (SALK_ roots and root Ohnishi et At4g32120 00754) hairs, radial al., 2015 (HPGT2) hpgt2-1 expansion of At2g25300 (SALK_ the cells in the (HPGT3) 07036) root tip, small hpgt3-1 leaves, shorter (SALK_ inflorescence 00940) stems, reduced fertility and shorter siliques

245

Table 4.9 continued AT1G7780 β-1,3- GT31 - Qu et al., galactosyl 2008 transferase At1g32930 β-1,6- galt31A Embryo lethal Geshi et al., (GALT31A) galactosyl (FLAG_ mutant 2013 transferase 379B0) At5g39990 β-1,6- glcat14a-1 Enhanced cell Knoch et al., (GlcAT14A) glucuronosy (SALK_ elongation in 2013; At5g15050 ltransferase 06433) seedlings Dilokpimol (GlcAT14B) glcat14a-2 and Geshi, At2g37585 (SALK_ 2014 (GlcAT14C) 043905) At1g70630 β-arabino ray1-1 (SALK_ Reduced root Gille et al., (RAY1) furanosyl 053158) growth, rosette 2013 transferase ray1-2 (GABI_ size and 001C0) inflorescence At2g15390 α-1,2-fucosyl fut4-1 Enhanced cell Wu et al., (FUT4) transferase (SAIL_284_B) elongation in 2010; Liang fut4-2 (SALK_ seedlings et al., 2013; 12530) Tryfona et At1g14080 fut6-1 (SALK_ al., 2014 (FUT6) 0783) fut6-2 (SALK_ 09950) At3g46550 Fasciclin sos5-2 Reduced root growth Shi et al., (FLA4) like AGPs (SALK_125874) under salt stress 2003 At2g24450 Fasciclin fla3 Development of male Li et al., 2010 (FLA3) like AGPs (SALK_016582) reproductive organs

246

Table 4.9 continued At5g55730 Fasciclin fla1-1 Shoot regeneration Johnson et al., (FLA1) like AGPs (CS01810, WS) 2011 fla1-2 (SALK_058964) At5g03170 Fasciclin fla11 Synthesis of MacMillan et (FLA11) like AGPs (SALK_046976) secondary cell wall al., 2010 At5g60490 fla12 (FLA12) SM.15162 At5g14380 Classical RNAi and Growth and Coimbra et (AGP6) AGP amiRNA development of pollen al., 2011 lines grain and pollen tube At3g01700 Classical RNAi and Growth and Coimbra et (AGP11) AGP amiRNA development of pollen al., 2011 lines grain and pollen tube At2g23130 Lys-rich agp17 influence Agrobacterium Yang et al., (AGP17) classical (SALK_101062) binding 2007

4.5. Materials and Methods

4.5.1. In silico analysis of six Hyp-O-GALTs

Protein sequences from GALT1, GALT3, GALT4 and GALT6 were run through several prediction programs (TMHMM 2.0, TargetP 1.1, SignalP v2.0.b2 server to obtain information on their putative subcellular localization and topology (Nielsen et al.

1999, Krogh et al. 2001, Emanuelsson et al. 2000). Hydrophobic cluster analysis (HCA) plots were generated using the drawhca server on the Internet

247

(http://smi.snv.jussieu.fr/hca/hca-form.html) and was analysed as described by Breton et al. (1998). The coexpression network among GT31 member genes was illustrated using the program GENEMANIA (www.genemania.org).

4.5.2. Plant lines and plant growth conditions

Arabidopsis thaliana accession Columbia-0 (Col-0) and two T-DNA insertion lines for At1g26180-(galt1-1, Sail_170_A08 and galt1-2, Salk_006871), At3g06440 (galt3-1,

Salk_085633 and galt3-2, Salk_005178), At1g127120 (galt4-1, Salk_136251 and galt4-

2, Salk_131723), and At5g62620 galt6-1, Sail_59_D08 and galt6-2, Sail_70_B02) were obtained from the Arabidopsis Biological Resource Center (ABRC, Ohio State

University). All plants used in this study were germinated after 4 days of stratification in the dark at 4°C were grown under long-day conditions (16h of light/8h of dark, 22 °C,

60% humidity) in growth chambers or growth rooms. Sequencing of the amplified fragment led to the accurate determination of the T-DNA insertion site.

4.5.3 Mutant confirmation by PCR and RT-PCR

Genomic DNA was isolated from leaves of the mutants and WT plants was extracted using 2X CTAB method and subsequent PCR analysis was carried out using gene specific primers in in conjunction with the T-DNA primers. The primer location is indicated in Fig. 4, and the corresponding primer sequences are listed in Table 4.10. For sequencing purposes, PCR products were purified by gel extraction (Wizard® SV Gel and PCR Clean-Up System, Promega, Madison, WI, USA) with and sequenced by the

Ohio University Genomics (http://www.dna.ohio.edu/) Facility. To analyze transcript levels of GALT1, GALT3, GALT4 and GALT6, total RNA was isolated from 2 week

248 old seedlings of wild-type and mutant plants using Trizol (Life Technologies, Grand

Island, NY, USA) and Direct-zol™ RNA MiniPrep (Zymo Research, Irvine, CA, USA).

First-strand cDNA synthesis was performed from 2 µg of total RNA using oligo-dT

(IDT) and GoScript reverse transcriptase (Promega, Madison, WI, USA). RT-PCR was performed using OneTaq DNA polymerase (New England Biolabs, Ipswich, MA, USA) and gene-specific RT primers (Table 4.10). The number of amplification cycles was 28 to evaluate and quantify differences among transcript levels before the reaction reached saturation.

For qRT-PCR the cDNAs were amplified using Brilliant II SYBR Green QRT-PCR

Master Mix with ROX (Agilent Technologies, La Jolla, CA, USA) in an MX3000P real- time PCR instrument (Agilent Technologies). PCR was optimized and reactions were performed in triplicate. The transcript level was standardized based on cDNA amplification of ubiquitin 10 (UBQ10, At4g05320) as a reference. Primer sequences are listed in Table 4.10.

Table 4.10 List of primers used in this study Purpose Forward Reverse Cloning GALT1 CACCATGCATCATCATCATC TTACCATTCGCGGCAGC ATCACAAGAGATTTTATGGA AAA GGGCTTC GALT3 CACCATGCATCATCATCATC TTATTCGCAGCAAATAG ATCACAAGCAATTCATGTCA ATTGGTTC GTGGT

249

Table 4.10 continued GALT4 CACCATGCATCATCATCAT TCATCTCATGTTGCAGCA CATCACAAGAAGTCTAAAC TTG TCGATAATTC GALT6 CACCATGCATCATCATCAT TCATCTCATGTTGCAGCA CATCACAGGAAGCCCAAGT CTG TGTCA Subcellular localization GALT3 CAGGACGTCTAGATGAAGC CATGACCGTCGACTTTTC AATTCATGTCAGTGGTGAG GCAGCAAATAGATTGGTT ATTC CTC GALT4 CAGGACGTCTAGATGAAGA CATGACCGTCGACTTTCT AGTCTAAACTCGAT CATGTTGCAGCATTG GALT6 CAGGACGTCTAGATGAGGC CATGACCGTCGACTTTCT CCAAGTTGTCA CATGTTGCAGCACTG Screening for T-DNA galt1-1 GALT1-1RP GALT1-1LP TTTTTCACAGCCGAAAAT TTGGGAACTTGTTTTTACCC CAC C galt1-2 GALT1-2RP GALT1-2LP GAGTTCCAGTAGCCAGGG TTCGAATAGGTTGAGAGTC AAG GG galt3-1 GALT3-1RP- GALT3-1LP- AGGCAAATGGAATAACTT TGGGGTTACTTCGCTTACAT GGC G galt3-2 GALT3-2RP- GALT3-2LP- ACTGGTTTCTTCGTGGTTG TGAATTGGTGCAGAAAGGA TG TC

250

Table 4.10 continued galt4-1 GALT4-1RP GALT4-1LP GATTAAACCCGAATCGAGT TTTGAACTTGGAATTTGGT CC CC galt4-2 GALT4-2RP GALT4-2LP GACTTCCTTTCTTGCATGCT GGACTCGATTCGGGTTTAA G TC galt6-1 GALT6-1RP GALT6-1LP AGAACACGAGTTTGTCCCA TTTTGGTCGATTTGCTTAA TG CC galt6-2 GALT6-2RP GALT6-2LP GATGCAAAGGTGTCACACA CTCGAGTTTTGACAACTTG TG GG LBa1.3 ATTTTGCCGATTTCGGAAC LB3 TAGCATCTGAATTTCATAA CCAATCTCGATACAC RT-PCR GALT1 RTF- RTR- CATCTTCGGGACAGAGGTT AAACCAAACGCTCTCTTTG G CTGCRTR- GALT2 RTF- RTR- TCTTTGTTGCACTTAATCCA TGTGGTCGACCTTTCAACA A AATTAT GAAG RTR- GALT3 RTF- TTATTCGCAGCAAATAGAT GTTGACTACTATGGTTTACT TGGTTCTC TAGCTTG

251

Table 4.10 continued GALT4 RTF- RTR- CTTTGTGGCATTGCATGCAA CCATCTTGAATAATCTTA GAAAG ATCGCTTTTG GALT5 RTF- RTR- TATGTGAACACGGAGCTCTT TCCATCTTGAACAGCCGT GCATTC AATTTATGTCT GALT6 RTF- RTR- GCAATTTGCGAGTACGGGGC CGCCGTCAAGTAATTCTC TCATCAG TATGC

UBQ10 RTF- RTR- TCGACCCTTCACTTGGTGT ATAAGCTGGTGTTGACAG GCA QRT-PCR GALT1 QPCRF- QPCRR- TCTTAGACATCGTCCTCTTAGA ACACAGCTGGAAATTT TGCC GALT2 QPCRF- QPCRR- TCTTAGACATCGTCCTCTTAGA ACACAGCTGGAAATTT TGCC GALT3 QPCRF- QPCRR- AGCGAAATTT GTGGTGAAG CCAATCTATTTGCTGC GAATAA GALT4 QPCRF- QPCRR- AAGCGATTAAGATTATTCAA CCAATGCTGCAACATG GATGG AGATGA

252

Table 4.10 continued GALT5 QPCRF- QPCRR ACATAAATTACGGCTGTTCAA TCTTAAGAGCTTATCC GATGGA CATAAGCATA GALT6 QPCRF- QPCRR- ACACAAATTAAGGATGTTCAAA TGCCTGTCAACACCA ATGG GCTTAT

4.5.4. Heterologous expression of GALT1, GALT3, GALT4 and GALT6 and assay for AGP galactosyltransferase activity

The coding regions of GALT1, GALT3, and GALT4 were obtained from the RIKEN

Bioresource center and GALT6 was obtained from The French Plant Genomic Resource

Center (http://cnrgv.toulouse.inra.fr/ ). N-terminal 6x-His tag fusion gene constructs were amplified using Q5 high fidelity DNA taq polymerase ((New England Biolabs,

Ipswich, MA, USA) and cloned into the pENTR/D-TOPO vector (Life technologies,

Grand Island, NY, USA) and eventually cloned into the destination vector pMDC32 gateway vector by LR clonase enzyme mix (Life technologies Grand Island, NY, USA).

The primers used for amplification is listed in Table 4.10.The constructs were then transformed into Agrobacterium strain GV3101 by freeze thaw method and the transformants were subsequently grown over night in Luria-Bertani (LB) medium. The bacterial cells were harvested by centrifugation and suspended in a buffer containing 10 mM MES, 10 mM MgCl2, and 50 µM acetosyringone (OD600 = 0.2). Leaves from 6- week-old wild-type N. tabacum cv. Petit Havana were used for Agrobacterium-mediated transient expression. Four days after infiltration, the leaves were harvested, and

253 microsomes were prepared according to the method described by Liang et al. (2010) with minor modifications.

4.5.5. Fluorescent protein fusion and subcellular localization

Full length GALT3, GALT4 and GALT6 devoid of stop codon was cloned into the pENTR/D-TOPO vector (Life technologies, Grand Island, NY, USA) and sequenced.

The resulting plasmids were cloned in the destination vector pEarlyGate 101 by gateway cloning strategy, using LR clonase enzyme mix (Life technologies Grand Island, NY,

USA) to generate the YFP N-terminal fusion constructs. These constructs were then transformed into Agrobacterium strain GV3101 and was infiltrated in tobacco leaves as described in the above section expect that the bacterial concentration was lower (OD600

= 0.05). The AtGALT3-YFP, AtGALT4-YFP and AtGALT6-YFP constructs were co- expressed with either the ER marker GFP-HDEL or the Golgi marker sialic acid transferase (ST)-GFP to ascribe subcellular localization. Transformed plants were incubated under normal growth conditions and imaged after 2 days post-infiltration using an upright Zeiss LSM 510 META laser scanning microscope (Jena, Germany), using a 40 X oil immersion lens and an argon laser. For imaging the expression of YFP constructs, the excitation line was 514 nm, and emission data were collected at 535–590 nm, whereas for GFP constructs, the excitation line was 458 nm and the emission data were collected at 505-530 nm.

254

4.5.6. Galactosyltranferase assay with microsomal preparations from transiently expressed GALT1, GALT3, GALT4, GALT6 in tobacco epidermal cells

The standard GALT reaction was performed as described in Basu et al. (2013) using detergent permealized microsomes from transiently expressed AtGALT1, AtGALT3,

AtGALT4, AtGALT6. Three reactions were included as controls, one with no substrate acceptor and one with permeabilized microsomal membranes from the WT tobacco leaves infiltrated with Agrobacteria GV3101 transformed with the empty expression vector (pMDC32) and tobacco leaves infiltrated with ST-GFP constructs as negative control.

4.5.7. Purification of Hyp-GALT5 reaction products by reverse-phase HPLC

The GALT reaction products were purified by RP-HPLC as described by Liang et al.

(2010).

4.5.8. Determination of substrate specificity of the GALT1, GALT3, GALT4 and

GALT6 enzyme activity

Microsomal fractions from tobacco leaves expressing GALT3 and GALT6 constructs were used for determination of substrate specificity as described by Basu et al. 2013.

4.5.9. Characterization of GALT2-GALT6 [AO]7:GALT activity

The effect of divalent ions on the AGP specific GALT activity of the candidate Hyp-

O-GALTs (GALT2-GALT6), was examined as described as Basu et al. (2015). Two controls were included, tobacco leaves infiltrated with empty vector (pMDC32) was considered as WT background and tobacco leaves expressing GALT1 was used as a negative control.

255

4.5.10. Isolation of Golgi-enriched plant microsomal membranes

Plant microsomal membranes were extracted from WT, galt1-1, galt1-2, galt3-1, galt3-2, galt4-1, galt4-2, galt6-1 and galt6-2 according to Basu et al. (2013) with minor modifications.

4.5.11. Extraction of AGPs

AGPs were extracted from the WT, galt1-1, galt1-2, galt3-1, galt3-2, galt4-1, galt4-

2, galt6-1and galt6-2 mutant plants as described in Schultz et al. (2000) and quantified as described by Gao et al. (1999). AGP profiling was conducted as described by Youl et al. (1998) with modifications. AGPs were obtained from eight grams of plant material, precipitated by β -Yariv reagent and dissolved in 1 mL of deionized water before applying 100 µl onto a polymeric reverse-phase column (PRP-1, 5 µm, 4.1 × 150 mm;

Hamilton) equilibrated with buffer A (0.1% trifluoroacetic acid). Fifty µg of [AO]7 was used as a control to monitor the retention time of a pure AGP peptide. Samples were eluted from the column following a linear gradient with solvent B (0.1% trifluoroacetic acid in 80% acetonitrile): 0 to 30% solvent B in 30 min, then 30 to 100% in 30 min at a flow rate of 0.5 mL/min. Chromatography was monitored by absorption at 215 and 280 nm.

AGPs from silique, flower, inflorescence stem, senescent leaves were extracted as described by Lamport, (2013) with minor modifications. To extract AGPs from siliques

(2.5g), flowers (1.5g), inflorescence stem (5g), senescent leaves (5g), were ground to a fine powder in liquid nitrogen. Ground tissue was added to extraction buffer of CaCl2

(2% w/v) at a volume of 2 ml for each gram of tissue, and stirred for 3h at room

256 temperature. Samples were centrifuged for 30 min at 10,000g at room temperature. The supernatant was freeze dried overnight and resuspended in 1 ml of 2% CaCl2 and transferred to 2 ml microcentrifuge tubes. AGPs were precipitated with equal volume of the β -Yariv reagent (2 mg/ml in 2% w/v CaCl2) and leaving overnight at 4°C. The insoluble Yariv–AGP complex was collected by centrifugation at 10,000g in a microcentrifuge for 1hr. The β-Yariv was removed by washing the pellet three times in

2% (w/v) CaCl2 and then twice in methanol. The pellet was dried, dissolved in 100 µl of water mixed with 25 ng of solid sodium dithionite and incubated for around 30 min at

~50 °C until the mixture decolourised. The resulting solution was then desalted on a PD-

10 column (Pharmacia) that had been equilibrated with water, and the eluate was freeze- dried.

4.5.12. Evaluation of seed set

Mature siliques from 6-weeks old WT, galt1, galt3, galt4 and galt6 plants were collected and silique length and weight were measured. For seed number, siliques were decolorized by incubation in 100% ethanol at 37°C overnight before dissection of the siliques. For reciprocal cross pollination, 10 flowers from WT, galt4-1 and galt4-2 were selected at stage 12. These flowers and were emasculated before pollinating them with fresh pollen obtained from flower stage 13. After 10 d siliques collected from these flowers to examined seed set.

4.5.13. Root growth measurements

Root growth in response to Yariv reagent was monitored as described in Basu et al.

(2015).

257

4.5.14. In Vitro pollen germination assay

Flowers collected from WT, galt1-1, galt1-2, galt3-1, galt3-2, galt4-1, galt4-2, galt6-1 and galt6-2 plants 1 to 2 weeks after bolting were used for the examination of pollen tube phenotypes. The pollen tube elongation was analysed as desceibed in Basu et al. (2015).

4.5.15. Aberrant root hair morphology

Root hair length from 8-d-old plants grown on agar plates was determined on low- magnification (×10) digital images captured using a CCD camera and image analysis freeware (ImageJ; http://rsb.info.nih.gov/ij/). To ensure comparable results, the area 3 to

5 mm behind the root tip was analyzed. Plants grown on agar plates were carefully removed in ∼100 µL of half-strength MS medium on microscope slides for analysis.

Quantification data are the means of 50 to 75 values representing 15 root hairs each of

20 to 35 individual plants measured.

4.5.16. Cytochemical staining of seeds and determination of adherent mucilage size and mass

Seeds of all the indicated genotypes were prehydrated in water and stained either with 0.01% ruthenium red. The staining was performed as described by Willats et al.

(2001) and Harpaz-Saad et al. (2011). Imaging was done using a Zeiss LSM 510 confocal microscope. The volume of adherent mucilage was measured using method described by Yu et al. (2014).

258

4.5.17. Measurement of chlorophyll content

Chlorophyll was extracted from WT, galt6-1 and galt6-2 leaves by immersion in 1 mL of N, N-dimethylformamide for 48 h in the dark at 4 oC. Absorbance was recorded at

664 and 647 nm, and total chlorophyll concentration was calculated (Xiao et al., 2004).

The total chlorophyll content was measured and normalized per gram fresh weight of sample.

259

CHAPTER 5: GLYCOSYLATED SALT-OVERLY SENSITIVE5 MEDIATES

ROOT GROWTH AND SEED COAT MUCILAGE ADHERENCE VIA FEI/FEI2

KINASE PATHWAY

This work will be submitted to the journal of Plos One for review.

Debarati Basu, Tayler Debrosse, Emily Poirier, Kirk Emch, Wuda Wang, Lu Tian Hayley Herock, Andrew Travers and Allan M. Showalter 5.1. Abstract

The fundamental processes that underpin plant growth and development depend crucially on the action and assembly of the cell wall, a dynamic structure that changes in response to both developmental and environmental cues. While much is known about cell wall structure and biosynthesis, much less is known about the functions of the individual wall components, particularly with respect to their potential roles in cellular signaling. Loss-of-function mutants of two arabinogalactan-protein (AGP)-specific galactosyltransferases namely, GALT2 and GALT5, confer pleiotropic growth and development phenotypes indicating the important contributions of carbohydrate moieties towards AGP function. Notably, galt2galt5 double mutants displayed impaired root growth and root tip swelling in response to salt, likely as a result of decreased cellulose synthesis. These mutants phenocopy a salt-overly sensitive mutant called sos5, which lacks a fasciclin-like AGP (SOS5/FLA4) as well as a fei1fei2 double mutant, which lacks two cell wall-associated leucine-rich repeat receptor-like kinases. Additionally, galt2gal5 as well as sos5 and fei2 showed reduced seed mucilage adherence. Quintuple galt2galt5sos5fei1fei2 mutants were produced and provided evidence that these genes

260 act in a single, linear genetic pathway. Further genetic and biochemical analysis of the quintuple mutant demonstrated involvement of these genes with the interplay between cellulose biosynthesis and two plant growth regulators, ethylene and ABA, in modulating cell wall integrity.ethylene and ABA, in modulating cell wall integrity

5.2 Introduction

Growing plant cells control the biogenesis, deposition, and remodeling of the cell wall. The plant cell wall is a dynamic and complex structure composed of and proteins that play crucial roles in all aspects of plant life. To avoid loss of integrity, which may lead to growth cessation, continuous cellular surveillance is required to sense cell wall perturbations and coordinate cell wall performance with the internal growth machinery (Boisson-Dernier et al., 2013). Mounting evidence indicates that a dedicated plant cell wall integrity (CWI) maintenance mechanism exists in plants (Wolf et al.,

2012). This mechanism monitors and maintains the functional integrity of the cell wall during different biological processes, including exposure to abiotic and biotic stress

(Hamann and Denness 2011; Hamann, 2015). While our understanding of the mechanisms involved in biosynthesis of cell wall polymers have increased significantly, current understanding of the components and mechanisms involved in the perception and regulation of processes maintaining CWI is limited.

Plant cell walls are composite structures composed of cellulose and matrix polysaccharides such as hemicelluloses and pectins (Somerville et al., 2004). Besides the polysaccharides, cell wall proteins represent another important component of plant cell walls and are involved in wall structure, support, signaling, and interactions with other

261 wall components and with the plasma membrane (Jamet et al., 2006). Arabinogalactan- proteins (AGPs) are one such cell wall glycoprotein family that are implicated to function in various aspects of plant growth and development, including root growth and development, somatic embryogenesis, hormone responses, xylem differentiation, pollen tube growth and guidance, programmed cell death, cell expansion, salt tolerance, and cellular signaling (Showalter, 1993; Showalter, 2001; Seifert and Roberts, 2007; Ellis et al., 2010; Tan et al., 2012). Evidence for these functions is circumstantial, indirect, or based on genetic mutant analysis. While these functional implications speak to the importance of AGPs in plants, there are virtually no studies that delve into the biochemical mechanism of action for these proposed AGP functions.

Fasciclin-like arabinogalactan-proteins (FLAs) are a subfamily of AGPs characterized by the presence of one or two fasciclin (Fas 1) domains, which are cell adhesion domains (Johnson et al., 2003). They are frequently predicted to have a glycosylphosphatidylinositol (GPI) anchor, which would allow for its localization to the plasma membrane, making them ideal candidates for signal perception and transduction.

They are implicated in cell wall biosynthesis, cell wall remodeling, and signaling

(Seifert and Roberts, 2007). The FLA11 and FLA12 genes are preferentially expressed in secondary cell wall forming cells (Ito et al., 2005; Persson et al., 2005), similar to their orthologues from other plant species (MacMillan et al., 2010; Liu et al., 2013).

Moreover, fla11fla12 double mutants display a substantial reduction in cellulose content accompanied by reduced tensile strength and tensile modulus of elasticity. Another well characterized fasciclin-like AGP is SOS5 (salt overly sensitive), also known as FLA4,

262 reported to be involved in root growth under elevated salt or sucrose containing media

(Shi et al., 2003). Xu et al. (2008) demonstrated that SOS5 acts upstream of cell wall deposition in a non-additive genetic pathway by interacting with FEI1/FEI2, two cell wall leucine rich repeat-like receptor kinases (RLKs), based on the identical phenotypes displayed by single, double and triple mutants, namely, sos5, fei1fei2, and sos5fei1fei2 loss-of-function mutants (Xu et al., 2008). Specifically, these mutants demonstrated reduced cell elongation accompanied by impaired anisotropic growth (i.e., root tip swelling) when grown under the restrictive conditions of elevated salt or sucrose.

Furthermore, the swollen root tip phenotype was reversed in all these mutants by blocking ethylene biosynthesis, but not ethylene perception (Xu et al., 2008). The recent characterization of two genetic mutants namely, galt2 and galt5, which encode two galactosyltransferase (GALT) enzymes responsible for adding the first Gal onto AGP protein backbones also displayed identical phenotypes as observed in sos5 and fei1 fei2 mutants (Basu et al., 2015). Since these galt mutants have reduced glycosylation of

AGPs, it is hypothesized that glycosylation (i.e., sugar addition) of the AGP domain of

SOS5 is required for it to function in the SOS5/FEI1-FEI2 pathway to signal normal root growth under non-permissive conditions.

Harpaz-Saad et al. (2011) also demonstrated that SOS5 and FEI2 play a novel and non-conditional role in the formation of the transverse cellulosic rays deposited across the inner layer of seed coat mucilage. The malformation of these cellulosic rays in these mutants coincides with an increase in the solubility of the pectinaceous component of seed coat mucilage. Disruption of GALT2 and GALT5 in galt2galt5 double mutants

263 resulted in an identical seed coat mucilage phenotype (Basu et al., 2015). Based on the above information, we hypothesize that glycosylation of the AGP domain of SOS5, which is catalyzed by GALT2 and GALT5, is essential for the cellular signaling of normal root growth, cellulosic ray formation, and extrusion of seed coat mucilage in

Arabidopsis. Here, we address this hypothesis by generating quintuple mutant plants

(galt2galt5sos5fei1fei2) and conducting their phenotypic functional characteri characterization with respect to other related mutants, such as galt2galt5 double mutants and sos5fei1fei2 triple mutants.

5.3. Results

5.3.1. Expression patterns of GALT2, GALT5, SOS5 and FEI1, FEI2

RNA transcript abundance profiles of GALT2, GALT5, FEI1, FEI2 and SOS5 were examined in various vegetative and floral tissues using Arabidopsis PlaNet, a publicly available dataset (Figure 5.1). SOS5 was not included in this analysis due to its absence from the ATH1 Affymetrix chip Analysis revealed ubiquitous but distinct expression patterns for the four genes throughout plant development (Mutwil et al., 2011). Since

SOS5 is not present on the Affymetrix ATH1 array, the relative expression levels was assessed using qRT-PCR. In addition, transcriptome analyses using RNA extracted from laser-capture dissected seed coat tissue indicated that the transcript levels are high in the seed coat during early embryogenesis (Figure 5.2; Le et al., 2010; http://seedgenenetwork.net/arabidopsis; Dean et al., 2011; http://bar.utoronto.ca/efp_seedcoat/cgi-bin/efpWeb.cgi). Higher transcript abundance of

GALT2, GALT5 and SOS5 was displayed in primary roots and in siliques during early

264 stages of seed development (Figure 5.3A). Further qRT-PCR was performed to assess the expression pattern of these five genes during the course of seed development

(Figure 5.3B). The expression of GALT2, GALT5 peaks at approximately 4-7 DPA

(days post anthesis) which coincides with the timing of seed coat mucilage production.

Even though, FEI1 has a higher abundance than FEI2, FEI1 did not exhibit aberrant seed coat mucilage phenotype (Harpaz-Saad et al., 2011). Furthermore, co-expression analysis using the GeneCAT (http://genecat.mpg.de) tool indicated that the expression of

FEI1 and FEI2 are tightly correlated with the expression of AGP-specific GTs as well as numerous candidate AGPs (Mutwil et al., 2008; Table 5.1). Collectively, these data indicate the concerted involvement of these five genes in both root tip development and formation of seed coat mucilage. To assess their roles in salt stress, the transcript abundance of these five genes in root tips were determined by growing wild type plants in MS supplemented with 100 mM of NaCl. All five genes were induced at least 2.5 fold in response to this salt treatment (Figure 5.3C).

265

Table 5.1 List of candidate genes coexpressed with FEI1 as query genes using the Gene CAT coexpression tool Gene Name Predicted function Pearson’s correlation coefficient At1g31420 (FEI1) At5g09870 Cellulose synthase (CESA5) 0.70353 At4g32410 Cellulose synthase (CESA1) 0.60589 At5g05170 Cellulose synthase (CESA3) 0.56805 At5g64740 Cellulose synthase (CESA6) 0.52679 At4g39350 Cellulose synthase (CESA2) 0.44579 At2g35620 Cell wall receptor like kinase (FEI2) 0.47394 At4g26940 Galactosyltransferase GT 31 family 0.34069 At3g06440 Galactosyltransferase GT 31 family 0.35562 (GALT3) At1g53290 Galactosyltransferase GT 31 family 0.19700 At1g74800 Galactosyltransferase GT 31 family 0.16268 (GALT5) At1g05170 Galactosyltransferase GT 31 family 0.14670 At2g32430 Galactosyltransferase GT 31 family 0.14442 At5g53340 Galactosyltransferase GT 31 family 0.07022 (HPGT1)

At1g11730 Galactosyltransferase GT 31 family 0.00675 At4g32120 Galactosyltransferase GT 31 family 0.05157 (HPGT2) At5g55730 Fasciclin-like arabinogalactan protein 0.34251 (FLA1)

266

Table 5.1 continued At4g12730 Fasciclin-like arabinogalactan-protein 0.54985 (FLA2) At2g45470 Fasciclin-like arabinogalactan-protein 0.54608 (FLA8) At5g44130 Fasciclin-like arabinogalactan-protein 0.48029 (FLA 13) At1g03870 Fasciclin-like arabinogalactan-protein 0.45305 (FLA9) At2g04780 Fasciclin-like arabinogalactan-protein 0.36382 (FLA7) At5g60490 Fasciclin-like arabinogalactan-protein 0.19956 (FLA12) At2g23130 Arabinogalactan-protein (AGP17) 0.37150 At2g46330 Arabinogalactan-protein (AGP16) 0.19776 At1g02730 ATCSLD5, CELLULOSE SYNTHASE 0.13064 LIKE D5 At5g65390 Arabinogalactan-protein (AGP7) 0.14295 At5g10430 Arabinogalactan-protein (AGP4) 0.11422 At2g15390 Fucosyltransferase (FUT4) 0.01729

267

Figure 5.1 Transcript profiling of GALT2, GALT5, FEI1, and FEI2 throughout the different developmental stages in Arabidopsis as depicted by PlaNet. Y-axis: expression values, x-axis: tissues/treatments, red dots: average expression, green dots: expression from individual microarray/RNAseq experiments. The horizontal red lines indicate the expression patterns during root and seed development.

268

Figure 5.2 Expression analysis of indicated genes during the course of seed development. Analysis includes CESA gene family, COBL2, GALT2, GALT5, FEI1 and FEI2 through the course of seed development as depicted by the Bio-Array Resource eFP browser (Winter et al., 2007) based on the data set generated by Le et al. (2010) following gene expression in laser-capture micro-dissected seeds.

269

Figure 5.3 Transcript profiling of FEI1, FEI2, SOS5, GALT2 and GALT5 genes from different organs and developmental stages and in response to salinity stress. (A) qRT-PCR analysis of the five genes in different plant organs/tissues. Expression levels are the mean ± SE of three technical replicates relative to the Ubiquitin 10 (UBQ10) reference gene. The expression value of GALT2 in leaf was considered as 1. (B) Expression analysis of the five genes during the course of seed development. Total RNA was isolated from seeds 4, 7, and 11 DPA. Relative expression was normalized using GAPC as a reference gene. The expression value of GALT2 at 11 DPA was

270 considered as 1. (C) Induction of transcript abundance of GALT2, GALT5, FEI1, FEI2, and SOS5 in response to salt treatment. Total RNA was extracted from root tips of 250 WT and mutant seedlings grown in MS media and in MS media supplemented with 100 mM NaCl for 7d. Experiments in (B) and (C) are the mean ± SE of three technical replicates.

5.3.2. Quintuple mutants display cell expansion defects

It is important to mention that Shi et al. (2003) characterized SOS5 using a point mutant having a substitution from Ser to Phe at the junctional region between the second fasciclin-like domain and the second AGP-like domain. But it is unclear whether the point mutation affects the fasciclin domain, or AGP domain. The point mutation can have three possible effects, it can reduce the transcription of the gene or if the mutation affects the AGP-like domain then the glycosyaltion of SOS5 would be affected or if the mutation affects the fasciclin-like domain then it might have an effect that is yet to be characterized. Here in this manuscript, a T-DNA insertional sos5 mutant was used that renders it a functional knock out. The roots of sos5, fei1fei2 and sos5fei1fei2 behave similarly to that of galt2galt5 in their response to 100mM NaCl (Basu et al., 2015).

To test for genetic interaction among GALT2, GALT5, FEI1, FEI2, and SOS5 a quintuple mutant was generated by crossing sos5fei1fei2 triple mutants with galt2galt5 double mutants. The qRT-PCR analysis confirmed that the quintuple mutant was a functional knock-out of all five genes (Figure 5.4). The quintuple mutant displayed decrease root elongation and root tip swelling compared to the wild type plants and phenocopied both parental lines, galt2galt5 and sos5fei1fei2, in response to elevated salt and sucrose (Figure 5.5 and 5.6). The increased diameter and reduced elongation observed in quintuple mutant roots indicates that anisotropic expansion is defective in

271 the quintuple mutant root cells similar to their parental lines. The quintuple mutants, like their parental lines, also displayed swollen roots on medium that contains an elevated concentration of sucrose (Figure 5.6 B and 5.6D).

272

B

Figure 5.4 RT-PCR analysis of the quintuple mutants to confirm null status. (A) Gene structures and T-DNA insertion sites for SOS5, FEI1, and FEI2. (B) RT-PCR analysis of galt2galt5sos5fei1fei2 quintuple mutants to confirm null status. Total RNA was extracted from rosette leaves of 2-week-old WT and homozygous mutant plants of the indicated genotypes. UBQ10 was used as the loading control. Experiments were repeated at least twice with virtually identical results. Arrows indicate the locations of primers used for RT-PCR.

273

Figure 5.5 Quintuple mutants exhibits reduced root elongation in response to elevated NaCl. (A) WT and mutant seedlings were grown on MS medium containing 1% sucrose for 5 d and then transferred to MS medium containing 100 mM NaCl. (B) WT and mutant seedlings were grown on MS medium containing 0% sucrose for 5 d and then transferred to MS medium containing 4.5% sucrose. Quantification of root elongation after transfer to non-permissive conditions was recorded after 7, 14 and 21 d. Data are means ± SE; n ≥ 25. Asterisks above the bars indicate significant differences relative to the WT as determined by on one way ANOVA and Tukey's post-hoc comparisons (*P < 0.05; **P < 0.01).

274

Figure 5.6 Quintuple mutant displays conditional root anisotropic growth defect. (A) Root tips of the indicated seedlings four days after transfer from MS medium containing 0% sucrose to MS medium containing 100 mM NaCl as imaged by dark field microscopy. (B) Root tips of the indicated seedlings four days after transfer from MS medium containing 0% sucrose to MS medium containing 4.5 % sucrose as imaged by dark field microscopy. (C) Graphical representation of root tip width in response to 100 mM NaCl. (D) Graphical representation of root tip width in response to 4.5% sucrose. The stripped bars indicate seedlings grown on control unsupplemented MS plates. Quantitation of root tip width was measured at the level of the youngest root hair using ImageJ software. Values are the means (n>15) ± SE. Asterisks above the bars indicate significant differences relative to the WT as determined by on one way ANOVA and Tukey's post-hoc comparisons (*P < 0.05). Scale bar= 1mm.

275

Like roots, hypocotyls of the etiolated seedlings are composed of cells that primarily undergo longitudinal expansion. The hypocotyls of etiolated quintuple mutant seedlings exhibited significantly (27-45%) wider hypocotyls (Figure 5.7A and 5.7C). However, unlike the root growth, the hypocotyl phenotype was not accompanied by a decrease in the overall length of the hypocotyl except in the case of galt2galt5 (Figure 5.7B).

As an additional test of the sensitivity of the mutants towards elevated NaCl, the root bending assay was used (Zhu et al., 2010; Baek et al., 2011; Golldack et al., 2014). This assay is based on the ability of roots to sense and reorientation themselves relative to gravity by means of a differential growth response achieved by cell expansion occurring in the elongation zone (Swarup et al., 2005). This assay also revealed that the quintuple mutants, like their parental lines, were salt hypersensitive as indicated by the increased root curvature displayed by the mutants due to their delayed reorientation of root growth

(Figure 5.8).

276 A

Figure 5.7 Hypocotyl phenotype of the quintuple mutants. (A) Quantification of hypocotyl widths from WT and mutant seedlings of the indicated genotypes were grown for four days in the dark on MS medium with 0% sucrose. Values represent the mean (n = 10) ± SE. (B) Measurement of hypocotyl lengths of seedlings grown as indicated above. Asterisks above the bars indicate significant differences relative to the WT as determined by on one way ANOVA and Tukey's post-hoc comparisons (*P < 0.05; **P < 0.01). (C) Representative images of hypocotyls from WT and mutants of the indicated genotypes. Scale bar = 1mm.

277

Figure 5.8 Salt hypersensitivity assessed by root bending assay. (A) Five-day-old seedlings grown on MS plates were transferred to MS plates with 100 mM NaCl and reoriented at an angle of 180° (upside down). The images were taken 5 d after seedling transfer. (B) Analysis of root curvature of the indicated seedlings were measured using ImageJ softeware. Values represent the mean (n = 5) ± SE. Asterisks indicate significant differences relative to the wild type (*P<0.05) as determined by Tukey’s HSD method.

278

5.3.3. The quintuple mutant is defective in cellulose biosynthesis

There have been reports linking cell expansion with impaired cell wall structure

(Schindelman et al., 2001; Wiedemeier et al., 2002). The altered pattern of cell expansion in the quintuple mutant may arises from a cell wall defect as fei1fei2 and sos5 mutants displayed impaired cell expansion (Xu et al., 2008; Shi et al., 2003). To assess the cell wall properties, the effect of isoxaben, an inhibitor of cellulose synthase, on fei1fei2, galt2galt5, sos5, and the quintuple mutant was examined. The quintuple mutants displayed increased sensitivity to isoxaben after 72h compared to the wild type seedlings, similar to that of sos5fei1fei2, sos5 and galt2galt5 seedlings (Figure 5.9).

This suggests, like fei1fei2, sos5 and galt2galt5, the quintuple mutant perturbs either the biosynthesis or function of cellulose. In addition, roots of sos5fei1fei2, galt2galt5 and quintuple mutant seedlings grown in non-permissive conditions produced ectopic lignin

(Figure 5.10A and 5.10B) under both elevated salt and sucrose. These findings are consistent with impaired cellulose biosynthesis as presence of ectopic lignin deposition is generally correlated with a decreased level of crystalline cellulose (Humphrey et al.,

2007).

279

Figure 5.9 Mutants display hypersensitivity towards isoxaben. Representative root tips were imaged using dark field microscopy. Root tips of indicated seedlings were germinated and grown for five days on MS medium with 1% sucrose and then transferred for 72 hours to MS medium supplemented with either 1 nM, 2 nM, 5 nM and 10 nM isoxaben. The width of the roots was measured at the level of the youngest root hair using ImageJ software. Values are the means (n>15) ± SE. Bar = 1 mm.

280

Figure 5.10 Phloroglucinol staining to detect lignin. (A) Seedlings of indicated genotypes were grown on MS medium containing 0% sucrose for 5 d and then transferred to MS medium containing 2 nM isoxaben + 100 mM NaCl or (B) 2 nM isoxaben 4.5% sucrose for 3d (C) 100 mM NaCl (D) 4.5% sucrose for 7 d.. Scale bar = 1 mm.

281

Cellulose synthesis was further analyzed by measuring the incorporation of [14C] Glc into crystalline and non-crystalline cellulosic cell wall fractions obtained from excised root tips of wild type and quintuple mutant seedlings grown under non-permissive conditions. The quintuple mutant along with fei1fei2, sos5, sos5fei1fei2 and galt2galt5 roots displayed a striking reduction in cellulose biosynthesis, in response to NaCl (i.e.,

15-27%) and sucrose treatment (i.e., 11-23%) as measured by the incorporation of radiolabeled [14C] UDP Glc into both acid-insoluble material (crystalline cellulose; Peng et al., 2000). The acid-soluble material (non-crystalline cellulose and other wall polymers) also exhibited a reduction in response to NaCl (i.e., 18-27%) and sucrose (i.e.,

13-28%) treatment (Heim et al., 1998) (Figure 5.11). Similar results were reported for fei1fei2 roots (Xu et al., 2008).

282

Figure 5.11 The quintuple mutant displays cellulose deficiency under elevated salt stress. Incorporation of [14C]Glc into acid-soluble and acid-insoluble fractions from excised root tips from WT and mutant seedlings in response to 100 mM NaCl, (C) and (D) in response to 4.5% sucrose and (E) and (F) in response to AIB. The seedlings were grown on 0% sucrose for 5 d and then transferred, different treatment conditions for 7d. (A), (C), and (E) indicate incorporation of [14C]Glc into acid soluble fraction and (B), (D), and (F) indicate incorporation of [14C]Glc into acid soluble fraction. Values are means ±

283

SE from two biological replicates, and the experiments were repeated at least two times with similar results. Asterisks indicate statistical differences between WT and mutants based on one way ANOVA and Tukey's post-hoc comparisons (*P < 0.05; **P < 0.01).

5.3.4. Role of ACC in AGP-FEI1/FEI2 mediated cell expansion

To investigate possible mediation of ethylene in elevated salt or sucrose induced altered cell elongation quintuple mutants and its parental lines are subjected to elevated salt and sucrose together with ethylene biosynthetic and signaling inhibitors. Further, previous reports have demonstrated that inhibition of ethylene biosynthesis or perception can partially revert the swollen phenotypes of certain root morphology mutants, such as sabre (Aeschbacher et al., 1995) and cev1 (Ellis et al., 2002), thereby indicating its crucial role in cell expansion. This information provided the impetus to determine the effect of blocking ethylene signaling or biosynthesis on swollen root tip phenotype.

Towards this goal two ethylene biosynthesis inhibitor cobalt chloride (CoCl2), amino isobutyric (AIB), and aminoethoxyvinyl glycine (AVG) was used. Both CoCl2 and AIB block ACC oxidase, but they employ a different mechanism. The Co2+ ions inhibit ACC oxidase required for the conversion of ACC to ethylene, thus blocking ethylene synthesis, whereas AIB is a structural analog of ACC that blocks ACC oxidase activity by acting as a competitive inhibitor. In contrast, AVG is an inhibitor of pyridoxal phosphate required for ACS activity (Yang and Hoffman, 1984). As AVG and CoCl2 both block ethylene biosynthesis by distinctly different mechanisms, it is unlikely that this phenotypic reversion of sos5fei1fei2, sos5, galt2galt5 and quintuple mutants is due to off-target effects. Surprisingly, silver ions (silver nitrate), which block ethylene perception, had no appreciable effect on the root phenotype of quintuple mutants

284

(Figure 5.12). Taken together, these results demonstrated that a signaling pathway involving ethylene, participates in the reduction of root cell elongation when mutants were subjected to elevated levels of salt and sucrose. In addition these phenotypes in the quintuple mutant were nearly identical to that of sos5fei1fei2, sos5, and galt2galt5 mutants indicating that these genes are likely in the same genetic pathway.

Seifert et al. (2014) reported that SOS5 acts synergistically with abscisic acid

(ABA) signaling to control root growth. Consequently, the role of ABA on mutant phenotype was assessed, which revealed that ABA suppresses the swollen root tip phenotype in all mutants in presence of 5 µM ABA supplemented media containing 100 mM NaCl (Figure 5.12).

285

Figure 5.12 Role of ethylene and ABA on conditional root phenotype. Seedlings grown on MS medium containing 0% sucrose for 4 d and transferred to MS medium containing 100 mM NaCl supplemented with AgNO3, AVG, CoCl2, and AIB as indicated. Scale bar = 1 mm.

286

5.3.5. Altered seed coat pectin mucilage organization in fei2, sos5, galt2galt5 and quintuple mutant

Disruption of FEI2, SOS5 or GALT2 and GALT5 resulted in defective seed coat pectin mucilage organization (Harpaz-Saad et al., 2011; Basu et al., 2015). This prompted an examination and comparison of the mucilage in the quintuple mutant seeds with previously characterized sos5, sos5fei1fei2 and galt2galt5 mutants using the cationic dye ruthenium red, which stains for acidic pectins. Examination of seeds after hydration via ruthenium red staining without shaking displayed substantially expanded outer mucilage layer in the quintuple mutants as well as the parental line scompared to wild type seeds (Figure 5.13). Furthermore, quantitative mucilage determination revealed that the galt2galt5, cesa5, cesa5sos5 and quintuple mutant displayed an increased mucilage content compared to wild type (Figure 5.14). In contrast, upon removal of the non-adherent outer layer by mild shaking, substantially thinner inner adherent mucilage was observed in the mutants (Figure 5.15). In order to investigate the effect of cation chelators on the quintuple mutant seed coat mucilage phenotype, ruthenium red staining was conducted following treatment with the divalent cation chelator EDTA. The removal of Ca2+ ions may lead to disruption of the ionic cross- linking of galactorunic acid residues in pectin, allowing for separation of pectin under agitation which will be reflected by rapid loss of ruthenium red staining of seeds

(Western et al., 2001). Reduced thickness of the inner mucilage layer relative to the wild type, as in water, was observed which is identical to that of fei2, sos5 and galt2galt5 seed mucilage phenotype indicating an increase in soluble mucilage (Figure 5.14).

287

In order to confirm and quantify the changes in soluble versus adherent mucilage, the outer (soluble) and adherent mucilage from WT and mutant seeds were analyzed by sequential extraction of seeds with ammonium oxalate, 0.2 N NaOH, and 2 N NaOH

(Table 5.2). The fei1fei2, sos5, galt2galt5, sos5fei1fei2, and quintuple mutants had a significant increase in total sugars in the ammonium oxalate extract (18-35%) and the

0.2 N NaOH extract (24-55%) (i.e., soluble and weakly attached pectins) as compared with WT seeds, which is consistent with the increase in soluble mucilage observed by ruthenium red staining. In contrast, quintuple mutants and their parental lines displayed a decrease in total sugars in 2 N NaOH extracts, which represent the majority of adherent mucilage (i.e., strongly linked pectins and cross-linking glycans/hemicelluloses) (Arsovski et al., 2009; Huang et al., 2011).

288

Figure 5.13 Staining of seeds with ruthenium red with no shaking, demonstrating both adherent and nonadherent layers of seed mucilage. (A) Seeds were placed in ruthenium red solution and photographed after 2 min. (B) seeds shaken in water and then stained with ruthenium red. Bar in figure (A) is 100 µm and in figure (B) is 50 µm.

289

Figure 5.14 Aberrant adherent mucilage structure as depicted by ruthenium red staining. Ten milligrams of seeds of the wild type (WT) and indicated mutants was hydrated in water and occasionally shaken prior to the ruthenium red staining. Bar= 0.2 mm.

290

Figure 5.15 Effect of cationic chelator on seed mucilage phenotype. Ruthenium red staining for pectins following pre-treatment with EDTA and gentle shaking. Bar= 0.2 mm.

291

Table 5.2 Quantification of total sugars from WT, galt2-1, galt5-1, galt2galt5, sos5, sos5fei1fei2, and quintuple mucilage sequentially extracted using ammonium oxalate, 0.2 N NaOH, and 2 N NaOH Extracts Genotype Ammonium 0.2 N NaOH 2 N NaOH oxalate WT 0.80+0.02 1.14+0.38 0.81+0.05 galt2-1 0.74+0.09 1.14+0.27 0.46+0.03 galt5-1 0.66+0.07 1.26+0.09 0.47+0.05 galt2galt5 0.95+0.02a 1.42+0.05a 0.41+0.03 sos5 0.92+0.09a 1.59+0.21a 0.39+0.04 sos5fei1fei2 1.03+0.11a 1.63+0.25b 0.39+0.03 quintuple 1.11+0.17a 1.73+0.24b 0.38+0.01 Intact seeds were extracted sequentially with 0.2% ammonium oxalate, 0.2 N NaOH and 2 N NaOH, neutralized, and assayed by the phenol-sulfuric acid method against glucose standards. The results are shown as µg/mg of seeds. Analyses were performed in triplicate and results are given as µg/mg seed + SE. All genotypes were grown, harvested and stored together. Letters ‘a’ and ‘b’ denote a significantly difference from the wild type (P <0.05; P<0.01).

5.3.6. Quintuple mutants display reduced cellulosic rays in the adherent layer of seed mucilage

In order to investigate cellulose deposition in seed mucilage, and in particular formation of the cellulosic rays, seed coat cellulosic structures were visualized by staining Pontamine Fast Scarlet S4B, a cellulose-specific dye and Calcofluor white, a fluorescent probe for β -glycans. The rays in the quintuple mutant seeds appeared substantially reduced and malformed compared with wild type seeds, but similar to its parental lines, sos5fei1fei2 and galt2galt5 (Figure 5.16). To more specifically investigate cellulose biosynthesis in quintuple mutant seeds, the crystalline cellulose

292 content of whole seeds was quantified (Figure 5.17). Consistent with the pattern observed with Pontamine staining, cellulose content in the seeds of the galt2galt5 (12%) sos5fei1fei2 (19%) and quintuple mutants (23%) were significantly reduced (12~24%) compared to the wild type. Such a decrease in crystalline cellulose content can be attributed to a reduction of crystalline cellulose deposition in the rays. Furthermore, the quintuple mutant phenotype was compared with the known seed specific CESAs, namely CESA2, CESA5 and CESA9 that are involved in secondary cell wall synthesis

(Mendu et al., 2011). The decrease in crystalline cellulose from seeds is less than that of the cesa2cesa5cesa9 triple mutants suggesting that the quintuple mutants can only affect the cellulosic rays as opposed to the secondary cell wall and cellulosic rays in case of the three CESAs (Sullivan et al., 2011).

293

Figure 5.16 Reduced cellulosic rays observed in mutant seed coat mucilage. Pontamine scarlet red and calcofluor staining of cellulosic rays of WT and mutant seeds was performed following gentle shaking in water. Error bars indicate SE of three replicates. Based on one way ANOVA and Tukey's post-hoc comparisons (*P < 0.05; **P < 0.01). Scale bar = 50 µm.

294

Figure 5.17 Impaired cellulosic rays in the mutants. Crystalline cellulose content determined from WT and mutant seeds. Similar results were obtained in two biological repeats. Asterisks indicate statistical differences between mutants and WT based on one way ANOVA and Tukey's post-hoc comparisons (*P < 0.05; **P < 0.01; ***P< 0.001).

A number of mutants with reduced seed mucilage like .myb61 (disruption of which causes impaired deposition and extrusion of mucilage) and atsbt1.7 (subtilisin-like serine proteases that do not release mucilage upon hydration) display defects in germination in conditions of reduced water potential (Penfield et al., 2001; Rautengarten et al., 2008). Consequently, the effect of the altered mucilage in the quintuple mutant on seed germination was examined. This was done in the presence of increasing concentrations of polyethylene glycol (PEG) in order to vary the water potential of the

295 media. No significant differences were observed in germination frequency or in seedling establishment in quintuple mutants seeds compared to wild type at any concentration of

PEG, indicating that the defects in seed mucilage observed in these mutants did not affect germination in dry conditions (Figure 5.18).

Figure 5.18 Seedling establishment percentage of the mutants belonging to the indicated genotypes on media containing increasing concentration of polyethylene glycol (PEG). Germination percentages were determined from two independent experiments, with more than 200 seeds per line for each experiment. Values are means ± SE.

296

5.4. Discussion

5.4.1. GALT2, GALT5, SOS5, FEI1, FEI2 act in a single genetic signaling pathway

Three separate studies in Arabidopsis have reported that loss-of-function mutants of

SOS5, a GPI anchored fasciclin-like AGP, two cell wall receptor kinases, FEI1 and

FEI2, and two AGP specific galactosyltransferase, GALT2 and GALT5, share similar phenotypes (Xu et al., 2008; Harpaz-Saad et al., 2011; Basu et al., 2015). Moreover, sos5 fei1fei2 triple mutants were shown to act in a single, non-additive genetic pathway

(Xu et al., 2008). This information led to the hypothesis that these five genes act in the same linear genetic pathway, as envisioned in our proposed model (Figure 5.19). In order to test this hypothesis and the model, quintuple mutants were generated and a through functional characterization and comparison of their phenotypes with that of the two parental lines, galt2galt5 and sos5fei1fei2, was performed. Here, genetic evidence is provided that indicates the five genes act in a single, non-additive genetic pathway, consistent with our model in which GALT2 and GALT5 function in the glycosylation of

SOS5, which in turn interacts with the FEI1/FEI2 cell wall receptor-like kinases to signal cell elongation in roots in the presence of salt, as well as extrusion and production of seed coat mucilage.

297

Figure 5.19 Proposed model linking GALT2 and GALT5 with SOS5/FEI1/FEI2 in cellular signaling of root growth. Signaling of normal root growth involves glycosylated SOS5 binding FEI1/2, possibly inducing dimer formation and activation of the kinase domain as well as allowing for the binding of FEI1/2 to ACC synthase. Such binding will inhibit the production of ACC, a potential signaling molecule and ethylene precursor, which directly or indirectly, inhibits cellulose biosynthesis independent of ethylene. In contrast, when SOS5 glycosylation is inhibited or SOS5 is mutated or FEI1/2 is mutated, ACC synthase can no longer bind to FEI; thus unbound ACC synthase produces ACC, which inhibits cellulose synthesis as well as leads to the production of ethylene.

298

5.4.2. SOS5-FEI1/FEI2 pathway is required for anisotropic growth in the root

Phenotypic analysis of the knock-out quintuple mutants indicated that glycosylated

SOS5 and FEI1/FEI2 are necessary for anisotropic cell expansion in Arabidopsis root cells and also play a role in cell expansion in hypocotyls of etiolated seedlings. It is important to mention that Genetic evidence coupled with biochemical analyses revealed that these genes modulate cell wall function by positively regulating the biosynthesis of cellulose, a wall component crucial for anisotropic expansion. Several receptor-like kinases (RLKs) were identified that likely act as sensors of cell wall integrity, regulating cell growth in response to cell wall perturbations (Steinwand et al., 2010; Steinwand et al., 2014). In Arabidopsis, there are as many as 600 members of receptor-like kinases

(RLKs), many of which were shown to act in a variety of different signaling pathways that function throughout plant development (Gish et al. 2011). Two other well- documented receptor kinases implicated in regulating cell wall function include the wall-associated kinases (WAKs) and four members of the Catharanthus roseus RLK1-

Like kinases (CrRLK1L). The four members include FERONIA (FER), THESEUS1

(THE1), HERCULES1 (HERK1), and HERK2 (Decreux and Messiaen, 2005; Hematy and Hofte, 2008; Guo et al., 2009). WAKs are implicated to function in cell expansion and in cell wall signaling in response to pathogen and stress (He et al., 1998; Kohorn and Kohorn, 2012). Interestingly, wak2 knock-out mutants also displayed impaired cell expansion in root, but unlike FEI1/FEI2 mutants, they exhibit a dependence on sugars and salts for seedling growth (Kohorn et al., 2006). Hematy and Hofte (2008) identified the THESEUS1 (THE1) RLK as a suppressor of cesA6 and suggested that it functions in

299 sensing and regulating cell wall integrity. In contrast, disruption of the two partially redundant FEI1 and FEI2 RLKs was found to cause root growth arrest, root tip swelling, and a reduction in crystalline cellulose production in root tips when grown under restrictive conditions of either elevated sucrose or salt (Xu et al., 2008).

One potential ligand for the FEI1 and FEI2 RLKs is the extracellular FLA, SOS5, which is likely glycosylated by the action of GALT2 and GALT5. Several lines of genetic evidence suggested that glycosylated SOS5 functions, likely interacts with

FEI1/FEI2 either directly or indirectly from the : 1) the patterns of expression of the

GALT2, GALT5, SOS5, FEI1, and FEI2 and are largely overlapping, 2) all these mutants have a very similar root-elongation phenotype in response to elevated levels of salt and sucrose, 3) they all exhibit nearly identical radial swelling of root tips as well as wider hypocotyls, 4) substantial reduction of cellulose biosynthesis in the root tip region as well as reduced cellulosic rays in seed coat mucilage layer, 4) suppression of radial swelling by inhibitors of ethylene biosynthesis by AVG, AIB and CoCl2, but not by genetic disruption of ethylene perception by silver nitrate, 5) ectopic lignin deposition in the roots, 6) the effect of sucrose/salt on these mutants was not an outcome of increased osmotic potential of the medium, and 7) aberrant non adherent mucilage extrusion. 8) double, triple and quintuple mutants displayed non-additive phenotypes, providing genetic evidence which indicates these genes and their encoded proteins, may regulate cell expansion in a linear pathway.

300

5.4.2. Role of ethylene and ABA in GALT2/GALT5-SOS5/FEIs Pathway

Mounting lines of evidence support the link between ethylene and several mutants that affect root growth anisotropy, including sabre, cev1, and lue1 (Aeschbacher et al.,

1995; Ellis et al., 2002; Bouquin et al., 2003). But unlike fei1fei2, the swollen root tip phenotype of these mutants can be rescued by only disrupting either ethylene signaling or by both ethylene action and perception (Aeschbacher et al., 1995; Ellis et al., 2002).

In the root, ethylene strongly inhibits root elongation, but radial expansion is only modestly increased and microtubules appear to be unaffected (Baskin and Williamson,

1992). Thus, in the root, ethylene appears to act primarily by inhibiting the overall amount of cell expansion, not its orientation. In addition, Xu et al. (2008) demonstrated that the cytoplasmic domain of FEI2 interacts with ACC synthase (ACS2/5) proteins, leading to the hypothesis that SOS5 and the FEIs might act in a linear genetic pathway that depends on ACC, but not on ethylene signaling, upstream of cellulose deposition

(Xu et al., 2008). This indicates that swelling in the absence of FEI1/FEI2,

GALT2/GALT5, or SOS5 depends on a hitherto undiscovered pathway for ethylene perception or that ACC itself acts as a signaling molecule.

One potential mechanism that can explain the phenotype observed in case of the impaired GALT2GALT5/SOS5/FEI1FEI2 pathway can be attributed to the elevated levels of ROS in the elongation zone of Arabidopsis roots in response to ACC, which leads to the cross-linking of Hyp-rich glycoproteins and callose deposition in the cell wall, both of which may contribute to reduced cell expansion (De Cnodder et al., 2005).

301

Related to this, Xue et al. (2015) demonstrated that sos5 mutants indeed exhibit production of higher levels of ROS compared to wild type plants.

Together these results suggest that FEI1/FEI2 act redundantly and non-additively and might act in the same genetic pathway, where glycosylated SOS5 acts as a ligand.

Further experiments are needed confirm biochemical interaction between the extracellular domains of FEIs with glycosylated AGPs. It is important to note that there is no direct evidence of glycosylation of SOS5 is catalyzed specifically by GALT2 and

GALT5, however, the HPLC profiles of β -Yariv precipitable AGPs examined from galt2galt5 mutants revealed that virtually all AGPs, as opposed to a single or subset of these AGPs, were affected (Basu et al., 2015). The proposed model for the

GALT2GALT5/SOS5/FEI1FEI2 pathway provides a basis for testing and manipulating the signaling process to better understand and enhance root growth (Figure 5.19). In such a model, AGPs, may bind an unknown signaling ligand or may directly act as plasma membrane/cell wall pressure sensor to relay information to FEI1/FEI2. Indeed,

AGPs are suggested to form an electrostatic cushion between the relatively rigid cell wall and the plasma membrane (Serpe and Nothangle, 1994; Lamport et al., 2006). It is possible that they could effectively act as integrity sensors by sensing differences in turgor pressure when subjected to elevated salt or sucrose.

5.4.3. Effect of glycosylated SOS5-/FEIs pathway on cellulose biosynthesis

The reduction of the cellulosic rays in the quintuple mutants is identical to its parental lines and coincides with an increase in the solubility of the pectinaceous component of seed coat mucilage. This indicates that cellulose plays a structural role in

302 anchoring the pectinaceous mucilage to the seed surface. Although the exact mechanism is unknown, it likely involves interaction with other mucilage components such as pectin, SOS5, and other uncharacterized AGPs (Griffiths et al., 2014).

In conclusion, this study corroborates, links, extends previous biochemical and genetic studies designed to understand the roles of FEI1/FEI2, SOS5, and

GALT2/GALT5 (Xu et al., 2008; Shi et al., 2003; Basu et al., 2015). The current work provides genetic evidence supporting role of AGPs, particularly with respect to their carbohydrate moieties or AG polysaccharides, in cellular signaling mediated by FEIs that regulates both root growth and seed coat mucilage adherence in Arabidopsis.

5.5. Materials and Methods

5.5.1. Plant material

The Columbia (Col-0) ecotype of Arabidopsis thaliana was used in this study. The galt2-1 and galt5-1 mutant lines used in this study were characterized in Basu et al.

(2013, 2015). The triple mutant sos5fei1fei2 was kindly gifted by Dr. Keiber, University of North Carolina Chapel Hill, North Carolina, USA. The quintuple mutant was generated by crossing a double galt2galt5 with a sos5fei1fei2 triple mutant. The cesa5 and cesa5sos5 double mutant were kindly gifted by Dr. Haughn, University of British

Columbia, Vancouver, British Columbia, Canada. The cesa6 and cesa2cesa5cesa9 seeds were kindly gifted by Dr. DeBolt, university of Kentucky, Lexington. The insertion sites all were confirmed by DNA sequencing of PCR-amplified products using gene-specific primers in conjunction with T-DNA primers (Table 5.3) from the respective lines.

303

5.5.2. Plant growth conditions

Arabidopsis thaliana ecotype Col was used throughout this study. Seeds were surface sterilized for 10 min in 30% bleach + 0.01% Tween-20, excessively rinsed in distilled water and plated on 0.6% agar plates containing 1× MS salts pH 5.7, + 1% sucrose. The plates were incubated in the dark at 4°C for 3 d and were subsequently transferred to a 16-h-light/8-h-dark cycle at 22°C for light-grown seedlings.

5.5.3. qRT-PCR analysis

Arabidopsis siliques were staged by marking flowers on the DPA and sampled at 4,

7, 10, and 14 DPA. Total RNA was isolated using the RNeasy Plant Mini Kit (Qiagen) and the RNase-Free DNase Set (Qiagen). First-strand cDNA synthesis was performed from 2 µg of total RNA using oligo-dT (IDT) and Go Script reverse transcriptase

(Promega, Madison, WI, USA). For qRT-PCR, the cDNAs were amplified using

Brilliant II SYBR Green QRT-PCR Master Mix with ROX (Agilent Technologies, La

Jolla, CA, USA) in an MX3000P real-time PCR instrument (Agilent Technologies). The reactions were performed in duplicate. The transcript level was standardized based on cDNA amplification of GAPC was used as the reference gene for qRT-PCR (Huang et al., 2011). The gene-specific primers used are listed in Table 5.3.

5.5.4. RT-PCR

Total RNA was isolated from 14-d-old WT, fei1, fei2, sos5, fei1fei2, sos5fei1fei2, galt2galt5, quintuple mutant seedlings using the RNeasy kit (Qiagen). First-strand cDNA was synthesized from 1 µg of the total RNA pretreated with RNase-free DNase

(Promega) using the Go script reverse transcriptase (Promega) with oligo dT, according

304 to the manufacturer's instructions. qRT-PCR was performed with Brilliant III SYBR

MM with ROX according to the manufacturer's instructions (Agilent Technologies) using gene-specific primers (Table 5.3). Expression levels were calculated using the comparative CT method, which involves normalizing against the geometric mean of the two housekeeping genes (UBQ10 (At4g05320) and PP2A (At1g13320) for each tissue type (Schmittgen and Livak, 2008).

Table 5.3 Primers used in this study Purpose Forward Reverse Screening of homozygous mutants GALT2 TCACTTGGTCATTCCCTTTTG CAAATCGATGGAGTCTCT CCA GALT5 TTTCCACTTTCGACAATTTGG CTAATTACATGGTTTTGCG GG Lba1.3 ATTTTGCCGATTTCGGAAC SOS5 ATGGCCGCCGCAATTAACGT GCCGGAAGAAACTATCTC CACC ACGC FEI1 GAAGCTGGAAATGTTGAATG TTAATCAGAGCTGGAATC AAGA ATAAAATTC FEI2 ACAAATCGATATTGTGTGCA TCAATCGGAGCTGGAGTC ATGACAG GTAGAAG T-DNA left GGCAATCAGCTGTTGCCCGT border primer- CTCACTGGTG JMLB1(fei1) T-DNA left TTACCCAACTTAATCGCCTTG border primer- CAGCACAT JMLB1(fei2) GALT2 TCTTAGACATCGTCCTCTTAG ACACAGCTGGAAATTTTG A CC

305

Table 5.3 continued GALT2 TCTTAGACATCGTCCTCTTA ACACAGCTGGAAATTTTG GA CC GALT5 TATGTGAACACGGAGCTCTT TCCATCTTGAACAGCCGT GCATTC AATTTATGTCT SOS5 CACCATGGCGAACGTAATCT TACCAAAACATAACAAA CAATTTCC ATGCTATAC FEI1 ATATGGAGCAATACCTACA TTAATCAGAGCTGGAATC GC ATAAAATTC FEI2 GAAACTGGAATCTCTTAATG TCAATCGGAGCTGGAGT AAGAGC CGTAGAAGT UBQ10 GTCGACCCTTCACTTGGTGT ATCCTCAAGCTGCTTTCC AG QPCR GALT2 CATAAGCTTAGGCTATTCAA GGTCGACCTTTCAACAAA TTAT GALT5 GATGGACATAAATTACGGCT TCTTAAGAGCTTATCCCA GTTCAAGATGGA TAAGCATA SOS5 TCGGAGTATCCAAAGTTCTT TCATACCAAAACATAAC TTG A AAATGCTA FEI1 AATCACGAAAAACGGCCAA TTAATCAGAGCTGGAATC GGGATA ATAAAATTC FEI2 TAAACTTCTTAATCAGCGAA TCAATCGGAGCTGGAGT AACCGGG CGTAGAAGT GAPC TCAGACTCGAGAAAGCTGCT GATCAAGTCGACCACAC AC GG

306

5.5.5. Cellulose synthesis assays

Cellulose synthesis was determined by [14C]Glc labeling as described by Fagard et al. (2000) and Xu et al. (2008) with following modifications. The WT, galt2galt5, sos5fei1fei2 and quintuple seedlings were grown on 0% sucrose MS plates for 4 d and then transferred to MS medium containing either 4.5% sucrose or 100mM NaCl for 5d.

Root tips (~1.5 cm) were cut and washed three times with 3 mL of glucose-free MS medium. Forty root tips were then incubated in 1 mL of MS medium containing 0.1

µCi/mL [14C]Glc for 1 h in the dark at 22°C in glass tubes. Subsequently, these roots were washed three times with 1 mL of glucose-free MS medium. Next, the roots were extracted three times with 1 mL of boiling absolute ethanol for 20 min, and total aliquots were collected (ethanol-soluble fraction). Roots were then resuspended in 1 mL of chloroform:methanol (1:1, v/v), extracted for 20 min at 45°C, and finally resuspended in

1 mL of acetone for 15 min at room temperature with gentle shaking. The remaining material was resuspended in 500 µL of an acetic acid:nitric acid:water solution (8:1:2, v/v/v) for 1 h in a boiling-water bath. Acid-soluble material and acid-insoluble material were separated by glass microfiber filters (GF/A; 2.5 cm diameter), after which the filters were washed with 3 mL of water. The acid wash and water wash constitute the acid-soluble fraction. The filters yield the acid-insoluble fraction. The amount of label in each fraction was determined by scintillation counting using liquid scintillation fluid.

The percentage of label incorporation was expressed as 100× the ratio of the amount of label in each fraction to the total amount of label (ethanol plus acid-soluble plus acid- insoluble fractions).

307

5.5.6. Germination assay

Mature seeds were placed on filter paper in petri dishes and moistened with an aqueous solution of PEG 8000 at 12–21% concentration (w/v). The seeds were cold treated at 4 0C for 4 days in the dark and then grown at 22 0C in 100 µE constant light for another 5 days. Germination was scored by radical protrusion.

5.5.7. Crystalline cellulose determination

Crystalline cellulose levels were determined based on a microscale modification by

Updegraff (1969) as described in Griffiths et al. (2014). Twenty five milligrams of seeds was frozen in liquid nitrogen, ground using mortar and pestle, and then dried at 50°C overnight. Ground seeds were then treated with 1 mL of the acetic-nitric acid reagent

(stock solution, 150 mL of 80% [v/v] glacial acetic acid diluted with water and 15 mL of

[70%] concentrated nitric acid) and vortexed. Samples were heated at 100°C for 1 h, centrifuged for 5 min at 3,000 rpm, and washed. Samples were then treated with 1 mL

72% (w/v) H2SO4 vortex, incubated at room temperature for 1 h, centrifuged for 5 min at

10,000 rpm, and then diluted 10 times in distilled water (duplicated samples). One hundred microliters of diluted sample was treated with 200 µL of cold freshly prepared anthrone reagent (0.2% [w/v] anthrone [Sigma-Aldrich] in concentrated H2SO4) and vortexed. Anthrone mixtures were incubated for 15 min at 100°C, and duplicate samples were measured two times for A620 in a spectrophotometer. A standard curve was prepared from a standard dilution of dextrose. Total amounts of cellulose were calculated per weight of dry seed mass.

308

5.5.8. Cell wall preparation

Hundred milligram of WT, sos5fei1fei2, quintuple cesa5sos5, sos5, seeds were extracted sequentially with 0.2% ammonium oxalate, 0.2N and 2N sodium hydroxide for

1 h each with vigorous shaking at 37 oC. Both sodium hydroxide extractions were neutralized with acetic acid. Total sugar (µg/ mg seed) was determined with a phenol- sulfuric assay based on that of Dubois et al. (1956). In short, 200 µl of resuspended extract was incubated with 100 µl freshly made 5% (v/v) aqueous phenol and 1 ml concentrated sulfuric acid for 2 h at 30 oC. Absorbance was detected at 500 nm against glucose standards 0.5, 2.5, 5, 7.5, 10, 15, 25 µg for which a linear response curve was obtained.

5.5.9. Seed staining and visualization

Seeds of all the indicated genotypes were prehydrated in water and stained either with 0.01% Pontamine Fast Scarlet S4B. In both cases, staining was performed as described by Willats et al. (2001) and Harpaz-Saad et al. (2011). Imaging was done using a Zeiss LSM 510 confocal microscope.

5.5.10. Phenotypic analysis of the mutants in response to salt

Analysis of mutant root phenotype in response to salt stress was analyses as described in Basu et al. (2015).

5.5.11. Phloroglucinol Staining

Lignin staining was performed using phloroglucinol as described (Cano-Delgado et al., 2003). Seedlings were fixed in a solution of three parts ethanol to one part acetic

309 acid for 15 min then seedlings were then stained with lignin in a 2% phloroglucinol-HCl solution for 5 min.

310

CHAPTER 6: GENERAL CONCLUSIONS

AGPs are fascinating and undoubtedly one of the most complex families of proteins found in plants. Their complexity arises not only from the incredible diversity of the glycans decorating the protein backbone but also from diversity of protein backbones containing AG glycomodules. In general, the key distinguishing features (with notable exceptions) of AGPs includes –i) their protein backbone typically constitutes of 10%

(w/w) and is rich in Hyp/Pro, Ala, Ser, and Thr, with the dipeptide motifs Ala-Hyp, Ser-

Hyp, Thr-Hyp, Val-Pro, Gly-Pro and Thr-Pro, ii) type II AGs that are O-glycosidically linked to Hyp residues on the protein backbone. These type II AGs have (1-3)-β-D- linked Galp residues that form a backbone substituted at C6 by galactosyl side chains, usually terminating in Araf, Rhap, and Galp residues, giving rise to a glycan chain, iii) their ability (with few exceptions) to bind a class of synthetic chemical dyes, the β-Yariv reagents (Figure 3.25; Yariv et al., 1967). β -Yariv reagent has proven very useful in their detection, quantification, and precipitation. The glycan structures of AGPs remain a conundrum yet to be resolved. Tan et al. (2010) reported that the AGP glycan is composed of repeated units of β-1,3-galactotriose with or without side chains connected by β -1,6-linkages based on NMR analysis of the synthetic glycosylated peptide expressed in tobacco which mimics an AGP protein backbone. On the contrary, longer

β-1,6-galactan side chains are reported for AGPs from radish root (Haque et al., 2005), wheat flour (Tryfona et al., 2010), and Arabidopsis leaf (Tryfona et al., 2012) supported by a combination of analytical approaches, including the use of enzymes to release oligosaccharides specifically from AGs, high-energy matrix-assisted laser desorption

311 ionization (MALDI)-collision-induced dissociation (CID) mass spectrometry (MS), and polysaccharide analysis by carbohydrate gel electrophoresis. One plausible explanation may be that the structures of AGs on heterologously expressed proteins may be different from those found on endogenous AGPs due to overload of substrates (Tan et al., 2004).

Despite the fact that plants contain substantial amounts of AGPs, there is a lack of knowledge on the enzymology of AGP glycan biosynthesis. AGP biosynthesis involves

O-galactosylation of Hyp residues followed by a stepwise elongation of the complex sugar chains. Thus, AGP biosynthesis requires many different proteins, and identifying these proteins and determining their biochemical activity is challenging. This dissertation research focused on the identification of AGP-specific Hyp-O-GALTs involved in the initial glycosylation of AGPs and the functional characterization of Hyp-

O-GALT mutants. The overall goal was to examine enzymatic synthesis of the carbohydrate moieties decorating AGPs and evaluate the contribution of AGP glycosylation to biological functions. Moreover an attempt was made to elucidate the role of AGPs in cellular signaling. Overall, my intent was to unravel the structure, function, and biosynthesis of AGPs. A diversified approach utilizing biochemistry, molecular genetics, physiology, proteomics, and cell biology were employed to determine the function of the carbohydrate moieties of AGPs in plant growth, development and reproduction.

Prior to this research, only two fucosyltransferases involved in addition of terminal fucose residues and one β-(1,3) GALT (At1g77810) involved in the synthesis of the β-

(1,3)-galactan backbone of AGPs glycans were characterized (Wu et al., 2008; Qu et al.,

312

2008). Initiation of AGP biosynthesis requires the action of Hyp-O-GALTs, to initiate the glycosylation of AGP backbone. In an attempt to decipher the glycosylation machinery involved in the biosynthesis of glycan structures of AGPs, I investigated the

CAZy family GT31 (www.cazy.org), a GT family implicated in the biosynthesis of

AGPs (Qu et al. 2008; Egelund et al., 2011). This family has 33 members in

Arabidopsis, 20 of which contain a putative GALT domain (GALT; Pfam01762, http://pfam.sanger.ac.uk/) and one of which (GALT1) was previously identified as a β-

(1,3)-GALT involved in the biosynthesis of a protein-bound N-linked oligosaccharide

(Strasser et al., 2007). Interestingly, only six of these 20 putative GALTs, contain both a

GALT catalytic domain and a GALECTIN carbohydrate binding domain. The

GALECTIN domain was previously identified in UDP-N-acetyl D- galactosamine:polypeptide N-acetylgalactosaminyltransferase-6 (ppGalNAc-T6) GT 27 family involved in catalyzing the first steps of the glycosylation of mammalian mucins, which are animal extracellular proteins analogous to AGPs (Hassan et al., 2000;

Wandall et al., 2007). My approach in identifying candidate Hyp-O-GALT enzymes involved in the biosynthesis of AGPs builds upon a robust in vitro AGP-specific Hyp-

GALT assay developed by Liang et al. (2010) in the Showalter laboratory. Towards this goal, these six putative GALTs with GALECTIN domains were chosen as the best candidates for adding the first Gal residue directly onto the AGP protein backbone by analogy to animal GT 27 family members. These six genes were designated as GALT1-

GALT6, to be consistent with the nomenclature of one of its members, GALT1.

313

Utilizing the in vitro GALT assay, heterologously expressed five out of six, namely

GALT2-GALT6, but not in GALT1, demonstrated addition of a single galactose (from

UDP-Gal) specifically onto AGP peptide substrates. Extensive product characterization of [AO]7 :GALT2 and [AO]7 :GALT5 reaction products, using reverse phase HPLC, acid hydrolysis/Dionex, base hydrolysis/Bio-gel P2 supported these findings (Figures 2.3,

2.5, 3.3 and 3.5). In addition, consistent with their activity, these five GALTs fused with

YFP colocalized exclusively with the known Golgi marker ST-GFP for except GALT2, which is localized to both the ER and Golgi vesicles ((Boevink et al., 1998; Figures 2.9,

3.8 and 4.8).

The five Arabidopsis AGP Hyp-O-GALT genes show overlapping but distinct expression profiles which indicates tissue/cell-specific expression patterns of these five genes and highlights gene redundancy and compensation mechanisms as depicted in

Figures 3.10, 3.11 and 4.5. This also explains why a reduced content of β -Yariv precipitable AGPs could be observed for all the loss-of function galt mutants, while all the galt mutants retained significant amounts of residual AGPs and a more severe reduction of AGPs was observed in the galt2galt5 double mutant (Table 3.2). Several lines of evidence support the notion that all five GT31 members encode AGP-specific

Hyp-O-GALTs –i) These five proteins share very similar protein structures with amino acid identities ranging from 35%-70% (Table 4.2). ii) Incorporation of [14C]Gal from

14 UDP-[ C]Gal into the [AO]7 was observed following HPLC fractionation of the Hyp-

O-GALT reaction products. iii) Hyp-[14C]Gal monosaccharide was identified in base hydrolysates of the Hyp-O-GALT reaction iv) All five Hyp-O-GALTs share similar

314 substrate specificity i.e. exclusive preference for peptidyl Hyp in AGP sequences. v) All five Hyp-O-GALTs were localized to the endomembrane system vi) Significant reduction in the endogenous Hyp-O-galactosylation activity was observed in their mutants. vii) A significant reduction of β-Yariv precipitable AGPs was observed in the single mutants and more severe reductions were observed in the double mutants compared to WT. viii) Less sensitivity towards β-Yariv reagent was seen in the loss of function mutants. ix) Substantial reductions of AGP epitopes were found in the knock- out mutants in immunostaining analysis using AGP-specific antibodies.

Recently, using affinity chromatography-based purification and molecular cloning, three additional Hyp-O-GALTs namely HPGT1-3, belonging to the CAZy GT31 family were identified (Ogawa-Ohnishi and Matsubayashi, 2015; Figure 2.1). The three

HPGTs and the five Hyp-O-GALTs are conserved throughout the land plants including mosses. This indicates that plants employ at least two structurally distinct classes of enzymes for Hyp O-galactosylation.

In addition to Hyp-O-GALTs and β -(1,3)-GALTs, AGP glycan synthesis also requires a number of other GTs like β-1,6-GALT, α-1,3- and α-1,5-arabinosyltransferase

(Ara-T), β-glucuronosyltransferase (GlcA-T) rhamnosyltransferases and xylosyltransferases (Figure 3.25). To date several GTs have been identified as listed in

Table 3.3 and 4.9. However, a number of GTs still await identification and characterization including rhamnosyltransferases and xylosyltransferases and other β-

(1,3)-GALTs.

315

For GALT2-6, the detected enzymatic activity in both the heterologous system, was much lower compared with the microsomes extracted from Arabidopsis suspension- cultured cells. This phenomenon has three possible explanations: (i) additional GALT(s) with the same function may exist, (ii) some unknown cofactor(s) or protein(s) serve to enhance the enzyme activity of the five Hyp-O-GALTs in Arabidopsis, and (iii) they may work in concert in a multi-enzyme complex. Indeed, it is possible that a membrane complex of AGP glycan synthase may exist in plants. In plants, growing evidence indicates that multi-enzyme complexes are responsible for biosynthesis of cell wall polysaccharides and glycoproteins. Indications come from the biosynthesis of pectin

(GAUT1 and GAUT7; Atmodjo et al., 2011), (ARAD1 and ARAD2; Harholt et al.,

2012), XyG (CSLC4, XXT1/XXT2, and XXT5; Chou et al., 2012), glucuronoarabinoxylan (IRX10 and IRX14; Zeng et al., 2010), and N-glycosylated protein (GMI, GnTI, GMII and XylT ; Schoberer et al., 2013; and HRGPs by P4H5,

P4H 2, P4H 13; Velasquez et al., 2014). Based on subcellular co-localization approaches, FRET acceptor photo-bleaching techniques as well as immuno-precipitation techniques, Dilokpimol et al. (2014) reported that GALT31A and GALT29A were organized into heterodimer complexes, which had enhanced β -1,6-GALT enzymatic activity. This cooperative action between AtGALT31A and AtGALT29A plays a potential regulatory mechanism for producing β-1,6-galactan side chains of type II AG during plant development and provide further evidence for an enzyme complex involved in the biosynthesis of AG polysaccharides. Both GALT31A and GALT29A are type II transmembrane proteins located in the Golgi vescicles, suggesting that addition of β-1,6-

316 galactose residues to the side chains of AGPs occurs later during the transit of nascent

HRGPs into Golgi stacks.

Additionally, three β -glucuronosyltransferases (GlcAT), GlcAT14A, GlcAT14B, and GlcAT14C, which are responsible for the addition of glucuronic acid (GlcA) residues to AGPs were identified (Knoch et al. 2013; Dilokpimol and Geshi 2014). All three enzymes are members of the CAZy GT14 family and are reported to add GlcA to both β-(1,6)- and β-(1,3)-galactan chains. Even though, GlcAT14A is co-expressed with

GALT31A and co-localize in the Golgi apparatus, the FRET photo-bleaching acceptor technique revealed that they do not physically interact. Thus, these findings suggest that all the enzymes involved in AGP glycan synthesis, although probably co-regulated, are not necessarily part of a multi-protein complex.

Ara, after Gal, is the second most abundant sugar found in AGP glycans. Recently,

Gille et al. (2013) identified the RAY1 gene, a member of CAZy GT77, which encodes an arabinosyltransferase. Another important aspect of AGP glycan biosynthesis is its sequential addition of the sugar moieties along the compartments of endomembrane system, unlike N-glycan biosynthesis. Biochemical evidence that supports this notion was the observation that de-arabinosylation of tobacco AGP glycans by a specific arabinofuranosidase prevented the addition of fucose residues to the glycan, suggesting that arabinosylation was required for further addition of fucose by FUT4 and FUT6 (

Wu et al., 2010). Collectively, these studies suggest that for the assembly of AG chains, more than one Hyp-O-GALTs, β-(1,3)-GALTs, and β-(1,6)-GALTs are required along with GTs that decorated the AG termini. Further, glycan biosynthesis is initiated in the

317

ER and continues in the Golgi apparatus in a sequential manner, with elongation or branching of the AGP glycans probably occurring in different Golgi sub-compartments before their export to the cell surface. In this regard, localization of GALT2 in both ER and Gogi may be critical for transport of glycosylated AGP from ER to Golgi but further experiments are needed with different ER and Golgi markers to confirm these findings

(Figure 2.9). Indeed, more experiments are needed to identify and demonstrate the biochemical action of additional genes and enzymes still involved in the complex process of AG synthesis.

It will be interesting to determine the function of the GALECTIN domain found in

GALT2-6, especially whether and how it influences GALT activity. It can be speculated the GALECTIN domain in Hyp-O-GALTs behaves similar to that of human ppGalNAc-

T6 based on the fact that they both initiate O-glycosylation (Clausen and Bennett 1996).

Recent reports strongly indicate that ppGalNAc-T6 has glycopeptide specificity while maintaining peptide substrate specificity by using a series of novel random glycopeptide substrates containing a single GalNAc-O-Thr residue placed near either the N or C terminus of the glycopeptide substrate (Gerken et al., 2013). This can also explain why

20 isoforms of ppGalNAc-T6 are required in humans, although they catalyze the same enzymatic step, as there exists subtle tissue and cell specific differences and preferences towards substrates. Taken together, this indicates that O-glycosylation may in some instances be exquisitely controlled in humans and the same may be true for plants.

Despite much evidence supporting the involvement of AGPs in multiple aspects of plant growth and development, functional evaluation of specific AGPs remain elusive.

318

Genetic mutant analysis has proven fruitful and fundamental to the functional analysis of

Hyp-O-GALTs as the elucidation of the enzymes involved in the biosynthesis of AG glycans is expected to facilitate an understanding of the function of AGPs. To determine the functions of the Hyp-O-GALT in planta, I identified and analyzed two allelic loss- of-function mutants in each of the five AGP Hyp-O-GALTs. These mutants displayed pleiotropic growth and developmental defects (Table 3.2). The single galt mutants did not have dramatic phenotypes under normal growth conditions, except for reduced root hairs, less seeds in the siliques of galt4 and galt6, reduced β-Yariv precipitable AGPs, and reduced AGP specific Hyp-O-GALT activity (Figures 3.15, 4.5, 4.6, 4.11, 4.12 and

Tables 2.1 and 3.1). This suggests functional redundancy within the Arabidopsis Hyp-

O-GALT family. More interestingly, the galt2galt5 double mutant displayed not only substantially greater reductions of β-Yariv precipitable AGPs and AGP specific Hyp-O-

GALT activity, but also it displayed delayed flowering, aberrant seed coat mucilage extrusion, and reduced cellulosic rays in the adherent seed coat mucilage layer (Figures

3.24, 5.15, 5.17, 5.16, 5.17 and Table 3.1). In addition, significant reductions in the distribution and intensity of immunofluorescent labeling of AGP epitopes with a suite of

AGP-directed monoclonal antibodies was observed in galt2galt5 double mutants compared to WT plants (Figure 3.13).

Conditional mutant phenotypes were also observed, including salt-hypersensitive reduced root growth and root tip swelling as well as reduced inhibition of pollen tube growth and root growth in response to β-Yariv reagent (Figures 3.17, 3.18 4.16 and

4.17). Like the loss of function mutants of Hyp-O-GALTs, hpgt triple knock-out mutants

319 exhibited similar but distinct pleiotropic phenotypes, namely longer lateral roots, longer root hairs, and radial expansion of the cells in the root tip, smaller leaves, shorter inflorescence stems, reduced fertility and shorter siliques compared to WT.

Interestingly, hpgt triple mutants displayed longer root hairs contrary to galt single and galt2galt5 double mutants (Ogawa-Ohnishi and Matsubayashi, 2015; Basu et al., 2015).

One possible explanation may be that GALT2-6 and HPGTs have some preferences for glycosylating certain AGP backbones over the others, in other words they may glycosylate subsets of AGPs.

It is important to mention that GALECTIN domain in the five Hyp-O-GALTs, is only predicted to bind Gal and needs to be confirmed through structural and biochemical approaches. Hence, any phenotypes of these Hyp-O-GALTs may be attributed to the disruption of their GALECTIN domain as opposed to GALT domain. One way to conclusively address this will be to express GALECTIN domain in a galt mutant background and look for complementation and repeat the same experiment with the

GALT domain. Nonetheless one argument against such a notion is that three recently characterized Hyp-O-GALTs, HPGT1-3 share pleiotropic phenotypes similar to that of five Hyp-O-GALTs even though they do not have the GALECTIN domain. Thus, it is unlikely that different domains of the same protein will confer divergent physiological roles in plants. Nonetheless, further experiments will help in confirming such a hypothesis.

To date, there appears to be a variety of functions exhibited by mutants in AGP- glycosylation, ranging from embryo lethality to conditional effects on root growth, cell

320 elongation, and pollen tube growth. But the results available thus far are inconclusive concerning the molecular roles of a particular portion or sugar of AGP glycans in plant growth and development. Nonetheless, this mutant work clearly indicates that AGP glycans are essential for optimal plant growth and development. Perhaps this work will be come clearer as more than one redundant gene is knocked out in a single plant to reveal more drastic phenotypes during normal growth and development. It will be also interesting to analyze the structure of AGPs in the mutants which will in turn help to relate the role of glycosylation to AGP function. And finally it will be important to unravel the mode of action of AGPs in regulating these pleitropic functions

Another important aspect addressed in the current work is the determination of exactly which AGPs out of the 85 different AGPs predicted in Arabidopsis are being acted on by these Hyp-O-GALTs. Towards this goal, I compared HPLC profiles of β-

Yariv precipitable AGPs obtained from 14-d old mutant seedlings with that of WT seedlings. Disruption of GALTs in both single and double mutants revealed that virtually all these AGPs, as opposed to a single or subset of these AGPs, were affected in the single and double mutants, with the double mutant being more severely affected (Figure

3.12). Consistent with the ubiquitous expression of these five Hyp-O-GALTs, HPLC profiling of β -Yariv precipitable AGPs further corroborate that these Hyp-O-GALTs globally affect AGP glycosylation in the galt mutants. However, the possibility of unique glycosylation properties of these Hyp-O-GALTs in other organs or tissues cannot be excluded. Nonetheless, as Hyp-O-galactosylation is a highly specific post translational modification to AGPs and indispensable for further elongation of the AG

321 chains, the results in this dissertation unambiguously demonstrate the critical importance of Hyp-O–galactosylation to biological function.

Finally, I have provided genetic evidence for a likely non-additive linear signaling pathway involving interactions among FEI1 and FEI2 (two partially redundant cell wall receptor-like kinases), SOS5 (a GPI-anchored fasciclin-like AGP), and GALT2 and

GALT5 (two AGP specific Hyp-O-GALTs). It is important to mention that further biochemical evidence is required to confirm such interactions especially whether SOS5 is glycosylated by GALT2 and GALT5 using deglycosylated SOS5 as substrates. My work indicates that this FEIs-glycosylated SOS5 pathway controls cell wall function in both Arabidopsis roots and seed coat mucilage by regulating cellulose biosynthesis. In roots, disruption of FEI1, FEI2 , SOS5 or both GALT2 and GALT5 results in virtually identical conditional phenotypes, which includes swollen root tips, and reduced root elongation in response to elevated salt or sucrose (Shi et al., 2003; Xu et al., 2008; Basu et al., 2015; Figures 5.6, 5.7 and 5.8). The phenotype of the galt2galt5 double mutant is slightly less severe compared to fei1fei2 or sos5, which can be attributed to gene redundancy among Hyp-O-GALTs. Impaired cellulose biosynthesis at the root tips is likely responsible for such defective growth phenotypes. The root phenotype of fei1fei2, sos5 and galt2galt5 can be suppressed by inhibitors of ethylene biosynthesis, but not by disruption of the ethylene signaling pathway, indicating a role for ACC as a signaling molecule in this pathway (Figures 5.12). This sheds light into phytohormonal regulation of cell wall integrity.

322

Harpaz-Saad et al. (2011; 2012) found cellulose to be as an essential component of seed mucilage and identified CESA5, FEI2 and SOS5 as regulators of cellulose deposition in the form of rays radiating from the seed across the inner layer of seed mucilage. Other studies defined additional, partially redundant functions for CESA2,

CESA5 and CESA9 in cellulose deposition into the radial cell walls of mucilage secretory cells in seed (Sullivan et al., 2011; Mendu et al 2011; Stork et al., 2010).

Results here indicate that glycosylated SOS5, synthesized by the action of GALT2 and

GALT5, is essential for the FEIs-SOS5 pathway based on the identical non-additive phenotypes displayed by quintuple (galt2galt5sos5fei1fei2) mutants (Figures 5.19). One possible mechanism which may explain this pathway relates to lipids rafts. It has long been suggested that AGPs with GPI anchors may influence perception of signaling molecules. One proposed hypothesis is that it involves lipid rafts where organizing proteins such as GPI anchored AGPs and cell wall receptor kinases with transmembrane domains along with CSC (cellulose synthase complexes) may exist in ideal microenvironments that facilitate signaling events critical for the maintenance of cell wall integrity (Simons and Toomre, 2000). Additionally, lipid rafts were suggested to connect AGPs with microtubules perhaps aided by membrane-binding or membrane- spanning, microtubule-associated proteins as exemplified by COBRA, a GPI-anchored protein associated with cortical microtubules (Sardar et al., 2006; Gardiner et al., 2001).

Thus, future research aimed at uncovering new molecules regulating cellulose biosynthesis will benefit from the findings here. Whether direct or indirect physical interaction occurs between the extracellular domains of FEI1/FEI2 and glycosylated

323

SOS5 remains an important question for future studies. Mounting evidence supports the role of AGP glycans in regulating AGP function. Indications for the importance of the carbohydrate chains in determining biological function comes from the analysis of mur-

1 and reb1-1 mutants (van Hengel and Roberts 2002; Andème-Onzighi et al., 2002). In addition, Liang et al. (2013) demonstrated that disruption of FUT4 and FUT6 not only resulted in reduced root growth but also affect intramolecular interactions between

AGPs and other wall components. The mutation in GALT31A caused the arrested embryo development at the globular stage thereby linking the requirement of correctly glycosylated AGPs with the progression of embryogenesis beyond the globular stage

(Geshi et al., 2013). However, unlike GALT31A, under optimal conditions, galt2galt5 showed only subtle phenotypes suggesting gene redundancy and compensation. Xi et al.

(2014) speculated that AGP9 may affect xylem vessel differentiation and vessel cell expansion by interacting with transmembrane proteins, such as receptor kinases. In another study, Costa et al. (2013) postulated that AGPs can act as receptors for extracellular growth signals and interact with transmembrane proteins, possibly receptor kinases, to mediate pollen tube elongation.

These findings in this dissertation provide important insights into AGP biosynthesis and raise several interesting questions for future studies. One of them is why eight Hyp-

O-GALTs exist for initiation of AGP glycan biosynthesis (Ogawa-Ohnishi and

Matsubayashi, 2015; Basu et al., 2015 unpublished data Basu et al.). A plausible explanation may be that these Hyp-O-GALTs have cell/tissue-specific expression patterns regulated temporally and spatially for proper functioning in specific cell types.

324

Transcriptomic analyses using RNA extracted from laser-capture dissected seed coat tissue supports such hypothesis (Le et al., 2010; Figures 4.7 and 5.2). Another unresolved question is why both GALT2 and GALT5 and the three HPGT catalyze addition of only a single Gal residue onto the penultimate Hyp residue. This suggests the existence of other β-(1,3)-GALTs which will elongate the AG polysaccharide (Ogawa-

Ohnishi and Matsubayashi, 2015; Liang et al., 2010). Transient or stable expression of

Hyp-O-GALTs along with the one characterized β -(1,3)-GALTs and two β -(1,6)-

GALTs in tobacco followed by product characterization may address this hypothesis.

Another strategy will be to synthesize potential substrate acceptors, either chemically or by purifying native AGPs followed by treatment with specific microbial glycosyl hydrolases or chemicals (e.g. partial acid hydrolysis) and using them for monitoring β-

(1,3)-GALT activity. Although Liang et al. (2010) provided evidence of a second Hyp-

[14C]Gal disaccharide based upon base hydrolysis of the GALT reaction products using

Arabidopsis microsomal membranes, still no AGP specific β -(1,3)-GALTs were identified except At1g77810 (Qu et al., 2008). Furthermore, an extensive phenotypic comparison of all the knock-out mutants of AGP GTs in conjunction with biochemical analysis of AGP glycosylation is required to conclusively unveil the role of specific glycan structures.

Taken together, recent identification and biochemical characterization of a number of AGP specific GTs has advanced our understanding of AGP biosynthesis. However, how these enzymes are spatially organized and assembled within different compartments of the endomembrane system and how they are regulated during vegetative and

325 reproductive plant growth is not fully understood and offers exciting research opportunities for the future. Even more challenging will be to examine the potential relationships between a given glycan/oligosaccharide structure and its biological function in a given tissue. This research is important and has implications not only for elucidating the basic functions of AGPs in terms of plant growth and development, but also for enhancing crop productivity for agriculture and biomass production for the biofuel industry.

REFERENCES

1. Aeschbacher R, Hauser M.-T, Feldmann KA, Benfey PN. (1995) The SABRE

gene is required for normal cell expansion in Arabidopsis. Genes Dev. 9: 330–340.

2. Adams DO, Yang SF. (1977) Methionine metabolism in apple tissue: Implication

of S-adenosylmethionine as an intermediate in the conversion of methionine to ethylene.

Plant Physiol. 60: 892–896.

3. Albenne C, Canut H, Jamet E. (2013) Plant cell wall proteomics: the leadership

of Arabidopsis thaliana. Front Plant Sci. 4: 111.

4. Albersheim P, Darvill A, Roberts K, Sederoff R, Staehelin A (2011) Plant Cell

Walls. (New York: Garland Science, Taylor and Francis Group).

5. Andème-Onzighi C, Sivaguru M, Judy-March J, Baskin TI, Driouich A. (2002)

The reb1-1 mutation of Arabidopsis alters the morphology of trichoblasts, the

expression of arabinogalactan proteins and the organisation of cortical microtubules.

Planta 215: 949-958.

6. Arioli T, Peng L, Betzner AS, Burn J, Wittke W, Herth W, Camilleri C, Höfte H,

Plazinski J, Birch R, Cork A, Glover J, Redmond J, Williamson RE. (1998) Molecular analysis of cellulose biosynthesis in Arabidopsis. Science 279: 717–20.

7. Atmodjo MA, Sakuragi Y, Zhu X, Burrell AJ, Mohanty SS, Atwood JA, Orlando

R, Scheller HV, Mohnen D. (2011) Galacturonosyltransferase (GAUT) 1 and GAUT7

are the core of a plant cell wall pectin biosynthetic

homogalacturonan:galacturonosyltransferase complex. Proc. Natl. Acad. Sci. USA 108:

20225–20230.

327

8. Arsovski AA, Popma TM, Haughn GW, Carpita NC, McCann MC, Western T.

(2009) AtBXL1 encodes a bifunctional β -D-xylosidase/α-L-arabinofuranosidase required for pectic arabinan modification in Arabidopsismucilage secretory cells. Plant

Physiol. 150: 1219–1234.

9. Bacic A, Churms SC, Stephen A M, Cohen PB, Fincher GB. (1987) Fine- structure of the arabinogalactan protein form Lolium multiflorum. Carbohydr Res. 162:

85–93.

10. Baek D, Jiang J, Chung J-S, Wang B, Chen J, Xin Z, Shi H. (2011) Regulated

AtHKT1 Gene Expression by a Distal Enhancer Element and DNA Methylation in the

Promoter Plays an Important Role in Salt Tolerance Plant Cell Physiol. 52: 149-161.

11. Baldwin TC, McCann MC, Roberts K (1993) A novel hydroxyproline-deficient arabinogalactan protein secreted by suspension-cultured cells of

(purification and partial characterization). Plant Physiol. 103: 115–123.

12. Basu D, Liang Y, Liu X, Himmeldirk K, Faik A, Kieliszewski M, Held M,

Showalter AM. (2013) Functional identification of a hydroxyproline-O- galactosyltransferase specific for arabinogalactan protein biosynthesis in Arabidopsis. J.

Biol. Chem. 288: 10132–10143.

13. Basu D, Wang W, Ma S, DeBrosse T, Poirier E, Emch K, Soukup E, Tian L,

Showalter AM. (2015) Two Hydroxyproline Galactosyltransferases, GALT5 and

GALT2, Function in Arabinogalactan-Protein Glycosylation, Growth and Development in Arabidopsis. PLoS ONE 10(5): e0125624.

328

14. Baskin TI, Williamson R.E. (1992) Ethylene, microtubules and root morphology in wild-type and mutant Arabidopsis seedlings. Plant Biochemistry and Physiology

Symposium 11: 118–130.

15. Batoko H, Zheng H Q, Hawes C, Moore I. (2000) A Rab1 GTPase is required for transport between the endoplasmic reticulum and Golgi apparatus and for normal Golgi movement in plants. Plant Cell 12: 2201–2217.

16. Baumberger N, Steiner M, Ryser U, Keller B, Ringli C. (2003) Synergistic interaction of the two paralogous Arabidopsis genes LRX1 and LRX2 in cell wall formation during root hair development. Plant J. 35: 71–81.

17. Ben-Tov D, Abraham Y, Stav S, Thompson K, Loraine A, Elbaum R, de Souza

A, Pauly M, Kieber J.J, Harpaz-Saad S. (2015) COBRA-LIKE2, a Member of the

Glycosylphosphatidylinositol-Anchored COBRA-LIKE Family, Plays a Role in

Cellulose Deposition in Arabidopsis Seed Coat Mucilage Secretory Cells. Plant Physiol.

167: 711-724.

18. Bennett EP, Mandel U, Clausen H, Gerken TA, Fritz TA, Tabak LA. (2011)

Control of mucin-type O-glycosylation. A classification of the polypeptide GalNAc- transferase gene family. Glycobiology 22: 736-756.

19. Bleecker AB, Kende H. (2000) Ethylene: a gaseous signal molecule in plants.

Annu Rev Cell Dev Biol. 16: 1–18

20. Boerjan W, Ralph J, Baucher M. (2003) Lignin biosynthesis. Annu. Rev. Plant

Biol. 54: 519–546.

329

21. Boavida LC, McCormick S. (2007) Temperature as a determinant factor for increased and reproducible in vitro pollen germination in Arabidopsis thaliana. Plant J.

52: 570–582.

22. Boevink P, Oparka K, Santa Cruz S, Martin B, Betteridge A, Hawes C. (1998)

Stacks on tracks: The plant Golgi apparatus traffics on an actin/ER network. Plant J. 15:

441–447.

23. Boisson-Dernier A, Lituiev DS, Nestorova A, Franck CM, Thirugnanarajah S,

Grossniklaus U. (2013) ANXUR Receptor-Like Kinases Coordinate Cell Wall Integrity with Growth at the Pollen Tube Tip Via NADPH Oxidases. PLoS Biology 11: e1001719.

24. Bonawitz ND, Chapple C. (2010) The genetics of lignin biosynthesis: connecting genotype to phenotype. Annu Rev Genet. 44: 337-363.

25. Borner GHH, Sh.errier DJ, Weimar T, Michaelson LV, Hawkins ND, MacAskill

A, Napier JA, Beale MH, Lilley KS, Dupree P. (2005) Analysis of detergent-resistant membranes in Arabidopsis: evidence for plasma membrane lipid rafts. Plant Physiol

137: 104–116.

26. Boron AK, Van Orden, J, Nektarios Markakis, M, Mouille, G., Adriaensen, D,

Verbelen, JP, Hofte, H, Vissenberg, K. (2014) Proline-rich protein-like PRPL1 controls elongation of root hairs in Arabidopsis thaliana. J. Exp. Bot. 65: 5485–5495.

27. Boratyn GM, Schäffer AA, Agarwala R, Altschul SF, Lipman DJ, Madden TL.

(2012) Domain enhanced lookup time accelerated BLAST. Biol. Direct. 7: 12.

330

28. Bouquin T, Mattsson O, Naested H, Foster R, Mundy J. (2003) The Arabidopsis lue1 mutant defines a katanin p60 ortholog involved in hormonal control of microtubule orientation during cell growth. J. Cell Sci. 116: 791–801.

29. Boyes DC, Zayed AM, Ascenzi R, McCaskill AJ, Hoffman NE, Davis K,

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

30. Boughanmi N, Thibault F, Decou R, Fleurat-Lessard P, Bere E, Costa G, and

Lhernould S. (2010) NaCl effect on the distribution of wall ingrowth polymers and arabinogalactan proteins in type A transfer cells of Medicago sativa GabSs leaves.

Protoplasma 242: 69-80.

31. Bouton S, Leboeuf E, Mouille G, Leydecker MT, Talbotec J, Granier F, Lahaye

M, Höfte H, Truong HN. (2002) QUASIMODO1 encodes a putative membrane-bound glycosyltransferase required for normal pectin synthesis and cell adhesion in

Arabidopsis. Plant Cell 14: 2577–90.

32. Bradley DJ, Kjellbom P, Lamb CJ. (1992) Elicitor- and wound induced oxidative cross-linking of a proline-rich plant cell wall protein: a novel, rapid defense response.

Cell 70: 21–30.

33. Breton C, Bettler E, Joziasse DH, Geremia RA, Imberty A. (1998) Sequence- function relationship of prokaryotic and eukaryotic galactosyltransferases. J Biochem

Tokyo 123: 1000–1009.

331

34. Brown DM, Goubet F, Vicky WWA, Goodacre R, Stephens E, Dupree P, Turner

SR. (2007) Comparison of five xylan synthesis mutants reveals new insight into the mechanisms of xylan synthesis. Plant J. 52: 1154–68.

35. Brown DM, Zhang ZN, Stephens E, Dupree P, Turner SR. (2009)

Characterization of IRX10 and IRX10- like reveals an essential role in glucuronoxylan biosynthesis in Arabidopsis. Plant J. 57: 732–46

36. Buckner B, Johal GS, Janick-Buckner D. (2000) Cell death in maize. Physiologia

Plantarum 108: 231– 239.

37. Burton RA, Shirley NJ, King BJ, Harvey AJ, Fincher GB (2004) The CesA gene family of barley. Quantitative analysis of transcripts reveals two groups of co-expressed genes. Plant Physiol. 134: 224–236.

38. Burton RA, Wilson SM, Hrmova M, Harvey AJ, Shirley NJ, Medhurst A, Stone

BA, Newbigin EJ, Bacic A, Fincher GB. (2006) Cellulose synthase-like CslF genes mediate the synthesis of cell wall (1,3;1,4)-beta-D-glucans. Science 311: 1940–1942.

39. Busk PK, Lange L. (2015) Classification of fungal and bacterial lytic polysaccharide monooxygenases. B.MC Genomics 16: 368.

40. Campbell JA, Davies GJ, Bulone V, Henrissat B. (1997) Biochem J. 326: 929–

939.

41. Cannon MC, Terneus K., Hall Q, Tan L, Wang Y, Wegenhart B L, Chen L,

Lamport DTA, Chen Y, Kieliszewski MJ. (2008) Self-assembly of the plant cell wall requires an extensin scaffold. Proc. Natl. Acad. Sci. U.S.A. 105: 2226–2231.

332

42. Can˜ o-Delgado A, Penfield S, Smith C, Catley M, and Bevan M. (2003)

Reduced cellulose synthesis invokes lignification and defense responses in Arabidopsis thaliana. Plant J. 34: 351–362.

43. Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, Henrissat B.

(2009) The Carbohydrate-Active EnZymes database (CAZy): an expert resource for glycogenomics. Nucleic Acids Res 37: D233–D238.

44. Carpita N. (1996) Structure and biogenesis of the cell walls of grasses. Annu Rev

Plant Mol Biol 47: 445–476.

45. Carpita N, Tierney M, Campbell M. (2001) Molecular biology of the plant cell wall: searching for the genes that define structure, architecture and dynamics. Plant Mol

Biol. 47: 1–5.

46. Carpita NC, McCann MC. (2010) The maize mixed-linkage (1,3), (1,4)-beta-D- glucan polysaccharide is synthesized at the Golgi membrane. Plant Physiol. 153: 1362–

1371.

47. Castilho A, Neumann L, Daskalova S, Mason SH, Steinkellner H, Altmann F,

Strasser, R. (2012) Engineering of sialylated mucin-type O-glycosylation in plants. J.

Biol. Chem. 287: 36518-36526.

48. Cavalier DM, Keegstra K. (2006) Two xyloglucan xylosyltransferases catalyze the addition of multiple xylosyl residues to cellohexaose. J. Biol. Chem. 281: 34197–

34207.

333

49. Chaves I, Regalado AP, Chen M, Ricardo CP, Showalter AM. (2002)

Programmed cell death induced by (β-D-galactosyl)3 Yariv reagent in Nicotiana tabacum BY-2 suspensioncultured cells. Physiol Plant. 116: 548-553.

50. Chen Z, Hong X, Zhang H, Wang Y, Li X, Zhu JK, Gong Z. (2005) Disruption of the cellulose synthase gene, AtCesA8/IRX1, enhances drought and osmotic stress tolerance in Arabidopsis. Plant J. 43: 273–283.

51. Cheung AY, Chen CY, Glaven RH, de Graaf BH, Vidali L, Hepler PK, Wu HM.

(2002) Rab2 GTPase regulates vesicle trafficking between the endoplasmic reticulum and the Golgi bodies and is important to pollen tube growth. Plant Cell 14: 945–962.

52. Clarke A, Gleeson P, Harrison S, Knox RB. (1979) Pollen-stigma interactions, identification and characterization of surface components with recognition potential.

Proc. Natl. Acad. Sci. USA. 76: 3358-3362.

53. Clausen H, Bennett EP. (1996) A family of UDP-GalNAc: polypeptide N- acetylgalactosaminyl-transferases control the initiation of mucin-type O-linked glycosylation. Glycobiology 6: 635–646.

54. Cocuron JC, Lerouxel O, Drakakaki G, Alonso AP, Liepman AH, Keegstra K,

Raikhel N,Wilkerson CG. (2007) A gene from the cellulose synthase-like C family encodes a b-1,4 glucan synthase. Proc Natl Acad Sci U S A 104: 8550–8555

55. Coimbra S, Costa ML, Mendes MA, Pereira A, Pinto J, Pereira LG. (2010) Early germination of Arabidopsispollen in a double null mutant for the arabinogalactan protein genes AGP6 and AGP11. Sexual Plant Reproduction. 23: 199–205.

334

56. Cosgrove DJ. (2005) Growth of the plant cell wall. Nat. Rev Mol Cell Biol. 6:

850-861.

57. Dardelle F, Lehner A, Ramdani Y, Bardor M, Lerouge P, Driouich A, Mollet

JC. (2010) Biochemical and immunocytological characterizations of Arabidopsis pollen tube cell wall. Plant Physiol. 153: 1563-1576.

58. Darley CP, Forrester AM, McQueen-Mason SJ. (2001) The molecular basis of plant cell wall extension. Plant Mol. Biol. 47: 179–195.

59. De Cnodder T, Vissenberg K, Van Der Straeten D, Verbelen JP. (2005)

Regulation of cell length in the Arabidopsis thaliana root by the ethylene precursor 1- aminocyclopropane-1-carboxylic acid: A matter of apoplastic reactions. New Phytol.

168: 541–550.

60. Decreux A., and Messiaen J. (2005) Wall-associated kinase WAK1 interacts with cell wall pectins in a calcium-induced conformation. Plant Cell Physiol. 46: 268–

278.

61. DeBolt S, Gutierrez R, Ehrhardt DW, Melo CV, Ross L, Cutler SR, Somerville

C, Bonetta D. (2007) Morlin, an inhibitor of cortical microtubule dynamics and cellulose synthase movement. Proc. Natl. Acad. Sci. USA. 104: 5854-5859.

62. Decreux A., Messiaen J. (2005) Wall-associated kinase WAK1 interacts with cell wall pectins in a calcium-induced conformation. Plant Cell Physiol. 46: 268–278.

63. Demesa-Arévalo E, Vielle-Calzada JP. (2013) The classical arabinogalactan protein AGP18 mediates megaspore selection in Arabidopsis. Plant Cell 25: 1274-1287.

335

64. Dereeper A, Guignon V, Blanc G, Audic S, Buffet S, Chevenet F, Dufayard JF,

Guindon S, Lefort V, Lescot M, Claverie JM, Gascuel O. (2008) Phylogeny.fr.Robust phylogenetic analysis for the non-specialist. Nucleic Acids Res. 36: W465–W469.

65. Dhugga KS, Barreiro R, Whitten B, Stecca K, Hazebroek J, Randhawa GS,

Dolan M, Kinney AJ, Tomes D, Nichols S, Anderson P. (2004) Guar seed beta-mannan synthase is a member of the cellulose synthase super gene family. Science 303: 363–

366.

66. Dilokpimol A, Poulsen CP, Vereb G, Kaneko S, Schulz A, Geshi N. (2014)

Galactosyltransferases from Arabidopsis thaliana in the biosynthesis of type II arabinogalactan: molecular interaction enhances enzyme activity. BMC Plant Biol. 14:

90.

67. Dilokpimol A, Geshi N. (2014) Arabidopsis thaliana glucuronosyltransferase in family GT14. Plant Signal Behav. PMCID: PMC4091549.

68. Ding L, Zhu J-K. (1997) A role for arabinogalactan-proteins in root epidermal cell expansion. Planta 203: 289-294.

69. Djerbi S, Lindskog M, Arvestad L, Sterky F, Teeri TT (2005) The genome sequence of black cottonwood (Populus trichocarpa) reveals 18 conserved cellulose synthase (CesA) genes. Planta 221: 739–746.

70. Doblin MS, Pettolino FA, Wilson SM, Campbell R, Burton RA, Fincher GB,

Newbigin E, Bacic A. (2009) A barley cellulose synthaselike CSLH gene mediates

(1,3;1,4)-beta-D-glucan synthesis in transgenic Arabidopsis. Proc. Natl. Acad. Sci. USA

106: 5996–6001.

336

71. Dodd RB, Drickamer K. (2001) Lectin-like proteins in model organisms: implications for evolution of carbohydrate-binding activity. Glycobiology 11: 71–79

72. dos Santos ALW, Wietholter N, Gueddari NE, Moerschbacher BM. (2006)

Protein expression during seed development in Araucaria angustifolia: transient accumulation of class IV chitinases and arabinogalactan proteins. Physiologia Plantarum

127: 138–148.

73. Driouich A, Baskin TI. (2008) Intercourse between cell wall and cytoplasm exemplified by arabinogalactan proteins and cortical microtubules. Am J Bot. 95: 1491-

1497.

74. Dubos C, Stracke R, Grotewold E, Weisshaar B, Martin C, Lepiniec L.

(2010) MYB transcription factors in Arabidopsis. Trends Plant Sci. 15: 573–581.

75. Ebringerova A, and Heinze T. (2000) Xylan and xylan derivatives–biopolymers with valuable properties, Naturally occurring xylans structures, isolation procedures and properties. Rapid Commun. 21: 542–556.

76. Egelund J, Ellis MA, Doblin M S, Qu Y, Bacic A. (2011) “Genes and enzymes of the GT31family: towards unraveling the function(s) of the plant glycosyltransferase family members,” in Plant Polysaccharides: Biosynthesis and Bioengineering, ed.

Oxford: Wiley-Blackwell River Street, Hoboken, NJ.

77. Egelund J, Damager I, Faber K, Olsen CE, Ulvskov P, Petersen BL. (2008)

Functional characterisation of a putative rhamnogalacturonan II specific xylosyltransferase. FEBS Lett. 582: 3217–3222.

337

78. Egelund J, Petersen BL, Motawia MS, Damager I, Faik A, Olsen CE, Ishii T,

Clausen H, Ulvskov P, Geshi N. (2006) Arabidopsis thaliana RGXT1 and RGXT2 encode Golgi-localized (1,3)-α-D-xylosyltransferases involved in the synthesis of pectic rhamnogalacturonan-II. Plant Cell 18: 2593–607.

79. Ellis C, Karafyllidis I, Wasternack C, Turner JG. (2002) The Arabidopsis mutant cev 1 links cell wall signalling to jasmonate and ethylene responses. Plant Cell 14:

1557–1566.

80. Ellis M, Egelund J, Schultz CJ, Bacic A. (2010) Arabinogalactan-proteins: Key regulators at the cell surface? Plant Physiol. 153: 403-419.

81. Emanuelsson O, Brunak S, von Heijne G, Nielsen H. (2007) Locating proteins in the cell using TargetP, SignalP and related tools. Nat Protoc. 2: 953–971.

82. Fagard M, Desnos T, Desprez T, Goubet F, Refregier G, Mouille G, McCann M,

Rayon C, Vernhettes S, Höfte H. (2000) PROCUSTE1 encodes a cellulose synthase required for normal cell elongation specifically in roots and dark-grown hypocotyls of

Arabidopsis. Plant Cell 12: 2409–2423.

83. Faik A, Price NJ, Raikhel NV, Keegstra K. (2002) An Arabidopsis gene encoding an α-xylosyltransferase involved in xyloglucan biosynthesis. Proc. Natl. Acad.

Sci. USA 99: 7797–802.

84. Fincher GB, Stone BA, Clarke AE. (1983) Arabinogalactan-proteins: structure, biosynthesis, and function. Annu. Rev. Plant Physiol. 34: 47–70.

85. Fry SC. (1989) The structure and functions of xyloglucan. J. Exp. Bot. 40: 1-11.

338

86. Galvan-Ampudia CS, Testerink C. (2011) Salt stress signals shape the plant root.

Curr Opin Plant Biol. 14: 296-302.

87. Garcia-Muniz N, Martinez-Izquierdo JA, Puigdomenech P. (1998) Induction of mRNA accumulation corresponding to a gene encoding a cell wall hydroxyproline-rich glycoprotein by fungal elicitors. Plant Mol. Biol. 38: 623–632.

88. Gardiner M, Chrispeels MJ. (1975) Involvement of the Golgi apparatus in the synthesis and secretion of Hydroxyproline-rich cell wall glycoproteins. Plant Physiol.

55: 536–541.

89. Gao M, Showalter AM. (1999) Yariv reagent treatment induces programmed cell death in arabidopsis cell cultures and implicates arabinogalactan protein involvement.

Plant J. 19: 321–331.

90. Gerken TA, Revoredo L, Thome JJC. (2013) The Lectin Domain of the

Polypeptide GalNAc Transferase Family of (ppGalNAc Ts) Acts as a switch directing glycopeptide substrate glycosylation in an N- or C-terminal direction, further controlling mucin type O-Glycosylation. J. Biol. Chem. 288: 19900-

19914.

91. Geshi N, Pauly M, Ulvskov P. (2002) Solubilization of galactosyltransferase that synthesizes 1, 4-β-galactan side chains in pectic rhamnogalacturonan I. Plant Physiol.

114: 540–548.

92. Geshi N, Johansen JN, Dilokpimol A, Rolland A, Belcram K, Verger S, Kotake

T, Tsumuraya Y, Kaneko S, Tryfona T, Dupree P, Scheller HV, Höfte H, Mouille G.

339

(2013) A galactosyltransferase acting on arabinogalactan protein glycans is essential for embryo development in Arabidopsis. Plant J. 76: 128–137.

93. Gibeaut DM, Carpita NC. (1993) Synthesis of (1–>3), (1–>4)-beta-D-glucan in the Golgi apparatus of maize coleoptiles. Proc Natl Acad Sci USA 90: 3850–3854.

94. Gille S, Sharma V, Baidoo EEK, Keasling JD, Scheller HV, Pauly M. (2013)

Arabinosylation of a Yariv-precipitable cell wall polymer impacts plant growth as exemplified by the Arabidopsis glycosyltransferase mutant ray1. Molecular Plant 6:

1369-1372.

95. Gish L, Clark S. (2011) The RLK/Pelle family of kinases. Plant J. 66: 117–127.

96. Golldack D, Li C, Mohan H, Probst N. (2014) Tolerance to drought and salt stress in plants: Unraveling the signaling networks, Front. Plant Sci. 5.

97. Goto M. (2007) Protein O-glycosylation in fungi: diverse structures and multiple functions. Biosci. Biotechnol. Biochem. 71: 1415–1427.

98. Goubet F, Misrahi A, Park SK, Zhang ZN, Twell D, Dupree P. (2003) AtCSLA7, a cellulose synthase-like putative glycosyltransferase, is important for pollen tube growth and embryogenesis in Arabidopsis. Plant Physiol. 131: 547–57.

99. Griffiths JS, Tsai AY, Xue H, Voiniciuc C, Sola K, Seifert G, Mansfield SD,

Haughn GW. (2014) SALT-OVERLY SENSITIVE5 mediates Arabidopsis seed coat mucilage adherence and organization through pectins. Plant Physiol. 165: 991-1004.

100. Guan Y, Nothnagel EA. (2004) Binding of arabinogalactan proteins by Yariv phenylglycoside triggers wound-like responses in arabidopsis cell cultures. Plant

Physiol. 135: 1346–1366.

340

101. Guo H, Li L, Ye H, Yu X, Algreen A, Yin Y. (2009) Three related receptor-like kinases are required for optimal cell elongation in Arabidopsis thaliana. Proc Natl Acad

Sci USA 106: 7648–7653.

102. Hamann T, Denness L. (2011) Cell wall integrity maintenance in plants:

Lessons to be learned from yeast? Plant Signal. Behav. 6: 1706-1709.

103. HamannT. (2015) The plant cell wall integrity maintenance mechanism–A case study of a cell wall plasma membrane signaling network. Phytochemistry 112, 100-109.

104. Hanke DE, Northcote DH. (1975) Molecular visualization of pectin and DNA by ruthenium red. Biopolymers 14: 1–17.

105. Hanisch FG. (2001) O-glycosylation of the mucin type. Biol Chem. 382: 143–

149.

106. Harholt J, Suttangkakul A, Scheller HV. (2010) Biosynthesis of pectins. Plant

Physiol. 153: 384–395.

107. Harholt J, Jensen JK, Verhertbruggen Y, Søgaard C, Bernard S, Nafisi M,

Poulsen CP, Geshi N, Sakuragi Y, Driouich A, Knox JP, Scheller HV. (2012) ARAD proteins associated with pectic Arabinan biosynthesis form complexes when transiently overexpressed in planta. Planta 236: 115–128.

108. Harpaz-Saad S, McFarlane HE, Xu S, Divi UK, Forward B, Western TL, Kieber

JJ. (2011) Cellulose synthesis via the FEI2 RLK/SOS5 pathway and CELLULOSE

SYNTHASE 5 is required for the structure of seed coat mucilage in Arabidopsis. Plant J.

68: 941-953.

341

109. Hassan H., Reis C A, Bennett EP. (2000) The lectin domain of UDP-N-acetyl-D- galactosamine: polypeptide N-acetylgalactosaminyltransferase-T4 directs its glycopeptide specificities. J. Biol. Chem. 275: 38197–38205.

110. He Z-H, Fujiki M, Kohorn BD. (1996) A cell wall-associated, receptor-like protein kinase. J Biol Chem. 271: 19789–19793.

111. Heim DR, Skomp JR, Tschabold EE, Larrinua I. (1990) Isoxaben inhibits the synthesis of acid-insoluble cell wall materials in Arabidopsis thaliana. Plant Physiol. 93:

695–700.

112. Hieta R, Myllyharj J. (2002) Cloning and characterization of a low molecular weight prolyl 4-hydroxylase from Arabidopsis thaliana. Effective hydroxylation of proline-rich, collagen-like, and hypoxiainducible transcription factor alpha-like peptides.

J. Biol. Chem. 277: 23965–23971.

113. Held MA, Be E, Zemelis S, Withers S, Wilkerson C, Brandizzi F. (2011) CGR3: a Golgi-localized protein influencing homogalacturonan methylesterification. Mol. Plant

4: 832–44

114. Hématy K, Höfte H. (2008) Novel receptor kinases involved in growth regulation. Curr Opin Plant Biol. 11: 321–328.

115. Henrissat B, and Davies GJ. (2000) Glycoside hydrolases and glycosyltransferases. Families, modules, and implications for genomics. Plant Physiol.

124: 1515–1519.

342

116. Hijazi M, Velasquez SM, Jamet E, Estevez JM, Albenne C. (2014) An update on post-translational modifications of hydroxyproline-rich glycoproteins: toward a model highlighting their contribution to plant cell wall architecture. Front Plant Sci. 5: 395.

117. Hoffman CS, Winston F. (1987) A ten-minute DNA preparation from yeast efficiently releases autonomous plasmids for transformation of Escherichia coli. Gene

(Amst.) 57: 267 –272.

118. Holland N, Holland D, Helentjaris T, Dhugga KS, Xoconostle-Cazares B,

Delmer DP. (2000) A comparative analysis of the plant cellulose synthase (CesA) gene family. Plant Physiol. 123: 1313–1324.

119. Huang J, DeBowles D, Esfandiari E, Dean G, Carpita NC, Haughn GW. (2011)

The Arabidopsis transcription factor LUH/MUM1 is required for extrusion of seed coat mucilage. Plant Physiol. 156: 491–502.

120. Humphrey TV, Bonetta DT, Goring DR. (2007) Sentinels at the wall: Cell wall receptors and sensors. New Phytol. 176: 7–21.

121. Immerzeel P, Eppink MM, De Vries SC, Schols HA, Voragen AGJ. (2006)

Carrot arabinogalactan proteins are interlinked with pectins. Physiologia Plantarum 128:

18–28.

122. Iraki NM, Singh N, Bressan RA, Carpita NC. (1989) Cell-walls of tobacco cells and changes in composition associated with reduced growth upon adaptation to water and saline stress. Plant Physiol. 91: 48–53.

343

123. Ito S, Suzuki Y, Miyamoto K, Ueda J, Yamaguchi I. (2005) AtFLA11, a fasciclin-like arabinogalactan-protein, specifically localized in sclerenchyma cells.

Bioscience, Biotechnology, and Biochemistry. 69: 1963-1969.

124. Jamet E, Albenne C, Boudart G, Irshad M., Canut H, and Pont Lezica R. (2008)

Recent advances in plant cell wall proteomics. Proteomics 8: 893–908.

125. Jauh GY, and Lord EM. (1996). Localizationof pectins and arabinogalactan protein in lily (Lilium longiflorum L.) pollen tube and style, and their possible roles in pollination. Planta 199: 251–261.

126. Jensen JK, Sørensen SO, Harholt J, Geshi N, Sakuragi Y, Møller I, Zandleven J,

Bernal AJ, Jensen NB, Sørensen C, Pauly M, Beldman G, Willats WGT, Scheller HV.

(2008) Identification of a xylogalacturonan xylosyltransferase involved in pectin biosynthesis in Arabidopsis. Plant Cell 20: 1289–1302.

127. Johnson KL, Jones BJ, Bacic A, Schultz CJ. (2003). The fasciclin-like arabinogalactan proteins of arabidopsis. A multigene family of putative cell adhesion molecules. Plant Physiol. 133: 1911–1925.

128. Johnson KL, Kibble NAJ, Bacic A, Schultz CJ. (2011) A Fasciclin-Like

Arabinogalactan-protein FLA mutant ofArabidopsis thaliana, fla1, shows defects in shoot regeneration. PLoS ONE 6: e25154.

129. Jia Q S, Zhu J, Xu XF, Lou Y, Zhang ZL, Zhang ZP, Yang ZY. (2014)

Arabidopsis AT-hook protein TEK positively regulates the expression of arabinogalactan proteins in controlling nexine layer formation in the pollen wall. Mol.

Plant 8: 251–260.

344

130. Kang JS, Frank J, Kang CH, Kajiura H, Vikram M, Ueda A, Kim S, Bahk JD,

Triplett B, Fujiyama K, Lee SY, von Schaewen A, Koiwa H. (2008) Salt tolerance of

Arabidopsis thaliana requires maturation of N-glycosylated proteins in the Golgi apparatus. Proc. Natl. Acad. Sci. USA 105: 5933–5938.

131. Kato H, Takeuchi Y, Tsumuraya Y, Hashimoto Y, Nakano H, Kovac P. (2003)

In vitro biosynthesis of by membrane-bound galactosyltransferase from radish

(Raphanus sativus L.) seedlings. Planta 217: 271–282.

132. Kelley LA, Sternberg MJ. (2009) Protein structure prediction on the Web: a case study using the Phyre server. Nat. Protoc. 4: 363–371.

133. Kieliszewski MJ, O’Neill M, Leykam J, Orlando R. (1995) Tandem mass- spectrometry and structural elucidation of glycopeptides from a hydroxyproline-rich plant-cell wall glycoprotein indicate that contiguous hydroxyproline residues are the major sites of hydroxyproline O-arabinosylation. J. Biol. Chem. 270: 2541–2549.

134. Kieliszewski MJ, Shpak E. (2001) Synthetic genes for the elucidation of glycosylation codes for arabinogalactan-proteins and other hydroxyproline-rich glycoproteins. Cell and Molecular Life Sciences 58: 1386–1398.

135. Kitazawa K, Tryfona T, Yoshimi Y, Hayashi Y, Kawauchi S, Antonov L,

Tanaka H, Takahashi T, Kaneko S, Dupree P, Tsumuraya Y, Kotake T. (2013) β - galactosyl yariv reagent binds to the β -1,3-galactan of arabinogalactan proteins. Plant

Physiol. 161: 1117-1126.

345

136. Kjellbom P, Snogerup L, Stöhr C, Reuzeau C, McCabe PF, Pennell RI. (1997)

Oxidative cross- linking of plasma membrane arabinogalactan proteins. Plant J.

12:1189–1196.

137. Knoch E, Dilokpimol A, Tryfona T, Poulsen CP, Xiong G, Harholt J, Petersen

BL, Ulvskov P, Hadi MZ, Kotake T, Tsumuraya Y, Pauly M, Dupree P, Geshi N. (2013)

A β–glucuronosyltransferase from Arabidopsis thaliana involved in biosynthesis of type

II arabinogalactan has a role in cell elongation during seedling growth. Plant J. 76:

1016–1029.

138. Knoch E Dilokpimol A Geshi N. (2014) Arabinogalactan proteins: focus on carbohydrate active enzymes. Front. Plant Sci. 11: 198.

139. Kohorn BD, Kobayashi M, Johansen S, Riese J, Huang L.F, Koch K, Fu S,

Dotson A, Byers NR. (2006) An Arabidopsis cell wall-associated kinase required for invertase activity and cell growth. Plant J. 46: 307–316.

140. Kohorn BD, Kohorn SL. (2012) The cell wall associated kinases, WAKs, as pectin receptors. Front. Plant Sci. 3: 88.

141. Koornneef M. (1981) The complex syndrome of the ttg mutants. Arabidopsis Inf

Serv. 18: 45–51

142. Krogh A, Larsson B, von Heijne G, Sonnhammer EL. (2001) Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 305: 567–580.

143. Lairson LL, Henrissat B, Davies GJ, Withers SG. (2008) Glycosyltransferases: structures, functions, and mechanisms. Annu Rev Biochem. 77: 521–555.

346

144. Lamport DT, Kieliszewski MJ, Showalter AM. (2006) Salt stress upregulates periplasmic arabinogalactan proteins: using salt stress to analyse AGP function. New

Phytol. 169: 479-492.

145. Langan KJ, Nothnagel EA. (1997) Cell surface arabinogalactan-proteins and their relation to cell proliferation and viability. Protoplasma 196: 87–98.

146. Lazzaro MD, Donohue JM, Soodavar FM. (2003) Disruption of cellulose synthesis by isoxaben causes tip swelling and disorganizes cortical microtubules in elongating conifer pollen tubes. Protoplasma 220: 201-207.

147. Levitin B, Richter D, Markovich I, Zik M. (2008) Arabinogalactan proteins 6 and 11 are required for stamen and pollen function in Arabidopsis. Plant J. 56: 351–363.

148. Le BH, Cheng C, Bui AQ, Wagmaister JA, Henry KF, Pelletier J, Kwong L,

Belmonte M, Kirkbride R, Horvath S, Drewsd GN, Fischer RL, Okamuro JK, Harada JJ,

Goldberg RB. (2010) Global analysis of gene activity during Arabidopsis seed development and identification of seed-specific transcription factors. Proc Natl Acad Sci

USA 107: 8063–8070.

149. Lee KJD, Sakata Y, Mao S-L, Pettolino F, Bacic A, Quatrano RS, Knight CD,

Knox JP. (2005) Arabinogalactan proteins are required for apical cell extension in the moss Physcomitrella patens. Plant Cell 17: 3051–3065.

150. Lee CH, Teng Q, Huang WL, Zhong RQ, Ye ZH. (2009) The F8H glycosyltransferase is a functional paralog of fra8 involved in glucuronoxylan biosynthesis in Arabidopsis. Plant Cell Physiol. 50: 8128–8127.

347

151. Lee CH, O’Neill MA, Tsumuraya Y, Darvill AG, Ye ZH. (2007) The irregular xylem9 mutant is deficient in xylan xylosyltransferase activity. Plant Cell Physiol. 48:

1624–1634.

152. Liepman AH, Wilkerson CG, Keegstra K. (2005) Expression of cellulose synthase-like (Csl) genes in insect cells reveals that CslA family members encode mannan synthases. Proc. Natl. Acad. Sci. USA 102: 2221–2226.

153. Liwanag AJM, Ebert B, Verhertbruggen Y, Rennie EA, Rautengarten C,

Oikawaa A, Andersend MCF, Clausend MH, Scheller HV. (2012) Pectin biosynthesis:

GALS1 in Arabidopsis thaliana is a β -1,4-galactan β -1,4-galactosyltransferase. Plant

Cell 24: 5024–5036.

154. Levy S, York WS, Stuikeprill R, Meyer B, Staehelin LA. (1991) Xyloglucan-the role of the fucosylated side-chain in surface-specific side-chain folding. Plant J. 1: 195–

215.

155. Liang Y, Faik A, Kieliszewski M, Tan L, Xu WL, Showalter AM. (2010)

Identification and characterization of in vitro galactosyltransferase activities involved in arabinogalactan-protein glycosylation in tobacco and Arabidopsis. Plant Physiol. 154:

632–642.

156. Li J, Yu MA, Geng LL, Zhao J. (2010) The fasciclin-like arabinogalactan protein gene, FLA3, is involved in microspore development of Arabidopsis. Plant J. 64: 482–

497.

348

157. Liu C, Mehdy M. (2007) A nonclassical arabinogalactan protein gene highly expressed in vascular tissues, AGP31, is transcriptionally repressed by methyl jasmonic acid in Arabidopsis. Plant Physiol. 145: 863–874.

158. Liu H, Shi R, Wang X, Pan Y, Li Z, Yang X, Zhang G, Ma Z. (2013)

Characterization and expression analysis of a fiber differentially expressed fasciclin-like arabinogalactan protein gene in Sea Island cotton fibers. PLoS One 8: e70185.

159. Liu XL, Liu L, Niu QK, Xia C, Yang KZ, Li R, Chen L-Q, Zhang X-Q, Zhou Y,

Ye D. (2011) MALE GAMETOPHYTE DEFECTIVE 4 encodes a rhamnogalacturonan

II xylosyltransferase and is important for growth of pollen tubes and roots in

Arabidopsis. Plant J. 65: 647–60

160. Lombard V, Ramulu H G, Drula E, Coutinho P M, Henrissat B. (2014) The carbohydrate-active enzymes database (CAZy) in Nucl. Acids Res. 42: D490-D495.

161. Lu H, Chen M, Showalter AM. (2001) Developmental expression and perturbation of arabinogalactan-proteins during seed germination and seedling growth in tomato. Physiologia Plantarum 112: 442–450.

162. Ma H, Zhao J. (2010) Genome-wide identification, classification, and expression analysis of the arabinogalactan protein gene family in rice (Oryza sativa L.). J. Exp. Bot.

61: 2647–2668.

163. Ma B, Yin C-C, He S-J, Lu X, Zhang W-K, Lu T-G, Chen S-L, Zhang J-S.

(2014) Ethylene-Induced Inhibition of Root Growth Requires Abscisic Acid Function in

Rice (Oryza sativa L.) Seedlings. PLoS Genet 10: e1004701.

349

164. MacMillan CP, Mansfield SD, Stachurski ZH, Evans R, Southerton SG (2010)

Fasciclin-like arabinogalactan proteins: specialization for stem biomechanics and cell wall architecture in Arabidopsis and Eucalyptus. Plant J. 62: 689–703.

165. Madson M, Dunand C, Li X, Verma R, Vanzin GF, Caplan J, Shoue DA, Carpita

NC, Reiter WD. (2003) The MUR3 gene of Arabidopsis encodes a xyloglucan galactosyltransferase that is evolutionarily related to animal exostosin. Plant Cell 15:

1662–1670.

166. Majewska-Sawka A, Nothnagel EA. (2000) The multiple roles of arabinogalactan proteins in plant development. Plant Physiol. 122: 3–10.

167. Mendu V, Griffiths JS, Persson S, Stork J, Downie AB, Voiniciuc C, Haughn

GW, DeBolt S. (2011) Subfunctionalization of cellulose synthases in seed coat epidermal cells mediates secondary radial wall synthesis and mucilage attachment. Plant

Physiol. 157: 441–453.

168. Miao Y, Li HY, Shen J, Wang J, Jiang L. (2011) QUASIMODO 3 (QUA3) is a putative homogalacturonan methyltransferase regulating cell wall biosynthesis in

Arabidopsis suspension-cultured cells. J. Exp. Bot. 62: 5063–78

169. Molhoj M, Verma R., Reiter WD. (2004) The biosynthesis of D-galacturonate in plants. Functional cloning and characterization of a membrane-anchored UDP-D- glucuronate 4-epimerase from Arabidopsis. Plant Physiol. 135: 1221-1230.

170. Mollet JC, Kim S, Jauh GY, Lord EM. (2002) Arabinogalactan proteins pollen tube growth, and the reversible effects of Yariv phenylglycoside. Protoplasma 219: 89–

98.

350

171. Moreau C, Aksenov N, Lorenzo M, Segerman B, Funk C, Nilsson P, Jansson S,

Tuominen H. (2005) A genomic approach to investigate developmental cell death in woody tissues of Populus trees. Genome Biol. 6: R34.

172. Motose H, Sugiyama M, Fukuda H. (2004) A mediates inductive interaction during plant vascular development. Nature 429: 873–878.

173. Mouille G, Ralet MC, Cavelier C, Eland C, Effroy D, Hématy K, McCartney L,

Truong HN, Gaudon V, Thibault JF, Marchant A, Höfte H. (2007) Homogalacturonan synthesis in Arabidopsis thaliana requires a Golgi-localized protein with a putative methyltransferase domain. Plant J. 50: 605–14.

174. Mutwil M, Obro J, Willats WGT, Persson S (2008) GeneCAT: Novel webtools that combine BLAST and co-expression analyses. Nucleic Acids Res. 36: W320–326.

175. Mutwil M, Klie S, Tohge T, Giorgi FM, Wilkins O, Campbell MM, Fernie AR,

Usadel B, Nikoloski Z, Persson S. (2011) PlaNet: combined sequence and expression comparisons across plant networks derived from seven species. The Plant Cell 23: 895–

910.

176. Newman AM, Cooper JB. (2011) Global Analysis of Proline-Rich Tandem

Repeat Proteins Reveals Broad Phylogenetic Diversity in Plant Secretomes. PLoS ONE

6: e23167.

177. Nikolovski N, Rubtsov D, Segura MP, Miles GP, Stevens TJ, Dunkley TP,

Munro S, Lilley KS, Dupree P. (2012) Putative glycosyltransferases and other plant

Golgi apparatus proteins are revealed by LOPIT proteomics. Plant Physiol. 160: 1037-

1051.

351

178. Nguema-Ona E, Vicré-Gibouin M, Cannesan MA, Driouich A. (2013)

Arabinogalactan proteins in root–microbe interactions. Trends Plant Sci. 18: 440–449.

179. Nothnagel EA. (1997) Proteoglycans and related components in plant cells. Int.

Rev. Cytol. 174: 195–291.

180. Ogawa-Ohnishi M, Matsubayashi Y. (2015) Identification of three potent hydroxyproline O-galactosyltransferases in Arabidopsis. Plant J. 81: 736–746.

181. Oka T, Saito F, Shimma Y, Yoko-o T, Nomura Y, Matsuoka K., Jigami Y.

(2010) Characterization of endoplasmic reticulum-localized UDP-D-galactose.

Hydroxyproline O-galactosyltransferase using synthetic peptide substrates in

Arabidopsis. Plant Physiol. 152: 332–340.

182. O'Neill M A, Eberhard S, Albersheim P, Darvill AG. (2001) Requirement of borate cross-linking of cell wall rhamnogalacturonan II for Arabidopsis growth. Science

294: 846-849.

183. Pagant S, Bichet A, Sugimoto K, Lerouxel O, Desprez T, McCann M, Lerouge

P, Vernhettes S, Höfte H. (2002) KOBITO1 Encodes a Novel Plasma Membrane

Protein Necessary for Normal Synthesis of Cellulose during Cell Expansion in

Arabidopsis. Plant Cell 14: 2001-2013.

184. Paredez AR, Somerville CR, Ehrhardt DW. (2006) Visualization of cellulose synthase demonstrates functional association with microtubules. Science 312: 1491-

1495.

352

185. Pedersen JW, Bennett EP., Schjoldager KTBG, Sjoberg A, Levery SB, Meldal

M, Clausen H, Wandall HH. (2011) Lectin domains of polypeptide GalNAc-Ts exhibit glycopeptide binding specificity. J. Biol. Chem. 286: 32684–32696.

186. Pena MJ, Zhong R, Zhou GK, Richardson EA, O’Neill MA, Darvill AG, York

WS, Ye ZH. (2007) Arabidopsis irregular xylem8 and irregular xylem9: Implications for the complexity of glucuronoxylan biosynthesis. Plant Cell 19: 549-563.

187. Penfield S, Meissner RC, Shoue DA, Carpita NC, Bevan MW. (2001) MYB61 is required for mucilage deposition and extrusion in the Arabidopsis seed coat. Plant Cell

13: 2777-2791.

188. Peng LC, Hocart CH, Redmond JW, Williamson RE. (2000) Fractionation of carbohydrates in Arabidopsis root cell walls shows that three radial swelling loci are specifically involved in cellulose production. Planta 211: 406–414.

189. Penning BW, Hunter CT, Tayengwa R, Eveland AL, Dugard CK, Olek AT,

Vermerris W, Koch KE, McCarty DR, Davis MF, Thomas SR, McCann MC, Carpita

NC. (2009) Genetic resources for maize cell wall biology. Plant Physiol. 151: 1703–

1728.

190. Pennell RI, Janniche L, Kjellbom P, Scofield GN, Peart JM, Roberts K. (1991).

Developmental regulation of a plasma membrane arabinogalactan protein epitope in oilseed rape flowers. Plant Cell 3: 1317–1326.

191. Perrin RM, DeRocher AE, Bar-Peled M, Zeng W, Norambuena L, Orellana A,

Raikhel NV, Keegstra K. (1999) Xyloglucan fucosyltransferase, an enzyme involved in plant cell wall biosynthesis. Science 284: 1976–1979.

353

192. Persson S, Raab T. (2004) Toward a systems approach to understanding plant cell walls. Science 306: 2206–2211.

193. Persson S, Paredez A, Carroll A, Palsdottir H, Doblin M, Poindexter P, Khitrov

N, Auer M, Somerville CR (2007) Genetic evidence for three unique components in primary cell-wall cellulose synthase complexes in Arabidopsis. Proc Natl Acad Sci

USA. 104: 15566–15571.

194. Pereira LG, Coimbra S, Oliveira H, Monteiro L, Sottomayor M. (2006)

Expression of arabinogalactan protein genes in pollen tubes of Arabidopsis thaliana.

Planta 223: 374–380.

195. Persson S, Paredez A, Carroll A, Palsdottir H, Doblin M, Poindexter P, Khitrov

N, Auer M, Somerville CR. (2007) Genetic evidence for three unique components in primary cell-wall cellulose synthase complexes in Arabidopsis. Proc. Natl. Acad. Sci.

USA 104: 15566–15571.

196. Peugnet I, Goubet F, Bruyant-Vannier M P, Thoiron B, Morvan C, Schols H A,

Voragen AGJ. (2001) Solubilization of rhamnogalacturonan I galactosyltransferases from membranes of a flax cell suspension. Planta 213: 435–445.

197. Pirovano W, Feenstra KA, Heringa J. (2008) PRALINE™: a strategy for improved multiple alignment of transmembrane proteins. Bioinformatics 24: 492-497.

198. Poulsen CP, Dilokpimol A, Mouille G, Burow M, Geshi N. (2014)

Arabinogalactan glycosyltransferases target to a unique subcellular compartment that may function in unconventional secretion in plants. Traffic doi: 10.1111/tra.12203.

354

199. Qu Y, Egelund J, Gilson PR,, Houghton F, Gleeson P A, Schultz CJ, Bacic A.

(2008) Identification of a novel group of putative Arabidopsis thaliana β-(1,3)-galactosyl transferases. Plant Mol. Biol. 68: 43–59.

200. Rautengarten C, Usadel B, Neumetzler L, Hartmann J, Büssis D, Altmann T.

(2008) A subtilisin-like serine protease essential for mucilage release from Arabidopsis seed coats. Plant J. 54: 466-480.

201. Rencoret J, Gutiérrez A, Nieto L, Jiménez-Barbero J, Faulds C B, Kim H, Ralph

J, Martínez ÁT, Río del JC. (2011) Lignin Composition and Structure in Young versus

Adult Eucalyptus globulus Plants. Plant Physiology 155: 667-682.

202. Refregier G, Pelletier S, Jaillard D, Hofte H. (2004) Interaction between Wall

Deposition and Cell Elongation in Dark-Grown Hypocotyl Cells in Arabidopsis. Plant

Physiol. 135: 959-968.

203. Richmond TA, Somerville CR. (2000) The cellulose synthase superfamily. Plant

Physiol. 124: 495–498.

204. Robinson DG, Glas R. (1982) Secretion kinetics of hydroxyproline-containing macromolecules in carrot root discs. Plant Cell Rep. 5: 197-198.

205. Roy A, Kucukural A, Zhang Y. (2010) I-TASSER A unified platform for automated protein structure and function prediction. Nat. Protoc. 5: 725–738.

206. Saint-Jore CM, Evins J, Batoko H, Brandizzi F, Moore I, Hawes C. (2002)

Redistribution of membrane proteins between the Golgi apparatus and endoplasmic reticulum in plants is reversible and not dependent on cytoskeletal networks. Plant J. 29:

661-678.

355

207. Saito F, Suyama A, Oka T, Yoko-O T, Matsuoka K, Jigami Y, Shimma YI.

(2014) Identification of novel peptidyl serine -galactosyltransferase gene family in plants. J Biol Chem 289: 20405–20420

208. Sardar HS, Yang J, Showalter AM. (2006) Molecular interactions of arabinogalactan proteins with cortical microtubules and F-actin in Bright Yellow-2 tobacco cultured cells. Plant Physiol. 142: 1469–1479.

209. Sarkar P, Bosneaga E, Auer M. (2009) Plant cell walls throughout evolution: towards a molecular understanding of their design principles. J. Exp. Bot. 60: 3615-

3635.

210. Sauer M, Paciorek T, Benkova E, Friml J. (2006) Immunocytochemical techniques for whole mount in situ protein localization in plants. Nat. Protoc. 1: 98-

103.

211. Schilmiller AL, Koo AJ, Howe GA. (2007) Functional diversification of acyl- coenzyme A oxidases in jasmonic acid biosynthesis and action. Plant Physiol. 143: 812–

824.

212. Schindler T, Bergfeld R, Schopfer P. (1995) Arabinogalactan proteins in maize coleoptiles: developmental relationship to cell death during xylem differentiation but not to extension growth.Plant J 7: 25–36

213. Schmittgen TD, Livak KJ. (2008). Analyzing real-time PCR data by the comparative C (T) method. Nat. Protoc. 3: 1101–1108.

214. Schultz JC, Kim LJ, Currie G, Bacic A. (2000) The classical arabinogalactan protein gene family of Arabidopsis. Plant Cell 12: 1751-1768.

356

215. Schmittgen TD, Livak KJ. (2008) Analyzing real-time PCR data by the comparative CT method. Nat Protoc 3: 1101–1108.

216. Schwacke R, Schneider A, Van Der Graaff E, Fischer K, Catoni E, Desimone M,

Frommer WB, Flügge UI, Kunze R. (2003) ARAMEMNON, a novel database for

Arabidopsis integral membrane proteins. Plant Physiol. 131: 16-26.

217. Seifert GJ, Roberts K. (2007) The biology of arabinogalactan proteins. Annu.

Rev. Plant Biol. 58: 137–161.

218. Seifert GJ, Xue H, Acet T. (2014) The Arabidopsis thaliana FASCICLIN LIKE

ARABINOGALACTAN PROTEIN4 gene acts synergistically with abscisic acid signaling to control root growth. Ann Bot; PMID: 24603604.

219. Serpe MD, Nothnagel EA. (1994) Effects of Yariv phenylglycosides on Rosa cell suspensions: Evidence for the involvement of arabinogalactan-proteins in cell proliferation. Planta 193: 542–550.

220. Shi H, Kim Y, Guo Y, Stevenson B, Zhu J-K. (2003) The Arabidopsis SOS5 locus encodes a putative cell surface adhesion protein and is required for normal cell expansion. Plant Cell 15: 19–32.

221. Shpak E, Barbar E, Leykam JF, Kieliszewski MJ. (2001) Contiguous hydroxyproline residues direct hydroxyproline arabinosylation in Nicotiana tabacum. J.

Biol. Chem. 276: 11272-11278.

222. Showalter AM (1993) Structure and function of plant cell wall proteins. Plant

Cell 5: 9-23.

357

223. Showalter AM (2001) Arabinogalactan-proteins: structure, expression, and function. Cellular and Molecular Life Sciences 58: 1399-1417.

224. Showalter AM, Keppler B, Lichtenberg J, Gu D, Welch LR. (2010) A bioinformatics approach to the identification, classification, and analysis of hydroxyproline-rich glycoproteins. Plant Physiol. 153: 485–513.

225. Somerville C, Bauer S, Brininstool G, Facette M, Hamann T, Milne J, Osborne E,

Paredez A, Persson S, Raab T. (2004) Toward a systems approach to understanding plant cell walls. Science 306: 2206–2211.

226. Speranza A, Taddei AR, Gambellini G, Ovidi E, Scoccianti V. (2009) The cell wall of kiwifruit pollen tubes is a target for chromium toxicity: alterations to morphology, callose pattern and arabinogalactan protein distribution. Plant Biol. 11:

179–193.

227. Steinwand BJ, Kieber JJ. (2010) The Role of Receptor-Like Kinases in

Regulating Cell Wall Function. Plant Physiol. 153: 479-484.

228. Steinwand BJ, Xu S, Polko JK, Doctor SM, Westafer M, Keiber JJ. (2014)

Alterations in auxin homeostasis suppress defects in cell wall function. PLoSONE 9: e98193.

229. Sterling JD, Atmodjo MA, Inwood SE, Kumar Kolli VS, Quigley HF, Hahn MG,

Mohnen D. (2006) Functional identification of an Arabidopsis pectin biosynthetic homogalacturonan galacturonosyltransferase. Proc. Natl. Acad. Sci. USA 103:5236–41.

230. Stork J, Harris D, Griffiths J, Williams B, Beisson F, Li-Beisson Y, Mendu V,

Haughn G, DeBolt S. (2010) Cellulose synthase9 serves a nonredundant role in

358 secondary cell wall synthesis in Arabidopsis epidermal testa cells. Plant Physiol. 153:

580–589.

231. Strahm A, Amado R, Neukom H. (1981) Hydroxyproline-galactoside as a protein polysaccharide linkage in a water soluble arabinogalactan-peptide from wheat endosperm. Phytochemistry 20: 1061-1063.

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

Stadlmann J, Altmann F, Steinkellner H, Mach L. (2007) A unique 1,3- galactosyltransferase is indispensable for the biosynthesis of N-glycans containing

Lewis a structures in Arabidopsis thaliana. Plant Cell 19: 2278–2292.

233. Sullivan S, Ralet MC, Berger A, Diatloff E, Bischoff V, Gonneau M, Marion-

Poll A, North HM (2011) CESA5 is required for the synthesis of cellulose with a role in structuring the adherent mucilage of Arabidopsis seeds. Plant Physiol. 156: 1725–1739.

234. Sun W, Zhao ZD, Hare MC, Kieliszewski MJ, Showalter AM. (2004) Tomato

LeAGP-1 is a plasma membrane-bound, glycosylphophatidylinositol-anchored arabinogalactan-protein. Plant Physiol. 120: 319–327.

235. Swarup R, Kramer EM, Perry P, Knox K, Leyser HM, Haseloff J, Beemster GT,

Bhalerao R, Bennett MJ. (2005) Root gravitropism requires lateral root cap and epidermal cells for transport and response to a mobile auxin signal. Nat. Cell Biol. 7:

1057–1065.

236. Tan L, Qiu F, Lamport DT, Kieliszewski MJ. (2004) Structure of a hydroxyproline (Hyp)-arabinogalactan polysaccharide from repetitive Ala-Hyp expressed in transgenic Nicotiana tabacum. J Biol. Chem. 279: 13156-13165.

359

237. Tan L, Varnai P, Lamport DTA, Yuan C, Xu J, Qiu F, Kieliszewski MJ. (2010)

Plant O-glycosylation are composed of repeating trigalactosyl subunits with short bifurcated side chains. J. Biol. Chem. 285: 24575–24583.

238. Tan L, Showalter AM, Egelund J, Hernandez-Sanchez A, Doblin MS, Bacic A.

(2012) Arabinogalactan-proteins and the research challenges for these enigmatic plant cell surface proteoglycans. Front Plant Sci. 3: 1-10.

239. Tan L, Eberhard S, Pattathil S, Warder C, Glushka J, Yuan C, Hao Z, Zhu X,

Avci U, Miller JS, Baldwin D, Pham C, Orlando R, Darvill A, Hahn MG, Kieliszewski

MJ, Mohnen D. (2013) An Arabidopsis cell wall proteoglycan consists of pectin and arabinoxylan covalently linked to an arabinogalactan protein. Plant Cell 25: 270–287.

240. Taylor NG, Scheible WR, Cutler S, Somerville CR, Turner SR. (1999) The irregular xylem 3 locus of Arabidopsis encodes a cellulose synthase required for secondary cell wall synthesis. Plant Cell 11: 769–780.

241. Taylor NG, Laurie S, Turner SR. (2000) Multiple cellulose synthase catalytic subunits are required for cellulose synthesis in Arabidopsis. Plant Cell 12, 2529-2540.

242. Tryfona T, Liang H-C, Kotake T, Kaneko S, Marsh J, Ichinose H, Lovegrove A,

Tsumuraya Y, Shewry PR, Stephens E, Dupree P. (2010) Carbohydrate structural analysis of wheat flour arabinogalactan protein. Carbohydrate Res. 345: 2648–2656.

243. Tryfona T, Liang HC, Kotake T, Tsumuraya Y, Stephens E, Dupree P. (2012)

Structural characterisation of Arabidopsis leaf arabinogalactan polysaccharides. Plant

Physiol. 160: 653-666.

360

244. Tryfona T, Theys TE, Wagner T, Stott K, Keegstra K, Dupree P. (2014)

Characterisation of FUT4 and FUT6 α-(1→2)-fucosyltransferases reveals that absence of root arabinogalactan fucosylation increases Arabidopsis root growth salt sensitivity.

PLoS ONE 9: e93291.

245. Tsumuraya Y, Ogura K, Hashimoto Y, Mukoyama H, Yamamoto S. (1988)

Arabinogalactan-proteins from primary and mature roots of radish (Raphanus sativus

L.). Plant Physiol. 86: 155–160.

246. Updegraff DM. (1969) Semimicro determination of cellulose in biological materials. Anal Biochem. 32: 420–424

247. van Buuren ML, Maldonado-Mendoza IE, Trieu AT, Blaylock LA, Harrison MJ.

(1999) Novel genes induced during an arbuscular mycorrhizal (AM) symbiosis formed between Medicago truncatula and Glomus versiforme. Molecular Plant-Microbe

Interactions 12: 171–181.

248. Vanholme R, Van Acker R, Boerjan W. (2010) Potential of Arabidopsis systems biology to advance the biofuel field. Trends Biotechnol. 28: 543–547.

249. van Hengel AJ, Roberts K. (2002) Fucosylated arabinogalactan-proteins are required for full root cell elongation in Arabidopsis. Plant J. 32: 105-113.

250. Vanzin GF, Madson M, Carpita NC, Raikhel NV, Keegstra K, Dieter Reiter W.

(2002) The mur2 mutant ofArabidopsis thaliana lacks fucosylated xyloglucan because of a lesion in fucosyltransferase AtFUT1. Proc. Natl. Acad. Sci. USA 99: 3340–45.

251. Velasquez SM, Ricardi MM, Dorosz JG, Fernandez PV, Nadra AD, Pol-Fachin

L, Egelund J, Gille S, Harholt J, Ciancia M, Verli H, Pauly M, Bacic A, Olsen CE,

361

Ulvskov P, Petersen BL, Somerville C, Iusem ND, Estevez JM. (2011) O-glycosylated cell wall proteins are essential in root hair growth. Science 332: 1401-1403.

252. Vlad F, Spano T, Vlad D, Daher FB, Ouelhadj A, Fragkostefanakis S, Kalaitzis

P. (2007) Involvement of Arabidopsis prolyl 4 hydroxylases in hypoxia, anoxia and mechanical wounding. Plant Signal. Behav. 2: 368–369 10.

253. Wandall HH, Irazoqui F, Tarp MA, Bennett E P, Mandel U, Takeuchi H, Kato

K, Irimura T, Suryanarayanan G, Hollingsworth MA. (2007) The lectin domains of polypeptide GalNAc-transferases exhibit carbohydrate binding specificity for GalNAc:

Lectin binding to GalNAc-glycopeptide substrates is required for high density GalNAc-

O-glycosylation. Glycobiology 17: 374–387.

254. Wang Y, Mortimer JC, Davis J, Dupree P, Keegstra K. (2013) Identification of an additional protein involved in mannan biosynthesis. Plant J. 73: 105–117.

255. Western TL, Skinner DJ, Haughn GW. (2000) Differentiation of mucilage secretory cells of the Arabidopsis seed coat. Plant Physiol 122: 345–356

256. Wightman R, Turner S. (2010) Trafficking of the plant cellulose synthase complex. Plant Physiol. 153: 427–432.

257. Willats GTW, Knox JP. (1996) A role for arabinogalactan-proteins in plant cell expansion: evidence from studies on the interaction of β -glucosyl Yariv reagent with seedlings of Arabidopsis thaliana. Plant J. 9: 919–925.

258. Willats WGT, McCartney L, Knox JP. (2001) In-situ analysis of pectic polysaccharides in seed mucilage and at the root surface of Arabidopsis thaliana. Planta

213: 37-44.

362

259. Winter D, Vinegar B, Nahal H, Ammar R, Wilson GV, Provart NJ. (2007) An

'electronic fluorescent pictograph' browser for exploring and analyzing large-scale biological data sets. PLoS One 2: e718.

260. Wolf S, Hématy K, Höfte H. (2012) Growth control and cell wall signaling in plants. Annu Rev Plant Biol. 63: 381–407.

261. Wu Y, Williams M, Bernard S, Driouich A, Showalter AM, Faik A. (2010)

Functional identification of two nonredundant Arabidopsis (1,2) fucosyltransferases specific to arabinogalactan proteins. J. Biol. Chem. 285, 13638–13645.

262. Wycoff KL, Powell PA, Gonzales RA, Corbin DR, Dixon RA. (1995) Stress activation of a bean hydroxyproline-rich glycoprotein promoter is superimposed on a pattern of tissue-specific developmental expression. Plant Physiol. 109: 41-52.

263. Xiao S, Dai LY, Liu FQ, Wang ZL, Peng W, Xie DX. (2004) COS1:

An Arabidopsis coronatine insensitive1 suppressor essential for regulation of jasmonate- mediated plant senescence and defense. Plant Cell 16: 1132–1142.

264. Xu SL, Rahman A, Baskin TI, Kieber JJ. (2008) Two leucine-rich repeat receptor kinases mediate signaling, linking cell wall biosynthesis and ACC synthase in

Arabidopsis. Plant Cell 20: 3065–3079.

265. Xue H, Seifert GJ. (2015) FASCICLIN LIKE ARABINOGALACTAN

PROTEIN 4 and RESPIRATORY BURST OXIDASE HOMOLOG D and F independently modulate abscisic acid signaling, Plant Signal. Behav. 10: e989064.

363

266. Yang J, Sardar HS, McGovern KR, Zhang YZ, Showalter AM. (2007) A lysine- rich arabinogalactan protein in Arabidopsis is essential for plant growth and development, including cell division and expansion. Plant J. 49: 629–640

267. Yang Z, Drew DP, Jørgensen B, Mandel U, Bach SS, Ulvskov P, Levery SB,

Bennett EP, Clausen H, Petersen BL. (2012) Engineering mammalian mucin-type O- glycosylation in plants. J. Biol. Chem. 287: 11911–11923.

268. Yang SF, Hoffman NE. (1984) Ethylene biosynthesis and its regulation in higher plants. Annu. Rev. Plant Physiol. 35: 155–189.

269. Yariv J, Lis H, Katchalski E. (1967) Precipitation of arabic acid and some seed polysaccharides by glycosylphenylazo dyes. J. Biochem. 105: 1C–2C.

270. York WS, O’Neill MA. (2008) Biochemical control of xylan biosynthesis-which end is up? Curr. Opin. Plant Biol. 11: 258–65

271. Youl JJ, Bacic A, Oxley D. (1998) Arabinogalactan-proteins from Nicotiana alata and Pyrus communis contain glycosylphosphatidylinositol membrane anchors.

Proc Natl Acad Sci USA 95: 7921–7926.

272. Yu L, Shi D, Li J, Kong Y, Yu U, Chai G, Hu R, Wang J, Hahn MG, Zhou G.

(2014) CELLULOSE SYNTHASE-LIKEA2, a glucomannans synthase, is involved in maintaining adherent mucilage structure in Arabidopsis seed. Plant Physiol. 164: 1842-

56.

273. Zagorchev L, Kamenova P, Odjakova M. (2014) The role of plant cell wall proteins in response to salt stress. The Scientific World Journal, vol. 2014.

364

274. Zhao Q, Wang H, Yin Y, Xu Y, Chen F, Dixon RA. (2010) Syringyl lignin biosynthesis is directly regulated by a secondary cell wall master switch. Proc. Natl.

Acad. Sci. U.S.A. 107: 14496–14501.

275. Zhan X, Wang B, Li H, Liu R, Kalia RK, Zhu JK, Chinnusamy V. (2012)

Arabidopsis proline-rich protein important for development and abiotic stress tolerance is involved in microRNA biogenesis. Proc. Natl. Acad. Sci. U.S.A.109: 18198-18203.

276. Zhang Y. (2008) I-TASSER server for protein 3D structure prediction. BMC

Bioinformatics 9: 40.

277. Zhang Y, Yang J, Showalter AM. (2011a) AtAGP18, a lysine-rich arabinogalactan protein in Arabidopsis thaliana, functions in plant growth and development as a putative co-receptor for signal transduction. Plant Signal. Behav. 6:

855-857.

278. Zhang YZ, Yang J, Showalter AM. (2011b) AtAGP18 is localized at the plasma membrane and functions in plant growth and development. Planta 233: 675-683.

279. Zhong R, Ye Z H. (2012) MYB46 and MYB83 bind to the SMRE sites and directly activate a suit of transcription factors and secondary wall biosynthetic genes. Plant Cell Physiol. 53: 368–380.

280. Zhong R, Lee C, Zhou J, McCarthy R L, Ye Z H. (2008) A battery of transcription factors involved in the regulation of secondary cell wall biosynthesis in Arabidopsis. Plant Cell 20: 2763–2782.

281. Zhu J-K, Liu J, Xiong L. (1998) Genetic analysis of salt tolerance in Arabidopsis evidence for a critical role of potassium nutrition. Plant Cell 10: 1181–1191.

365

282. Zimmermann P, Hirsch-Hoffmann M, Hennig L, Gruissem W. (2004)

GENEVESTIGATOR. Arabidopsis microarray database and analysis toolbox. Plant

Physiol. 136: 2621–2632.

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

Thesis and Dissertation Services ! !