E.
.y.«..w.,..§. .
I!“ A .Q
. .4. 3.4%?
t

u... z 3.4
‘ .
mg

.
.

.
.

Jain ‘
3a 3 . a...
gnarwmmfismfiwmfi

. i.“ ‘fiunftfi

l
, 15

h .313"
Pin-12.9.
1.. .v .(z.

y

.. 35.:
.1... . .

‘ «'1...

2...: I :x

.nrsxi

tit-3|

ii

'2 v
4 ‘ ,1
VJ». .n...u.$.. a.
53.111 in?
z
y . 6
.. .wv‘... l
.‘ II

 

3‘3;- 2}. 1
.110}...
i:

\Q.‘

A. .1
122‘!!! i
.9. .3: ‘i‘ 5

vs.)

.7»

‘ , n:)u!..l

1. ‘.v4 1... .-.|. l
.v $2. .

‘ I... v. .

0 ‘I ‘Au?ra
v. ¢.!! 1

1| V .

1! . 1
:v...r!..,

 

 

VHE'BIS

1

lIBRARY
Michigan State
University

This is to certify that the

dissertation entitled

XYLOGLUCAN FUCOSYLTRANSFERASE,
A PLANT CELL WALL BIOSYNTHETIC ENZYME

presented by

Robyn Michele Perrin

has been accepted towards fulfillment
of the requirements for

Ph.D. degreein Botan; and Plant Pathology

 

ijMVOIJJ .Qfiagagzk

Major professor I

 

Date lineman]

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6101 cJCIRCIDateDuepes-p. 15

XYLOGLUCAN F UCOSYLTRANSFERASE, A PLANT CELL WALL
BIOSYNTHETIC ENZYME

by

Robyn Michele Perrin

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Botany and Plant Pathology Department

2001

ABSTRACT

XYLOGLUCAN FUCOSYLTRANSFERASE, A PLANT CELL WALL
BIOSYNTHETIC ENZYME

by
Robyn Michele Perrin
Cell walls are major determinants of plant growth, morphology, development, and
interactions with the environment. Components of the plant cell wall include
polysaccharides such as cellulose, pectins, and hemicelluloses as well as structural
proteins. Although much is known about the architecture of the plant cell wall, far less is
known about the biosynthesis of its structural components. The focus of this project is
the biosynthesis of a hemicellulosic polysaccharide component of the plant primary cell
wall, xyloglucan. Xyloglucan is comprised of a beta-(1,4)-glucan backbone decorated
with side chains consisting of xylose, xylose and galactose, or xylose, galactose and
fucose. An enzyme that adds the terminal fucose to galactose on xyloglucan in an alpha-
(1,2)-linkage, xyloglucan fucosyltransferase (XyG FUTase), was purified from rapidly
elongating segments of etiolated pea epicotyls. Amino acid sequence data from this
purified enzyme allowed the identification of an expressed sequence tagged cDNA (EST)
encoding an Arabidopsis XyG FUTase. A full-length cDNA was obtained for this
Arabidopsis gene and named Arabidopsis F ucosyltransflrase 1 (A W UT 1 ). The
biochemical function of AtF UT 1 was confirmed by 1) using antibodies against the protein
encoded by this cDNA to immunoprecipitate XyG FUTase activity fi'om Arabidopsis
microsomal proteins, and 2) expressing AtF UT 1 in mammalian Cos-7 cells, thereby

conferring XyG FUTase activity.

The regulation of XyG fucosylation was studied in Arabidopsis as a model cell wall
biosynthetic process. The expression of AtFUTI, the distribution of XyG FUTase
activity, and the fucosylation state of XyG was studied in seven Arabidopsis tissues.
Gene expression and enzyme activity levels were highest in the uppermost regions of the
expanding Arabidopsis inflorescence stem and lowest in fully expanded tissue such as
mature rosette leaves. AtFUTI ::GUS transgenic plants were used to detemrine the spatial
regulation of AtF UT 1 expression. Results generally paralleled gene expression data

obtained by other techniques. A gradient of reporter gene expression was observed in the

elongating inflorescence stem.

35S: :AtF UT 1 transgenic plants were generated to study the effect of strong constitutive
AtFUTI expression on XyG structure. These plants show higher expression of AtFUTI
and XyG FUTase activity compared to wild-type controls. Altered XyG structure is also
observed in the transgenic plants in comparison to wild-type controls; however, these
alterations consist not of highly elevated levels of fucose content, but of approximately

two-fold higher acetylation of galactose.

AtF UT 1 is one of a small number of genes encoding a cell wall biosynthetic enzyme with
confirmed biochemical fimction. The sequence information from this gene and enzyme,
as well as gene expression data, should assist in the identification and evaluation of other

plant cell wall biosynthetic genes.

ACKNOWLEDGMENTS

I have a confession to make. I always read acknowledgements — in books, in
dissertations, in presentations. They allow insight into the author’s experiences and
emphasize how progress towards any goal is a collective effort. So the first person I’d
like to thank is you, the reader, for your audience. The documentation of work such as
this would be neither necessary nor meaningful if there were no parties interested in its

conclusions.

To my parents, I offer deep gratitude. I have been tremendously fortunate to not only live
in an age and a culture in which education (and particularly the education of women) was
held in high regard, but to be born into a family where learning was respected. My
parents encouraged me to read, explore, and ask questions while growing up. They
enabled me to pursue a bachelor’s degree, and have been staunch supporters of my efforts
in graduate school. Thank you, Mom and Dad, for giving me confidence and the freedom

to pursue my goals.

There were many educators, particularly science teachers, who spurred my interest in

biology. In particular I’d like to thank Dr. James Perley, whose Plant Biology course at

The College of Wooster convinced me that I wanted to do research on plants.

At the Plant Research laboratory, there have been many people with whom it has been a

genuine pleasure to work. In the Cell Wall group: Dr. Amy DeRocher, who put a

iv

tremendous amount of effort into purification of pea xyloglucan fucosyltransferase; Dr.
Maor Bar-Peled, who also worked on purification and characterization of the pea
enzyme; Weiqing Zeng, who has worked on several different projects and who is always
willing to offer advice or assistance; Dr. Ahmed F aik, whose biochemical and
carbohydrate analysis expertise is invaluable; Dr. Rodrigo Sarria-Milan, who worked in
Natasha’s lab before moving on to a position with BASF; Dr. Tanya Wagner, who has
provided excellent input into this project and who is always ready and willing to
collaborate on experiments; Sybil Meyers, whose technical skills are exemplary; Dr.
Curtis Wilkerson, our resident bioinformatics and molecular biology guru who has
provided critical experimental advice on many occasions; Dr. Natasha V. Raikhel, who
has provided energy and insight towards this study as a co-advisor; and my primary
advisor, Dr. Kenneth Keegstra. Ken has been a tremendous help and influence over the
past five years. His outstanding mentoring has taught me a great deal not only about
effective research, but also about effective management technique. He has demonstrated
time and again that it is possible to conduct highly rigorous scientific research, and yet
still take time to socialize with the lab or pursue hobbies. He has also provided an
excellent example of how to be diplomatic and courteous, always treating others with
respect even when he disagreed with a scientific point of discussion. Countless thanks to
other members of the Keegstra lab, past and present: Dr. Jenny Davila-Aponte, Dr.
Sigrun Reumann, Dr. John Froehlich, Dr. Lynda Fitzpatrick, Dr. Mitsuru Akita, Dr.
Kentaro Inoue, and Diane Jackson. All have made this lab a great place in which to
work. And more thanks go to other members of the MSU-DOE PRL, though there are

too many to mention everyone by name: Linda Danhof, Nikki LeBrasseur, Dr. John

Scott-Craig, Dr. J ianping Yu, Dr. Anton Sanderfoot, Dr. Diane Bassham, Emily Avila-
Teeguarden, and many others. Finally, I’d like to thank the support staff at the PRL
(Karen Bird, Karen Cline, Zita Schneider, Jackie Malkins, Uwe Rossbach, Marlene

Cameron and Kurt Stepnitz), all wonderfill people who make everything run smoothly.

My last and deepest thanks go to my husband, Clint Thayer. You have been steadfast and
patient, always confident in my abilities even when the inevitable bad days happened.
Without your support, this effort would have been so much more difficult. Thank you for

all of your encouragement and understanding.

vi

TABLE OF CONTENTS

LIST OF TABLES .................................................................................................. x
LIST OF FIGURES .................................................................................................. xi
INTRODUCTION ................................................................................................... 1
CHAPTER ONE - The Plant Cell Wall: Composition, Synthesis and Biological Roles
1.1 Plant cell wall structure ..................................................................................... 4
1.1.1 Primary Cell Wall Structural Components .................................................. 4
1.1.2 Models for the Organization of Primary Cell Wall Components ................. 15
1.1.3 Biological Function of Xyloglucan .............................................................. 20
1.2 The Cell Wall and Grth ................................................................................ 26
1.2.1 Cell Division ................................................................................................ 26
1.2.2 Cell Elongation and Expansion ................................................................... 28
1.3 The Cell Wall and Environmental Interactions ................................................ 31
1.4 Cell Wall Events During the Life of a Cell ...................................................... 35
1.4.1 Plant Cell Wall Polysaccharide Biosynthesis .............................................. 35
1.4.2 Organization and Remodeling of the Cell Wall ........................................... 42
1.5 Approaches to Identify and Evaluate Cell Wall Biosynthetic Enzymes

and Genes ........................................................................................................... 46
1.6 Conclusions ........................................................................................................ 50
References ................................................................................................................. 50

CHAPTER TWO - Purification and Biochemical Characterization of Pea Xyloglucan

F ucosyltransferase ..................................................................................................... 64
2.1 Introduction ......................................................................................................... 65
2.2 Materials and Methods ........................................................................................ 67
2.2.1 Purification of Pea Xyloglucan Fucosyltransferase ...................................... 67
2.2.2 Activity Assays ............................................................................................. 68
2.2.3 In-gel Assay and Two-Dimensional Electrophoresis ................................... 68
2.2.4 Carbohydrate Methylation Linkage Analysis ............................................... 69
2.2.5 Amino Acid Sequencing ............................................................................... 69
2.2.6 Amino Acid Modification ............................................................................. 69
2.3 Results ................................................................................................................. 70
2.3.1 Purification of Pea Xyloglucan Fucosyltransferase ...................................... 70
2.3.2 Analysis of Purified Pea Xyloglucan Fucosyltransferase Activity ............ '... 74
2.3.3 Product Characterization ............................................................................... 76
2.3.4 Determination of Peptide Amino Acid Sequence ......................................... 78
2.3.5 Characterization of Amino Acid Residue Types Essential for Activity ....... 78
2.4 Discussion ........................................................................................................... 80
References ................................................................................................................. ‘ 85

vii

CHAPTER THREE - Identification and Functional Analysis of an Arabidopsis thaliana

Gene Encoding a Xyloglucan Fucosyltransferase .................................................... 89
3.1 Introduction ......................................................................................................... 90
3.2 Materials and Methods ........................................................................................ 91
3.2.1 RNA Gel Blot Analysis ................................................................................. 91
3.2.2 Library Screening .......................................................................................... 92
3.2.3 In Vitro Transcription and Translation .......................................................... 92
3.2.4 Activity Assays .............................................................................................. 93
3.2.5 Antibody Preparation ..................................................................................... 93
3.2.6 Irnmunoprecipitation ...................................................................................... 94
3.2.7 Heterologous Expression ............................................................................... 94
3.3 Results .................................................................................................................. 95
3.3.1 Cloning and Analysis of Arabidopsis Xyloglucan Fucosyltransferase .......... 95
3.3.2 In Vitro Transcription and Translation ........................................................... 102
3.3.3 Immunoprecipitation of XyG F UTase Activity ............................................. 104
3.3.4 Expression of AIF UT 1 in Mammalian Cells .................................................. 106
3.4 Discussion ............................................................................................................ 108
References .................................................................................................................. 1 1 1

CHAPTER FOUR - Analysis of Xyloglucan Fucosylation as a Model Cell Wall

Biosynthetic Process .................................................................................................. 113
4.1 Introduction .......................................................................................................... 114
4.2 Materials and Methods ......................................................................................... 116
4.2.1 Plant Growth Conditions ............................................................................... 116
4.2.2 Collection of Tissue Panels ........................................................................... 118
4.2.3 RNA Preparation ........................................................................................... 118
4.2.4 Expression Analysis by RT-PCR ................................................................... 119
4.2.5 Golgi Vesicle Preparations ............................................................................ 120
4.2.6 Activity Assays .............................................................................................. 121
4.2.7 Xyloglucan Analysis ...................................................................................... 121
4.2.8 Generation of AtFUTl ::GUS-GF P Transgenic Plants .................................. 124
4.2.9 GUS Staining Conditions ............................................................................... 125
4.2.10 Treatments with Hormones or Abiotic Conditions ........................................ 125
4.2.11 Generation of 358::AtFUT1 Transgenic Plants ............................................. 126
4.2.12 Quantification of Elongation in Inflorescence Stem Regions ........................ 127
4.3 Results .................................................................................................................. 127
4.3.1 Gene Expression, Enzyme Activity, and XyG Structural Analyses of
Arabidopsis Tissues ....................................................................................... 127
4.3.2 Expression Profiling by Northern Blot and Microarray Analysis ................. 137
4.3.3 Expression Analysis Using Transgenic AtFUTl ::GUS-GFP Plants ............. 141
4.3.4 Quantification of Elongation Rate in Inflorescence Stern Regions ............... 150
4.3.5 Analysis of Cis-acting Promoter Element Motifs in the AtF UT 1 Promoter
Region ............................................................................................................ 152
4.3.6 Production and Analysis of 3SS::AtFUT1 Transgenic Plants ....................... 157
4.4 Discussion ............................................................................................................ 161

viii

References .................................................................................................................. l 75

CHAPTER FIVE - Conclusions and Future Directions ............................................ 177
5.1 Summary of Research Significance ..................................................................... 177
5.2 Future Directions ................................................................................................. 178
References .................................................................................................................. 181

ix

LIST OF TABLES

Table l: Xyloglucan nomenclature ........................................................................... 9
Table 2: Product methylation linkage analysis ......................................................... 77
Table 3: Effect of group-specific reagents on purified pea XyG-FUTase

activity ............................................................................................................ 79
Table 4: Comparison of pea xyloglucan fucosytransferase peptides and corresponding

Arabidopsis peptides ..................................................................................... 96
Table 5: Comparison of motifs found in alpha-1,2 and alpha-1,6

fiicosyltransferases ......................................................................................... 100
Table 6: Promoter cis-elements of AtF UT 1 .............................................................. 154
Table 7: Analysis of XyG from 358::AtFUT1 transgenic plants ............................. 160
Table 8: Summary of data on AtFUTI expression from public database of microarray

experiments .................................................................................................... 172

LIST OF FIGURES

Figure 1: General structure of xyloglucan ................................................................ 8
Figure 2: Composition of Arabidopsis leaf cell walls .............................................. 19
Figure 3: Xyloglucan fucosyltransferase activity assay ........................................... 71
Figure 4: Biochemical purification of XyG FUTase from pea ................................. 72
Figure 5: Identification of proteins that co-migrate with fucosyltransferase

activity ............................................................................................................ 74
Figure 6: RNA gel blot using an EST for AtF UT 1 as a probe .................................. 97
Figure 7: Kyte-Doolittle hydrophobicity plot of AtFUTl ........................................ 101
Figure 8: Sodium carbonate treatment of in Vitro-transcribed and translated

AtFUTl .......................................................................................................... 103
Figure 9: Immunopreciptation of XyG FUTase activity from Arabidopsis

microsomal proteins ...................................................................................... 105
Figure 10: AtF UT 1 expressed in a COS cell line shows XG-specific FUTase

activity ........................................................................................................... 107
Figure 11: Tissue panel composition ........................................................................ 129
Figure 12: Expression profile of AtF UT 1 by RT-PCR ............................................. 131
Figure 13: XyG FUTase activity of Golgi vesicle samples ...................................... 133
Figure 14: Procedure for analysis of XyG fucosylation ........................................... 135
Figure 15: Analysis of XyG fucosylation ................................................................. 136
Figure 16: Expression profiling of AtF UT 1 by RNA gel blot analysis .................... 139
Figure 17: Expression of AIF UT 1 and XyG FUTase activity levels in leaves and

roots of Arabidopsis grown in liquid culture ................................................. 140
Figure 18: Constructs generated for promoter-reporter analysis .............................. 143

Figure 19: Expression of AtF UT 1 in 3-day-old promoter-GUS transgenic plants... 144
Figure 20: Expression of AtF UT 1 in S-day-old promoter-GUS transgenic plants... 145
Figure 21: Expression of AIF UT 1 in 10-day-old promoter-GUS transgenic plants. 146
Figure 22: Expression of AIF UT 1 in 14-day-old promoter-GUS transgenic plants. 147
Figure 23: Expression of AIF UT 1 in promoter-GUS transgenic plants and mature

organs. ........................................................................................................... 148
Figure 24: Expression of AtF UT 1 in promoter-GUS transgenic plants and intact

stems ............................................................................................................. 149
Figure 25: Elongation in upper and lower regions of Arabidopsis

inflorescence stems ....................................................................................... 151
Figure 26: Expression of AIF UTI in hormone-treated or etiolated

promoter-GUS transgenic plants ................................................................... 156
Figure 27: Expression analysis of AIF UT 1 in 35S::AtF UT 1 transgenic plants by

RT-PCR ........................................................................................................ 158
Figure 28: XyG FUTase activity in Golgi vesicles derived from 35S::AtF UT 1 transgenic

plants ............................................................................................................. 159
Figure 29: XyG fucosylation in 35S::AtF UT 1 transgenic plants ............................. 161
Figure 30: Structure of genomic region upstream of AtF UT 1 ................................. 165

xi

INTRODUCTION
Were it not for the plant cell wall, the mechanisms of plant growth and development
would be fundamentally different. The plant cell wall acts as a cellular exoskeleton,
providing support and positional context to the cytoplasm of each cell. Herbaceous
plants adopt an upright form because of the turgor pressure exerted against the rigid cell
wall. Turgor pressure is not possible, in cells that lack a Wall. Ultimately, cell walls are
responsible for the morphology characteristic of each organ so familiar in nature. The
extensibility of the cell wall delimits cellular elongation, and continued elongation or
expansion requires synthesis and deposition of new cell wall material. Thus, growth and
development is determined by properties of the cell wall in plants. The mechanism of
growth is much different in animals, which lack a cell wall. In short, cell walls provide

structure to the cell, leading to formation of forces that sculpt plant form.

In addition to being required for normal growth and development, the cell wall is
absolutely required for cell division in plants. Plant cells divide by forming a new cell
plate at the phragrnoplast, a structure that forms at the point of division, where callose
and nascent cell wall materials are deposited to partition the cytoplasm into two daughter
cells. This mechanism is distinct from animal cells, which pinch in two to form new
cells. However, the plant cell cannot proliferate by any other means. Naked cells or

“protoplasts” that lack a cell wall cannot divide.

The wall is alSo the means by which plants interact with their environment. Pathogens

that attack plants must either penetrate the cell wall or grow within the apoplast. Perhaps

 

arising as a response to the cell wall lesiOns that can accompany pathogen attack,
fragments of cell wall material can serve as signaling compounds or “elicitors” of plant
defense responses. Plants perceive abiotic .signals through the cell wall, resulting in
differential cell wall extensibility (as in gravotropism or phototropism) or cell wall
reinforcement (as in responses to touch or wind stress.) Many signals and hormones
involved in the perception of and response to environmental conditions must negotiate
the apoplastic region of the plant before binding to specific receptors, some of which may

have apoplastic domains.

In the first chapter of this thesis, I will describe the structure and function of the plant cell
wall in more detail. I will consider polysaccharide biosynthesis, including the enzymes
involved in this process, and then will focus on the process of cell wall polysaccharide
biosynthesis specifically. I will discuss the regulation of cell wall biosynthesis as it
relates to plant growth. The cellular localization of cell wall biosynthetic processes will
be considered, and known genes and enzymes involved in these processes will be
described. I will also briefly summarize known mechanisms for remodeling of the cell
wall during the life of the plant. Throughout this discussion, I will focus in detail upon
one polysaccharide component of the primary cell wall, xyloglucan. The second chapter
will describe the purification and biochemical characterization of a plant cell wall
biosynthetic enzyme, xyloglucan fucosyltransferase, from garden pea. The third chapter
will describe the identification of the gene encoding xyloglucan fircosyltransferase in
Arabidopsis thaliana and the confirmation of its biochemical function. The fourth

chapter will describe studies in which the regulation of xyloglucan fucosylation was

assessed in Arabidopsis as a model cell wall biosynthetic process. Finally, the fifth

chapter will present conclusions and remarks on possible future research directions.

CHAPTER ONE
The Plant Cell Wall: Composition, Synthesis and Biological Roles

1.1 Plant cell wall structure

The plant cell wall can be divided into the primary cell wall, which is associated with
growing cells, and the secondary cell wall that forms later in some cell types and is
associated with cessation of growth, structural reinforcement or specialized
differentiation (as in vascular cells). These two wall types differ in composition, and
only the primary cell wall will be considered in detail in this chapter. In general the plant
cell wall consists primarily of polysaccharides, though other components include proteins
and phenolic compounds. The structure of the cell wall is complex, but ordered. Some
polymers in the plant cell wall are among the most intricate found in nature and, for that
reason, challenging to analyze structurally. Because of this complexity, it has taken
decades of study of cell wall material for a comprehensive picture of wall architecture to
emerge. This section will begin with a survey of cell wall polymers, and will then

consider models of primary cell wall architecture and techniques used to study cell walls.
1.1.] Primary Cell Wall Structural Components

Polysaccharides

Cellulose I

Cellulose is the most characteristic feature of the plant cell wall and is the most abundant
biopolymer on Earth. This polymer is comprised of B-(l ,4)-glucose units linked together

to form chains that are up to at least 15,000 units long (Brett and Waldron, 1996).

Although the structure of cellulose appears simple at first glance, these polymers are
highly ordered into a metastable crystalline form called cellulose I through hydrogen
bonding. Coalesced molecules of cellulose form into microfibrils, long structures with a
width of about 10 nm in higher plants (though they may be as thin as 2 nm in young
cells) (Brett and Waldron, 1996). Solid-state NMR studies showed that native cellulose
is composed of two allomorphs, called Ia and 113, with the 10 form being most prominent
in higher plants (Atalla and VanderHart, 1984; Delmer, 1999). Subsequent studies
determined that individual glucan chains in cellulose are arranged parallel to each other,
with all reducing ends at the same end of the microfibril (Delmer, 1999). Additionally,
the Ia alloform was shown to have a triclinic unit cell while the 113 form has a monoclinic

unit cell (Delmer, 1999).

Unlike other cell wall polysaccharides, cellulose is synthesized at the plasma membrane
by a catalytic complex of proteins visualized in electron microscopy as rosettes (Kimura
et al., 1999). In the past five years there has been a great deal of research on cellulose
biosynthesis, and a family of enzymes have been identified that are thought to act as the
catalytic subunit of the cellulose synthase complex (Arioli et al., 1998; Holland et al.,
2000; Pear et al., 1996). More information about cellulose synthesis will be presented

later in this chapter.

Hemicelluloses: Xyloglucan
Hemicellulose is a generic term for polymers that interact closely with cellulose

microfibrils to form a cellulose-hemicellulose network. Traditionally, ' these polymers

have been defined as those cell wall polymers that can be extracted with molar
concentrations of alkali, no matter what their structures are (McCann and Carpita, 2000).
Since these polysaccharides aren’t really “half” (“hemi”) of a cellulose molecule, a more
appropriate term might be “pericellulose”, although for historical reasons the term
hemicellulose continues to be used. Xyloglucan (XyG) is the major hemicellulose found
in dicotyledonous plants and all monocotyledonous plants with the exception of
Commelinoids, which include most notably the grasses but also bromeliads, palms,
gingers, and cypresses (Hayashi, 1989). These latter species have a type 11 cell wall
(described below) and utilize glucuronoarabinoxylan (GAX) as the major cross-linking
hemicellulose. Thus, in the majority of higher plants, xyloglucan is the most abundant

polymer that serves to cross-link cellulose microfibrils.

XyG has a backbone structure that is identical to cellulose. Unlike cellulose, however,
glucose units are substituted with side chains (Hayashi, 1989). The substitution patterns
vary, and the most common side chains consist of Xyl alone, Gal-Xyl disaccharides, or
Fuc-Gal-Xyl trisaccharides (see Figure 1). A shorthand nomenclature is used to describe
XyG side chain substitutions, shown in Table 1. In most dicots and noncommelinoid
monocots, the structural pattern tends to consist of three substituted Glc residues
followed by a fourth unsubstituted Glc (McCann and Carpita, 2000). The predominant
subunits tend to be XXXG and XXFG, but XLF G is also observed (Guillen et al., 1995).
In some species, rare instances of XFFG or XFLG have been observed, but fucosylation
at the second position of the subunit is not common (Hisamatsu et al., 1991). Additional

modifications may include acetylation of galactose residues and addition of or-L-Ara

(Brett and Waldron, 1996; Mch and Carpita, 2000). Solonaceous species (tomato,
tobacco, peppers) and peppermint have arabino-xyloglucan, where fucose is absent but
Ara may be added to Xyl. Major subunits in these cases are AXGG, XAGG, or AAGG
(York et al., 1996). A very small proportion of the commelinoid monocot primary cell
wall may contain XyG, but little substitution has been observed other than irregular Xyl
residues (McCann and Carpita, 2000). Some species, notably tamarind (T amarindus
indica) and nasttu'titun (Tropaeolum maj us), accrunulate non-fucosylated XyG in the
periplasmic space between the plasma membrane and cell wall of endosperrn cells (York
et al., 1990). This polysaccharide is broken down during seed germination and is thought
to serve as a seed storage carbohydrate. Thus, it may be used as an acceptor substrate for
XyG fucosylation assays. Non-fucosylated seed storage XyG is distinct from the

endogenous cell wall XyG found in tamarind and nasturtium, which does contain fucose.

0‘l
2
Gal
1
Bl
2
Xyl Xyl Xyl XY'
1 1 1 1
“I 0‘I 0‘l 0‘l
6 6 6 6
4G101E4Glg1-l-3-4GICl-E4GIC1E GIC1E4GIg1-E4GIC1E4GIC1B
01| «I
1 1
Xyl Xyl
Nonasaccharide Heptasaccharide

Figure 1. General structure of xyloglucan. Xyloglucan consists of a B (1,4) glucan
backbone decorated with side chains of xylose, xylose and galactose, or xylose, galactose

and fucose. The shorthand descriptions of these units are: XXFG (nonasaccharide),
XXXG (heptasaccharide) (see Table 1.)

Table l. Xyloglucan nomenclature. An abbreviated nomenclature has been adopted to
describe different subunit structures in xyloglucan. The most common structures are in
bold. Modified from (Fry et al., 1993b).

 

 

 

 

 

 

 

 

 

 

 

Abbreviation Structure
G B-D-Glcp
X a-D-Xylp-(l,6)-B-D-Glcp
L B-D-Galp—(1,2)- a-D-Xylp—(l ,6)-[3-D-Glcp
F a-L-Fucp-(1,2)- [i-D-GalI-QJ} a-D-Xylp-(1,6)-B-D-Gch
A or-L-Araf-(l,2)- B-D-Glcp
I
a-D-Xylp-(l ,6)
B B-D-Xylp-(1,2)- )- B-D-Glcp
l
a-D-Xylp-(l ,6)
C or-L-Araf-(l,3)- B-D-Xylp-(l,2)- B-D-Glcp
l
a—D-Xylp-(l ,6)
S (found in oc-L-Araf-(l,3)- B-D-Xylp-(l,6)- B-D-Glcp
Solonaceous
Spa)

 

 

Hemicelluloses: Glucuronoarabinoxylans

Glucuronoarabinoxylans (GAXs) are the major cross-linking polymers in commelinoid
monocots, although cell walls from all angiospenns also have some amount of GAXs
(McCann and Carpita, 2000). GAXs have a backbone of B-(1,4)-linked Xyl and are
substituted with single-sugar additions of a-L-Ara and a—D-Gch units (V erbruggen et
al., 1998). In commelinoid monocots, the Ara units are most commonly added at the 0-2
position; the Gch residues are attached at the 0-2 position of different Xyl units
(McCann and Carpita, 2000). The Gch units may be methylated at the 4-0 position

(Rizk et al., 2000).

Hemicelluloses: ( I , 3), (1,4)-flD-glucans

Grasses (Order Poales) also have (1,3), (1,4)-B-D-glucans as an additional hemicellulose.
This polymer is also known as mixed-linkage glucan. The structure of this polymer
consists of three-Glc (cellotriose) and four-Glc (cellotetraose) units, present in a ratio of
about 2:1, connected by B-(1,3)-linkages (Buckeridge et al., 2001). The B-(1,4) regions
adopt an extended conformation, while the B-(1,3) linkages introduce kinks to ultimately

result in interrupted corkscrew-shaped polymers (Brett and Waldron, 1996; McCann and

Carpita, 2000).

Other hemicelluloses
Other cross-linking polymers, usually found to a lesser extent in the cell wall, include
g1 ucomannans, galactomannans, mannans, xylans, glucuronomannans, and

gaIaCtoglucomannans (Brett and Waldron, 1996; McCann and Carpita, 2000; Mohnen,

10

 

1999; Sims et al., 1997). Glucomannans are the major hemicellulose in gymnosperrns
and have an unsubstituted B-(1,4)-linked backbone of irregular alterations of Glc and
Man. In other angiosperms, members of this collection of hemicelluloses are present in

fairly low abundance.

Pectins

Pectins form a gel-like substance and are abundant in the middle lamella region that is
sandwiched between adjacent cell walls. This group of polysaccharides is rich in GalA
and may be extracted from the wall by treatment with Ca2+ chelating agents such as
EDTA, EGTA, ammonium oxalate or cyclohexane diamine tetraacetate (Mohnen, 1999).
As might be expected, many pectins interact with Ca2+ ions, and the functional
implications of this interaction is described in more detail below. Pectins are often used

as gelling agents in the food industry.

Pectins: Homogalacturonan and Rhamnogalacturonan I
Homogalacturonan (HGA) is a fundamental pectin structure consisting of oL-(l ,4)-D-
GalA polymers that contain up to 200 units (though polymers as long as 2000 units have
been reported) and reach around 100 nm in length (Brett and Waldron, 1996; McCann
and Carpita, 2000; Willats et al., 2001). Rhamnogalacturonan I (RG I) consists of
repeating (1,2)-a-L-Rha-(1,4)-a-D—GalA units, forming a rod-like polymer (Willats et
81 ., 2001). RG I may be covalently connected to HG, and the GalA units on RG I may be
modified by methyl esterification (Brett and Waldron, 1996). Rha units of RG 1 may also

be modified by side chains of variable structure, but principally containing Gal and Ara,

11

4 V ‘

though Fuc has also been reported (An et al., 1994; Brett and Waldron, 1996). Together,

HG and RG I are the most abundant pectins in most plant species.

Pectins: Rhamnogalacturonan II

Rhamnogalacturonan H (RG 11) is a highly substituted, highly structurally complex
modified HGA (Whitcombe et al., 1995). Monosaccharides present in RG 11 include
GalA, Rha, Ara and Gal (present in a ratio of 10:7 :5:5) with smaller amounts of unusual
monosaccharides such as apiose, aceric acid, 2-0-methyl-Fuc, KDO and DHA (Brett and
Waldron, 1996; Herve du Penhoat et al., 1999; McCann and Carpita, 2000; Whitcombe et
al., 1995). It is present in smaller amounts in the cell wall, but its intricate structure is
conserved throughout the angiosperms (Brett and Waldron, 1996; McCann and Carpita,
2000). RG H may dimerize by forming diester bonds with boron, suggesting that it may
serve an important functional role despite its low abundance in the cell wall (O'Neill et
al., 1996). Recently, it has been shown that RG-II isolated from a mutant with a lesion in
the GDP-Fuc biosynthetic pathway, murI , has a reduced rate of borate cross-linked dimer
formation (O'Neill et al., 2001). Thus, it would appear that Fuc or the 2-0-methyl-Xyl
linked to F no in this molecule is required for dimerization. Another recent study

described covalent connections between RG H and HG (Ishii and Matsunaga, 2001).

Other Pectins
Arabinan (or-(1,5)-D-Ara chains substituted with a-(1,3)- or a-(1,2)-linked Ara),
galaCtans (B-(l,4)-D-Gal chains) or arabinogalactans (galactan chains substituted with

shorter a-(1,5)-D-Ara chains or terminal Ara units) are also found in cell walls (Eriksson

12

 

et al., 1996). Type II arabinogalactans ((1,3) and (1,6)-B-D-Gal chains attached by (1,3),
(1,6) branch points) are found associated with particular proteins to form arabinogalactan

proteins, described in detail below.

Non-pob'sacclraride components of the cell wall

Cell Wall Proteins

In addition to polysaccharides, the cell wall may also contain a complement of proteins,
many of which are thought to have a structural role and impart reinforcement after most
grth has occurred by cross-linking to other wall components (Showalter, 1993). Three
major classes of cell wall proteins are hydroxyproline-rich glycoproteins (HRGPs),

proline-rich proteins (PRPs) and glycine-rich proteins (GRPs).

Proteins: Hydroxyproline—rich Proteins
l-IRGPs, as the name suggests, contain an abundance of the unusual amino acid
hydroxyproline. Extensin is the best-studied HRGP and contains repeating Ser-(Hyp)4
and Tyr-Lys-Tyr sequences, forming a helical structure that appears rodlike in electron
microscopy studies (Kieliszewski and Lamport, 1994). Extensin may be glycosylated
extensively with Ara chains attached to Hyp; there may also be single Gal substitutions
attached to Ser residues. Tyr residues may participate in intra- or intermolecular
crosslinking to form isodityrosine linkages. Extensin glycosylation patterns differ
between dicots and monocots, with less glycosylation occurring in monocots (McCann

and Carpita, 2000).

13

Proteins: Praline-rich Proteins

Like HRGPs, PRPs form a large multigene family in most species. PRPs described thus
far form two classes: those which have tandem repeats of Pro-Pro-Val-X-(Asp/Thr),
where X is often Tyr, His, or Glu, or those which have an N-terminal Pro rich domain and
a C-terminal domain that is not rich in Pro (Fowler et al., 1999). PRPs are not heavily
glycosylated. They are thought to provide structural reinforcement during development

and in response to damage from wounding or pathogen infection (Hong et al., 1989).

 

Proteins: Glycine-rich Proteins

GRPs may contain over 70% Gly and form B-pleated sheets (Sachetto-Martins et al.,
2000). These proteins often are present at the interface of the cell wall and the plasma
membrane, and have been suggested to be nucleation sites for lignin formation (Sachetto-

Martins et al., 2000). Like PRPs, GRPs are not heavily glycosylated.

Phenolic Compounds
Cell walls may also contain phenolic compounds, normally thought of as being
constituents of lignin in the secondary cell wall but also present. in some primary cell
walls. In commelinoid monocots and plants in the Chenopodiaceae family such as sugar
beet and spinach, phenolic compounds such as hydroxycinnarnic acids are abundant
(Wallace and Fry, 1994). Because of these compounds, cell walls of such species will
fluoresce under ultraviolet light. Ferulic acid and para-coumeric acid are the most
common hydroxycinnarnic acids. Cross-links between these units and GAX serve to

structurally reinforce grass cell walls (Wallace and Fry, 1994).

14

 

1.1.2 Models for the Organization of Primary Cell Wall Components
Some of the first models of cell wall structure were based on data fi'om cell wall
fractionation studies and suggested mechanisms for growth hormone-induced wall
loosening. One model suggested that auxin-induced extension of the plant cell wall
occurred either by a process of transglycosylation of polymers or by irreversible
hydrolysis of covalent polysaccharide bonds (Cleland, 1981). In each proposed
mechanism, slippage of the cell wall polymers was suggested to occur during
“stretching”, and new covalent connections between polymers were thought to form
before cessation of elongation. This model did not propose specific polymers tha‘twould
be targets for transglycosylation or hydrolysis, instead putting forth a generalized scheme
for events possibly required during elongation growth. An earlier study developed a
classic primary cell wall model, proposed by the Albersheim group (Keegstra et al.,
1973). In this model, a structure of the primary cell wall was proposed based on covalent
interconnections between cell wall components as suggested by extraction and
fractionation experiments. Xyloglucans (Xsz) were suggested to interact with cellulose
microfibrils via hydrogen bonds, but authors suggested that the XyG polymers covalently
connected to pectic polysaccharides, with further covalent connections between pectins
and arabinogalactan chains attached to wall proteins through serine residues (Keegstra et
al., 1973). In this model, the proposed mechanism for extension of the cell wall was
transient dissociation between XyG polymers and cellulose as the hydrogen bonds were
broken and then reformed during cell wall slippage. Several subsequent studies

Questioned the possibility of covalent bonding between XyG and pectin, and this aspect

15

 

of the model was generally disputed (Hayashi, 1989; McCann and Carpita, 2000; McNeil
et al., 1984). A recent study, however, indicated evidence for covalent linkage between
about one third of the XyG component and pectin in suspension-cultured rose cell walls,

and suggested that the Keegstra et al. model be re-visited (Thompson and Fry, 2000).

Carpita and Gibeau later proposed another model of plant primary wall structure (Carpita
and Gibeaut, 1993). The authors differentiated between the Type I primary cell wall,
characteristic of all dicots and nongraminaceous monocots examined to date, from the
Type H primary cell wall, characteristic of the grasses. The model is also generalized, as

walls of specific tissues or from certain species may show structural variation.

In the Type 1 cell wall model, there are three components: the cellulose-xyloglucan
framework, a gel-like matrix of pectins, and structural proteins. The main structural
features of the wall are suggested to form by the interaction of xyloglucan and cellulose
polymers by hydrogen bonding. One fraction of XyG is suggested to interact closely
with microfibrils, perhaps even integrating into the microfibrils themselves in some
cases. Other XyG polymers are suggested to interact less directly with cellulose, serving
instead to bridge the distance between adjacent microfibrils and thus serve as a network
of hemicelluloses throughout the wall, with several XyG strands coalescing together in
some regions. Evidence for the two fractions of XyG is provided by extraction
experiments, in which both easily extracted and tightly associated portions of XyG are

derived fi'om cell wall preparations (Pauly et al., 1999). Microscopic studies have also

16

allowed visualization of cell walls from which pectins have been removed, showing lacy

interconnections of polymers between cellulose microfibrils (Mch et al., 1990).

The second domain of the cell wall consists of pectins that form a gel-like matrix. The
porosity of the cell wall is thought to be controlled in large part by pectins, as the degree
of methyl esterification of GalA residues in homogalacturonan (HGA) as well as the
concentration of Ca2+ present in the wall may influence the formation of “junction
zones”. These zones are regions of association between antiparallel chains of HGA
cross-linked with Ca2+ ions. If esterification occurs, interaction of HGA chains with
calcium ions, and thus the formation of junction zones, is prevented. Other
manipulations of pectin structure may also influence physical characteristics. For
example, pore size limiting unrestricted diffusion in the primary cell wall was shown to
be about 4 nm in diameter (Carpita et al., 1979), but enzymatic digestion of pectins may
increase the pore size to over 10 nm (Baron-Epel et al., 1988). This effectively means
proteins over 100 kDa might be able to move through these 10 nm pores via unrestricted

diffusion.

The third domain of the Type I primary cell wall in this model consists of structural
proteins, many of which are rich in glycine or hydroxyproline and have characteristic
repeating units (Carpita and Gibeaut, 1993). The protein domain is thought to lock the

primary cell wall structure into place, and may serve an important role in reinforcement

of the wall after elongation.

l7

 

One analysis of cell wall structure in Arabidopsis has been published (Zablackis et al.,
1995). This study utilized leaf material from four-week-old Col-0 Arabidopsis thaliana
plants to determine the content and structures of major cell wall components. It, should
be noted that Zeblackis et al. only determined cell wall structure for one type of tissue,
rosette leaves, and that the leaves of four-week-old Arabidopsis plants might contain both
primary cell wall material and secondary cell wall material (from vascular tissue). Leaf
cell walls of Arabidopsis contain a high proportion (34% of the total wall material) of
phosphate buffer-soluble pectins relative to other plant cell walls that have been
analyzed, indicating that pectins are both fairly abundant and fairly weakly held within
the wall in this species. The composition of the Arabidopsis leaf cell wall is shown in
Figure 2. Pectins include rhamnogalacturonan I (RG I), rhamnogalacturonan H (RG H),
and homogalacturonan (HGA). Xyloglucan is the primary hemicellulose, though a small
amount of glucuronoarabinoxylan (GAX) is also present. Cellulose comprises 14% of
the cell wall. Structural proteins also occur in Arabidopsis leaf cell walls in relatively
high abundance, and the amino acid composition of these proteins differs somewhat from
that observed in other species (Zablackis et al., 1995). With the exception of the protein
content of Arabidopsis leaf cell walls and the high content of phosphate buffer-soluble
pectins, the only other notable difference observed was that Arabidopsis leaf RG 1 is less
highly branched than RG 1 from sycamore suspension cells. The authors concluded that
in all other respects, cell walls from Arabidopsis leaves resembled cell walls from other

species that have been analyzed (Zablackis et al., 1995).

18

I RG ll
:3 Homogalacturonan

! I Xyloglucan

. I Glucuronoarabinoxylan
l I cellulose
1392151- __

‘mRGl-

 

 

20% ,, .,

 

Figure 2. Composition of Arabidopsis leaf cell walls. Data from Zeblackis et al., 1995.

Some cell wall models have been necessarily static, seeking to establish a generalized cell
wall structure applicable to many types of plant cells. More recent approaches, however,
have allowed biologists to appreciate that different types of cell walls exist within the
plant, and sometimes even in different regions of a single cell (Roberts, 2001).
Antibodies against specific carbohydrate epitopes and other technical advances will
continue to allow a conceptual shift in our view of the cell wall from a general structure
to a dynamic one. Another technique, Fourier Transform Infrared (F TR)
microspectroscopy, is emerging as a possible means of examining cell wall architecture
in growing regions, different tissues and cell types, and plants subjected to environmental
or genetic modifications (McCann etal., 2001; McCann and Carpita, 2000). Researchers
are seeking to establish a sufficiently robust FTIR spectra database to allow more detailed
and more rapid analysis of a variety of different cell wall structures (N. Carpita, personal

communication.)

1.1.3 Biological Function of Xyloglucan

Xyloglucan (XyG) is known to associate with cellulose. There are thought to be different
fractions of XyG: a portion that interacts intimately with cellulose microfibrils, coating
the surface and even penetrating the outer region of the fibrils, and a portion that serves
to link adjacent microfibril chains (Hayashi, 1989; Pauly et al., 1999). The existence of
the closely-associated portion has been demonstrated by fractionation studies. Analysis
of etiolated peas showed that a portion of XyG can be removed with an XyGéspecific
endoglucanase (XEG), another portion requires treatment with concentrated alkali for

removal, and a third portion can only be extricated from cellulose by treating the XEG-

20

and KOH-treated cell walls with cellulase (Pauly et al., 1999). XyG chains that span the
distance between microfibrils have been directly visualized by microscopy studies
(Fujino et al., 2000; Itoh and Ogawa, 1993; McCann et al., 1990). Rapid freezing-deep
etching (RFDE) microscopy involves cryofixation of a cell wall sample, removal of ice
crystals under a vacuum, covering the surface of the sample with platinum and carbon
layers, dissolving away the biological material, and visualization under the electron
microscope. Crosslinking strands can be seen as narrow interconnections between the
thicker cellulose microfibrils using the RF DE technique. Fuj ino et al. have shown that
these strands are resistant to treatment with protein denaturing agents or weak alkali
solution, but disappear completely upon treatment with XyG-digesting endo-(1,4)-[3-
glucanase (Fujino et al., 2000). Anti-XyG antibodies also labeled the crosslinking
structures, providing further evidence that they are composed of xyloglucan (Fuj ino et al.,
2000). Attempts to determine the smallest unit of XyG that will bind to cellulose have
concluded that oligosaccharides with at least four glucose backbone units are necessary
(Hayashi et al., 1994b; Vincken et al., 1995). Another study also indicated that the extent
of in Vitro binding of XyG oligosaccharides to cellulose was greater than that observed
for cellulose oligosaccharides of equivalent backbone length, indicating that the presence
of side chains on a B-(l,4)-glucan molecule does impart specific binding properties

(Hayashi eta1., 1994b).

The biological function of XyG, therefore, clearly involves a structural role. Because of
the nature of this thesis, the relevance of the fucose residue to the function of XyG is of

interest. In that regard, some contradiction is found within the literature. When efforts

21

were made to predict the conformation of xyloglucan structures by computer modeling,
calculations predicted a distinct effect of side-chain presence and composition on the
conformation assumed by the XyG polymer (Levy et al., 1997; Levy et al., 1991).
Metropolis Monte Carlo simulations of conformational forms of XyG structures indicated
that lowest-energy states could be achieved when fucosylated trisaccharide side-chains
were present and folded back to interact with one face of the polymer, causing the
backbone of the molecule to adopt a flattened shape in this region (Levy et al., 1997;
Levy et al., 1991). This in turn was suggested to facilitate binding of XyG to cellulose
(Levy et al., 1997; Levy et al., 1991). When tested, researchers found that there was a
twofold higher rate of in Vitro binding between fucosylated pea XyG and cellulose
(avicel) compared to nonfucosylated tamarind seed XyG, lending biochemical suppon to
the theory that the presence of fucose on side-chains has structural consequences (Levy et
al., 1997). Hayashi et al. also found that fucosylated XyG had a higher adsorption
constant for cellulose than tamarind XyG when polymers of equal backbone chain length
were compared (Hayashi et al., 1994a). Some speculations about the role of terminal Fuc
in Zinnia have also been made, as XyG in tissues that require high tensile strength
contain Fuc while XyG in tissues that do not are nonfucosylated (McCann and Roberts,
1994). In isolated Zinnia mesophyll cells, fucosylated XyG is not produced until
elongation has halted, indicating a specialized function for this structure (McCann and

Roberts, 1994).

However, another group using a different experimental system has speculated that the

presence of fucosylated side-chains bears little impact on the function of XyG (Whitney

22

et al., 1995; Whitney et al., 1999). The bacterium Acetobacter aceti ssp. xylinwn
produces cellulose to form extracellular pellicles during fermentation. Researchers have
added exogenous tamarind XyG and other polymers to the fermentation media and
examined the resulting products using microscopic and materials science analytical
methods (Whitney et al., 1995; Whitney et al., 1999). In the presence of tamarind XyG, a
material was formed that contained cellulose microfibrils cross-linked with
interconnecting strands. Authors concluded that terminal fucose residues are not
necessary for formation of cellulose-XyG cross-links, and questioned the hypothesis
posed by Levy et al. (Whitney et al., 1995). Subsequent studies examined the uniaxial
tensile strength, small deformation oscillatory rheology, and action of cell wall modifying
proteins on this material (Whitney et al., 2000; Whitney et al., 1999). Compared to
cellulose synthesized in absence of tamarind XyG, the material was found to have greater
extensibility during uniaxial tension tests and extended in response to a cell wall
modifying protein, indicating that there are some structural differences imparted by the
presence of XyG (Whitney et al., 2000; Whitney et al., 1999). However, no data were
presented in these later studies that compared composites generated using fucosylated
forms of XyG to composites generated using tamarind XyG. It is possible that the
presence of Fuc on XyG affects properties that would be of consequence in planta but not
detectable in the Acetobacter system. For example, the presence or absenCe of Fire on
XyG may affect the suitability of XyG as a substrate for a wall modifying enzyme(s) not
investigated in these experiments; some evidence for this has already been described
(Steele and Fry, 2000; Steele et al., 2001). It is also possible that the structure of XyG

side chains affects the interaction of XyG with other polymers. A recent study detected

23

binding between radiolabeled glucuronoarabinoxylan and pectin to XyG (Rizk et al.,
2000). This binding was greatly reduced when nonfucosylated XyG was used in assays
or when terminal Fuc was removed fi'om pea XyG by mild acid treatment (Rizk et al.,

20001

In addition to its proposed structural roles, XyG has also been suggested to serve a role as
a signaling molecule within the plant. In 1984, York et al. digested XyG isolated from
sycamore cell suspension cultures with T richoderma endocellulase, a hydrolytic enzyme
that cleaves XyG at unsubstituted Glc residues to generate XyG oligosaccharides (York
et al., 1984). These oligosaccharides were purified, and surprisingly some of them
showed an ability to inhibit auxin-induced elongation of pea epicotyl segments at
nanomolar concentrations (York et al., 1984). Thus, fragments of XyG were proposed to
act as “oligosaccharins”, or small units of polysaccharides that elicit specific biological
responses when applied at very low concentrations (Fry et al., 1990). Other instances of
carbohydrates acting as signaling molecules have been reported, such as fragments of
plant or fungal cell walls serving to elicit plant defense responses against pathogens (Fry
et al., 1993a; John et al., 1997). Specific structural requirements for biological activity of
XyG fragments have been established, and further studies have shown that differing
responses occur at different levels of XyG oligosaccharide application (Aldington et al.,
1991; Fry et al., 1993a; Fry et al., 1990; McDougall and Fry, 1989). At low levels (1-10
nM), a fucose-containing nonasaccharide (subunit structure XXF G) inhibits both auxin-
induced and acid-induced growth in pea (Fry et al., 1990; York et al., 1984). The Fuc

residue is absolutely required for this effect, since oligosaccharides that lack F uc cannot

24

inhibit auxin-induced elongation. However, an oligosaccharide containing 2-0-methyl-
F uc did not act as an anti-auxin agent, indicating that Fuc is necessary but not sufficient
for activity (McDougall and Fry, 1989). At higher levels (1 uM), XyG oligosaccharides
promote growth in pea, presumably by acting as substrates for a wall-modifying enzyme
called xyloglucan endotransglycosylase (XET) that can incorporate new fi'agments of
XyG into the existing cell wall (Fry et al., 1990). Because the concentration of fragments
needed to provoke the biological response is much higher than that required for the anti-
auxin response, fragments in this case are not thought to be acting as signaling molecules
per se. The structural requirements for this activity differ, with the most active XyG
oligosaccharides actually lacking Fuc (Fry et al., 1990). The order of effectiveness of
XyG fragments matches the order of substrate preference of XEG, supporting the theory
that XET-assisted incorporation of fragments into existing polymers leads to extension

(Fry et al., 1993a).

Despite this foundation of evidence indicating that XyG oligosaccharides may act as
signaling molecules in garden pea, very little additional work has been published that
clarifies the nature of this event. No XyG nonasaccharide receptor has been reported, and
the significance of these observations remains unclear. Interestingly, a cell wall-localized
a—fucosidase has been purified from pea and characterized biochemically (Augur et al.,
1993). This enzyme was shown to be localized to the primary cell wall and is
developmentally regulated (Augur et al., 1993). It acts only on XyG oligosaccharides,
lacking the ability to cleave Fuc from large polysaccharides (Augur et al., 1993). The

gene encoding this pea enzyme has been cloned (Augur et al., 1995). However, no

25

further work describing this gene and enzyr‘ne and their possible roles in generation of

XyG oligosaccharins has been reported.

It should be noted that the activity of XyG oligosaccharides as signaling molecules has
only been studied in pea. Members of the legume family may be predisposed to the use
of carbohydrate signaling molecules, as carbohydrate structures play an important role in
the lipochitooligosaccharide Nod signaling factors necessary for nodulation by nitrogen-
fixing bacteria. It is possible that carbohydrate signaling factors may not play a

significant role in other plant species, such as Arabidopsis.

1.2 The Cell Wall and Growth

1.2.] Cell Division

It is impossible to consider plant cell division in absence of the cell wall. Early during
cell division, a cytoskeletal structure called the preprophase band forms and determines
the firture site of attachment of the cell plate to existing parental cell walls. The emerging
cell plate forms as polysaccharides are deposited at the division plane after biosynthesis
in the Golgi apparatus (for reviews, see Assaad, 2001; Smith, 1999; Sylvester, 2000).
These polysaccharides are deposited at the phragmoplast, an equatorial cytoskeletal array,
via secretory vesicles. The nascent cell plate contains cell wall materials such as
xyloglucan, synthesized in the Golgi apparatus, and an abundance of callose (beta 1,3-
glucan), synthesized at the cell plate. The presence of callose causes the immature cell
plate to be fluid, a property necessary for later events (Assaad, 2001). The cell plate

expands towards the periphery of the cell and contacts existing walls at the point

26

previously determined by the preprophase band. This process is achieved in part by
action of microtubules and is regulated by a signal transduction pathway that involves a
MAP kinase cascade (Assaad, 2001). Recent research has begun to identify some of the
molecular mechanisms for cell plate formation and positional fixation. For example, a
high molecular weight dynamin-like protein called phragmoplastin (or ADLl in
Arabidopsis) may be involved in vesicle-tubule-vesicle fusion events in the cell plate (Gu
and Venna, 1996; Kang et al., 2001; Zhang et al., 2000). Vesicle trafficking elements
required to shuttle cell wall materials to the phragrnoplast include a syntaxin called
Knolle (Lukowitz et al., 1996) and a Secl-like protein called Keule (Assaad et al., 1996).
Upon maturation of the cell plate, callose is replaced by cellulose (beta 1,4-g1ucan) and
pectin (Assaad, 2001). After the cell plate matures it eventually becomes the site of the
middle lamella, a pectin-filled structure shared by adjoining cell walls. Two enzymes
that appear to be essential for the process of cell plate maturation are Cytl, a Man-1-
phosphate guanylyltransferase required for N-glycosylation (Lukowitz et al.,, 2001;
Nickle and Meinke, 1998), and Korrigan, an endo-l,4-beta-glucanase (Zuo et al., 2000).
Additional primary cell wall material is deposited next, forming a layer that is
distinguishable in ultrastructural studies (for example, see McCann et al., 1990). In

certain cells, secondary cell wall polysaccharides may be deposited at a later point.

It is clear that the cell wall is essential for cytokinesis in plants, because cells lacking a
wall cannot divide (Cooper et al., 1994; Meyer and Abel, 1975; Suzuki et al., 1998;
Wojtaszek, 2000). If the cell wall is digested using polysaccharide hydrolytic enzymes,

the resulting naked cell is called a protoplast. Cultured protoplasts will not divide

27

without first forming a cell wall (Meyer and Abel, 1975). In one case, the application of
3,4-dehydroxyproline, which inhibits activity of prolyl 4-hydroxylase and prevents
hydroxylation and thus glycosylation of hydroxylproline-containing cell wall proteins,
prevents execution of mitosis (Cooper et al., 1994). This indicates that these cell wall
components are essential for wall formation and cell division. In contrast, application of
2,6-dichlorobenzonitrile (DCB), a cellulose biosynthesis inhibitor, did not prevent
monocot or dicot cells in suspension culture from dividing (Shedletzky et al., 1990;
Shedletzky et al., 1992). Instead, these cells, which lacked a cellulose-hemicellulose
network, developed greatly altered cell wall composition. Dicot (tomato and tobacco)
cell walls were reinforced with pectin linked with phenolic structures, while
graminaceous monocot (barley) cell walls were reinforced with increased levels of
hemicelluloses, also linked to phenolic compounds (Shedletzky et al., 1990; Shedletzky
et al., 1992). Experiments such as these indicate that plant cells can tolerate remarkable

alterations in cell wall structure and show a great deal of flexibility in adaptation.

1.2.2 Cell Elongation and Expansion

In addition to being essential for cell division, the cell wall is of fundamental importance
for growth processes. During elongation, the cell wall both permits growth and delimits
it. If sufficient water is available, water will move into the cytoplasm, where osmotically
active solutes are present. Water tends to move from a region of less negative to more
negative water potential (reviewed in Hsiao and Bradford, 1983). This eventually causes
internal pressure as the cell contents abut against the defined volume dictated by the rigid

cell wall. If the cell wall remains inflexible, turgor pressure is maintained and there is no

28

net movement of water. If the cell wall becomes extensible, however, additional water
may move into the cell, which will then begin to increase in volume. The direction of
increase can be controlled by the orientation of cell wall components and the localization
of cell wall softening. Cell walls in which cellulose microfibrils, a major component of
the wall, are not directionally organized will experience an increase in all directions,
termed isodiametric growth. Cell walls in which microfibrils have been deposited in a
directional manner will elongate along the axis perpendicular to the direction of
microfibril alignment, resulting in anisodiametric growth. It is also theoretically possible
for localized cell wall softening to dictate the dimensions of cell expansion. Plant cell
expansion, therefore, occurs through an interplay between the amount of water available
to the cell, the amount of osmotically active solutes contained within the cell, and the
extensibility of the cell wall. The cell cannot continue to expand and be viable, however,
without at some point synthesizing and depositing new cell wall material. If no new
material was added the wall would thin to the point of breakage, as occurs in isolated cell

walls undergoing extension in Vitro (Cosgrove, 2000a).

The force exerted on the cell wall by a turgid cell is not insignificant. Carpita and
Gibeaut (Carpita and Gibeaut, 1993) estimated that a turgid, spherical cell with a radius
of 50 um and a cell wall thickness of 0.1 um exerts a pressure of 2500 bars upon the cell
wall. If the cell is of equivalent volume but is cylindrical rather than spherical, it will
develop 5000 bars of tension in the tangential ads but the longitudinal tension will

remain 2500 bars. Clearly, the cell wall must maintain structural integrity in face of great

force.

29

Plant cell expansion and cell wall biosynthesis are generally well coordinated, with
increased rates of cell wall deposition occuning in regions of expansion (Cosgrove,
1997a). A stimulation of arabinan synthetase has been observed in Phaseolus vulgaris
during extension growth (Bolwell and Northcote, 1981) and subcultured bean suspension
cell cultures (Bolwell and Northcote, 1983). Labeling experiments showed a positive
correlation between the rate of deposition of radiolabeled uronides in the cell wall, which
was taken as a measure of cell wall deposition, and the growth rate in regions of maize

root tips (Silk et al., 1984).

The relationship between biosynthesis and elongation can be complex, particularly in the
case of auxin-induced elongation. For example, Bret-Harte et al. noted that application
of auxin to stem segments promoted cell wall deposition when glucose levels were
sufficient, but auxin-induced elongation during limiting glucose conditions was not
correlated with deposition and instead resulted in cell wall thinning (Bret-Harte et al.,
1991). The authors suggested that auxin promotes cell wall synthesis and cell wall
loosening by different mechanisms (Bret-Harte et al., 1991), which is not inconsistent
with later work on extension of cell walls by action of expansins (discussion below). In
an earlier study, it was observed that auxin or fusicoccin treatments promoted glucose
uptake, but not incorporation, into the cell wall under conditions in which elongation
occurred; however, synthesis was enhanced under conditions that limited elongation
(Brummell and Hall, 1983). Thus, the effect of auxin on cell wall synthesis may depend

upon the conditions occurring at the time of exposure to the growth hormone.

3O

Several possible outcomes exist as a result of cell wall deposition. If deposition rates are
sufficiently high and elongation rates are low, the wall will thicken; if deposition is
insufficient during elongation, the wall will thin (Brummell and Hall, 1983; Cosgrove,
1997a). Alternatively, deposition of new material and cell expansion may reach an
equilibrium such that the thickness of the wall is maintained despite an increase in cell
volume (Brummell and Hall, 1983). In short, the synthesis and deposition rates of new
cell wall material may be correlated with growth, particularly with the longer-term

maintenance of growth, but are not by themselves sufficient to cause growth.

1.3 The Cell Wall and Environmental Interactions

As sessile organisms, plants face a particular challenge. They must respond to
environmental factors such as biotic and abiotic stress without being able to evade them
through relocation. Since the cell wall is the first part of the plant body that interacts with
the environment, it is frequently involved in responses to the plant’s surroundings.

Several examples of cell wall participation in reactions to the environment will be

described.

Plant responses to abiotic signals frequently require differential growth to occur. For
example, phototropism involves the perception of a light signal and growth towards a
light source. Similarly, gravitropism allows plants to perceive gravitational forces and
orient growth patterns such that shoot tissue grows upwards and root tissue grows

downwards. In both cases, growth curvature may occur. While components of the signal

31

transduction pathway differ between phototropism and gravitropism, auxin is thought to
be involved in both responses. A gradient of auxin in a plant organ such as an elongating
shoot may result in differential expansion of cells on one side versus the other (Salisbury
and Ross, 1992). This ultimately results in curvature of the organ. Alternatively, cells on
one side of the organ may be more sensitive to auxin than those on the other side, also
resulting in curvature (Salisbury and Ross, 1992). Since auxin affects the extensibility of
the cell wall, the capability of the plant to respond effectively to the signal depends not
only on the ability to perceive the signal and transduce it, but for its cell walls to undergo

sufficient reorganization.

Another differential growth response to an abiotic signal occurs when plants face water
shortages. Under limiting water conditions, growth rates in shoot tissues are sharply
curtailed while growth rates in root systems are not inhibited. This is an adaptive
advantage to the plant, which can continue to mine the soil for water. Recently,
mechanisms for this differential grth occurrence were proposed based on events
occurring in the cell wall (Wu and Cosgrove, 2000; Wu et al., 1996). Maize roots grown
at low water potential produce more expansin, a cell wall protein that causes extensibility
(discussed in more detail below) compared to roots grown at normal water potential (Wu
and Cosgrove, 2000; Wu et al., 1996). Cell walls of apical elongation zone regions of
low water potential roots were also more responsive to expansin, indicating that structural
changes had taken place in the wall (Wu and Cosgrove, 2000; Wu et al., 1996). This was
in contrast to the basal elongation zone region of the root, which showed less

responsiveness to expansin and in which structural modifications had occurred that

32

caused wall “stiffening”. The exact nature of the structural modifications that occur in
the apical and basal regions is not known, though stiffening has been proposed to arise by
oxidative cross-linking through the action of apoplastic peroxidase (Wu and Cosgrove,

2000).

Some types of environmental responses may involve differential cell wall reinforcement
rather than cell wall extensibility. One such example is the modification of the cell wall
in response to wind stress or other forms of mechanical stress. Plants exposed to wind or
mechanical manipulation may become shorter, stockier, and stronger, which confers an
adaptive advantage to firrther or increased stress (Braam et al., 1997 ; Salisbury and Ross,
1992). Certainly wind stress can be of significant consequence in agriculture, where
plants too fragile to withstand strong winds may be flattened. The morphological pattern
associated with wind or mechanical responses is called thigrnomorphogenesis.
Investigations into the molecular control of thigrnomorphogenesis began when a group of
differentially expressed Arabidopsis genes were identified as being expressed in response
to slight mechanical stimulation such as touch, wind, and water spray (Braam and Davis,
1990). This subset of genes was called T CH (Touch) genes, and several of the cDNAs
were sequenced. Multiple cDNAs encoded calmodulin isoforms, indicating that calcium
acts as a second messenger during mechanical stress signal transduction (Braam and
Davis, 1990; Braam et al., 1997). Another TCH gene, T CH4, encoded a xyloglucan
endotransglycosylase (XET), an enzyme that cleaves xyloglucan chains and ligates them
to adjacent XyG chains (Xu et al., 1995). XETs have been proposed to play a role in

incorporation of nascent XyG into the existing cell wall, possibly affecting cell wall

33

structural properties (Thompson et al., 1997). Thus, it was not perhaps surprising that
mechanical stimulation induced a wall-modifying enzyme. The localization of Tch4
protein in stem tissues of wind-treated plants was consistent with a role in morphological

changes occurring after mechanostimulation (Antosievvicz et al., 1997).

A second type of cell wall reinforcement response occurs in woody plants. Reaction
wood is a type of xylem-rich tissue that is formed in response to mechanical strain
(Salisbury and Ross, 1992). Softwood trees form reaction wood (called compression
wood) on the underside of branches, causing the branch to be propped upright, while
hardwood trees form reaction wood (called tension wood) on the upper side of branches,
pulling the branch towards the tree body (Salisbury and Ross, 1992). Reaction wood may
also be formed when a tree or branch is subjected to strain, such as the artificial strain
imposed by tying a sapling into a U-shaped form or the natural strain that occurs when
tree limbs bear the weight of heavy snow or ice. In all cases, the reaction wood involves
in part alterations to cell wall structure. The alterations differ among the species, but
frequently secondary cell wall thickening is observed, and increased lignification may

occur (Salisbury and Ross, 1992).

A final example of cell wall structural modification in response to the environment
occurs when a plant mounts a defense against attacking pathogens. In response to
invasion by pathogens, plants may alter cell wall structures in an attempt to prevent
spread of the pathogen. Such structural modifications may include deposition of callose

(thought to be a wounding response), lignification, cross-linking of aromatic compounds

34

such as ferulate, and increased incorporation of other phenolic compounds (Hardham and

Mitchell, 1998; Wojtaszek, 2000).

1.4 Cell Wall Events During the Life of a Cell

The cell wall experiences a natural progression during the life of the cell. At appropriate
times, the cell wall is 1) biosynthesized, 2) deposited and assembled, 3) remodeled in
response to endogenous or exogenous conditions, and 4) disassembled during senescence
or cell death. The biosynthesis of the cell wall will be considered in detail below. Little
is actually known about the assembly processes that occur to establish the structure of the
cell wall, and this process will not be considered in detail for this reason. Remodeling of
the wall may occur during growth, and two major proteins that are capable of cell wall
modifications will be reviewed. Because of the nature of this thesis, the process

hydrolysis and degradation of cell wall components will not be reviewed.

1.4.] Plant Cell Wall Polysaccharide Biosyntheer

Polysaccharide biosynthesis in general relies on the specificities of enzymes that link
precursor units into polymers. Thus, polysaccharide structure is not determined as a
primary outcome of gene activity, but as a secondary outcome through action of
carbohydrate-active enzymes (Perrin et al., 2001; Wojtaszek, 2000). Synthesis may be
affected by the presence of biosynthetic enzymes, the order in which these enzymes are

encountered, the affinity and specificities of these enzymes, and the availability of

35

substrate molecules such as sugar nucleotide donors (Perrin et al., 2001). These enzymes

tend to be integral membrane proteins.

Several aspects of plant cell wall polysaccharide biosynthesis are notable. There are two
main sites of synthesis within the cell. Cellulose ([3-(1,4)-glucan) and callose (Ii-(1,3)-
glucan) are synthesized at the plasma membrane with extrusion of the product into the
apoplast, while all other cell wall polysaccharides are synthesized within the Golgi
apparatus and transferred to the apoplast by secretory vesicles. Despite a detailed
knowledge of cell wall structural features, there is a relatively small body of knowledge
about the enzymes that make cell wall polysaccharides, and an even smaller body of
knowledge about the genes that encode these enzymes. In part, this is because these
families of enzymes share higher-order structural features, but may not share extensive
sequence identity. Thus, it has not been particularly fruitful to attempt to identify plant-
specific polysaccharide biosynthesis enzymes by homology to identified enzymes from
animals or microbes. I will define the two main types of cell wall biosynthetic enzymes

and give an overview of existing examples of plant enzymes for each type.

Glycan synthases

Glycan syntheses are processive enzymes devoted to constructing the backbone of
polysaccharides (Perrin et al., 2001). For example, cellulose synthase, callose synthase,
and the glucan synthase that makes the backbone of XyG are all examples of glycan

synthases.

36

Within the past five years, genes encoding the cellulose synthase catalytic subunit have
been identified. This has not been a trivial effort. The lability of rosettes, cellulose-
synthesizing enzyme complexes located in the plasma membrane, has caused difficulty
for biochemical approaches to studying this enzyme. Identification of cellulose synthase
at the molecular level first occurred in the cellulose-producing bacteria Acetobacter
xylinum and A grobacterium tumefaciens. The use of a product entrapment technique first
allowed the purification of a cellulose synthase catalytic subunit enzyme from A. xylinum,
followed by identification of the corresponding gene (reviewed in Dehner, 1999). A
similar gene was subsequently isolated from A. tumefaciens. Though it was hoped that
these bacterial genes might be used as hybridization probes to identify corresponding
plant genes, several attempts to do so were not fruitful (Dehner, 1999). However, a
careful study of the bacterial genes using hydrophobic cluster analysis identified some
highly conserved regions. Within these regions are three conserved aspartate residues
followed by a QXXRW motif (Delmer and Amor, 1995; Saxena et al., 1995). These
features seemed characteristic of cellulose synthases, and more a recent structural study
on a glycosyltransferase from Bacillus subtilus indicated that portions of such regions
may be involved in binding to UDP-glucose (Chamock and Davies, 1999). Candidates
for plant celluloSe synthase genes were derived by random sequencing of cDNA libraries
from developing cotton fibers, collected at a developmental stage during which a massive
amount of cellulose is synthesized. Two candidate genes were identified that had
similarity to the bacterial cellulose synthases and contained the D,D,D,QXXRW motif
(Pear et al., 1996). However, the plant homologs contained inserted regions that were not

present in the bacterial enzymes, explaining why attempts to utilize the bacterial genes as

37

probes had failed. Six transmembrane domains were predicted for the cotton enzymes,
two at the amino terminus and four near the carboxy terminus. The central soluble region
was capable of binding UDP-glucose in a Mg2+-dependent manner, consistent with
previous biochemical studies (Pear et al., 1996). Originally named CelA-I and CeIA-Z,

these genes have recently been designated GhCesA-l and GhCesA-Z (Dehner, 1999).

In vivo support that CesA genes do indeed constitute the cellulose synthase catalytic
subunit rests largely upon genetic evidence. In the first example, the rsz mutant
described above was shown to have a point mutation in a gene homologous to the cotton
CesA genes, causing cell wall abnormalities throughout the plant when grown at elevated
temperatures (Arioli et al., 1998). In the second example, one Arabidopsis mutant
identified on the basis of collapsed vascular (xylem) cell phenotype was found also to
have a mutation in a CesA homolog (Taylor et al., 1999). The irx3 mutant lacks
sufficient cellulose deposition in vascular tissues to withstand the negative pressure
involved in water conductance, and thus its vasculature collapses. Immunological
evidence that CesA is present in rosettes was shown in an elegant microscopy study that
utilized a freeze fracture replica labelling technique, applied to plants for the first time, to
demonstrate that antibodies directed against a cotton CesA could label rosettes of adzuki

bean, a vascular plant (Kimura et al., 1999).

With the sequencing of the Arabidopsis genome now complete, it has become clear that
that there is much genomic redundancy. Many genes appear to be present as members of

families rather than as single copies. This has also been the case for cellulose synthase

38

genes, but to an extreme. There appear to be approximately 40 genes in Arabidopsis that
bear similarity to the original CesA genes, comprising a large superfarnily. This
superfamily has been divided into one family of ten “true” CesA genes and six families of
cellulose synthase-like (Cs!) genes that are less closely related (Richmond and
Somerville, 2000). A web site is maintained by Richmond and Somerville to document
sequence data for cellulose synthase and cellulose synthase-like genes
(http://cellwall.stanford.edu). It is possible that some of the Csl genes may be involved in
synthesis of non-cellulosic glucan polymers, such as the backbone of XyG. Recently,
evidence has been generated to suggest that multiple isoforms of cellulose synthase are
needed in the same cell for the formation of functional dimeric complexes, indicating

why there might be a need for so many CesA isoforms in plants (Perrin, 2001).

Glycosyltransferases

Unlike glycan synthases, glycosyltransferases are non-processive enzymes.
Glycosyltransferases add single sugar units to polysaccharides, resulting in side-chain
structures (Perrin et al., 2001). Typically, these enzymes are type H integral membrane
proteins with an N-tenninal cytoplasmic tail, a signal-anchor transmembrane domain that
spans the Golgi apparatus membrane, and a stem region and C-terminal catalytic domain
located in the lumen of the Golgi (Kleene and Berger, 1993). Two plant
glycosyltransferases involved in synthesis of cell wall polysaccharides have been

identified and will be described.

39

Galactosyltransferases
Legume seeds utilize galactomannans as seed storage polysaccharides. Galactomannan
(GM) consist of a backbone of B-D-(1,4)-Man with substitutions of d—(1,6)-Gal. Using
endosperrn of hand-dissected germinating fenugreek seeds, Edwards et al. solubilized
GM galactosyltransferase (GalT) and demonstrated its activity in an in Vitro assay using
either mannan oligosaccharides with a degree of polymerization of five or more units or
GM with a low level of Gal substitution as exogenous acceptor substrates (Edwards et al.,
1999). Non-denaturing isoelectric focusing combined with an in-gel assay, followed by
SDS-PAGE analysis, indicated that enzyme activity was correlated with a protein of
approximately 51 kD with a pI of 6.5 (Edwards et al., 1999). Amino acid sequence data
was obtained from the 51 kD protein and used to design degenerate oligonucleotide
primers for isolation of the corresponding cDNA by RT-PCR (Edwards et al., 1999).
Confirmation of the cDNA as encoding a functional GM galactosyltransferase was
achieved by expressing the cDNA in the methylotropic yeast Pichia pastoris as a
truncated, secreted protein. This protein was purified from the yeast growth media and

used in the assay, showing high galactosyltransferase activity (Edwards et al., 1999).

F ucosyltransferases
Previous studies had established certain facts about XyG fucosyltransferase (F UT) and
the FUT gene family as a whole. It had been shown by immuno-electron microscopy
using anti-a-(l ,2)-Fuc-Gal antibodies that fucosylated XyG first appears in the lumen of
the trans-Golgi and trans Golgi network before vesicle-mediated secretion to the cell wall

(Moore et al., 1991; Puhlmann et al., 1994). This activity appeared to be spatially distinct

40

from galactosyl- and xylosyltransferase activity (Brummell et al., 1990; Camirand et al.,
1987). The Golgi localization of activity agreed with similar studies of
fucosyltransferases in mammals (e.g Borsig et al., 1996; Kimura et al., 1995;Narimatsu
et al., 1996). FUTs received early attention from the medical community because of their
role in synthesis of glycoprotein blood group antigens (such as the ABO blood group)
and glycolipids (Field and Wainwright, 1995; Mollicone et al., 1995). Some Fuc-
containing antigens and FUT activities also appeared to be correlated with tumor
formation, serving as clinical indicators of certain types of cancer (Bolscher et al., 1989;
Shah Reddy et al., 1982; Staudacher et al., 1999; Weiser and Wilson, 1981). A rash of
cloning and characterization studies in mammalian systems in the early to mid-19905
identified a number of F UTs critical to glycobiological processes. Expression-based
cloning proved to be a valuable technique to identify candidate genes in many cases
(Ernst et al., 1989; Fukuda et al., 1996). Mammalian studies also established the
structure of FUTs, which matches the general structure of glycosyltransferases described
above. The critical nature of the C-terminus in catalysis was illustrated by deletion
studies showing that removal of as little as one amino acid from the C-terminal end could
completely abolish all enzymatic activity, while deletion of as much as one third of the
N-terminus had little effect (Xu et al., 1996). With time, FUTs were also identified in
parasitic organisms such as ulcer-causing bacteria, apparently serving to synthesize
mammalian-type antigens so the pathogen can evade detection by the host (Ge et al.,
1997; Martin et al., 1997). A F UT involved in synthesis of Nod factors in nitrogen-fixing
bacteria was also identified (Quesada et al., 1997; Quinto et al., 1997). None showed

high homology to any plant genomic or cDNA sequences in public databases.

41

Early studies of plant XyG FUT used rapidly elongating regions of etiolated pea epicotyls
as a source of enzyme for partial purification (Hanna et al., 1991). Later studies utilized
solubilized, carbonate-washed pea microsomal membrane proteins followed by GDP
affinity chromatography and size exclusion chromatography (F aik et al., 2000; Perrin et
al., 1999). Chapters two through five of this thesis will describe the purification of XyG
FUTase from pea, identification of a gene encoding XyG FUTase from Arabidopsis and
confirmation of its biochemical function, and analysis of the fucosylation of XyG as a

model cell wall biosynthetic process.

1.4.2 Organization and Remodeling of the Cell Wall

After biosynthesis and deposition of the plant cell wall, remodeling may take place to
affect wall extensibility and thus growth. Two classes of cell wall proteins that are
involved in alterations of wall structure will be described briefly: expansins and

xyloglucan endotransglycosylases (XETs).

Expansins

It has long been observed that plant cell walls can undergo extension upon exposure to
either auxin or acidic solutions, but the mechanism of such extension was not fully
understood (Cleland, 1981). It had been noted that isolated cell walls were capable of
undergoing acid-induced elongation, but lost this property if the walls were subjected to
protein denaturing conditions (Cosgrove, 1998). Thus, it appeared that acid-induced

growth was mediated by a proteinaceous component. McQueen-Mason et al. (1992) used

42

crude protein preparations from cucumber cell walls to determine whether acid-induced
extensibility could be restored to boiled cell wall strips, and identified two proteins that
showed activity in the assay. Neither protein caused detectable hydrolysis of wall

polymers (McQueen-Mason et al., 1992). These proteins were named “expansins”.

Cucumber hypocotyl expansin can promote separation of the pure cellulose microfibrils
found in filter paper, indicating that expansins can disrupt hydrogen bonds between
glucan chains (McQueen-Mason and Cosgrove, 1994). Based on binding studies
indicating that expansin acts at the interface of cellulose and one or more hemicellulose,
it was suggested that the point of action of expansins dicots is the cellulose-xyloglucan
interface, although this speculation has not been confirmed structurally (McQueen-Mason
and Cosgrove, 1995). Since expansins cause mechanical alterations in the interaction
between polymers but do not exhibit hydrolytic properties, they are not thought to be
enzymes in the sense of being able to break or make covalent bonds, but rather are
proteins that alter structural properties of biological components to speed up a biological
activity (extension) by lowering activation energy of the process (McQueen-Mason,

1995)

Expansins comprise a large gene family in plants and have been divided into two major
groups: the (rt-expansins, which was the first group identified, and the B-expansins, which
are more distantly related proteins first observed as major pollen allergens in grass
species (Cosgrove et al., 1997). Beta-expansins do have acid growth-promoting activity

on monocot cell walls but not dicot cell walls, and (Jr-expansins show higher activity on

43

dicot cell walls, indicating that each class of expansin may act on different polymers
(Cosgrove, 1998; Cosgrove et al., 1997). Recently, even more distantly related proteins
have been identified as expansin-related genes and expansin-like genes (Cosgrove,
personal communication.) There are 30-40 or— and B—expansin, expansin-like, and
expansin-related genes in Arabidopsis (http://www.bio.psu.edu/expansins/). Expression
analysis has shown that a variety of expression patterns are seen amongst members of the
gene family, and some show very specific developmental or tissue-specific patterns of
expression (Cho and Cosgrove, 2000). Expansins appear to play a role not only in
elongation growth, but also abscission, fruit ripening, leaf initiation, pollen tube growth,
and function of specific cell types such as guard cells (Cosgrove, 1997b; Reinhardt et al.,
1998). Expansins are plant-specific genes and have been identified throughout the plant
kingdom in organisms ranging from ferns to gyrnnospenns to monocots and dicots (Cho

and Kende, 1997; Cosgrove, 2000b; Kim et al., 2000; Shcherban et al., 1995).

X yloglucan endotransglycosylases (XE Ts)

XETs are enzymes that cleave XyG chains internally and transfer them to other XyG
chains (Fry et al., 1992). The accepting chain may be either a polysaccharide or an
oligosaccharide. XETs are specific for XyG and do not cleave other glucans, and thus
are distinct from endocellulases (Fry et al., 1992). Because the products of this reaction
are the same as the reactants, tracking the activity of XETs has been challenging,
although several assays have been developed to do so (Fry et al., 1992; Sulova et al.,
1995). Some XET isoforms have been reported to be induced by auxin treatment, and a

physiological role for XET has been proposed during fi'uit ripening and responses to wind

44

stress (Catala et al., 1997 ; Purugganan et al., 1997; Schroder et al., 1998). A maize XET
has also been reported to be induced by flooding and ethylene, and may be involved in
formation of root aerenchyma (Saab and Sachs, 1996). A localized increase in XET
activity has also been detected at the sites of root hair formation in Arabidopsis

(V issenberg et al., 2001).

XETs have been proposed to play a fundamental role in modulating cell wall
extensibility, but application of XET to boiled cell wall preparations showed that it was
insufficient to promote elongation under acidic conditions (McQueen-Mason et al.,
1993). Therefore, it seems more likely that XETs serve to integrate newly formed XyG
into existing chains in the cell wall, perhaps serving to modulate wall properties in the
process. Rather extensive biochemical characterization of XETs has been conducted.
The enzyme forms a fairly stable glycosyl-enzyme intermediate with polymeric XyG that
may be disrupted upon addition of XyG oligosaccharides, an event which has allowed
development of simple enzyme purification strategies (Steele and Fry, 1999; Sulova et
al., 1998). Like expansins, XETs form large gene families in plant genomes. Substrate
preference and catalytic properties have been evaluated for different XET isomers, and it
has been suggested that different family members serve different physiological roles

(Steele and Fry, 2000; Steele et al., 2001).

45

1.5 Approaches to Identify and Evaluate Cell Wall Biosynthetic Enzymes and
Candidate Genes

It is important to note that among the full complement of polysaccharides present in the
plant cell wall, there are hundreds of different linkages. In general, each linkage is
thought to be directed by a specific enzyme. Although there have been occasional
examples of multifunctional enzymes that can catalyze two different linkages, they are
the exception rather than the rule (e.g. Lee and Spicer, 2000; Nguyen et al., 1998). Thus,
there are expected to be hundreds of genes in plant genomes that are devoted to the
synthesis of the plant cell wall (V arner and Lin, 1989). One recent review estimated that
there are over 1,000 genes in the Arabidopsis genome that are involved in making,
assembling, and remodeling the cell wall and conferring specific properties to it (Roberts,
2001). A database that catalogs known and predicted carbohydrate-active enzymes
(http://afmb.cnrs-mrs.fi'/~pedro/CAZY/db.html) shows that Arabidopsis has at least 140
Arabidopsis ORFs within families predicted to be involved in polysaccharide and
glycoprotein biosynthesis; this estimation does not include the many enzymes that
conjugate sugar units to small metabolites such as hormones (Perrin et al., 2001).
However, while it is clear that much of the plant genome is devoted to synthesis of the
cell wall, it has been challenging to identify these genes and corresponding enzymes with

precision and confidence in their functional roles.
Several different approaches have been or may be taken to identify cell wall biosynthetic

genes and enzymes. One approach is to purify a cell wall biosynthetic enzyme to relative

homogeneity, obtain amino acid sequence from the purified protein, and use this

46

information to identify the corresponding gene. Ideally, the role of the enzyme encoded
by the gene will be confirmed by a functional test. The major advantage of this approach
is that, in general, the results can be definitive. However, biochemical purification
projects are inherently low-throughput, as each problem must often be approached anew.
In addition, cell wall polysaccharide biosynthetic enzymes are frequently labile and lose
activity upon manipulation (Perrin et al., 2001). This approach will also not be amenable
to enzymes that require multiple interacting partners for activity, unless the entire
complex can be co-purified. A suitable enzymatic assay is needed in order to track
activity of the enzyme of interest during the purification procedure, and this may be a
limiting factor when working with cell wall enzymes. The complicated structures of cell
wall components may make it difficult to obtain suitable acceptor substrates of defined
structure, and in some cases the sugar nucleotide donor may not be available
commercially (Perrin et al., 2001). Despite these limitations, two cell wall biosynthetic
genes have been identified using this approach; one is the subject of this thesis. These

projects were described in the proceeding section.

A second approach towards identifying cell wall biosynthetic genes is “forward”
genetics, so termed because a mutation characterizing a phenotype is identified first and
then the gene responsible for the phenotype is identified by positional cloning (that is, the
researcher moves from phenotype forward to gene identification.) This approach has
been successful for the identification of genes encoding enzymes that are members of
functional complexes (e.g. Arioli et al., 1998). Forward genetics approaches require a

feasible screening strategy in which mutant phenotypes are identified. Screens are

47

usually designed to be as labor-free as possible in order to screen large populations of
mutagenized plants quickly. However, mutations in essential genes will result in lethal
phenotypes, and it may be expected that certain members of the cell wall biosynthetic
machinery may be essential. This requires the identification of conditional alleles.
Despite these limitations, several screens designed to identify cell wall biosynthetic genes
have been conducted, and a number of genes have cloned as a result (Arioli et al., 1998;

Reiter et al., 1997; Turner and Somerville, 1997).

Tools have begun to emerge for an additional approach. Functional genomics strategies
identify candidate genes as possibly encoding cell wall biosynthetic enzymes.
Candidates are identified using bioinformatics and expression profiling techniques,
making use of nucleotide and protein databases and information from expressed
sequenced tagged (EST) sequencing projects or microarray experiments. Candidate
genes are then further studied, often by analyzing the effects of mutations in these genes
(in an approach called “reverse genetics”, because the researcher moves from gene
towards gene firnction) or by expressing the gene in a heterologous system and assaying
the recombinant enzyme for biochemical function. The functional genomics approach
requires a baseline of knowledge for execution, and in the case of cell wall biosynthetic
genes this foundation is still being established. Knowledge of similar enzymes is helpful,
as is an understanding of motifs and protein structural features that are necessary for
activity. Likewise, knowledge of expression patterns of related genes is helpful. The
advantage of the functional genomics method is that, unlike the other two methods

described above, there is potential for high-throughput or moderately-high throughput

48

approaches. However, there are limitations. In the area of cell wall biosynthesis, the
baseline of knowledge is still relatively small, which can make selection of gene
candidates more challenging. Reverse genetic approaches can be of limited use if there is
insufficient information as to the role of the gene. For example, a candidate gene may be
identified as having a UDP—arabinose binding motif, and a mutant is obtained in hopes
that the gene will affect synthesis of a particular arabinogalactan. If the mutant shows no
alteration in structure for this arabinogalactan, though, the researcher is then faced with
having to systematically analyze all structures containing arabinose — and there is the
possibility that, although the candidate protein appeared to have a UDP-Ara binding
motif, it actually binds to another sugar nucleotide altogether. Additionally, the
redundancy that has been observed in several plant genomes (including Arabidopsis) can
hamper reverse genetic analysis, as mutations affected in all members of the gene family
may have to be obtained in order to observe a phenotypic effect. Finally, even the
functional genomics approach may require an enzymatic assay for definitive evaluation
of a candidate protein, bringing up the limitations associated with assays mentioned
above. It is clear that biochemical techniques will continue to be relevant in the “post-
genomic” era. Despite these drawbacks, however, the functional genomics approach may
be a feasible method for identifying genes devoted to cell wall synthesis -- particularly as
the foundation of knowledge about these genes and enzymes continues to become more

robust.

49

1.6 Conclusions

Plant protoplasts are encased in cell walls that are indispensable for normal growth,
development, and interactions within the plant itself and the rest of its surroundings. The
wall provides positional context to the plant body, permits and delimits growth, and
mediates communication both within the plant and between the plant and its
environment. Plant cell walls are composed of a complement of polysaccharide
molecules, structural proteins that may themselves be glycosylated, and phenolic
compounds. Cell wall polysaccharide synthesis occurs by action of enzymes such as
glycan synthases and glycosyltransferases. A major goal in the field of plant cell wall
biology is to identify all genes encoding enzymes devoted to synthesis of cell wall
polysaccharides, but thus far only a few have been identified unequivocally. This thesis
is focused upon the identification of an enzyme involved in the synthesis of a plant cell

wall polysaccharide, xyloglucan fucosyltransferase.

References

Aldington, S., McDougall, G. J. and Fry, S. C. (1991). Structure-activity relationships
of biologically active oligosaccharides. Plant, Cell and Environment 14, 625-636.

An, J., Zhang, L., O'Neill, M. A., Albersheim, P. and Darvill, A. G. (1994). Isolation
and structural characterization of endo-rhamnogalacturonase—generated fragments of the
backbone of rhamnogalacturonan I. Carbohydr Res 264, 83-96.

Antosiewicz, D. M., Purugganan, M. M., Polisensky, D. H. and Bream, J. (1997).
Cellular localization of Arabidopsis xyloglucan endotransglycosylase- related proteins
during development and after wind stimulation. Plant Physiol. 115, 1319-1328.

Arioli, T., Peng, L., Betzner, A. 8., Burn, J., Wittke, W., Herth, W., Camilleri, C.,

Hofte, H., Plazinski, J., Birch, R. et a1. (1998). Molecular analysis of cellulose
biosynthesis in Arabidopsis. Science 279, 717-720.

50

Assaad, F. F. (2001). Plant cytokinesis: exploring the links. Plant Physiology 126, 509-
516.

Assaad, F. F., Mayer, U., Wanner, G. and Jurgens, G. (1996). The KEULE gene is
involved in cytokinesis in Arabidopsis. Molecular and General Genetics 253, 267-277.

Atalla, R. H. and VanderHart, D. L. (1984). Native cellulose: a composite of two
distinct crystalline forms. Science 223, 283-285.

Augur, C., Benhamou, N., Darvill, A. and Albersheim, P. (1993). Purification,
characterization, and cell wall localization of an alpha-fucosidase that inactivates a
xyloglucan oligosaccharin. Plant J 3, 415-26.

Augur, C., Stiefel, V., Darvill, A., Albersheim, P. and Puigdomenech, P. (1995).
Molecular cloning and pattern of expression of an alpha-L-fucosidase gene from pea
seedlings. J Biol Chem 270, 24839-43.

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

Bolscher, J. G., Bruyneel, E. A., Van Rooy, H., Schallier, D. C., Marcel, M. M. and
Smets, L. A. (1989). Decreased fucose incorporation in cell surface carbohydrates is
associated with inhibition of invasion. Clin Exp Metastasis 7, 557-69.

Bolwell, G. P. and Northcote, D. H. (1981). Control of hemicellulose and pectin
synthesis during differentiation of vascular tissue in bean (Phaseolus vulgaris) callus and
in bean hypocotyl. Planta 152, 225-233.

Bolwell, G. P. and Northcote, D. H. (1983). Induction by growth factors of
polysaccharide synthases in bean cell suspension cultures. Biochem J 210, 509-515.

Borsig, L., Kleene, R., Dinter, A. and Berger, E. G. (1996). Irnmunodetection of alpha
1-3 fucosyltransferase (F ucT-V). Eur J Cell Biol 70, 42-53.

Braam, J. and Davis, R. W. (1990). Rain-, wind-, and touch-induced expression of
calmodulin and calmodulin-related genes in Arabidopsis. Cell 60, 357-364.

Braam, J., Sistrunk, M. L., Polisensky, D. H., Xu, W., Purugganan, M. M.,
Antosiewicz, D. M., Campbell, P. and Johnson, K. A. (1997). Plant responses to
environmental stress: regulation and functions of the Arabidopsis TCH genes. Planta 203
Suppl, 835-41.

Bret-Harte, M. S., Baskin, T. I. and Green, P. B. (1991). Auxin stimulates both

deposition and breakdown of material in the pea outer epidermal cell wall, as measured
interferometrically. Planta 185, 462-471.

51

Brett, C. T. and Waldron, K. (1996). Physiology and biochemistry of plant cell walls.
London: Chapman and Hall."

Brummell, D. A., Camihand, A. and Maclachlan, G. A. (1990). Differential
distribution of xyloglucan dansferases in pea Golgi dictyosomes and secretory vesicles.
Journal of Cell Science 96, 705-710.

Brummell, D. A. and Hall, J. L. (1983). Regulation of cell wall synthesis by auxin and
fusicoccin in different tissues of pea stem segments. Physiologia Plantarum 59, 627-634.

Buckeridge, M. S., Vergara, C. E. and Carpita, N. C. (2001). Insight into multi-site
mechanisms of glycosyl transfer in (1-->4)beta-D-glycans provided by the cereal mixed-
linkage (1-->3),(1-->4)beta-D-glucan synthase. Phytochemistry 57, 1045-1053.

Camirand, A., Brummell, D. A. and MacLachlan, G. A. (1987). Xyloglucan
fucosyltransferase activity is localized in golgi dictyosomes.

Carpita, N. C. and Gibeaut, D. M. (1993). Structural models of primary cell walls in
flowering plants: consistency of molecular structure with the physical properties of the
walls during growth. PlantJ. 3, 1-30.

Carpita, N. C., Sabularse, D., Montezinos, D. and Delmer, D. P. (1979).
Determination of the pore size of cell walls of living plant cells. Science 205, 1144-1147.

Catala, C., Rose, J. K. and Bennett, A. B. (1997). Auxin regulation and spatial
localization of an endo-l ,4-beta-D-glucanase and a xyloglucan endotransglycosylase in
expanding tomato hypocotyls. Plant j 12, 417-26.

Chamock, S. J. and Davies, G. J. (1999). Structure of the nucleotide-diphospho-sugar

transferase, SpsA from Bacillus subtilis, in native and nucleotide-complexed forms.
Biochemistry 38, 6380-63 85.

Cho, H. T. and Kende, H. (1997). Expansins in deepwater rice intemodes. Plant
Physiology 113, 1137-1143.

Cleland, R. E. (1981). Wall extensibility: hormones and wall extension. In Encyclopedia
of Plant Physiology, New Series, vol. 13B (ed. W. Tanner and F. A. Loewus), pp. 255-
273. Berlin: Springer-Verlag.

Cooper, J. B., Heuser, J. E. and Varner, J. E. (1994). 3,4-Dehydroproline inhibits cell
wall assembly and cell division in tobacco protoplasts. Plant Physiology 104, 747-752.
Cosgrove, D. J. (1997a). Assembly and enlargement of the primary cell wall in plants.
Annu. Rev. Cell Dev. Biol. 13, 171-201.

Cosgrove, D. J. (1997b). Crreping walls, softening fi'uit, and penetrating pollen tubes:
the growing role of expansins. Proc Natl Acad Sci U S A 94, 5504-5505.

52

Cosgrove, D. J. (1998). Cell wall loosening by expansins. Plant Physiology 118, 333-
339.

Cosgrove, D. J. (2000a). Expansive growth of plant cell walls. Plant Physiology and
Biochemistry 38, 109-124.

Cosgrove, D. J. (2000b). New genes and new biological roles for expansins. Current
Opinion in Plant Biology 3, 73-78.

Cosgrove, D. J., Bedinger, P. A. and Durachko, D. M. (1997). Group I allergens of
grass pollen as cell wall loosening agents. Proc Natl Acad Sci U S A 94, 6559-6564.

Delmer, D. (1999). Cellulose biosynthesis: exciting times for a difficult field of study.
Annual Review of Plant Physiology and Plant Molecular Biology 50, 245-276.

Delmer, D. P. and Amor, Y. (1995). Cellulose biosynthesis. Plant Cell 7, 987-1000.

Edwards, M. E., Dickson, C. A., Chengappa, S., Sidebottom, C., Gidley, M. J. and
Reid, J. S. (1999). Molecular characterisation of a membrane-bound

galactosyltransferase of plant cell wall matrix polysaccharide biosynthesis. Plant Journal
19, 691-7.

Eriksson, 1., Andersson, R., Westerlund, E., Andersson, R. and Anna, P. (1996).
Structural features of an arabinan fragment isolated from the water-soluble fraction of
dehulled rapeseed. Carbohydr Res 281, 161-172.

Ernst, L. K., Rajan, V. P., Larsen, R. D., Ruff, M. M. and Lowe, J. B. (1989). Stable
expression of blood group H determinants and GDP-L-fucose: beta-D-galactoside 2-
alpha-L-fucosyltransferase in mouse cells after transfection with human DNA. J Biol
Chem 264, 3436-47.

Faik, A., Bar Peled, M., DeRocher, A. E., Zeng, W., Perrin, R. M., Wilkerson, C.,
Raikhel, N. V. and Keegstra, K. (2000). Biochemical characterization and molecular
cloning of an alpha-1,2-fircosyltransferase that catalyzes the last step of cell wall
xyloglucan biosynthesis in pea. J Biol Chem 275, 15082-9.

Field, M. C. and Wainwright, L. J. (1995). Molecular cloning of eukaryotic
glycoprotein and glycolipid glycosyltransferases: a survey. Glycobiology 5, 463-72.

Fowler, T. J., Bernhardt, C. and Tierney, M. L. (1999). Characterization and

expression of four proline-rich cell wall protein genes in Arabidopsis encoding two
distinct subsets of multiple domain proteins. Plant Physiology 121, 1081-1091.

Fry, S. C., Aldington, S., Hetherington, P. R. and Aitken, J. (1993a). Oligosaccharides
as signals and substrates in the plant cell wall. Plant Physiol. 103, 1-5.

53

Fry, S. C., McDougall, G. J., Lorences, E. P., Biggs, K. J. and Smith, R. C. (1990).
Oligosaccharins from xyloglucan and cellulose: modulators of the action of auxin and H+
on plant growth. Symp Soc Exp Biol 44, 285-98.

Fry, S. C., Smith, R. C., Renwick, K. F., Martin, D. J., Hodge, S. K. and Matthews,
K. J. (1992). Xyloglucan endotransglycosylase, a new wall- -loosening enzyme activity
from plants. Biochem J 282, 821-8.

Fry, S. C., York, W. S., Albersheim, P., Darvill, A., Hayashi, T., Joseleau, J.-P.,
Kato, Y., Lorences, E. P., Maclachlan, G. A., McNeil, M. et al. (1993b). An

unambiguous nomenclature for xyloglucan-derived oligosaccharides. Physiol. Plant. 89,
1-3.

Fujino, T., Sone, Y., Mitsuishi, Y. and Itoh, T. (2000). Characterization of cross-links
between cellulose microfibrils, and their occurrence during elongation growth in pea
epicotyl. Plant Cell Physiol 41, 486-94.

Fukuda, M., Bierhuizen, M. F. and Nakayama, J. (1996). Expression cloning of
glycosyltransferases. Glycobiology 6, 683- 9.

Ge, Z., Chan, N. W., Palcic, M. M. and Taylor, D. E. (1997). Cloning and
heterologous expression of an alphal,3-fi1cosyltransferase gene from the gastric pathogen
Helicobacter pylori. J Biol Chem 272, 21357- 63.

Ga, X. and Verma, D. P. (1996). Phragrnoplastin, a dynamin-like protein associated
with cell plate formation in plants. EMBO J 15, 695-704.

Guillen, R., York, W. S., Pauly, M., An, J., Impallomeni, G., Albersheim, P. and
Darvill, A. G. (1995). Metabolism of xyloglucan generates xylose-deficient
oligosaccharide subunits of this polysaccharide in etiolated peas. Carbohydr Res 277,
291-311.

Hanna, R., Brummell, D. A., Camirand, A., Hensel, A., Russell, E. F. and
Maclachlan, G. A. (1991). Solubilization and properties of GDP-fucose: xyloglucan 1,2-

alpha-L-fucosyltransferase from pea epicotyl membranes. Arch Biochem Biophys 290, 7-
13.

Hardham, A. R. and Mitchell, H. J. (1998). Use of molecular cytology to study the
structure and biOlogy of phytopathogenic and mycorrhizal fungi. Fungal Genetics and
Biology 24, 252-284.

Hayashi, T. (1989). Xyloglucans in the primary cell wall. Annual Review of Plant
Physiology and Plant Molecular Biology 40, 139-168.

54

Hayashi, T., Ogawa, K. and Mitsuishi, Y. (1994a). Characterization of the adsorption
of xyloglucan to cellulose. Plant Cell Physiol. 35, 1199-1205.

Hayashi, T., Takeda, T., Ogawa, K. and Mitsuishi, Y. (1994b). Effects of the degree of
polymerization on the binding of xyloglucans to cellulose. Plant Cell Physiol. 35, 893-
899. ‘

Herve du Penhoat, C., Gey, C., Pellerin, P. and Perez, S. (1999). An NMR solution
study of the mega-oligosaccharide, rhamnogalacturonan II. J Biomol NMR 14, 253-271.

Hisamatsu, M., Impallomeni, G., York, W. S., Albersheim, P. and Darvill, A. G.
(1991). A new undecasaccharide subunit of xyloglucans with two alpha-L-fucosyl
residues. Carbohydr.Res. 211, 117-129.

Holland, N., Holland, D., Helentjaris, T., Dhugga, K. S., Xoconostle-Cazares, B. and
Delmer, D. (2000). A comparative analysis of the plant cellulose synthase (CesA) gene
family. Plant Physiology 123, 1313-1323.

Hong, J. C., Nagao, R. T. and Key, J. L. (1989). Developmentally regulated expression
of soybean proline-rich cell wall protein genes. Plant Cell 1, 947-943.

Hsiao, T. C. and Bradford, K. J. (1983). Physiological consequences of cellular water
deficits. In Limitations to Efi’icient Water Use in Crop Production, (ed. H. M. Taylor W.
R. Jordan and T. R. Sinclair), pp. 227-265. Madison, WI: American Society of
Agronomy. '

Ishii, T. and Matsunaga, T. (2001). Pectic polysaccharide rhamnogalacturonan II is
covalently linked to homogalacturonan. Phytochemistry 57, 969-974.

Itoh, T. and Ogawa, T. (1993). Molecular architecture of the cell wall of poplar cells in
suspension culture, as revealed by rapid-freezing and deep-etching techniques. Plant Cell
Physiol. 34, 1 187-1196.

John, M., Rohrig, H., Schmidt, J., Walden, R. and Schnell, J. (1997). Cell signalling
by oligosaccharides. Trends in Plant Sciences 2, 111-115.

Kang, B. H., Busse, J. S., Dickey, C., Rancour, D. M. and Bednarek, S. Y. (2001).
The arabidopsis cell plate-associated dynamin-like protein, ADLlAp, is required for
multiple stages of plant growth and development. Plant Physiology 126, 47-68.

Keegstra, K., Talmadge, K. W., Bauer, W. D. and Albersheim, P. (1973). The
Structure of Plant Cell Walls. HI. A Model of the Walls of Suspension-cultured

Sycamore Cells Based on the Interconnections of the Macromolecular Components.
Plant Physiology 51, 188-196.

55

Kieliszewski, M. J. and Lamport, D. T. A. (1994). Extensin: Repetitive motifs,
functional sites, post-translational codes, and phylogeny. Plant J. 5, 157-172.

Kim, J. H., Cho, H. T. and Kende, H. (2000). Alpha-expansins in the semiaquatic ferns
Marsilea quadrifolia and Regnellidium diphyllum: evolutionary aspects and physiological
role in rachis elongation. Planta 212, 85-92.

Kimura, H., Kudo, T., Nishihara, S., Iwasaki, H., Shinya, N., Watanabe, R., Honda,
H., Takemura, F. and Narimatsu, H. (1995). Murine monoclonal antibody recognizing
human alpha(1,3/1,4) fucosyltransferase. Glycoconjugate Journal 12, 802-812.

Kimura, S., Laosinchai, W., Itoh, T., Cui, X., Linder, C. R. and Brown, R. M.
(1999). lmmunogold labeling of rosette terminal cellulose-synthesizing complexes in the
vascular plant vigna angularis. Plant Cell 11, 2075-2086.

Kleene, R. and Berger, E. G. (1993). The molecular and cell biology of
glycosyltransferases. Biochim Biophys Acta 1154, 283-325.

Lee, J. Y. and Spicer, A. P. (2000). Hyaluronan: a multifunctional, megaDalton, stealth
molecule. Current Opinion in Cell Biology 12, 581-586.

Levy, 8., Maclachlan, G. and Staehelin, L. A. (1997). Xyloglucan sidechains modulate

binding to cellulose during in vitro binding assays as predicted by conformational
dynamics simulations. Plant J. 11, 373-386.

Levy, 8., York, W. S., Stuike Prill, R., Meyer, B. and Staehelin, L. A. (1991).
Simulations of the static and dynamic molecular conformations of xyloglucan. The role
of the fucosylated sidechain in surface-specific sidechain folding. Plant J 1, 195-215.

Lukowitz, W., Mayer, U. and Jurgens, G. (1996). Cytokinesis in the Arabidopsis
embryo involves the syntaxin-related KN OLLE gene product. Cell 84, 61-71.

Lukowitz, W., Nickle, T. C., Meinke, D. W., Last, R. L., Conklin, P. L. and
Sommerville, C. R. (2001). Arabdiopsis cytl mutants are deficient in a mannose-l-
phosphate guanylyltransferase and point to a requirement of N-linked glycosylation for
cellulose biosynthesis. Proc Natl Acad Sci U S A 98, 2262-2267.

Martin, S. L., Edbrooke, M. R., Hodgman, T. C., van den Eijnden, D. H. and Bird,
M. I. (1997). Lewis X biosynthesis in Helicobacter pylori. Molecular cloning of an
alpha(l,3)-fircosyltransferase gene. J Biol Chem 272, 21349-56.

McCann, M. C., Bush, M., Milioni, D., Sado, P., Stacey, N. J., Catchpole, G.,
Defernez, D., Carpita, N. C., Hofte, H., Ulvskov, P. et al. (2001). Approaches to
understanding the functional architecture of the plant cell wall. Phytochemistry 57, 811-
821.

56

McCann, M. C. and Carpita, N. (2000). The cell wall. In Biochemistry and Molecular
Biology of Plants, (ed. B. B. Buchanan W. Gruissem and R. L. Jones), pp. 52-108.
Rockville, MD, U.S.A.: American Society of Plant Physiologists.

McCann, M. C. and Roberts, K. (1994). Changes in cell wall architecture during cell
elongation. J.Exp. Bot. 45 Suppl., 1683-1691.

McCann, M. C., Wells, B. and Roberts, K. (1990). Direct visualization of cross-links in
the primary cell wall. J Cell Science 96, 323-334.

McDougall, G. J. and Fry, S. C. (1989). Anti-auxin activity of xyloglucan
oligosaccharides: the role of groups other than the terminal alpha-L-fucose residue.
Journal of Experimental Botany 40, 233-238.

McNeil, M., Darvill, A. G., Fry, S. C. and Albersheim, P. (1984). Structure and
function of the primary cell walls of plants. Annu Rev Biochem 53, 625-63.

McQueen-Mason, S. and Cosgrove, D. J. (1994). Disruption of hydrogen bonding
between plant cell wall polymers by proteins that induce wall extension.
Proc.Natl.Acad.Sci. USA 91, 6574-6578.

McQueen-Mason, S. and Cosgrove, D. J. (1995). Expansin mode of action on cell
walls: analysis of wall hydrolysis, stress relaxation, and binding. Plant Physiology 107,
87-100. '

McQueen-Mason, S., Durachko, D. M. and Cosgrove, D. J. (1992). Two endogenous
proteins that induce cell wall extension in plants. Plant Cell 4, 1425-1433.

McQueen-Mason, S. J. (1995). Expansins and cell wall expansion. J.Exp.Bot. 46, 1639-
1650.

McQueen-Mason, S. J., Fry, S. C., Durachko, D. M. and Cosgrove, D. J. (1993). The
relationship between xyloglucan endotransglycosylase and in-vitro cell wall extension in
cucumber hypocotyls. Planta 190, 327-31.

Meyer, Y. and Abel, W. D. (1975). Importance of the cell wall for cell division and in
the activity of the cytoplasm in cultured tobacco protoplasts. Planta 123, 33-40.

Mohnen, D. (1999). Biosynthesis of pectins and galactomannans. In Carbohydrates and
their derivatives including tannins, cellulose, and related lignins, vol. 3 (ed. P. B. M.),
pp. 497-527. Amsterdam: Elsevier.

Mollicone, R., Cailleau, A. and Oriol, R. (1995). Molecular genetics of H, Se, Lewis
and other fucosyltransferase genes. Transfits Clin Biol 2, 235-42.

57

Moore, P. J., Swords, K. M., Lynch, M. A. and Staehelin, L. A. (1991). Spatial
organization of the assembly pathways of glycoproteins and complex polysaccharides in
the Golgi apparatus of plants. J Cell Biol 112, 589-602.

Narimatsu, H., Iwasaki, H., Nishihara, S., Kimura, H., Kudo, T., Yamauchi, Y. and
Hirohashi, S. (1996). Genetic evidence for the Lewis enzyme, which synthesizes type-1

Lewis antigens in colon tissue, and intracellular localization of the enzyme. Cancer Res
56, 330-8.

Nguyen, A. T., Holmes, E. H., Whitaker, J. M., Ho, 8., Shetterly, S. and Macher, B.
A. (1998). Human alpha1,3/4-fi1cosyltransferases. 1. Identification of amino acids
involved in acceptor substrate binding by site-directed mutagenesis. Journal of Biological
Chemistry 273, 25244-25249.

Nickle, T. and Meinke, D. W. (1998). A cytokinesis-defective mutant of Arabidopsis
(cytl) characterized by embryonic lethality, incomplete walls, and excessive callose
accumulation. Plant Journal 15, 321 -332.

O'Neill, M. A., Warrenfeltz, D., Kates, K., Pellerin, P., Doco, T., Darvill, A. G. and
Albersheim, P. (1996). Rhamnogalacturonan-H, a pectic polysaccharide in the walls of
growing plant cell, forms a dimer that is covalently cross-linked by a borate ester. In vitro
conditions for the formation and hydrolysis of the dimer. J Biol Chem 271, 22923-30.

Pauly, M., Albersheim, P., Darvill, A. and York, W. S. (1999). Molecular domains of
the cellulose/xyloglucan network in the cell walls of higher plants. Plant Journal 20, 629-
639.

Pear, J. R., Kawagoe, Y., Schreckengost, W. E., Delmer, D. P. and Stalker, D. M.
(1996). Higher plants contain homologs of the bacterial ceIA genes encoding the catalytic
subunit of cellulose synthase. Proc.Natl.Acad. Sci. USA 93, 12637-12642.

Perrin, R. M. (2001). Cellulose: how many cellulose synthases to make a plant? Current
Biology 11, R213-216.

Perrin, R. M., DeRocher, A. E., Bar Peled, M., Zeng, W., Norambuena, L., Orellana,
A., Raikhel, N. V. and Keegstra, K. (1999). Xyloglucan fucosyltransferase, an enzyme
involved in plant cell wall biosynthesis. Science 284, 1976-9.

Perrin, R. M. Wilkerson, C. and Keegstra, K. (2001). Golgi enzymes that synthesize
plant cell wall polysaccharides: finding and evaluating candidates in the genomic era.
Plant Molecular Biology 47, 115130.

Puhlmann, J., Bucheli, E., Swain, M. J., Dunning, N., Albersheim, P., Darvill, A. G.
and Hahn, M. G. (1994). Generation of monoclonal antibodies against plant cell-wall
polysaccharides. 1. Characterization of a monoclonal antibody to a terminal alpha-(1--
>2)-linked fucosyl-containing epitope. Plant Physiol 104, 699-710.

58

Purugganan, M. M., Braam, J. and Fry, S. C. (1997). The Arabidopsis TCH4
xyloglucan endotransglycosylase. Substrate specificity, pH optimum, and cold tolerance.
Plant Physiol 115, 181-190.

Quesada, V. D., Fellay, R., Nasim, T., Viprey, V., Burger, U., Prome, J. C.,
Broughton, W. J. and Jabbouri, S. (1997). Rhizobiurn sp. strain NGR234 NodZ protein
is a fircosyltransferase. Journal of Bacteriology 179, 5087-5093.

Quinta, C., Wijfjes, A. H. M., Bloemberg, G. V., Blok Tip, L., Lopez Lara, I. M.,
Lugtenberg, B. J., Thomas Oates, J. E. and Spaink, H. P. (1997). Bacterial nodulation
protein NodZ is a chitin oligosaccharide fucosyltransferase which can also recognize
related substrates of animal origin. Proc Natl Acad Sci U S A 94, 4336-41.

Reinhardt, D., Wittwer, F., Mandel, T. and Kuhlemeier, C. (1998). Localized
upregulation of a new expansin gene predicts the site of leaf formation in the tomato
meristem. Plant Cell 10, 1427-37.

Reiter, W. D., Chapple, C. and Somerville, C. R. (1997). Mutants of Arabidopsis
thaliana with altered cell wall polysaccharide composition. Plant J. 12, 335-345.

Richmond, T. A. and Somerville, C. R. (2000). The cellulose synthase superfamily.
Plant Physiology 124, 495-498.

Rizk, S. E., Abdel-Massih, R. M., Baydoun, E. A. H. and Brett, C. T. (2000). Protein-
and pH-dependent binding of nacent pectin and glucuronoarabinoxylan to xyloglucan in
pea. Planta 211, 423-429.

Roberts, K. (2001). How the cell wall acquired a cellular context. Plant Physiology 125,
127-130.

Saab, I. N. and Sachs, M. M. (1996). A flooding-induced xyloglucan endo-
transglycosylase homolog in maize is responsive to ethylene and associated with
aerenchyma. Plant Physiol 112, 385-91. '

Sachetto-Martins, G., Franco, L. 0. and de Oliveira, D. E. (2000). Plant glycine-rich
proteins: a family or just proteins with a common motif? Biochim Biophys Acta 1492, 1-
14.

Salisbury, F. B. and Ross, C. W. (1992). Plant Physiology. Behnont, California:
Wadsworth.

Saxena, I. M., Brown, R. M. J., Fevre, M., Geremia, R. A. and Henrissat, B. (1995).

Multidomain architecture of beta-glycosyl transferases: implications for mechanism of
action. J Bacteriol 177, 1419-24.

59

Schroder, R., Atkinson, R. G., Langenkamper, G. and Redgwell, R. J. (1998).
Biochemical and molecular characterisation of xyloglucan endotransglycosylase from
ripe kiwifruit. Planta 204, 242-51.

Shah Reddy, 1., Kessel, D. H., Chou, T. H., Mirchandani, I. and Khilanani, U.
(1982). Plasma fucosyltransferase as an indicator of imminent blastic crisis. Am J
Hematol 12, 29-37.

Shcherban, T. Y., Shi, J., Durachko, D. M., Guiltinan, M. J., McQueen-Mason, S. J.,
Shieh, M. and Cosgrove, D. J. (1995). Molecular cloning and sequence analysis of
expansins - A highly conserved, multigene family of proteins that mediate cell wall
extension in plants. Proc.Natl.Acad Sci. USA 92, 9245-9249.

Shedletzky, E., Shmuel, M., Delmer, D. and Lamport, D. T. (1990). Adaptation and
growth of tomato cells on the herbicide 2,6-dichlorobenzonitrile leads to production of

unique cell walls virtually lacking a cellulose-xyloglucan network. Plant Physiology 100,
120-130.

Shedletzky, E., Shmuel, M., Trainin, T., Kalman, S. and Delmer, D. (1992). Cell wall
structure in cells adapted to growth on the cellulose- synthesis inhibitor 2,6-
dichlorobenzonitrile. A comparison between two dicotyledonous plants and a
grarninaceous monocot. Plant Physiol. 100, 120-130.

Showalter, A. M. (1993). Structure and function of plant cell wall proteins. Plant Cell 5,
9-23.

Silk, W. K., Walker, R. C. and Labavitch, J. (1984). Uronide deposition rates in the
primary root of Zea mays. Plant Physiol 74, 721-726.

Sims, I. M., Craik, D. J. and Bacic, A. (1997). Structural characterisation of
galactoglucomannan secreted by suspension-cultured cells of Nicotiana plumbaginifolia.
Carbohydr Res 303, 79-92.

Smith, L. G. (1999). Divide and conquer: cytokinesis 1n plant cells. Current Opinion in
Plant Biology 2, 447-453.

Staudacher, E., Altmann, E, Wilson, I. B. and Man, L. (1999). Fucose in N-glycans:
from plant to man. Biochim Biophys Acta 1473, 216-36.

Steele, N. M. and Fry, S. C. (1999). Purification of xyloglucan endotransglycosylases
(XETs): a generally applicable and simple method based on reversible formation of an
enzyme-substrate complex. Biochemical Journal 340, 207-211.

Steele, N. M. and Fry, S. C. (2000). Difi‘erences in catalytic properties between native
isoenzymes of xyloglucan endotransglycosylase (XET). Phytochemistry 54, 667-680.

60

Steele, N. M., Sulova, 2., Campbell, P., Braam, J., Farkas, V. and Fry, S. C. (2001).

Ten isoenzymes of xyloglucan endotransglycosylase from plant cell walls select and
cleave the donor substrate stochastically. Biochemical Journal 355, 671-679.

Sulova, Z., Lednicka, M. and Farkas, V. (1995). A colorimetric assay for xyloglucan-
endotransglycosylase from germinating seeds. Anal Biochem 229, 80-5.

Sulova, Z., Takacova, M., Steele, N. M., Fry, S. C. and Farkas, V. (1998). Xyloglucan
endotransglycosylase: evidence for the existence of a relatively stable glycosyl-enzyme
intermediate. Biochem J 330, 1475-80.

Suzuki, K., Itoh, T. and Sasamoto, H. (1998). Cell wall architecture prerequisite for the
cell division in the protoplasts of white poplar, Populus alba L. Plant and Cell
Physiology 39, 632-638.

Sylvester, A. W. (2000). Division decisions and the spatial regulation of cytokinesis.
Current Opinion in Plant Biology 3, 5 8-66.

Taylor, N. G., Scheible, W. R., Cutler, S., Somerville, C. R. and Turner, S. R. (1999).
The irregular xylem3 locus of arabidopsis encodes a cellulose synthase required for
secondary cell wall synthesis. Plant Cell 11, 769-779.

Thompson, J. E. and Fry, S. C. (2000). Evidence for covalent linkage between
xyloglucan and acidic pectins in suspension-cultured rose cells. Planta 211, 275-286.

Thompson, J. E., Smith, R. C. and Fry, S. C. (1997). Xyloglucan undergoes
interpolymeric transglycosylation during binding to the plant cell wall in vivo: Evidence
from l3C/3H dual labelling and isopycnic centrifugation in caesium trifluoroacetate.
Biocheml 327, 699-708.

Turner, S. R. and Somerville, C. R. (1997). Collapsed xylem phenotype of Arabidopsis
identifies mutants deficient in cellulose deposition in the secondary cell wall. Plant Cell
9, 689-701 .

Varner, J. E. and Lin, L. S. (1989). Plant cell wall architecture. Cell 56, 231-9.

Verbruggen, M. A., Spronk, B. A., Schols, H. A., Beldman, G., Voragen, A. G.,
Thomas, J. R., Kamerling, J. P. and Vliegenthart, J. F. (1998). Structures of
enzymically derived oligosaccharides from sorghum glucuronoarabinoxylan. Carbohydr
Res 306, 265-274. ,

Vincken, J.-P., De Keizer, A., Beldman, G. and Voragen, A. G. J. (1995).

Fractionation of xyloglucan fiagrnents and their interaction with cellulose. Plant Physiol.
108, 1579-1585.

61

Wallace, G. and Fry, S. C. (1994). Phenolic components of the plant cell wall.
Int. Rev. Cytol. 151, 229-268. '

Weiser, M. M. and Wilson, J. R. (1981) Serum levels of glycosyltransferases and

related glycoproteins as indicators of cancer: biological and clinical implications. Crit
Rev Clin Lab Sci 14,189-239.

Whitcombe, A. J., O'Neill, M. A., Steffan, W., Albersheim, P. and Darvill, A. G.
(1995). Structural characterization of the pectic polysaccharide, rhanmogalacturonan-II.
Carbohydr. Res. 271, 15-29.

Whitney, S. E. C., Brigham, J. E., Darke, A. H., Reid, J. S. G. and Gidley, M. J.
(1995). In vitro assembly of cellulose/xyloglucan networks: Ultrastructural and
molecular aspects. Plant J. 8, 491-504.

Whitney, S. E. C., Gidley, M. J. and McQueen Mason, S. (2000). Probing expansin
action using cellulose/hemicellulose composites. Plant Journal 22, 327-334.

Whitney, S. E. C., Gothard, M. G. E., Mitchell, J. T. and Gidley, M. J. (1999). Roles
of cellulose and xyloglucan in determining the mechanical properties of primary plant
cell walls. Plant Physiology 121, 657-663.

Willats, W. G., McCartney, L., Mackie, W. and Knox, J. P. (2001). Pectin: cell
biology and prospects for functional analysis. Plant Mol Bio] 47, 9-27.

Wojtaszek, P. (2000). Genes and plant cell walls: a difficult relationship. Biol Rev Camb
Philos Soc 75, 43 7475.

Wu, Y. and Cosgrove, D. J. (2000). Adaptation of roots to low water potentials by

changes in cell wall extensibility and cell wall proteins. Journal of Experimental Botany
51, 1543-1553.

Wu, Y., Sharp, R. E., Durachko, D. M. and Cosgrove, D. J. (1996). Growth
maintenance of the maize primary root at low water potentials involves increases in cell-

wall extension properties, expansin activity, and wall susceptibility to expansins. Plant
Physiology 111.

Xu, W., Purugganan, M. M., Polisensky, D. H., Antosiewiu, D. M., Fry, S. C. and
Braam, J. (1995). Arabidopsis TCH4, regulated by hormones and the environment,
encodes a xyloglucan endotransglycosylase. Plant Cell 7, 1555-1567.

Xu, Z., Vo, L. and Macher, B. A. (1996). Structure-function analysis of human

alphal,3-fucosyltransferase. Amino acids involved in acceptor substrate specificity. J
Biol Chem 271, 8818-23.

62

York, W. S., Darvill, A. G. and Alberscheim, P. (1984). Inhibition of 2,4-
Dichlorophenoxyacetic Acid-Stimulated Elongation of Pea Stem Segments by a
Xyloglucan Oligosaccharide. Plant Physiol. 75, 295-297.

Zablackis, E., Huang, J., Miifler, B., Darvill, A. G. and Albersheim, P. (1995).
Characterization of the cell wall polysaccharides of Arabidopsis thaliana leaves. Plant
Physiology 107, 1129-1138.

Zhang, Z., Hong, Z. and Verma, D. P. (2000). Phragrnoplastin polyrnerizes into spiral
coiled structures via intermolecular interaction of two self-assembly domains. Journal of
Biological Chemistry 275, 8779-8784.

Zuo, J., Nin, Q. W., Nishizawa, N., Wu, Y., Kost, B. and Chua, N. H. (2000).

KORRIGAN, an Arabidopsis endo-l,4--glucanase localizes to the cell plate by polarized
targeting and is essential for cytokinesis. Plant Cell 12, 1137-1145.

63

CHAPTER TWO

Purification and Biochemical Characterization of Pea Xyloglucan
Fucosyltransferase

Credits for experiments described in this chapter are as follows:

Drs. Amy DeRocher and Maor Bar-Peled developed techniques for the
purification of pea xyloglucan fucosyltransferase. I analyzed samples that were
purified using these techniques by an in-gel assay and two-dimensional SDS-
PAGE.

I assisted Drs. DeRocher and Bar-Peled with scaling up the purification technique
to obtain enough protein for amino acid sequencing. Dr. DeRocher and I
prepared the protein sample used for amino acid sequence determination. The
amino acid sequencing was performed as a paid service by Harvard
Microchemistry (Cambridge, MA).

Methylation linkage analysis of product generated from an assay performed by
Dr. DeRocher was performed as a paid service by the Complex Carbohydrate
Research Center (Athens, GA).

I used protein samples prepared by Dr. Bar-Peled to characterize essential amino

acid residue types using group-specific modifying agents.

2.1 Introduction

Polysaccharides are a major structural feature of plant cell walls, imparting specific
mechanical properties that ultimately affect the ability of the wall to extend or maintain
structural integrity. These polysaccharides are often separated into different components
thought to represent domains within the wall: matrix polysaccharides, which cross-link

to cellulose, and pectins, which form a viscous gel.

Matrix polysaccharides include hemicelluloses that bind tightly but non-covalently to
cellulose microfibrils, crosslinking them into a complex network. The hemicellulose
xyloglucan (XyG) comprises approximately 20% of the total cell wall in dicot and
nongraminaceous monocot plants and forms a load-bearing network by associating to the
surfaces of surrounding cellulose microfibrils through hydrogen bonds (Hayashi, 1989;
Zablackis et al., 1995). Xyloglucan contains a [3-1,4-glucan backbone decorated with
side chains of xylose alone, xylose and galactose, and xylose, galactose, and fucose
(Figure 1, Chapter 1). The presence or absence of the terminal fucose residue may have
structural and biological significance. Some models suggest that the presence or absence
of this fucose residue will determine whether the xyloglucan conformation is planar and
thus better able to bind to cellulose (Levy et al., 1997; Levy et al., 1991), though
contradicting evidence has been described (Whitney et al., 1995). XyG networks may be
modified by xyloglucan endotransglycosylase (XET), an enzyme which cleaves and
rejoins adjacent XyG chains. A recombinant XET demonstrated different activity rates
for fucosylated versus nonfucosylated XyG oligosaccharide acceptors, indicating that

fucosylation state may affect XET modification of the cell wall (Purugganan et al., 1997).

65

In addition, oligosaccharides consisting of an XyG nonasaccharide prevent auxin-
promoted elongation of pea stems if these oligosaccharides contain fucose, but not if they
lack fucose (Creelman and Mullet, 1997; Fry et al., 1993; McDougall and Fry, 1989;
York et al., 1984). Thus, it is possible that XG fragments act as signalling molecules in

vivo.

Most matrix polysaccharides are branched molecules modified by various sugars. These
modifications are important because they allow heterogeneity in the shape of matrix
polysaccharides and patterns of cross-links, resulting in a dynamic and porous cell wall.
These polysaccharide modifications occur via glycosyltransferase reactions, many of
which occur in the Golgi (Colley, 1997; Driouich et al., 1993; Moore et al., 1991;
Puhlmann et al., 1994). Attempts to clone plant glycosyltransferases using sequences
derived from bacterial or mammalian glycosyltransferases have been unsuccessful (AED,
NVR, KK, data not shown.) This is not entirely unexpected, for although Golgi
glycosyltransferases ofien have similar general structural features, they rarely share

extensive sequence similarity (Kleene and Berger, 1993).

This chapter describes the biochemical purification of a xyloglucan fucosyltransferase
from garden pea (Pisum sativum) and analysis of the corresponding protein by an in-gel
assay and two-dimensional electrophoresis. Carbohydrate analysis data on the product of
the enzyme are presented. Partial amino acid sequence was determined from the purified
enzyme, and amino acid modifying agents were used to analyze residue types necessary

for biochemical fimction.

66

V —-m" ‘3:

2.2 Materials and Methods

2.2.1 Purification of Pea Xyloglucan F ucosyltransferase

2-cm segments of etiolated Pisum sativum, cv Alaska, excised just below the apical hook,
were collected and homogenized in 1.5 volumes buffer (50 mM Hepes pH 7.5, 1 mM
EDTA pH 8.0, 0.4 M sucrose, 1 mM DTT, 0.1 mM PMSF, 1 ug/mL each aprotinin,
leupeptin, and pepstatin). The homogenate was filtered and centrifuged at 2,000 g for 15
minutes, and the supernatant was centrifuged at 100,000 g for 1 hour. The resulting
pellets were washed and homogenized in the presence of 0.1 M Na2C03 to strip away
peripheral membrane proteins (Fujiki et al., 1982). The suspension was centrifuged at
100,000 g for 1 hour and the resulting pellets were washed and resuspended in buffer (50
mM Pipes-KOH pH 6.2, 20% glycerol, 1 mM EDTA, 1 mM DTT, 0.1 M PMSF, 1
ug/mL each aprotinin, leupeptin, and pepstatin). The suspension was homogenized,
Triton X-100 was added to a final volume of 0.8%, and the sample was stirred for 1-2
hours to solubilize membrane proteins before centrifugation a final time at 100,000 g for
1 hour. Supernatant was collected and saved. Pea carbonate-washed supematants were
pooled and separated on a GDP-HA agarose affinity chromatography column and GDP-
binding proteins were eluted using excess free GDP. Protein levels were monitored by
A230. The protein samples were desalted on a Sephadex G-25 column, concentrated, and
firrther separated on a Phenomenex SEC 4000 size exclusion column. Some samples
were further purified using a Poros QE or Resource Q anion exchange column and

subsequently separated by SDS-PAGE electrophoresis.

67

2. 2. 2. Activity Assays

Samples were incubated at room temperature for the specified time (30 min in most
cases) with 25 mM Pipes-KOH pH 6.2, 0.5 mg/mL tamarind xyloglucan, and GDP-[3H )-
fircose to a final concentration of 3.3 nM (573 GBq/mmol, NEN, Boston, MA). Most
assays also contained 50 11M non-radiolabeled GDP-fucose. Reactions were precipitated
using 70% ethanol and 3 H incorporation was measured by scintillation counting. The
amount of fucose incorporated into the product was used to calculate activity in nanokats

(nKat = nMoles substrate incorporated into product per second).

2.2.3 In-gel Assay and Two-Dimensional Electrophoresis

Anion exchange purified samples were loaded into 2 lanes of a nondenaturing
polyacrylamide gel that had 0.1 mg / ml tamarind xyloglucan and 0.1%
decyhnaltopyranoside incorporated into the gel matrix. The gel was run at 4°C. The two
lanes were excised from the gel. One was equilibrated in 25 mM Pipes KOH, pH 6.2 and
incubated overnight in 25 mM Pipes KOH, pH 6.2 with 0.1% 1"'C GDP fucose (NEN,
10.5 GBq/mMole, 0.37 MBq/ml). Unincorporated 14C was washed out of the gel, leaving
only the 1”'C fucose that had been incorporated into the xyloglucan. The gel was then
fixed, enhanced, and fluorographed. The other gel strip was equilibrated in 2% SDS, 60
mM Tris, pH 8.0, and 2% beta mercaptoethanol, overlaid on a 10 % acrylarnide SDS gel,
and proteins in the gel were separated by electrophoresis. Protein in the gel was
visualized by silver staining. Proteins that co-migrated with the fircosyltransferase
activity were identified by aligning the axes of the fluorograph and the stained two-

dimensional gel and normalizing for gel shrinkage and expansion.

68

2.2.4 Carbohydrate Methylation Linkage Analysis

Tamarind seed xyloglucan was fucosylated by purified pea FUTase equal to 33 pKat total
activity (assay conditions: 1 mg/mL tamarind XyG, 1.5 mM GDP-fucose, 50 mM Pipes-
KOH pH 6.2). XyG product was precipitated with ethanol, re-suspended in water, re-
precipitated, and sent to the Complex Carbohydrate Research Center (Athens, GA) for
linkage analysis. An equal amount of tamarind xyloglucan was also submitted for

linkage analysis.

2.2.5 Amino Acid Sequencing

Proteins in size exclusion column eluate fractions containing peak amounts of
fucosyltransferase (F UTase) activity were concentrated using a Millipore 4-mL 10-kDa
concentrator and separated by electrophoresis. After brief staining with Coomassie
Brilliant Blue and destaining, the separated proteins were excised, rinsed in 50%

acetonitrile, stored at —80° C and sent to Harvard Microchemistry (Cambridge, MA) for

tryptic peptide sequencing.

2.2. 6 Amino Acid Modification

Purified enzyme was subjected to amino acid group-specific modification. The following
reagents were used: N—ethylmaleimide (N EM), diethylpyrocarbonate (DEPC), pyridoxal
5'-phosphate (PLP), and N-bromosuccinimide (NBS). Modifying reagents were used
according to the methods described by Britten and Bird (Britten and Bird, 1997).

Standard assays were conducted as described above in the presence of up to 5 mM of each

69

modifying reagent using GDP-[3H]Fuc. The reactions were halted by adding 67% ethanol
and washed as described above before assessing label incorporation. Controls were

conducted in the absence of these modifying reagents.

2.3 Results

2.3.1 Purification of Pea Xyloglucan F ucosyltransferase

The terminal fucosyl residue on XyG side chains is added by a fucosyltransferase
(F UTase). We purified enough of this FUTase from pea epicotyls to determine partial
amino acid sequences from the enzyme. Microsomes were prepared from the pea
epicotyls, carbonate-washed to enrich for membrane proteins, and solubilized with
nonionic detergent such as Triton X-100. A specific assay for this enzyme had
previously been developed using tamarind seed storage xyloglucan, which lacks fucosyl
residues, as an acceptor molecule and radiolabeled GDP-fucose as a donor (Figure 3)
(Camirand et al., 1987; Farkas and Maclachlan, 1988). GDP-agarose affinity
chromatography, size exclusion chromatography, and anion exchange chromatography
were used in conjunction with FUTase activity assays to purify and detect the enzyme
(Figure 4). The fucosyltransferase activity co-eluted with 60-kDa and 65-kDa proteins
(Figure 48). Other proteins are also present in various fiactions, but their elution profile
did not match the profile of the enzyme activity. It was possible to purify XyG FUTase
1400-fold after size-exclusion chromatography, resulting in a total of 50 pg protein

containing 70 nKat XyG FUTase activity. Analysis of the purification data indicates that

70

XyG FUTase is present at very low levels in the cells (less than 0.01% of total protein)

(Faik et al., 2000).

71

 

 

 

 

. 0 fucose
Tamarind
Seed Q galactose
Xyloglucan
. glucose
Tamarind xyloglucan . xylose
+ GDP- 3H]-fucos.e+
enzyme
3H
xyloglucan

Figure 3. Xyloglucan fucosyltransferase activity assay. Nonfucosylated forms of
XyG accumulate as seed storage polysaccharides in a few species, notably tamarind and
nasturtium. Tamarind seed storage XyG (top) is available commercially, and is incubated
with GDP-[3H]-fucose and enzyme during a XyG FUTase activity assay. The product of
the reaction is precipitated with 66% ethanol (bottom) and the amount of radiolabel
incorporated is quantified by scintillation counting (Camirand et al., 1987; Farkas and
Maclachlan, 1988; Perrin et al., 1999).

72

GDP-afllnlty
Size 5ch
8

   

“.f‘f

 

03
g

 

b
1

N
1

RelativeAcuvnympmer‘) B

 

o-
Fractionz234567 91011121314151617

Figure 4. Biochemical purification of XyG FUTase from pea. A. Silver-stained SDS-
PAGE gels showing protein profiles from 1) carbonate-washed, detergent-solubilized
microsomes from pea epicotyls (Sol. Microsomes), or the fi'actions containing peaks of
FUTase activity from 2) a GDP-agarose affinity column (GDP-affinity) and 3) a size
exclusion column (Size Excl.). Arrows indicate 60 and 65 kDa polypetides (see text).B.
Top: Silver-stained SDS-PAGE gel showing the protein profile from several fractions of
an anion-exchange column. Bottom: Total XyG FUTase activity in dpm x 10'5 for each
traction of the anion exchange column eluate. Note that bars align with the
corresponding SDS-PAGE profile above.

73

2.3.2 Analysis of Purified Pea Xyloglucan F ucosyltransferase Activity

A second method was used to determine which polypeptide(s) in the size exclusion
column eluate co-migrated with the fucosyltransferase activity. For two-dimensional gel
analysis, nondenaturing gel electrophoresis of identical samples was used in the first
dimension followed by either fucosyltransferase activity staining or SDS-PAGE in the
second dimension. Activity in the native gel co-migrated with two abundant proteins with
molecular masses of 60 and 65 kDa (Figure 5). Other, less abundant polypeptides were
also detected on the silver-stained gel, but they did not coincide with the

fucosyltransferase activity.

74

" 97 kD

'66 kD

'45 kD

 

Figure 5. Identification of proteins that co-migrate with fucosyltransferase activity.
Duplicate samples of anion exchange-purified sample were separated by native page with
tamarind xyloglucan incorporated into the gel matrix. Fucosyltransferase activity in one
lane was visualized by incorporation of MC labeled fucose into the xyloglucan, followed
by fluorography (A). The protein profile of the other lane was analyzed by 2-D gel
analysis using SDS PAGE and silver staining (B). The native gel dimensions of panels A
and B were aligned to identify proteins that co-migrate with fucosyltransferase activity.

75

2.3.3 Product Characterization

To confirm that the purified pea protein synthesizes an (Jr-1,2 fucosezgalactose linkage,
carbohydrate analysis was performed 'on the product resulting from in vitro fucosylation
of tamarind xyloglucan (thG) by purified FUTase (Table 2). Linkage analysis indicated
that incubation of thG with purified FUTase resulted in a decrease in the mole
percentage of terminal galactose and the appearance of 2-galactose and terminal fucose,

thus confirming the activity of the purified enzyme (Table 2).

76

Table 2. Product methylation linkage analysis. Carbohydrate linkage analysis of
tamarind xyloglucan before (tamarind XyG) and afier (fucosylated XyG) incubation with
purified pea FUTase.

 

 

 

 

 

 

 

 

 

Sugar Residue % tamarind XyG % fucosylated XyG
4-Glucose 16.4 17.5
4,6-Glucose 37.0 31.5
terminal Xylose 19.0 13.5
2-Xylose 15.0 14.3
terminal Galactose 12.6 5.5
2-Galactose - ' 9.0
terminal Fucose - 8.7

 

 

 

 

77

2.3.4 Determination of Peptide Amino Acid Sequence

After biochemical purification and subsequent analysis, two polypeptides of
approximately 65 kD and 60 kD in size were observed to co-purify with XyG FUTase
activity (Figure 4). Limited peptide sequence was obtained from both proteins. The 65-
kDa peptide was identified as a homolog of BiP, a molecular chaperone usually localized
in the ER. It remains unclear whether copurification of BiP with FUTase represents a
significant interaction. Six peptide sequences were obtained for the approximately 60-
kDa protein: VFGFLGR, YLLHPTNNVWGLVVR, AVLITSLSSGYFEK,
YYDAYLAK, LLGGLLADGFDEK, and ESILPDVNR. These peptides were not highly
homologous to any proteins of known function in public databases, but an Arabidopsis
thaliana expressed sequence tag (EST) entry was identified that encodes a homolog of

the purified pea protein (described in chapter 3.)

2.3.5 Characterization of Amino Acid Residue Types Essential for Activity

To gain some insight into the structure-firnction relationship in XyG-FUTase, amino acid
residue types essential for catalytic activity were identified. NEM had relatively little
effect on the XyG-FUTase activity, with 5 mM giving ~40% inhibition of activity (Table
3). However, N-bromosuccinimide (NBS), DEPC, and PLP reagents (which react with
tryptophan, histidine and lysine residues, respectively) decreased the activity by about 90-
99% when applied at levels of up to 5 mM (Table 3). Preincubation of the XyG-FUTase

with GDP-Fuc did not prevent inactivation by NBS treatment (data not shown).

78

 

 

Table 3. Effect of group-specific reagents on purified pea XyG-FUTase E
activity. The purified enzyme (~0.2 pg) was pre-incubated with the indicated :
modifying reagent for 20 min and assayed for [3H]-fucose transfer to tamarind ,
XyG under standard assay conditions. Values are expressed as % inactivation. .
Shown are results from one experiment that was repeated three times.

 

 

 

Inactivation (%)
Reagent Acts on 0.5 mM 2.5 mM 5 mM
NEM Cysteine 5.9 34.4 41.5
PLP Lysine 51.4 81.0 87.9
DEPC Histidine 63.5 88.5 95.1
NBS Tryptophan 96.0 96.3 96.4

 

79

 

2.4 Discussion

Elucidation of plant cell wall biosynthesis has been a slow process. Despite decades of
work by organic chemists which have helped describe the polysaccharide and protein
components of the cell wall, and despite the fundamental importance of plant cell walls in
agriculture, nutrition, materials science, textiles, and as mediators of plant growth in
general, only one gene involved in cell wall biosynthesis (CesA , the catalytic subunit of
cellulose synthase) had been cloned at the time that this study was initiated (Arioli et al.,
1998; Pear et al., 1996). None of the glycosyltransferases that synthesize non-cellulosic
cell wall polysaccharides or any of the carbohydrate moieties of plant cell wall
glycoproteins had been identified at the molecular level (Brett and Waldron, 1996). In
part, this was because these integral membrane proteins have been recalcitrant to
purification; many efforts to purify them have resulted in loss of activity such that it
became impossible to follow their purification through subsequent steps (Reiter, 1998).
Genetic approaches have been attempted, but have both advantages and limitations for
identification of cell wall biosynthetic enzymes (for a review, see Chapter 1.) The
growing number of glycosyltransferase sequences from other organisms involved in
protein glycosylation, bacterial Nod factor synthesis, or bacterial lipopolysaccharide
synthesis available in the databases at the time had not yet assisted plant researchers in
identifying plant glycosyltransferases by homology alone. This indicated that plant
glycosyltransferases might have primary structures rather unique from

glycosyltransferases in other organisms.

80

Although most attempts to purify glycosyltransferases and solubilize them have resulted
in a loss of activity, this was not the case with a XyG FUTase. Maclachlan's group had
shown that it was possible to purify this enzyme fi'om pea epicotyls and assay activity of
solubilized microsomes using tamarind or nasturtium seed xyloglucan (F arkas and
Maclachlan, 1988; Hanna et al., 1991; Maclachlan et al., 1992). In seeds of some species
such as tamarind or nasturtium, xyloglucan is used as a storage polysaccharide. Storage
XyG lacks the terminal fucose residue and may be used in an assay as an acceptor along
with radiolabeled GDP-fucose as a sugar nucleotide donor. Purification techniques and
the assay were used to obtain enough enzyme to determine limited amino acid sequence
data. The purification scheme involved the preparation of detergent-solubilized
microsomes from pea epicotyls (enriched in integral membrane proteins during a sodium
carbonate wash step), GDP agarose affinity chromatography, and size exclusion
chromatography. Anion exchange chromatography was also used as a further
purification step in some experiments, though it was not necessary for obtaining
sufficiently pure protein for amino acid analysis. By analyzing the xyloglucan-specific
fucosyltransferase activity of the chromatographic fractions, we observed two
polypeptides that were found consistently in the most active fiactions (Figure 4).
Sequencing tryptic peptides from these proteins indicated that the larger polypeptide,
approximately 65 kDa in size, is highly homologous to BiP, the ER-localized molecular
chaperone in the hsp70 family. Whether this was co-purified due to artifactual or
physiological reasons is unknown, though it is interesting to consider that the enzyme
was subjected to harsh treatment during the purification procedure (for example,

exposure to a pH of 11 for over an hour during the carbonate wash) and association with

81

a chaperone protein could have prevented denaturation and thus loss of activity. Peptides
derived from the second (~ 60 kDa) polypeptide were not homologous to any known
proteins in the databases but did allow us to identify an Arabidopsis Expressed Sequence
Tagged (EST) clone which contained these peptide sequences with 63%-86%
conservation of identity (described in Chapters 3 and 4). Gel filtration and SDS-PAGE
data indicated that the native form of purified pea XyG-FUTase exists as an oligomer of

approximately 250 kDa.

Isopycnic centrifugation had previously provided evidence that the XyG-FUTase is
present in both a dense Golgi fraction and in vesicles from the trans-Golgi network
(Brummell et al., 1990). Later studies using antibodies to the fucosylated region of XyG
determined that the fucose-containing product is observed in the trans-Golgi cistemae
and in the trans-Golgi network (Zhang and Staehelin, 1992). Taken together, these
studies provide evidence that fucosylation occurs after XyG backbone biosynthesis has
been completed. Thus, the purified pea XyG FUTase would be expected to be an integral
membrane protein localized to the Golgi. In vitro transcription and translation of the
Arabidopsis XyG FUTase in the presence of canine pancreatic microsomes indicated that
it is an integral membrane protein (R. Perrin, data presented in Chapter 3) and

localization of this enzyme to the Golgi has been confirmed (R. Sarria, unpublished data.)

Several structure-function studies on FUTase enzymes in other systems have been
initiated by first examining the effects of various amino acid modifying reagents on

enzyme activity and substrate binding, later followed by point mutagenesis, deletion, or

82

domain swapping experiments (for a review, see de Vries et al., 2001). As a first
approach towards identifying amino acid residue types that are critically important for
catalytic activity, activity assays with purified pea XyG FUTase were conducted in the
presence of a series of amino acid modifying reagents (Table 3). The sulfhydryl reagent
NEM, a cysteine-specific modifying agent, has been widely used in discriminating
between glycoprotein fucosyltransferase isotypes (de Vries et al., 1995). Pea XyG
FUTase is not NEM-sensitive, indicating that Cys residues may not be essential for
substrate binding or catalysis. Similar studies on human or-(l,3) fucosyltransferases
coupled with site-directed mutagenesis and amino acid sequence alignments of various
isoforms have led other researchers to conclude that FUTases that are NEM-insensitive
contain either a Ser or Thr residue in place of conserved Cys residues in the GDP-Fuc

binding domain (de Vries et al., 2001; Holmes et al., 1995).

Purified pea XyG FUTase was extremely sensitive to modification of Trp residues (Table
3). There are eight Trp residues present in pea XyG FUTase, six of which (75% of the
total) are found in the C-terminal half of the protein. In fact, five Trp residues (62.5% of
the total) are within the most C-terminal quarter of the protein. The general structure of
glycosyltransferases, including fucosyltransferases, consists of a variable N-terminal
region (a short cytoplasmic fiagment, the transmembrane domain, and a extended stem
region located in the hunen of the Golgi), followed by a more conserved globular
catalytic region at the carboxy-terminal portion of the protein (Breton et al., 1998). The
catalytic region is also within the lumen of the Golgi. It appears that the distribution of

Trp residues in pea XyG FUTase shows bias towards the region of the protein expected

83

to constitute the catalytic domain. A similar distribution of Trp residues is found in the
Arabidopsis XyG FUTase sequence, though the Arabidopsis protein has an additional Trp
in the middle region. Because incubation of pea XyG FUTase with GDP-Fuc before
exposure to NBS does not prevent inhibition of activity (data not shown), it is likely that
the critical NBS-sensitive Trp residue(s) is involved in binding to XyG or, alternatively,
is required to assume a catalytically competent structural conformation. It is interesting
to note that Trp residues have previously been noted to be critical for determining
acceptor substrate specificity in other FUTases (Dupuy et al., 1999; Elmgren et al., 1997).
In one study, a conserved Trp residue in the variable stem region of vertebrate FUTases
determined acceptor specificity (Dupuy et al., 1999). Changing this Trp to an Arg in the
enzyme encoded by the FUT 3 gene was sufficient to shift acceptor specificity from

Lewis H-type 1 to Lewis H-type 2 acceptors (Dupuy et al., 1999).

In addition, pea XyG FUTase was moderately sensitive to chemical modification of Lys
or His residues (Table 3). There are 14 His residues in this protein, and the distribution is
biased towards the middle and C-terminal regions of the protein. By contrast, the 33 Lys
residues are distributed evenly throughout the protein. More studies using site-directed
mutagenesis are needed to determine the locations and the roles of all of these essential

residues.

Other biochemical characterization of the purified pea XyG FUTase has been conducted
and published (F aik et al., 2000). This characterization included kinetic analysis of

enzymatic activity, determination of pH optimum, requirement for divalent cations and

84

inhibition of activity by the byproduct GDP (data generated by M. Bar-Peled and A.
Faik.) Substrate specificity assessment indicated that the enzyme is specific for XyG and
does not transfer fucose to other fucose-containing cell wall or glycoprotein components
(data generated by A. Faik.) This study also reported the identification of a pea cDNA

sequence encoding this enzyme (data generated by W. Zeng.)

The purification of this enzyme allowed a detailed study of the biochemical
characteristics of XyG fucosylation. It also allowed the identification of both the gene
encoding pea XyG F UTase (W. Zeng) and the gene encoding Arabidopsis XyG FUTase
(R Perrin, described in Chapters 3 and 4.) This availability of the purified pea enzyme
and the pea and Arabidopsis genes potentially enables the identification of other plant-
specific glycosyltransferases. Hundreds to thousands of different genes are needed to
synthesize the various polysaccharides that compose the cell wall. Substrate acceptors
and assays remain unavailable for many of these enzymes. Identification of other
carbohydrate transferases, perhaps by sequence similarity, could lead to tailored in vitro
production of carbohydrates, as well as an understanding of how the complex plant cell

wall is biosynthesized.

References

Arioli, T., Peng, L., Betzner, A. 8., Burn, J., Wittke, W., Herth, W., Camilleri, C.,
Hofte, H., Plazinski, J., Birch, R. et al. (1998). Molecular analysis of cellulose
biosynthesis in Arabidopsis. Science 279, 717-720.

Brett, C. T. and Waldron, K. (1996). Physiology and biochemistry of plant cell walls.
London: Chapman and Hall.

85

Britten, C. J. and Bird, M. I. (1997). Chemical modification of an alpha-3-
fucosyltransferase; definition of amino acid residues essential for enzyme activity.
Biochimica et Biophysica Acta 1334, 57-64.

Brummell, D. A., Camirand, A. and Maclachlan, G. A. (1990). Differential
distribution of xyloglucan transferases in pea Golgi dictyosomes and secretory vesicles.
Journal of Cell Science 96, 705-710.

Camirand, A., Brummell, D. A. and Maclachlan, G. (1987). Fucosylation of
xyloglucan: localization of the transferase in dictyosomes of pea stem cells. Plant Physiol
84, 753-756.

Colley, K. J. (1997). Golgi localization of glycosyltransferases: more questions than
answers. Glycobiolog/ 7, 1-13.

Creelman, R. A. and Mullet, J. E. (1997). Oligosaccharins, brassinolides, and
jasmonates: nontraditional regulators of plant growth, development, and gene expression.
Plant Cell 9, 1211-23.

de Vries, T., Srnka, C. A., Palcic, M. M., Swiedler, S. J., van den Eijnden, D. H. and
Macher, B. A. (1995). Acceptor specificity of different length constructs of human
recombinant alpha IBM-fucosyltransferases. Replacement of the stem region and the
transmembrane domain of fircosyltransferase V by protein A results in an enzyme with
GDP-fucose hydrolyzing activity. J Biol Chem 270, 8712-22.

Driouich, A., Faye, L. and Staehelin, L. A. (1993). The plant Golgi apparatus: A
factory for complex polysaccharides and glycoproteins. Trends BiochemSci. 18, 210-
214.

Faik, A., Bar Peled, M., DeRocher, A. E., Zeng, W., Perrin, R. M., Wilkerson, C.,
Raikhel, N. V. and Keegstra, K. (2000). Biochemical characterization and molecular
cloning of an alpha-1,2-fi1cosyltransferase that catalyzes the last step of cell wall
xyloglucan biosynthesis in pea. J Biol Chem 275, 15082-9.

Farkas, V. and Maclachlan, G. (1988). Fucosylation of exogenous xyloglucans by pea
microsomal membranes. Arch Biochem Biophys 264, 48-53.

Fry, S. C., Aldington, S., Hetherington, P. R. and Aitken, J. (1993). Oligosaccharides
as signals and substrates in the plant cell wall. Plant Physio]. 103, 1-5.

Fujiki, Y., Hubbard, A. L., Fowler, S. and Lazarow, P. B. (1982). Isolation of
intracellular membranes by means of sodium carbonate treatment: application to
endoplasmic reticulum. Journal of Cell Biology 93.

Hanna, R., Brummell, D. A., Camirand, A., Hensel, A., Russell, E. F. and
Maclachlan, G. A. (1991). Solubilization and properties of GDP-fucose: xyloglucan 1,2-

86

 

alpha-L-fucosyltransferase from pea epicotyl membranes. Arch Biochem Biophys 290, 7-
13.

Hayashi, T. (1989). Xyloglucans in the primary cell wall. Annual Review of Plant
Physiology and Plant Molecular Biology 40, 139-168.

Kleene, R. and Berger, E. G. (1993). The molecular and cell biology of
glycosyltransferases. Biochim Biophys Acta 1154, 283-325.

Levy, 8., Maclachlan, G. and Staehelin, L. A. (1997). Xyloglucan sidechains modulate
binding to cellulose during in vitro binding assays as predicted by conformational
dynamics simulations. PlantJ. 11, 373-386.

Levy, 8., York, W. S., Stuike Prill, R., Meyer, B. and Staehelin, L. A. (1991).
Simulations of the static and dynamic molecular conformations of xyloglucan. The role
of the fucosylated sidechain in surface-specific sidechain folding. Plant j 1, 195-215.

Maclachlan, G., Levy, B. and Farkas, V. (1992). Acceptor requirements for GDP-
fucosezxyloglucan 1,2-alpha-L-fi1cosyltransferase activity solubilized from pea epicotyl
membranes. Arch Biochem Biophys 294, ZOO-205.

McDougall, G. J. and Fry, S. C. (1989). Anti-auxin activity of xyloglucan
oligosaccharides: the role of groups other than the terminal alpha-L-fucose residue.
Journal of Experimental Botany 40, 233-238.

Moore, P. J., Swords, K. M., Lynch, M. A. and Staehelin, L. A. (1991). Spatial
organization of the assembly pathways of glycoproteins and complex polysaccharides in
the Golgi apparatus of plants. J Cell Biol 112, 589-602.

Pear, J. R., Kawagoe, Y., Schreckengost, W. E., Delmer, D. P. and Stalker, D. M.
(1996). Higher plants contain homologs of the bacterial celA genes encoding the catalytic
subunit of cellulose synthase. Proc.Natl.Acad.Sci. USA 93, 12637-12642.

Puhlmann, J., Bucheli, E., Swain, M. J., Dunning, N., Albersheim, P., Darvill, A. G.
and Hahn, M. G. (1994). Generation of monoclonal antibodies against plant cell-wall
polysaccharides. 1. Characterization of a monoclonal antibody to a terminal alpha-(l--
>2)-linked fucosyl-containing epitope. Plant Physiol 104, 699-710.

Purugganan, M. M., Braam, J. and Fry, S. C. (1997). The Arabidopsis TCH4
xyloglucan endotransglycosylase. Substrate specificity, pH optimum, and cold tolerance.
Plant Physiol 1 15, 181-190.

Reiter, W. D. (1998). Arabidopsis thaliana as a model system to study synthesis,
structure, and function of the plant cell wall. Plant Physio]. Biochem. 36, 167-176.

87

Tm}

Whitney, S. E. C., Brigham, J. E., Darke, A. H., Reid, J. S. G. and Gidley, M. J.
(1995). In vitro assembly of cellulose/xyloglucan networks: Ultrastructural and
molecular aspects. PlantJ. 8, 491-504.

York, W. S., Darvill, A. G. and Alberscheim, P. (1984). Inhibition of 2,4-
Dichlorophenoxyacetic Acid-Stimulated Elongation of Pea Stem Segments by a
Xyloglucan Oligosaccharide. Plant Physiol. 75, 295-297.

Zablackis, E., Huang, J., Miifler, B., Darvill, A. G. and Albersheim, P. (1995).
Characterization of the cell wall polysaccharides of Arabidopsis thaliana leaves. Plant
Physiology 107, 1129-1138.

Zhang, G. F. and Staehelin, L. A. (1992). Functional complementation of the Golgi
apparatus of plant cells. Plant Physiology 99, 1070-1083.

 

88

CHAPTER THREE

Identification and Functional Analysis of an Arabidopsis thaliana Gene Encoding a

Xyloglucan Fucosyltransferase

Credits for the work presented in this chapter are as follows:

The pea xyloglucan fucosyltransferase peptide sequence that allowed the
identification of an Arabidopsis EST was determined by joint efforts of Drs. Amy
DeRocher, Maor Bar-Peled, and myself as described in Chapter 2 of this thesis.

I used this Arabidopsis EST as a probe for RNA gel analysis.

I obtained a full-length cDNA clone for AtF UT 1 by library screening.

1 performed in vitro transcription and translation experiments with this cDNA
clone to demonstrate that it encodes an integral membrane protein.

Weiqing Zeng prepared antibodies that recognize a His,-tagged recombinant
version of AtFUTl. I used these antibodies to immunoprecipitate xyloglucan
fucosyltransferase activity from solubilized Arabidopsis microsomes.

Dr. Ariel Orellena and Lorena Norambuena used the AtF UT 1 cDNA clone for

heterologous expression experiments in Cos-7 cells.

89

3.1 Introduction

The previous chapter described the purification of a pea xyloglucan fucosyltransferase
(XyG FUTase) and identification of limited amino acid sequences from this protein.
Although pea is an advantageous model system for biochemical analyses because of the
biomass that can be generated in a relatively short period of time, Arabidopsis is the
organism of choice for plant genetic and molecular biological analysis. The complete
sequence of the Arabidopsis genome was determined during the course of this research
project (Arabidopsis Genome Initiative, 2000), and over 113,000 expressed sequence
tagged (ESTs) have been deposited in public databases as of September, 2001. These
ESTs are a valuable resource for identification of expressed genes, and the number of
ESTs now available from various tissues and plant treatments also allows researchers to

determine gene expression patterns “in silico”.

The amino acid sequence data described in the previous chapter fi'om purified pea XyG
F UTase allowed us to identify an Arabidopsis EST that encoded peptides with a high
degree of similarity to the pea protein. This, in turn, allowed the identification of a full-
length clone of the Arabidopsis gene by cDNA library screening. Sequence analysis of
this clone, originally named AtFTI and later re-narned AtF UT 1 , indicated that it
contained features expected of a Golgi-localized glycosyltransferase. Attempts to use
this clone to produce active enzyme by in vitro transcription and translation were not
successful, but translation in the presence of canine pancreatic microsomes indicated that
the translation product is an integral membrane protein. Two methods were used to

confirm the function of AtFUTl as a xyloglucan fucosyltransferase. In the first method,

90

antibodies to HisG-AtFUTl were used to immunoprecipitate XyG FUTase activity from
solubilized Arabidopsis microsomal proteins. In the second case, AtF UT 1 cDNA was

expressed in mammalian Cos-7 cells, conferring XyG FUTase activity to these cells.

3.2 Materials and Methods

3. 2.1 RNA Ge] Blot Analysis

Northern analysis was performed using standard techniques (Sarnbrook et al., 1989). The
858-nt insert of EST 191A6T6 was labeled with 32P-dATP in a random prime labeling
reaction (Sarnbrook et al., 1989). Twenty micrograms of total RNA isolated from
Arabidopsis 7-day-old seedlings, organs of mature plants (flowers, leaves, roots, stems),
and cell culture was subjected to electrophoresis in a 1% agarose formaldehyde gel.
RNA samples were transferred to Hybond N+ membrane (Amersham Pharmacia Biotech,
Pisscataway, NJ) using capillary transfer in 10X SSC (sodium chloride/sodium citrate)
buffer. RNA was cross-linked to the blot by UV illumination. The blot was hybridized
at 50° C for nine hours. The probe was denatured by adding 1/10 volume 3 M NaOH and
was added to the blot. Hybridization occurred at 50° C for 13 hours. The blot was
washed twice with 2X SSC/0.1% SDS (50° C, 20 minutes each), once with 2X SSC/0.1%
SDS (45° C, 30 minutes), once with 0.1X SSC/0.1% SDS (45° C, 30 minutes), and a final
time with 0.1X SSC/0.1% SDS (50° C, 30 minutes.) Autoradiography fihn was exposed

to the blot for 40.5 h.

91

3. 2.2 Library Screening

191A6T7 was labeled with digoxigenin using a DNA labeling and detection kit (C/N
1093657, Boehringer Mannheim, Basel, Switzerland) and used as a probe to screen the
CD4-15 portion of a size-fractionated Arabidopsis cDNA library (Kieber et al., 1993) at
high stringency. The library was titred and subjected to primary, secondary and tertiary

screening. In vivo excision was performed to isolate positive clones.

3. 2.3 In Vitro Transcription and Translation

In vitro transcription was performed using a full-length cDNA AtFUTl clone in
pBluescript KS' vector. Commercial reagents were obtained from Promega (Madison,
WI.) The transcription reaction included two micrograms of linearized AtFUTl cDNA,
1X in vitro transcription buffer, 0.5 mM rNTP (final concentration, minus rGTP), 1/10
volume methylated cap, 10 mM DTT, 0.05 mg/ml BSA, 100 units RNasin and 25-50
units T3 polymerase. The reaction was incubated for 10 minutes at 37° C, GTP was
added to a concentration of 0.5 mM, and the reaction continued for another 90 minutes at
37° C. An aliquot of the reaction was analyzed on an agarose gel stained with ethidium
bromide. In vitro translation reaction compositions were as recommended by the
manufacturer (Promega, Madison, WI) and included rabbit reticulocyte lysate, RNasin,
canine pancreatic microsomes, amino acid mix minus methionine, and 35S-Met. In
addition to AtFUTl , beta-lactarnase mRN A (provided by manufacturer) was used as a
positive control for translation and a reaction with DEPC-treated ddH20 was used as a
negative control to determine background levels of translation in this system. The

reactions were incubated at 30° C for 1 hour. After incubation, translation reactions were

92

recovered by centrifugation through a l M sucrose cushion in 50 mM Pipes-KOH pH 6.2.
AtFUTl samples were divided into an aliquot that received no further treatment (“total”)
and an aliquot that was washed with 0.1 M sodium carbonate on ice for thirty minutes
with homogenization. The carbonate-washed AtFUTl sample was centrifuged at
100,000 g for 1 hour. The protein was precipitated with TCA from the supernatent. The
pellet was analyzed directly. All samples were subjected to electrophoresis on a 10%

SDS-PAGE gel.

3. 2.4 Activity Assays

Activity assays were conducted as described previously (see Chapter 2.)

3. 2.5 Antibody Preparation

The portion of AtFTI encoding amino acids 73 to 566 was PCR-amplified using
appropriate primers and cloned into the pET28a expression vector (N ovagen, Madison
WI). The resulting insoluble fusion protein was purified by washing inclusion bodies
four times with 1% Triton X-100, 50 mM Hepes-KOH pH 7.6, 10 mM Mng and one
time with 25 mM Hepes-KOH pH 7.0, 8 M urea. The pellet was resuspended in 6 M
guanidine-HCl and protein was precipitated from the supernatant with 10% TCA. The
protein was emulsified with Titerrnax adjuvent (Cthx Corporation, Norcross, GA) and

injected into a rabbit (Cocalico Biologicals, Rearnstown, PA).

93

3.2. 6 Immunoprecipitation

For imrnunoprecipitation reactions, solid NaCl was added to carbonate-washed
solubilized Arabidopsis protein to a final concentration of 200 mM. The Arabidopsis
protein was pre-cleared by incubation with 1/10 volume of 50% slurry of protein A-
Sepharose beads (Pharmacia) in buffer A (25 mM Pipes-KOH pH 7.5, 50 mM NaCl, 2
mM EDTA pH 8.0). The resulting supematants were incubated with 50 p1 of immune or
pre-immune anti-AtF T1 serum for 1 hour. 1/5 volume of protein A-Sepharose slurry was
added to precipitate the antigen-antibody complexes and the samples were incubated for
an additional 3 hours with rocking at 4° C. Samples were then centrifuged, washed five
times in buffer A containing 1% Triton X-100 and two times in buffer A without
detergent. The pellets were resuspended in buffer A to a final volume of 120 pl and

assayed for AtFUTase activity.

3. 2. 7 Heterologous Expression

Cos-7 cells were grown on 100-mm plates in DMEM-10% fetal bovine serum. Cells
were transfected with different plasmids using LipofectarnineTM reagent (Life
Technologies) following the manufacturer's instructions using 9 pg of DNA and 72 pg of
Lipofectamine. Cells were incubated for 24 hours in the medium containing DNA-
Lipofectamine without fetal bovine serum. The medium was changed to DMEM-10%
fetal bovine serum and incubated for another 48 hours. The cells were scraped off the
dish in 0.25 M sucrose, 10 mM Tris-HCl pH 7.5 and 0.4% CHAPS. XyG-FUTase
activity was measured using 50 pg of protein in the absence (-XG) or presence (+XG) of

100 pg tamarind xyloglucan. The incubation was carried out in a volume of 0.1 mL in

94

the presence of 1 pM GDP-Fuc (93,000 dpm), 10 mM MnClz, 20 mM Hepes pH 7.0,
0.05% Triton X-100 at 25°C for 90 min. The reaction was halted by addition of ethanol
to a final concentration of 70%. Samples were incubated at 4°C and filtered through 1.5-
pm glass fiber filters. The filters were washed with 70% ethanol containing 1 mM

EDTA. The filters were dried and radioactivity determined by liquid scintillation. A

biological control using pea Golgi vesicles was carried out in parallel.

3.3 Results

3.3.] Cloning and Analysis of Arabidopsis Xyloglucan F ucosyltransferase

The previous chapter of this thesis described the purification of XyG FUTase from pea
and the sequence of several peptides from this purified protein. Peptides derived from
this approximately 60 kDa pea polypeptide were not homologous to any known proteins
in the databases but did allow us to identify an Arabidopsis Expressed Sequence Tagged
(EST) clone which contained these peptide sequences with 63%—86% conservation of
identity (Table 4). A RNA gel blot performed with the EST 191A6T7 as a probe (858
bp in length) indicated a transcript approximately 2.0 kb long, from which we concluded

that the EST was not a full-length cDNA clone (Figure 6).

95

Table 4. Comparison of pea xyloglucan fucosytransferase peptides and
corresponding Arabidopsis peptides.

Pea Peptides Arabidogsis Pegtide % Identity

VFGFLGR VFHHLGR 8 6 %
YLLH PTNNVWGLVVR YLFH PTNQVWGLVTR 8 0 it
AVLITSLSSGYFEK AVLVTSLNAGYAEN 6 4 it
YYDAYLAK YYEAYL SH 6 3 %
LLGGLLADGFDEK LLGGLLASGFDED 8 5 %

96

ell culture

    

Figure 6. RNA gel blot using an EST for AtFUTI as a probe. RNA fi'om Arabidopsis
cell suspension culture, stem, roots, leaves, flowers, or 7-day old mdlings was used for
Northern blot analysis with radiolabeled AtF UT 1 as a probe.

97

The EST was used as a probe to screen an Arabidopsis cDNA library. Two cDNA clones
were isolated, the longest containing a 1768-bp insert. Both lacked 13 nucleotides of the
3’ UTR and the poly-A tail found in 191A6T7. The lack of the poly-A tail and part of the
3’ UTR was not surprising, as inserts from the library used for screening had been
described as beginning 8-10 nucleotides from the polyA-tail due to Klenow 3’-5’
exonuclease activity during library construction. The cDNA clones contain a 1677-nt
open reading frame encoding a 63.5 kDa protein and correspond to a region of the fully
sequenced Arabidopsis bacterial artificial chromosome (BAC) T18El2 derived from
chromosome H (nucleotides 41209-41503, 41780-43252), which has been fully
sequenced by the Arabidopsis Genome Initiative. There is an AATAAA consensus
polyadenylation signal eight nucleotides from the 3’ end of the library-derived clones.
The translation product of this ORF also contains a region near the N-terminus similar to

a fifth pea peptide (see Chapter 2).

The cDNA and corresponding genomic clone were originally designated AtFTI ;
however, the name was changed to AtF UT 1 in July, 2001, to avoid consfusion with an
Arabidopsis gene named Flowering Locus T (FT) or Arabidopsis farnesyltransferases
(also abbreviated FT) (Kardailsky et al., 1999; Yalovsky et al., 2000). Analysis of the
BAC indicates that there may be a second glycosyltransferase approximately 300 bp
downstream from AtFTI that is 63% similar to AtFTI at the amino acid level. In addition
to this gene, named AtF UT 2, eight other AtF UT 1 -like genes have been identified

(described in Sarria et al., 2001.) Thus, Arabidopsis may have a family of FUTases,

 

each differentially regulated by such factors as environmental stress, tissue localization,

or developmental stage, or specific to different acceptors (Sarria et al., 2001, in press).

The protein encoded by AtF UT 1 is not significantly homologous to other proteins in the
databases, though low (40-60% over short regions) similarity to NodZ is observed. NodZ
is a bacterial alpha 1,6-fucosyltransferase involved in the synthesis of nodulation factors
in Rhizobium (Quesada Vincens et al., 1997). A motif which is found in all alpha-1,2 and
alpha-1,6 fucosyltransferases described to date is also present in the putative Arabidopsis
fucosyltransferase (Table 5) (Breton et al., 1998). Subsequently, comparison to other
FUTases demonstrated that AtFUTl contains a motif that was previously thought to
discriminate between alpha-1,2 and alpha-1,6 F UTases (analysis conducted by C.
Wilkerson and A. Faik) (Faik et al., 2000). A Kyte-Doolittle hydrophobicity plot of
AtFU T1 indicated that this protein contains a short hydrophilic extreme N-terminus

followed by a hydrophobic region and a largely hydrophilic C-terminus (Figure 7).

99

Table 5. Comparison of motifs found in alpha-1,2 and alpha-1,6
fucosyltransferases. (Motif described in Breton et al., 1998).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Protein Motif Sequence
Lig or-(l ,6)-FucT IGVHVRRTD
Hum a—(l ,6)-FucT IGVHVRRTD
Azorhizobium caulinodans NodZ (or- IGVHIRFGN
1,6)
Bradyrhizobiumjaponicum NodZ (a- IGIHVRHGN
1,6)
Yersinia enterocolitica a—(l ,2) FucT VGIHIRRGD
Rabbit Se-III (or—(1,2) FucT) VGVHVRRGD
Rabbit Se-II (tr—(1,2) FucT) VGVHVRRGD
Mouse Secl (or—(1,2) FucT) VGVHVRRGD
Pig H (or-(1,9 FucT) VGVHVRRGD
Mouse H (or—(1,2) FucT) VGVHVRRGD
Human Se (or—(1,2) FucT) VGVHVRRGD
Rabbit H (tr—(1,2) FucT) VGVHVRRGD
Human H (or-(1,2) FucT) VGVHVRRGD
Consensus (V/I)G(V/I)H(V/I)Rr(G/T) (D/
N)
AIF U T I IGIQVRVPD

 

 

100

 

flilllllllllllllllllll1
50 100 150 200 250 300 350 400 450 500 550

 

 

Figure 7. Kyte-Doolittle hydrophobicity plot of AtFUTl. Plot was generated using
the Protean component of the DNA" software package. Positive values are hydrophilic,
negative values are hydrophobic.

101

3.3.2 In vitro Transcription and Translation

The AtF UT 1 cDNA clone was used for in vitro transcription and translation experiments.
Translation reactions using 35 S-methionine produced several major products, presumably
corresponding to translational initiation at various N-terminal Met residues. Translation
was conducted in the presence of canine pancreatic microsomes. Such microsomes have
been used to promote in vitro translation of integral membrane proteins by allowing
association with the membranes (for example, see Bayle et al., 1995). Following
translation, a portion of the reaction was subjected to treatment with 0.1 M sodium
carbonate, which strips away proteins that are not integrally inserted into membranes
(Fujiki et al., 1982). The sodium carbonate-washed sample was separated by
centrifugation into supematent and pellet components and analyzed by electrophoresis
and autoradiography (Figure 8.) The largest molecular weight polypeptide from the
translation product, representing the full-length protein, is only present in the pellet

sample after treatment with sodium carbonate.

102

 

s P r LH,o

'5. i

— 207 kD

—121kD

— 81kD

 

- 51.2 kD

 

"' 33.6 kD
- 28.6 kD

Figure 8. Sodium carbonate treatment of in vitro-transcribed and translated
AtFUT 1. In vitro transcription and translation followed by sodium carbonate treatment
indicates the highest molecular-weight translation product (arrow) is a membrane protein.
Translation reactions were conducted in the presence of canine pancreatic microsomes.
After translation, an aliquot was removed (“T” for Total), and the remainder was
homogenized in the presence of 0.1 M sodium carbonate on ice for 30 min. The sodium
carbonate-treated sample was centrifuged at 100,000 g for l h to separate soluble (“S”)
and pellet (“P”) samples. “L”, B-lactamase - positive control (precursor, 31.5 kD; after
processing, 28.9 kD), “deO”, water negative control.

103

3.3.3 Immunoprecipitation of XyG FU T use Activity

To confirm the identity of AtF UT 1 as encoding a fucosyltransferase using XyG as an
acceptor, we prepared polyclonal antibodies directed against AtF UT 1 overexpressed in E.
coli and used them to immunoprecipitate proteins from carbonate-washed, detergent-
solubilized Arabidopsis proteins. These antibodies were shown to recognize a protein of
the expected molecular mass in a sample of Arabidopsis microsomal proteins by
immunoblotting (analysis by Weiqing Zeng) (Perrin et al., 1999). The
immunoprecipitated proteins were then assayed for XyG FUTase activity. FUTase
activity found in pellets derived from immunoprecipitation reactions was 2.6-fold higher

when immune antiserum was used compared to preimmune serum (Figure 9).

104

 

Activity (nKat x 10‘)
h

 

 

 

o _
Preimmune Immune
Pellet Pellet

Figure 9. Immunopreciptation of XyG FUTase activity from Arabidopsis
microsomal proteins. Polyclonal antibodies raised against AtFUTl were used in
irnmunopreciptation reactions with carbonate-washed Arabidopsis microsomal proteins.
The resulting pellets were washed extensively and assayed for XyG FUTase activity.
This is an example similar to results seen in seven different replicates.

105

3.3.4 Expression of AtFUTl in Mammalian Cells

As a second means of demonstrating the function of AtFUTl , a cDNA encoding this
protein was expressed in the Cos-7 mammalian cell line. This line has no endogenous
fucosyltransferase activity towards XyG. Upon expression of AtF UT 1 in this system,
substrate-dependent FUTase activity was observed that was 41 times higher than in

control lines transfected with empty vector (Figure 10.)

106

 

 

 

 

t. 6-
,_
X
1'6 _
g 4
5
2‘
z: 2‘
H
O
<

0..

-XG +XG -XG +XG -XG +XG -XG +X

LCos-T‘I LCos-7J l'Cos-7'-l L Pea
+vector +AtFUT1 Golgi

Figure 10. AtF U T1 expressed in a COS cell line shows XG-specific FUTase activity.
Activity in nKat is shown in the presence or absence of tamarind XG for untransformed
Cos-7 cells, cells transfonned with vector DNA (Cos-7 vector), cells transformed with
vector containing AtF T1 (Cos-7 vector AtFTI), or solubilized pea Golgi vesicles (Pea
Golgi). For graphs, error bars show i 1 standard deviation; if no error bars are visible,
standard deviations are contained within the width of the plot element.

107

3.4 Discussion

The identification of an Arabidopsis homolog of the purified pea XyG FUTase allowed
molecular and genetic approaches to the study of XyG biosynthesis. The resources
available to Arabidopsis researchers include a fully sequenced genome, publically
available EST databases, collections of T-DNA and transposon mutants, expression
analysis using DNA microarrays or DNA chips, an assortment of service facilities, and
countless available protocols for genetic, molecular, physiological, cell and biochemical
analysis. One of the most valuable resources, though, is an active community of

researchers working on this model organism.

Despite the tools that are available, however, few cell wall biosynthetic genes had been
identified in Arabidopsis at the time that this project was initiated. Even at the point of
conclusion of this thesis work, AtFUTI joins only a handful of other identified cell wall
biosynthetic genes, many of which have been identified by homology to other genes

(such as the collections of cellulose synthase and cellulose synthase-like genes).

Identification of a full-length version of AtFUTI began by performing a Northern blot to
determine how much of the coding region was covered by the existing EST. At that time,
it was not known how many other fircosyltransferase-like genes existed in the
Arabidopsis genome. The Northern blot indicated a transcript approximately 2 kb long in
various tissues, which indicated that the EST was not full-length. However, it should be
noted that a genome-wide survey for AtF UT 1 -like genes has demonstrated the presence

of nine other family members (Sarria et al., 2001). Because the entire EST was used as a

108

hybridization probe for the Northern blot shown in Figure 6, this blot most likely
demonstrates a composite expression profile for multiple family members — a situation
which is very likely considering the heterogenous transcript sizes observed between
samples. Other expression data will be shown in Chapter 4. However, it is true that EST

191A6T7 does not encode a full-length cDNA for AtFUTI .

The library used to identify cDNA clones for AtFUTI was originally constructed using
RNA from 3-day old Arabidopsis seedling hypocotyls (Kieber et al., 1993). cDNAs
derived from this RNA had been size-fractionated, and the library subset containing 2-3
kb cDNA inserts was used for screening. Two clones were identified encoding AtFU T1.
Sequence analysis of the encoded protein was promising, in that a motif characteristic of
or-(1,2) and or-(l,6) F UTases is present (Table 5). The gross structural features were also
characteristic of a fucosyltransferase; for example, a hydrophobic region is present at the
N-terminus that could act as a signal-anchor domain (Figure 7). In vitro transcription and
translation supported the idea that AtFUTl is an integral membrane protein (Figure 8),
and other studies have shown that this protein is Golgi-localized (Sarria et al., 2001). The
absence of the highest molecular weight product from the soluble fraction indicates that
the full-length protein has an integral transmembrane domain and cannot be disassociated
from the canine pancreatic membranes by sodium carbonate treatment. However, protein
arising from in vitro translation experiments did not possess XyG FUTase activity (data
not shown), possibly indicating that the protein could not assume a conformation

necessary for catalytic activity or that other interacting proteins were required.

109

Two approaches were taken to determine the biochemical firnction of AtFUTl.
Immunoprecipitation experiments showed that polyclonal antibodies directed against a
recombinant, tagged version of AtFUTl could precipitate XyG FUTase activity from
solubilized membrane proteins, indirectly showing that the cDNA clone did encode a
functional XyG F UTase enzyme (Figure 9). Heterologus expression experiments also
directly demonstrated that the AtF UT 1 cDNA can confer XyG FUTase activity to a
mammalian cell line which is normally incapable of transferring fucose to XyG (Figure
10). Subsequent studies have also used tobacco cell cultures and Pichia pastoris as
heterologous expression systems for AtFUTI (A. Faik, T. Wagner). Tobacco, like other
Solanaceous species, does not possess fucosylated XyG and no background XyG

fucosylation activity is detected in tobacco cell cultures.

Identification of an Arabidopsis XyG FUTase allows several areas of study to be
undertaken, several of which will be described in Chapter 4. One of the most important
consequences is that identification of this cell wall biosynthetic gene should allow other
candidate genes to be identified in Arabidopsis or other plant systems. Expression
profiling should aid in assessing cell wall biosynthetic gene candidates, as it may
reasonably be expected that genes involved in primary cell wall synthesis will be
coordinately regulated. Analysis of plants affected in this gene will also provide insight

into the roles of primary cell biosynthesis in general and XyG fucosylation in particular.

110

References

Bayle, D., Weeks, D. and Sachs, G. (1995). The membrane topology of the rat
sarcoplasmic and endoplasmic reticulum calcium ATPases by in vitro scanning. Journal
of Biological Chemistry 270, 25678-25684.

Breton, C., Oriol, R. and Imberty, A. (1998). Conserved structural features in
eukaryotic and prokaryotic fucosyltransferases. Glycobiology 8, 1-8.

Faik, A., Bar Peled, M., DeRocher, A. E., Zeng, W., Perrin, R. M., Wilkerson, C.,
Raikhel, N. V. and Keegstra, K. (2000). Biochemical characterization and molecular
cloning of an alpha-1,2-fircosyltransferase that catalyzes the last step of cell wall
xyloglucan biosynthesis in pea. J Biol Chem 275, 15082-9.

Fujiki, Y., Hubbard, A. L., Fowler, S. and Lazarow, P. B. (1982). Isolation of
intracellular membranes by means of sodium carbonate treatment: application to
endoplasmic reticulum. Journal of Cell Biology 93.

Arabidopsis Genome Initiative. (2000). Analysis of the genome sequence of the
flowering plant Arabidopsis thaliana. Nature 408, 796-815.

Kardailsky, 1., Shukla, V. K., Ahn, J. H., Dagenais, N., Christensen, S. K., Nguyen,
J. T., Chory, J., Harrison, M. J. and Weigel, D. (1999). Activation tagging of the floral
inducer FT. Science 286, 1962-1965.

Kieber, J. J., Rothenberg, M., Roman, G., Feldmann, K. A. and Ecker, J. R. (1993).
CTRl , a negative regulator of the ethylene response pathway in Arabidopsis, encodes a
member of the raf family of protein kinases. Cell 72, 427-441.

Perrin, R. M., DeRocher, A. E., Bar Peled, M., Zeng, W., Norambuena, L., Orellana,
A., Raikhel, N. V. and Keegstra, K. (1999). Xyloglucan fircosyltransferase, an enzyme
involved in plant cell wall biosynthesis. Science 284, 1976-9.

Quesada Vincens, D., Fellay, R., Nasim, T., Viprey, V., Burger, U., Prome, J. C.,
Broughton, W. J. and Jabbouri, S. (1997). Rhizobium sp. strain NGR234 NodZ protein
is a fucosyltransferase. J Bacteriol 179, 5087-5093.

Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning: A
Laboratory Manual. Cold Springs Harbor, NY: Cold Springs Harbor Press.

Sarria, R., Wagner, T.A., O'Neill, M., Faik, A., Wilkerson, C., Keegstra, K.,

Raikhel, N.V. (2001) Characterization of a family of Arabidopsis genes related to
xyloglucan fucosyltransferasel. Plant Physiol. (in press).

111

 

Yalovsky, S., Kulukian, A., Rodriguez-Concepcion, M., Young, C. A. and Gruissem,
W. (2000). Functional requirement of plant famesyltransferase during development in
Arabidopsis. Plant Cell 12, 1267-1278.

112

“A?

 

CHAPTER FOUR

Analysis of Xyloglucan Fucosylation as a Model Cell Wall Biosynthetic Process

lrnages in this dissertation are presented in color.

I performed all experiments described in this chapter with the exception of xyloglucan
analyses performed in collaboration with Drs. William York and Zhonghua Jia (Complex

Carbohydrate Research Center, University of Georgia, Athens GA.)

113

4.1 Introduction

The previous chapter described the identification of Arabidopsis xyloglucan (XyG)
fucosyltransferase (AtF UT 1 ), a cell wall biosynthetic enzyme that adds a terminal fucose
residue to galactose units on XyG in an or-(l,2)- linkage. This constitutes one of the last
steps of XyG biosynthesis before the polymer is exported from its site of synthesis in the

Golgi to the apoplast and incorporated into the growing cell wall.

AtFUTl encodes one of only two biosynthetic enzymes at the time of this writing for
which a suitable assay have been developed, whose carbohydrate product can be readily
analyzed, and whose gene has been identified. Only galactomannan galactosyltransferase
also falls into this category (Edwards et al., 1999). Few genes encoding cell wall
biosynthetic enzymes have been identified at the molecular level, and for some enzyme
activities (such as cellulose synthase), no assays have been developed to demonstrate
their biochemical function. Several reasons for these limitations have been outlined in
previous chapters and include enzyme lability, unavailability of substrates for assays, and
a lack of sequence similarity between glycosyltransferases in general that hinders gene
identification. Thus, XyG FUTase is a suitable model system to study cell wall
biosynthesis at the levels of gene expression, protein activity, and carbohydrate product

structure.

There is evidence to support the notion that cell wall biosynthesis is a regulated process.

As tissue expands and differentiates, cells comprising that tissue ofien enlarge, elongate

or adapt characteristic shapes. Specific types of cell wall material are required in all of

114

. -5“M q

these cases. Biosynthesis alone is only half of the overall picture during the growth
process, because increased deposition of wall material without a corresponding increase
in extensibility or means to incorporate new polymers into existing networks would only
result in thicker walls, not necessarily increased cell size. However, if the cell were
unable to increase biosynthesis of wall material during expansion, it would eventually
suffer a loss of wall integrity if the wall thinned to the point of breakage (Cosgrove,

2000)

Determination of expression patterns of AIF UT 1 will also enable identification and
evaluation of other candidate genes encoding cell wall biosynthetic enzymes. Co-
regulation has been described for other biosynthetic processes in plants. In peppermint,
the enzyme activity and gene expression profiles of several members of the monoterpene
biosynthesis pathway showed coordinate regulation during the development of oil-
producing trichome glands (McConkey et al., 2000). Several genes involved in the
biosynthesis of carotenoids show similar expression pattern profiles in various cultivars
of marigold, with deeply colored cultivars having up-regulation of multiple pathway
members and white cultivars having little expression (Moehs et al., 2001). The
completion of the Arabidopsis genome sequencing project in combination with global
profiling techniques, such as DNA microarray or Affymetiix chip technologies, should
further facilitate the identification of co-regulated sets of genes. Careful investigation of
the regulation of AtF UT 1 gene expression will allow the identification and evaluation of
other gene encoding cell wall biosynthetic enzymes. Although expression profiling by

itself is not diagnostic of gene function, cell wall biosynthetic gene candidates that have

115

expression profiles similar to AtF UT 1 , a known enzyme involved in primary cell wall
biosynthesis, may reasonably be considered to be stronger candidates than those that do

not.

In this chapter, the regulation of AtF UT 1 expression and enzyme activity are investigated.
Sets of tissue from Col-0 Arabidopsis were generated and used to study AtFUTI
expression, XyG F UTase activity, and XyG fucosylation. The genomic region upstream
of AtF UT 1 was isolated for preparation of AtF UT 1 .° :6 US-GFP promoter-reporter
constructs. Regulation of AtF UT 1 expression was analyzed in various tissues and in
response to various conditions using AtFUT1::GUS-GFP transgenic plants. In addition,
355: :AtF UT 1 over-expressing transgenic plants were generated and analyzed for AtFUTI

gene expression, XyG F UTase activity, and XyG structure.

4.2 Materials and Methods

lrnages in this dissertation are presented in color.

4.2.] Plant Growth Conditions

Arabidopsis plants were grown under hydroponic conditions as previously described
(Gibeaut et al., 1997). Opaque plastic storage boxes (Rubbermaid, Wooster OH) were
filled with hydroponic growth medium (1.25 mM KN03, 1.5 mM Ca(NO3)2, 0.75 mM
MgSO4, 0.5 mM KH2P04, 50 pM KCl, 50 pM H3BO3, 10 pM MnSO4, 2 pM ZnSOi, 1.5
pM CuSO4, 0.075 pM (NH4)6M07024, 0.1 mM Na203Si, 72 pM Fe-diethylenetriamine
pentaacetate (Sequestrene 330, Becker Underwood Inc., Ames IA)) and plugs were

prepared from Rockwool (GrodanHP, Agro Dynamics, East Brunswick, NJ) with a cork

116

borer. These plugs were inserted into holes drilled into lids for the plastic containers.
Glass Pasteur pipets were used to drill a channel through the middle of the Rockwool
plugs. WT Col-0 seeds were suspended in 0.1% agarose and placed onto the center of the
plugs. The tanks were placed in growth chambers (16 h days, 22 ° C) and aeration of the
nutrient solution was provided using small aquarium pumps. After germination, the
plants were thinned to one per Rockwool plug. Plants were allowed to grow until

maturity, approximately 8 weeks under these conditions.

Soil-grown plants were grown on greenhouse-prepared soil media suitable for
Arabidopsis under the same growth chamber conditions. Seeds were suspended in 0.1%
agarose, distributed onto soil surface, and subjected to 4° C for three days to promote

uniform germination before transferring pots to the growth chamber.

Agarose medium for plant growth (4.3 g/L MS salts, 100 pg myo-inositol, 10 pg
thiamine-HCl, 1 pg nicotinic acid, 1 pg pyridoxine-HCl, 1% sucrose, 2.6 mM MES, pH
adjusted to 5.7 with KOH, 1.5% phytagar) was used to prepare sterile plates. For
selective plates, hygromycin (50 pl/ml) was added after the media had cooled. In some
cases, vancomycin (500 mg/L) was added to control contamination. Seeds were
sterilized with a series of washes, all 20 minutes long, on a platform rocker: one wash in
ethanol solution (80% ethanol, 0.1% Triton X-100) one wash in bleach solution (30%
bleach, 0.1% Triton X-IOO), and three washes in sterile distilled water. After

sterilization, seeds were suspended in sterile 0.1% agarose and plated on growth medium.

117

4.2.2 Collection of Tissue Panels

Hydroponically- grown Arabidopsis was harvested at maturity and tissue was collected on
ice. Samples harvested were flowers, the uppermost 6 cm of the inflorescence stem, the
lowest 6 cm of the inflorescence stem, siliques, cauline leaves, rosette leaves, and roots.
Tissue samples were weighed and divided into portions used for Golgi vesicle
preparations, RNA preparations, and cell wall preparations. Golgi vesicles were prepared
using fresh tissue, while RNA and cell wall samples were prepared using tissue that was

frozen in liquid nitrogen and stored at —80° C.

4.2.3 RNA Preparation

RNA was prepared as previously reported in a method designed for use with pine tree
samples (Plant Molecular Biology Reporter (1993) 11:113-116) with revisions for
Arabidopsis (C. Wilkerson and S. Myers, MSU, East Lansing, MI). Extraction buffer
(2% hexadecyltrimetlrylammonium bromide, 2% polyvinylpyrrolidone K 30, 100 mM
Tris-Hcl pH 8.0, 25 mM EDTA, 2.0 M NaCl, 0.5 g/L spermidine, 2% B-mercaptoethanol
added just before use) was kept at 65° C. Tissue was ground under liquid nitrogen to a
fine powder and added to the pre-warmed extraction buffer (5 ml buffer per gram of
tissue), then mixed by inverting the tube and vortexing. Samples were extracted twice
with an equal volume of chloroform. One-quarter volume of 10 M LiCl was added to the
supematent to precipitate RNA at 4° C for several hours. The samples were centrifuged
(7000 rpm, 5800 g) at 4° C for 20 minutes and pellets were resuspended in Tris-EDTA
solution (0.5% sodium dodecyl sulfate, 10 mM Tris pH 7.5 (DEPC-treated), 1 mM

EDTA (DEPC-treated». An equal volume of phenol/chloroform/isoamyl alcohol

118

(25:24zl) was added to the samples, which were vortexed and centrifuged at room
temperature for 15 minutes (3000 rpm, 1000 g). The upper phase was removed and 1/4
volume 10 M LiCl was added. Samples were incubated at 4° C for several hours, then
centrifuged (7000 rpm, 5800 g) for 20 minutes at 4° C to pellet RNA. The pellets were
resuspended in 2 m1 DEPC-treated water, and 1/10 volume 3M sodium acetate pH 5.2
and 2.5 voltnnes ethanol were added. The samples were centrifuged again (7000 rpm,
5800 g) for 15 minutes to pellet RNA. Supematent was removed, an appropriate amount
of DEPC-treated water was used to resuspend the pellet, samples were centrifuged briefly
to remove any carbohydrate contamination, and an aliquot was removed for

quantification.

4.2.4 Expression Analysis by RT-PCR

RNA samples were treated with RNase-free DNase I using the DNA-flee kit (Ambion,
Austin, TX) according to the manufacturer’s instructions. RNA (1 pg) samples were
used in a reverse transcription reaction with 0.5 pg oligo dT primer (Invitrogen, Carlsbad,
CA), 1X first strand buffer, 10 mM DTT, 5 mM dNTPs, and 200 U Superscript II reverse
transcriptase (Life Technologies, Rockville, MD). RNA, water and the reverse primer
were incubated at 70° C for ten minutes, transferred to ice for 2 minutes, centrifuged, and
first strand buffer, DTT and dNTPs were added. The samples were incubated at 42° C for
1 minute and reverse transcriptase was added. The samples were finther incubated at 42°
C for 50 minutes and then at 70° C for 15 minutes. PCR was performed using an equal

volume (0.8 p1) of the original reverse transcription reactions as template in 20 pl PCR

reactions with 1X PCR buffer, 150 pM MgClz, 125 pM each forward and reverse primer,

119

and 5 U Amplitaq (Roche, Mannheim, Germany). Primers used to amplify AtFUTl
were: Foward primer: 5’ GAA GGG CTA CTT GCT TCT GGT TTT 3’; Reverse
primer: 5’ CCC GAT GAA TGT TTG GTC TCC TT 3’. These primers amplify a 578 bp
fragment of AtFUTI from nucleotides 571-1149. Thermal cycling parameters for
amplification of AtF UT 1 were: 94° C, 1 minute (hot start); 92° C, 30 seconds, 58° C, 1
minute, 72° C, 1 minute 30 seconds for 22 to 24 cycles (which is within the linear range
of amplification); 72° C, 5 minutes. Primers used to amplify cytochrome c were:
Forward primer: 5’ TCG CTT ATI‘ TGA AGG AAG TG 3’; Reverse primer: 5’ CTC
TTC ACA TCA ATA GCGT AAT 3’. These primers amplify a 212 bp fragment of
cytochrome c (AtGI number TC87099), located on chromosome 4. Thermal cycling
parameters for the amplification of cytochrome c were: 94° C, 1 minute (hot start); 94° C,
30 seconds, 55° C, 30 seconds, 72° C, 1 minute for 28 cycles (within the linear range of
amplification); 4° C. After amplification, samples were analyzed by electrophoresis on a

1% agarose/1X TAB gel.

4.2.5 Golgi Vesicle Preparations

Golgi vesicles were prepared using the method of Munoz et al., adapted for use in
Arabidopsis (Munoz et al., 1996). Tissue was weighed and 1 volume (g/ml) of 0.5 M
sucrose solution was added (0.5 M sucrose, 0.1 M KH2P04 pH 6.65, 5 mM MgC12, 1 mM
DTT, and EDTA-free Complete protease inhibitor cocktail tablets (Roche, Mannheim,
Germany) according to the manufacturer’s instructions). The tissue was chopped by
hand with a razor blade, ground with a mortar and pestle, and homogenized with a Pyrex

Dounce homogenizer. The homogenate was filtered through Miracloth and centrifuged

120

at 1000 g for 5 minutes. The supematent was removed and layered on top of a 1.3 M
sucrose solution cushion (1.3 M sucrose, 0.1 M KH2P04 pH 6.65, 5 mM MgC12) and
centrifuged at 100,000 g for 90 minutes in a swinging bucket rotor. The upper phase was
removed without disturbing the interface and 1.1 M sucrose solution (1.1 M sucrose, 0.1
M KH2P04 pH 6.65, 5 mM MgC12) was layered on top, followed by a layer of 0.25 M
sucrose solution (0.25 M sucrose, 0.1 M KH2P04 pH 6.65, 5 mM MgClz). The samples
were centrifuged at 100,000 g for 100 minutes in a swinging bucket rotor. The 0.25
M/1.1 M sucrose interface fraction was collected, 1 volume of water was added, and the
samples were centrifirged at 100,000 g for 50 minutes in a fixed angle rotor. The
supematent was discarded and the pellet was resuspended in STM buffer (0.5 M sucrose,
2 mM MgC12, 20 mM Tris-HCl pH 7.5, and protease inhibitor cocktail tablets.) Aliquots

of the preparations were reserved for quantification by the Bradford assay.

4. 2. 6 Activity Assays
Activity assays were conducted as described in Chapter 2 of this thesis. All assays of
Golgi vesicle preparations included 1% Triton X-100 to permeabilize the vesicles,

allowing accessibility of exogenous substrates to the enzyme.

4. 2. 7 Xyloglucan Analysis

I . Cell Wall Preparations

Alcohol-insoluble residue (AIR) preparations were performed to gather cell wall material
for analysis. Tissue was collected, weighed, and fiozen in liquid nitrogen. The tissue

was ground to a powder under liquid nitrogen using a mortar and pestle, and 80% ethanol

121

If

was added (approximately 20 ml per gram of tissue). The tissue was further disrupted
using a polytron homogenizer. Cell wall material was collected using nylon mesh and
washed with 80% ethanol, then 100% ethanol. The washed material was suspended in
methanol: chloroform (1:1) and stirred for one hour at room temperature. The insoluble

material was collected on Whatrnan paper by filtration and washed with acetone.

Further preparation of cell wall material and analysis of XyG structure was performed at

the Complex Carbohydrate Research Center (Athens, GA).

2. Endo-polygalacturonase (EPG) and pectin methylesterase (PME) treatment. AIR
samples were treated simultaneously with EPG and PME to solubilize pectin oligo- and
polysaccharides as follows. AIR samples were suspended in 10 mL of 50 mM NaOAc,
pH 5, containing 0.01% thimerosal. EPG (5 units, where one unit releases 1 pM of
reducing galacturonate/min at 30°C), purified from Aspergillus niger (provided by Dr.
Carl Bergmann, CCRC), and PME (5 units) from A. oryzae (supplied by Novozymes
A/S) were added. The suspension was incubated for 24 h in a shaking incubator at 24°C.
The suspensions were filtered and the solid residues were treated a second time with EPG

and PME.

3. XG-specific endo-fl-I,4-glucanase (XEG) treatment. The residues obtained after EPG
and PME treatment were suspended in 10 mL 20 mM NaOAc, pH 5, containing 0.01%
thimerosal. XEG (10 units) was added and the suspension was incubated for 24 h in a

shaking incubator at 24°C, and then filtered. The filtrate, containing XG

122

 

oligosaccharides, was applied to a C-18 cartridge (Supelco Supelcelan LC-18 SPE Tube),
which was washed with water (10 mL) to remove salts. The oligosaccharides were then
eluted from the cartridge with 25% aq. MeOH (10 mL). The eluant was concentrated
under vacuum and lyophilized. The oligosaccharides were further purified on a Superdex
75 column (Pharmacia Biotech) and eluted with 50 mM ammonium formate, pH 5.0, at a
flow rate of 0.5 mL/min. A refractive index (RI) detector was used to monitor
carbohydrate in the eluant. Fractions containing the oligosaccharides were pooled and
lyophilized several times to remove volatile ammonium formate salts, affording pure

XEG-released oligosaccharides.

4. IN KOH and 4N KOH treatment. The residues from XEG digestion were suspended
in 10 mL of 1N KOH containing 1% NaBI-I4 and stirred for 24 h at room temperature.
The suspensions were filtered and the insoluble residues were then suspended in 10 mL
of 4N KOH containing 1% NaBH4. The suspensions were stirred for 24 h at room
temperature and then filtered. The filtrate was neutralized (pH 5) with glacial AcOH and
dialyzed (3,500 MWCO tubing; spectrum) against six changes of deionized water over 2
days. The retentates were lyophilized to afford the 4N KOH-solubilized XG fractions,
which were digested with XEG (as in step 3) to generate XG oligosaccharides. XG

oligosaccharides were isolated from the reaction mixture as in step 3.

5. Spectral Analysis. MALDI-TOF MS and NMR were performed on the
oligosaccharides purified from steps 3 and 4. MALDI-TOF MS (Matrix assisted laser

desorption time-of-flight MS) was performed with a Hewlett Packard LDI 1700XP

123

spectrometer, operating at an accelerating voltage of 30 kV, an extractor voltage of 9 kV,
and a source pressure of approximately 9 x 10-7 torr. The matrix used was a mixture of
2,5-dihydroxybenzoic acid (DHB, 0.2 M) and l-hydroxyisoquinoline (HIC, 0.06 M), both
in 50% aqueous acetonitrile. Oligosaccharides were dissolved in 0.6 mL of D20 (99.9%,
Cambridge Isotope Laboratories) and NMR spectra were recorded using a Varian Inova

600 (600 MHz for 1H) at 25°C.

4.2.8 Generation of AtFU T1 :: G US-GFP Transgenic Plants

Regions upstream of the AtF UT 1 ORF were amplified fiom Arabidopsis genomic DNA.
Three fragments were generated: 1) 1987 bp upstream of the ORF, 2) 3103 bp upstream
of the ORF, and 3) a 2934 bp fragment that included a 2 kb region upstream of the ORF,
the first exon, and the only intron of AtFUTI . The primers used for amplification were:
for fiagment l, 5’ GGG GGA TCC CTA TAG TGG CTG TCT GCT TGA GGA 3’
(primer F2) and 5’ GGGCCA TGG ATT GCT CTT GAG GGA 3’(primer KK506); for
fragment 2, 5’ GGG GGA TCC GTT TCT TGG GTG GAT GTT TGT T'I‘T 3’ (primer
F1) and primer KK506; for fiagment 3, primer F2 and 5’ TCG AAA CCC GGT ATT
AAC CA 3’. Primers included restriction sites for BamHI and NcoI digestion. Thermal
cycling parameters were: 96° C for 2 minutes (hot start); 94° C for 30 seconds, 57° C for
30 seconds, 72° C for 2 minutes for a total of 35 cycles; 4° C. Fragments were cloned in
pGemT-Easy vector (Promega, Madison, WI) and then into BamHI and NcoI sites in the
pCAMBIA 1303 vector. This caused a fusion of the first ATG codon originating from
AtF UT 1 with the gusA-mGFPS double reporter present in the pCAMBIA 1303 vector.

Selection in bacterial hosts and transgenic plants was provided by the kanamycin

124

resistance marker the hygromycin resistance marker, respectively. The constructs were
introduced into Agrobacterium tumefaciens by standard transformation methods and
transgenic Arabidopsis plants were generated using the vacuum infiltration method

(Bechtold and Pelletier, 1998).

4.2.9 GUS Staining Conditions

Whole plant seedlings or plant tissues were harvested and placed into cold 90% acetone
for 20 minutes on ice to permeabilize tissue. Samples were washed three times for 15
minutes each in working solution (100 mM NaH2P04, 10 mM EDTA, 0.5 mM
ferrocyarride, 0.5 mM ferricyanide, 0.1% Triton X-100). Samples were then incubated in
staining solution (working solution with 2 mM X-Gluc (Rose Scientific, Edmonton,
Canada)) at 37° C for 16-36 hours. Samples were cleared of chlorophyll by several
washes in 70% ethanol, then photographed using a Nikon Coolpix 995 digital camera
attached to a conventional light microscope or stereoscope. Whole organs of mature

plants were photographed without magnification.

4.2.10 T reatrnents with Hormones or Abiotic Conditions

AtF UT 1 : : GUS-GFP plants were treated with various hormones or environmental
conditions and stained for GUS as described above. Several treatments involved growth
of seedlings on vertical grth media plates for 9 days, then transfer to incubation media
(4.3 g MS salts/L, 10 mM Pipes pH 6.0, 50 pg/ml chloramphenicol) with or without
hormones (50 pM 1-naphthaleneacetic acid (NAA), 50 pM abscisic acid (ABA), 0.03%

BHT (a salicylic acid analog), 100 ‘pM gibberellic acid (GA3), 1 mM 1-

125

aminocyclopropane-l-carboxylic acid (ACC) or 2 pM epibrassinolide. Methyl jasmonate
(MeJ A) treatments were performed by adding 2 pl concentrated Me] A to 100 pl ethanol
on a growth plate containing seedlings in an air-tight plexiglass box. All incubations
occurred for 24 hours. Cytokinin treatment consisted of adding 6-benzylaminopurine
(BAP) to plate grth media to a final concentration of 4 pM. Seeds were plated on this
media and allowed to grow for 10 days before staining. Cold treatment involved
incubation of 9-day old plants at 4° C under growth lights for 48 hours. Etiolated growth

involved growing seeds on vertical growth plates wrapped in aluminum foil for 10 days.

4.2.11 Generation of 35S::AtF U TI Transgenic Plants

A cDNA fragment encoding the ORF of AtF UT 1 was subcloned into the NcoI/Xhol sites
of pETl4-b. The AtF UT 1 cDNA was then cloned into the SalI/Xbal sites of pCAMBIA
1300 MCS (a derivative of pCAMBIA with additional sites in the MCS generated by Dr.
Anton Sanderfoot, MSU, East Lansing, MI). This generated a construct in which the 35S
CaMV strong constitutive promoter controls the expression of AtF UT 1. The construct
was introduced into A grobacterium tumefaciens by standard transformation methods and
transgenic Arabidopsis plants were generated using the vacuum infiltration method
(Bechtold and Pelletier, 1998). Transgenic plants were selected using hygromycin and
the presence of the transgene was confirmed by PCR amplification across the AtF UT 1
intron, which results in a 489 bp product for the genomic version and a 212 bp product
for the transgene. The primers used for amplification were: forward primer, 5’ TCG
ACG CCG GAG T'I‘T TC 3’; reverse primer, 5’ AT‘T CAG TAC CCG GAC CAC ATC

3’. The thermal cycling parameters were 95° C for 1 minute (hot start); 95° C for 1

126

minute, 50° C for 1 minute, 72° C for 1 minute for a total of 30 cycles; 4° C. Products
were analyzed on agarose gels. RNA was prepared from T2 plants using the RNeasy
Mini Plant Kit (Qiagen, Germantown, MD) and used to analyze AtF UT 1 expression by
RT-PCR as described above. Alcohol-insoluble residue preparations were used to

analyze XyG structure as described above. 7

4.2. 12 Quantification of Elongation in Regions of Arabidopsis Inflorescence Stems

Growing inflorescence stems of 15 Arabidopsis plants were marked at four locations
using activated charcoal mixed with water, applied with a paintbrush. At time 0, two
marks 0.5 cm apart were made at the top of the inflorescence stem just below the flower
bud and two marks 0.5 cm apart were made at the bottom of the stem just above the
rosette. The distance between each set of marks was determined 48 hours later. For
measuring purposes, distance was recorded between the top of one mark and the top of

the second member of the set.

4.3 Results

Images in this dissertation are presented in color.

4.3.1 Gene Expression, Enzyme Activity, and XyG Structural Analyses of Arabidopsis
Tissues

Wild-type (Col-0) Arabidopsis plants were grown hydroponically and dissected at
maturity into various tissue types (Figure 11). These tissue samples were divided into
three portions: one for RNA extraction, one for preparation of Golgi vesicles, and one

for preparation of cell walls. Tissues collected were flower buds, defined as the cluster of

127

flowers at the end of an inflorescence stem; the 6 cm of inflorescence stem directly
beneath the flower buds; siliques; cauline leaves; the lowermost 6 cm of the inflorescence

stem; rosette leaves; and total roots.

128

flowers

top 6
cm of

stem Activity analysis

cauline
leaves

Plants dissected
and tissues
collected

 

Expression analysis

lowest
6 cm of
“em Carbohydrate (XyG)
analysis

rosette

leaves

 

roots

Figure 11. Tissue panel composition. Arabidopsis plants were grown in a hydroponic
system to maturity and dissected into seven different tissue types, as indicated. These
pooled tissues were separated into three different portions for Golgi vesicle preparation
and XyG FUTase activity assays, expression analysis by RT-PCR, and XyG structural
analysis.

129

The expression of AtFUTI in these tissues was assessed using RT-PCR (reverse
transcription-polymerase chain reaction). RNA was prepared from each tissue sample,
quantified, and treated with RNase-free DNase to remove any genomic DNA
contamination. One microgram of RNA from each tissue was used in a RT reaction with
oligo-dT as primer for synthesis of the first-strand cDNA. As a control to detect DNA
contamination of the original RNA preparations, one RT reaction contained leaf RNA,
but no reverse transcriptase (“No RT”). An equal volume of each RT reaction was used
as template for PCR amplification of either AtF UT 1 or cytochrome c. The primers used
to amplify AtF UT 1 have been determined to be specific for amplification of this gene
(Sarria et al., 2001.) In earlier applications of these primers for AtF UT 1 expression
analysis by RT-PCR, amplification was allowed to proceed for thirty cycles (Sarria et al.,
2001 ), which is beyond the linear range of product accumulation (data not shown).
Experiments were performed for both the AtF UT 1 primer set and the cytochrome c
primer set to determine the optimum cycle number endpoint for qualitative comparisons
of starting transcript amount (data not shown). Following PCR, the products were
analyzed on an agarose gel and stained with ethidium bromide (Figure 12). The highest
level of expression was detected in the uppermost region of the inflorescence stem,
followed by roots. It should be noted that previous analyses of gene expression
experiments have, at times, shown high levels of expression of AtF UT 1 in RNA derived

from flowers.

130

Top stem
Low stem

or
iii
a
.9
i7)

Cauline
Rosette
Roots

 

 

 

Figure 12. Expression profile of AtFU T1 by RT-PCR. RNA prepared fiom
Arabidopsis tissues and used for reverse transcription (RT) reactions primed with oligo-
dT. An equal volume of reverse the RT reactions was used for PCR with primers specific
for either AtF UT 1 (top) or cytochrome c as a standard gene control (bottom). PCR was
performed for 24 (AtFUTl) or 28 (cytochrome 0) cycles, which was within the linear
range of amplification as previously determined (data not shown). Multiple technical and
biological replicates yielded similar profiles (data not shown).

131

A second portion of the Arabidopsis tissue samples was used to prepare Golgi
vesicles. Xyloglucan fircosyltransferase is a Golgi-localized enzyme, and fucosylated
XyG accumulates primarily in the trans-Golgi and trans-Golgi network (Brummell et
al., 1990). Thus, Golgi vesicles have been used as an enriched source of XyG
FUTase enzyme activity (Wulff et al., 2000). The protein content of Golgi vesicle
extracts was quantified, and the samples were assayed for XyG FUTase activity in the
presence of detergent (1% Triton X-100). The detergent permits accessibility of the
nonfucosylated acceptor substrate to the enzyme, which is oriented such that the
active site is within the lumen of the Golgi (Wulff et al., 2000). The highest activity
levels occur in flowers, the upper portion of the inflorescence stem, and (to a lesser
extent) in the roots (Figure 13). Little or no activity could be detected in fully
expanded rosette leaves. Over fivefold higher activity was observed in the upper
region of the inflorescence stem compared to the lower-most region, which at the

time of collection had already lignified.

132

total cpmlpg protein

 

0 flowers siliques cauline upper lower rosettes root
stem stem

Figure 13. XyG FUTase activity of Golgi vesicle samples. Golgi vesicles were
isolated fiom the indicated tissues using the method of Munoz et al. (Munoz et al., 1996).
Xyloglucan fucosyltransferase assays were performed as described previously (Chapter
2). Data are averages of three replicates. An independent biological replicate resulted in
a nearly identical profile (data not shown). Error bars indicate one standard deviation.

133

The final portion of tissue samples was used for preparation of cell walls and analysis of
XyG fucosylation. Preparations of alcohol-insoluble residues (AIR) were made from the
various tissue samples, and xyloglucan was extracted from these preparations with
enzymatic and alkali treatment, after enzymatic removal of pectins (Figure 14).
Extracted XyG fractions were then digested with a XyG-specific endoglucanase that
cleaves XyG at unsubstituted glucan residues, releasing XyG oligosaccharides. NMR
and MS analyses were performed on the samples to determine the amount of fucose
present per unit of oligosaccharide, using spectral peaks indicative of the presence of Fuc
(Figure 15). Two independent biological replicates were performed (both shown in
Figure 15). In the first set, the amount of fucose present on XyG oligosaccharides
derived from 4 M KOH extraction of rosette leaves was approximately half the amount
present in other tissue samples. In the second biological replicate, the amount of Fuc
present was approximately constant among all tissue types. Analysis of XyG derived
from a series of developing leaves indicated no major differences in the XyG F no content

of leaves ranging from 1 week to 5 weeks old (data not shown).

134

 

 

Analysis of the Xyloglucans from Arabidopsis Cell Wall Materials]

 

 

Flcohol Insoluble Residue (AIR)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EPG/PME : EPG-1
50 111M NaOAc (pectin)
pH 5, 24 II
EPG/PME > EPG-II
so 111M NaOAc (pectin) a!!!“ min
1’“ 5. 24 I 005 3mg, Superdex-SD75

H20 and 25% 50 IM formic acid

- It H 5

XEG A, M6011 [1,0 XE G released (N319 )- P : Xyloglucan
so mM NIOAc soluble Lyophilized XGO' Lyopliilized 01180! ‘
pH 5. Stir 24 h

Adjust to pH

NMR

1” K0“ ; 1 iliilifeon M 11.4 16.911 t rial] MALDI-MS'
0.1% N 3 3° " ”NY“ 3" " m‘ ° air: sea
Stir, 24.. H‘ Lyophilized 'y

Adjust
4 M KOH ‘ 4 M KOH to pH 5-6 4 M KOH 1' XEG "an“: Xyloglucan
Stir, 24 Ii Lyophilized 3. Purification

4. Lyopliilbed

 

 

Cell wall residue
[Tullulosfl ]

Figure 14. Procedure for analysis of XyG fucosylation. Xyloglucan was extracted
from cell walls of Arabidopsis tissues and digested into oligosaccharide fragments. The
fucose content of xyloglucan oligosaccharide fragments was determined using NMR and
MALDI-MS analyses.

135

 

 

 

 

 

 

 

 

 

 

 

 

e 0.7

E

a

g 0.6## A

o

8 o -

a .a

8

= 0. ‘

O 4

‘3, o.3~——47

X

o _

'5 0.2

E

3 0.1‘

3

u. . . .

'3 Rosettes Top Stem Lower Stem Siliques Roots Cauline

5 Leaves
0.

 

 

 

 

 

Rosettes Top Stein Lower Stem Siliques Roots Cauline Flowers
Leaves

Mole Fuc I Mole XyG Oligosaccharide

Figure 15. Analysis of XyG fucosylation. Data were obtained by NMR and MALDI-
MS techniques that quantify the amount of fucose present per unit of XyG
oligosaccharide. Two independent biological replicates were performed (first replicate,
top; second replicate, bottom.)

136

4.3.2 Expression Profiling by RNA Gel Blot and Microarray Analysis

The expression of AtF UT 1 is very low and, therefore, diffith to assess using
hybridization techniques. As evidence of this, only one EST for AtF UT 1 (191A6T7,
from the mixed tissue library PRL2) exists in the public databases, despite the fact that
over 100,000 Arabidopsis ESTs have been sequenced. This makes expression profiling
of this gene by inventory of EST distribution (“electronic Northems”) impossible at

present.

Some hybridization techniques have been applied to detect AtF UT 1 expression, however.
Figure 16 shows a RNA gel blot in which 15 pg samples of total RNA were hybridized
with an AtF UT 1 -specific probe. Total RNA was extracted from total aboveground tissue
of soil-grown wild type (Col-0) plants one to six weeks old, as well as fiom flowers
(defined as fully opened individual flowers cut fiom the inflorescence stem), floral buds
(defined as unopened flowers present on the inflorescence apex), four different segments
of the inflorescence stem, rosette leaves, and roots (from liquid-grown plants). The
strongest signals for AtF UT 1 were observed in floral buds, the top of the stem, and roots.
Little if any signal was detected in rosette leaves and 6-week old plants. The blot was
then stripped and re-probed to detect eIF4a, a translation initiation factor, as a loading

control.

AtF UT 1 expression has also been detected using poly(A)+ RNA. Figure 17 shows the
result of hybridizing an AtF UT 1 -specific probe to poly(A)+ RNA derived from 250 pg of

total RNA extracted from leaf and root tissue of Arabidopsis grown in liquid medium.

137

Strong signal was detected in the root sample, while very little expression was observed
in leaves. Golgi vesicles were prepared from identical tissue (leaves or roots of
Arabidopsis grown in liquid media) and used for enzyme assays (Figure 17).
Approximately fivefold higher activity was associated with Golgi vesicles prepared from
root tissue versus leaf tissues. Thus, the steady-state levels of AtFUTI transcript
determined by RNA gel blot analysis positively correlate with the levels of XyG FUTase

enzyme activity for these tissues.

138

m
a "N
mm m .D SEE

xxxi:§€..§883m
oouooooofimmmgm
ooooouggmuu >4...
3:33:3°°%._._3 “8
~NMVm~eC¢~EE£2L
.,.. 1;“ a..' ‘3; 'M“"'T"‘ is)”:

AtFUTl

 

Figure 16. Expression profiling of AtFUTI by RNA gel blot analysis. Total RNA

 

was prepared from total aboveground tissue of soil-grown Arabidopsis at ages 1 through
6 weeks, or from organs of mature plants (flowers, flower buds, top 0.5 cm of
inflorescence stem (top stem), two successive 1.0-cm segments below this (mid stem 1
and mid stem 2), the remaining stem from the bottom of the mid stem 2 section to the
base of the inflorescence (low stem), rosette leaves, and roots). Fifteen micrograms of
total RNA was used for RNA gel blot analysis with an AtF UT 1 -specific hybridization
probe. The blot was then stripped and hybridized with a probe to detect eIF4a as a
loading control.

139

 

A33 BESO
02 3
— 240

a
330
E
8
v20
E
.2
.510
(I

 

 

leaf root

Figure 17. Expression of AtFUTI and XyG FUTase activity levels in leaves and
roots of Arabidopsis grown in liquid culture. Total RNA was prepared from leaf and
root tissue of Arabidopsis grown in liquid grth medium. Poly(A)+ RNA was prepared
from 250 pg of total RNA and used for RNA gel blot analysis with a probe specific for
AtF UT 1 (panel A). The same tissue was also used for Golgi vesicle preparations and
activity assays (panel B).

140

4.3.3 Expression Analysis Using Transgenic AtFUT1::GUS-GFP Plants

Three different constructs were made for generation of promoter-reporter transgenic
plants (Figure 18). These transgenic lines were used for in planta analysis of AtFUTI
gene expression. At the time that the AtF UT 1 : :GUS-GFP constructs were made, the
nearest predicted ORF upstream of AtFUTI was several kilobases away, which did not
allow delineation of the AtF UT 1 promoter region with certainty. The constructs
generated included 2-kb or 3.1-kb fragments of the genomic region upstream of the
AtF UT 1 start codon fused with the GUS-GFP double reporter present in the pCAMBIA
1303 vector. A third construct also included the first exon and the first and only intron of
AtFUTI , as introns have been reported to contain regulatory regions in rare instances
(e.g. Gidekel et al., 1996). It should be noted that the first exon contains the signal
anchor region that directs insertion into the Golgi membrane, and thus this could result in
Golgi-lumenal localization of the reporter proteins. This would have implications for the
accessibility of the X-Gluc substrate to the GUS reporter in transgenic plants containing

this construct.

Transgenic Arabidopsis plants containing these constructs were subsequently analyzed
for GUS expression (Figures 19-23.) Only lines with the construct containing the 2-kb
promoter fragment was used for systematic analysis of reporter expression (see
Discussion). Transgenic plants homozygous for single insertions of the transgene were
identified by analysis of resistance ratios for the hygromycin marker. T3 lines were used

for systematic gene expression analysis in seedlings (3 days, 5 days, 10 days, or 14 days

141

old) and in organs of mature soil-grown plants. GUS expression was first detected in the
veins of emerging cotyledons (Figures 19-24). Signal was later detected in emerging true
leaf buds and at the base and tips of true leaves (Figures 20 and 21). Some central vein
staining was occasionally noted in the hypocotyl. Little, if any, GUS expression was
detected in roots (Figure 19). In: mature organs, GUS expression was most notable at the
top of the inflorescence stem (Figure 24). Although signal was high in the uppermost
portion of the stem, it was not detectable in floral tissue itself. In true leaves, staining
could be observed in the midvein at the base of the leaf and at the leaf margins,
presumably representing hydathodes (Figure 23). Little staining was observed in cauline

leaves (Figure 23).

142

A.
[— 3 kb of AtFUT1 promoter - . a
[ 2 kb of AtFUT1 promoter
[ 2kb ofAtFUT1 ‘1’" .2"d ‘ gusA
, . ,P'°m°t°' _ _ , . 9x99 [239". ‘

 

 

 

 

 

 

Hin dIII(11600)

Sph I(11598)

B Psi I(11592
' Sal I(11582)
Xba I(11576)

Bum HI(1157O ‘

Kpn I(11565)

Sac I(11559) ,

Eco RI(llS49) 3:

Bst XI(11306) ‘
canvass promoter

Xho I (10519)

  
      
    
     
   
    
 
  
  
 
 

lacZ alpha
CaMV 35$ promoter

Nco I(l)
3g: 11 (8) f“ ("3589)
Spe I(13358-0 Glycosylatlon alto mutation

Pm! I(2151),mGFP5‘

hygromyoln (R) Nhe I(2541)||stldlno tag

Xho I (9425)

CaMV35$ polyA N
poly-A
T-Bordor (loft) -Bordor (right)
Sac ll (8907) Sph I(2979)

kanamycin (R)

[)8 R322 orl
[:3 R322 bom

Figure 18. Constructs generated for promoter-reporter analysis. Various fragments
of the AtFUT1 genomic region upstream of the coding region were amplified from
genomic DNA (panel A). These fragments were fused to the gusA-mGFP double
reporter in pCAMBIA 1303 (panel B).

143

 

D. E.

Figure 19. Expression of AtFUT1 in 3-day-old promoter-GUS transgenic plants.
Representative images are shown from staining of four different P2.0 transgenic lines.
Staining occurred in the veins of emerging cotyledons (A, B) and was largely absent fi'om
the hypocotyl-root junction (C) as well as roots (D). Occasional staining could be seen in
veins of the hypocotyl (E). GUS staining was variable, with 10% to 50% of individuals
in each homozygous single-insertion line showing visible stain accumulation. Bar = 100
um.

 

D. E.

Figure 20. Expression of AtFUT1 in 5-day-old promoter-GUS transgenic plants.
Representative images are shown from staining of four different P2.0 transgenic lines.
Staining occurred in veins of cotyledons (AD) and at base of emerging leaf buds (A, E).
GUS staining was variable, with 10% to 50% of individuals in each homozygous single-
insertion line showing visible stain accumulation. Bar = 1 mm for A-C, 100 pm for D
and E.

145

 

D F. F

Figure 21. Expression of AtFUT1 in 10-day-old promoter-GUS transgenic plants.
Representative images are shown from staining of four different P2.0 transgenic lines.
Staining occurred in veins of cotyledons (A, D) and at base and tips of emerging leaf
buds (A-F). GUS staining was variable, with 5% to 66% of individuals in each
homozygous single-insertion line showing visible stain accumulation. Bar = 1 mm for A-
D, 100 pm for E and F.

146

 

Figure 22. Expression of AtF UT 1 in l4-day-old promoter-GUS transgenic plants.
Representative images are shown from staining of four different P2.0 transgenic lines.
Staining occurred in veins of cotyledons (A-C) and in the central vein and tips and
midveins of true leaves (D, E). GUS staining was variable and generally low, with 0%-
10% of individuals in each homozygous single-insertion line showing visible stain
accumulation. Bar = 1 mm.

147

 

Figure 23. Expression of AtF U T 1 in promoter-GUS transgenic plants and organs.
Representative images are shown from staining of four different P2.0 transgenic lines.
Plants were grown on soil for 6 weeks and intact organs were stained for GUS activity.
Left to right: mature rosette leaf, upper inflorescence stem (just below flower buds),
lower inflorescence stem (just above rosette), flowers, silique, cauline leaf.

148

 

Figure 24. Expression of AIF U T 1 in intact stems of promoter-GUS transgenic
plants. Representative images are shown from staining of four different P2.0 transgenic
lines. Plants were grown on soil for 6 weeks and entire inflorescence stems were stained
for GUS activity. Left, overview of stem; right, close-up of flowers.

149

4. 3.4 Quantification of Elongation in Inflorescence Stem Regions

Since high expression of the reporter gene was detected in the uppermost region of
the inflorescence stem and little expression was observed in the lowermost regions,
elongation in these two areas was quantified. Marks were made on inflorescence
stems of growing Arabidopsis plants at the top of the stem, just below the flowers,
and at the bottom of the stern, just above the rosette. The initial distance between the
two markings was 5 mm. The distance between these sets of marks was recorded 48
hours later. Figure 25 shows the average distance between each set of markings at O

and 48 h.

150

 

M
(II

 

 

 

 

N
O

 

 

.3
0|

 

 

.L
O

interval distance (mm)
0|

 

 

 

0

top, 0 h top, 48 h bottom, 0 h bottom, 48 h

Figure 25. Elongation in upper and lower regions of Arabidopsis inflorescence
stems. Sets of marks 0.5 cm apart were made in the upper part of growing inflorescence
stems just below the flowers and the bottommost part of the stem, respectively, on 15
Arabidopsis plants. After 48 hours, the distance between each set of marks was recorded
and used to calculate the elongation rates of each region in mm/h. Error bars indicate
standard deviation.

151

4. 3.5 Analysis of Cis-acting Promoter Element Motifs in the AtFUT1 Promoter Region
The genomic sequence 1060 nt upstream of the AtF UT 1 ORF (see Discussion) was used
to search for cis-acting regulatory element motifs in two plant databases, PLACE
(http://www.dna.affrc.go.jp/htdocs/PLACE/) and PlantCARE
(http://sphinx.rug.ac.be:8080/PlantCARE/cgi/index.html). Promoter element motifs were
noted, compiled and annotated (Table 6). Several major categories of motifs were
observed, including motifs related to regulation by auxin, light, seed-specific expression,
pollen-specific expression, or other types of regulation including regulation by salicylic
acid (SA), elicitors, cold temperature, jasmonate or wounding. A number of consensus

sequences for transcription factor binding sites were also noted.

To investigate the regulation of AtF UT 1 in response to various hormones and growth
conditions, promoter-reporter seedlings were grown on plates for 9 days and then
transferred to incubation medium with or without various growth hormones (abscisic
acid, auxin (l-naphthaleneacetic acid (NAA)), BHT (a salicylic acid analog), gibberellic
acid (GA3), epibrassinolide, or methyl jasmonate (MeJA) (Figure 26)), or were grown in
darkness (Figure 26), subjected to cold treatment (data not shown), or wounded (data not
shown). Plants were treated with MeJA by adding 2 ul concentrated MeJA in 100 pl
ethanol to the surface of an agar plate containing 9-day old seedlings, and allowing the
MeJA to volatilize within an air-tight plexiglass box. All hormone-treated plants with the
exception of cytokinin-treated plants were incubated for 24 hours. Exposure to cytokinin
was performed by germinating seeds on plates containing 4 uM 6-benzylaminopurine

(BAP) and collecting seedlings after 10 days. Etiolated plants were grown on plates

152

wrapped with aluminum foil for 10 days. Plants were subjected to cold treatment by
transferring 9-day old seedlings to 4° C for 48 hours. Plant were wounded by cutting
mature leaf tissue with scissors. After treatments, all plants were stained for GUS
activity. The results (Figure 26) indicate that cytokinin and, to a lesser extent, the SA
analog BHT appear to up-regulate expression of the reporter, while MeJA may repress its
expression. Auxin treatment resulted in a different spatial pattern of reporter, with
increased staining apparent in the stem, but did not appear to enhance the overall amount
of staining. All other treatments resulted in GUS staining patterns that resembled those
observed in control plants, i.e., variable staining concentrated primarily in the cotyledon
veins and central vein. Data for cold and wounding treatments are not shown, but no
induction of expression was observed. It should be noted that the increased GUS activity
elicited by cytokinin was not variable, with 100% of the individuals in three different
promoter-reporter transgenic lines showing increased reporter activity. Likewise, the
repression of GUS activity by MeJA was also not variable, with all plants showing an

absence of staining.

153

Table 6. Promoter cis-elements of AtF U T I . 1060 nt upstream of the AtF UT 1 ORF
was used to search two plant promoter element databases, PLACE
(http://www.dna.affrc.go.jp/htdocs/PLACE/) and PlantCARE
(http://sphinx.rug.ac.be:8080/PlantCARE/cgi/index.html). Cis elements were
grouped into auxin elements (yellow), light responsiveness (green), transcription
factor (TF) binding sites (orange), pollen expression (pink) or other (blue).

   

 

    

   

  

 
    

 

cis element function number of repetitions
‘ ' 'J'ré; i‘eSp_o"13\/e’~ ' ' "1
A81 mo‘lll‘ auxin SA responsiveness “I
N’I‘BBF'EAWC‘D auxin responsiveness 2
A‘uxRR—oore ‘ auxin responsiveness ‘I
6.. ~ auxin tesaomsieras " ' l
i I l
i
U
:o ’ ' EAIME‘HQ l?
.1?
A 3
mot. UXPI‘CSSICN‘, .
2l
3i
3i
1,
15
:3
J
MUG—front ll
u‘Tz rmn‘ 2
Al'vli; ”53121311“ 3
Jli-IE ' Humor «l
CCC-rum Ll
Emilia/(ls. ‘I

‘.C

'l

1 , , P .1»
l:» r , t

 

154

Table 6, continued.

 

155

 

Auxin (NAA)

 

 

BTH (SA analog) ACC

  

a 1X57
\ l

   
 

, I

. . ('
.. ,
5': < . ‘

      

etiolated

Figure 26. Expression of AtFUT1 in hormone-treated or etiolated promoter-GUS
transgenic plants. Representative images are shown from staining of three different
P2.0 transgenic lines. Seedlings were grown for 10 days on growth plates and then
transferred to incubation media (control) or treated with methyl jasmonate (MeJA, see
methods), 50 M l-napthalene acetic acid (auxin), 50 uM abscisic acid (ABA), 0.03%
BHT, 100 uM gibberellic acid (GA3), 1 mM 1-aminocyclopropane-l-carboxylic acid
(ACC), 2 uM epibrassinolide (epiBR), or cytokinin. All treatments except that with
cytokinin took place for 24 hours. Cytokinin (6-benzylaminopurine, BAP) treatment was
performed by growing plants on standard growth plates supplemented with 4 uM BAP
for 10 days. Etiolated plants were grown on plates wrapped with aluminum foil. Bar =
100 um (etiolated), 1 mm (all others).

156

4.3.6 Production and Analysis of 355::AtF U T I Transgenic Plants

A construct containing cDNA encoding AtF UT 1 regulated by the CaMV 358 strong
constitutive promoter was used to generate transgenic Arabidopsis plants. The presence
of the transgene in hygromycin-resistant plants was confirmed by PCR amplification
using primers that span the intron of AtFUT1, allowing one to distinguish between the
transgene and the endogenous genomic version of the gene (data not shown). RNA was
prepared from T2 transgenic plants and used to examine expression of AIF UT 1 by RT-
PCR (Figure 27) in leaves, which have very low levels of AtF UT 1 expression in wild-
type plants. Multiple lines with elevated expression of AtF UT 1 were identified, and
several were chosen for further analysis. Golgi vesicles were prepared from leaves of
these lines, and activity assays were performed using this protein (Figure 28). The
transgenic lines showed elevated levels of XyG FUTase activity compared to Golgi

vesicles derived from wild-type leaves.

Cell wall preparations were made from these lines and analyzed for XyG Fuc content at
the Complex Carbohydrate Research Center (Figure 29.) Preliminary analyses indicate
that the level of F uc present in XyG oligosaccharides of the over-expressing transgenic
lines is slightly elevated over wild type levels in xyloglucan endoglucanase (XEG)-
extracted fractions, but is not elevated in 4M KOH-extracted fi-actions. There is also an
unexpected increase in the amount of XyG subunits containing acetylated Gal, with
levels of acetylated Gal units being approximately two-fold higher in the over-expressing

transgenic lines as compared to wild type plants.

157

[5 HI 1 kb ladder
Pi11:
\Arr
1-1
1-2
1-3
1.4
1-5
1-10
1-11
1-12
1-13
1-14
1-15
1-16
1-17
1-18
1-21
2-11
2-13

   
 

3-2
3—4
3-10
4-1 1

InexNeeec
'T'T'T'T'I"?
NNNNNN

3-1

 

Figure 27. Expression analysis of AtF UT 1 in 35S::AtF UT 1 transgenic plants by RT-
PCR. RNA was prepared from leaf tissue of wild-type or transgenic lines. One
microgram of RNA was used as template in reverse transcription reactions, and an equal
volume of RT reaction was used for PCR amplification of AIF UT 1. Samples were
analyzed on an agarose gel stained with ethidium bromide.

158

350-1
300 . - - '
250 TM_»“,_ —
zoo-e—Al -
150=4
100

5°“- I i ll.
04 ,

line 1 line 2 line 3 line 4 line 5 line 6

 

 

 

 

 

 

 

 

   

 

 

 

 

total cpmlmicrogram protein

16000“
14000 DEERE..-
12000
10000-w-

3000

6000
4000-l————

2000 “"“I”
o +———-.~—

WT 1-4-3 1-4-4 1-4-5 3-4-4 3-4-5 3-10-1

 

 

 

 

 

 

 

 

 

 

  

total cpmlg fresh weight

 

Figure 28. XyG FUTase activity in Golgi vesicles derived from 35S::AtFUT1
transgenic plants. Golgi vesicles were prepared from leaves of transgenic plants and
used in XyG FUTase activity assays. Assays were conducted at room temperature for
one hour. Duplicate samples were performed. Results are expressed as total cpm/ug
protein (top) or total cpm/ g fiesh weight (bottom). Error bars indicate one standard
deviation.

159

Table 7. Analysis of XyG from 35S::AtFUT1 transgenic plants. XyG was extracted
fiom depectinated cell wall preparations of wild-type and over-expressing transgenic
lines. Depectinated cell wall material was first treated with an XyG-specific
endoglucanase (XEG fraction), then with 1 M KOH (fraction not analyzed further), and
finally with 4 M KOH. XyG polymers fi'om the XEG and 4 M KOH fractions were
digested into oligosaccharides and analyzed by lH-NMR to quantify resonance signals
characteristic of specific XyG side chain structures. Data were normalized by setting the
total amount of xylose residues per unit of oligosaccharide to 3. Data are expressed as
the number of residues or side-chains per oligosaccharide subunit. Fuc, total fucose per
XyG oligosaccharide subunit; L, total galactose per XyG oligosaccharide subunit; L(Ac),
Gal(Ac)-Xyl side-chains; F, F uc-Gal-Xyl side-chains (with no acetate substitutions on
Gal); F(Ac), Fuc-Gal(Ac)-Xyl side-chains; X, Xyl residues; F(Ac)a, Fuc-Gal(Ac)-Xyl
calculated from the area of the acetate methyl signal; F(Ac)b, Fuc-Gal(Ac)-Xyl
calculated from the area of the H6/H6' of acetylated Gal in F. Acetate modifications are
only detectable in the XEG fraction, because acetate groups are hydrolyzed in 4 M KOH.

 

"xé’ assess}; .. .4 g ,,

1;: Sample F116... .. L . . MAC! .2. :. F ...-FlA :11-:..x... £51159}: Fléelfi
wild-type 0.45 0.44 0.08 0.22 0.23 2.03 0.21 0.18
1.4-2 0.46 0.41 0.11 0.11 0.35 2.02 0.33 0.32
1.4-3 0.47 0.4 0.09 0.09 0.38 2.04 0.36 0.36
1-4-4 0.47 0.4 0.08 0.1 0.37 2.06 0.36 0.35
145 0.46 0.4 0.1 0.1 0.36 2.04 0.34 0.35
34.1 0.47 0.41 0.09 0.12 0.35 2.03 0.33 0.33
34.2 0.48 0.39 0.1 0.12 0.36 2.03 0.35 0.32
3.43 0.49 0.4 0.06 0.15 0.34 2.04 0.34 0.34
3.4-4 0.48 0.38 0.04 0.16 0.32 2.1 0.32 0.33
345 0.48 0.4 0.05 0.14 0.34 2.07 0.34 0.34
3.101 0.46 0.41 0.09 0.09 0.37 2.04 0.33 0.34

41116011 11am .. _

-.§_ample...-. "Ft-9.1“"‘trll.-.::’; “1:1”?
wild-type 0. 5 0. 47 0. 52 2. 01
142 0.49 0.45 0.52 2.03
143 0.51 0.44 0.52 2.04
144 0.52 0.46 0.53 2.01
145 0.51 0.45 0.54 2.01
34.1 0.5 0.47 0.52 2
341-2 0.5 0.45 0.51 2.04
343 0.49 0.43 0.53 2.04
3-44 0.53 0.45 0.53 2.02
34.5 0.52 0.46 0.54 2
3.101 0.51 0.46 0.53 2.01

160

4.4 Discussion

Cell wall biosynthesis is a critical process during plant growth and tissue differentiation.
There are, however, limited systems at present that can be used to study cell wall
biosynthetic processes at multiple levels. xyG FUTase offers a means of investigating
one cell wall biosynthetic event at the level of gene regulation, enzyme activity, and

product analysis.

It is important to note that various regions of the mature Arabidopsis plant will be
undergoing different growth rates at a single point in time. Rosette leaves, for example,
go through a period of expansion and growth early during the life of the plant, but cease
increasing in size once the plant reaches vegetative maturity and begins to produce a
reproductive inflorescence stem. The stem itself grows by division of cells in the shoot
apical meristem at the top of the inflorescence stem. These cells elongate and then
differentiate. As a result, growth may still be occurring at the top of the stem at the same
time that cells at the bottom of the stem have ceased to increase in size. A similar
gradient of growth occurs in the roots. Thus, there are inherent differences in growth
kinetics among various tissues of mature Arabidopsis, and this may be reflected by
variations in the rates of cell wall biosynthesis occuning in these tissues. It is also
possible, however, that differences exist in the cell wall composition and structure of
various tissue and cell types. Such differences have been described in different plant
species and are reviewed in Chapter 1. Therefore, it must be determined whether any

differences in the regulation of a cell wall biosynthetic process between an assortment of

161

tissue types are due to variations in growth kinetics or variations in cell wall polymer

structure.

Levels of AtFUT1 gene expression and XyG FUTase activity were assessed in seven
different Arabidopsis tissues. Distinct differences were observed in gene expression and
enzyme activity among the samples. In general, the upper region of the stem showed high
gene expression and activity levels, while the lower region of the stem showed less
activity. Fully expanded rosette leaves showed a near-absence of gene expression and
enzyme activity. Flowers showed high enzyme activity, but RT-PCR analysis of the
same sample showed less gene expression, although analysis of other floral samples by

RT-PCR and hybridization techniques have shown increased gene expression in flowers.

Although the structure of Arabidopsis cell walls has been a subject of previous study
(Zablackis et al., 1995), only leaves were used for analysis in their experiments. No
systematic comparison of XyG structure from cell walls of different tissues in
Arabidopsis has yet been published. Although the structure of XyG has been determined
for a number of species, it is not known whether there may be subtle variations in the
structure of this polymer throughout different regions of the plant. Some small
differences have been detected, for example, in the structure of XyG isolated from
growing and non-growing regions of etiolated pea epicotyls (Pauly et al., 2001b).
Analysis of XyG fucosylation patterns in oligosaccharides derived from various
Arabidopsis tissues indicated no major differences, with the amount of Fuc present per

unit of XyG oligosaccharide remaining at a fairly constant ratio of around 0.5. Although

162

one rosette leaf sample indicated about half as much Fuc present per XyG unit compared
to other tissues, subsequent analysis of independent biological replicates and a leaf
developmental series did not show significant differences. It is assumed that biological
variability was the source of the initial lower ratio. Thus, the Fuc content of XyG appears

to remain fairly constant throughout different tissues of Arabidopsis.

One disadvantage of RT-PCR and hybridization-based expression analysis techniques is
their limited ability to discern cell-type or tissue-specific patterns of gene regulation.
Two methods commonly used to overcome this limitation are in situ hybridization and
promoter-reporter analysis. Because of the inherently low expression levels of AtFUT1
and the fact that it is a member of a family containing nine other AtF UT 1 -like genes,

promoter-reporter analysis was used to determine spatial patterns of gene expression.

In order to make appropriate promoter-reporter constructs, the delineations of the
AtFUT1 promoter region had to be considered. At the time that the constructs were
planned, the nearest predicted ORF upstream of AtF UT 1 was several kilobases away,
which did not allow the delineations of the promoter to be determined with any degree of
certainty. Between the time that the constructs were generated and the time at which the
promoter-reporter transgenic plants were analyzed, however, the genomic region
upstream of AtF UT 1 was re-annotated by the Arabidopsis Genome Initiative (Figure 30).
There are two predicted ORFs upstream of the AtFUT1 ORF, the closest of which is
currently annotated as a 458 nt-long hypothetical protein (AT2G03230). No ESTs in

exist in public databases that correspond to the hypothetical protein, and there is no

163

evidence that the gene encoding it is transcribed. The predicted hypothetical protein
would be 159 aa long protein and approximately 20 kD in size. The next nearest
upstream predicted ORF is currently annotated as an unknown protein (AT2G03240).
An ORF prediction for the unknown protein did not exist at the time that the promoter-
reporter constructs were made, probably because of the complex intron-exon structure of
this gene; there are 10 predicted exons and 9 introns over a span of 3924 nt in the
unknown protein gene. A BLAST search using the predicted cDNA for this gene returns
a partial mRN A sequence for a transcript detected during early response to nematode
infection with Z-score of 119 and E value of le '23, indicating that this gene may encode
a nematode response protein and is probably transcribed. The end of the nematode
response ORF is 1060 bp upstream from start of the AtFUT1 ORF. Thus, it is expected
that the AtF UT 1 promoter region may be 1060 bp long or less, unless regulatory regions
exist within the coding region of the nematode response gene. For this reason, only the
transgenic plants containing the 2.0 kb AtF UT 1 promoter fragment were used for

extensive analysis of gene expression.

164

Nematode infection

protein

10 exons, 9 introns At UTl
Probably transcribed 1060

l
_r—J —

Hypothetical protein
No evidence of
transcription

Figure 30. Structure of genomic region upstream of AtFUT1. Two ORFs lie
upstream of AtF UT 1 on Chromosome 2. Directly upstream is a small predicted ORF
(hypothetical protein) that is 458 nt long, shown as a white box. No ESTs corresponding
to this hypothetical protein exist in public databases. Further upstream is an ORF
encoding a gene that is induced upon infection by nematodes. This gene has a complex
intron-exon structure, with 10 exons and nine introns over a total of 3924 nt. There is
evidence that the nematode response gene is transcribed. The end of the nematode
response gene ORF is 1060 nt upstream of the beginning of the AtF UT 1 ORF, which
itself is 2043 nt long. This figure is drawn to scale.

165

Systematic analysis of transgenic promoter-reporter plants indicates that AtF UT 1 gene
expression is regulated. The observations generally agree with gene expression data
obtained by RT—PCR and hybridization techniques, with high reporter activity occurring
in the top of the inflorescence stem and lower activity occurring in fully expanded rosette
leaves (limited mainly to a small portion of the midvein). One unexpected finding was
the extent of expression correlated with vascular tissue in cotyledons, regions of the true
leaf vasculature, and the central vascular cylinder of the hypocotyl. Fully differentiated
vascular tissue is generally thought to be comprised of secondary cell walls, which are
not rich in XyG. It is possible that at the early stages of development, the vascular tissue
showing increased reporter gene expression had not yet fully progressed into secondary
cell wall synthesis. Alternatively, it is possible that the vascular patterns observed were
artifacts of the promoter-reporter technique, although ferricyanide and ferrocyanide were
included in the staining buffer in an attempt to reduce dispersal of the stain (for example,
through the vascular system). However, the overall expression profile does correlate with
that observed using other methods, and it is possible that the patterns observed in the

transgenic plants are indicative of spatial gene expression regulation in vivo.

One important observation is that the expression pattern of AtF UT 1 in the inflorescence
stem does correlate very well to the elongation rate in regions of the stem. In other
words, high gene expression is observed in the top of the stem, which shows over an
eightfold increase in elongation rate compared to the bottom of the stem. The increase in

expression in'this region of the stem was confirmed by three different methods (RT-PCR,

166

northern analysis, and promoter-GUS staining). A similar profile was observed for XyG
FUTase enzyme activity levels in these regions. This provides strong evidence that XyG
fucosylation, a primary cell wall biosynthetic process, is regulated in a manner that is

positively correlated with growth in the inflorescence stem.

A similar correlation could not be detected in the other major region of elongation growth
in the plant, the root. However, different methods did indicate some instances in which
AtF UT 1 expression and XyG FUTase activity were elevated in roots. It is possible that
the level of expression in roots of promoter-reporter plants was too low to be detectable
by this method. It was not possible to dissect Arabidopsis roots, which are thinner than
the width of a human hair, very fragile, and prone to entanglement, into elongating and
differentiated regions. The variable results in expression and activity levels may be
indicative of the ratio of elongating versus differentiated tissue in various biological
samples. It should be noted that AtFUT1 is thought to be the sole gene responsible for
fucosylation of XyG, as a T-DNA knockout (Sarria et al., 2001, in press) and the mur2
mutant affected in AtFUT1 (V anzin et al., 2001, submitted) lack Fuc on XyG in all

tissues examined, including (in the case of Vanzin et al.) the roots.

XyG FUTase activity profiles showed high activity levels in flowers, but various gene
expression techniques did not show strong up-regulation of AtFUT1 in floral tissue. One
possible explanation is that the high activity levels are due not to the presence of floral
tissue itself, but the presence of fi'agments of inflorescence stern collected during tissue

harvesting. If, as the promoter-reporter analysis indicates, high gene expression occurs in

167

the uppermost region of the inflorescence stem, the amount of stem present in a given
“floral” tissue sample could have a large effect on the amount of XyG FUTase activity
and AtFUT1 gene expression. Thus, variation between different samples could be due to
differences in harvesting technique and exact sites of tissue dissection. With this in mind,
it is interesting to note the difference in AtF UT 1 expression between “flowers” and
“floral buds” as detected by northern blot (Figure 6). In this case, flowers were defined
as individual fully opened flowers cut away from the inflorescence stem (although some
amount of floret stem is present, holding the flower together at the base.) Flower buds
were defined as the unopened flowers at the top of the inflorescence. Because of the
manner in which these samples were harvested, the “floral buds” would contain more
inflorescence stem than the “flowers”. Greater AtFUT1 expression is observed in the
floral bud sample versus the flower sample (Figure 6). An alternative explanation for the
variability in gene expression of flower samples would be that expression could be
regulated in a pollen-specific manner, and the extent of detection would be due to how
much pollen was present after harvesting, handling, and experimental manipulation.
Pollen-specific expression was observed in promoter-reporter analysis of a gene encoding
an isoform of GDP-D-mannose 4,6 dehydratase, which synthesizes GDP-Fuc (Bonin et
al., in press.) However, no pollen staining has been observed in promoter-reporter

analysis of AtF UT 1 gene expression.
Promoter-reporter analysis also allows gene regulation to be assessed easily after

exposure to various treatments, such as hormones or different growth treatments.

Incubation of promoter-reporter plants with a series of different growth hormones at

168

levels used in gene regulation experiments reported in the literature indicated that
AtF UT 1 may be up-regulated in expression by the cytokinin BAP and, to a lesser extent,
by the SA analog BHT. AtF UT 1 expression may also be repressed upon exposure to the
volatile signaling molecule MeJA. Cytokinin is a growth hormone involved in cell
division and developmental regulation, while SA is thought to be a key signaling
molecule for certain plant defense responses. MeJA also participates in defense
responses, but seems to play a major role in defense against herbivores rather than fimgal,
bacterial or viral pathogens. The biological significance of these observations is not
clear, but further testing and verification by independent techniques would assist in
assessing the regulation of AtF UT 1 in response to cytokinins, SA and MeJA. The
expression of AtF UT 1 in various mutants defective in the ability to produce cytokinin,

SA and Me] A would also enable further characterization of these results.

In addition to the AtF UT 1 gene expression data described here, I also assessed other
publicly available data. During the course of this thesis research, the Arabidopsis
Functional Genomics Consortium performed microarray experiments for individual
researchers wishing to investigate global gene expression under conditions of interest or
in various genetic backgrounds. A cDNA encoding AtF UT 1 was present on the arrays
used for these experiments, and all experimental data from these efforts are made
available to the public after an initial waiting period. These data sets were analyzed for
AtF UT 1 expression levels. Table 7 indicates several experiments in which a reciprocal
set of microarrays (for example, flowers versus leaves and leaves versus flowers)

indicated differential expression of AtFUTI between the two samples studied. The

169

criteria used were that one slide showed a normalized green/red signal for AtFUT1
greater than Mo on one slide and the slide for the inverse labeling experiment showed a
low, if not exactly reciprocal to the first, normalized green/red signal. In the first study,
researchers compared gene expression in flower versus leaf tissues (Jeff Landgraf,
personal communication). AtF UT 1 expression results were consistent to other expression
data shown above (that is, AtF UT 1 expression was higher in flowers than leaves). In the
second study, gene expression profiles were compared between the ABA-insensitive
mutant abiI to wild type Ler plants. However, these experiments have been repeated
using Affymetrix DNA array chips, and the trend noted in the microarray experiments did
not hold (Julian Schroeder and Majid Ghassemian, personal communication). Thus, it
must be noted that trends observed in microarray experiments should be confirmed by
independent techniques, such as hybridization analysis or RT-PCR, before conclusions
can be drawn with certainty. In a third experiment, researchers attempted to identify
genes that are targets of the homeobox gene knotted] (an) using transgenic plants
expressing kn] under the control of the glucocorticoid receptor and induced with
dexamethasone to wild-type plants sprayed with dexamethosone (Sarah Hake and Naomi
Ori, personal communication). KN CITED] is a transcription factor that is involved in
leaf development and meristem identity, and was the first homeodomain-containing
protein identified in plants (V ollbrecht et al., 1991). AtF UT 1 expression shows a
threefold increase in normalized G/R ratio in the induced kn] transgenic plants. These
experiments are also being repeated using the Affymetrix system, and expression of
AtF UT 1 will be assessed using this technique. Gene expression profile patterns were

examined in the fourth study in an attempt to identify genes necessary for adaptation to

170

low availability of chlorophyll using a leaky magnesium chelatase mutant, cchI. This
mutant becomes pale green when transferred to moderate light (200 mol photons 111'1
sec”), whereas wild type plants remain green under these conditions (Judy Brusslan,
personal communication). Genes expressed in wild type and cchI plants under these
light conditions were compared, and AtFUT1 is expressed more highly in wild-type
plants than in the cchI mutant. It is interesting to note that the promoter region of
AtF UT 1 does contain a number of Cis-acting elements previously characterized in light-
regulated genes (Table 6.) In the sixth experiment, investigators compared gene
expression profiles in the enhanced disease resistance 1 (edrI) mutant to those of wild-
type plants (Roger Innes and Dingzhong Tang, personal communication). The edrI
mutant shows enhanced resistance to the fungal pathogen that causes powdery mildew.
Germination of asexual spores is inhibited in the mutants, but initial stages of infection
are not affected (Frye and Innes, 1998). AtF UT 1 expression is higher in edrI plants than
in wild type. It is possible that AtF UT 1 expression is altered during response to pathogen
infection, though this possibility needs to be investigated by other methods. This might
be explained, for example, if pathogen defense includes alteration of cell wall
biosynthetic processes. Cell wall structural alterations, such as cross-linking phenolic
compounds or localized lignification, have been described previously (Hardham and

Mitchell, 1998; Wojtaszek, 2000).

171

Table 8. Summary of data on AtFUT1 expression from public database of
microarray experiments. Only data with a green channel: red channel normalized (G/R
normalized) ratio of >20 in one slide and a low G/R signal in an inverse experiment are
shown. *, independent Affymetrix chip data are available, and they do not support the

trend suggested by the microarray experiments.

TM “1:24-er ( In" .1 *2»: W 329-12".- 1 'zzn-tvrr,‘-1)'.;m 153‘ I'M. 1

Flowers vs. Leaves
Leaves vs. Flowers

abi1 slide 1*
abi1 slide 2*

downstream genes of RM slide 1 (induced kn1_GR green.
induced control red)

downstream genes of RM slide 2 (induced control green,
induced kn1_GR red)

genes in chlorophyll starvation slide 1 (WT green, och1 red)
genes in chlorophyll starvation slide 2 (cch1 green. WT red)

Pathogen Response 2 - reverse
Pathogen Response 1

edr1 Mutant 2 - reverse (6dr1 green, WT red)
edr1 Mutant1 (strong spatial bias) (WT green. edr1 red)

172

' “ ,-.,. :2 fifi'W'flML Herman.

ggrns y; 5.4 "it -'~ “‘.1’.\‘*"':‘- -“ g --
1...“ m («W.mzzi-uist r. .Expadmsnt .-.-. - w.~.~....Gl.Bm

 
 

”2.169

0.346

6.897
0.674

3.174

0.668

2.347
0.265

2.146
0.654

2.05
0.584

Ectopic expression is Often used to study the effects of gene manipulation in planta.
Transgenic plants expressing AtFUT1 under control of a strong constitutive promoter
show increased gene expression and XyG FUTase activity compared to wild type
controls. Preliminary analysis of XyG structure indicates that Fuc levels are slightly
elevated in the transgenic plants. A more surprising observation is that the level of
acetylated Gal units (for example, side chains containing Fuc-Gal-Xyl with an acetate
group present on Gal) is approximately two-fold higher in the over—expressing plants.
Acetylation of Gal has been observed at the 0-6, O-4, and O-3 positions in sycamore
XyG (York et al., 1988). Mono- or di-acetylation may occur, and in sycamore the
predominant form was mono-acetylation at the 0-6 position (York et al., 1988).
Although it is known that acetyl Co-A is the donor substrate for acetylation (Pauly and
Scheller, 2000), the enzyme responsible for the acetylation of XyG or the gene that
encodes it has not been identified. The biological role of acetylated Gal units in XyG is
not known, although it has been speculated to affect hindrance of polymer breakdown
(Pauly and Scheller, 2000). Interestingly, a recent structural analysis of XyG derived
from the mar] mutant indicated that XyG derived from leaf tissue is less acetylated in the
mutant than in wild type (Pauly et al., 2001a). The marl mutant is deficient in an
isoform of GDP-D-mannose-4,6-dehydratase and thus partially lacks the ability to
synthesize GDP-Fuc. Analysis of mur2, a mutant affected in AtFUT1, or a mutant with a
T-DNA insertion in AtFUT1, also indicates reduced levels of Gal acetylation (W. York,
personal communication.) When preliminary data on XyG structure from 35S::AtFUT1
plants are compared to the mar] , mur2, and AtF UT 1 knockout studies, it appears that

there may be a correlation between the presence of F no on XyG and the extent of Gal

173

acetylation. One explanation for this could be that the presence of F uc permits a
particular structural conformation to be assumed that facilitates acetylation.
Alternatively, there could be a process-based relationship between fucosylation and
acetylation. For example, if the residence time of the polymer in the Golgi determined
the exposure of the polymer to the acetyltransferase, and if over-expression of AtF UT 1
caused an increase in this residence time, an increased level of Gal acetylation could
result. Additional studies would have to be conducted in order to evaluate these

possibilities, however.

No differences in phenotype are observed for these plants under normal growth
conditions. However, it is possible that strong constitutive expression of AtF UT 1 results
in phenotypes only apparent under certain conditions, such as exposure to pathogens,
herbivores, or abiotic stresses. Further characterization under a series of conditions may
allow identification of situations in which alteration of AtF UT 1 expression results in a

demonstrable consequence for the plant.

In conclusion, AtFUT1 expression and XyG FUTase activity were found to be positively
correlated with growth in aboveground regions of Arabidopsis. Determination of the
gene expression profile of AtF UT 1 should assist in the evaluation of other primary cell
wall biosynthetic gene candidates, as has indeed been the case within our research group
(T. Wagner, C. Wilkerson, S. Meyer, unpublished results.) Various hormone treatments
may affect the expression of AtFUT1, possibly allowing the plant to regulate this cell .

wall biosynthetic process in response to these signals. Manipulation of the expression of

174

AtF UT 1 using the strong constitutive 358 promoter results in increased XyG FUTase
activity and altered XyG acetylation patterns. It is possible that these changes may have
phenotypic consequences under certain conditions, but no phenotypic effects were
observed when plants were grown in growth chambers. Overall, this work should lead to
a more precise understanding of xyloglucan fucosylation as a model cell wall

biosynthetic process.

References

Bechtold, N. and Pelletier, G. (1998). In planta Agrobacterium-mediated transformation
of adult Arabidopsis thaliana plants by vacuum infiltration. Methods Mol Biol 82.

Brummell, D. A., Camirand, A. and Maclachlan, G. A. (1990). Differential
distribution of xyloglucan transferases in pea Golgi dictyosomes and secretory vesicles.
Journal of Cell Science 96, 705-710.

Cosgrove, D. J. (2000). Expansive growth of plant cell walls. Plant Physiology and
Biochemistry 38, 109-124.

Edwards, M. E., Dickson, C. A., Chengappa, S., Sidebottom, C., Gidley, M. J. and
Reid, J. S. (1999). Molecular characterisation of a membrane-bound
galactosyltransferase of plant cell wall matrix polysaccharide biosynthesis. Plant Journal
19, 691-7.

Frye, C. A. and Innes, R. W. (1998). An Arabidopsis mutant with enhanced resistance
to powdery mildew. Plant Cell 10, 947-956.

Gibeaut, D. M., Hulett, J., Cramer, C. R. and Seemann, J. R. (1997). Maximal

biomas of Arabidopsis thaliana using a simple, low-maintenance hydroponic method and
favorable environmental conditions. Plant Physiology 115, 317-319.

Gidekel, M., Jimenez, B. and Herrera-Estrella, L. (1996). The first intron of the
Arabidopsis thaliana gene coding for elongation factor 1 beta contains an enhancer-like
element. Gene 170, 201-206.

McConkey, M. E., Gershenzon, J. and Croteau, R. B. (2000). Developmental

regulation of monoterpene biosynthesis in the glandular trichomes of peppermint. Plant
Physiol 122.

175

Moehs, C. P., Tian, L., Osteryoung, K. W. and Dellapenna, D. (2001). Analysis of
carotenoid biosynthetic gene expression during marigold petal development. Plant
Molecular Biology 45.

Munoz, P., Norambuena, L. and Orellana, A. (1996). Evidence for a UDP-glucose
transporter in Golgi apparatus- derived vesicles from pea and its possible role in
polysaccharide biosynthesis. Plant Physiol. 112, 1585-1594.

Pauly, M., Qin, Q., Greene, H., Albersheim, P., Darvill, A. and York, W. S. (2001).
Changes in the structure of xyloglucan during cell elongation. Planta 212, 842-850.

Vollbrecht, E., Veit, B., Sinha, N. and Hake, S. (1991). The developmental gene
Knotted-1 is a member of a maize homeobox gene family. Nature 350, 241-243.

Wulff, C., Norambuena, L. and Orellana, A. (2000). GDP-fucose uptake into the Golgi
apparatus during xyloglucan biosynthesis requires the activity of a transporter-like
protein other than the UDP-glucose transporter. Plant Physiology 122, 867-877.

Zablackis, E., Huang, J., Muller, B., Darvill, A. G. and Albersheim, P. (1995).

Characterization of the cell wall polysaccharides of Arabidopsis thaliana leaves. Plant
Physiology 107, 1129-1138.

176

CHAPTER FIVE
Conclusions and Future Directions

5.] Summary of Research Significance

At the beginning of this thesis, no cell wall biosynthetic genes had been identified at the
molecular level with certainty. Although there were hints about conserved sequence
motifs and plant genes that might be important for cellulose synthesis in 1996 (Pear et al.,
1996), it was not until 1998 that genetic evidence was published confirming the necessity
of the CesA isoforms for manufacturing cellulose (Arioli et al., 1998). Although there
were a number of FUTases identified in various animal systems (for review, see Field
and Wainwright, 1995), no plant sequences in public databases showed any similarity to

the mammalian proteins.

And yet, a great deal had been determined regarding the structure of the plant cell wall
through several decades of investigation, and evidence was beginning to suggest that cell
wall biosynthetic glycosyltransferases could be identified. An assay had been developed
to detect XyG FUTase activity (Hanna et al., 1991) and preliminary results from this
research group indicated that purification of this enzyme was indeed feasible (data from
A. DeRocher, M. Bar-Peled, N. V. Raikhel, K. Keegstra.) During the course of this
thesis research, XyG FUTase has been purified by a team effort involving several
researchers in the Keegstra and Raikhel laboratories, the homologous gene encoding this
enzyme has been cloned in Arabidopsis (R. Perrin) and later in pea (W. Zeng), and the
regulation of this gene and enzyme system have been studied (R. Perrin). Other research

groups have identified additional cell wall biosynthetic genes and enzymes during the

177

past five years. These include galactomannan galactosyltransferase (Edwards et al.,
1999), the cellulose synthase catalytic subunit (Arioli et al., 1998), and cellulose

synthase-like genes (Richmond and Somerville, 2001).

This thesis research seeks to contribute some groundwork to the field of plant cell wall
biosynthesis, allowing the information on the regulation of AIF UT 1 expression and XyG
FUTase enzyme activity to be used to evaluate other candidate genes for cell wall
synthesis. The present work has also confirmed that the process of XyG fucosylation is
correlated with elongation growth in aboveground regions of Arabidopsis, an idea that
has long been presumed and is now confirmed at the molecular level. This indicates that
XyG fucosylation may be studied as a model primary cell wall biosynthetic event. In
addition, it is hoped that this research will begin to provide some insight into the
biological significance of XyG fucosylation. Manipulation of the expression of AIF UT 1
by strong constitutive expression causes an increase in XyG FUTase activity, alteration
of XyG structure in terms of Gal acetylation (though not, interestingly, in terms of
fucosylation), and could have phenotypic consequences under specific circumstances

although no major differences are observed under normal growth conditions.

5.2 Future Directions

There are a number of additional experimental lines that may be followed to extend this
work, a few of which will be described. One area of study would utilize the AtF UT 1
over-expressing transgenic lines described in this thesis, as well as a T-DNA insertional

AtF UT 1 mutant completely lacking F uc on XyG, to do more physiological studies

178

comparing the performance of these plants to wild type controls under various conditions.
These studies would seek to determine the function of F uc on XyG. Fuc modification of
XyG is a well-conserved structure found throughout all dicot and nongraminaceous
monocot plant species (although Solenaceous species replace F uc modification with other
structures such as Ara). The relative regularity of Fuc modification would seem to
indicate that there is some sort of selective pressure for this structure in nature. Although
both T-DNA knockout and over-expressing plants do not show altered phenotypes under
normal growth conditions in greenhouses or growth chambers, it is entirely possible that
the presence of F no on XyG is advantageous under certain conditions such as challenge
with pathogens or herbivores. Alternatively, subtle differences in growth or
developmental rates might not be discemable under nonsystematic observation, but could
be identified by careful growth kinetic characterization. Such analysis is being done on
the T-DNA knockout by Paradigm Genetics and could also be performed using the over-

expressing lines as well.

Another line of research would involve confirmation of the cytokinin and SA inducibility
and MeJA repression of AIF UT 1 by an independent technique such as RT-PCR This
could be followed by whole plant-level experiments to determine the biological
significance of these observations. At present, there are indications that these signaling
molecules may regulate the expression of AtFUT1, but the consequences of this

regulation have not yet been determined.

179

 

Although microarray technology has been used during the past five years to investigate
global gene expression patterns, Affymetrix arrays are beginning to emerge as the leading
technology in these types of experiments. Theoretically, it would be possible to use
Affymetrix arrays and the complete Arabidopsis genome sequence information to
continue to develop a more thorough expression profiling data set for AtFUT1, which
could be used to identify other co-regulated genes. By comparing promoter sequences of
such co-regulated genes, it would be possible to identify Cis-acting sequences putatively
required for coordinately regulated expression. This could allow another means of
identifying and assessing other candidate genes, if they have such regulatory elements in
their promoters, and could also help to identify trans-acting factors involved in regulating

the expression of cell wall biosynthetic genes.

Finally, existing information about AtF UT 1 and XyG F UTase will continue to assist
identification and evaluation of other genes involved in synthesis of XyG and other cell
wall polymers. lnforrnation on protein structural features and motifs have allowed
identification of gene candidates (C. Wilkerson), while expression profiling data is
currently providing another means of evaluating several gene candidates of unknown
function that may be involved in cell wall biosynthesis (T. Wagner, S. Meyers, C.
Wilkerson.) This should provide a complement to other approaches to identify cell wall
biosynthetic genes, such as biochemical characterization of candidates expressed in
heterologous systems (A. F aik, T. Wagner), suppression of candidate genes or gene
families by RNAi or insertional mutagenesis followed by phenotypic characterization

(W. Zeng, T. Wagner, S. Hazen, J. Scott-Craig), or identification of quantitative trait loci

180

associated with cell wall structural features (S. Hazen). Although cell wall biosynthesis
research is and will continue to be a challenging area, there are now many different
approaches available. It is almost certain that a large number of cell wall biosynthetic

genes will be identified and characterized within the next five to ten years.

References

Arioli, T., Peng, L., Betzner, A. 8., Burn, J., Wittke, W., Herth, W., Camilleri, C.,
Hofte, H., Plazinski, J., Birch, R. et al. (1998). Molecular analysis of cellulose
biosynthesis in Arabidopsis. Science 279, 717-720.

Edwards, M. E., Dickson, C. A., Chengappa, S., Sidebottom, C., Gidley, M. J. and
Reid, J. S. (1999). Molecular characterisation of a membrane-bound

galactosyltransferase of plant cell wall matrix polysaccharide biosynthesis. Plant Journal
19, 691 -7.

Field, M. C. and Wainwright, L. J. (1995). Molecular cloning of eukaryotic
glycoprotein and glycolipid glycosyltransferases: a survey. Glycobiology 5, 463 -72.

Hanna, R., Brummell, D. A., Camirand, A., Hensel, A., Russell, E. F. and
Maclachlan, G. A. (1991). Solubilization and properties of GDP-fucose: xyloglucan 1,2-

alpha-L-fucosyltransferase from pea epicotyl membranes. Arch Biochem Biophys 290, 7 -
13.

Pear, J. R., Kawagoe, Y., Schreckengost, W. E., Delmer, D. P. and Stalker, D. M.
(1996). Higher plants contain homologs of the bacterial celA genes encoding the catalytic
subunit of cellulose synthase. Proc.Natl.Acad. Sci. USA 93, 12637-12642.

Richmond, T. A. and Somerville, C. (2001). Integrative approaches to determining Csl
function. Plant Molecular Biology 47, 131-143.

181