hu-

 

 

'HE-fiis

5707 $074

This is to certify that the
dissertation entitled

XYLOGLUCAN BIOSYNTHESIS: IDENTIFICATION AND
CHARACTERIZATION OF FUCOSYLTRANSFERASE AND

CELLULOSE SYNTHASE-LIKE GENES

presented by

WEIQING ZENG

has been accepted towards fulfillment
of the requirements for the

degree In Cell and Molecular Biology

 

 

PhD/(WM Keg/$2

Major Professor’ s/Signature

[flu/L 22 2 (.3091
U I I

Date

MSU is an Affirmative Action/Equal Opportunity Institution

— -.—--.-.n--c—.-.-on._ —

 

LIBRARY
Michigan State
University

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 C:/CIFIC/DateDue.p65-p.15

XYLOGLUCAN BIOSYNTHESIS: IDENTIFICATION AND CHARACTERIZATION
OF FUCOSYLTRANSFERASE AND CELLULOSE SYNTHASE-LIKE GENES

By

Weiqing Zeng

A DISSERTATION
Submitted to
Michigan State University

In partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Cell and Molecular Biology Program

2004

ABSTRACT

XYLOGLUCAN BIOSYNTHESIS: IDENTIFICATION AND CHARACTERIZATION
OF FUCOSYLTRANSFERASE AND CELLULOSE SYNTHASE-LIKE GENES

By

Weiqing Zeng

Plant cell walls provide strength and shape to plant cells, and play a critical role in
intra- and intercellular communications during plant growth and development.
Xyloglucan is the major hemicellulose in the primary wall of dicots and non-
graminaceous monocots. XyG is composed of a B-(1,4)-glucan backbone that is usually
decorated with a-(1,6)-xylose, B-(l ,2)-galactose and a-(1,2)-fucose residues. Until very
recently, none of the genes encoding xyloglucan synthetic enzymes had been identified. I
am therefore interested in identification of these genes and in exploration of their
expression patterns during plant growth and development.

Utilizing sequence information obtained from a partially purified xyloglucan
fucosyltransferase enzyme from pea epicotyls, the full-length cDNAs encoding
Arabidopsis and pea XyG fucosyltransferase, AtFutl and PsFutl, were identified. The
biochemical function of AtFUTl was confirmed by the immunoprecipitation of XyG
fucosyltransferase activities from Arabidopsis microsomes using antibodies against E.
col-expressed AtFUTl proteins and, by the presence of XyG-specific fucosyltransferase
activity in COS7 cells expressing the AtFutl gene. The identity of the PsFUTl was
confirmed through sequence comparisons with the purified enzyme, its sequence

similarity with AtFUTl, and its expression in COS7 cells. In young pea plants, the

expression level of PsFutl was found to be highly correlated with the intemode
elongation rate of etiolated pea seedlings.

Cellulose synthase-like proteins of Arabidopsis show 7% to 46% identity with
cellulose synthases and, therefore, have been hypothesized to be involved in the
biosynthesis of wall matrix polysaccharides such as XyG. Expression profiling of AtCsl
genes was performed by examining EST libraries, MPSS libraries and microarray
databases, but did not reveal any strong candidate genes for encoding the xyloglucan
backbone glucan synthase.

Among all AtCsl genes, those of the D subfamily have the highest level of
identity to AtCesA genes. To distinguish whether AtCle proteins might be isoforms of
AtCesA proteins or enzymes involved in the biosynthesis of matrix polysaccharides, a
polyclonal antibody against an AtCle2 peptide was generated and utilized to locate the
proteins at Golgi membranes using cellular membrane fractionation techniques such as
sucrose density gradient and two-phase partitioning. Therefore AtCle genes are likely to
encode enzymes involved in the matrix polysaccharide biosynthesis instead of CesA
proteins.

In another attempt to identify functions for AtCle genes, transgenic plants
containing RNAi constructs targeting AtCle transcripts were generated. These plants
showed deformed cotyledons, were dwarfed as adult plants, and had shortened roots with
more branches. However, the transcript suppression of AtCle genes was only clearly
seen in cotyledons, and no clear alteration in cell wall sugar composition was detected.
Therefore, the exact biochemical functions of AtCle proteins still could not be

determined.

ACKNOWLEDGMENTS

First and the most important, I would like to thank my thesis supervisor and
mentor Dr. Ken Keegstra, for I have learned so much from his broad scientific knowledge,
his logical scientific thinking and his personality. Ken has always been supportive and
willing to help on trivial questions coming from everyday’s bench work, on the design of
experiments and projects, as well as on career development. I have therefore gained
valuable experiences on how to develop and tackle scientific questions, how to think and
communicate in a more scientific and more efficient way. Ken has been a role model for
me not only as a successful and outstanding scientist, but also yet as a parent, a grand
parent tending family matters.

I would also like to thank other members of my graduate committee, Dr.
Christoph Benning, Dr. Hans Kende, Dr. Natasha Raikhel, Dr. Melvin Schindler, and Dr.
Jonathan Walton. Their continuous support and supervision have helped tremendously to
keep my thesis career alive and to reach this final stage. I also have learned a great deal
from them about how to survive in a scientific career.

Members of the Keegstra lab have made my life in the last five years the most
precious experience for me. Almost all of my techniques learned here were from these
fi'iendly people. They have been very supportive in both scientific matters and in daily
life. Specially, Dr. Sigrun Reumann has showed me how to be organized and manage
multiple gigantic experiments, at the same time. Dr. Mitsuru Akita, Dr. Kentaro Inoue,
and Dr. Teruko Konishi have been very generous sharing their expertise on protein and

enzyme biochemistry. Dr. Lynda Fitzpatrick has always been a fun and caring colleague

iv

and has made my lab life so delightful. Other current and past members including Dr.
Aaron Liepman, Dr. John Frohlich, Dr. Ahmed Faik, Dr. Robyn Perrin, Dr. Dianne
Jackson Constan, Dr. Curtis Wilkerson, Dr. Tanya Wagner, Sybil Myers, Linda Denhorf
all have made my life in the lab much easier and full of fun with their assistance on my
work and their companionship in the lab.

I would also have to thank the entire Plant Research Laboratory at Michigan
State University. The people here have made PRL a scientifically challenging place, and
as yet, have been constantly supportive and encouraging. Tremendous amount of help
have come from the supporting staffs that have made my life so much easier over the
years. Especially I want to thank Dr. Tony Sanderfoot and Dr. Jianping Yu for their
constant help on my daily bench work as well as experiment designs. I am also grateful
for my home department, the Cell and Molecular Biology Program, and its director Dr.
Susan Conrad, its secretary Angie Zell, for taking care of many of trivial and important
matters critical to a Ph.D student’s maturation.

Finally I would like to thank my family for their understanding and support,
especially my wife Qiong Wang for her patience, encouragement and her emotional
support. Qiong’s taking over of the entire house work has made the accomplishment of

this thesis feasible.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... ix
LIST OF FIGURES ................................................................................... x
ABBREVIATIONS ................................................................................ xiii
CHAPTER]: The Plant Cell Wall, its Structure, Function, and Biosynthesis ............... 1
Plant cell wall structure and function ..................................................... 2
Plant cell wall functions ........................................................... 2
Plant cell wall structures .......................................................... 3
Xyloglucan structure and function ....................................................... 10
Xyloglucan structures ............................................................ 10
Structural roles of xyloglucan ................................................... 12
Xyloglucan’s role during cell expansion ....................................... 13
Functions of xyloglucan fragments ............................................. l4
Xyloglucan as a reserve carbohydrate .......................................... 15
Identification of genes involved in wall biosynthesis ................................. l7
Genes involved in xyloglucan biosynthesis .................................... l7
Genes involved in the biosynthesis of other wall polysaccharides ......... 20
Functional studies of AtCsl genes ............................................... 24
Regulation of cell wall metabolism ...................................................... 26
References ................................................................................... 28
CHAPTER 2: Cloning and Expression Studies of Xyloglucan Fucosyltransferase .......... 42
Abstract ...................................................................................... 43
Introduction ................................................................................. 44
Materials and methods ...................................................................... 48
Results ............................................. y ........................................... 58
Isolation of full length cDNA clone encoding AtFUTl ...................... 58
Expression of AtFUTl partial polypeptide in E. coli cells and the
generation of antibodies against it ..................................... 61
Cloning of full length coding region, intron and 5’, 3’ UTRs of PsFutl...67
Sequence analysis of PsFutl ..................................................... 77
Specificity analysis of clone PsSUTR as probe for Northern blotting... . ..79
Intemode elongation curve of etiolated seedlings ............................. 79
Expression of PsFutl in three intemodes of etiolated pea seedlings at 7
days old .................................................................... 82
Expression of PsFutl in the third intemode at different growth stages. . ..84
Discussion ................................................................................... 88
References ................................................................................... 93

vi

CHAPTER 3: Sequence and Expression Pattern Analysis of Arabidopsis Cellulose

Synthase-Like Genes ................................................................................ 97
Preface ....................................................................................... 98
Abstract ..................................................................................... 99
Introduction ................................................................................ 1 00
Materials and methods ................................................................... 105
Results ...................................................................................... 108

Sequence extraction, annotation and comparison analysis ................. 108

Expression patterns of AtCsl genes revealed by EST abundance. . . . . .....124
Expression patterns of AtCsl genes revealed by MPSS database. . . . . .....l27
Expression patterns of AtCsl genes revealed by tissue panel microarrays

from Stanford Microarray Database .................................. 133
Regulated expressions of AtCsl genes revealed by microarray
experiments .................................................................................. 134
Discussion .................................................................................. 138
References ................................................................................. 141
CHAPTER 4: Localization Studies of AtCle2 Proteins ..................................... 144
Abstract ..................................................................................... 145
Introduction ................................................................................ 146
Materials and methods ................................................................... 152
Results ...................................................................................... 160
Generation, purification and characterization of antibodies against
AtCle2 .................................................................. 160
AtCle2 is an integral membrane protein .................................... 166
AtCle2 migrates differently from a plasma membrane marker PMA2
during linear sucrose gradient centrifugation ....................... 170
AtCle2 is not a plasma membrane protein as revealed by two phase
partition ................................................................... 173
AtCle2 protein is localized at Golgi rich vesicles during floatation
sucrose gradient centrifugation ....................................... 173
Discussion .................................................................................. 1 80
References ................................................................................. l 84

CHAPTER 5: Generation and Characterization of RNAi Transgenic Plants for AtCle

Genes ................................................................................................. 189
Abstract .................................................................................... 190
Introduction ................................................................................ 191
Materials and methods ................................................................... 197
Results ...................................................................................... 202

Generation of RNAi constructs ................................................. 202

Generation of transgenic plants containing AtCsl RNAi constructs. . ....210

RNAi transgenic plants containing construct D1235f3 or D23t2 showed
deformed cotyledons and reduced adhesion of cotyledon cells...212

D1235f3 transgenic plants showing deformed cotyledons have
corresponding suppression of AtCle gene transcripts ........... 218

vii

D1235f3 transgenic plants showing deformed cotyledons also showed
dwarfed adult plants and shortened roots with more branching..222
Cotyledon deformation is inheritable for D1235f3 transgenic plants but

weakens when passed to the next generation ........................ 227

D1235f3 RNAi transgenic plants showed decreased glucose .............. 236
Discussion .................................................................................. 241
References ................................................................................. 248

viii

LIST OF TABLES

Table 2.1. Sequence alignment among the PsFutl -5UTR and its equivalents from all

AtFut genes .......................................................................................... 80
Table 2.2. PsFUTl activity during the elongation of third intemode between day 7, 9 and
1 1 ...................................................................................................... 86
Table 3.1. Summary of reannotated AtCsl genes ............................................. 112
Table 3.2. Characteristics of AtCsl proteins ................................................... 120
Table 3.3. Summary of sequence identities among all AtCsl proteins ..................... 123
Table 3.4. Summary of ESTs identified in the dbEST for AtCsl genes ..................... 125
Table 3.5. EST abundance for AtCsl gene families ........................................... 126

Table 3.6. Numbers of MPSS tags that found from five different Arabidopsis cDNA
libraries for all AtCsl genes and AtFutl , AtGalT and AtXTl ............................... 128

Table 3.7. Tissue expression profile of AtCsl genes revealed by Stanford Microarray
Database ............................................................................................. 135

Table 4.1. Specific activity of XyG fucosyltransferase and G811 for membrane vesicles
present in each fraction from the sucrose step gradient (Figure 5.7) ........................ 178

Table 5.1. Design of RNAi constructs targeting different AtCsl genes and primers used
to create them ....................................................................................... 204

Table 5.2. Primers used for PCR reactions to confirm the existence of constructs in
transgenic plants ................................................................................... 205

Table 5.3. Primers used for RT-PCR reactions to confirm the suppression of
transcripts ........................................................................................................................ 206

Table 5.4. Summary of the inheritance of cotyledon deformation phenotype from T1 to
T2 generation for D1234f3 RNAi lines ......................................................... 232

Table 5.5. Inheritance and segregation of cotyledon deformation phenotype and Basta
resistance marker for D1235f3 RNAi plants ................................................... 233

ix

LIST OF FIGURES

Figure 1.1. Structure of a typical xyloglucan building block ................................... 6
Figure 2.1. Cloning of full-length coding region of AtFutl .................................. 59
Figure 2.2. Expression of peptides AtFUTl-l .2 and AtFUTl-2.3 in BL21 (DE3)-pLysS
cells ................................................................................................... 62
Figure 2.3. Anti-AtFUTl polyclonal antibodies recognize an approximately 63-kDa
polypeptide in solubilized membrane proteins of Arabidopsis ................................ 66
Figure 2.4. Isolation of PsFutl fragment PsFutl -a .............................................. 68
Figure 2.5. Second round of PCR reactions amplifying cDNA fragments toward the 5’
and 3’ ofPsFutl-a ................................................................................... 71
Figure 2.6. RACE reactions amplifying the 5’ and 3’ ends ofPsFutl . .........73
Figure 2.7. DNA and protein sequences related to PsFutl and PsFUTl , and the sequence
and position of primers used for DNA fragment amplifications .............................. 74
Figure 2.8. PsFutl specific DNA probe PsSUTR .............................................. 81
Figure 2.9. Growth curve of etiolated pea intemodes .......................................... 83
Figure 2.10. Expression of PsFutl in etiolated pea intemodes at day 7 ..................... 85

Figure 2.11. Expression of PsFUTl in the third intemode of the etiolated pea at 7, 9 and

11 days old ........................................................................................... 87
Figure 3.1. Alignment of AtCslA proteins using sequences before (A) and after (B)
manual re-annotation .............................................................................. 109
Figure 3.2. Chromosomal locations of AtCsl genes ........................................... 115
Figure 3.3. Predicted intron-exon structure of AtCsl genes according to the re-annotated
sequences ............................................................................................ 116
Figure 3.4. A rooted phylogenetic tree of all AtCsl proteins ................................ 121

Figure 3.5. MPSS expression profile of all AtCsl genes and AtFutl, AtGalT,
AtXTl ............................................................................................................................. 131

Figure 4.1. Expression and detection of peptides P2 and P3 in BL21 (DE3) cells ....... 161

Figure 4.2. Specificity of AtCle2 antibodies against Arabidopsis total protein
extracts .............................................................................................. 164

Figure 4.3. Immunoblotting using purified AtCle2 antibodies against Arabidopsis crude
membranes (A) and purified Golgi complexes (B) ........................................... 167

Figure 4.4. Solubilization of AtCle2 from Arabidopsis microsomal membranes using
various reagents .................................................................................... 168

Figure 4.5. Migration of AtCle2 and selected marker proteins on a linear sucrose
density gradient ..................................................................................... 171

Figure 4.6. Two phase partitioning of Arabidopsis microsomal membranes. . . . . . . . ......174

Figure 4.7. Fractionation of microsomal vesicles by floatation centrifugation on a

sucrose step gradient ............................................................................... 177
Figure 5.1. Design of RNAi constructs for AtCsl genes ...................................... 207
Figure 5.2. Sequence similarities between RNAi construct D1235f3, D23f2 and AtCle

genes ................................................................................................. 209
Figure 5.3. Transformation and screening of RNAi plants ................................... 211

Figure 5.4. Determination of the presence of RNAi constructs in transgenic plants... . .213

Figure 5.5. Cotyledon deformation phenotype of transgenic plants carrying construct
D1235f3 or D23t2 .................................................................................. 215

Figure 5.6. Abaxial side of cotyledons from T2 plant of control lines (Control) and
D1235f3 lines (Normal, Inter I, Inter II and Strong) .......................................... 217

Figure 5.7. Cells within deformed cotyledons of D1 2358 plants showed less
adherence ........................................................................................... 219

Figure 5.8. Transcript abundance of AtCle genes in D1235f3 RNAi plants examined by

two step RT-PCR reactions ...................................................................... 220
Figure 5.9. Root length of T1 transgenic plants containing RNAi constructs ............ 223
Figure 5.10. Root length of T1 plants containing construct D1235f3 ..................... 224

Figure 5.11. T] plants containing construct D1235f3 showing deformed cotyledons also
showed more branched roots ..................................................................... 225

xi

 

r~
.
.r
hwy"
..u.‘

Figure 5.12. D1235f3 Tl plants had deformed cotyledons also yielded dwarfed adult
plants ................................................................................................. 228

Figure 5.13. Adult T2 plants containing RNAi construct D1235f3 at 6 weeks old ...... 230

Figure 5.14. Cell wall sugar composition of rosette leaves from T1 D1235f3 plants
showing different cotyledon deformation ...................................................... 237

Figure 5.15. Sugar composition analysis of T2 plants of D1235f3 RNAi lines. . . ...238

xii

611

ER?
1161
HELP
III
1155
W
“N
ill
Film}
lGl
RCII

2,4-D
AGP
Ara
AtFutl
AtGalT
AtXTl
BSA
CIP
Csl
DTT
EM
ER
Fuc
GA
Gal
GAX
Glc
GRP
HGA
HRGP
Man
MPSS
NASCArrays
PMSF
PRP
PsFutl
RG I
RG 11
Rha
SDS
SMD
TCA
TGN
UTR
XTH
XyG
Xyl

ABBREVIATIONS

2,4-Dichlorophenoxyacetic acid
Arabinogalactan protein

Arabinose

Arabidopsis xyloglucan fucosyltransferase
Arabidopsis xyloglucan galactosyltransferase
Arabidopsis xyloglucan xylosyltransferase 1
Bovine serum albumin

Calf intestinal alkaline phosphatase
Cellulose synthase-like

Dithiothreitol

Electron microscopy

Endoplasmic reticulum

Fucose

Gibberellic acid

Galactose

Glucuronoarabinoxylan

Glucose

Glycine-rich protein

Homogalacturonic acid
Hydroxyproline-rich glycoproteins
Mannose

Massively Parallel Signature Sequencing
Nottingham Arabidopsis Stock Center’s microarray database
Phenylmethanesulfonyl fluoride
Proline-rich protein

Pea xyloglucan fucosyltransferase
Rhamnogalacturonan I
Rhamnogalacturonan II

Rhamnose

Sodium dodecyl sulfate

Stanford Microarray Database
Trichloroacetic acid

Trans-Golgi network

Untranslated region

Xyloglucan endotransglucosylase/hydrolase
Xyloglucan

Xylose

* Images in this thesis are presented in color

xiii

CHAPTER 1

The Plant Cell Wall, its Structure, Function, and Biosynthesis

PLANT CELL WALL STRUCTURE AND FUNCTION
Plant cell wall functions

The presence of a cell wall is one of the unique features that distinguish plant
cells from animal cells. The walls, on one hand, provide strength, shape and rigidity to
plant cells and contribute to the structural integrity of plants. On the other hand, the walls
also allow controlled flexibility to plant cells via a controlled weakening and loosening,
thereby allowing cell expansion (Brett, 1983).

Plant cell walls limit the ability of macromolecules, such as large proteins or
nucleic acids, to enter or leave cells. Walls are also involved in the apoplastic transport of
small molecules and ions, including some small proteins and polysaccharides (Carpita et
al., 1979). The movement through plant cell walls is also charge-dependent because the
walls bear a net negative charge (Carpita et al., 1979).

Plants are constantly exposed to attacks from many microorganisms that are
potentially pathogenic. Many plant defense mechanisms involve the cell wall. An intact
cell wall is an effective physical banier, protecting against pathogen penetration and
spread (Stacey et al., 1992). In addition, the cross—linking of cell wall structural proteins
strengthens the cell walls as a rapid response in the initial stages of the hypersensitive
response before transcription-dependent defense mechanisms occur (Brisson et al., 1994).
The formation of effective papillae at the site of infection provides further mechanical
barriers to penetration (Israel et al., 1980). Another role for cell walls is the elicitation of
phytoalexin production in response to pathogen attacks. Many endogenous elicitors of
phytoalexin production are derivatives of degraded wall polysaccharides created by

pathogen attack or abiotic stresses (Stacey et al., 1992). Oligosaccharide products derived

by degradation of wall polysaccharides may even diffuse to neighboring cells to trigger
defense responses, or sometimes act as signal molecules regulating cell expansion and

plant development (Aldington et al., 1991).

Plant cell wall structures

Electron microscopic examination showed that many plant cell walls consist of
three layers, a middle lamella, a primary wall and sometimes a secondary wall (Brett and
Waldron, 1996). The middle lamella is the first layer formed immediately after cell
division and is derived from the cell plate. The primary wall is the next major layer
deposited from daughter cells onto the middle lamella after cell division is completed
(Brett and Waldron, 1996). The plant cells continue to deposit primary wall as long as the
cell enlarges and its surface area grows (Brett and Waldron, 1996). The primary wall
usually reaches a thickness of 50-100 nm and remains extensible for possible further
growth (Kaczkowski, 2003). Many cells only have two layers of walls. For other cell
types, a third layer of wall, the secondary wall, is deposited afier cell expansion stops and
differentiation begins. The thickness of the secondary wall is not as homogeneous as the
primary wall and varies significantly even for different parts of a cell. This secondary cell
wall contributes to many structurally and morphologically different cell types with
unique functions (Preston, 1974; Bacic et al., 1988). For example, cells building up
xylem elements or sclerenchyma have thick secondary walls and are mechanically more
resistant (Kaczkowski, 2003). The deposition of secondary wall is usually accompanied
by lignification to make the walls even stronger (Kaczkowski, 2003). Although a

secondary wall is composed mostly of cellulose and lignin, other polysaccharides such as

xylans and glucomannans are also found in many cases to be deposited along with the
cellulose (Fincher and Stone, 1982; Northcote, 1985; Gibeaut and Carpita, 1994).

Biochemically, the primary cell wall consists of two phases, a microfibrillar
skeleton and a matrix (Carpita and Gibeaut, 1993). The microfibrillar phase has a high
degree of crystallinity and is formed by microfibril structures that are composed solely of
cellulose, a highly structured collection of B-l,4 glucan chains (Carpita and Gibeaut,
1993). The glucan chains are arranged in a crystalline lattice within each microfibril
through intramolecular or intermolecular hydrogen bonding, thereby giving rise to the
high tensile strength of the microfibrils (Carpita and Gibeaut, 1993).

The wall matrix phase is chemically complex and consists of many different
polysaccharides, proteins and phenolic compounds (Brett and Waldron, 1996). There are
two major categories of wall matrix polysaccharides, hemicellulose and pectin,
characterized by their extractabilities by different reagents (Selvendran, 1985). Pectins
are readily extracted by hot water with a chelating agent such as ammonium oxalate,
EDTA, or dilute acid solution, whereas hemicelluloses are extracted only with alkaline
solution (Selvendran, 1985).

Pectins are a mixture of variably branched and strongly hydrated polysaccharides
that are rich in galacturonic acid, rhamnose, arabinose and galactose (Selvendran, 1985 ).
The presence of pectins controls the wall porosity and charge. Pectic polysaccharides
help cells adhere to form tissues, keep wall enzymes in position, and also keep tissues at
proper hydration level (Kaczkowski, 2003). Pectin polysaccharides also act as signal
molecules in response of symbiotic organisms, pathogens, and insects (McNeil et al.,

1984). The pectic polysaccharides comprise a group of acidic polysaccharides, including

homogalacturonan (HG), rhamnogalacturonan I (RG 1), and rhamnogalacturonan 11 (RG
11), as well as several neutral oligosaccharides and polysaccharides, including arabinan,
galactan, and arabinogalactan that usually appear as side chains of the former three types
(Bacic et al., 1988; Gibeaut and Carpita, 1994). RG I and RG 11 are extremely complex
polysaccharides with numerous side-chain modifications of arabinans or galactans
(Mohnen, 1999).

Overall, hemicelluloses generally have simpler side chains compared to pectic
polysaccharides. Their backbones, on the other hand, are similar to B-glucan chains of
cellulose so that hemicelluloses can form numerous hydrogen bonds with the cellulose
molecules (Gibeaut and Carpita, 1994). This cross-linking ability of hemicelluloses is
their main function and helps provide wall structural integrity. Major hemicelluloses
include xyloglucan (XyG), [3-(1,3)-(1,4) mixed-linkage glucan, glucuronoarabinoxylan
(GAX), arabinoxylan, xylan, glucomannan, galactomannan, glucuronomannan, and
mannan (Carpita and Gibeaut, 1993). The types of hemicellulose vary greatly in different
plant species, as does the structure or amount of a particular hemicellulose (Carpita and
Gibeaut, 1993).

XyG is the major hemicellulose in the primary cell wall of most dicot plants and
consists of a backbone of a B-(l ,4)-linked glucan chain that is modified with 0t-( 1,6)-
xylose, B-(1,2)-galactose, and sometimes a-(1,2)-fucose (Figure 1.1) (Hayashi, 1989).
The mixed-linkage glucans are glucose chains with B-(1,3) and B-(l,4) linkages. Xylans
and mannans are polymers of single sugar residues of [3-(1 ,4)-xylose or B-(l ,4)-mannose,
respectively (Brett and Waldron, 1996). Other xylan- and mannan- derived

hemicelluloses either have more than one sugar residue in their backbones

Fuc
a(1,2)
1

Gal Gal
3 130.2)
.. it it
Xyl Xyl Xyl -

lamp) loc(l,6) a(1.6)
I,

_- Glc _p G10 —5 Glc —>Glc -—
130.4) 130,4) 130,4)
_ 3-20k

 

 

 

 

. . . . *0 Absent in some Xyloglucans

—’ Present in all Xyloglucans

Figure 1.1: Structure of a typical xyloglucan building block. Adapted and modified

from Hayashi, 1989 and Fry, 1989.

(glucomannan,), or have side chain modifications (galactomannan, arabinoxylan, GAX),
or both (glucuronomannan) (Brett and Waldron, 1996).

Cell walls contain five major classes of structural proteins, hydroxyproline-rich
glycoproteins (HRGPs) such as extensins, glycine-rich proteins (GRPs), proline-rich
proteins (PRPs), solanaceous lectins, and arabinogalactan proteins (AGPs) (Showalter,
1993). There are also many enzymes associated with the walls such as peroxidases,
phosphatases, invertases, proteases and many polysaccharide-degrading enzymes
(Showalter, 1993).

The composition of the wall matrix varies significantly from plant to plant, cell to
cell, and even at different developmental stages of the same cell type. Most of the
compositional variation comes from differences in the categories, proportions, and
structural modifications of the polysaccharides present in the wall matrix (McNeil et al.,
1984). The primary walls of dicot plants all contain a high proportion of pectic
polysaccharides and a smaller amount of Xsz (Bacic et al., 1988). Gymnosperrns have a
similar primary wall matrix with the presence of pectins and Xsz (Burke et al., 1974;
Thomas et al., 1987). In the Poaceae plants, the primary walls usually contain highly
substituted GAX as the major matrix polysaccharide as well as variable amounts of B-
(1,3)-(1,4) mixed-linkage glucans. The pectic polysaccharides and Xsz are usually
minor components (Bacic et al., 1988). Most non-graminaceous monocot plants have a
cell wall matrix composition similar to that of dicot plants, while some of them showed
similarity with graminaceous plants (Bacic et al., 1988).

Therefore, according to the different compositions, the primary walls in higher

plants were classified into two distinct types (Carpita and Gibeaut, 1993). The type I wall

 

.-

 

 

 

is representative of all Dicotyledonae and some Monocotyledonae. This type of wall has
a wall skeleton structure of cellulose microfibrils cross-linked by the hemicellulose Xsz
(Carpita and Gibeaut, 1993). Other non-cellulosic polysaccharides having similar inter-
linking functions, such as glucomannan, galactomannan, glucuronomannan, B-(1,3)-
glucans, and GAXs are also found in some type I walls but in very low amounts (Maltby
et al., 1979; Darvill et al., 1980; Meier and Reid, 1982; Bacic et al., 1988). Xsz bind
tightly to the surface of cellulose microfibrils through hydrogen bonds and connect
different microfibrils (McCann and Roberts, 1992; Carpita and Gibeaut, 1993). Type I
primary walls comprise three domains that are structurally different but interacting with
each other (Carpita and Gibeaut, 1993). The first domain, cellulose-XyG network, is the
basic wall structure framework that is embedded in a second domain of matrix pectin
polysaccharides including mainly polygalacturonic acids (PGA) and R63 (Carpita and
Gibeaut, 1993; Gibeaut and Carpita, 1994). Chains of PGA are cross-linked with Ca2+ to
form junction zones for chain linkings. The size of j unction zones and the size and
frequency of side chains on RGs control the wall porosity (Brett and Waldron, 1996). The
third wall structural domain consists of the structural proteins (Carpita and Gibeaut,
1993)

The type II wall is found only in Poaceae and closely related monocot families
(Carpita and Gibeaut, 1993). Structurally the type 11 primary wall is based on the same
cellulose microfibrils that are cross linked by GAXs instead of Xsz. The GAXs have a
backbone of B-(1,4)-D-xylose residues that are modified with single arabinose residues at
0-3 positions or glucuronosyl acid (Gch) residues at the 0-2 positions of the backbone

xyloses (Gibeaut and Carpita, 1994). The degree of side chain modification varies greatly

among different plants. The variation determines the abilities of GAXs to bind with each
other. The presence of high amount of side chains prevents the hydrogen bonding among
themselves or with cellulose microfibrils (Carpita and Gibeaut, 1993). The type II walls
not only have very little amount of pectins but also display a developmental preference
for methyl esterified or unesterified pectins in specific cell types (Knox et al., 1990). The
esterified PGAs are usually found in vascular tissues while the unesterified PGAs are
usually found in the coritcal cells of root tips and parenchyma cells of seedlings and

coleoptiles (Knox et al., 1990).

XYLOGLUCAN STRUCTURE AND FUNCTION
X yloglucan structures

Xsz have a cellulose-like backbone composed of B-( l ,4) linked D-glucosyl
residues (Figure 1.1). Some of the backbone glucosyl residues are decorated with a-D-
xylosyl residues in 1,6 linkages in a regular pattern that varies depending upon the plant
species. The combination of backbone glucosyl residues and their immediate xylose side
chains constitute the core building blocks for each XyG molecule. Some of the xylose
residues of Xsz are further substituted at the C-2 positions with mono-, or disaccharides
containing B—D-galactose, a-L-arabinose, and a—L-fucose residues (Vierhuis et al., 2001 ).
The constitution and distribution of side chains again is plant-species and tissue-type
specific (Fry, 1989; York et al., 1996; Vincken et al., 1997; Pauly et al., 2001).

A concise nomenclature was developed to unambiguously designate XyG
structures (Fry et al., 1993). XyG molecules are divided into segments consisting of only
a single backbone glucosyl residue and its side chain residues attached. Each segment
with a different overall structure is designated with a different code letter. The letter “G”
refers to the unbranched backbone B-D-Glc, “X” refers to an a-D-Xyl-(l ,6)-B-D-Glc
disaccharide segment, “L” refers to the trisaccharide B-D-Gal-(l,2)-0t-D-Xyl-(1,6)-B-D-
Glc, and “F” refers to the tetrasaccharide a-L-Fuc-(l ,2)- B-D-Gal-(l ,2)-0t-D-Xyl-(l,6)-B-
D-Glc (Fry et al., 1993) (Figure 1.1). XyG from solanaceous plants usually contains
terminal L-arabinose instead of fucose and is therefore designated as arabinoxyloglucan
(Vincken et al., 1997; Vierhuis et al., 2001). The trisaccharide a-L-Ara-(l,2)-0t-D-Xyl-
(1 ,6)-B-D-Glc then is referred as “S”, and the tetrasaccharide B-L-Ara-(l,3)-0t-L-Ara-

(l,2)-0t-D-Xyl-(l,6)-B-D-Glc is referred as “T” (Fry et al., 1993; Vincken et al., 1997).

10

Using this convenient nomenclature system, the two major core building blocks of
XyG are named as XXXG and XXGG (Vincken et al., 1997). XXXG type of Xsz
usually have 75% of the backbone glucoses substituted at their 6 positions with a-D-
xylose at a regular pattern of three substituted residues followed by one unsubstituted
residue (Figure 1.1). XXGG type of Xsz usually have clusters of two instead of three
branched glucose residues alternating with a two glucose residues without any
substituents (Vincken et al., 1996; York et al., 1996).

XXXG type of XyG is the major hemicellulose in the primary wall of
dicotyledonous and non-graminaceous monocotyledonous plants (Pauly et al., 2001).
They usually contain modifications of B-D-Gal or disaccharide a-L-Fuc-(l ,2)-B-D-Gal
on the xylose residues and yield structures of XXLG, XLLG, XXFG and XLFG (York et
al., 1990; Hisamatsu et al., 1991). Furthermore the presence of fucose residues is
regarded as a specific feature of dicotyledonous Xsz, as Xsz from graminaceous
monocotyledonous plants generally lack the fucose residues. However some exceptions
have been reported (Hayashi, 1989; McDougall and Fry, 1994).

The XXXG type of XyG is also found in seeds as storage polysaccharide,
including seeds from T amarindus indica (York et al., 1990), cyclamen (Braccini et al.,
1995), T ropaeolum majus (McDougall and Fry, 1990), and Detarium senegalense (Wang
et al., 1996). These Xsz rarely bear any other substituents except some B-D-galactose
on some of the xylose residues, to give structures XXLG and XLLG.

The XXGG type of Xsz are usually found in solanaceous plants as well as
Poaceae plants. One of the unique feature of this type of Xsz is the presence of terminal

a-L-arabinose to form core structures XSGG (Hayashi, 1989; Vincken et al., 1996; York

11

et al., 1996). The presence of fucose has not yet been found in this type of Xsz

(Vincken et al., 1997).

Structural roles of xyloglucan

XyG forms a load-bearing network with cellulose. It not only adheres to the
surface of cellulose microfibrils, but also can connect different cellulose microfibrils at
its two ends to contribute to wall plasticity and elasticity (Fry, 1989). The linkages
between XyG and cellulose microfibrils are non-covalent hydrogen bonds (Thompson
and Fry, 2000).

It was proposed many years ago that in dicot plants, Xsz and pectic
polysaccharides, as well as the structural glycoproteins, are covalently joined together to
form the wall matrix phase (Keegstra et al., 1973). Evidence obtained more recently
fiirther supports this hypothesis. XyG and pectic polysaccharides were repeatedly co-
extracted from the wall and found to be inseparable from each other by ethanol
precipitation, gel permeation chromatography, and ion exchange chromatography.
(Chambat et al., 1984; Joseleau and Chambat, 1984; Selvendran, 1985; Selvendran et al.,
1985; Femenia et al., 1999). Treatment of those XyG-pectin complexes with endo-
glycosidases specific for a glycosyl linkage in one polymer would always cause
molecular weight decrease of both XyG and pectic components, strongly indicating an
association of XyG and pectic polysaccharides (Femenia et al., 1999; Watt et al., 1999;
Thompson and Fry, 2000). It has been suggested that the covalent linkage between Xsz
and pectic polysaccharides could be alkali-stable bonds of glycosidic or other forms

(Keegstra et al., 1973; Thompson and Fry, 2000), alkali-labile ester bonds (Mueller et al.,

12

1976; Kim and Carpita, 1992; Brown and Fry, 1993; Brown and Fry, 1993), or through
oxidatively coupled phenolics such as diferuloyl esters (Grabber et al., 1995; Parr et al.,

1996).

X yloglucan ’s role during cell expansion

The growing plant cell wall is a dynamic structure with constant metabolism of its
constituent polysaccharides during cell growth and expansion. Apart from the usual
degradation of existing molecules and the synthesis of new molecules, there is an
additional key aspect of XyG metabolism. As one of the major components of the wall
network connecting cellulose microfibrils and pectic polysaccharides, XyG constitutes
20-25% of the dry weight of the primary cell wall in dicot plants and contributes
substantially to the mechanical strength of the cell wall fabric (Fry, 1989). Therefore
XyG has to be disconnected from other polysaccharides in the wall or degraded to allow
the loosening of the existing wall so that more newly synthesized wall materials can be
incorporated during expansion. Two key enzymes, expansins and XyG
endotransglucosylases/hydrolases (XTHs), are thought to be involved in these processes.

Expansins are cell wall-associated proteins that weaken the non-covalent binding,
namely hydrogen bonds, between XyG and cellulose microfibrils, to allow turgor-driven
polymer creep (Rayle and Cleland, 1992; Cosgrove, 2000). This reaction would then
separates existing wall polysaccharides and allow new wall materials to be deposited and
inserted among the previously existing ones.

XTH was first identified as an enzyme catalyzing the transglycosylation of Xsz

(Fry et al., 1992; Nishitani and Tominaga, 1992). In contrast to expansins, which do not

13

possess hydrolytic activities, the XTHs cleave the glycosidic linkages in the backbone of
XyG molecules. Instead of transferring the sugar to water, yielding hydrolysis of the
glycosidic bond, XTHs transfer the cleavage product to another XyG molecule to form a
new glycosidic linkage (McQueen-Mason, 1997). It is proposed that, by doing so, XTHs
loosen the connections mediated by XyG molecules and allow temporary wall weakening.
The rejoining between different existing XyG molecules or between newly deposited
XyG molecules and old ones accomplish the expansion of walls as well as the

incorporation of new wall materials, yet reconstitutes the wall network structure (Rose et

aL,2002)

Functions of xyloglucan fragments

Oligosaccharin is a name for oligosaccharides derived from the degradation of
fungal and plant cell wall polysaccharides that have the ability to modify plant growth or
defense responses (Ozeretskovskaya and Romenskaya, 1996). The XyG degradation
product, XXF G, is one of the earliest identified oligosaccharins. It was shown to have an
effect on plant growth and development (York et al., 1984; McDougall and Fry, 1991).

In excised pea stem segments or whole shoots, the nonasaccharide XXF G at a
concentration of 10’11 to 10'8 M can inhibit 2,4-dichlorophenoxyacetic acid (2,4-D)-
induced elongation (York et al., 1984; McDougall and Fry, 1988; Emmerling and Seitz,
1990; Augur et al., 1992). At this concentration, XXF G also inhibits H+-induced growth
(Lorences et al., 1990), gibberellic acid (GA3)-induced growth (Warneck and Seitz, 1993)
and endogenous growth (Warneck and Seitz, 1993). The growth inhibition effect of

XXF G is dependent on the terminal or-L-fucose or the disaccharide a-L-fucose-(l,2)-[3-

l4

D-galactose, since neither XXLG nor XXXG showed similar growth inhibition abilities
(McDougall and Fry, 1989). On the other hand the non-reducing terminal a-D-xylose and
the reducing terminal D-glucose are not essential for the inhibition activity because the
oligosaccharides GXFG and XXFGol are active growth inhibitors (Augur et al., 1992).
Interestingly, XXF G also shows a growth-promoting effect in pea stem segments
when applied at a higher concentration of about 10'6 M and in the absence of exogenous
phytohormones (McDougall and Fry, 1990; Warneck and Seitz, 1993; Cutillas and
Lorences, 1997). This growth promotion ability is not dependent on the fucose-galactose
side chains but requires a Xyl/Glc-rich backbone (Warneck et al., 1998). The mechanism
and signal transduction pathway of the XyG Oligosaccharin controlling plant growth and

development are still poorly characterized.

X yloglucan as a reserve carbohydrate

The seeds of plants have nutrition stored in the form of carbohydrate, lipid or
protein. Those substances are mobilized during seed germination before the seedlings
become self-sufficient, to allow for energy generation and to produce raw materials for
building cells and tissues in the germinating seedlings (Mayer and Poljakoff-Mayber,
1975). The most common carbohydrate stored in seeds is starch. Other storage
carbohydrates include fructan and cell wall polysaccharides such as galactomannans,
Xsz and galactans (Buckeridge et al., 2000). These storage polysaccharides are normal
components of the cell walls of other tissues but in much lower quantities.

XyG is the storage polysaccharide accumulated in the cotyledonary tissues of

several dicot plants including tamarind (T amarz‘ndus indica) (Kooiman, 1961), nasturtium

15

(T ropaeolum majus) (Edwards et al., 1985), Hymenaea courbaril (Buckeridge and
Dietrich, 1990) and Copaifera langsdorflii (Buckeridge et al., 1992). Comparative studies
of Xsz from seeds of these plants showed the same basic structural units XXXG,
XLXG, XXLG and XLLG (Buckeridge et al., 1992). It was also found that the ratios
between these units changes depending on the fine structures shown by different species
and different populations of the same species growing under different environmental
conditions (Buckeridge et al., 1992). However all seed Xsz showed similar
monosaccharides proportions, indicating that a similar degree of branching is preserved,
but the distribution of blocks with different degree of galactosylation determines the fine
structure of Xsz from different sources (Buckeridge et al., 2000).

Studies in T. majus (Edwards et al., 1985), T. indica (Reis et al., 1987), C.
langsdorffii (Buckeridge et al., 1992), and H. courbaril (Tine et al., 2000) have shown
that four degradative enzyme activities are involved in the mobilization of storage XyG.
All four types of enzymes have been isolated from T. majus. They are a XyG-specific
endo B-(1,4)-D-glucanase or XTH (Edwards et al., 1986; Fanutti et al., 1993), a B-
galactosidase with high specificity for XyG (Edwards et al., 1988), a XyG-

oligosaccharide specific a-xylosidase or oligoxyloglucan exo-xylohydrolase (F anutti et

al., 1991), and a transglycosylating B-glucosidase (Crombie et al., 1998).

16

IDENTIFICATION OF GENES INVOLVED IN WALL BIOSYNTHESIS

Genes involved in xyloglucan biosynthesis

According to the chemical structure of XyG one could predict the need for at least
four different enzymes involved in its biosynthesis: a B-glucan synthase(s) to make the
backbone, plus a-xylosyltransferase(s), 13-galactosyltransferase(s) and or-
fucosyltransferase to add the side chains. When I began my thesis work, none of the
genes encoding these enzymes had been identified. I joined the first efforts to identify the
gene encoding the XyG fucosyltransferase (Perrin et al., 1999). An enzyme activity assay
for XyG fucosyltransferase was developed using the tamarind XyG, which naturally lacks
the terminal fucose, as the acceptor substrate, and 14C labeled GDP-fucose as the sugar
donor (Farkas and Maclachlan, 1988). The XyG fucosyltransferase was purified from
etiolated pea epicotyls through a GDP-affinity and size exclusion column with the XyG
fucosyltransferase activity monitored at each step (Perrin et al., 1999). The peptide
sequences obtained from the purified enzyme led to the identification of an Arabidopsis
EST clone, which subsequently led to the identification of the full-length cDNA sequence
for Arabidopsis XyG fucosyltransferase (AtFutl) (Perrin et al., 1999). The pea version of
the XyG fucosyltransferase gene, PsFutl, was also identified from a cDNA library
utilizing the peptide sequences identified from the purified enzyme (Faik et al., 2000).
The identity of the AtFutl was confirmed by the precipitation of XyG fucosyltransferase
activity from detergent solubilized Arabidopsis microsomes with antibodies against an E.
coli-expressed peptide encoded by AtFutl , and the detection of XyG fucosyltransferase

activity from AtFUTl proteins expressed in COS7 cells (Perrin et al., 1999).

17

An AtFutl mutant mur2, was identified from a screen for altered monosaccharide
Composition in cell walls (Reiter et al., 1997). These mutants have a 50% reduction in
total cell wall L-fucose content and showed a reduction in XyG fucosylation to about 1%
of wild type due to the mutation in the AtFutI (Vanzin et al., 2002). Furthermore, XyG is
the only polysaccharide affected in the mur2 mutant plants (Vanzin et al., 2002).
However, there is no visible phenotype observed for mur2, and its wall strength was
indistinguishable from that of the wild-type plants (Vanzin et al., 2002). A total of nine
DNA sequences that have significant sequence identity with AtFutl have been identified

in the Arabidopsis genome database (Sarria et al., 2001). Together with AtFUTl and the
PsFUTl, proteins encoded by these nine genes constitute a new glycosyltransferase
family in the CAZY database (family 37, http://afmb.cnrs—mrs.fr/~pedro/CAZY)
(Henrissat and Davies, 2000). AtF UT proteins other than AtFUTl did not show XyG-
SPecific fucosyltransferase activities when they were expressed in tobacco BY2 cell
cultures (Sarria et al., 2001). These AtF UT proteins showed very low identity with
fucosyltransferases involved in the biosynthesis of N-linked glycoproteins and were
proposed to be fucosyltransferases forming fucose linkages present in cell wall
Carbohydrates other than XyG, namely, RG 1, RG 11 and AGPs (Sarria et al., 2001).
The XyG galactosyltransferase was identified from a forward genetic approach.
The mur3 mutant was identified in the same screen that identified murZ (Reiter et al.,
1997). The mur3 plants also showed a 50% reduction in cell wall fucose content just as
mur2 Plants do (Reiter et al., 1997). When the cell wall materials from mur3 and wild-
type Plants were fractionated into pectin- and XyG- rich fractions, significant decreases

of fuCose and galactose contents in both 1 M KOH and 4 M KOH soluble fractions were

18

 

 

.‘Q

.H. A...

 

 

'h—. .

s.
. ‘d‘

 

observed, indicating the possibility that the disaccharide side chain a-L-fucose-(1,2)-B-
D- galactose was missing (Madson et al., 2003). When the Xsz from mur3 plants were
digested with endo-B-D-glucanase, two of the normally seen XyG building blocks XXFG
and )(LFG were missing (Madson et al., 2003). The gene mur3 was identified through
positional cloning. When expressed in Pichia pastoris, MUR3 showed XyG-specific

gal actosyltransferase activity and therefore confirmed the identity of MUR3 to be a XyG

gal actosyltransferase (Madson et al., 2003).

The XyG xylosyltransferase was identified through a candidate gene approach. In

XyG the a-D-xylose is added to the B-glucan backbone in a 1,6 linkage. This glycosidic
reaction is very similar to the one catalyzed by the galactomannan-specific a-(1,6)

gal actosyltransferase (Faik et al., 2002). When the sequence of the galactomannan a-(l,6)
gal actosyltransferase gene from fenugreek (Edwards et al., 1999) was used to search the
Arabidopsis genome database, a group of seven genes that share low sequence similarity
With it was identified. When these genes were expressed in Pichia pastoris, one of them,
AtXTl, showed XyG specific xylosyltransferase activity (F aik et al., 2002). This XyG-
SpeCific xylosyltransferase activity assay used cellopentose as substrate acceptor, and the
I)rOCluct hydrolysis analysis indicated the xylose residue was added onto the second
g1 uCose residue from the non—reducing end (Faik et al., 2002). The other six homologs of
Athl did not show this XyG-specific xylosyltransferase activity during the same study,

and their functions are still not known (Faik et al., 2002).

A gene encoding the XyG backbone glucan synthase has not yet been identified
ai 1113/ due to the lack of a specific enzyme activity assay method. Currently, several

onps, including ours, are working vigorously in trying to identify this gene by

19

developing the XyG glucan synthase assay and using reverse genetics and genomic
candidate gene approaches. One attractive hypothesis is that one or more of the cellulose
synthase-like genes encode(s) the XyG glucan synthases. Supporting evidence includes
the fact that the chemical structures of cellulose and XyG backbone are exactly the same,
and the fact that cellulose synthase-like genes share significant sequence identity with

genes encoding cellulose synthases (Doblin et al., 2001).

Gen es involved in the biosynthesis of other wall polysaccharides
The identification of genes encoding enzymes involved in plant cell wall
polysaccharide biosynthesis began only in the mid—to-late 1990’s with the identification
of the first cellulose synthase from cotton (Pear et al., 1996). Even at the present time,
only a few genes have been identified and functionally characterized even though
hundreds of enzymes are predicted to participate in the biosynthesis of the various cell
Wall polysaccharides. Beside the genes that encode enzymes involved in XyG
bioSynthesis, as described above, other genes identified include those encoding the
cellulose synthase catalytic subunits (Williamson et al., 2002), two genes encoding
enZyl'nes for the biosynthesis of galactomannan (Edwards et al., 1999) (Dhugga, personal
Corrll'nunication), and two genes encoding enzymes involved in pectin polysaccharide
biosynthesis (Bouton et al., 2002; Iwai et al., 2002).
The first cellulose synthase gene was identified through an effort of random
Seqtlencing of clones from a cDNA library constructed from developing cotton fiber at 21
days post anthesis, at which time cotton fibers make primarily cellulose in very high

quatltities (Pear et al., 1996). A full length clone identified showed sequence similarities

20

with the bacterial cellulose synthase, celA. When expressed in E. coli the proteins
encoded by this cDNA specifically bound to UDP-glucose, the presumed substrate for
cellulose synthase (Pear et al., 1996). An Arabidopsis mutant rsz was identified later.
The mutant plants grown at a restrictive temperature of 30°C contained only half the
amount of cellulose present in wild type plants (Arioli et al., 1998). The mutant turned
out to be defective in a gene encoding a glycosyltransferase closely related to the enzyme
identified as cellulose synthase in cotton fibers (Arioli et al., 1998). Combining the
protein sequence comparison analysis with bacterial cellulose synthases and the
cellulose-deficient mutant analysis, the first Arabidopsis cellulose synthase gene was
finally identified and designated as the cellulose synthase catalytic subunit, AtCesAl.
The Arabidopsis genome contains a cellulose synthase gene family of ten genes.

Most of these ten genes have been linked to cellulose production by characterizing their
corresponding mutant plants. AtCesAl (Arioli et al., 1998), AtCesA3 (Burn et al., 2002),
AtCesA4 (Fagard et al., 2000), AtCesA6 (Taylor et al., 1999), AtCesA7 (Zhong et al.,

2003), and AtCesA8 (Taylor et al., 2000) all showed decreased cellulose production in

their corresponding mutant or antisense plants. AtCesA3 and AtCesA6 mutations also

resulted in resistance to isoxaben, a herbicide that selectively inhibits cellulose synthesis

(Scheible et al., 2001; Desprez et al., 2002). Mutant plants for all these AtCesA genes,

p his the antisense plant of AtCesA2 (Burn et al., 2002) all showed morphological

p hen0types that are consistent with a cellulose deficiency. Therefore all ten AtCesA
genes seem to have a function involved in cellulose synthesis.

Galactomannan is a hemicellulose that is synthesized as the major storage

p 0 1ySaccharide and deposited in large quantity in seed endosperm at a certain stage of

21

legume seed development (Reid and Meier, 1970; Meier and Reid, 1977; Edwards et al.,

1992). Galactomannan has a B-(l ,4)-mannan backbone decorated with CX-(l,6) linked

single galactose residues. A gene encoding the a-(l,6)-galactosyltransferase involved in
the attachment of the galactose residues to galactomannan backbone was first identified
from fenugreek (Trigonellafoenum-graecum L) seed endosperm (Edwards et al., 1999).
Microsomal membranes were prepared from developing fenugreek seed endosperm and
solubilized with Triton X-100. With a combination of non-denaturing isoelectric focusing
and SDS-PAGE, a protein band in the gel was identified to be associated with the
galactosyltransferase activity (Edwards et al., 1999). This protein was subjected to
sequencing and the peptide sequence information obtained allowed the isolation of a
CDNA clone which, when expressed in Pichia pastoris, showed high activity of
galactomannan specific galactosyltransferase (Edwards et al., 1999).

Very recently the gene encoding the enzyme involved in the backbone
bioSynthesis of galactomannan, the B-mannan synthase, was identified in guar seed
(Kanwarpal S. Dhugga, personal communication). Again, the rapid deposition of
galactomannan in large amounts in developing guar seeds was utilized as an advantage.
An EST library was constructed from the developing seeds. One of the cDNA clones
iSolated, when expressed in soybean somatic embryos, showed high level of activity for
B‘l‘l’lannan synthase (Kanwarpal S. Dhugga, personal communication). The hydrolysis
analysis of the enzyme reaction product was also confirmed to be B-mannan (Kanwarpal
8‘ Dhugga, personal communication).

A RG II-specific glucuronyltransferase gene (NpGUTl) was identified from

Nicotiana plumbagim'folia through T-DNA tagging (Iwai et al., 2002). NpGUTl is

22

similar to the gene encoding the catalytic domains of animal glucuronyltransferases
involved in heparan sulfate synthesis (Iwai et al., 2002). When the expression of
NpGUTl was suppressed by an antisense gene, the glucuronic acid content level was
remarkably reduced in the NazCO3-soluble pectic fractions. In particular, the glucuronic
acid in RG 11 of cell wall materials was totally missing, confirming a firnction involved in
the formation of a rather specific sugar linkage (Iwai et al., 2002).

Another gene encoding an enzyme involved in pectic polysaccharide biosynthesis
was identified through the characterization of an Arabidopsis T-DNA insertion mutant
QUA SIMODOI (qua!) (Bouton et al., 2002). The qua] plants were dwarfed and showed
reduced cell adhesion (Bouton et al., 2002). The biosynthesis of homogalacturonic acid
(HGA) in qua] plants was affected since cell walls from the mutant plants had 25% less
uronic acid and galacturonic acid, and a strong decrease of HGA backbone epitopes
SDeCifically recognizable by monoclonal antibodies JIMS and JIM7 (Bouton et al., 2002).
The cloned qual gene encodes a type II transmembrane protein that belongs to family 8
of glycosyltransferases (Bouton et al., 2002).

It is still a daunting task and a frustratingly slow process to identify all of the
g1 ycosyltransferases involved in the biosynthesis of wall matrix polysaccharides, given
the large number of different sugar linkages present (Perrin et al., 2001). From these a
few successful gene identifications described here, it is realized that the biochemical
approach and the genomic candidate approach are both feasible and very efficient,
especially when these two approaches are combined (Edwards et al., 1999; Perrin et al.,
I 999; Faik et al., 2002) (Kanwarpal S. Dhugga, personal communication). In addition,

e . . .
nzyme actrvrty assays seem to be the most valuable tool and almost always a necessrty

23

since the biochemical function of each glycosyltransferase needs to be confirmed
individually. On the other hand, a wise choice of plant species and plant tissue is critical,

especially for the success of the genomic candidate gene approach.

Functional studies of AtCsl genes

In an effort to identify genes encoding enzymes involved in wall matrix
pol ysaccharide biosynthesis, genes showing significant sequence identity with known
wall biosynthesis genes have drawn particular interest. The identification of AtCesA
genes through genomic (Pear et al., 1996) and genetic approaches (Arioli et al., 1998)
was a major breakthrough that allowed additional gene identification studies through
Comparative genomics (Cutler and Somerville, 1997; Richmond and Somerville, 2000;
Saxena and Brown, 2000).

From the Arabidopsis genomic database, a Blast search using Arabidopsis
cellulose synthase 1 (AtCesAl, Atstl) as query sequence identified a superfamily of
40 genes. Among them 10 cellulose synthase genes showed more than 50% identity with
each other and are proposed to have similar functions. Most of these ten genes were later
identi fied as encoding true cellulose synthases through mutant plant characterizations
(Williamson et al., 2002). The other 30 homologs showed less than 50% identity with
AtCesA and were named as Arabidopsis cellulose synthase-like (AtCsl) genes
(Richmond and Somerville, 2000; Saxena and Brown, 2000). The proteins encoded by
“1386 AtCsl genes are sufficiently divergent from AtCesA proteins that they are not very

h.kely cellulose synthase catalytic subunit isoforms (Richmond and Somerville, 2001).

AtCsl proteins not only share identity with AtCesA proteins in the primary sequences, but

24

 

,._. v-

i.. “-21

‘7: ‘1”‘1

us.. but.

3150 contain a motif D..D. . .D..QxxRW that is indicative of B processive

g1 ycosyltransferases and thought to be important in the catalytic site of these enzymes
(Saxena et al., 1995; Pear et al., 1996; Richmond and Somerville, 2001). Like CesA
proteins, AtCsl proteins all appear to belong to family 2 of inverting nucleotide-
diphospho-sugar glycosyltransferases (Campbell et al., 1997). These AtCsl proteins all
contain multiple transmembrane domains organized in a way similar to AtCesA proteins
with l or 2 transmembrane domains in the amino-terminal region and 3 to 6 of them in
the carboxyl-terminal region (Richmond and Somerville, 2001).

Given their structural features, AtCsl proteins have been proposed to be enzymes
involved in the biosynthesis of B-linked polysaccharide backbones, such as XyG, mannan,
galactomannan, and GAX (Richmond and Somerville, 2001), or even cellulose (Doblin et
al., 2001). Direct evidence supporting this hypothesis came from the recent observation
that a guar CslA gene encodes the guar galactomannan-specific mannan synthase

(Kanwarpal S. Dhugga, personal communication).

25

»
...u\ui

un-‘~-

--.h. .

I
'1...
u
*v ‘0
\ .
.g-k.
u
a.
to '
.'-~'«
. ‘
I.
' .

 

 

‘ t.
‘ D-
a.
"l
t

I.
.r~‘
u,;
q'
*4.
~‘ q
4.
§.‘ ‘

REGULATION OF CELL WALL METABOLISM

The plant cell wall, with its dynamic functions, not only maintains the integrity of
p1ant cells but also participates in the regulation of cell growth. Consequently, wall
structure is under frequent modulation by endogenous and external stimuli. The term
“cell wall metabolism” used here describes the process of degradation of existing wall
materials, synthesis and cross-linking of newly deposited wall materials, and also
changes in individual wall polysaccharides including modification of their chemical
structure and changes in the levels of individual polysaccharides.

The metabolism of cell walls is regulated during natural growth and
differentiation (Bolwell, 1988; Stolle-Smits et al., 1999; Bush et al., 2001; Obel et al.,
2002; Tokumoto et al., 2002). Among the factors that influence wall properties are
endogenous or externally applied plant growth regulators such as auxin (Rubery and
Northcote, 1970; Loescher and Nevins, 1972; Nishitani and Masuda, 1981; Masuda, 1985;
Talbott and Ray, 1992), gibberellic acid (Fry, 1980; Potter and Fry, 1994; Nishitani,

1995), ethylene (Abeles and F orrence, 1970; Hayashi and Maclachlan, 1984), abscisic
acid (Wu et al., 1994; Maheswari, 1999), jasmonic acid (Takahashi et al., 1995; Ueda et
31-, 1 996; Moore et al., 2003), cytokinin (Robertson et al., 1999), salicylic acid

(Ni Shimura et al., 2003), and brassinosteroid (Zurek and Clouse, 1994; Xu et al., 1995).
Environmental stresses that induce plant responses often also induce changes in cell wall
metabolism. Environmental signals that affect wall metabolism include light (Walton and
Ray, 1982; Miyamoto et al., 1997; Shieh etal., 1997), temperature (Kubacka-Zebalska
and Kacperska, 1999; Nakarnura et al., 2003), drought (Gershenzon, 1984; Maheswari,

1 999), osmotic stress (Boffey and Northcote, 1975; Singh et al., 1985; Talbott and Ray,

26

1992), mechanical stress (J affe et al., 1985), and wounding (Wu and Bradford, 2003).
Plant cell walls also participate actively in the pathogenesis process, including pathogen
recognition, signal transduction, and the subsequent plant response (Bolwell, 1986;
Esquerre-Tugaye et al., 2000). Fruit ripening is another process that involves significant
wall metabolism, mostly degradation of cell wall polysaccharides (Wakabayashi, 2000).

Many genes encoding enzymes involved in the degradation or reorganization of

cell wall polysaccharides have been identified. Consequently, antibodies have been
generated against the proteins, and probes have been created for the genes. These
reagents allowed the regulation of wall degradation and reorganization to be investigated.
Examples include enzymes involved in wall degradation during fruit ripening (Redondo-
Nevado et al., 2001; Tateishi et al., 2001; Tateishi et al., 2002; Chen and Paull, 2003),
Wall metabolism during auxin-induced rapid stem elongation (Inouhe and Nevins, 1991),
as Well as expansins and XTHs that are involved in wall reorganization (del Campillo and
Lewis, 1992; Zurek and Clouse, 1994; Reidy et al., 2001; Lee et al., 2003). However,
only a few genes involved in wall polysaccharide biosynthesis have been identified, and
mOSt of them have been identified very recently (Arioli et al., 1998; Edwards et al., 1999).
Without the genes as molecular tools, very few studies on the regulation of wall
bioSynthetic enzymes have been performed. We are, therefore, interested in the

identi fication of genes involved in XyG biosynthesis as the first step. In the second step
We Will generate necessary molecular tools such as gene—specific probes and enzyme-
Specific antibodies, and use them to explore the relationships between wall biosynthesis

and plant grth and development.

27

REFERENCES

Abeles, F. B. and Forrence, L. E. (1970). "Temporal and hormonal control of beta-1 ,3-
glucanase in Phaseolus vulgaris L." Plant Physiol. 45 (4): 395.

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

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

Augur, C., Yu, L., Sakai, K., Ogawa, T., Sinay, P., Darvill, A. G. and Albersheim, P.
( l 992). "Further studies of the ability of xyloglucan oligosaccharides to inhibit auxin-
stimulated growth." Plant Physiol. 99 (1): 180-185.

Bacic, A., Harris, P. J. and Stone, B. A. (1988). "Structure and function of plant cell
walls". The Biochemistry of Plants. J. Preiss, New York, Academic Press. 14: 297-371.

Boffey, S. A. and Northcote, D. H. (1975). "Pectin synthesis during wall regeneration of
Plasmolyzed tobacco leaf cells." Biochem. J. 150 (3): 433.

Bolwell, G. P. (1986). "Microsomal arabinosylation of polysaccharide and elicitor-

iIlduced carbohydrate-binding glycoprotein in french bean." Phytochem. 25 (8): 1807-
1 8 1 3 -

BOIWell, G. P. (1988). "Synthesis of cell wall components: Aspects of control."
Phytochem. 27 (5): 1235-1253.

Bo“ton, S., Leboeuf, E., Mouille, G., Leydecker, M.-T., Talbotec, J., Granier, F.,
Lah aye, M., Hofte, H. and Truong, H.-N. (2002). "QUASIMODOI encodes a putative
mernbrane-bound glycosyltransferase required for normal pectin synthesis and cell
adhesion in Arabidopsis." Plant Cell 14: 25 77-2590.

Bl‘accini, 1., Du, P. C. H., Michon, V., Goldberg, R., Clochard, M., Jarvis, M. C.,
Pang, Z. H. and Gage, D. A. (1995). "Structural analysis of cyclamen seed xyloglucan

011gosaccharides using cellulase digestion and spectroscopic methods." Carbohydr. Res.

276 ( 1): 167-181.

Brett, C. T. (1983). "Cell walls do not a prison make." New Scientist 99: 693-695.

arett, C. T. and Waldron, K W. (1996). Physiology and Biochemistry of Plant Cell
8118. Chapman & Hall.

28

-w

Bro
Buc

Em

Bu:
Bu

Bu

5

(a

Brisson, L. F., Tenhaken, R. and Lamb, C. (1994). "Function of oxidative cross-
]inking of cell wall structural proteins in plant disease resistance." Plant Cell 6: 1703-

1712.

Brown, J. A. and Fry, S. C. (1993). "Novel O-D-galacturonoyl esters in the pectic
polysaccharides of suspension-cultured plant cells." Plant Physiol. 103: 993-999.

Brown, J. A. and Fry, S. C. (1993). "The preparatioun and susceptibility to hydrolysis
of novel O-galacturonoyl derivatives of carbohydrates." Carbohydr. Res. 240: 95-106.

Buckeridge, M. S. and Dietrich, S. M. C. (1990). "Galactomannans from Brazilian
legume seeds." Revta. Brasi]. Bot. 13: 109-1 12.

Buckeridge, M. S., Pessoa, d. S. H. and Tine, M. A. S. (2000). "Mobilisation of storage
cell wall polysaccharides in seeds." Plant Physiol. Biochem. 38 (1-2): 141-156.

Buckeridge, M. S., Rocha, D. C., Reid, J. S. G. and Dietrich, S. M. C. (1992).
"Xyloglucan structure and post-germinative metabolism in seeds of Copaifera Iangsdorfii
from savanna and forest populations." Physiol. Plant. 86(1): 145-151.

Burke, D., Kaufman, P., McNeil, M. and Albersheim, P. (1974). "Structure of plant
cell walls. 6. Survey of walls of suspension-cultured monocots." Plant Physiol. 54 (1):
1 09- 1 15.

Burn, J., Hocart, C. H., Birch, R. J., Cork, A. C. and Williamson, R. E. (2002).
"Functional analysis of the cellulose synthase genes CesA 1, CesAZ, CesA3 in
Ar abidopsis." Plant Physio]. 129: 797-807.

Bush, M. s., Marry, M., Huxham, r. M., Jarvis, M. c. and McCann, M. C. (2001).

'S'Developmental regulation of pectic epitopes during potato tuberisation." Planta 213:
69-880.

Campbell, J. A., Davies, G. J., Bulone, V. and Henrissat, B. (1997). "A classification
of nucleotide-diphospho-sugar glycosyltransferases based on amino acid sequence
SlrI1ilarities."Biochem. J. 326: 929-939.

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

Carpita, N. C. and Gibeaut, D. M. (1993). "Structural models of primary cell walls in
0Vfr'ering plants: consistency of molecular structure with the physical properties of walls
dunng growth." Plant.]. 3: 1-30.

Ch ambat, G., Barnoud, F. and Joseleau, J. P. (1984). "Structure of the primary cell

:Valls of suspension cultured Rosa glauca cells 1. Polysaccharides associated with
‘31 lulose." Plant Physiol. 74 (3): 687-693.

29

Chen, N. J. and Paull, R. E. (2003). "Endoxylanase expressed durig papaya fruit
ripening: purification, cloning and characterization." F unct. Plant Biol. 30 (4): 433-441.

Cosgrove, D. J. (2000). "Loosening of plant cell walls by expansins." Nature 407: 321-
326 .

Crombie, H. J., Chengappa, S., Hellyer, A. and Reid, J. S. G. (1998). "A xyloglucan
oligosaccharide—active, transglycosylating beta-D-glucosidase from the cotyledons of
nasturtium (T ropaeolum majus L.) seedlings-purification, properties and characterization
ofa cDNA clone." PlantJ. 15(1): 27-38.

Cutillas, I. A. and Lorences, E. P. (1997). "Effect of xyloglucan oligosaccharides on
growth, viscoelastic properties, and long-term extension of pea shoots." Plant Physiol.

113 (1):103-109.

Cutler, S. and Somerville, C. R. (1997). "Cellulose synthesis: Cloning in silico." Curr.
Biol. 7: R108-R111.

Darvill, J. E., McNeil, M., Darvill, A. G. and Albersheim, P. (1980). "Structure of
plant cell walls XI. Glucuronoarabinoxylan, a second hemicellulose in primary cell walls
0f suspension-cultured sycamore cells." Plant Physio]. 66: 1135-1139.

del Campillo, E. and Lewis, L. N. (1992). "Occurrence of 9.5 cellulase and other
hYdrolases in flower reproductive organs undergoing major cell wall disruption." Plant
Physiol. 99 (1015-1020).

Desprez, T., Vernhettes, S., F agard, M., Refregier, G., Desnos, T., Aletti, E., Py, N.,
Pelletier, S. and Hofte, H. (2002). "Resistance against herbicide isoxaben and cellulose
deficiency caused by distinct mutations in same cellulose synthase isoforrn CESA6."

Plant Physiol. 128: 482-490.

Doblin, M. S., Melis, L. D., Newbigin, E., Bacic, A. and Read, S. M. (2001). "Pollen
t“hes of Nicotiana alata express two genes from different beta-glucan synthase families.‘
Plant Physiol. 125: 2040-2052.

EdvVards, M., Bowman, Y. J. L., Dea, I. C. M. and Reid, J. S. G. (1988). "A beta-D

galactosidase from Nasturtium T ropaeolum majus L. cotyledons purification properties

:gd demonstration that xyloglucan is the natural substrate." J. Biol. Chem. 263 (9): 4333-
3 7 .

Edwards, M., Dea, I. c. M., Bulpin, P. v. and Reid, J. s. G. (1985). "Xyloglucan

amYIoid mobilization in the cotyledons of T ropaeolum majus seeds following
germination." Planta 163(1): 133-140.

30

Edwards, M., Dea, I. C. M., Bulpin, P. V. and Reid, J. S. G. (1986). "Purification and
properties of a novel xyloglucan-specific endo-l-4-beta-D-glucanase from germinated
Nasturtium seeds T ropaeolum majus." J. Biol. Chem. 261 (20): 9489-9494.

Edwards, M., Scott, C., Gidley, M. J. and Reid, J. S. G. (1992). "Control of
mannose/galactose ratio during galactomannan formation in developing legume seeds."
Planta 187 (67-74).

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

galactosyltransferase of plant cell wall matrix polysaccharide biosynthesis." Plant J. 19:
691-697.

Emmerling, M. and Seitz, H. U. (1990). "Influence of a specific xyloglucan
nonasaccharide derived from cell walls of suspension-cultured cells of Daucus carota L.
on regenerating protoplasts." Planta 182: 174-180.

Esquerre—Tugaye, M.-T., Boudart, G. and Dumas, B. (2000). "Cell wall degrading
enzymes, inhibitory proteins, and oligosaccharides participate in the moleculare dialogue
between plants and pathogens." Plant Physiol. Biochem. 38 (1-2): 157-163.

Fagard, M., Desnos, T., Desprez, T., Goubet, F., Refregier, G., Mouille, G., McCann,
M., Rayon, C., Vernhettes, S. and Hofte, H. (2000). "PROC UST E 1 encodes a cellulose
synthase required for normal cell elongation specifically in roots and dark-grown
hypocotyls of Arabidopsis." Plant Cell 12: 2409-2432.

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-fucosyltransferase that catalyzes the last step of cell wall
xyloglucan biosynthesis in pea." J. Biol. Chem. 275 (20): 15082-15089.

Faik, A., Price, N. J., Raikhel, N. V. and Keegstra, K. (2002). "An Arabidopsis gene
encoding an alpha-xylosyltransferase involved in xyloglucan biosynthesis." Proc. Natl
Acad. Sci. USA 99 (l 1): 7797-7802.

Fanutti, C., Gidley, M. J. and Reid, J. S. G. (1991). "A xyloglucan oligosaccharide-
specific alpha-D-xylosidase or exooligoxyloglucan-alpha-xylohydrolase from germinated
Nasturtium T ropaeolum majus L. seeds: Purification properties and its interaction with a
xyloglucan-specific endo-1-4-beta-D glucanase and other hydrolases during storage
xyloglucan mobilization." Planta 184 (1): 137-147.

Fanutti, C., Gidley, M. J. and Reid, J. S. G. (1993). "Action of a pure xyloglucan endo-

transglycosilase (formerly called xyloglucan-specific endo-(1,4)-beta-D-glucanase) from
the cotyledons of germinated nasturtium seeds." Plant J. 3: 691-700.

31

Farkas, V. and Maclachlan, G. (1988). "Fucosylation of exogenous xyloglucans by pea
microsomal membranes." Arch. Biochem. Biophysi. 264 (1): 48-53.

Femenia, A., Rigby, N. M., Selvendran, R. R. and Waldron, K. W. (1999).
"Investigation of the occurrence of pectic-xylan-xyloglucan complexes in the cell walls
of cauliflower stem tissues." Carbohydrate Polymers 39 (2): 151-164.

F incher, G. B. and Stone, B. A. (1982). "Metabolism of noncellulosic polysaccharides".
Encyclopedia of Plant Physiology. W. Tanner and F. A. Loewus, Berlin, Springer-Verlag.
13B: 68-132.

Fry, S. C. (1980). "Gibberellin-controlled pectinic acid and protein secretion in growing-
cells." Phytochemistry 19 (5): 735-740.

Fry, S. C. (1989). "Cellulose, hemicellulose and auxin-stimulated growth: a possible
relationship." Physiol. Plant. 75 (4): 532-536.

Fry, S. C. (1989). "The structure and functions of xyloglucan." J. Exp. Bot. 40 (210): 1-
ll.

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 (3): 821-828.

Fry, S. C., York, W. S., Albersheim, P., Darvill, A., Hayashi, T., Joseleau, J. P., Kato,
Y., Perez, L. E., Maclachlan, G. A. and et a1. (1993). "An unambiguous nomenclature
for xyloglucan-derived oligosaccharides." Physiol. Plant. 89 (1): 1-3.

Gershenzon, J. (1984). Rec. Adv. Phytochem. 18: 274.

Gibeaut, D. M. and Carpita, N. C. (1994). "Biosynthesis of plant cell wall
polysaccharides." FASEB J. 8 (12): 904-915.

Grabber, J. H., Hatfield, R. D., Ralph, J., Zon, J. and Amrhein, N. (1995). "Ferulate
cross-linking in cell walls isolated from maize cell suspensions." Phytochemistry 40:
1077-1082.

Hayashi, T. (1989). "Xyloglucans in the primary cell wall." Ann. Rev. Plant Physiol.
Plant Mol. Bio]. 40: 139-168.

Hayashi, T. and Maclachlan, G. (1984). "Pea Pisum sativum cultivar Alaska xyloglucan
and cellulose 3. Metabolism during lateral expansion of pea epicotyl cells." Plant Physiol.

76 (3): 739-742.

Henrissat, B. and Davies, G. J. (2000). "Glycoside hydrolases and glycosyltransferases:
families, modules, and implications for genomics." Plant Physiol. 124: 1515-1519.

32

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 (1): 117-130.

Inouhe, M. and Nevins, D. J. (1991). "Inhibition of auxin-induced cell elongation of
maize coleoptiles by antibodies specific for cell wall glycanases." Plant Physiol. 96 (426-
431)

Israel, H. W., Wilson, R. G., Aist, J. R. and Kunoh, H. (1980). "Cell wall appositions
and plant disease resistance: Acoustic microscopy of Papillae that block fungal ingress."
Proc. Natl Acad. Sci. USA 77: 2046-2049.

Iwai, H., Masaoka, N., Ishii, T. and Satoh, S. (2002). "A pectin glucuronyltransferase
gene is essential for intercellular attachment in the plant meristem." Proc. Natl Acad. Sci.
USA 99 (25): 16319-16324.

Jaffe, M. J., Wakefield, A. H., Telewski, F., Gulley, E. and Biro, R. (1985).
"Computer-assisted image-analysis of plant-growth, thigmomorphogenesis, and
gravitropism." Plant Physiol. 77 (3): 722-730.

Joseleau, J. P. and Chambat, G. (1984). "Structure of the primary cell walls of
suspension cultured Rosa glauca cells 2. Multiple forms of xyloglucans." Plant Physiol.
74 (3): 694-700.

Kaczkowski, J. (2003). "Structure, function and metabolism of plant cell wall." Acta
Physio]. Plant. 25 (3): 287-305.

Keegstra, K., Talmadge, K. W., Bauer, W. D. and Albersheim, P. (1973). "The
structure of plant cell walls. 111. A model of the walls of suspension-cultured sycamore

cells bases on the interconnections of the macromolecular components." Plant Physiol.
51: 188-196.

Kim, J. B. and Carpita, N. C. (1992). "Changes in esterification of the uronic acid
groups of cell wall polysaccharides during elongation of maize coleoptiles." Plant
Physiol. 98: 646-653.

Knox, J. P., Linstead, P. J., King, J., Cooper, C. and Roberts, K. (1990). "Pectin
esterification is spatially regulated both within cell walls and between developing tissues

of root apices." Planta 181: 512-521.

Kooiman, P. (1961). "The constitution of T amarindus amyloid." Rev. T rav. C him. 80:
849-865.

33

92‘. I"
,H
,5.»-
v. j
I. .....

Kubacka-Zebalska, M. and Kacperska, A. (1999). "Low temperature-induced
modifications of cell wall content and polysaccharide composition in leaves of winter
oilseed rape (Brassica napus L. var. oleifera L.)." Plant Sci. 148: 59-67.

Lee, D.-K., Ahn, J. H., Song, S.-K., Choi, Y. D. and Lee, J. S. (2003). "Expression of
an expansin gene is correlated with root elongation in soybean." Plant Physiol. 131: 985-
997.

Loescher, W. H. and Nevins, D. J. (1972). "Auxin-induced changes in Avena coleoptile
cell wall composiiton." Plant Physiol. 50: 556-563.

Lorences, E. P., McDougall, G. J. and Fry, S. C. (1990). "Xyloglucan and cello-
oligosaccharides antagonists of the growth-promoting effect of hydrogen ion." Physiol.
Plant. 80(1): 109-113.

Madson, M., Dunand, C., Li, X., Verma, R., Vanzin, G. F., Caplan, J., Shoue, D. A.,
Carpita, N. C. and Reiter, W.-D. (2003). "The MUR3 gene of Arabidopsis encodes a
xyloglucan galactosyltransferase that is evolutionarily related to animal exostosins."
Plant Cell 15: 1662-1670.

Maheswari, M. (1999). "Effect of GA, ABA and water stress on leaf elongation and
XET activity in barley Hordeum vulgare L." Indian J. Exp. Biol. 37 (10): 1001-1004.

Maltby, D., Carpita, N. C., Montezinos, D., Kulow, C. and Delmer, D. P. (1979).
"Beta-1,3-glucan in developing cotton fibers. Structure, localization, and relationship of
synthesis to that of secondary wall cellulose." Plant Physiol. 63: 1158-1164.

Masuda, Y. (1985). "Cell-wall modifications during auxin-induced cell estension in
monocotyledonous and dicotyledonous plants." Biol. Plant. 27 (2-3): 119-124.

Mayer, A. M. and Poljakoff-Mayber, A. (1975). The Germination of Seeds. Pergamon
Press.

McCann, M. C. and Roberts, K. (1992). "Architecture of the primary cell wall". The
Cytoskeletal Basis of Plant Growth and Form. C. W. Lloyd, London, Academic: 109-129.

McDougaII, G. J. and Fry, S. C. (1988). "Inhibition of auxin-stimulated growth of pea
stem segments by a specific nonasaccharide of xyloglucan." Planta 175 (3): 412-416.

McDougall, G. J. and Fry, S. C. (1989). "Structure-activity relationships for xyloglucan
oligosaccharides with antiauxin activity." Plant Physiol. 89 (3): 883-887.

McDougall, G. J. and Fry, S. C. (1990). "Xyloglucan oligosaccharides promote growth

and activate cellulase evidence for a role of cellulase in cell expansion." Plant Physiol. 93
(3): 1042-1048.

34

McDougall, G. J. and Fry, S. C. (1991). "Xyloglucan nonasaccharide a naturally-
occurring oligosaccharin arises in-vivo by polysaccharide breakdown." J. Plant Physiol.
137(3): 332-336.

McDougall, G. J. and Fry, S. C. (1994). "Fucosylated xyloglucan in suspension-
cultured cells of the graminaceous monocotyledon F estuca arundinacea." J. Plant
Physio]. 143 (6): 591-595.

McNeil, M., Darvill, A. G., Fry, S. C. and Albersheim, P. (1984). "Structure and
function of the primary cell walls of plants." Ann. Rev. Biochem. 53: 652-683.

McQueen-Mason, S. (1997). "Plant cell wall and the control of growth." Biochemical
Society Transactions 25: 204-214.

Meier, H. and Reid, J. S. G. (1977). "Morphological aspects of the galactomannan
formation in the endosperm of T rigonella foenum-graecum L. (Leguminosae)." Planta
133: 243-248.

Meier, H. and Reid, J. S. G. (1982). "Reserve polysaccharides other than starch in
higher plants". Encyclopedia of Plant Physiology. F. A. Loewus and W. Tanner, Berlin,
Springer-Verlag. 13A: 418-471 .

Miyamoto, K., Mitani, Y., Soga, K., Ueda, J ., Wakabayashi, K., Hoson, T.,
Kamisaka, S. and Masuda, Y. (1997). "Modification of chemical properties of cell wall
polysaccharides in the inner tissues by white light in relation to the decrease in tissue
tension in Pisum sativum epicotyls." Physio]. Plant. 101 (1): 38-44.

Mohnen, D. (1999). "Biosynthesis of pectins and galactomannans". Carbohydrates and
Their Derivatives Including Tannins, Cellulose, and Related Lignins. B. M. Pinto,
Elsevier Science LTD, Oxford. 3: 497-527.

Moore, J. P., Paul, N. D., Whittaker, J. B. and Taylor, J. E. (2003). "Exogenous
jasmonic acid mimics herbivore-induced systemic increase in cell wall bound peroxidase
activity and reduction in leaf expansion." F unct. Eco]. 17(4): 549-554.

Mueller, S. C., Brown, R. M. J. and Scott, T. K. (1976). "Cellulosic microfibrils:
nascent stages of synthesis in a higher plant cell." Science 194: 949-951.

Nakamura, Y., Wakabayashi, K. and Hoson, T. (2003). "Temperature modulates the
cell wall mechanical properties of rice coleoptiles by altering the molecular mass of
hemicellulosic polysaccharides." Physio]. Plant. 118: 597-604.

Nishimura, M. T., Stein, M., Hon, B.-H., Vogel, J. P., Edwards, H. and Somerville, S.

C. (2003). "Loss of a callose synthase results in salicylic acid-dependent disease
resistance." Science 301: 969-972.

35

Nishitani, K. (1995). "Endo-xyloglucan transferase, a new class of transferase involved
in cell wall construction." J. Plant Res. 108 (1089): 137-148.

Nishitani, K. and Masuda, Y. (1981). "Auxin-induced changes in the cell wall structure:
changes in the sugar compositions, intrinsic viscosity and molecular weight distributions
of matrix polysaccharides of the epicotyl cell wall of Vigna angularis." Physiol. Plant. 52:
482-494.

Nishitani, K. and Tominaga, R. (1992). "Endoxyloglucan transferase, a novel class of
glycosyltransferase that catalyzes transfer of a segment of xyloglucan molecule to
another xyloglucan molecule." J. Biol. Chem. 267 (29): 21058-21064.

Northcote, D. H. (1985). "Cell organelles and their filnctio in biosynthesis of cell wall
components: control of wall assembly during differentiation". Biosynthesis and
Degradation of Wood Components. T. Higuchi, Orlando, Florida, Academic: 87-108.

Obel, N., Porchia, A. C. and Scheller, H. V. (2002). "Dynamic changes in cell wall
polysaccharides during wheat seedling development." Phytochem. 60: 603-610.

Ozeretskovskaya, O. L. and Romenskaya, I. G. (1996). "Oligosaccharins as regulatory
molecules of plants." Russian J. Plant. Physiol. 43 (3): 648-655.

Parr, A. J., Waldron, K. W., N g, A. and Parker, M. L. (1996). "The wall-bound
phenolics of Chinese water chestnut (Eleocharis dulcis)." J. Sci Food Agric 71: 501-507.

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 (5-6): 842-
850.

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.

Perrin, R., Wilkerson, C. and Keegstra, K. (2001). "Golgi enzymes that synthesize
plant cell wall polysaccharides: finding and evaluating candidates in the genome era."
Plant Mol. Bio]. 47: 115-130.

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 (5422): 1976-1979.

Potter, 1. and Fry, S. C. (1994). "Changes in xyloglucan endotransglycosylase (XET)
activity during hormone-induced growth in lettuce and cucumber hypocotyls and spinach
cell suspension cultures." J. Exp. Bot. 45 (280 SPEC. ISSUE): 1703-1710.

36

Preston, R. D. (1974). "Plant cell walls". Dynamic Aspects of Plant Ultrastructure. A. W.
Robards, Maidenhead, McGraw-Hill: 256-309.

Rayle, D. L. and Cleland, R. E. (1992). "The acid growth theory of auxin-induced cell
elongation is alive and well." Plant Physio]. 99: 1271-1274.

Redondo-Nevado, J., Moyano, E., Medina-Escobar, N., Caballero, J. L. and Munoz-
Blanco, J. (2001). "A fruit-specific and developmentally regulated
endopolygalacturonase gene from strawberry (F ragaria x ananassa cv. Chandler)" J.
Exp. Bot. 52 (362): 1941-1945.

Reid, J. S. G. and Meier, H. (1970). "Formation of reserve galactomannan in the seeds
of T rigonella foenum-graecum." Phytochem. 9: 513-520.

Reidy, B., McQueen, M. S., Nosberger, J. and Fleming, A. (2001). "Differential
expression of alpha- and beta-expansin genes in the elongating leaf of F estuca pratensis."
Plant Mol. Bio]. 46 (4): 491-504.

Reis, D., Vian, B., Darzens, D. and Roland, J. C. (1987). "Sequential patterns of
intramural digestion of galactoxyloglucan in tamarind seedlings." Planta 170(1): 60-73.

Reiter, W. D., Chapple, C. and Somerville, C. R. (1997). "Mutants of Arabidopsis
thaliana with altered cell wall polysaccharide composition." Plant J. 12 (2): 335-345.

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

Richmond, T. A. and Somerville, C. R. (2001). "Integrative approaches to determining
Csl fimction." Plant Mol. Bio]. 47: 131-143.

Robertson, D., Wojtaszek, P. and Bolwell, G. P. (1999). "Stimulation of cell wall
biosynthesis and structural changes in response to cytokinin- and elicitor-treatments of
suspension-cultured Phaseolus vulgaris cells." Plant Physiol. Biochem. 37 (7-8): 611-622.

Rose, J. K. C., Braam, J., Fry, S. C. and Nishitani, K. (2002). "The XTH family of
enzymes involved in xyloglucan endotransglucosylation and endohydrolysis: Current
perspectives and a new unifying nomenclature." Plant Cell Physiol. 43 (12): 1421-1435.

Rubery, P. H. and Northcote, D. H. (1970). "Effect of Auxin (2,4-
Dichlorophenoxyacetic acid) on synthesis of cell wall polysaccharides in cultured
sycamore cells." Biochim. Biophys. Acta. 222 (1): 95.

Sal'ria, R., Wagner, T. A., O, N. M. A., Faik, A., Wilkerson, C. G., Keegstra, K. and

Raikhel, N. V. (2001). "Characterization of a family of Arabidopsis genes related to
xyloglucan fucosyltransferasel." Plant Physio]. 127 (4): 1595-1606.

37

Saxena, I. M. and Brown, R. M. J. (2000). "Cellulose synthases and related enzymes."
Curr. Opin. Plant Biol. 3: 523-531.

Saxena, I. M., Brown, R. M. J., Fevre, M., Geremia, R. A. and Henrissat, B. (1995).
"Multidomain architecture of beta-glycosyltransferases: implications for mechanism of
action." J. Bacteriol. 177: 1419-1424.

Scheible, W.-R., Eshed, R., Richmond, T. A., Delmer, D. and Somerville, C. R.
(2001). "Modifications of cellulose synthase confer resistance to isoxaben and
thiazolidinone herbicides in Arabidopsis Ixr] mutants." Proc. Natl Acad. Sci. USA 98
(18): 10079-10084.

Selvendran, R. R. (1985). "Developments in the chemistry and biochemistry of pectic
and hemicellulosic polymers." J. Cell Sci. Supp]. 2: 51-88.

Selvendran, R. R., Stevens, B. J. H. and 0' Neill, M. A. (1985). "Developments in the
isolation and analysis of cell walls from edible plants". Biochemistry of plant cell walls.
C. T. Brett and J. R. Hillman, Cambridge University Press: 39-78.

Shieh, M. W., Shi, J. and Cosgrove, D. J. (1997). "Developmental, hormonal and light
regulation of the transcript for the cell wall loosening protein expansin." Plant Physiol.
114(3): 341. Suppl. S.

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

Singh, N. K., Handa, A. K., Hasegawa, P. M. and Bressan, R. A. (1985). "Proteins
associated with adaptation of cultured tobacco cells to NaCl." Plant Physio]. 79 (1): 126-
137.

Stacey, G., Greshoff, P. M. and Keen, N. T. (1992). "Friends and foes: new insights
into plant-microbe interactions." Plant Cell 4: 1173-1179.

Stolle-Smits, T., Beekhuizen, J. G., Kok, M. T. C., Pijnenburg, M., Recourt, K.,
Derksen, J. and Voragen, A. G. J. (1999). "Changes in cell wall polysaccharides of
green bean pods during development." Plant Physiol. 121: 363-372.

Takahashi, K., Fujino, K., Kikuta, Y. and Koda, Y. (1995). "Involvement of the
accumulation of sucrose and the synthesis of cell-wall polysaccharides in the expansion
of potato cells in response to jasmonic acid." Plant Sci. 111 (1): 11-18.

Talbott, L. D. and Ray, P. M. (1992). "Changes in molecular size of previously

deposited and newly synthesized pea cell wall matrix polysaccharides effects of auxin
and turgor." Plant Physio]. 98 (1): 369-379.

38

Tateishi, A., Inoue, H., Shiba, H. and Yamaki, S. (2001). "Molecular cloning of beta-
galactosidase from Japanese pear (Pyrus pyrifolia) and its gene expression with fruit
ripening." Plant Cell Physio]. 42 (5): 492-498.

Tateishi, A., Inoue, H. and Yamaki, S. (2002). "Cloning and expression of beta-
galactosidase cDNA related to softening of avocado (Persea americana) finit." J. Jpn.
Soc. Hortic. Sci. 71 (1): 48-55.

Taylor, N. G., Laurie, S. and Truner, S. R. (2000). "Multiple cellulose synthase

catalytic subunits are required for cellulose synthesis in Arabidopsis." Plant Cell 12:
2529-2539.

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.

Thomas, J. R., McNeil, M., Darvill, A. G. and Albersheim, P. (1987). "Structure of
plant cell walls XIX. Isolation and characterization of wall polysaccharides from
suspension-cultured Douglas fir cells." Plant Physiol. 83 (3): 659-671.

Thompson, J. E. and Fry, S. C. (2000). "Evidence for covalent linkage between
xyloglucan and acidic pectins in suspension-cultured rose cells." Planta 211 (2): 275-286.

Tine, M. A. S., Cortelazzo, A. L. and Buckeridge, M. S. (2000). "Xyloglucan
mobilisation in cotyledons of developing plantlets of Hymenaea courbari] L.
(Leguminosae-Caesalpinoideae)." Plant Sci. 154 (2): 117-126.

Tokumoto, H., Wakabayashi, K., Kamisaka, S. and Hoson, T. (2002). "Changes in the
sugar composition and molecular mass distribution of matrix polysaccharides during
cotton fiber development." Plant Cell Physio]. 43 (4): 411-418.

Ueda, J., Miyamoto, K. and Hashimoto, M. (1996). "Jasmonates promote abscission in
bean petiole explants: Its relationship to the metabolism of cell wall polysaccharides and
cellulase actgivity." J. Plant Growth Regal. 15 (4): 189-195.

Vanzin, G. F., Madson, M., Carpita, N. C., Raikhel, N. V., Keegstra, K. and Reiter,
W. D. (2002). "The mur2 mutant of Arabidopsis thaliana lacks fucosylated xyloglucan
because of a lesion in fucosyltransferase AtFUTl." Proc. Natl Acad. Sci. USA 99 (5):
3340-3345.

Vierhuis, E., York, W. S., Kolli, V. S. K., Vincken, J. P., Schols, H. A., Van, A. G. J.
W. M. and Voragen, A. G. J. (2001). "Structural analyses of two arabinose containing
oligosaccharides derived from olive fruit xyloglucan: XXSG and XLSG." Carbohydr.
Res. 332 (3): 285-297.

39

Vincken, J. P., Wijsman, A. J. M., Beldman, G., Niessen, W. M. A. and Voragen, A.
G. J. (1996). "Potato xyloglucan is built from XXGG-type subunits." Carbohydr. Res.
288 (0): 219-232.

Vincken, J. P., York, W. S., Beldman, G. and Voragen, A. G. J. (1997). "Two general
branching patterns of xyloglucan, XXXG and XXGG." Plant Physiol. 114 (1): 9-13.

Wakabayashi, K. (2000). "Changes in cell wall polysaccharides during fruit ripening." J.
ofPlant Res. 113 (1111): 231-237.

Walton, J. D. and Ray, P. M. (1982). "Inhibition by light of growth and Golgi-localized
glucan synthetase in the maize mesocotyl." Planta 156: 302-308.

Wang, Q., Ellis, P. R., Ross, M. S. B. and Reid, J. S. G. (1996). "A new polysaccharide
from a traditional Nigerian plant food: Detarium senegalense Gmelin." Carbohydr. Res.
284 (2): 229-239.

Warneck, H. and Seitz, H. U. (1993). "Inhibition of gibberellic acid-induced elongation-
growth of pea epicotyls by xyloglucan oligosaccharides." J. Exp. Bot. 44 (264): 1105-
1109.

Warneck, H. M., Fulton, D. C., Seitz, H. U. and Fry, S. C. (1998). "Transport,
degradation and cell wall-integration of XXFGol, a growth-regulating nonasaccharide of
xyloglucan, in pea stems." Planta 204 (1): 78-85.

Watt, D. K., Brasch, D. J., Larsen, D. S. and Melton, L. D. (1999). "Isolation,
characterisation, and NMR study of xyloglucan from enzymatically depectinised and
non-depectinised apple pomace." Carbohydr. Polymers 39 (2): 165-180.

Williamson, R. E., Burn, J. E. and Hocart, C. H. (2002). "Towards mechanism of
cellulose synthesis." Trends Plant Sci. 7 (10): 461-467.

Wu, C. T. and Bradford, K. J. (2003). "Class I chitinase and beta-1,3-glucanase are
differentially regulated by wounding, methyl jasmonate, ethylene, and gibberellin in
tomato seeds and leaves." Plant Physiol. 133 (1): 263-273.

Wu, Y., Spollen, W. G., Sharp, R. E., Hetherington, P. R. and Fry, S. C. (1994).
"Root growth maintenance at low water potentials: Increased activity of xyloglucan

endotransglycosylase and its possible regulation by abscisic acid." Plant Physio]. 106 (2):
607-615.

Xu, W., Purugganan, M. M., Polisensky, D. H., Antosiewicz, D. M., Fry, S. C. and
Braam, J. (1995). "Arabidopsis TCH4, regulated by hormones and the environment,
encodes a xyloglucan endotransglycosylase." Plant Cell 7 (10): 1555-1567.

40

York, W. S., Darvill, A. G. and Albersheim, P. (1984). "Inhibition of 2,4-D stimulated
elongation of pea stem segments by a xyloglucan Oligosaccharide." Plant Physiol. 75 (2):
295-297.

York, W. S., Kolli, V. S. K., Orlando, R., Albersheim, P. and Darvill, A. G. (1996).
"The structures of arabinoxyloglucans produced by solanaceous plants." C arbohydr. Res.
285(0): 99-128.

York, W. S., Van Halbeek, H., Darvill, A. G. and Albersheim, P. (1990). "Structural
analysis of xyloglucan oligosaccharides by proton NMR spectroscopy and fast-atom-
bombardment mass spectrometry." Carbohydr. Res. 200: 9-32.

Zhong, R., Morrison 111, W. H., Freshour, G. D., Hahn, M. G. and Ye, Z.-H. (2003).
"Expression of a mutant form of cellulose synthase AtCesA7 causes dominant negative
effect on cellulose biosynthesis." Plant Physio]. 132: 786-795.

Zurek, D. M. and Clouse, S. D. (1994). "Molecular cloning and characterization of a

brassinosteroid-regulated gene from elongating soybean (Glycine max L.) epicotyls."
Plant Physio]. 104 (1): 161-170.

41

CHAPTER 2

Cloning and Expression Studies of Xyloglucan Fucosyltransferase

Part of the work presented in this chapter has been published:

Perrin, R. M., DeRocher, A. E., Bar-Peled, M., Zeng, W., Norambuena, L., Orellana, A.,

Raikhel, N. V. and Keegstra, K. (1999). Science 284 (5422): 1976-1979.

F aik, A., Bar-Peled, M., DeRocher, A. E., Zeng, W., Perrin, R. M., Wilkerson, C.,

Raikhel, N. v. and Keegstra, K (2000). J. Biol. Chem. 275 (20): 15082-15089.

42

ABSTRACT

XyG is the major hemicellulose in the primary cell wall of dicot and
nongraminaceous monocot plants. However, the genes involved in its biosynthesis are
still largely unknown. In our previous effort to identify genes involved in XyG
biosynthesis, XyG fucosyltransferase was purified from etiolated pea epicotyls, and a
partial-length cDNA clone was identified from Arabidopsis using the protein sequence
information obtained. The first part of this chapter describes the identification of a full-
length cDNA encoding the Arabidopsis XyG fucosyltransferase, using sequence
information obtained from a genomic BAC clone, T18E12. Part of the polypeptide
encoded by this cDNA was expressed in E. coli cells, and a polyclonal antibody against it
was generated. This antibody was used to immunoprecipitate XyG fucosyltransferase
activity from detergent-solubilized Arabidopsis membrane proteins. Heterologous
expression of this cDNA in mammalian COS7 cells or tobacco BY2 cells produced XyG
fucosyltransferase activity, thereby confirming the identity of this cDNA clone.
Furthermore, the tryptic peptide sequence information obtained from the purified pea
XyG fucosyltransferase protein led to the isolation of a full-length pea cDNA encoding
this enzyme. Sequence comparison revealed that pea XyG fucosyltransferase and its
Arabidopsis equivalent are very close ortholo gs. The second part of this chapter describes
the rapid elongation of the intemodes of etiolated pea seedlings at early growth stages
and the expression of pea XyG fucosyltransferase during the growth process. The data
presented here have established a correlation between the intemode elongation rate and

XyG fiicosyltransferase expression.

43

 

 
 

, .

‘W

1.
It.

Introduction

XyG has the same B-(l,4)-glucan backbone as cellulose, but with side chain
decorations of L-Fuc-(l,2)-D-Gal-(1,2)-D-Xy1-(l,6); D-Gal-(1,2)-D-Xyl-(1,6); and D-
Xyl-(l,6) (Figure 1.1) (Kato and Matsuda, 1980; Fry, 1989). The terminal fucose residue
is thought to have an important function in stabilizing a planar configuration of the XyG
backbone that is essential for cellulose binding as determined by molecular modeling and
in vitro binding studies (Hayashi et al., 1994; Cosgrove, 1997; Levy et al., 1997),
although contradictory evidence has also been presented (Whitney et al., 1995). The
terminal fucose has also been suggested to be involved in signal transduction during plant
growth and development. Some oligosaccharides produced by partial digestion of XyG
influence the growth rate of intact plant tissue at very low concentrations (Darvill et al.,
1992; Aldington and Fry, 1993). One Oligosaccharide, XXF G, is able to antagonize the
growth promoting effect of auxin at a concentration of 10'9 M (York et al., 1984), but
loses its effectiveness when the terminal fucose is absent (McDougall and Fry, 1989).
Therefore the fucose was hypothesized to determine the ability of XyG oligosaccharides
to act as signaling molecules in vivo (McDougall and Fry, 1989).

GDP-fucose2XyG 1,2-0L-L-fucosyltransferase is the enzyme responsible for the
attachment of fucose onto galactose of the XyG sidechains. Given the importance that the
terminal fucose might have, we sought to identify the gene encoding the enzyme that
adds it to XyG. A biochemical approach was used to purify the enzyme while monitoring
its activity using an in vitro assay developed by Maclachlan’s group (Farkas and
Maclachlan, 1988). This assay uses radiolabeled GDP-fucose as donor and tamarind seed

XG that naturally lacks the terminal fucose as acceptor (Farkas and Maclachlan, 1988;

44

Hanna et al., 1991; Desveaux et al., 1998). The enzyme activity of XyG
fucosyltransferase has been partially purified from pea epicotyl microsomal membranes
(Hanna et al., 1991; Desveaux et al., 1998). The activity co-migrated with a 150 kD
peptide observed on a nondenaturing gel (Hanna etal., 1991 ). However, previous
attempts to purify the protein to homogeneity or to clone the XyG fucosyltransferase
gene had not been successful. Neither had any other genes encoding enzymes involved in
XyG biosynthesis been cloned.

Etiolated pea epicotyls contain significant amount of XyG fucosyltransferase
activity (Farkas and Maclachlan, 1988; Hanna et al., 1991; Maclachlan et al., 1992).
Consequently this tissue was used by Drs. Amy E. DeRocher and Maor Bar-Peled in our
group to successfully purify this enzyme, utilizing a combination of GDP affinity and
anion exchange chromatography (Perrin et al., 1999). Information from the purified
protein led to the identification of a partial cDNA clone from Arabidopsis. The first half
of this chapter describes the cloning of the full-length coding region of the Arabidopsis
XyG fucosyltransferase(AtFut1) and its pea equivalent (PsFutl) based on peptide
sequence information obtained from the purified enzyme.

For the work on the identification of Arabidopsis XyG fucosyltransferase (Perrin
et al., 1999), my role was the isolation of a full-length coding region of AtFutl, and the
generation and characterization of a polyclonal antibody against E. coli-expressed AtFUT
peptide. For the work on pea (F aik et al., 2000), my role was the isolation of a full-length
coding region of PsFutl , as well a genomic clone with the intron and the untranslated

regions (UTR).

45

With the identification of XyG fucosyltransferase genes (Perrin et al., 1999; Faik
et al., 2000), it is feasible to study for the first time the relationship between plant growth
and development and XyG biosynthesis at the molecular level. At the time this work was
done, XyG fucosyltransferase was the only molecular tool available with its gene cloned
and with an in vitro enzyme assay.

The process of cell growth often involves up to a thousand fold increase in cell
volume (Pennell, 1998). The cell enlargement is brought about by a process of relaxation
of existing cell walls, the synthesis and deposition of new cell wall polysaccharides, and
the cross linking of new and old wall materials. Therefore enzymes involved in these
three aspects of cell wall growth, wall degradation, wall modification, and wall
biosynthesis, are all thought to be delicately regulated during plant growth.

XyG plays a key role in controlling cell wall expansion by providing mechanical
strength and physical integrity to cell walls via cross-linking the cellulose microfibrils
(Cosgrove, 2000). As the major hemicellulose of the primary cell wall for dicot and
nongraminaceous plants, XyG undergoes rapid synthesis and also turnover during plant
grth and development (Pauly et al., 2001). An a-xylosidase that degrades XyG
oligosaccharides isolated from Arabidopsis showed higher enzyme activity and
messenger RNA abundance in younger leaves, where XyG turnover is higher (Sampedro
et al., 2001). XyG endotransglycosylase (XET) and expansin are two additional types of
wall modifying proteins. XET cleaves XyG mid-chain and then transfers the cleaved
XyG molecule to the non-reducing end of another XyG chain (Sulova et al., 1998; Steele
and Fry, 1999). Expansins are thought to loosen the polysaccharide associations, mainly

cellulose — XyG association or XyG — XyG association, to allow slippage or the

46

If}

...-

 

x.‘ "
J-
.b\
'1
-.--
’I
&.\-.
..
e
H' '3
1.1.l¥

 

 

 

 

 

deposition of new wall materials (Cosgrove, 2000). More XET mRNAs were detected in
elongating stems, during phytohonnone-induced plant growth and during fruit softening
(Steele et al., 2001). A root-specific soybean expansin, GmEXPl, also shows higher
transcript abundance during root rapid elongation (Lee et al., 2003). Even though there
are clear indications that enzyme activities involved in XyG biosynthesis (XyG synthase
and XyG fucosyltransferase) are regulated during plant development and primary wall
biosynthesis (Hayashi and Matsuda, 1981; Pauly et al., 2001), there has never been any
molecular studies on the regulation of gene expression, because none of the genes
involved in XyG biosynthesis had been cloned. I tried to address this question in the
second part of this chapter utilizing the molecular tool of pea XyG fucosyltransferase.
Many scientists have used etiolated pea as a model system to study the expression
of enzymes involved in cell wall biosynthesis, including the development of enzyme
activity assays and the partial purification of XyG fucosyltransferase (Camirand and
Maclachlan, 1986; F arkas and Maclachlan, 1988; Hanna et al., 1991; Maclachlan et al.,
1992). Etiolated pea seedlings have a relatively simple morphology and large amount of
rapidly growing tissue. At a young stage of growth etiolated pea seedlings contain both
fast growing and non-growing tissues. Seedling growth occurs in a visible manner. Its
X yG oligosaccharides were also well characterized (Hayashi and Maclachlan, 1984;
Guillen et al., 1995). This plant has therefore become an attractive model for both
biochemists and molecular biologist (Labavitch and Ray, 1974; Talbott and Ray, 1992),
and In the second half of this chapter, etiolated pea seedlings were used to correlate the

expression patterns of the XyG fucosyltransferase with plant growth stages.

47

MATERIALS AND METHOD
Plant material
Pisum sativum cv. Alaska seeds were sterilized with 10% bleach at room
temperature for 15 min and rinsed 4 to 5 times with running water before being imbibed
overnight (~12 hr) with aeration. Seeds were planted in coarse vermiculite and grown in
the dark at 25°C. Intemode length was measured in the dark under green safe light with a
ruler. Pea tissues were also collected in the dark under safe light and stored on ice until

used.

Preparation of pea microsomes

Pea seedling samples were cut into small pieces and homogenized in extraction
buffer (40 mM Hepes'KOH, pH6.5, 0.2 M sucrose, 1 mM dithiothreitol [DTT], 1 mM
MgC12, 0.2 mM phenylmethanesulfonyl fluoride [PMSF]) with a Polytron for 15 seconds
four times on ice. The homogenate was filtered through a double layer of miracloth and
centrifuged at 10,000x g for 10 minutes at 4°C. The supernatant was collected as the 810
fraction and centrifuged again at 100,000x g for 1 hour at 4°C. The pellet, labeled P100,
was resuspended in extraction buffer using an ice-cold Dounce homogenizer. The protein
concentration of P100 microsomes was determined with the BCA protein assay reagent

(PIERCE, Rockford, Illinois).

Oligonucleotide prim er synthesis and PCR product sequencing
All oligonucleotide primers were synthesized at the Macromolecular Sequencing

and Synthesis Facility in the Department of Biochemistry, Michigan State University. All

48

PCR reactions were carried out in a Peltier thermal cycler-200 (MJ Research). PCR
products were all purified from low melting point agarose gel and cloned into pGEM-T
Easy vector (Promega, Madison, WI) for sequencing. Sequence alignments were carried
out using the software package DNASTAR version 4.0 (Madison, WI) and GCG
programs (Madison, WI). DNA sequencing was done either at the Genomics Technology
Support Facility at Michigan State University or the W. M. Keck Biotechnology

Resource Lab, Yale University.

RNA isolation, gel electrophoresis and Northern blotting

Total RNA from pea seedlings was isolated according to a standard protocol
(Houdebine and Puissant, 1990). Messenger RNA was isolated from total RNA in the
same way as for mRNAs used for RACE reactions. RNA separation by formaldehyde gel
electrophoresis, RNA blotting to N+ Hybrid-bond Nylon membrane, and Northern
hybridization all followed standard protocols (Sambrook et al., 1989) except that the
hybridization buffer used was 0.25 M NazHPO4, 7% sodium dodecyl sulfate (SDS), 1

mM EDTA, pH 7.6.

Preparation of E. coli competent cells

A single colony of DHSCX. or BL21 (pLysS) was picked from a fresh LB plate and
inoculated in 10 m1 LB for overnight culturing at 37°C. On the second day, 3 m1 of the
mini culture was inoculated into 300 ml of new LB medium (at 1:100 dilution) and
allowed to grow at 30°C with brisk shaking (250mm) until the OD600 reached 0.3 to

0.6. The bacterial culture was chilled on ice for 10 to 15 minutes and all following steps

49

were carried out on ice. Cells were collected by centrifugation at 1,000x g for 15 minutes
at 4°C and resuspended by swirling in 100 m1 (1/3 of the starting culture volume) of RFI
buffer (100 mM RbCl, 50 mM MnClg, 30 mM KOAc-pH7.5, 10 mM CaClz, 15%
glycerol, pH 6.0 with acetic acid, filter sterilized) that was pre-chilled on ice. The cell
suspension in RFl buffer was incubated on ice for 0.5 to 2 hours. Cells were then pelleted
again under the same conditions and resuspended again in 24 ml (1/ 12.5 to 1/20 of the
starting culture volume, depending on the OD600 value when the cells were collected) of
pre-chilled RF2 buffer (10 mM MOPS-pH6.8, 10 mM RbCl, 75 mM CaClz and 15%
glycerol, filter sterilized) by swirling and incubated on ice for 15 minutes up to 12 hours.
Cells were then divided into aliquots and placed into pre-chilled eppendorf tubes on ice in
the cold room and quickly frozen by dropping them into liquid nitrogen. Competent cells

were stored at -80°C for up to four years.

5 ’and 3 ’ RACE

Poly(A)+ RNA was isolated from total RNA using the polyATtract mRNA
Isolation System IV (Promega). RACE reactions were carried out using the 5'-RACE
System, version 2.0 (Life Technologies, Inc.) according to the manufacturer's instructions.
For 5'-RACE, kee258 (5'-ACTCCCGGGTCAACTAGTAGAA-3') was used for cDNA
synthesis, and kee305 (5'-GGCTGACTGATACCTGCTGAGACAAGATTTTT-3') or
kee304 (5’-GTTTGTGTCGGGCTTCATATT-3’) were used as the nested primers. For
3'-RACE, poly (dT) primer was used for cDNA synthesis, and kee302 (5'-
CCAAAAAGCCTGGGCAGAAAT-3') was the nested primer. For both reactions 1.5 pg

mRNA were used for reverse transcription.

50

Construction of plasmid pBS-A tF utl

Oligonucleotide primers kee213 (5’-
GATCCAGAGTCTCCCTCAAGAGCAATCCATGGATCAG-3’) and kee214 (5’-
AATTCTGATCCATGGATTGCTCTTGAGGGAGACTCTG-3’) were designed to
contain the missing sequence of the partial Arabidopsis XyG fucosyltransferase cDNA
clone 6-1B1 (Perrin et al., 1999) and to contain restriction sites BamHI and EcoRI at their
ends and a Ncol site in the middle (Figure 2.1 A). About 2 pg of each primer were
vacuum dried and redissolved separately in polynucleotide kinase (PNK) buffer
(Boehringer Mannheim) containing 1 mM ATP in a total volume of 20 ul each. Primers
were phosphorylated at 37°C for 15 minutes by 10 units of PNK followed by 15 minutes
incubation at 70°C to inactivate the enzyme. Phosphorylated primers were mixed
together and denatured at 95°C for 30 seconds, then transferred to a water bath at 90°C.
The primer mixture in the water bath then was allowed to air cool to 30°C on a lab bench.
The annealed oligonucleotide primers were inserted into vector pBlueScript SK (+) via
restriction enzymes BamHI and EcoRI. A positive clone was selected by restriction
digestion of plasmids with NcoI and named as pBS-AtFutl-Ncol. Its identity was further
confirmed by sequencing.

Plasmid DNA of pBS-AtFutl-Ncol was linearized with EcoRI digestion, de-
phosphorylated with calf intestinal alkaline phosphatase (CIP) and purified via low
melting point agarose gel. The insert released from cDNA clone 6-1B1 by EcoRI
digestion was purified the same way and ligated into the treated pBS-AtFutl-NcoI vector,

giving rise to the construct of pBS-AtFutl containing the full coding region for AtFutl.

51

Construct pBS-AtFutl was screened by double digestion with NcoI and BglII. After
digestion the construct with insert in the right direction gave rise to two fragments of 1.2
kb and 3.52 kb while constructs containing the insert in the opposite direction gave rise to

two fragments of 0.56 kb and 4.16 kb.

Construction of expression plasmids pET-AtFut1-1.2 and pE T -AtF at] -2.3

DNA fragments of AtF ut1-1 .2 and AtFut1-2.3 were amplified by PCR reactions
using primers kee176 (5 ’-AACTACTACATATGAAGTATCTCAGCT-3’, NdeI site
underlined) and keel 78 (5’-CACAAGAATTCATACTAGCTTAAGTC-3’, EcoRI site
underlined), or primers keel 77 (5’-TAAATTGGCATATGGGTTTCGCCGAA-3’, NdeI
site underlined) and keel 78, and plasmid DNA of pBS-AtFutl as template. Gel purified
PCR products were inserted into expression vector plasmid pET28a through restriction
sites NdeI and EcoRI to give rise to the expression plasmids pET-AtFut1-1.2 and pET-
AtFut1-2.3. Positive clones were screened by a combination of digestions using BglII
alone, XhoI alone, NdeI plus EcoRI, or KpnI plus SacI. Positive clones selected were

sequenced to confirm their identities.

Expression of pE T -AtF at] -I .2 and pE T -A tF ml -2.3

Expression plasmids pET-AtFut1-1.2, pET-AtFut1-2.3, and vector control
pETZSa were transformed into E. coli expression strain BL21 (DE3)-pLysS. One colony
was picked up for each construct for protein expression and inoculated in 10 ml LB with
50 pg/ml kanamycin for overnight culture at 37°C. On the second day new cultures of

200 ml were inoculated with 2 ml overnight culture. Cells were induced with 1 mM IPTG

52

at 37°C when their OD600 value reached 0.4 to 0.6. One milliliter of cell culture was
collected before induction and every hour after induction for 4 hr. After measuring the
OD600 value, cells were collected and resuspended in adequate amount of 2x SDS-
polyacrylarnide gel electrophoresis (PAGE) loading buffer. The volumes of the loading
buffer used were adjusted according to the OD600 values of individual cell culture so that
the final concentration of the cells roughly equaled to each other. Total proteins from
equal amount of cells collected from different induction time points were resolved by
SDS-PAGE and visualized by Commassie staining or western blotting with antibodies
against the 6xHis tag. SDS-PAGE loading buffer contained 62.5 mM Tris-HCl (pH6.8),

2% SDS, 10% glycerol, 5% B-mercaptoethanol and 0.01% bromphenol blue.

Solubility test of peptide AtF U T I -2.3

After being induced for 4 hr, a 2 ml culture of cells containing pET-AtFut1-2.3
was collected and washed once with PBS buffer and resuspended in 200 pl PBS buffer.
Afier sonication at a 10 seconds duration for a total of l min in an output of 50% on ice,
an aliquot of 20 ul was taken out as total protein. The sonicated solution was spun at 4°C
for 5 minutes at 10,000x g. The pellet was resuspended in 180 pl PBS buffer by
sonication and designated as the P10 fraction. The supernatant was saved as the 810
fraction. After removing an aliquot of 20 ul, the S10 fraction was centrifiiged again at
100,000x g for 30 minutes at 4°C. This time the supernatant was saved as 8100, and the
pellet was resuspended in 160 pl PBS buffer by sonication and designated as P100.

Aliquots of 20 ul from each fraction was used for SDS-PAGE.

53

h

Purification of peptide AtF U T 1 -2.3

A colony that showed high expression of peptide AtFUT1-2.3 was inoculated into
1 liter of LB medium at 37°C and induced with 1 mM IPTG when the OD600 reached
0.47. After 5 hr, the cells were collected and washed once with SOmM Tris-HCI (pH7.8)
buffer containing 5 mM MgC12, and resuspended in 20 ml of the same buffer. Cells were
lysed by adding lysozyme to a final concentration of 100 ug/ml and DNase I to a final
concentration of 10 U/ml followed by sonication on ice for 1 min at 50% output. The
protein inclusion bodies were pelleted at 10,000x g for 10 min and washed four times
with 1% Triton X-100 and four times with H20. This protein pellet was then solubilized
in 12 ml of 6 M guanidine hydrochloride buffer and precipitated with trichloroacetic acid
(TCA). The pellet was resuspended in 2 m1 of 6 M guanidine hydrochloride and analyzed

on SDS-PAGE with bovine serum albumin (BSA) as standard.

Generation of polyclonal antibodies against purified AtF U T I -2.3 peptide

About 600 ug of protein in 6 M guanidine hydrochloride buffer were emulsified
with Titermax adjuvant (Cthx Corporation, Norcross GA) and injected into two New
Zealand White rabbits (MSU Clinical Center Animal Service) No. 61 and No. 86. The
rabbits were each boosted with 50 ug protein two weeks after the first bleed and the
second bleed was collected after two more weeks. All blood samples were allowed to clot
at 37°C for 30 minutes. The clots stuck to the wall of tubes were scraped down with
Pasteur pipette. The clotted blood was allowed to sit at 4°C overnight and then spun for

10 min at 1800x g in a Sorvall RT 6000D centrifuge. The supernatant was collected as

54

serum and sodium azide was added to a final concentration of 0.02%. Sera were stored at

-20°C for long term and 4°C for short term use.

Isolation of full-length coding region and genomic clone of PsF utI

All five tryptic peptide sequences obtained from purified pea XyG
fucosyltransferase were conceptually reverse translated into nucleotide sequences. After
analyzing the degenerate rate of each nucleotide sequence and the length of possible PCR
amplification products from using each two of them as primers, two of these five peptides
P5 and P2 were chosen for the design of degenerate primers. Part of the P5 peptide
sequence (ADGFDEK) was used to design the upstream degenerate primer kee230 (5’-
GCIGAYGGITTYGAYGARAA-3’; I = inosine; R = purines, A or G; Y = pyrimidines,
C or T), and part of the P2 peptide sequence (HPTNNVWG) was used to design the
downstream degenerate primer kee232 (5’-ICCCCAIACRTTRTTIGTIGGRTG-3’). The
template used for fragment PsFutl -a amplification is a plasmid cDNA library derived
from apical hooks of 4-hour-auxin-treated etiolated pea seedlings in vector pBluescript
SK (+) and was kindly provided by Dr. Hans Kende (Department of Energy Plant
Research Laboratory, Michigan State University). In a PCR reaction of 50 ul, 0.1 ul of
cDNA library template and 50 pmol of each primer were used. A discontinuous
touchdown PCR program was designed with the reaction carried for 3 cycles at annealing
temperature of 60°C, then10 cycles at 55 °C, and then 50 °C for 25 cycles. Taq
polymerase, and dNTPs were added after the denaturing step during the first

amplification cycle to hot start the reaction.

55

Second round PCR reactions were performed under the same conditions using
PsFutl-a specific primers (kee257: 5'-TTCTACTAGTTGACCCGGGAGT-3'; kee258:
5'-ACTCCCGGGTCAACTAGTAGAA-3') and the cDNA library vector specific primers
(REV: 5'-TCAGGAAACAGCTATGACCATG-3'; DIR: 5'-
GTAATACGACTCACTATAGGGC-3'). Primers kee25 7 and kee258 were designed
from a 21 bp region from PsFutl -a that bears unique restriction digestion sites for HincII,
Aval, Xmal, HapII, MspI and SmaI. Primers REV and DIR were derived from the
sequences outside the polylinker area of the vector plasmid pBluescript SK (+) in which
the template cDNA library was constructed. The same parameters and template were
used as for the amplification of fragment PsFutl -a with the program being slightly
modified to 10 cycles of amplification at an annealing temperature of 65°C, 20 cycles at
60°C and 10 cycles at 55°C.

The full-length cDNA coding region of PsFutl was amplified using Pwo
polymerase (Roche Molecular Biochemicals), end-specific primers (kee322: 5'-
AATGAATATGCTGATAAAGAGAGTC-3' and kee323: 5'-
CTAATTGTCTACGAGCTTAAGGC—3', start and stop codons are bolded), and the
same cDNA library as template under the same PCR conditions for PsFutl-a
amplification. Clone PsFutl -geno was amplified by the same reaction using 300 ng pea
genomic DNA as template. Clone PsFutl-SUTR was amplified with primers kee324 (5’-
ACGAGACATTCGTTCTAATTTCACG-3’) and kee325 (5’-

CAAAGTCAAAGGGCATGTTGTAC-3’) with 300 ng genomic DNA as template.

X yloglucan fucosyltransferase enzyme activity assay

56

Microsomal membrane samples were solubilized with 1% Triton X-100 for 15
minutes on ice immediately before enzyme assays were performed. In a
fiicosyltransferase activity assay reaction of 50 pl, 10 pl of solubilized P100 membranes
(9 p1 P100 sample plus 1 pl 10% Triton X-100) was added with 20 p1 of 2.5 x Reaction
Buffer (100 mM HEPES-KOH pH6.8, 0.5 M Sucrose, 2.5 mM DTT, 2.5 mM MgClz), 10
pl of 1% tamarind seed XyG and 2.5 p] (diluted to 10 pl with H20 before use) of GDP
14C-fucose (10 pCi/ml and 271 mCi/mmol) (NEN Life Science Products, Inc.). Reactions
were carried out at room temperature and stopped with 1 ml of 70% ethanol. Products of
the enzyme assays were precipitated at 4°C or room temperature for at least 2 hours and
washed three times with 70% ethanol. After the last washing, the pellet was briefly air

dried and resuspended in 200 pl H20 for scintillation counting.

 

Images in this thesis are presented in color

57

RESULTS
Isolation of full length cDNA clone encoding AtF U T 1

XyG fucosyltransferase was purified from microsomes derived from the epicotyls
of etiolated pea (Pisum Sativum var. Alaska) seedlings by Drs. Amy DeRocher and Maor
Bar-Peled. Although a 1400>< enrichment over microsomal membranes was achieved, two
proteins were present in the final preparation. Tryptic peptides derived from each purified
protein were subjected to sequencing (Perrin et al., 1999). One of the proteins was
identified as a homolog of Hsp70 and therefore was not pursued further. Database
searches using the resulting sequence information from the other protein led to the
identification of an Arabidopsis clone from the PRL-2 EST library that encoded four of
the six tryptic peptides. Screening of a size-fractionated Arabidopsis cDNA library
(Kieber et al., 1993) with this EST sequence yielded cDNA clone 6-1B1 (Perrin et al.,
1999)

The sequence of this cDNA was compared with that of a newly deposited BAC,
T18E12. An open reading frame (ORF) from T18E12 contained all the sequence of 6-
131 plus regions encoding three amino acids at the N-terrninus, and a region before that
presumably to be the 5’ untranslated region. I began working on the project at about this
time and participated in generating a full-length clone for the gene we identified as
described here. A pair of complementary oligonucleotide primers designed to contain
sequences encoding the missing first three amino acids (Figure 2.1A) were annealed and
inserted into vector plasmid pBluescript SK (+). The AtFutl coding region from clone 6-
181 was then inserted into the same plasmid vector right after the annealed primers,

giving rise to the hill-length clone pBS-AtFutl (Figure 2.1B). The full-length clone pBS-

58

Figure 2.1: Cloning of full-length coding region of AtFutl. A. Complementary primers
containing the sequence encoding the missing first three amino acids (in bold) of AtFutl.
Restriction enzyme recognition sites are italic. B. Cloning procedure for pBS-AtFutl.
Primers kee213 and kee214 were annealed and inserted into pBluescript through BamHI
and EcoRI to form intermediate plasmid pBS-AtFutl-Ncol. The fragment of AtFutl
released from clone 6-1 Bl was inserted behind the annealed primers on pBS-AtFutl-Ncol
through restriction site EcoRI to form the construct pBS-AtFutl. When digested with
NcoI and BglII, pBS-AtFutl gave rise to two fragments of 1.2 kb and 3.52 kb. The
construct containing an opposite insert yielded two fragments of 0.56 kb and 4.16 kb

when digested with the same enzymes.

59

A

KEE213: 5’ GATCCAGAGTCTCCCTCAAGAGCAATCCATGGATCAG 3’
KEE214: 3’ GTCTCAGAGGGAGTTCTCGTTAGG’TACCTAGTCTTAA 5’
BamHI Ncol EcoRI
B

BamHl EcoRI

    

 

         
 

 

1.2 kb

BamHI EcoRl BamHI EcoRI —
BglII 1'

pBS-AtFutl 3’52 kb
pBluescript-SK C:> pBS-AtFutl-Ncol NcoI & BglII
BamHI EcoRI digestion
0.56 kb
NCOI —

BglII :

4.16 kb

) : annealed primers kee213 and kee214

N: EcoRI fragment released from clone 6-1B1

60

AtFutl was expressed in both COS-7 cell lines (Perrin et al., 1999) and tobacco BY2 cell
lines (Sarria et al., 2001). In both cases, the expressed proteins were able to transfer
fucose from GDP-fucose to tamarind XyG, thereby confirming the identity of this clone

as XyG fucosyltransferase.

Expression of A tF U T] partial polypeptide in E. coli cells and the generation of
antibodies against it

To further confirm the identity of the cDNA clone of AtFutl and explore its
expression patterns during plant growth, a polyclonal antibody was generated against an
AtFUTl polypeptide expressed in E. coli cells.

Two DNA fragments of AtFutl, AtFutl-1.2 and AtFutl-2.3, were amplified by
PCR reactions and inserted into vector pET28a behind a 6xHis tag. Clone pET-AtFutl-
1.2 encoded a peptide starting from Met 31 until the end of AtFUTl and contained the
putative transmembrane domain that is predicted for the region between residues 45 and
62. Clone pET-AtFut1-2.3 encoded a peptide starting from Met 73 until the end of
AtFUTl and did not contain the transmembrane domain.

Peptide AtFUT1-2.3 was expressed in large quantity after IPTG induction and
was detected easily with both Commassie stained SDS-PAGE gel (Figure 2.2B) and
antibodies against the 6xHis tag (Figure 2.2C). However the majority of the expressed
polypeptide was detected in inclusion bodies (Figure 2.2D). Attempts to prevent the
formation of inclusion bodies by lowering the induction temperature, shortening the

culture time after induction, or using lower concentration of the inducing reagent IPTG

61

Figure 2.2: Expression of peptides AtFUT1-1.2 and AtFUT1-2.3 in BL21 (DE3)-
pLysS cells. A: Growth curve after IPTG induction of BL21 cells containing plasmids
pET-AtFut1-2.3, pET-AtFut1-1.2 and vector control pET28a. B: Commassie Blue stained
10% SDS-PAGE showing the induction time course of AtFUTl -2.3, AtFUT1-1.2 and
vector control (1 mM IPTG at 37°C). C: Western blot of the same gel as in B using
antibodies against 6xHis Tag (PIERCE). D: The solubility test of overexpressed peptides
AtFUTl-2.3 (1. Total protein after 4 hour induction; 2. 10,000x g supernatant after
sonication; 3. 10,000x g pellet after sonication; 4. 100,000x g supernatant from sample 2;

5. 100,000xg pellet fi'om sample 2).

62

 

   
  

 

 

8 A: Vector control

\0

8 O : AtFUT1-2.3
2'25 ‘ I : AtFUT1-1.2
1.75 <
1.25
0.75
0.25 . i fl T %

0 1 2 3 4
Time after induction (hr)
B AtFUT1-2.3 AtFUT1-1.2 vector control

0 1

 

234012340123(hr)

   

- A . +207kD

63

207 kDa

121 kDa

81 kDa

51kDa

34 kDa

AtFUT 1 -2.3 AtFUTl -1 .2 Vector control

 

0

1234012340123(hr)

 

 

 

207 kDa

121kDa

81 kDa

51 kDa

34 kDa

 

64

 

all yielded negative results. XyG fucosyltransferase activity was not detected from either
the soluble protein fractions or the inclusion bodies.

Peptide AtFUTl-1.2 with the transmembrane domain was not expressed when the
plasmid was introduced into BL21 (DE3)—pLysS cells (Figure 2.2B). After induction
with IPTG the growth of the cell culture expressing construct pET-AtFut1-1.2 was highly
suppressed as its OD600 value had only very limited increase during the incubation time
(Figure 2.2A). No protein of the expected size could be detected by either coommassie
staining of SDS-PAGE gel (Figure 2.2B) or by immunoblotting with antibodies against
the 6xHis tag (Figure 2.2C). Different colonies containing plasmid pET-AtFutl-1.2 in
BL21 (DE3)-pLysS were tried for induction but yielded similar results. Efforts to induce
the expression of peptide AtFUT1-l.2 under lower culture temperature or with lower
IPTG concentration or under both conditions yielded no recombinant protein. This
toxicity seemed to be due to the presence of the transmembrane domain of the peptide
since peptide AtFUTl -2.3 lacking the transmembrane domain was expressed in E. coli
cells (Figure 2.2).

Peptide AtFUT1-2.3 in the inclusion bodies was washed with Triton X-100 and
water three times and denatured in 6 M guanidine hydrochloride. Denatured peptides
were resolved in SDS-PAGE, excised from the gel, and injected into rabbits to generate
polyclonal antibodies. The antiserum from the second bleed recognized the E. coli
antigen very strongly and also recognized a polypeptide of ~65 kDa (Figure 2.3) from
detergent-solubilized Arabidopsis microsome proteins.

This antibody was then used by Robyn Perrin of our group to successfully

precipitate the XyG fucosyltransferase activity from detergent-solubilized Arabidopsis

65

 

207 kDa

121kDa

81 kDa

51.2 kDa

33.6 kDa

28.6 kDa

 

Figure 2.3: Anti-AtFUTl polyclonal antibodies recognize an approximately 63 kDa
polypeptide in solubilized membrane proteins of Arabidopsis. A. Immunoblotting
using antibodies against E. coli expressed AtFUT1-2.3 peptides. B. Coomassie blue
stained immunoblot PVDF membrane used in A. Lane 1 is Arabidopsis carbonate-

washed, detergent-solubilized membrane proteins and lane 2 is antigen (80 ng).

66

microsome membranes (Perrin et al., 1999), and therefore confirmed that the cDNA
identified indeed encodes the XyG fucosyltransferase. Another piece of evidence
confirming the identity of this cDNA clone came from the fact that high XyG
fucosyltransferase activity was detected from the COS7 cells or BY2 cells expressing

AtFutl (Perrin et al., 1999; Sarria et al., 2001).

Cloning of full-length coding region, intron and 5’, 3 ’ U T Rs of PsF utI

Because the XyG fucosyltransferase was purified from pea (Perrin et al., 1999),
we decided to isolate the cDNA encoding the pea XyG fucosyltransferase. Analysis of
the tryptic peptides derived from the purified pea fucosyltransferase enzyme yielded
sequence information for six peptides (P1 to P6) (Perrin et al., 1999). Five of them
showed high sequence similarity with AtFUTl (Figure 2.4A). Two of the peptides, P2
and P5, were chosen for the design of degenerate primers to amplify fragments from a
pea cDNA library (Figure 2.4B). Based on the AtFutl sequence, these two primers were
expected to generate a PCR product of 726 bp (Figure 2.4A). After unsuccessful trials
with commonly used PCR programs, a discontinuous touchdown program with hot start
was developed (Figure 2.4C). Use of this protocol yielded a single band of about 700 bp;
controls with single primer or no primer showed no clear bands (Figure 2.4D). This PCR
product was designated as PsFutl-a. The sequence of PsFutl -a has 67% identity with the
corresponding region of AtFutl at the nucleotide level and the peptide it encods has
61.6% identity with the corresponding region of AtFUT] protein. The peptide encoded
by PsFutl-a also contains the amino acid sequence of tryptic peptide P1. A further round

of PCR reactions were designed to amplify cDNA fragments toward the 5' and 3' ends of

67

Figure 2.4: Isolation of PsFutl fragment PsFutl-a. A. Sequences of tryptic peptides
derived from purified pea XyG fucosyltransferase and the deduced sequences of
comparable regions from Arabidopsis. B. Degenerate primers kee230 and kee232 used
for the amplification of PsFutl-a were designed according to the sequences of peptides
P5 and P2. Their relative positions on the cDNA were predicted according to the
positions of their comparable regions on AtFutl cDNA. C. The discontinuous
touchdown PCR program with hot start used for the amplification of fragment PsFutl-a.
The amplification reaction was first carried out at an annealing temperature of 60°C for
three cycles, then 55°C for 10 cycles, followed by 50°C for 26 cycles. Taq polymerase
was added immediately after the denaturing step of the first amplification cycle for the
hot start. D. PCR reactions for the amplification of PsFutl-a with primers kee230 and
kee232 (1); with primer kee230 alone (2); with primer kee232 alone; and with no primer

added (4).

68

 

 

 

 

Pea Peptide Arabidopsis Peptide Identity
Pl VFGFLGR VFHHLGR 86%
P2 YLLHPTNNVWGLVVR YLFHPTNQVWGLVTR 80%
P3 AVLITSLSSGYFEK AVLVTSLNAGYAEN 64%
P4 YYDAYLAK YYEAYLSH 63%
P5 LLGGLLADGFDEK LLGGLLASGFDED 85%
P6 ESILPDVNR EKLLPEVDT 44%
B
P5: A D G F D E K

Kee230: S’G C I G A Y G G I T T Y G A Y G A R A A 3’

 

 

P2: H P T N N V W G
5'C A Y C C I A C I A A Y A A Y G T I T G G G G I ’
Kee232: 3’G T R G G I T G I T T R T T R C A I A C C C C I
Kee230
' 726bp
<
Kee232
D
C 94°C 3’
94°C 1’
3x 60°C 1’
72°C 2’
94°C 1’ 1.6 kb
10x ‘ 55°C 1’ 1kb
72°C 2’
940C 1, 0.5 kb
50°C 1’
26X 72°C 2’
72°C 10’

 

69

this gene from the same cDNA library using primers derived from the fragment of
PsFutl-a and regions derived from the plasmid vector (Figure 2.5). Among five major
PCR products generated and sequenced (Figure 2.58), PsFutl-j and PsFutl-k revealed
the longest unknown sequences 5’ and 3’ to the fragment PsFutl-a, respectively. Because
one of the primers used for the amplification of both PsFutl-j and PsFutl-k came from
the same region of PsFutl-a, the combined sequence of PsFutl-j and PsFutl-k
encompassed the entire sequence of PsFutl-a. The combined sequence encoded all six
peptide sequences previously obtained from the purified enzyme (Perrin et al., 1999) with
one amino acid change in peptide P1 (Figure 2.78), indicating these sequences were
indeed part of the cDNA clone encoding the pea XyG fucosyltransferase.

However, the sequence combined from cDNA fragments PsFutl-j, PsFutl-a and
PsFutl-k still lacked the 5’ end as compared with the sequence of AtFutl cDNA. Primers
were then designed according to the sequence of fragment PsFutl-j and PsFutl-k to
amplify the missing ends of PsFutl and possibly the 5’ and 3’ UTR regions. Five 5’
RACE products were generated using primers at different locations (Figure 2.6). These
products were sequenced. One of the fragments PsFut1-5R33 showed the longest
previously-unidentified sequence containing 73 bp of coding region and 194 bp of 5’UTR
region. At the same time, one of the 3’ RACE products, PsFut1-3Ra1, was sequenced and
showed a perfect match with fragment PsFutl -k plus about 90 bp sequence after the stop
codon. The combined sequence of cDNA fragments PsFut1-5R33, PsFutl-j, PsFutl-a,
PsFutl-k yielded the full coding region for PsFutl compared with AtFutl (Figure 2.7 A).

A fiill-length cDNA coding region was amplified with primers specific for the

ends of the predicted gene (PsFutl, Figure 2.7B). The same reaction was also performed

70

Figure 2.5: Second round of PCR reactions amplifying cDNA fragments toward the
5’ and 3’ of PsFutl-a. A. Location of primers used for the PCR reactions on the plasmid
from the cDNA library template. Broken line indicates the cDNA inserts in the library
constructed on vector pBluescript. Solid line on top of the broken line is the cloned
fragment PsFutl-a. Primers kee 257 and kee 258 were derived from PsFutl-a. DIR and
REV are primers derived from the vector plasmid pBluescript SK; B. PCR reactions with
primers kee257 and REV (l), kee258 and REV (2), kee257 and DIR (3), and kee258 and
DIR (4). Five major PCR products (indicated by arrows) were cloned onto pGEM-T easy
vector (Promega) and subjected to sequencing. Fragment j and k showed the longest

unknown sequences 5’ and 3’ to fragment PsFutl-a.

 

 

Kee257/258
1?
DIR f PsFutl-a fREV
pBluescript SK
B 1 2 3 4
1.6 kb
1.0 kb
0.5 kb

 

72

a b c
kee258 kee258 kee304 cDNA primer

kee304 kee305 kee305 Nested primer

4— 1.6kb

4— 1.0kb

+— 0.5 kb
4— 0.4 kb

 

Figure 2.6: RACE reactions amplifying the 5’ and 3’ ends of PsFutl. Primers used
include kee258 and kee304 (a); kee258 and kee305 (b); kee304 and kee305 (c). Five
major PCR products were cloned into pGEM-T easy for sequencing. Fragment PsFut1-
5R33 showed the longest unknown sequence and is marked with arrow. Primer sequences

and their positions on the cDNA were as indicated in Figure 2.7.

73

Figure 2.7: DNA and protein sequences related to PsFutl and PsFUTl, and the
sequence and positions of primers used for DNA fragment amplifications. A.
Sequence alignment of cDNA fragments PsFutl-al , PsFutl -j1, PsFutl-a and PsFut1-
5R33 by Pileup of GCG program. B. Sequence of PsFutl including cDNA coding region,
intron, and 5’ and 3’ UTR. Peptide sequence is underneath the cDNA coding region.
Primers used for cDNA fragments amplification are marked with lines above or
underneath the specific nucleotide sequences. Primers kee230 and kee232 were used for
the amplification of the first PCR fragment PsFutl -a. Kee257 and k66258 were used for
the second round PCR reactions. Primers kee258, kee304 and kee305 were used for 5’
RACE while kee302 was used for 3’RACE. Primers kee322 and kee323 were used for
the amplification of full length cDNA coding region PsFutl and the genomic clone
PsFutl-geno. Kee324 and kee325 were used for the amplification of clone PsFutl-SUTR.
Peptide sequences obtained from purified enzyme are underlined and marked as P1~6,
with one mismatch for P1 indicated in parenthesis. The sequence motif found in all or-
1,2-fucosy1transferases and a-l,6-fucosyltransferases is in bold and green, and a putative
transmembrane domain is in bold and italic. Asparagines that are possibly N-glycosylated

are marked with # underneath.

74

A

PsFutl—al

PSFutl-a + ___________________ >
PsFutl-jl + .............. >
PSFUt1'5R33 + -------------- >

CONSENSUS

B

61
121

181

241
18

301
38

361
58

421
78

481
541
601
661

721
87

781
100

841
120

901
140

961
160

102]_
180

1081
200

KuEM _
ACGAGACATTCGTTCTAATTTCACGTAGTGAATCAAATGACAAAGTGCCACATCAATTTA
CCGTGGCCACAGCATCACTGTACCAGGCAGTCACACACTCAGAGTCGGTGGTTGAAGTTT
TGAAGTACCATTTGAAGTACACCACTCTCTCTACTCTCAACTAGAATCAATCGTTGTGCA

Kw$n _,
TCTGTGGCAATGAATATGCTGATAAAGAGAGTCATTGCAATCAAGAATCCAAGGGGTGAT
M N M L I K R V I A I K N P R G D

GATAATAATAATAATAAACTCTCAGATTTGGAAACCCTCACCGACAAATGTACAACATGC
D N N N N K L S D L E T L T D K C T T C

Kufifi
CCTTTGACTTTGATGCGTGTAATGGCTTTCTTCGTCGTTTCTTTCATGCTCTTCTCCGTT
P L T L M R V .M' A F F V V S F .M' L F S V

 

 

CTCTTCTCACTCTCCGTCGTTCTTCGAGACCCTCCCTCTGATGCCGCTATTTCATCTACA
L F S L S V V L R D P P S D A A I S S T

ACTACTCTCTTCCAACTCAATCAGGGTATGTTCCTCTTTTTCCAAAACTATTTCCTTTTC
T T L P Q L N Q G

TCAATTCAGTTTACCTAACTTCATGAATAAGATTGTAGATCGGGATCAGCATCGGTGGAA
ATCTATCTAGTCATTTTTGTCTAATTAATTACAAATTTAACTAAAGGTTTTGAGTTTTTG
ACAAACAACGGGCTTGTTTAGAATCTGTATTCTTAAATTTGTTAGCTTTAGTTAGAAAAA
AGATGATATTTGCTTTTCTTGCAACTGCATCTCATTGCAGAGTTATGGAATTTCATCTTC

TCTGACTTGTTACTTGTAGGTTTGGGCAGTGATGACTTTGATTCTGTGGAGTTGCTAAAT
L G s D D F D s v E L L N
Ke6230 _ Kmos
GATAAACTGCTCGGTGGCCTrcTCGCTGATGGATTTGACGAAAAATCTTGTCTCAGCAGG
D K L L G G L L A D G F D E K s c L s R
p5
TATCAGTCAGCCATTTTCGGCAAAGGACTATCAGGAAAACCTTCTTCTTACCTTATTTCT
Y Q s A I F G K G L s G K P s s Y L I s
KuMM
AGACTGAGAAAATATGAAGCCCGACACAAACAATGTGGACCTTATACTGAATCTTATAAC
R L R K Y E A R H K Q c G P Y T E s Y E
AAAACGGTGAAAGAACTCGGTTCTGGTCAATTCTCTGAGTCCGTGGATTGTAAATATGTA
K T v K E L G s G Q F s E s v D c K Y v

 

 

 

 

GTATGGATTTCGTTTAGTGGGTTAGGAAATAGGATATTGACCCTAGTTTCAGCATTTCTT
V W I S F S G L G N R I L T L V S A F L
KeeZS7 A
TACGCTCTCCTCACTGATCGTGTTCTACTAGTTGACCCGGGAGTCGATATGACTGATCTT

Y A L L T D R V L L V D P G V D M T D L
Kuflm

 

 

75

60
120
180

240
17

300
37

360
57

420
77

480
86

540
600
660
720

780
99

840
119

900
139

960
159

1020
179

1080
199

1140
219

1141
220

1201
240

1261
260

1321
280

1381
300

1441
320

1501
340

1561
360

1621
380

1681
400

1741
420

1801
440

1861
460

1921
480

1981
500

2041
520

2101
540

2161
560

2221
2281

TTTTGTGAACCATTTCCTGACGCTTCCTGGTTTGTCCCTCCTGATTTTCCTCTTAACAGT
F C E P F P D A S W F V P P D F P L N S

CATTTGAATAATTTCAACCAGGAATCGAATCAGTGTCACGGGAAAATACTGAAAACAAAA
H L N N F N Q E S N Q C H G K I L K T K

TCAATCACAAATTCCACTGTGCCATCTTTTGTTTATCTTCATCTAGCCCATGATTACGAT
S I T N S T V P S F V Y L H L A H D Y D
it
GATCATGACAAGCTTTTCTTCTGTGATGAAGAGCAACTTTTTCTCCAAAATGTACCTTTG
D H D K L F F C D E E Q L F L Q N V P L

TTGATTATGAAAACCGATAATTATTTCATCCCATCTCTGTTCTTGATGCCATCTTTTGAG
L I M K T D N Y F I P S L F L M P S F E

CAGGAGTTGAATGATCTCTTCcCAAAGAAAGAAAAAGTTTTCCATTTCTTGGGTAGATAT
Q E L N D L F P K K E K v F H F L G R X
Kee232 (G) P1
CTGTTAgAccCCACAAACAATGTGTGGGGAcTTGTCGTCAGATACTATGATGCATACTTA
L L ‘H p T N N v w G L v v R Y Y D A Y L
P2 P4
GCTAAAGTTGATGAAAGAATAGGCATACAAATCAGAGTATTTGACACCGATCCCGGTCCA
A K v D E R I G I Q I R v F D T D P G P

 

 

 

TTTCAACATGTACTCGATCAGGTCTTAGCTTGTACTTTGAAGGAGAGTATTTTGCCTGAT
F Q H V L D Q V L A C T L K E S I L P D
P6
GTAAACCGCGAGCAGAACATTAATAGTTCATCAGGAACACCAAAGTCAAAAGCTGTACTG
V N R E Q N I g S S S G T P K S K A V L

 

ATAACATCGTTAAGTTcTGGTTATTTCGAAAAGGTAAGAGACATGTATTGGGAATTTCCT

I T s L s s G Y F E K V R D M Y w E F P
P3

ACAGAAACAGGAGAAGTGGTTGGCATTTATCAGCCAAGCCATGAAGGTTATCAACAAACT

T E T G E v v G r Y Q P s H E G Y Q Q T
Kee302 _

CAGAAGCAATTTCACAACCAAAAAGCCTGGGCAGAAATGTATCTCTTAAGCTTAACTGAT

Q K Q P H N Q K A w A E M Y L L s L T D

 

 

GTGCTAGTTACTAGCTCATGGTCTACTTTTGGATACGTGGCTCAGGGGCTTGGAGGGTTG
V L V T S S W S T F G Y V A Q G L G G L

AAACCATGGATACTCTACAAACCAGAGAATCGAACGGCCCCAAATCCGCCTTGTCAACGT
K P W I L Y K P E N R T A P N P P C Q R
#

GCCATGTCAATGGAGCCATGTTTTCATGCTCCTCCTTTTTATGACTGCAAGGCTAAGAGA
A M S M E P C F H A P P F Y D C K A K R

GGAACTGACACTGGTGCACTTGTTCCACATGTAAGGCACTGTGAGGATATGAGTTGGGGS
G T D T G A L V P H V R H C E D M S W G
KaflZS
CTTAAGCTCGTAGACAATTAGTTTTATCTAGATAGCTATGGCTTATGCTACTGTAACTGT
L K L V D N *

 

TAAAACTTGCGAAGTGTTGCAAATTTTGTTGTACAATAGATTGAGTAATTCCTCGTAAAA
AAAAAAAAAA

76

1200
239

1260
259

1320
279

1380
299

1440
319

1500
339

1560
359

1620
379

1680
399

1740
419

1800
439

1860
459

1920
479

1980
499

2040
519

2100
539

2160
559

2220
566

2280
2290

Y‘)

,. r ‘3
1&pr Lurk
b V
v" N '1
-.w
lithe--

T113431”
ms det

Sequence
It
S'IIR of

a“; .3“
Salt.

FL
*1

?;

using genomic DNA as template to amplify the coding region plus introns (PsFutl-geno,
Figure 2.7B). PsFutl-geno contains all the sequences of PsFutl, plus an intron of 294 bp.

The deduced amino acid sequence of PsFUTl is shown in Figure 2.7B.

Sequence analysis of PsF ml

The cDNA clone obtained for PsFutl contains a coding region of 1695 bp, a
5’UTR of 190 bp, and a 3’UTR of 110 bp. The deduced translation product is 565 amino
acid residues long, has a predicted molecular mass of 64 kDa, and a theoretical pI of
5.91 as determined by DNASTAR (Madison, WI). The coding region of PsFutl showed
an identity of 62.3% at the amino acid level and 63.8% at the nucleotide level to AtFutl.
The highest similarity was observed at the COOH terminus, which has been postulated in
other characterized fucosyltransferases to contain the catalytic domain (Kleene and
Berger, 1993). A single transmembrane domain from amino acids 43 to 65 was predicted
by the transmembrane helices prediction server TMHMM V. 2.0

(http://www.cbs.dtu.dk/services/TMHMM) (Figure 2.78). The single N-terrninal

 

transmembrane domain is consistent with the observation that Golgi-localized
glycosyltransferases are typically type II integral membrane proteins.

An intron of 294 bp was found in PsFutl-geno when its sequence was compared
with PsFutl (Figure 2.7B). The same region was predicted to be an intron when PsFut1-
geno was subjected to a web-based gene prediction program BCM Gene Finder
(http://dot.imgen.bcm.tmc.edu:933 l/gcne-finder/gf.html). PsFutl shows not only a high
sequence identity with AtFutl, but also similar gene structure. The genomic sequence of

AtFutl (from Arabidopsis BAC clone T18E12) has two exons of 277 bp and 1400 bp

77

with an intron of 276 bp in between. The genomic sequence of PsFutl (PsFutl-geno) has
exons of 258 bp and 1440 bp with an intron of 294 bp. The sequence similarity mostly
lies in the second exon with an identity of 67% while the first exon shares only 40%
identity. The first exon encodes for the very N-terminus, the transmembrane domain and
the stem region, and therefore it is more variable compared with the region coded by the
second exon, which contains conserved motifs for or- 1,2; 1,6 fucosyltransferases and the
catalytic domain. The introns from these two genes are located at similar position on their
respective genomic clones and also share a 47% identity, even higher than the identity
shared by the first exon. It seems likely that the intron was already present in the gene
before evolutionary divergence of peas and Arabidopsis.

Four N-glycosylation sites were predicted by the GCG program (Madison,
Wisconsin) (Figure 2.78). This program also predicted a cAMP/cGMP-dependent protein
kinase phosphorylation site at position 538-541, five casein kinase II phosphorylation
sites, a glycosaminoglycan attachment site at position 185-188, four N-myristoylation
sites, and seven protein kinase C phosphorylation sites. However all these predictions
need experimental confirmation to determine whether any of these posttranslational
modifications actually occur.

Although AtFUTl and PsFUTl proteins have very low overall sequence
similarity with other known fiicosyltransferases (less than 20% identity), three conserved
motifs were found using the PATTERNF IND program (http://hits.isb-sib.ch/cgi-
bin/hits_patsearch) (F aik et al., 2000). Among five motifs that were previously identified

(Oriol et al., 1999) to be present in all known :L-l,2- and oc-l ,6-fucosyltransferases, three

78

of them were found in both PsFUTl and AtFUTl (F aik et al., 2000).

Specificity analysis of clone Ps5 U T R as probe for Northern blotting

A DNA fragment named PsFutl -5UTR containing 189 bp of 5’UTR sequence
and 124 bp of 5’ end coding region of PsFutl was amplified from pea genomic DNA
(Figure 2.7B) and used as a probe for northern studies. To determine whether this region
is gene specific, the counterpart regions from all AtFut genes were identified from their
corresponding BAC clones and aligned by MultiAlign (DNAStar), a computer program
using ClustalW. Among the ten AtFut genes, this part of the sequence from AtFutl
shared only very low identity with other AtFut genes (5.1 to 16.3%) (Table 2.1). At this
low identity level it was very unlikely that the probe region from AtFutl would recognize
other AtFut genes. Therefore we postulated that PsFutl-SUTR would also be specific for
PsFutl when used as probe for northern blotting in peas. This assessment was further
tested by probing pea genomic DNA digested with EcoRV and HindIII using PsFut1-
5UTR as probe (Figure 2.8). There are only one EcoRV restriction site and two HindIII
restriction sites in the PsFutl genomic clone (Figure 2.8A). All three restriction sites
were in exon II (Fig2.8A). Therefore when PsFutl -5UTR was used as probe for southern
blotting, only one band was expected for both digestions. That was exactly what was
observed (Fig2.8B), supporting the conclusion that PsFutl -5UTR is a gene-specific probe

for PsFutl.

Intemode elongation curve of etiolated pea seedlings

79

I - ....-ln- .u..:-..J:...- [\ZA — Laid-.3”— uuu..-v< an: nan-vi.- l--.v-oa.I--vav I: urn-on VFIL um].- i-u..-lcl nun-av-nnoa --.v--~a§§-u .Utdnnaiunhvbml. uh-” idhniow.’

 

MFDmmm

 

 

 

 

 

 

 

 

 

 

 

 

93%< mm .13..

$.53. m; L cém ..cI.

tzuc< _.o 02 v.2 I...

93%.... 0.: odm mam m.w~ .22..

9452 QB v.3 9.0 9mm OGN 1....

Esau/x m6 o.o_ To a: wé m.m~ 1....

mgut< v.0 ow v.0 Nun v.0 NE “.2 .12..

NHL—£3. we 0.x _.m N:— _ m6 v.0 Qw m.» .23..

o::..:< Wm Yo Ym Wm Wm mm To 50 men .12..

2.52 v.0 Tm, us 3.3 _.m A.» ma: ox Pm me .12..

mFDmmL 93:/x M2252 Esq—3. 0:554 9.52 332.4 035/4 Sande. o L 31,344 ::%<

 

 

 

 

 

 

 

 

 

 

 

 

 

dusts—i=6 :5ch

2e 838:2 8052: 2:835 e .CSm/xzov :w:<£=2 .3 @233 can muse—o U<m wcmcconmotoo :2: 86¢ @8253

083 mosow Sn; :2“ com REES 2:8 2:. .Ssmmm mo 5&2 @550 25 ..n we as VS 98 MED .m we as am: 28:80 mHDm

-Ssmwm EoEmfiu <20 .85» 3&3. .1. E9... 3533:? a: use “Piméaammm weefia 2555:: 35.5% "flu 033—.

80

SWITR ExonI Inuon

 

 

 

ExonIl 3WJTR
189bp 258bp 294bp l440bp IOObp
' HindIII HindIII
Ps5UTR 1330bp EcoRV 1908bp
1498bp
> E
B' e 2
a E
6.0 kb
4.0 kb
3.0 kb
2.0 kb
1.6 kb
1.0 kb

 

Figure 2.8: PsFutl specific DNA probe Ps5UTR. A, The organization of clone

PsFUTgeno. Recognition sites for HindIII and EcoRV are indicated. Clone Ps5UTR

contains 189bp of 5’ UTR and 124bp of 5’ end coding region; B, Southern analysis of pea

genomic DNA using PsSUTR as probe. Fifteen pg of genomic DNA was used for each

digestion. The signals below the 1.0 kb marker are non-specific irregular spots.

81

I examined the expression of PsFutl during growth of etiolated pea in terms of
enzyme activity and messenger RNA abundance. The growth of etiolated pea seedlings
was monitored by measuring the intemode length daily from day 6 until day 11 (Figure
2.9A). At day 6 the first intemode had already accomplished its elongation as very little
additional length increase was seen in the following days (Figure 2.9). The second
intemode elongated slightly from day 6 to day 7 with an increase of 1.1 cm, and stopped
its elongation thereafter (Figure 2.9). However the third intemode showed dramatic
elongation between days 6 and 11, showing an increase in length of more than 10 fold.
The most rapid elongation for the third intemode occurred from day 7 to day 8 when the
length increased by 5.0 cm (Figure 2.9). The elongation rate for the third intemode then
decreased daily after this time point until it increased by only 0.6 cm from day 10 to day
11 (Figure 2.9B). Collectively, the results demonstrated that the third intemode had the
highest elongation rate at day 7 and that elongation of this intemode was nearly complete

atday 11.

Expression of PsF ml in three intemodes of etiolated pea seedlings at 7 days old

To correlate the expression of PsFutl with the elongation of etiolated pea
intemodes, microsomal membranes were isolated from the three intemodes of etiolated
seedlings at 7 days old and examined for XyG fucosyltransferase activity. Total RNA
was also isolated from the same tissues and examined for PsFutl messenger RNA
abundance. PsFutl -5UTR was used as probe for the northern blotting. As indicated in
Figure 2.9, only the third intemode showed dramatic elongation at day 7. Accordingly,

the specific activity of PsFUTl from the third intemode was found to be nearly 6 times

82

I 2 2nd intemode

  
  

 

 

14 i A: 1St intemode
A 12 . O I 3rd intemode
a l
3 ,
.x: 10 *
8'0 1
5 l
I; 8 1
1: .
o
E 6 ~
53 l
E 1

4 T

l
2 -
0 L

5 6 7 8 9 10 1 1 12
Days after planted

a

6
i . . .
5 r lSt intemode
l 2“d intemode
I 3rd intemode

Intemode elongation rate (cm/day)

 

 

 

’ 7-8 8-9 9-10 10-11

-1 3 Days after planting
Figure 2.9: Growth curve of etiolated pea internodes. All lengths were measured in the
dark under safe light, and were averages of 40 seedlings. A, intemode length of etiolated
pea seedlings from day 6 to day 11; B, elongation rate of etiolated pea intemodes

expressed as increased length (cm) per day comparing two consecutive days.

83

that from the first intemode and 3 times that from the second intemode (Figure 2.10A).
While the second intemode didn’t show much increase of PsFutl messenger RNA
compared with the first intemode at day 7, the third intemode had dramatically higher
expression of PsFutl as measured by northern blotting (Figure 2.108). This experiment
demonstrated a good correlation between pea intemode elongation rate and the

expression of PsFutl.

Expression of PsF ml in the third intemode at different growth stages

From day 7 to day l l, the elongation rate of the third intemode decreased almost
8-fold, although the absolute length increased almost 3-fold (Figure 2.9). There was a
223% increase in intemode length from day 7 to day 9 compared with only 75% increase
from day 9 to day 11 (Table 2.2). Accordingly, the elongation rate decreased dramatically
from 5.0 cm/day at day 7 to 0.6 cm/day at day 11 (Figure 2.9). The PsFUTl activity also
decreased during this time (Figure 2.11A). From day 7 to day 9, the specific activity of
PsFUTl decreased 48% while from day 9 to day 11, it further decreased by 22.5% (Table
2). Thus a correlation between the decrease of elongation rate and the decrease of
PsFUTl enzyme activity was observed. At the same time the PsFutl transcripts also
showed a dramatic decrease between day 7 and day 9, and a further, but smaller, decrease
between day 9 and day 11 (Figure 2.1 18). This again demonstrated a correlation between

the intemode elongation rate and PsFutl expression.

84

Specific activity of XyG fucosyltransferase
(pmol hr“1 mg ‘ microsomal protern)

 

lst intemode 2nd intemode 3rd intemode

1 st 2nd 3rd

PsSUTR ~ m

 

18$ rRNA [ k ;

Figure 2.10: Expression of PsFutl in etiolated pea intemodes at day 7. A, Specific

activity of XyG fucosyltransferase; B, northern blotting using PsSUTR as probe.

85

Table 2.2: PsFUTl activity decreases during the elongation of third internode between
day 7, 9 and 11.

 

 

 

 

After planted Intemode length (cm) Spec. act. (pmol'hr‘-mg‘1 micros. pro.)
7 days 3.59 increase 1686.4 decrease
9 days 11.6 223% 873.6 48.2%
11 days 14.29 75% 494.2 22.5%

 

 

 

 

 

 

 

86

>

2500 a

 

2000

1500

4.— 4; ‘44’_ .AL

1000 l

500

Specific activity of XyG fucosyltransferase
(pmol hr 1 mg ' microsomal protein)

     

7 days 9 days 1 1 days

Days after planted 7 9 ll
PsSUTR }. in «2.6a! W '

18S rRNA

 

Figure 2.1]: Expression of PsFUTl in the third intemode of the etiolated pea at 7, 9
and 11 days old. A, Specific activity of PsFUTl; B, Northern blotting using PsSUTR as

probe.

87

DISCUSSION
Isolation of cDNAs encoding the X yG fucosyltransferase

Identifying genes involved in the biosynthesis of plant cell wall polysaccharides is
the first step to understand the mechanism of plant growth and development. Only a few
genes with proven cell wall biosynthesis functions have been identified. Through the
efforts described in this chapter, the gene encoding the XyG fucosyltransferase was
identified from both Arabidopsis and pea. The identity of AtFutl was confirmed in two
ways. The first involved immunoprecipitation of XyG fucosyltransferase activity from
solubilized Arabidopsis microsomal membranes using antibodies against part of AtFUTl
protein. The second involved expression of AtFUTl in COS7 cells and BY2 cells and
detection of XyG fucosyltransferase activities in extracts of the transformed cells. The
identity of PsFutl was confirmed by the fact that the protein encoded by PsFutl contains
the sequences of all six peptides that we obtained from the microsequencing of the
purified PsFUTl.

The complete Arabidopsis genome has revealed the presence of a
fucosyltransferase family with nine additional AtFut genes in addition to AtFutl. AtFutl ,
however, is the only AtFut gene for which the function of the gene product has been
determined, i.e., adding a fucose residue to the side chain galactose of XyG in an a-1,2
glycosidic bond. All of the AtFut genes were placed into the same CAZy family 37
according to the sequence comparison of the proteins they encode. The other AtFUT
proteins showed less sequence identity with PsFUTl compared with AtFUTl , and are not

likely XyG fiicosyltransferases. Indeed XyG fucosyltransferase activity was not detected

88

in the BY2 cells expressing AtFUT3, 4 and 5 using non-fucosylated XyG as substrate
(Sarria et al., 2001).

The Kyte-Doolittle hydrophilicity plots suggested the presence of a
transmembrane domain at the NHz terminus for both AtFUTl and PsFUTl. Based on this,
it is likely that both AtFUTl and PsFUTl are type 11 membrane proteins (Singer, 1990).
The catalytic domains of AtFUTl and PsFUTl are likely to be facing the lumen and
similar to all other fucosyltransferases described to date. This conclusion is fiirther
supported by the fact that Triton X-100 solubilization of microsomes from both pea and
Arabidopsis could stimulate the XyG fucosyltransferase many times (Wulff et al., 2000).

An Arabidopsis mutant line mur2 was identified to carry a point mutation on the
XyG fucosyltransferase gene AtFutl (Vanzin et al., 2002). XyG isolated from leaves,
stems, flowers and roots of mur2 plants all contained only 2% of the wild-type amount of
fucose, while the fucose content of pectins and glycoproteins in mur2 plants are very
similar to that of the wild type plants (Vanzin et al., 2002). The mutant copy of AtF UT 1
expressed in mammalian cells also lost its ability to transfer fucose from GDP-fucose
onto the galactose residue of tamarind XyG (Vanzin et al., 2002). A T-DNA insertion
mutant line of AtFutl was also identified from the Arabidopsis Knockout Facility
(Madison, WI) (Perrin et al., 2003). This AtFut] mutant showed 50% - 75% less fucose
than wild type in mature rosette leaves, expanding leaves, and flowers (Perrin et al.,
2003). While the fucose content in RG1 and RGII of AtF utl mutant is comparable with
that of the wild type plants, no fucose was detected from the 4 M KOH-soluble XyG

fraction isolated from the mutant plants (Perrin et al., 2003). The characterizations of

89

AtFutl mutants further confirmed that AtFutl encodes for the enzyme responsible for the
attachment of the terminal fucose onto XyG.

The discontinuous touchdown PCR program (Figure 2.4C) developed for the
amplification of PsFutl -a is a powerful and effective program that easily amplifies low-
abundance targets, especially low copy number templates from cDNA libraries where
traditional PCR programs do not always work (Dr. Qin Du, personal communication).
This program was also used successfully to amplify target DNA fragments from non-
traditional organisms or from unpurified DNA samples contaminated with lipids and
proteins. One example was the successful utilization of this PCR program to amplify
various prokaryotic 16s rRNA fragments from the house fly intestinal extracts containing
various bacteria, where a various of traditional PCR programs didn’t work (Dr. Yunling
Xu, Dr. Edward Walker, personal communication). This PCR program starts with high
annealing temperature at 60°C for a small number of cycles to ensure the fidelity of the
initial amplification products. The product from the first few amplification cycles
therefore becomes part of the perfect match template pool for later cycles. Later cycles
using lower annealing temperature at 55 °C or 50°C ensured that enough products would
be generated. This combination of different annealing temperatures therefore highly
increases the probability that the amplification product from the true template become the

dominant product at the end of the PCR reactions.

Expression of PsF at] is correlated with intemode elongation rate

The growth and development of plants requires a series of biochemical processes,

including the biosynthesis and degradation of cell wall polysaccharides. Changes in the

90

plant cell wall occur in response to many developmental and environmental stimuli to
coordinate cell expansion and maturation. These changes are thought to occur by
regulating the expression of enzymes responsible for the metabolism of cell wall
polysaccharides. In this chapter I utilized the fast-growing intemodes from etiolated pea
seedlings during their early growth stages to correlate the expression of XyG
fucosyltransferase with intemode elongation rate. More XyG fucosyltransferase activity
and PsFutl transcripts were detected in the third intemode at 7 days old that showed the
highest elongation rates, compared with either the first and second intemodes of the same
age, or the third intemode at 9 or 11 days old.

Since XyG is one of the major structural components of the primary cell walls, its
metabolism plays a major role during wall expansion in plant development and growth
(Talbott and Ray, 1992). During auxin-induced rapid elongation of pea epicotyls, Xsz
are partially solubilized and their molecular weight decrease (Labavitch and Ray, 1974;
Hayashi and Maclachlan, 1984; Hoson, 1990; Talbott and Ray, 1992). The metabolism of
XyG consists of its degradation, biosynthesis and crosslinking into the wall. The
regulation of these three processes are all mediated by the regulated expressions of
corresponding enzymes. Endoglucanases (EGs) (Verma et al., 1975), OL-fucosidase
(Augur et al., 1995) and oc-xylosidase (Sampedro et al., 2001) are examples of XyG
degradation enzymes that show a higher expression in fast growing tissues or at fast
growing stages. XyG endotransglycosylases (XETs) and expansins are enzymes involved
in the incorporation of new polysaccharide molecules into cell walls. They also have been
shown to be up-regulated during the elongation of plant cells (Polisensky and Braam,

1996; Steele et al., 2001; Lee et al., 2003). Therefore the incorporation of XyG into the

91

expanding walls would be a net result of actions from enzymes involved in XyG

degradation, biosynthesis and reorganization.

92

REFERENCES
Aldington, S. and Fry, S. C. (1993). "Oligosaccharins." Adv. Bot. Res. 19: 1-101.

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 (42): 24839-24843.

Camirand, A. and Maclachlan, G. (1986). "Biosynthesis of the fucose-containing
xyloglucan nonasaccharide by pea Pisum sativum cultivar Alaska microsomal
membranes." Plant Physiol. 82 (2): 379-383.

Cosgrove, D. J. (1997). "Assembly and enlargement of the primary cell wall in plants."
Annu. Rev. Cell Dev. Biol. 13: 171-201.

Cosgrove, D. J. (2000). "Expansive growth of plant cell walls." Plant Physiol. Biochem.
38 (1-2): 109-124.

Darvill, A. G., Augur, C., Bergmann, C., Carlson, R. W., Cheong, J.-J., Eberhard, S.,
Hahn, M. G., Lo, V.-M., Marfa, V., Meyer, B., Mohnen, D., O' Neill, M. A., Spiro, M.
D., Van Halbeek, H., York, W. S. and Albersheim, P. (1992). "Oligosaccharins-
oligosaccharides that regulate growth, development and defence responses in plants."
Glycobiology 2: 181-198.

Desveaux, D., Faik, A. and Maclachlan, G. (1998). "Fucosyltransferase and the

biosynthesis of storage and structural xyloglucan in developing nasturtium fruits." Plant
Physiol. 118 (3): 885-894.

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-fucosyltransferase that catalyzes the last step of cell wall

xyloglucan biosynthesis in pea." J. Biol. Chem. 275 (20): 15082-15089.

Farkas, V. and Maclachlan, G. (1988). "Fucosylation of exogenous xyloglucans by pea
microsomal membranes." Arch. Biochem. Biophysi. 264 (1): 48-53.

Fry, S. C. (1989). "The structure and functions of xyloglucan." J. Exp. Bot. 40 (210): 1-
11.

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

(2): 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-

93

alpha-L-fucosyltransferase from pea epicotyl membranes." Arch. Biochem. Biophys. 290
(1): 7-13.

Hayashi, T. and Maclachlan, G. (1984). "Pea Pisum sativum cultivar Alaska xyloglucan
and cellulose 1. Macromolecular organization." Plant Physiol. 75 (3): 596-604.

Hayashi, T. and Maclachlan, G. (1984). "Pea Pisum sativum cultivar Alaska xyloglucan
and cellulose 3. Metabolism during lateral expansion of pea epicotyl cells." Plant Physio].
76(3): 739-742.

Hayashi, T. and Matsuda, K. (1981). "Biosynthesis of xyloglucan in suspension
cultured soybean cells: Occurrence and some properties of xyloglucan 4-beta-D-
glucosyltransferase and 6-alpha-D-xylosyltransferase." J. Biol. Chem. 256 (21): 11117-
11122.

Hayashi, T., Ogawa, K. and Mitsuishi, Y. (1994). "Characterization of the adsorption
of xyloglucan to cellulose." Plant Cell Physio]. 35 (8): 1199-1205.

Hoson, T. (1990). "Effect of auxin on autolysis of cell walls in Azuki bean epicotyls."
Plant Cell Physiol. 31: 281-287.

Houdebine, L.-M. and Puissant, C. (1990). "An improvement of the single-step method
of RNA isolation." BioTechniques 8 (2): 148-149.

Kato, Y. and Matsuda, K. (1980). "Structure of oligosaccharides obtained by controlled
degradation of mung bean xyloglucan with acid and Aspergillus oryzae enzyme
preparation." Agric. Biol. Chem. 44(8): 1751-1758.

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-kinase." Cell 72 (3): 427-441.

Kleene, R. and Berger, E. G. (1993). "The molecular and cell biology of
glycosyltransferases." Biochim. Biophys. Acta. 1154 (3-4): 283-325.

Labavitch, J. M. and Ray, P. M. (1974). "Relationship between promotion of
xyloglucan metabolism and induction of elongation by indoleacetic acid." Plant Physiol.
54: 499-502.

Lee, D.-K., Ahn, J. H., Song, S.-K., Choi, Y. D. and Lee, J. S. (2003). "Expression of
an expansin gene is correlated with root elongation in soybean." Plant Physio]. 131: 985-
997.

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 (3): 373-386.

94

Maclachlan, G., Levy, B. and Farkas, V. (1992). "Acceptor requirements for GDP-
fucose xyloglucan 1, 2-alpha-L-fucosyltransferase activity solubilized from pea epicotyl
membranes." Arch. Biochem. Biophys. 294 (1): 200-205.

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." J.
Exp. Bot. 20 (211): 233-238.

McDougall, G. J. and Fry, S. C. (1989). "Structure-activity relationships for xyloglucan
oligosaccharides with antiauxin activity." Plant Physiol. 89 (3): 883-887.

Oriol, R., Mollicone, R., Cailleau, A., Balanzino, L. and Breton, C. (1999).
"Divergent evolution of fucosyltransferase genes from vertebrates, invertebrates, and
bacteria." Glycobiology 9: 323-334.

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 (5-6): 842-
850.

Pennell, R. (1998). "Cell walls: structures and signals." C arr. Opin. Plant Biol. 1: 504-
510.

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 (5422): 1976-1979.

Perrin, R. M., Jia, Z., Wagner, T. A., O'Neill, M. A., Sarria, R., York, W. S., Raikhel,
N. V. and Keegstra, K. (2003). "Analysis of xyloglucan fucosylation in Arabidopsis."
Plant Physiol. 132: 1-1 1.

Polisensky, D. H. and Braam, J. (1996). "Cold-shock regulation of the Arabidopsis
TCH genes and the effects modulating intracellular calcium levels." Plant Physiol. 111
(4): 1271-1279.

Sambrook, J., F ritsch, E. F. and Maniatis, T. (1989). Molecular Cloning, a Laboratory
Manual. Cold Spring Harbor Laboratory Press.

Sampedro, J., Sieiro, C., Revilla, G., Gonzalez, V. T. and Zarra, I. (2001). "Cloning
and expression pattern of a gene encoding an alpha-xylosidase active against xyloglucan
oligosaccharides from Arabidopsis." Plant Physio]. 126 (2): 910-920.

Sarria, R., Wagner, T. A., O' N eill, M. A., Faik, A., Wilkerson, C. G., Keegstra, K.

and Raikhel, N. V. (2001). "Characterization of a family of Arabidopsis genes related to
xyloglucan fucosyltransferasel." Plant Physiol. 127 (4): 1595-1606.

95

Singer, S. J. (1990). "The structure and insertion of integral proteins in membranes."
Annu. Rev. Cell Biol. 6: 247-296.

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
enzyrne-substrate complex." Biochem. J. 340 (1): 207-211.

Steele, N. M., Sulova, Z., 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." Biochem. J. 355 (3): 671-679.

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 (3): 1475-1480.

Talbott, L. D. and Ray, P. M. (1992). "Changes in molecular size of previously
deposited and newly synthesized pea cell wall matrix polysaccharides effects of auxin
and turgor." Plant Physiol. 98 (1): 369-379.

Vanzin, G. F., Madson, M., Carpita, N. C., Raikhel, N. V., Keegstra, K. and Reiter,
W. D. (2002). "The mur2 mutant of Arabidopsis thaliana lacks fucosylated xyloglucan
because of a lesion in fucosyltransferase AtFUTl." Proc. Natl Acad. Sci. USA 99 (5):
3340-3345.

Verma, D. P. S., Maclachlan, G., Byrne, H. and Ewings, D. (1975). "Regulation and in
vitro translation of messenger ribonucleic acid for cellulase from auxin-treated pea
epicotyls." J. Biol. Chem. 250 (1019-1026).

Whitney, S. E. C., Brigham, J. E., Drake, 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 (4): 491-504.

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 Physio]. 122 (3): 867-877.

York, W. S., Darvill, A. G. and Albersheim, P. (1984). "Inhibition of 2,4-D stimulated

elongation of pea stem segments by a xyloglucan Oligosaccharide." Plant Physiol. 75 (2):
295-297.

96

CHAPTER 3

Sequence and Expression Pattern Analysis of Arabidopsis

Cellulose Synthase-Like Genes

97

product

”My
~.iz\l.l¥§

liter thr-

PREFACE

My research focus has shifted from XyG fucosyltransferase to another aspect of
XyG biosynthesis, i.e., efforts to identify the genes encoding the glucan synthase(s) that
produce(s) the glucan backbone of XyG. The effort to identify the genes encoding
enzymes involved in XyG biosynthesis began, and continue to operate, as a group project.
After the initial identification of fucosyltransferase genes, research efforts have been
divided into smaller new projects for different individuals involving new directions. I
decided to test the hypothesis that one or more members of the Arabidopsis cellulose

synthase-like gene families might be involved in the biosynthesis of the glucan backbone

of XyG.

98

ABSTRACT

Arabidopsis cellulose synthase-like genes are a group of 30 gene encoding
proteins that share identities ranging from 7% to 46% with Arabidopsis cellulose
synthase proteins. It has been hypothesized that the cellulose synthase-like genes may
encode proteins involved in the synthesis of other polysaccharides, such as the glucan
backbone of XyG. I started to evaluate this hypothesis by analyzing the sequences of
AtCsl genes and the proteins they encode. The automated annotation in the public
database resulted in unusual protein sequence alignments that most likely derived from
annotation mistakes. Therefore, the beginning of this chapter focuses on a manual re-
annotation and sequence analysis of AtCsl genes. We sought to investigate the possibility
that genes involved in the biosynthesis of the same polysaccharide might show similar
expression profiles. Therefore the expression profiles of AtCsl genes were compared with
genes known to be involved in cell wall biosynthesis. The dbEST and, Arabidopsis
MPSS databases, and the microarray data from Stanford Microarray Database and the
Nottingham Arabidopsis Stock Center Microarray Database were all utilized for the gene
expression analysis. Even though the expression studies did not reveal a tight link
between AtCsl genes and XyG fucosyltransferase, xylosyltransferase or
galactosyltransferase, several interesting observations were recorded. Specifically, some
AtCsl genes showed strong tissue expression preferences, and the expression of some
AtCsl genes changed in response to environmental signals. These sequence analyses and
gene expression profiling could be used in future studies on the fimctions of AtCsl genes

and the changes in wall biosynthesis in response to environmental signals.

99

INTRODUCTION

XyG is the major hemicellulose for dicot and nongraminaceous monocot plants.
Its most common structure is a B-(l ,4)-glucan backbone modified with side chains
composed of oc-(1,6)-xylose, B-(1,2)-galactose, and a-(l,2)-fucose residues (Figure 1.1)
(Hayashi and Maclachlan, 1984). XyG is proposed to be synthesized through the action
of at least four different enzymes, i.e., B-(l,4)-glucan synthase, a-(l,6)-xylosyltransferase,
[3-(1,2)-galactosyltransferase and (x-(l ,2)-fucosyltransferase (Camirand and Maclachlan,
1986; Camirand et al., 1987). XyG fucosyltransferase has already been identified through
earlier efforts of our group (Perrin et al., 1999; Faik et al., 2000). Other studies were also
being carried out by Dr. Ahmed Faik in our group for the identification of
xylosyltransferase (Faik et al., 2002) and Dr. Wolf Dieter Reiter’s group for the
identification of galactosyltransferase (Madson et al., 2003). Therefore my effort was
focused on the identification of the glucan synthase that is responsible for the
biosynthesis of the XyG backbone.

XyG backbone has the same chemical structure as cellulose, a rigid, straight chain
polymer of B—(l ,4)-linked D-glucose residues (Kato and Matsuda, 1980; Fry, 1989). It is
reasonable to postulate that the glucan synthase responsible for the synthesis of XyG
backbone will have significant sequence similarity to cellulose synthase, because it
recognizes the same donor substrate, i.e. UDP-glucose, and produces a similar product.
This was proved by the synthesis of XyG when UDP-glucose and UDP-xylose were
added into suspension cultured soybean cells at the same time (Hayashi and Matsuda,
1981). However, due to the existence of the sidechain sugar moieties on XyG, the

enzyme responsible for the backbone synthesis should also have some unique features.

100

Moreover, the XyG glucan synthase is hypothesized to operate in the Golgi complex
whereas cellulose synthase forms its product at the plasma membrane (Delmer and Stone,
1988). Initial attempts to identify homologs of cellulose synthase in Arabidopsis
identified a cellulose synthase-like (AtCsl) gene family (Cutler and Somerville, 1997),
encoding proteins of only about 30% identity to the cotton cellulose synthase encoded by
GhCesAl, which was the first cellulose synthase gene cloned in plants (Pear et al., 1996).

As sequencing of the Arabidopsis genome progressed, additional sequence data
allowed a more detailed and complete computational analysis of AtCsl genes. Richmond
and Somerville (2000) identified 30 AtCsl genes in the Arabidopsis genome through blast
searches (Richmond, http://cellwall.stanford.edu/cesa/index.shtrnl). All Arabidopsis
cellulose synthase (AtCesA) genes and Cs] genes were placed in a large cellulose
synthase superfamily. AtCsl genes were grouped into six families, namely AtCslA,
AtCslB, AtCle, AtCle, AtCle, and AtCslG, according to the distances among
sequences within the phylogenetic tree (Richmond, 2000; Richmond and Somerville,
2000; Saxena and Brown, 2000). Proteins encoded by the various Csl genes have become
the most prominent candidates for the synthesis of XyG backbone. It was postulated that
some of them might also function in the synthesis of other polysaccharide backbones
such as xylan, galactans, or mannans that are constructed with B-glycosidic linkages
(Cutler and Somerville, 1997). This chapter describes some efforts to explore this
hypothesis as it relates to the XyG glucan backbone.

When we began our work at the early stages in the Arabidopsis genome
sequencing project, many gene annotations were not accurate. When the sequences of

AtCsl genes were extracted from the GenBank and aligned, many of them showed

101

‘ . .\
wars) 1
h

1
1w "

Conclusmr

compare:

.6341 [0 ll'
01.:pr
‘ ‘1 r“ . .
15111111) t
leifft’m

110.?th

missing 5’ or 3’ coding regions or missing domains in the middle of the gene.
Conclusions from those sequence comparisons many times were misleading. Therefore
my first effort was to re-annotate the AtCsl gene sequences in the database as much as
possible. Most of the re-annotated sequences were confirmed to be correct by EST clones
or cDNA clones identified later on. The re-annotated sequences then were analyzed and
compared for gene or protein specific features and features common for each family.

Knowledge about where, when and to what extent a gene is expressed can often
lead to insights about the function of the gene and an understanding of the biological role
of the protein it encodes (Lockhart and Winzeler, 2000). Different tissues or cell types
usually express different sets of genes or the same genes to different extents. The
differences in gene expression determine the different cellular phenotypes and plant
morphologies. Transcriptional profiling therefore has become an important approach
during efforts to investigate the function of genes that are known only by their sequences.

The expressed sequence tags (ESTs) derived from large-scale sequencing of
cDNA libraries provide valuable information for the annotation of genomic sequences
and the gene expression profiles associated with specific tissues under specific growth
conditions. The frequency of the appearance of ESTs recovered from specific cDNA
libraries have often been used to estimate the expression pattern of individual genes
(White et al., 2000; F emandes et al., 2002).

MPSS (Massively Parallel Signature Sequencing) is another gene expression
analysis technique, generating gene-specific short sequence tags from a defined position
within a particular mRNA (Brenner et al., 2000; Brenner et al., 2000). The relative

abundance of these tags for a particular gene in a given library represents a quantitative

102

 

estimate of expression of that gene. An Arabidopsis MPSS database was built up by the

Meyers lab (http://dbix5001 .dbi.udel.edu/MPSS4/java.htm1), which includes 12.3 million

 

MPSS tags of 17 bp from five different libraries constructed from callus (1.96 million),
inflorescence (1.79 million), shoot (2.89 million), root (3.65 million) and silique (2.02
million). More than 95% of all Arabidopsis genes could be identified with unique MPSS

tags using these five MPSS libraries (hllLJ/(lleSOOl .dbi.udel.edu/MPSS4/iava.html).

 

Microarray analysis has been a powerful method for evaluating transcript
abundance on a genome scale. The expression profiles of genes of interest can be
examined at the same time for any given tissue under a given growth condition. Changes
in gene expression under various growth conditions or environmental stresses also can be
detected with microarray techniques. The Stanford Microarray Database (SMD,

http://genome-www.stanford.edu/microarray) contains data from over 3500 microarrays

 

from various experiments involving different model organisms including Arabidopsis
(Gollub et al., 2003). The Nottingham Arabidopsis Stock Center’s microarray database
(NASCArrays) is another database containing large amounts of data from many
Arabidopsis-related experiments using Affymetrix microarray chips
(http://ssbdjc2.nottinghamac.uk/narrays/experimentbrowsepl). These two databases
therefore are a rich source of gene expression profiles.

To investigate the hypothesis that one or more AtCsl genes might encode the XyG
glucan synthase, the expression patterns of AtCsl genes were compared with genes
encoding known XyG glycosyltransferases, i.e. fucosyltransferase (AtFutl),
galactosyltransferase (AtGalT) and xylosyltransferase (AtXTl). Assuming that genes that

cooperate during the biosynthesis of the same cell wall polysaccharides might show

103

similar expression patterns, any AtCsl genes that show similar expression patterns with
those encoding XyG glycosyltransferases would be very likely encoding XyG backbone
glucan synthase. For this purpose the tissue expression profiles of AtCsl genes were
analyzed by examining their corresponding EST clones from the dbEST and the
frequency of their specific MPSS signatures from the Arabidopsis MPSS database. The
relative expression levels among different tissues of Arabidopsis were also examined by
analyzing Arabidopsis tissue comparison data from the Stanford Microarray Database.
Finally, data from microarray experiments carried under conditions that might have
physiological effects on cell wall metabolism from both the SMD and the NASCArrays
were analyzed. The behaviors of all AtCsl genes appeared on those microarrays were

examined for their possible co-regulation with known cell wall biosynthesis-related genes.

104

MATERIALS AND METHODS

Sequence extraction and annotation

Sequences for all AtCsl genes were obtained in May of 2000 from a BlastN
search against the NCBI database using the sequence of AtCesAl (Atstl—l) as query.
The BAC clone IDs obtained from the blast search were compared with Dr. Todd
Richmond’s collection for the designation of each AtCsl gene. All cDNA sequences thus
obtained were translated into protein sequences using the EditSeq program of DNASTAR
(Madison, WI). Protein sequences were aligned by Multalin with default parameters

(httpz/fiirodestoulouse.inra.fr/multalin/multalin.html). For proteins that showed obvious

 

N-terminal, C-terrninal or internal deletions, or out-of-place sequence insertions, the
equivalent region from its BAC clone was identified and translated into six frames with
the Map program of GCG (Madison, WI). The possible amino acid sequences were then
examined against the protein sequences of other members of the same AtCsl family,
searching for new intron-exon boundaries. Re-annotated sequences were translated and

subjected to multiple alignments again.

EST database search

All AtCsl genes were searched using BlastN against the NCBI EST database-
others in July 2001 and updated a year later. Sequence alignments of AtCsl genes with
their EST clones were carried out with Multalin and manually examined. EST clones
showing more than 90% overall identity and those containing mismatches only in a single

region were selected as being derived from a particular gene. All EST clones found for

105

one particular AtCsl gene were aligned to make sure they all contain the same sequence

at either the 5’ or 3’ end.

Arabidopsis MPSS database search

A custom-made java script (Qiong Wang, Ribosome Database Center, Michigan
State University) was used to search all AtCsl genes, AtCesA genes, AtFutl, AtXTl and
AtGalT against the Arabidopsis MPSS database (http://mpss.ucdavis.edu/noniava.html)
in February of 2003. All related information about the MPSS signatures for each gene
query were extracted and input into a Microsoft Access database. The sequence and
position of each MPSS signature were manually examined, and only the true ones were

picked up according to instructions from the database website.

Microarray data retrieval and analysis

A custom-made S+ script (Rodrigo A Gutierrez, Department of Biochemistry,
Michigan State University) was used to search all AtCsl genes, AtCesA genes, AtFutl,
AtXTl and AtGalT against a custom microarray database made from the SMD
(http://genome-www.stanford.edu/microarray) raw data (Gutierrez, 2003). The custom
microarray database was constructed by selecting data only from good quality microarray
hybridizations considering the spatial bias of slides, R-square value of hybridizations,
data integrity, spot quality parameters, and reproducibility (Gutierrez, 2003). Data from
tissue expression profile arrays were analyzed separately. Data from experiments with no

hybridization repeats, and experiments comparing gene expression profiles of mutant

106

(non-cell wall related) and wild type plants were not selected. Only data from those
experiments with repeat arrays or reverse control arrays were kept and analyzed.

Data from microarray experiments involving environmental stress, pathogen
attack or possible cell wall metabolism were retrieved from NASCArray at August of
2003 (http://ssbdic2.nottinghamac.uk/narrays/experimentbrowsepl). Hybridization
signals and information of microarray chips were downloaded as cvs files and
incorporated into a Microsoft Access database for each selected microarray experiment
(Qiong Wang, Ribosome Database Center, Michigan State University). Hybridization
signals and chip information of AtCsl and related genes then were extracted, and the log2
ratios of signals from microarray chip pairs were calculated. However, only data with a D

Call of “1” were used.

107

Seqllt’lll‘r

ror-rcdu

. O
m‘vfi" '
‘ullk .l.ld‘~

{Ha
my

RESULTS
Sequence extraction, annotation and comparison analysis
All AtCesA and AtCsl genes were identified through a Blast search against the
non-redundant GenBank database using AtCesAl (Atstl —l) as query and using
information supplied by Dr. Todd Richmond at his website

(http://cellwall.stanford.edu/cesa/index.shtml), mainly the identification numbers of BAC

 

clones that bear one or more AtCsl genes. A simple alignment of the protein sequences
encoded by these cDNA sequences within the same individual AtCsl family showed
many missed N- or C-terminal regions, or missed internal and out-of-place sequence
fragments. An example is shown in Figure 3.1 for the manual sequence re-annotation of
AtCslA genes. The initial protein sequence alignment showed that the 5’ end of the gene
was missing for AtCslA] , 11 and 15; internal sequence was missing from AtCslAlO, 11
and 15; internal insertions were present in AtCslA7 and All; internal mis-annotation
occurred for AtCslA7 (Figure 3. l A). The genomic sequences from BAC clones of those
AtCsl genes were therefore re-examined for intron-exon boundaries and again translated
into protein sequences for alignment. After several rounds of careful manual annotation,
the alignment of AtCslA proteins showed a much improved consistency (Figure 3.18).
Sequences of other AtCsl families were re-annotated the same way. The re-annotated
sequences of AtCle3 and AtCslA9 matched perfectly with their cloned cDNA
sequences, respectively (F avery et al., 2001; Zhu et al., 2003) and, therefore, the re-
annotated sequences seemed to be much more accurate and reliable.

Information of the re-annotated AtCsl genes are summarized in Table 3.1,

showing the gene ID and the clone ID of the BAC clones where the AtCsl genes were

108

 

 

- OQ‘IIDCtr It s fl "l)-nllllIIQIU‘IoIIIIIOOIII‘vIII III'IIIIII ll!l.1|: rl It I a I IIIII'I‘Ij
an“.— ..VM........)z-z .. . — ...uwunqun." ”V. ”JD “nu.“ “Hun" n "Huh nun...T-..p.>>~. a) u .2)..:- ...u->>..- .1 ....u>- ...:~.-3: .5:..:3.:. - lH-nlns-MHIML-uu “.26.... n»,-.cu.u....\.w~.-~:.u ..t....\::flqr.?.;l‘ll
:— .DILCr-ofllnu . II. .-.—.=.-”. - -I)1b.-..5X: - 35> u>n>udw=>>~o ..u - 3.5.6: - -.->>= .9- . -.->->.-:¥> - 34-95..- .3 u. .1 . . . . 8:... -:.:.t.\.?.:
3) .51).. .1....l - a u llllllllllllllllllllll 1.9 IIIIIIIIII 6 lllllllllllllllllllllllllllll O llllllll 6 -.\-\== -\I:=
lllllllll .Ovrwrurlllllllnflflllllll- “Hr-Hilllllan-Mu. l A:- .-\ nun. Aves. a-It urn” nu... :- I
10.! n

6:3 E 2m fleeces? Co 62.3 38880 >6— 3305 $98 0:25 62 E 8e goo 5.5 208

109

 

mo 62.3 252028 :wE noaocm mEom oEE< .mwfizom BEEN—ma 5.8% 5:? 39.22.::1.::E...E_S_.:.F.c.c.EE.332seemocokfldzzv

 

 

52:22 wEms cocwzm 29$ moo—5:68 £305 .CEEm.x2.::smoo.331.38:23:3:_oo_._.x9:c cosmoEmme—o Eccocfiflm
260,—: 8 mEEooom 88m: can .2- _3m~:<v 348054 2:? .3562 :2: co momma ommnfimc Gamma—DO 028E 2: Soc @8888

203 moocoscom ...ezfieagxz =35:— Amv 3:: ES A3 9.82. 825.58 ”Em: .9589:— Smuz ..o LEE—age. "fin charm

O.O..0.OOOHOOUIatto—b..0'05“”..0...“....:C.°C.d’0u,'.°t...l-OIOOO‘OCOCCOOO“.-I-.¥J

HHUJ>H=HOO§II IIIIIII 1.88435:

gigs-.38..

gmm¢g>gLES“.Ema“.Qatar._._.IxE>HO>HOULJA>S>AJudgflthlllll>z§ A!»

< hgue>Odeugstua>d>ddhzuuura>auuaHfiuEHédtllévfiHu—Laih
wwhaahwm>tlg>gg>uanmaauag>4rhxuu>n5>>OUHudg>=HEgdflWLu—mrzeu—Hiwxdzzuul

pagans)“.nah-.3.—uA>L=Jah=a>HAG>rEOOAAguaswxmgfiumllIIXIOgKMJAXm

 

new. 8m cum on... 8m 8m ovm 8m Rm
be..asamxt.

  

cum com owe oov onv :mv omv ovv one owe adv oov
uoaatgerct.uxc.aa~o.x:.aec.usuxzu.auou>caezeu>p_Lex:eoooooe.ax:.ucpa....:uu has 1 ....cgmcduuorctdcuu..oc>............c.:xea>

 

 

uuwxgaakuucxuamlawm¥>m>az>u>u¥384u>B>CAng>hhgxaguigdxz>wchO xmu Lgxhwm—muOH-TIhurt.Amie-mghEAUXXGEHudJEWUHHEu¥ZMEO>
IIIIIIIIIIIIIIIIIIIII WAD—53Hu¥3¢dd¢>fl>¢4fl§m>hhgxgacwigtagm—CHOI «flu L53.ham-mgwzlllkzrlamtwafldhlAn—hzcstilllllllluuzmca.’
QUHIEOSKAOXNUEJngJOOH>A“6.28..h§>¢anlfiu>hh§=aumw24¢muzgucbgHhum-LAD u¢O>SAAUquH3§3>gigOOhngzaiilllwflmeO>
mum-.8366nun—cv—n—an—mevs....5:Hiking-adAO§>CJOEOw>hh§aucwwzglguxhw aha LghmamgalI:32>::5:nghxauxzcilllliliiuuaxca>
uungakm>snflwmLJEX§>O§>>u¥mO~=>§>CS=Om>HhgawcwxacxHu-swchw an“ mm..Hmmm—wxam—bltcuzrzamzuaHuPBUNOCZOIIIIIIIIIIII>¥=O¢O>
Gum-.8866“arty.uhwmdyx>m>nz>>>L¥flgauc¢>cdgw>hpégcwfisuaguchO “but LCZHmmOmwaalI>L=>24wzm6§P—fiuwzcosllllllllllIluxaxua.)
uumzxzaa>¢>¢¥uhw5wzx>¥dmwA>Jm¥2u¥4w¢¢>¢329m>hhmagcgfimszHOChw an.“ up;Karma—>8gllh§>admzwalflhlduwnml> lllllllll 1.5.3329;
aumziaOuKu—G“uhmmawmu—gJOOa»;u¥3¢5m¢m>¢JO=ON>hhmszgcwistzHOChw “hut L¢:hm3>mawhllhu=>=r_wzwgpdaugwgllllllllllllumaxca>
ouwzgau¢>¢¥u¥mmdgmbhJOZJ>>KXSO=JO>K>CJOEOM>~hgngngfimuabazhw “but ugthuwwaglIv.uzrzamlngFEAUXZ¢2>IIIIIIIIIIIIqux=O>

 

 

 

 

 

 

 

 

 

Omn Own Ohm Own Omm Ovn Ohm ONM OAm OOn OmN OON ONN HON
4:.0L22Aaul _ JLoLNOLa noun—Cm: . QDN. u :x>>w£ awesome—auct— CO... o. oACAux 03. 00. {Ace o . o o . . o . . . . .atautfiAar.A L! A: A. 03$de . .30 83-5
#:wmzzHaum>hflwa>nkw9ukOCOLHCHC Ollhx>>w=~=~g¥AGO—mu.>Oz¢lznzxwuxHZHuxhczvfl—n—mhmsim IIIIIIIIII :Hhmahmgdagnaznr—EIHdugunmdxu>wui>lm
JEELZZHJ maAGSA>OmmE gunman—AGE: OIlax.»>mzmzm§x4¢u¢¥>wzzlwaxEZHzgxmcahumtm>dm IIIIIIIIII zHcLOhwgggHAuaahamuguHWAwL>gi>§
.EJv—LzzHa lepiaunmam u EOE—LAC): OIIax>>m=mzuwMA=O¢¥>uz¢lznzHup:Hz>umxw3~50uxx>dwa¥xa>>uazhchmnthOd>OHHH¢2walmH¥O¢¢¢Hmauu>“wilt...
ACDsz—Saunngad . : .d a 5:: . q > > . allax>>z=xzuwxLCM—3x>OZKIZOEKBXHz>u¥mmguwhz>Ju IIIIIIIIII XHhanhmgrnggadnguccuHm34>u¥ma>xm
4"me zz>qul>h8mqmam we yucca—uh; 1.” thun>>w=¥2m§h acguawuuuazbmmxunh wawgmmww IIIIIIIIIIIIIIIIII thA>OHHzxna§flwd¥UguHmflange—wot?—
AMHEZZHALLHthauOLOo ”75:: m Acslm Ollax>>z§28¥Ach>uH~TZOKH wthzHuxmcgum¢2>Jw IIIIIIIIII xmcmnhmgggHrfi—Odg EECOHOSU.» uxw—‘zm
4:..KL2:..A>L>-§JRO:M : 5:: 3:. . q allmx>>mxxzamx Amway-flailZOKHN>¥HZ>O¥mE§muw=u>S IIIIIIIIII ¥HhLOhMOOJsEH>H30m§wduu¢cmeda>>wxwzmzl
AcHzmzznamn—Hwfi: .: . d . a 5:: .4 > > A n J|I=x>>m¢¥ Luwx decay—>6)filzmeO>8H2Ha¥m¢§Ouw>w>5 IIIIIIIIII ¥>PmOthOA>OH>JKQw§wduccuHmd¥>>wmm2L=L
JC>OLZZ>JHL>¢3HJ . : ... . a 5:: :5. .q JIIOX>>MOHA¢M¥2

 

 

 

 

 

 

 

 

8N SN OWN OmN ONN OAN OON OmA OOH ORA Om: OmA OVA AMA
dgflioo UHUuu‘I OGISJcoXtIXu-cbfigoo; ..- fl

 

Hggguawzlnsplyxrgwgugagg

Hg§$60w>lgmiw3¥>§ugdg§w>ugmz
Hgg...uTu..uHIOgua¢>>¥§u“SEE-In»HOWHn5.—deHan—LggEdeEwu—h:hhnumHsgmhuggwwmusn—EdmJadasau—Hdgduhh:
HOHgum—3WHSwXumUs.dead—had65Hgang"...#9590553.—HgduHulhamEuH>ud>llm§ugdhi
HanghZIGJm—ABEHLgi>¥flh§gh§>>Hm§mHfig>dHh>d¥JHLHL>LHIbOxxH>HHmuHHIlalgghh§4h§>dmummsmd>niduduhxdu§
Hangcmzlm—JWIHOfixHawfixuxih—gugdgHggwbgghdungfigHatflHggfi—UzhHIIOIOu—Igwhbsdu.
H315?—blamin—..WIgiHmmuguguggH)Hui—Emuggjagggggfithllgzl—LHEEEO:

OMA ONA OAH OOH O0 O0 Oh Ow Om 8 O6 ON OA A

O8AA§\NO¢
AOhahaude—u

AOhOhauAOc

nip-Ign—
O88§\AA¢

AOhOhOcude-u

110

t o o . o o o .3335. .96. .btouaAbA>>NAo o oax>bu>tbutAA>.m->¢. m
“whablhmmzlgrguguEHWOALA>>>A>hBOL>uO>>OUHLJ>IO>IHw

hOHu—OHOCHHt4§>J>H>d>h¥On$rl>mCUUJH>HO>£w
hgkguwHA§§LJ>quhzuuud>nluu4H>HO>2Hu
hO>uOuOm>LtbSAud>gah32h a> Eu :EOEM

 

 

Whm Ohm me 0mm Ovm Ohm ANO-

 

OAm OOW. Ow? OOV Oh? Ow? Om? 8' ON? ONV OAV OOV

>L¥§Hm>x¥2m HA wF—tvfl.Lg8uw3¢100uzrtxmxmta§>haa§>>LXIm-xdmgmgdsghhflmgggmufluzzfihwz”my...glhmamwagxuzrzdtytlhzduxzos

 

Omn Own Ohm Own Own Ovm Omm ONM OH” OOm OON 8N OhN Hum

 

 

 

OwN OmN OVN OmN ONN OAN OON OmA OOA ORA OOA Om..." OVA Am.“

 

 

«Op-Anode

“an-8:09

 

 

 

 

ddflfld HQ

 

«9:09.8-

111

Table 3.1: Summary of re—aunotated AtCsl gene. The gene ID, BAC clone ID were
found, and the match with corresponding cDNA clones found in the NCBI database and
the RH<EN cDNA database are shown. “M”: match; “nM”: not match; “no”: no cDNA
clone available at this time. All information was extracted from the public database.
Protein sequences encoded by the corresponding cDNAs were aligned to examine the
sequence match. Protein sequence alignments were carried out with the MultAlin program

(http://'prodes.toulouse.inra. fr/multalinx’multalin.hlml) using default parameters.

 

112

ne ID were
database and
': no cDNA
ambase.

mine the

41in progfii

TS.

 

 

 

 

 

 

 

Gene Gene_ID BAC_ID NCBI RIKEN
AtCslAl At4g16590 dl4320w nM M
AtCslAZ At5g22740 MDJ22.16 M M
AtCslA3 Atl g23480 F28C11.11 nM M
AtCslA7 At2g35650 T20F21.16 M M
AtCslA9 At5g03760 F17C15.180 M M
AtCslA] 0 At1g24070 T23E23.23 M no
AtCslAll At5g16190 T21H19.110 nM no
AtCslA14 At3g56000 F27 K1 9.1 80 M M
AtCslAlS At4g1341 0 T9E8.150 nM no
AtCle4 At3g28180 MIGl 0.8 M M
AtCleS At4g31590 F28M20.220 M M
AtCle6 At3g07330 F2103.4 M M
AtCle8 At2g24630 F25P17.7 M no
AtCle12 At4g07960 F1K3.3 M M
AtCslBl At2g32610 T26B15.17 nM no
AtCslBZ At2g32620 T26BlS.18 M no
AtCslB3 At2g32530 T26B1 5.9 M M
AtCslB4 At2g32540 T26B15.1 0 M no
AtCslBS At4g15290 FCAALL.256 nM M
AtCslB6 At4g15320 FCAALL.268 nM no
AtClel At2g33100 F25118.1 6 M no
AtCleZ At5g16910 F2K13.60 M no
AtCle3 At3g03050 T17822.26 M M
AtCle4 At4g381 90 F20D10.310 M nM
AtCleS At1g02730 F22D16.26 M no
AtCle6 At1g32180 F3C3.4 M no
AtClel At1g55850 F14J16.9 M M
AtCslGl At4g24010 T19F6.17 nM no
AtCslGZ At4g24000 T19F6.18 M M
AtCslG3 At4g23990 T19F6.19 nM no

 

113

IOU

AA

'3\C=

{UT-15‘.

found. The corresponding cDNA clones available for AtCsl genes were also identified
from the NCBI database and the RIKEN Arabidopsis Full Length cDNA database
(http://pfgweb.gsc.riken.go.jp/pub_data/index.html). The alignments of protein sequences
encoded by the re-annotated AtCsl cDNA coding regions with their counterparts from the
NCBI and RIKEN database indicated that most of the re-annotations were accurate
(Table 3.1). Even at the present time, there are still many mis-annotations for AtCsl genes
in the NCBI database. For example, the re—annotated sequence of AtCslAl and AtCslA3
matched with the corresponding cDNA sequences from the RIKEN database perfectly,
but their sequences from the NCBI database still show missing domains (Table 3.1).

Genes within the same family showed random chromosome locations with the
exception of AtCslB and AtCslG families (Figure 3.2). AtCslBl and B2 were found
adjacent to each other. The same was true for AtCslB3 and B4 and for AtCslBS and B6
(Figure 3.2). Furthermore AtCslBl and B2 are very close to AtCslB3 and B4, located on
the same BAC clone T26B15 (Figure 3.2). All three AtCslG genes were found together
on BAC clone T19F6 as a tandem array (Figure 3.2). This phenomenon could be due to
gene duplication events late in the gene evolution pathway. AtCle genes showed a very
interesting pattern in that AtCleS and D6 were found on chromosome 1, but at a
considerate distance apart, and the four other AtCle genes were each on a different
chromosome (Figure 3.2). This predicts that it should be relatively easy to obtain plants
with multiple AtCle genes knocked out by crossing plants bearing single AtCle gene
disruptions.

The intron-exon structures of genes within the same AtCsl family were highly

conserved (Figure 3.3). This was easily seen for genes of family G, B and C. AtCslGl,

114

6:303 mnnz Sat 3:63 803 8:03 U<m no action 05. .82: wEESEo 8:20 U<m of .3 20:82 38808030

2: 8 368on @8158 $3 meow EU? :03 .8 notion Baggage 2:. .moaow 7.03. .3 28:32 .aEemcEcEU "N6 «Sufi

329%
252

SEQ
8.62

GER:

2232 \I

@8sz

$23022 I:

2.62
320”: M
Ammo: \

 

 

>—

858":
332

I

5%”:
€7.32

62%,:

I

320“:

m _<_mU~<

5:”:
22.3.

 

 

:—

83:9:
3252

86:6
352 l

205;
532

$85.:
2:82

/i

 

 

=

A _ Eon:
2.62

$5.9:
2:62

 

 

 

5mg:

 

,:/.

/
2

wrzmbja

 

 

6:3“:
3:62 ll

Gun:
@962

9.88:
2252 III

$03
.252

GENE:
852 ll.

 

 

115

Figure 3.3: Predicted intron-exon structure of AtCsl genes according to the re-
annotated sequences. Some exons were numbered and linked with each other to indicate

their sequence equivalences as described in the text.

116

AtCslAl
AtCslA2
AtCslA3
AtCslA7
AtCslA9
AtCslAl O
AtCslAl 1
AtCslAl4
AtCslAl 5
AtCle4
AtCle5
AtCle6
AtCle8
AtCle12
AtCslBl
AtCslB2
AtCslB3
AtCslB4
AtCslBS
AtCslB6
AtClel
AtCle2
AtCle3
AtCle4
AtCleS
AtCle6
AtClel
AtCslG]
AtCslG2
AtCslG3

 

,i

 

I-fl-I-l—I-I—I—I-I-I
EI-I-I—I—I—I-I-I

 

100 hp

iii?»

 

117

G2 and G3 all contained five exons, and both the length and sequence of each exon were
very similar through out the family. The same phenomenon was observed for AtCslB and
AtCle families except that AtCslBS was missing the first exon found at the 5’ end of
other AtCslB genes, and the third exon of AtCle4 was the fusion of exon 3 and 4 of
other AtCle genes (Figure 3.3). The intron-exon structures of genes from AtCle and
AtCslA families were more complicated. After careful analysis, it was found that
AtCle2, D3, D5 and D6 seemed to have the most straightforward structure, containing
three exons, with the 5’ end of the first exon of AtCle6 missing (Figure 3.3). AtClel,
however, had 6 exons, but its exon 2 and 3 was the equivalent of the second exon of
AtCle2, 3, 5, 6; and its exon 4, 5 and 6 were equivalent of the third exon of AtCle2, 3,
5, 6 genes (Figure 3.3). AtCsl D4 also only showed three exons, but was more
complicated in that its second exon could be considered as three pieces with the first
piece as the equivalent of the 3’ part of the first exon of AtCleZ, 3, 5, 6; the middle
piece as the equivalent of the second exon of AtCle2, 3, 5, 6; and the 3’ piece when
fused with the third exon became the equivalent of the third exon of AtCle2, 3, 5, 6
(Figure 3.3). The intron-exon structure for AtCslA family was even more complicated.
As indicated in Figure 3.3, exon 5 of AtCslA3 was the equivalent of the fusion of exon 5
and 6 of AtCslAlO, and the fusion of exon 4 and 5 of all other AtCslA genes. Exon 6 of
AtCslA3 was equivalent to the fusion of exon 6 and 7 of AtCslA], 2, 9, 11, 14 and 15, or
exon 7 and 8 of AtCslAlO (Figure 3.3). Exon 7 of AtCs1A14 was equivalent to the fusion
of exon 7 and 8 ofAtCslAl, 2, 9, 11, and 15, or exon 8 and 9 ofAtCslAlO (Figure 3.3).
The diversity of the intron-exon structures of genes from AtCslA and AtCle family,

combined with their dispersed chromosome locations, strongly suggested that AtCslA

118

and AtCle were more ancient gene families and they have gone through more
evolutionary modifications.

The intron-exon structures of most of the re-annotated AtCsl genes matched with
Dr. Todd Richmond’s observations (http://cellwall.stanford.edu/cesa/index.shtml).
Exceptions were noted from this database for AtCslAl , A3, A10 and B5. Compared with
the re-annotated sequences described here, AtCslAl showed a size discrepancy in exons
3 and 4, AtCslA3 was missing the first exon and part of the 5’ end of the second exon,
AtCslAlO was missing the first exon, and AtCslBOS had an extra exon in the 5’ end.
Because the sequences described here for AtCslAl, 3, and B5 all matched with the cDNA
sequences from the RIKEN database, and sequence for AtCslAlO matched with the
cDNA sequence from the NCBI database (Table 3.1), the annotations presented here for
these four genes seemed more accurate.

The re-annotated cDNA sequences of AtCsl genes were translated into protein
sequences. The protein length and GenBank ID are summarized in Table 3.2. All AtCsl
proteins were predicted to have multiple transmembrane domains, with 1-3 near the N-
terminus and 3-6 close to the C-terminus (Table 3.2). A phylogenetic tree was
constructed for all AtCsl proteins (Figure 3.4). AtCslA and AtCle grouped together
whereas all other AtCsl proteins were grouped into another major clad. With all the
corrections of the gene annotations, the assignment of proteins to a family were still the
same as that done earlier (Richmond and Somerville, 2000; Richmond and Somerville,
2001). The relationships within each family were also mostly the same, except that
among the re-annotated sequences AtCslAl 5 was the closest relative of AtCslAlO

(Figure 3.4) instead of AtCslAll as shown before by Richmond and Somerville (2000,

119

Table 3.2: Characteristics of AtCsl proteins. Protein ID, predicted length, and the

predicted number and locations of transmembrane domains are included. All information

was extracted from the public database with manual annotation. Transmembrane domain

were predicted with server TMHMM (hup:f,"\m'w.cbs.dtudks‘scrvicos/TM1-1Mf\1/) with

default parameters.

 

 

Protein

Length

Protein_ ID

Transmembrane Domain [number (position)]

 

AtCslA]
AtCslA2
AtCslA3
AtCslA7
AtCslA9

55333
534213
556aa
5563a
533aa

AtCslAlO 553aa
AtCslAll 50033
AtCslAl4 53533

AtCslA 1 5

503321

CAB78701.I
BABl 1680.1
AAF79586.1
AAD15455.1
CA882941.1
AAF87149.1
CAC01860.1
CA887854.1
CAB40776.1

4 (70-92, 389-411, 431-453, 513-535)

4 (34-56,356-385,404-426,509-531)

5 (57-79,386-408,427-449,509-531 )

5 (72-94, 389-41 1, 426-448, 500-522, 527-549)
5 (35-57,363-385,404-426,481-503,507-529)

5 (70-92,394-415,430-452,51 1-529,533-551)

5 (l2-34,332-354,380-402,458-476,480-498)

4 (43-65,36 l -383,402-424,502-524)

4 (19-41,332-354,380-402,464-486)

 

AtCle4
AtCleS
AtCle6
AtCle8

AtCle12

673a
692a
6823a
690aa
6993a

BABO 1 433.1
CAA19764.1
AAF02144. 1
AAD23884.1
AAD15482.1

4 (469-487,494-512,531-553,628-650)

6 (67-85,105-127,178-200,488-510,522-540,647-669)
6 (67-85,106-128,l71-193,485-507,5 16-534,637-659)
6 (67-85,105-127,l 78-200,491-513,522-540,645-667)
5 (122-144,192-214,502-524,536-554,673-695)

 

AtCslBl
AtCslBZ
AtCslB3
AtCslB4
AtCslBS
AtCslB6

75 7aa
757aa
755aa
755a
664213
75933

102560
T02561
T02552
T02553
CAB78571.1
CAB78574.1

8 ( 15-37,49-71 ,525-547,560-578,597-619,662-684,699-721,728-746)
8 (20-38,47-69,533-555,570-592,605-627,67 l -693,706-724,733-755)
6 (7-25,491-513,526-544,628-650,657-679,688-710)

8 (20-42,51-69,531-553,572-594,614-636,671-693,700-722,731-753)
3 (622-644,655-677,686-708)

8 (24-42,49-71,108-126,264-282,3 l6-338,620-642,657-679,686-708)

 

AtClel

AtCleZ
AtCle3
AtCle4
AtCleS
AtCle6

1037aa

ll45aa
1145aa
llllaa
1181aa
979aa

AACO4910. I

CAC01704.1
AAF261 19.1
CAB37559.1
AAF02892.1
AAGZ3436.1

8 (179-201,208-230,815-837,848-866,885-903,936-958,962-984,1004-1022)

8 (288-310,323-341,923-945,952-974,993-1015,1046-1064,1079-1097,11 10-1 128)
8 (285-307,320-338,923-945,952-974,993-l015,1046-1064,1079-1097,1l 10-1 128)
8 (263-285,298-316,884-906,91 3-935,954-976,1009-103 1,1040-1058,1071-1089)
5(958-980,992-1010,1081-1lO3,1116-ll34,1144-1166)

7 (1 18-140,147-166,752-774,787-809,880-902,915-934,944-966)

 

AtClel

730aa

AAF79313.1

6 (29-51, 64-83, 51 1-533, 553-575, 645-667, 677-699)

 

AtCslGl
AtCslGZ
AtCslG3

771 aa
730aa
747aa

AAB63622.1
AAB63623.1
AAB63624.1

6 (ZS-50,537-559,566-584,603-625,659-681,693-71 1)
7 (21 -39,5 1 -69,520-542,554-5 76,641 -663,675-697,7 1 0-728)
8 (29-5l,55-77,531-553,572-594,607-625,664-686,693-715,727-745)

 

120

3:63 33:333.. .3 5:83: 2: 88:2: 8.: of 3 Eaton 2: 2 $3: 65 253 .68: 8:258
5036: 8:82: 05 $5638 6:285 3 :3 some .3 532 2C. .Aommfim $3628 :32on GEES: Rams/5 05 wfim:

3&3 98 60:02:33 <ZQo 30:30:56-8 on: .3 330:0 8: 363.5 8503.3 303. .3 3.: 33:99.13: 33?. 2. "Wm «Ema

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o 8 2. 8 cm 2: as o: OS 2: $2
662
662 825 82.3335 22 SE2
26.62
662
662
262
262
262 iullT—
3262
262
E62
:262 ,
23E -
22.62
3:62
862 1..__
8.62
2962 J
3:62 .iITIT
8.62
2.62
8.62
362 II_
3.62
862
862
862

 

 

2:62 UJ 1
E62
E62 l||||T_

 

 

121

2001). Another multiplealignment including AtCesA proteins showed that AtCle family
had the closest relationship to AtCesA showing sequence identity of 39.1-45.9%, while
AtCle family was the most distantly related with AtCesA proteins, showing sequence
identities of only 73-12% (Table 3.3). The highest sequence identities shared among
AtCsl protein families is between AtCslA and AtCle families at 36.3-43.5% (Table 3.3).
Two interesting genes were found during the Blast searches for AtCsl genes, a 6-
(l, 3)-glucosyltransferase from Bradyrhizobiumjaponicum (AAC62210) and a secondary
xylem cellulose synthase from Populus tremuloides (AAD03417). When the B—(l, 3)-
glucosyltransferase was used as the query for a Blast search against the non redundant
database, all AtCle and AtCslA genes were found in the hit list with E values ranging
from 2e-22 to 4e-16 but none of the members from other AtCsl gene families showed any
significant similarity. When the secondary xylem cellulose synthase was used as query
for the Blast search, all AtCle, AtCslB, AtCslG and AtCslB genes were found in the hit
list with 13 value ranging from O to 2e-61, but none of the AtCslA and AtCle genes were
found. Based on this analysis, AtCsl genes can be grouped into two categories, with
AtCslA and AtCle in one group and AtCle, B, E and G in another. This observation
leads to the prediction that the two categories may have distinct functions, but further
work will be needed to confirm this prediction. The sequence divergence showed from
the phylogenetic tree (Figure 3.3) fits this assessment well. Proteins encoded by AtCle,
B, G, E genes might synthesize B—(l, 4)-glycosidic bonds or using UDP-glucose as
substrate as AtCesA proteins do. The Bradyrhizobiumjaponicum 6—(1, 3)-
glucosyltransferase encodes an enzyme responsible for the synthesis of a 6-(1,3)- (1,6)-

glucan polymer, suggesting AtCslA and C proteins on the other hand might encode

122

Table 3.3: Summary of sequence identities among all AtCsl proteins. Re-annotated

cDNA sequences of AtCsl genes were translated into protein sequences and aligned with

the MegAlign program of DNAStar (Clustal W). The highest and lowest number of

protein identities from the comparisons of different proteins within the same family or

between two families were selected as the upper and lower range.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AtCesA AtCslA AtCslB AtCle AtCle AtCle AtCslG]
AtCesA 61.1-90.8 8.4-14.9 29.4-33.2 7.3—12.o 39.1-45.9 31.9-33.7 29.3-32.7
AtCslA 53.2-79.6 6.5-14.9 36.3-43.5 7.0-15.9 8.6-14.4 7.0-13.5
AtCslB 73.8-89.5 6.5-11.4 26.4-30.6 29.6-31.2 24.8-30.8
AtCle 59.8-88.1 6.4-12.9 7.4-10.8 7.3-10.1
AtCle 59.0-87.9 29.9-32.3 25.9-30.2
AtCle 35.5-36.3
AtCslG 79.7-80.7

 

 

123

enzymes responsible for sugar linkages different from B—(1,4) or using nucleotide sugar

substrate different from UDP-glucose.

Expression patterns of AtC sI genes revealed by EST abundance

While the functions of all AtCsl genes remain unknown, the expression patterns
of these genes were examined in different tissues, hoping that this information will shed
some light on their functions. The EST database was searched for the expression of AtCsl
genes associated with specific tissues or grth conditions. All EST clones available at
the time (July, 2002) for AtCsl genes were manually examined and only those with more
than 90% overall identity and no multiple mismatches were selected. AtCsl genes were
found to be expressed in a wide range of tissues such as siliques, flowers, roots, rosette
leaves, deve10ping seeds, soil grown or liquid grown seedlings etc (Table 3.4). AtCsl
genes within the same family showed different numbers of EST clones derived from
them. Those EST clones also come from a wide range of different tissues (Table 3.4).
AtCslAl, AtCle4 and C6, AtCle2 and D3, AtClel all showed more than 9 EST
clones in the database, indicating that these AtCsl genes might be expressed more
abundantly than other AtCsl genes (Table 3.4). Sequence alignments of EST clones with
AtCsl genes also confirmed the accuracy of many re-annotations of AtCsl gene ends
(data not shown).

Altogether there are 106 EST clones identified for all AtCsl genes. AtCle and D
family have the most EST clones with 31 for each. AtCslB and G have the least, only 2
and 4 respectively (Table 3.5). When the number of genes within each family is

considered, there are 3.5 EST clones on average for each AtCsl gene when all 30 AtCsl

124

 

 

 

 

 

 

 

&:::08 8:003 :-~ 8&3 :38w 0.3:: v mgmuz.
838 8:3: 33:: 803:8 :00& 8200? 3A.. 8&3 :::8w 0.6:: A: 5302
&:::08 33:30:32— 838 m mauve
8032.: :3 803:8 _ 2:802
&:::08 80:88 8:08 8320.0: 803:8 :00& 838 8:003 :-~ 8&3 :::8w 03:: m: MUD—80:2.
80:88 838 8:08 83203: 8:003 :-N 8&3 :::8w 0.3:: : N963.
8:: 323: _ 3:805.
838 83203: _ 20.802
838 83203: 80:88 N wUEUE
&:::08 838 803:8 :00& 8:3: 3.3: 8:003 :-~ 8&3 :::o& 03:: N: emu—mu}.
&:::08 80:88 8:08 83203: 82003 :-~ 8&3 :::8w 03:: : m0_mU:<
&:::088_bo3§: 80.32 838 803:8 :00& 8:003 :-~ 8&3 :::8w 0.3:: o: «.2802.
838 _ 3:62
80:88 _ 2:802.
838 _ 3562.
80:3 :3 8:3: 323: m 23803..
5.6: Ba 2.3:... H 2262
83:8 :00& : 3:80:82
&:::08 m 52:80}.
3:3 :3 88:38 v 9:65.
3:3 :3 80:88 838 : ~<_mo:<
&:::08 8:3: 330: 803:8 :00& a 3802
80:20 Hm: 3 083m 80:20 Hm: 0:00

 

.m . :0 3D .3 .v .N . E £32803. 8320:: 80:20 .5: .8: .3 338038 3: 803 :08: 3: m0:& 805.. 83.:

:0>3: 0:: 3:: 80:3 80:8: :3 80:20 km: 3 80:8: Z 80:& .803. 3.. ham—£0 0:: 2 35302 Mama .3 88555 3.». 23,—.

125

Table 3.5: EST abundance for AtCsl gene families. Numbers of EST clones for each
AtCsl family are summarized from Table 3.4. Percentage of EST clones is calculated as

the percentage of EST clones for each AtCsl family out of the total EST clones for all

 

 

 

 

 

 

 

 

AtCsl genes.

Gene Number Number of Percentage of Average EST
family of genes EST clones EST clones clones per gene
AtCslA 9 28 26.4% 3.1

AtCslB 6 2 1.9% 0.3

AtCle S 31 29.2% 6.2

AtCle 6 31 29.2% 5.2

AtCle 1 10 9.4% 10

AtCslG 3 4 3.8% 1.3

Total 30 106 100% 3.5

 

 

 

 

 

 

 

126

genes are included (Table 3.5). AtCle family showed the highest expression as it has 10
EST clones per gene. AtCle family follows it with 6.2 EST clones per gene, and AtCslB
family showed the least expression with only 0.3 EST clones per gene (Table 3.5).
Therefore AtCle, C and D family genes showed relatively high expression among AtCsl
genes indicated from the EST clones identified, while AtCslB and G genes showed
extremely low expression that were even difficult to detect by RT-PCR reactions (data

not shown).

Expression patterns of AtCsl genes revealed by MPSS database

The Arabidopsis MPSS database was searched for the expression of all AtCsl
genes, together with AtCesA genes, AtFutl, AtXTl and AtGalT genes. The comparison
of expression patterns of AtCsl genes with that of these genes that are known to be
involved in cellulose and XyG biosynthesis might yield some clues about the functions of
AtCsl genes. Most AtCesA genes showed reasonable expression with appearance of their
gene-specific MPSS tags ranging from 50 PPM (parts per million) to 250 PPM, while
AtCesA3 showed the highest expression, especially in flowers, at 1225 PPM (Table 3.6).
AtCesAl, 3, 5 and 6 all showed significantly more MPSS signature hits than AtCesA4, 7
and 8 (Table 3.6). Considering the fact that all five MPSS libraries were made from
vigorously growing tissues where primary wall is still the main wall biosynthesis event
(httpzl/dbiXSOOl .dbi.udel.edu/MPSS4/java.html, actively growing 10 weeks old callus,
mixed-stage buds and immature flowers, 14-day-old leaves, l4-day-old roots and siliques
one to two days post fertilization), this different expression pattern for different AtCesA

genes is consistent with the observation that AtCesAl , 3, 5, 6 are cellulose synthase for

127

 

 

 

Table 3.6: Numbers of MPSS tags that found from five different Arabidopsis cDNA
libraries for all AtCsl genes and AtFut], AtGalT and AtXTl. Numbers are in PPM
(parts per million) and were normalized according to the MPSS database

(http://’dbixs()()l .dbi.LulclcdufiVlPSS-‘lr’javahtml). All MPSS tags were manually

 

examined for their accuracy (Materials and Methods). Note that the tags for AtCslBS and
B6 could not be differentiated. Also AtCle2 showed two different tags that both seemed
valid according to criteria from the database, and therefore were both retained and named

as AtCle2 and AtCleZ’.

128

 

 

 

 

 

 

 

 

 

Gene_name Callus Flower Shoot Root Silique
AtCesAl 184 238 262 146 128
AtCesA2 4 120 64 48 34
AtCesA3 237 1225 366 604 451
AtCesA4 O 104 8 61 28
AtCesAS 38 190 153 75 61
AtCesA6 42 23 144 1 56 80
AtCesA7 10 9 5 47 38
AtCesA8 O 46 4 32 39
AtCesA9 O 24 3 2 2
AtCslAl 0 141 O O 18
AtCslA2 45 316 45 150 86
AtCslA3 1 1 1 15 14 7
AtCslA7 0 O 12 O 22
AtCle9 0 1 1 8 O 4
AtCslA 1 0 0 328 4 O 52
AtCslAl 1 3 1 19 O 7 14
AtCslA14 O 3 O 28 3
AtCslB3 O O O O 8
AtCslB4 O 4 4O 0 10
AtCslBS/B6 O 12 O O O
AtCleS 33 20 43 10 55
AtCle6 23 39 12 31 35
AtCle8 2 18 2 0 O
AtCle12 O 36 0 12 1
AtClel 2 25 O O 4
AtCle2 8O 10 1 1 22 O
AtCle2' 35 O 8 77 10
AtCle3 100 90 13 1 13 18
AtCle4 8 6 3 243 32
_AtCle5 52 81 14 12 13
_AtCle1 296 90 73 32 32
AtCslG] 0 14 1 O O
AtCslG2 8 53 O O 59
AtCslG3 27 38 3 O 6
AtFutl 9 45 27 50 29
AtGalT 14 O 2 4 19
fiXTl 154 61 49 16 54

 

129

primary walls expressed in expanding tissues, while AtCesA4, 7 and 8 are cellulose
synthases for the secondary walls expressed in developing vascular tissue (Williamson et
a1 - , 2002).

The overall expression of AtCsl genes is generally lower than that of AtCesA
genes as seen from the frequency of appearance of their gene-specific MPSS tags (Table
3 .6; Figure 3.5). AtCslA2 showed the highest overall expression with 316 PPM in
flowers and 150 PPM in roots (Table 3.6; Figure 3.5). AtCslAl, 10 and 11 also showed
moderate expression in flowers, at 141 PPM, 328 PPM and 119 PPM respectively (Table
3 . 6; Figure 3.5). Among other AtCsl genes, only AtCle3 in callus and roots, D4 in roots
and AtClel in callus showed an expression level of more than 100 PPM (Table 3.6;
Figure 3.5). AtCslB genes showed the lowest expression with a maximum value of 40
PPM in shoot for AtCslB4 (Table 3.6), which is consistent with the difficulties accounted
during the amplification of AtCslB transcripts by RT-PCR (data not shown). The MPSS
data also supported a root-preference expression of AtCle genes, including D2, D3 and

D4 (Table 3.6; Figure 3.5). AtCle2 and D3 also showed good expression levels in the

callus, indicating that they are responsive to environmental stimuli (Table 3.6; Figure
3.5)., since callus is a tissue type develops as a response to injury caused by physical or
chemical means and contains cells that are differentiated but unorganized

(http 3//Www.agls.uidaho.edu/plsc2O2/pdf/02notes-8.pdf).

AtCle proteins are the closest paralogs of AtCesA proteins within the AtCesA

superfalnily, but their sequence identities with AtCesA proteins are lower than the

identities shared among different AtCesA proteins. Since XyG has the same backbone

Chemical structure as cellulose, AtCle proteins are good candidates for encoding the

130

131

.Eon .5382 2 3:3an 20>» 98 850 some Soc woamccubtmu

3 Ho: 2:8 om new mmfiua e8 mwfi Esficflm .835 memoz E0832 was memuz no.“ mwfi Bancwmm 03“ 26>» 82: “m5

mccoz .m.m 2an Eot #835,523 2 8mm AHX~< €303. .:=m~< tau 35w 303‘ =a .3 2:9:— Ezmmoaxo mmm—Z ”ma v.5”:—

 

 

113 Flower ‘
i. Root i

lion]... 1
‘ I Shoot

‘I Silique 1

LLXIV
mow
Unfilv
EDISOIV

E zetsozv
Iolsow

IEIISDIV

 

9 CIISOJV

 

valsolv
SCIISDIV
.zcnsolv
3 zcnsow

I.
tatsow
ZIDISDJV
SOISDIV
931531V
SOISOIV
9/981831v

II.
SHISOIV
1 v I VISOW

I [VISOIV

 

—
0 [VISDJV

 

6V1531V
LVISDIV
{VISDIV

ZVISOJV

 

 

350

(de) uonuw 19d Suva

l a

[VISOJV

l

 

200 .

150

100
50 a
0

132

XyG B-(l, 4)-glucan synthase. Therefore the expression of AtCle genes were compared
with that of genes involved in XyG biosynthesis, AtFutl, AtGalT and AtXTl. However,
no obvious common expression profile was observed (Figure 3.5). This did not invalidate
the hypothesis though, since gene redundancy brings in complexity when gene expression
patterns are compared among gene families.

Two MPSS signature tags were recorded for AtCle2, one is 27 bp upstream of
the stop codon, and the other one is 143 bp downstream of the stop codon into the 3’UTR.
After careful analysis, no criteria could be used to eliminate either one of them as a
technical error. On the other hand, alternative splicing is always a possibility to generate
two different transcripts from the same gene. Without further experimental investigation,

both signatures were retained.

Expression patterns of AtCsl genes revealed by tissue panel microarrays from Stanford
Microarray Database

Genes involved in the biosynthesis of the same plant cell wall polysaccharide
might have similar expression pattems in different tissues. Any AtCsl gene encoding the
XyG glucan synthase might have similar expression profile with others involved in XyG
biosynthesis, i.e. AtFutl , AtGalT and AtXTl. Therefore the expression profiles of AtFutl,
AtGalT, AtXTl, AtCesA genes and all AtCsl genes were examined by analyzing
available microarray data comparing gene expression in different Arabidopsis tissues.

AtCsl gene expression data were extracted from microarray experiments
comparing gene expression levels among different tissues of Arabidopsis from the

Stanford Microarray Database (SMD). Only statistically significant and repeatable data

133

were selected (Table 3.7) after considering array quality, reproducibility and statistical
analysis (Gutierrez, 2003). Five AtCesA genes, AtCesAl, 3, 5, 6 and 7, showed a strong
expression bias in stems (Table 3.7). AtCesAl and 7 also showed higher expression in
siliques (Table 3.7). This observation is consistent with the concept that stems contain
large amounts of cellulose compared to other tissues. Whether the same is true for
siliques still remain to be seen. AtCesA2, however, showed an opposite trend with less
expression in siliques and roots compared with whole seedlings (Table 3.7). This
suggests that AtCesA2 has tissue specific expressions different from most other AtCesA
genes.

Among the expression profiles of AtCsl genes, a few tissue specific expressions
were noted (Table 3.7). Firstly AtCslAl and C5 showed higher expression in flowers
(Table 3.7). Secondly AtCslBl was expressed higher in both flowers and stems (Table
3.7). Finally AtCslAl and D2 showed higher expression in roots while the expression of
G3 was lower in roots (Table 3.7). These observations might suggest the existence of
gene redundancy of AtCsl genes, and different members within the same AtCsl family
might have different tissue preference of expression.

The expression of AtFutl and AtGalT, however, did not show very strong tissue
expression preferences, with maybe slightly higher expression of AtFutl in flowers and
AtGalT in siliques and stems (Table 3.7). This made it rather difficult to compare the
expression pattern of AtCsl genes with them and therefore no conclusions regarding

AtCsl function can be drawn from this analysis of tissue expression profiles.

Regulated expressions of AtC sl genes revealed by microarray experiments

134

ho 9o- moo owo who- who oos Ko- ao 8.7 :82
ww._ who- oho who- oho- who to woo woo who- who So. can?
8o 6.? Ioho So woo woo who- so So. who 8:02
So. wwo who we. who no.7 who- 8.? oho w_.o- wwo- who 6.52
wwo oho- owo who- loo- vwo- Iooo who- Q82.
So- who- woo who- who- to woo- Eo who So. who who- MQWUZ

Iloho- ooo woo So So who- oho ho- meoZ
llmoo- _ho- o:- Ro owo- wo._ woo- woo- who who- 3302.

no.0 hmd mod ood- cvd mod- mod- 3.10 090 3.0. ovd- vmd vU_mUo<
V0.0 5.0. Nvd- mod mNd- mm .0 Ned. vmd de- mvd mmd Q .o- hSmonw
090- Ed 320. de wmd- no.0 cmd- E .o cmd- Sud mSmquw

  

Efiifiiflaflflifildmvo- oho oho... who 2o.- So. 2302

-.... n‘u-o: ...-1.... .44 with]. II; n .- ...o-i. 5.... id: .-i .....(ii11i-l1tdu ....r‘dxnwi. ...1.

 

ohm 2 _- and. l mil-.2802
who ooo- owo So. who- who owo- who Who ow. T awho a o- 3802.
woo woo. So- ooo who. who ooo- wo._ mm ._ ww.._- fho oo._- 2802

 

who who ovo oho- who- we. so. o- _ho Eo- 2802
So who. who moo ooo owo who- wwo ,. So m:- 2802.
EHMWIJmho coo- oho- who- h o- mo .wnldljwfiooé.

. whoa-lo: w:- P2802

 

A: .o Sto- 5. 0. mm. o 2 .o- omd Swo- #50
525$ .32 Hoop mug-9 moo-mo, 530$ Manx m2 :55. 9.8. m> wag—mm. mm; m>
-m>.mmo_.m>-.5>>om -m>¢ou -39 -532.“ -m>¢8 -m>¢2.m>.mao_-m>¢o._ .5on -9..on -0336

 

 

62:27. 2o occasions»: 35>“: Ea Booze Son E 21 :2: 8:35 8 J :2: Home 2 one mwoc
owcmgo Rama 28 m 55 22: wEBofi 23:52 .88 9:0 zoo-5828 Bob 882% 22% .8 0:8 awn: $8502 98 flotooozv

cognate: 29$ 8% =< .8382— 52.39.22 38:5 .3 8.3.5.. 85» :03. .3 «ES.— .Smomoaxo «an; Kan 035.

135

In order to see if any of the AtCsl genes showed similar regulated expression
profile with genes involved in XyG biosynthesis, AtFutl, AtGalT and AtXTl, under
various internal or environmental conditions, the microarray data from the SMD and
NASCArray were examined for the expression patterns of AtCesA, AtCsl, AtFutl ,
AtGalT and AtXTl genes. There are 688 microarray hybridizations available at the time
(January 2002) from the Stanford Microarray Database. After discarding hybridizations
that correspond to technical experiments, analysis of other organisms, or did not meet
various quality control criteria (Gutierrez, 2003), data from 338 microarray
hybridizations were selected. Similarly data of 25 microarray hybridizations from the
Nottingham Arabidopsis Stock Center’s microarray database were retrieved and
incorporated with the ones from the SMD.

The expression of detectable AtCesA, AtCsl and AtFutl, AtGalT, and AtXTl
genes from these microarray hybridizations were selected and analyzed with hierarchical
clustering - Gene Cluster (http://rana.lbl.gov/EisenSoftware.htm). The relationships
among genes and experiments were seen through Treeview
(http://rana.lbl.gov/EisenSofiware.htm). However, the microarray data clustering analysis
of those selected data didn’t reveal any significant expression pattern linkage either
among AtCsl genes, or between AtCsl genes and genes of known functions (data not
shown). Part of the reason was that well-designed experiments with high quality
repeatable array chips and consistent signal detections are rare; and the signals of many
AtCsl genes detected on microarray chips were not significant due to their extremely low

expression. Furthermore, none of the AtCsl genes showed similar expression profile with

136

any of the AtCesA genes throughout these microarray experiments, consistent with the
hypothesis that AtCsl proteins have different function from cellulose synthase.

However, based on some well-designed and repeatable microarray experiments,
the expression of some AtCsl genes showed more alterations than others under a series of
environmental stress conditions (data not shown). If confirmed by RNA blottings or RT-
PCR, these AtCsl genes could well serve as links to connect different environmental

signal with plant cell wall metabolism.

137

 

DISCUSSION

Accurate gene annotation is essential for cloning genes via PCR, RT-PCR
reactions, and expressing genes for the production of active proteins. The sequencing of
the complete Arabidopsis genome was accomplished in 2001 and has become a rich
source for gene expression and gene functional studies for plant research
(http://www.arabidopsis.org/). However, inaccuracies of gene annotation have hampered
many research projects. The first part of this chapter described the re-annotation of AtCsl
genes from the Arabidopsis whole genome database. Many of the corrected sequences
were confirmed by comparison with cloned AtCsl cDNA sequences or EST clone
sequences. Among the 30 AtCsl genes, 9 of them from the NCBI database still show mis-
annotations compared with the sequences described in this chapter (Table 3.1). The
sequences of three of these 9 AtCsl genes were confirmed by the RIKEN cDNA clones.
The other 6 genes showed much improved sequence alignment when compared with their
corresponding homolo gs, but their annotations have still not been confirmed by
comparison with cDNA sequences available from any database.

The sequence alignment and structural analysis of AtCsl proteins indicated that
the substrate structure and catalytic mechanism for AtCslB, D, E and G proteins might be
more closely related to AtCesA proteins than that of AtCslA and C proteins (Table 3.3).
In other words, AtCslB, D, E, G proteins are more likely to use UDP-glucose as substrate
and form B-1,4 glycosidic bonds than the AtCslA and C proteins. Among them AtCle
proteins have the highest probabilities to have these enzymatic properties as they showed

the highest protein identity with AtCesA.

138

Several investigators have noted that all AtCsl proteins contain the conserved
aspartic acid residues and the QxxRW motif that are characteristic of processive B-
glycosyltransferases (Richmond and Somerville, 2000; Saxena and Brown, 2000; Doblin
et al., 2001). All AtCsl proteins described here showed the D,D,D and QxxRW motifs
even before re-annotation. It is therefore quite possible that all AtCsl proteins are
involved in the synthesis of different [3 linked polysaccharide backbones such as the
backbones of callose, XyG, GAX, mixed-linkage glucan, glucomannan, galactomannan
or galactoglucomannan. Doblin et al (2001) proposed that Cle proteins are involved in
cellulose biosynthesis based on gene expression studies in pollen. However, activity
assays and product analysis to support this conclusion are still lacking.

The second part of this chapter explored the relative expression levels and tissue
expression preferences of AtCsl genes revealed by digital northern approaches examining
EST clone abundance, the MPSS signature abundance, and microarray experiment
analysis. These efforts did not reveal any expression profile similarity between any AtCsl
genes and genes with known cell wall-related functions such as AtCesA, AtFutl , AtGalT
and AtXTl. Therefore, no immediate candidate can be identified as encoding the XyG
glucan synthase. We also could not make any new functional suggestions for AtCsl genes
based on the analysis shown here. However, because of the low expression of AtFutl ,
AtGalT, AtXTl, and most of the AtCsl genes, it is possible that the expression of some of
them may often not be detected. At the same time, because of the existence of family
members sharing high sequence identities, the expression of AtCsl genes may be
misrepresented by their close homologs. Techniques like real-time RT-PCR and RNA

blotting may reveal more meaningful expression patterns for AtCle genes.

139

A guar (C yamopsis tetragonoloba) homolog of AtCslA9 gene was recently found
to encode a B-mannan synthase (Kanwarpal S. Dhugga, personal communications). Guar
seeds contain 37% galactomannan as cell wall storage polysaccharide
(http://www.polvgal.clmaolL/ENGLISH/ROHSTO.HTM). A cDNA clone was isolated
from a guar seed EST database and showed sequence similarity with cellulose synthase-
like genes from other organisms (Kanwarpal S. Dhugga, personal communications). This
gene, when expressed in a soybean somatic embryo system, showed high level of activity
for B-mannan synthase (Kanwarpal S. Dhugga, personal communications). This study is
the first to provide experimental evidence for the speculation that Csl genes encode
glycan synthases or glycosyltransferases involved in cell wall polysaccharide
biosynthesis. While this provides a function for one gene in guar, it leaves unresolved the
functions of the Arabidopsis genes. It will be interesting to determine whether the

Arabidopsis homolog of the guar gene also encodes a mannan synthase.

140

REFERENCE

Brenner, S., Johnson, M., Bridgham, J., Golda, G., Lloyd, D. H., Johnson, D., Luo,
S., McCurdy, S., Foy, M., Ewan, M., Roth, R., George, D., Eletr, S., Albrecht, G.,
Vermaas, E., Williams, S. R., Moon, K., Burcham, T., Pallas, M., DuBridge, R. B.,
Kirchner, J., Fearon, K., Mao, J.-i. and Corcoran, K. (2000). "Gene expression

analysis by massively parallel signature sequencing (MPSS) on microbead arrays." Nat.
Biotechnol. 18: 630-634.

Brenner, S., Williams, S. R., Vermaas, E. H., Storck, T., Moon, K., McCollum, C.,
Mao, J.-i., Luo, S., Kirchner, J., Eletr, S., DuBridge, R. B., Burcham, T. and
Albrecht, G. (2000). "In vitro cloning of complex mixtures of DNA on microbeads:
Physical separation of differentially expressed cDNAs." Proc. Natl Acad. Sci. USA 97:
1665-1670.

Camirand, A., Brummell, D. and Maclachlan, G. (1987). "Fucosylation of xyloglucan
localization of the transferase in dictyosomes of pea stem cells." Plant Physio]. 84 (3):
753-756.

Camirand, A. and Maclachlan, G. (1986). "Biosynthesis of the fucose-containing
xyloglucan nonasaccharide by pea Pisum sativum cultivar Alaska microsomal
membranes." Plant Physiol. 82 (2): 379-383.

Cutler, S. and Somerville, C. R. (1997). "Cellulose synthesis: Cloning in silico." Curr.
Biol. 7: RIOS-R111.

Delmer, D. P. and Stone, B. A. (1988). "Biosynthesis of plant cell walls." The
Biochemistry of Plants. J. Priess, New York, Academic Press Inc. 14: 373-421.

Doblin, M. S., Melis, L. D., Newbigin, E., Bacic, A. and Read, S. M. (2001). "Pollen
tubes of Nicotiana alata express two genes from different beta-glucan synthase families."
Plant Physiol. 125: 2040-2052.

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-fucosyltransferase that catalyzes the last step of cell wall
xyloglucan biosynthesis in pea." J. Biol. Chem. 275 (20): 15082-15089.

Faik, A., Price, N. J., Raikhel, N. V. and Keegstra, K. (2002). "An Arabidopsis gene
encoding an alpha-xylosyltransferase involved in xyloglucan biosynthesis." Proc. Natl
Acad. Sci. USA 99 (l 1): 7797-7802.

Favery, B., Ryan, E., Foreman, J., Linstead, P., Boudonck, K., Steer, M., Shaw, P.
and Dolan, L. (2001). "KOJAK encodes a cellulose synthase-like protein required for
root hair cell morphogenesis in Arabidopsis." Gene. Dev. 15: 79-89.

141

Femandes, J., Brendel, V., Gai, X., Lal, S., Chandler, V. L., Elumalai, R. P.,
Galbraith, D. W., Pierson, E. A. and Walbot, V. (2002). "Comparison of RNA
expression profiles based on maize expressed sequence tag frequency analysis and micro-
array hybridization." Plant Physio] 128: 896-910.

Fry, S. C. (1989). "The structure and functions of xyloglucan." J. Exp. Bot. 40 (210): 1-
11.

Gollub, J., Ball, C. A., Binkley, G., Demeter, J., Finkelstein, D. B., Hebert, J. M.,
Hernandez-Boussard, T., Jin, H., Kaloper, M., Matese, J. C., Schroeder, M., Brown,
P. 0., Botstein, D. and Sherlock, G. (2003). "The Stanford Microarray Database: data
access and quality assessment tools." Nucleic Acids Res. 31 (1): 94-96.

Gutierrez, R. A. (2003). Inherent and regulated mRNA stability in A. thaliana. Ph.D.
Thesis. Department of Biochemistry and Molecular Biology, Michigan State University.

Hayashi, T. and Maclachlan, G. (1984). "Pea Pisum sativum cultivar Alaska xyloglucan
and cellulose 1. Macromolecular organization." Plant Physiol. 75 (3): 596-604.

Hayashi, T. and Matsuda, K. (1981). "Biosynthesis of xyloglucan in suspension
cultured soybean Glycine max cells. Synthesis of xyloglucan from UDP-glucose and
UDP-xylose in the cell-free system." Plant Cell Physiol. 22 (3): 517-524.

Kato, Y. and Matsuda, K. (1980). "Structure of oligosaccharides obtained by controlled

degradation of mung bean xyloglucan with acid and Aspergillus oryzae enzyme
preparation." Agric. Biol. Chem. 44(8): 1751-1758.

Lockhart, D. J. and Winzeler, E. A. (2000). "Genomics, gene expression and DNA
arrays." Nature 405: 827-836.

Madson, M., Dunand, C., Li, X., Verma, R., Vanzin, G. F., Caplan, J., Shoue, D. A.,
Carpita, N. C. and Reiter, W.-D. (2003). "The MUR3 gene of Arabidopsis encodes a
xyloglucan galactosyltransferase that is evolutionarily related to animal exostosins."
Plant Cell 15: 1662-1670.

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.

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 (5422): 1976-1979.

Richmond, T. A. (2000). "Higher plant cellulose synthases." Genome Biology 1 (4):
3001.l-.5.

142

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

Richmond, T. A. and Somerville, C. R. (2001). "Integrative approaches to determining
Csl function." Plant Mol. Biol. 47 : 131-143.

Saxena, I. M. and Brown, R. M. J. (2000). "Cellulose synthases and related enzymes."
Curr. Opin. Plant Biol. 3: 523-531.

White, J. A., Todd, J ., Newman, T., Focks, N., Girke, T., Martinez de Ilarduya, 0.,
Jaworski, J. G., Ohlrogge, J. B. and Benning, C. (2000). "A new set of Arabidopsis
expressed sequence tags from developing seeds. The metabolic pathway from
carbohydrates to seed oil." Plant Physiol 124: 1582-1594.

Williamson, R. E., Burn, J. E. and Hocart, C. H. (2002). "Towards mechanism of
cellulose synthesis." Trends Plant Sci. 7 (10): 461-467.

Zhu, Y., Nam, J., Humara, J. M., Mysore, K. S., Lee, L. Y., Cao, H., Valentine, L.,
Li, J., Kaiser, A. D., Kopecky, A. L., Hwang, H. H., Bhattacharjee, S., Rao, P. K.,
Tzfira, T., Rajagopal, J ., Yi, H., Veena, Yadav, B. S., Crane, Y. M., Lin, K., Larcher,
Y., Gelvin, M. J., Knue, M., Ramos, C., Zhao, X., Davis, S. J., Kim, S. 1., Ranjith-
Kumar, C. T., Choi, Y. J., Hallan, V. K., Chattopadhyay, S., Sui, X., Ziemienowicz,
A., Matthysse, A. G., Citovsky, V., Hohn, B. and Gelvin, S. B. (2003). "Identification
of Arabidopsis rat mutants." Plant Physio]. 132 (2): 494-505.

143

CHAPTER 4

Localization Studies of AtCleZ Proteins

144

ABSTRACT

Among the subfamilies of the AtCesA superfamily of genes, the AtCle
subfamily has the highest level of sequence identity to the cellulose synthase subfamily.
AtCsID proteins not only share significant sequence identity with AtCesA proteins but
also contain the B processive glycosyltransferase motifs conserved for AtCesA proteins.
However, the fimction of AtCle proteins is still not known. The two major hypotheses
are that they are involved in cellulose biosynthesis or the formation of matrix
polysaccharides that are similar in structure to cellulose, such as xyloglucan. In some
cases, unique tissue—specific expression of AtCle genes were observed, with high level
of expression in pollen tubes or roots. These observations led to the hypothesis that
AtCle proteins were isoforms of AtCesA proteins expressed in specific tissues. Among
the cell wall polysaccharides, only cellulose and callose are synthesized at the plasma
membrane while all other polysaccharides are synthesized in the Golgi complex. To test
this hypothesis and, thereby, shed some light on the functions of AtCle proteins,
antibodies against AtCle2 were generated and purified. Utilizing techniques of
endomembrane purification such as two-phase partitioning or sucrose density gradient
centrifugation, combined with localization of antibodies against marker enzymes for
different organelles, AtCle2 was localized in the Golgi complex. This observation made
it unlikely that AtCle proteins are involved in cellulose synthesis and more likely that

they are responsible for the synthesis of a matrix polysaccharide backbone.

145

INTRODUCTION

Plant cell walls consist of cellulose microfibrils embedded in a matrix of
polysaccharides. While cellulose microfibrils contain only B-(l, 4)-glucan chains, the cell
wall matrix is a heterogeneous complex including neutral polysaccharides such as
xyloglucan, mannan, galactomannan, and xylan, and acidic pectic polysaccharides such
as PGA, RG I and RG 11 (Staehelin and Moore, 1995). Another polysaccharide of plant
cell walls is callose, consisting of B-(1,3)-glucan chains.

A definitive feature that distinguishes enzymes involved in the biosynthesis of
cellulose and callose from those involved in the biosynthesis of other matrix
polysaccharides is their localization within the cell. Cellulose and callose are synthesized
at the plasma membrane and then incorporated into the wall while other matrix
polysaccharides are synthesized in Golgi complexes and then transported to the cell
surface for their incorporation into the cell wall (Delmer and Stone, 1988).

Cellulose synthase was first visualized as an organized macromolecular complex
(the terminal complex) using freeze—fracture techniques to examine the synthesis of
cellulose microfibrils in the plasma membrane of a green alga Oocystis apiculata (Brown
and Montezinos, 1976). The cellulose synthase terminal complex structure has been
observed at the plasma membrane using the same technique in many other organisms,
such as bacteria (Zaar, 1979), mosses and ferns (Rudolph et al., 1989), various algae
(Hotchkiss, 1989; Mizuta and Brown, 1992), and higher plants (Mueller et al., 1976;
Mueller and Brown, 1980). One of the latest localization studies of the cellulose synthase
rosette terminal complexes on the plasma membrane was done in the vascular plant

V igna angularis (Kimura et al., 1999). Authors of this study utilized an antibody against

146

an E. coli-expressed conserved region of the catalytic subunit CelA of cotton cellulose
synthase, combined with the freeze—fracture technique, to localize the terminal
complexes on the P-fracture face, i.e., the cytoplasmic side of the plasma membrane
(Kimura et al., 1999).

Callose is not a universal component of the cell wall matrix. It is often only seen
at limited locations such as cell plates of dividing cells, plasmodesmata, and sieve plates
(Northcote et al., 1989). It is also rapidly synthesized during wounding, pathogen attack,
and mechanical stress (Kauss, 1987; Bolwell, 1993; Benhamou, 1995). Callose synthase
activity has long been used as a classical marker for the plant plasma membrane (Larsson
et al., 1987; Widell and Larsson, 1990). This activity has been localized in plasma
membranes purified by linear sucrose density gradients from Nicotiana alata (Turner et
al., 1998) and by two-phase partitioning from Brassica oleracea (Fredrikson et al., 1991).
Using specific antibodies against (l,3)-B—glucan, the product itself, callose synthase
activity was also localized on cell walls of the yeasts Saccharomyces rouxii and
Schizosaccharomyces pombe, of the fungus Conidiobolus obscurus (Horisberger and
Rouvet-Vauthey, 1985; Horisberger et al., 1985; Latge et al., 1986), the roots of bean
(Phaseolus vulgaris) (Northcote et al., 1989), and Nicotiana alata pollen tubes (Meikle et
aL,199l)

It has been estimated that the biosynthesis of cell wall matrix polysaccharides
requires the activity of more than a hundred glycan synthases and glycosyltransferases
(Perrin et al., 2001). However, only a few genes encoding them have been identified and
characterized (Edwards et al., 1999; Perrin et al., 1999; Faik et al., 2000; Faik et al., 2002;

Madson et al., 2003). There are still no immunochemical studies utilizing antibodies

147

against specific enzymes to localize them to Golgi membranes. These glycan synthases
and glycosyltransferases have traditionally been localized to the Golgi complex by
localizing epitopes carried on their specific polysaccharides products. Polyclonal
antibodies against PGA/RG-I and XyG were generated and used to localize the complex
pectic polysaccharides PGA/RG-l to cis- and medial cistemae of the Golgi complexes,
and XyG to the trans cistemae and the transgolgi network (TGN). Well-studied examples
include the localization of these matrix polysaccharides in clover root tip cortical
parenchyma cells (Moore et al., 1991; Lynch and Staehelin, 1992) and suspension-
cultured sycamore maple (Acer pseudoplatanus L.) cells (Zhang and Staehelin, 1992).
Monoclonal antibody J IMS specifically recognizes unesterified PGA, while J 1M7
recognizes methyl-esterified PGA, and CCRC-Ml recognizes the terminal fucose residue
of XyG. These monoclonal antibodies were utilized to localize the corresponding wall
matrix polysaccharides to both the Golgi complex and secretory vesicles in winter vetch
(Vicia villosa Roth) (Sherrier and Vandenbosch, 1994) and sycamore maple cells (Zhang
and Staehelin, 1992). Recently another type of hemicellulose, mannan, was also localized
to the Golgi-rich vesicles from Arabidopsis callus using polyclonal antibodies against
mannans (Handford et al., 2003).

On the other hand, the activities of many glycan synthases and
glycosyltransferases have been detected in membrane vesicles derived from the Golgi
complexes. The presence of these activities in Golgi-rich fractions was used as evidence
to support a Golgi localization of these enzymes. Examples include XyG

fucosyltransferase, XyG galactosyltransferase, galactan galactosyltransferase from pea

148

(Brummell et al., 1990; Baydoun et al., 2001), and the arabinosyltransferase involved in
arabinoxylan synthesis from wheat (Porchia et al., 2002).

We are interested in identifying functions of AtCsl proteins as some of them
might be XyG glucan synthases (see chapter 3). The localization of AtCsl proteins
becomes critical for determining if they are glycosyltransferases involved in the matrix
polysaccharide biosynthesis or just cellulose synthases. The AtCle family has the
highest sequence similarity to AtCesA proteins within the AtCesA superfamily, showing
sequence identities of around 40~45°/o (see chapter 3). They are also very similar in size
and have a similar number and organization of predicted transmembrane domains (see
chapter 3). Moreover, AtCle and AtCesA proteins share structural features that have
been proposed to be indicative of processive glycosyltransferases (Saxena et al., 1995),
such as the D,D,D,QxxRW motif proposed to define the nucleotide-sugar binding domain
and the catalytic site (Richmond and Somerville, 2000). They both belong to family 2 of
inverting nucleotide-diphospho-sugar glycosyltransferases that synthesize repeating B-
glycosyl unit structures (Campbell et al., 1997; Richmond and Somerville, 2001).

Doblin et a1. (2001) provided evidence that led them to postulate that Cle genes
encode a specialized cellulose synthase activity. RT-PCR studies using primers designed
from conserved regions of AtCesA superfamily from Nicotiana alata pollen tubes
amplified NaClel as a major product and NaCesAl as a minor product (Doblin et al.,
2001). Expression studies revealed that NaClel was expressed abundantly in male
gametophyte tissues after pollen mitosis, and NaCesAl was expressed at a low level in a
range of floral and vegetative tissues, but not in the male gametophyte (pollen or anthers)

(Doblin et al., 2001). Given the fact that cellulose is a major component of pollen tube

149

cell walls, the authors hypothesized that the gene(s) encoding cellulose synthase in pollen
should have a relatively high expression. Since NaClel and NaCesAl are the only
members of the AtCesA superfamily found in pollen tubes, the authors proposed that
Cle proteins were actually isoforms of CesA proteins and that NaClel was responsible
for the deposition of cellulose in Nicotiana alata pollen tubes (Doblin et al., 2001). If this
prediction is correct, Cle proteins should be localized at the plasma membrane.

However in another study, AtCle3 was localized in the endoplasmic reticulum.
In this case, a AtCle3-GFP fusion protein under the constitutive 358 promoter was
transiently expressed in Nicotiana benthamiana leaf cells (Favery et al., 2001). GFP
fluorescence was only observed in the endoplasmic reticulum present in the thin layer of
cytoplasm between the plasma membrane and vacuole (Favery et al., 2001). Therefore
localization studies of AtCle proteins will be an important part of the attempt to clarify
the contradictory conclusions currently present in the literature.

Whether the AtCle proteins are localized at the Golgi complexes is critical for
understanding their functions. A convincing localization of AtCle proteins at the plasma
membrane would clearly support the hypothesis that they are cellulose synthases. On the
other hand, AtCle proteins are more likely to be enzymes involved in matrix
polysaccharides biosynthesis if they are located at the Golgi complex. Therefore,
antibodies against AtCle proteins were generated and used to detect those proteins in
Arabidopsis intracellular membrane organelles.

Sedimentation by centrifugation through sucrose density gradients has frequently
been used to separate intracellular membrane organelles including the plasma membrane,

ER, Golgi, mitochondria, chloroplast etc. An organelle purification protocol has been

150

developed in maize (Buckeridge et al., 1999) and pea (Teruko Konishi, personal
communication) to isolate relatively pure Golgi vesicles using sucrose floatation gradient
centrifugation. Two-phase partitioning is another technique used for plasma membrane
purification from other endomembrane systems in many organisms (Albertsson, 1977;
Larsson et al., 1987; Buckeridge et al., 1999; Lu and Hrabak, 2002).

This chapter describes the generation and characterization of a polyclonal
antibody against AtCle2 and the utilization of this antibody to localize AtCleZ

proteins in fractions enriched in membrane vesicles derived from Golgi complexes.

151

MATERIALS AND METHODS

Plant materials

Arabidopsis plants were grown under non-sterile hydroponic conditions (Gibeaut
et al., 1997) in environmentally controlled grth chamber on a 18-h day period and
day/night temperature of 22°C/20°C. Arabidopsis (ecotype Columbia) embryonic
suspension cell line T87-C33 (Axelos et al., 1992) were maintained in the dark at 25°C in
suspension culture medium (20 g L“1 sucrose, 4.3 g L71 Murashige-Skoog salts
[GIBCO-BRL], 0.5 g I.“l MES, 0.1 g L" myo-inositol, 1 mg L‘1 thiamine-HCI, 1 uM

2,4-D, and 1 mM KZHPO4, pH 5.7).

Preparation of total protein fraction and microsomal membranes from Arabidopsis
plants

Different tissues from 5-week-old hydroponically grown Arabidopsis plants were
cut, weighed and frozen in liquid nitrogen. Tissues were then ground in liquid nitrogen
and the resulting powder was resuspended in extraction buffer (150 mM Tris~HCl at
pH6.8, 3% SDS, 7.5% 2-mercaptoethanol) at 2 ml per gram fresh weight. The extracts
were boiled in a water bath for 15 minutes and centrifuged for 15 minutes at 10,000xg.
The supematants were saved as total protein fractions.

For microsomal membrane preparation, every 20 g tissues from 5-week-old
Arabidopsis plants were cut into small pieces with a Comfort Food Chopper (Zyliss,
Switzerland) and ground in 40 ml extraction buffer (SOmM Hepes-KOH, pH6.5, IOmM
potassium acetate, 2.5mM EDTA, 0.4M sucrose, 1 mM DTT, 0.2mM PMSF, 1%

Protease Inhibitors [Sigma]) with mortar and pestle on ice. After passage through a

152

double layer of miracloth, extracts were centrifuged at 10,000x g for 10 minutes at 4°C.
The pellet was called the P10 fraction. The supernatant S10 fraction was subjected to
further centrifugation at 100,000x g for 1 hour at 4°C. The supernatant from this step was

saved as the S100 fraction, and the pellet was resuspended in extraction buffer and

labeled as the P100 fraction.

Preparation of SI 0 protein fraction and microsomal membranes from Arabidopsis cell
suspension cultures

Arabidopsis cells cultured for 4 to 5 days after inoculation were collected on
miracloth using vacuum filtration. Usually 20 g of Cells were suspended at a
concentration of l g per 3 ml of extraction buffer (SOmM Hepes-KOH, pH6.5, 2mM
EDTA, 15% sucrose, 1 mM DTT, 0.2mM PMSF, 1% protease inhibitors [Sigma]). Cells
were disrupted in a blender with short bursts of 10 sec for a total of 1 min on ice. Cell
lysate was filtered through double layer of miracloth and subjected to centrifugation at
10,000x g for 10 min at 4°C. The supernatant was saved as the 810 protein fraction, and

the pellet was discarded. The supernatant solution was centrifuged at 100,000x g for 1

hour at 4 °C to yield the S100 fraction and a membrane pellet that was resuspended in

extraction buffer at 1/10 of the starting volume and labeled as the P100 fraction.

Solubilization of microsomal membrane proteins
A P100 microsomal fraction was isolated from either 5-week-old hydroponically
grown Arabidopsis above-ground tissues or from 4 toS-day-old Arabidopsis suspension

cultured cells. Equal amounts of P100 microsomal protein were resuspended in extraction

153

buffer (SOmM HepesKOH, pH6.5, 2mM EDTA, 15% sucrose, 1 mM DTT, 0.2mM
PMSF, 1% protease inhibitors [Sigma]) and incubated with equal volume of the same
buffer containing 4M Urea, 2M NaCl, 0.2M N32CO3, 0.4% Triton X-100, 0.4% SDS or
extraction buffer alone for 30 minutes on ice. After treatment all solutions were
centrifuged at 100,000x g for 1 hour at 4°C. Supematants were saved and the pellets were
resuspended in the same volume of extraction buffer. Equal volumes of each fraction

were used for SDS-PAGE and immunoblotting.

Sucrose density gradient fractionation of microsomal membranes

The 810 protein fraction isolated from Arabidopsis cultured cells was subjected to
two sequential sucrose gradient fractionation steps. The first sucrose gradient was made
in a SW28 centrifuge tube with 10 ml of extraction buffer containing 50% sucrose
overlaid with 10 ml of extraction buffer containing 15% sucrose. S10 protein fi'action was
laid on top of the gradient and subjected to centrifugation of 100,000xg for 1 hour at 4°C.
The interface (about 2 ml) between the two sucrose solutions was recovered and diluted
with the same buffer without sucrose until the sucrose concentration reached between
25% and 30%. This diluted interface was subjected to a second centrifugation step. The
second step used a linear gradient from 30% sucrose to 50% sucrose in the same
extraction buffer in a SW40 centrifuge tube. About 2.5m] of the diluted interface from the
first step was laid on top of the linear gradient and centrifuged at 100,000xg for 3 hours
at 4°C. Fractions of about 800 it] each were collected with an Auto Dense-F low
fractionator (Labconco). Sucrose concentrations was determined with a refractometer

(American Optical Corporation, Buffalo, NY).

154

Two phase partitioning

The two phase system developed by Larsson et a1 (1987) was used for the
purification of plasma membranes. The crude microsomal membranes (P100 pellet) from
suspension-cultured Arabidopsis cells were resuspended in buffer A (0.33 M sucrose, 3
mM KCl, 5 mM potassium phosphate at pH7.8, Protease Inhibitors [12100, Sigma]) in
1/10 of the original volume. For every 3 g of this microsome solution 9 g of buffer B
(buffer A plus 8.4% Dextran T500 and 8.4% PEG 3350) was added to make the final
concentration of both Dextran and PEG to 6.3%. After inverting the tubes 30 times for
mixing, the samples were centrifuged at 1,500x g for 5 minutes at 4°C in a swinging
bucket rotor. The upper phase was collected and called the U1 fraction and the lower
phase, the L1 fraction. Buffer C (buffer A plus 6.3% Dextran T500 and 6.3% PEG 3350)
was made and mixed and allowed to sit at 4°C for overnight to separate into two phases.
The U1 fraction was further extracted twice with an equal volume of the lower phase
from buffer C to produce phases U2, L2’, U3 and L3’ sequentially. The L1 fraction was
also further extracted twice with an equal volume of the upper phase from buffer C, to
produce the U2’, L2, U3’ and L3 phases. Equal volumes of each phase or equal amounts
of protein fi'om each phase were used for SDS-PAGE and enzyme activity assays. Protein

concentration was determined with Micro BCA kit (Pierce, Rockford, IL).

Preparation of Golgi rich microsomal membrane fractions

The centrifugation procedure developed by Buckeridge et a1 (1999) for maize was

adopted for use with Arabidopsis. The S10 protein fraction was extracted from 4 to 5

155

days old cell suspension culture the same way as used for other applications except in a
different buffer (84% sucrose [w/v], 20 mM Hepes-KOH, pH 7.0, 20 mM KCl, 5 mM
EDTA, 5 mM EGTA, 10 mM DTT, Protease Inhibitors [12100, Sigma]). The sucrose
concentration in the S10 fraction was checked with a refractometer (American Optical
Corporation, Buffalo, NY ) and adjusted to around 40% sucrose (37~45%) using the
same buffer containing no sucrose or 84% sucrose depending on the need. Sucrose
gradient buffer contained 20 mM Hepes-KCI, pH7.0 and protease inhibitors (1 : 100,
Sigma). For making the sucrose density gradient, 10 ml of 810 total protein with sucrose
concentration adjusted was laid on top of 5 ml of gradient buffer containing 50% (w/v)
sucrose in a SW28 centrifuge tube. On top of the sample, 8 ml of gradient buffer with
34% (w/v) sucrose, then 8 ml of 25% (w/v) sucrose in gradient buffer, and then 7 m1 of
18% (w/v) sucrose in gradient buffer were laid on sequentially. About 2 ml of gradient
buffer containing 9.5% (w/v) sucrose was used to make up the volume until the edge of
the tube. The gradient was centrifuged at 100,000x g for 1.5 to 3 hours at 4°C. After
centrifugation, the interface between layers with different sucrose concentrations were
collected and designated as fraction 1, 2, 3, 4, and 5 from the top to the bottom of the
gradient. According to the original protocol, developed and characterized in maize
(Buckeridge et al., 1999), those five fractions are enriched with tonoplast, ER, Golgi,
mitochondria, and plasma membrane, specifically. Here only fractions 3, 4 and 5 were
further characterized and used for the AtCle2 localization studies. The pellet fraction
was resolubilized in the same buffer without sucrose. Protein concentrations were

determined using the Micro BCA kit from PIERCE (Rockford, IL).

156

Expression of polypeptide P2 and P3 in E. coli cells

The DNA sequence encoding the first 272 amino acids of AtCleZ, fragment P2,
was amplified from Arabidopsis genomic DNA with primers kee9l9 (5’-
CCTACATATGGCATCTAATAAG-3’) and kee921 (5’-
TCATGAATTCTTGGGCCTCCAT-3’). The DNA sequence encoding the first 188
amino acids of AtCle3, fragment P3, was amplified from Arabidopsis genomic DNA
with primers kee920 (5’-TTAGCATATGGCGTCTAATAAT-‘3) and kee922 (5’-
CTGCTTGTTGTTCTCGAGCAAA-3’). PCR products were inserted into expression
plasmid pET28a via NdeI and EcoRI sites for DNA fragment encoding P2, and NdeI and
XhoI sites for DNA fragment encoding P3. The identity constructs encoding peptide
fragment P2 and P3 were confirmed by restriction digestion with NotI and sequencing.
Peptide fragment P2 and P3 were expressed in BL21 (DE3) cells according to
manufacturer’s guide (pET system manual, Novagen, Madison, WI). Total soluble

proteins were extracted and used for subsequent experiments.

Generation and purification of polyclonal antibodies

AtCle2-specific peptide IQEPGRPPAGHSVKFAQ(C), corresponding to a
region near the N-terminus (aa20~aa36), and AtCle3-specific peptide
(C)DAAEAERHQQPVSNSVTFA (aa19~aa3 7) were synthesized and conjugated with
keyhole limpet hemacyanin (KLH) for polyclonal antibody production in rabbits by
Alpha Diagnostic International, Inc. (San Antonio, TX). Both polyclonal antibodies were
affinity-purified using the SulfoLink Kit (PIERCE, Rockford, IL) according to the

manufacturer’s instruction. To prepare the affinity resin, one mg of either antigen peptide

157

with a Cys attached (lyophilized) was dissolved in 50 ul of DMSO plus 1 ml of coupling
buffer supplied with the kit and attached onto SulfoLink coupling gel. One ml of the
second bleed serum each from rabbit No. 6167 (against AtCle2) and No. 6169 (against
AtCle3) was diluted 5 times before application to the SulfoLink affinity resin
containing the specific peptides. The affinity-bound antibodies were eluted with 100 mM
glycine buffer at pH2.5. Eluted antibodies were concentrated and the buffer was changed
to PBS using the centrifugal filter device-Ultrafree (Millipore, Bedford, MA) with a
cutoff size of 50 kDa. Sodium azide was added to the purified antibodies to a final
concentration of 0.01%. Antibodies were placed at -20°C for long term storage and at 4°C

for short term use.

Antigen competition assay

Equal volumes of antibodies to be analyzed were incubated at room temperature
for 2 hours with variable quantities of E. coli total soluble protein extracts containing P2
or P3. After a centrifugation at 10,000x g for 10 minutes at 4°C, the supernatant was

saved and used for immunoblotting.

SDS-PA GE and immunoblotting

All SDS-PAGE and immunoblotting procedures followed standard protocols
(Harlow and Lane, 1988). Antibody against PMA2 (a Nicotiana plumbaginifolia plasma
membrane H+-ATPase) was a gift of Dr. Marc Boutry. Antibodies against AtSYP6l (a

Golgi marker), AtSYP21 (a PVC marker), AtELP (a TGN-PVC marker) and AtSEC12

158

(an ER marker) were gifts from Dr. Natasha Raikhel. Antibody against AtXTl was

generated by Dr. Ahmed Faik of our group.

Marker enzyme assays

The assay for B—glucan synthase II (callose synthase) activity was modified from
existing protocols (White et al., 1993; Konishi et al., 2001). The reaction mixture
contained 50 mM HepesKOH, pH7.0, 1 mM UDP-Glc, 5 mM CaClz, 4.36 uM UDP-
[14C]Glc (305.9 mCi/mmol). Reactions were carried out at room temperature for 45 min.
Xyloglucan carrier (5 ul of a 1% solution) was added to each reaction before adding 750
pl of 70% ethanol to give a final concentration of 67% ethanol to stop the reactions.
Reaction products were precipitated at room temperature for 30 minutes and washed
three times with 70% ethanol by vortexing and precipitation at 12,000 g for 5 min. XyG

fucosyltransferase activity assay was performed the same way as in chapter 2.

159

RESULTS
Generation, purification and characterization of antibodies against AtCle2

Because AtCle2 and AtCle3 are the two most highly expressed members
within the AtCle gene family (see chapter 3), we sought to prepare antibodies against
each of these two proteins. Short peptides of 18 aa and 20 aa, corresponding to a region
near the N-terminus of AtCle2 and AtCle3, respectively, were chemically synthesized,
coupled to a carrier protein and used as antigen to generate polyclonal antibodies. Both
antibodies against AtCle2 and D3 were affinity purified using an immobilized form of
their corresponding peptides that had been used as antigen.

To characterize the antibodies, a polypeptide fragment P2 containing the N-
terminal 272 amino acid from AtCle2 and a polypeptide fragment P3 containing the N-
terrninal 188 amino acids from AtCle3 were expressed in E. coli. Neither of the
fragments contained any of the predicted transmembrane domains contained within the
full-length protein. The expression of P2 and P3 in E. coli were confirmed by SDS-PAGE
followed by Commassie blue staining (Figure 4. 1 A) and immunoblotting using
antibodies against the His tag attached to the amino terminus of each expressed peptide
(Figure 4. 1 B). The peptide antibodies against AtCle2 and D3 were tested on the E. coli
extracts containing fragment P2 or P3. Both antibodies were confirmed to be protein
specific in that the antibodies against the AtCle2 peptide recognized expressed AtCleZ
fragment P2 but not P3 that was derived from AtCle3. Similarly antibodies against the
AtCle3 peptide recognized fragment P3 but not P2 (Figure 4.1C, D). Other proteins of
smaller molecular weights were also detected on the immunoblots (Figure 4.1C, D).

Those proteins might be just degradation products of fragments P2 or P3.

160

Figure 4.1: Expression and detection of peptides P2 and P3 in BL21 (DE3) cells. A:
Total E. coli protein extracts were resolved on 12.5% SDS-PAGE and stained with
Coomassie brilliant blue. Proteins were redissolved in SDS-PAGE loading buffer
according to their cell cultures’ OD600 values to give approximately equal protein
concentrations. Equal volumes of extracts (20 pl) from cells expressing P2 (1, 2) and P3
(3, 4) were loaded. Lanes 1,3: total protein extracts before induction; lanes 2, 4: total
protein extracts after 3 hours induction. The induced peptide P2 and P3 are indicated by
arrows. B, C, D: Immunoblots on E. coli extracts containing fragments P2 and P3 that
were resolved on a 12.5% SDS-PAGE, using antibodies against 6xHis tag (B); antibodies
against AtCle2 (C) and antibodies against AtCle3 (D). Same amount of proteins were

used as indicated in panel A.

161

198 kD—>

115kD—>
93 kD—>

SOkD"

 

 

 

 

162

 

 

 

 

 

 

 

 

I

I}. __

{If-14%!

 

 

 

Purified AtCle2 antibodies recognized four proteins (designated a, b, c and d in
Figure 4.2) in Arabidopsis protein extracts with protein “a” having the proper molecular
size. To explore if these bands are specific for AtCleZ, an antigen competition assay
was developed (Figure 4.2). In this assay, the purified antibodies against AtCle2 were
incubated with extracts from E. coli expressing fragment P2 or P3 before they were used
for immunoblotting against Arabidopsis total proteins. E. coli extracts containing
fragment P2 were able to absorb reactive antibodies so that no specific signal could be
detected when Arabidopsis total proteins were probed with the absorbed serum (Figure
4.2, lane C, band a). On the other hand, absorption with E. coli extracts containing
fragment P3 did not affect the ability of antibodies against AtCle2 to recognize proteins
from Arabidopsis extracts (Figure 4.2, lane D, band a). Other signals were also detected
(Figure 4.2, bands b, c, d) and were characterized later. Similar absorption experiments
with the purified antibodies against AtCle3 did not identify a specific protein in the
Arabidopsis extracts, and therefore the following experiments only used antibodies
against AtCle2.

Three protein signals (a, b, c) disappeared after antibody absorption using E. coli
expressed P2 fragment (Figure 4.2), indicating that they all contained the same epitopes
as the peptide antigen. Protein (1, at around 65 kDa, was still detected after antibody
absorption with E. coli proteins containing fragment P2, indicating it was a non-specific
recognition by this antibody. In later experiments, this protein was not detected on
immunoblots when fresher total protein samples or microsome membrane proteins were
used and when the washing time during immunoblotting was extended. When

Arabidopsis microsome membrane proteins were pelleted by centrifugation at 100,000x g

163

Figure 4.2: Specificity of AtCleZ antibodies against Arabidopsis total protein
extracts. Equal volumes (10 pl) of antibodies against AtCle2 were incubated with
specified E. coli total protein extracts for 2 hours at room temperature and centrifuged at
10,000xg for 10 minutes before they were used for immunoblotting at a dilution of
1:1,000 against Arabidopsis total protein extracts. Antibody preparations were incubated
with 100 pl of total protein extracts from BL21 (DE3) culture expressing vector plasmid
pET28a alone (A), 10 pl (B) or 100 pl (C) of total protein extracts from BL21 (DE3)
culture expressing fragment P2 and 100 pl of total protein extracts from BL21 (DE3)
culture expressing fragment P3 (D). The E. coli protein extracts used here were the same
as used in figure 4.1. Panels A, B, and C were from same gel containing 12.5%
acrylamide whereas panel D was in a separate gel containing 10% acrylamide. The four

major bands detected were designated as band a, b, c and (1, respectively.

164

A B C
100 p1

10 pl

 

c>\

 

<— 203 kDa

4—115kDa
<— 93kDa

{—48.2 kDa

{—34.7 kDa

«28.2 kDa

{—21.1kDa

 

 

165

b}!—

C>p

 

‘— 210 kDa

‘4—125kDa

{—101kDa

{—56.2 kDa

4—358 kDa

<— 29 kDa

 

(P100 fraction), proteins a and c sedimented with microsomal membranes while protein b
stayed in the soluble fraction (Figure 4.3A). Even though protein b showed similar
behavior as protein a in the antibody absorption assay, its smaller size (~90 kDa) and
fractionation with soluble proteins made it very unlikely to be AtCle2, which has 8
predicted transmembrane domains. Furthermore, when Arabidopsis proteins recovered in
the 10,000x g supernatant (S10 fraction) were fractionated by sucrose density gradient
centrifugation, proteins a, b and c clearly behaved differently. Protein a peaked in
fractions containing sucrose of 33% to 35% while protein b remained at the top of the
gradient. Protein c peaked in fractions containing sucrose of ~21% indicating it was
present in a different membrane system than protein a (data not shown). Finally, when
Golgi rich membranes were prepared from Arabidopsis cell suspension cultures, protein a
was the only one detected with the purified antibodies against AtCle2 (Figure 4.3B).
Therefore in purified Golgi rich membrane preparations this antibody specifically

recognized a membrane protein with a molecular size of about 120 kDa.

AtCleZ is an integral membrane protein

Sequence analysis predicted that AtCle2 protein contains 8 transmembrane
domains. After sequential centrifugation at 10,000x g and 100,000x g, the majority of
AtCle2 was detected in the P100 membrane fraction (Figure 4.3A), indicating AtCle2
is associated with membranes. To investigate whether AtCle2 is an integral membrane
protein, microsomal membranes from the P100 fraction were treated with different
reagents that affect the association of peripheral membrane proteins (Figure 4.4).

AtCle2 could not be removed from microsomal membranes by 2 M Urea, 1 M NaCl,

166

 

A B

$10 $100 P100 At Golgi

 

 

210 kDa—D +214kDa

125kDa—>
_‘ '-l ‘a’ "" 4-118kDa

101 kDa» —-i 4" <— 92kD
a a

 

 

56.2 kDa ->
<- 52.2 kDa

35.8 kDa -§
1 + 35.7 kDa

29 kDa —>
4- 28.9 kDa

 

 

 

 

 

 

Figure 4.3: Immunoblotting using purified AtCle2 antibodies against Arabidopsis
crude membranes (A) and purified Golgi complexes (B). S10, total protein extracts after
10,000x g centri fugation; S100 and P100, soluble proteins and membrane pellets after

100,000x g centrifugation. Immunoblotting signals a, b and c are the same as indicated in

figure 4.2.

167

AtCle2

PMA2

AtXT l

AtSYP61

AtSEC12

Buffer 2M Urea 1M NaCl 0,1M Na2C03 0.2% Triton 0.2% SDS
S P S P S P S P S P S P

.....l 3" F! o \- a El:

 

 

   

fit
‘3 m a
w— w —_ 1

Figure 4.4: Solubilization of AtCle2 proteins from Arabidopsis microsomal

membranes with various reagents. Equal amounts of protein from the P100 fraction

were incubated with buffer alone, or buffers containing 2 M Urea, 1 M NaCl, 0.1 M

NaQCO3, 0.2% Triton X-100 or 0.2% SDS on ice for 30 min, and centrifuged again at

100,000x g for 1 hr. Pellets were resuspended in the same volume of extraction buffer as

the supernatant. Equal volumes of each fraction were analyzed by SDS-PAGE and

immunoblottings using antibodies against AtCle2, plasma membrane marker PMA2,

Golgi marker AtXTl and AtSYP6l and ER marker AtSEC12. S, supernatant fractions; P,

pellet fractions.

168

 

0.1 M Na2C03 or 0.2% Triton X-100, which can dissociate peripheral membrane proteins.
Only 0.2% SDS completely solubilized AtCle2 from the microsomal membranes,
indicating that AtCleZ is an integral membrane protein.

Several control proteins were monitored to check whether the fractionation
experiments yielded the expected results. PMA2 is a plasma membrane H+-ATPase with
multiple transmembrane domains and a topology similar to AtCle2, four
transmembrane domains near the N-terminus and six close to the C-terminus (Morsomme
et al., 1998). PMA2 was also only removed by 0.2% SDS and behaved exactly like
AtCle2 during these treatments. AtXTl is predicted to be a type II integral membrane
protein. It was only completely solubilized by 0.2% SDS, but trace amounts of AtXTl
were also solubilized by all the other reagents used, except with the buffer alone,
indicating that AtXTl is also an integral membrane protein but might not associated with
the membranes as tightly as AtCle2 or PMA2. AtSEC12 is an ER-associated integral
membrane protein that could only be removed from membranes by nonionic or ionic
detergents but not by other treatments such as urea, salt or alkaline conditions (Bar-Peled
and Raikhel, 1997). The behavior of AtSEC 12 (Figure 4.4) was the same as observed
before (Bar-Peled and Raikhel, 1997) and served as a nice control. AtSEC12 also
demonstrated a looser integration with the membranes, because, in addition to nearly
complete solubilization by 0.2% SDS, AtSEC12 were also largely solubilized by 0.2%
Triton X-100. AtSYP61 is a trans-Golgi network (TGN) protein with a single
transmembrane domain near its C-terminus. It also showed a weaker association with
membranes compared with AtCle2 or PMA2, as AtSYP61 was also removed from the

membrane by 0.2% Triton X-100 (Figure 4.4).

169

AtCleZ migrates differently from a plasma membrane marker PMA2 during linear
sucrose gradient centrifugation

Sucrose density gradient centrifugation was employed in an attempt to resolve
different membranes such as plasma membrane, ER, Golgi, and mitochondria. As a first
step of enrichment, the 810 protein fraction from Arabidopsis suspension cultured cells
was subjected to a sucrose step gradient of 15% and 50%. The microsomes from the
interface between 15% and 50% sucrose were recovered and further fractionated on a
linear sucrose density gradient from 30% to 50%. The range of 30% to 50% was picked
because it produced a gradient covering the range of densities for most endomembrane
organelles. After centrifugation the sucrose density distribution was nearly linear (Figure
4.5A). XyG fucosyltransferase activity peaked in fraction 6 and 7 that contain 32% to
34% sucrose (Figure 4.5B). Immunoblotting demonstrated that AtXTl was detected in
the same fractions (Figure 4.5C), leading to the conclusion that Golgi membranes were
located in these fractions. AtCle2 was distributed most abundantly in fractions 5 to 8
that contain 31% to 35% of sucrose (Figure 4.5C). Therefore the conclusion was made
that AtCle2 was localized in the Golgi complex, but had a broader distribution than
XyG fucosyltransferase and xylosyltransferase. On the other hand the plasma membrane
marker PMA2 was distributed toward the bottom of the gradient among fractions 11 to 14
containing 38% to 41% sucrose (Figure 4.5C). The distribution of PMA2 detected by
immunoblotting matched with the second peak of the G811 activity distribution in the

gradient. GS 11 activity peaked twice in the gradient with its first peak at fractions 9 and

170

 

Figure 4.5: Migration of AtCle2 and selected marker proteins on a linear sucrose
density gradient. SIO fraction proteins from 4 to S-day-old Arabidopsis cultured cells
were placed on top of a sucrose step gradient of 15% and 50%. After centrifugation at
100,000x g for 1 hour, the interface was collected and its sucrose content was diluted to
between 25% and 35%, then loaded on a linear sucrose gradient from 30% to 50%.
Fractions of 0.8 ml were collected after centrifugation at 100,000x g for three hours.
Aliquots of each fractions were used for enzyme activity assays, or SDS-PAGE followed
by immunoblotting with various antibodies. A: Sucrose concentration of each fraction; B:
Enzyme activity assays for XyG fucosyltransferase and GS II; C: Irnmunodetection of
AtCle2 and marker proteins PMA2 (plasma membrane), AtXT] (Golgi), AtSEC12 (ER)

and AtSYP21 (PVC).

171

I < 3
' 1
Oz:

3>

Sucrose concentration (%)

cu

I XyG FUTse activity (DPM)

 

      

 

 

 

 

 

 

44 r‘ l
42 -~ ~——— —--———-———- i 1.1855
l —l— sucrose conc.| 1 1755
40 f ‘ + Suc. density J '
38 J. ’__'_—"h'— - 1.1655
36 4 - 1.1555
34 l - 1.1455
32 1 r 1.1355
30 1 r 1.1255
28 _ 1.1155
26 l—-~—~~-—~-*- -——e ~ - - - - .- 1,1055
1 2 3 4 5 6 7 8 9101112131415
Fractions
1800 T —- 700
1
1600 +
2- 600 A
1400 r E
200 i b 500 a
1 1 b
1000 4. er 400 E
800 -‘~ .L 300 8
00 '5 =
6 l 4- 200 8
400 ‘
200 i m ‘” 10°
0»:- ++ +* - - 1 l - -+-—- . -L .0
l 2 3 4 5 6 7 8 9 10 11 12 l3 14 15
Fractions
Tot l 2 3 4 5 6 7 8 9 10 11 12 13 14 15
- . - o 1 no N PMA2

WIQM - . . ' AtCle2

.....- AtXTl

‘u-MO-o. an...» .. AtSEC12

...-....-. ....- . ...- -- ~ -..... AtSYP21

172

Sucrose density (g°cm’3)

 

10 containing 35% to 37% sucrose. This could result from the presence of multiple

enzyme activities that catalyze the polymerization of UDP-glucose.

AtCleZ is not a plasma membrane protein as revealed by two-phase partitioning

Because density gradient centrifugation was not definitively conclusive with
respect to the localization of AtCle2, an independent approach to address this problem
was sought. The aqueous two-phase partitioning method (Figure 4.6) enriches the plasma
membrane in the upper phase and most other cellular membranes in the lower phase.
Three consecutive partitioning steps were employed. The upper and lower phase U3 and
L3 from the third partitioning were analyzed for XyG fucosyltransferase and GS 11
activity to confirm the effect of partitioning (Figure 4.6 B). The lower phase L3 showed
only about one fifth the level of GSII activity compared with the upper phase U3 (Figure
4.68), indicating that highly enriched plasma membrane fraction was obtained. All
phases then were analyzed by immunoblotting to locate AtCle2 protein, the Golgi
marker protein AtXTl, TGN marker protein SYP61, TGN-PVC marker protein AtELP,
plasma membrane marker protein PMA2, and ER marker protein AtSEC12. At the end of
the third partitioning AtCle2 was found primarily in the lower phase, which also
contained markers derived from ER, Golgi and TGN, while the plasma membrane marker,
PMA2, partitioned predominantly into the upper phase (Figure 4.6). From these

observations, it was concluded that AtCle2 is not localized in the plasma membrane.

AtCleZ protein is localized at Golgi rich vesicles during floatation sucrose gradient

centrifugation

173

Figure 4.6: Two-phase partitioning of Arabidopsis microsomal membranes. The S10
protein fraction prepared from Arabidopsis suspension cultured cells was centrifuged at
100,000x g for 1 hr to yield the microsomal membranes. The microsomal membranes
were then partitioned three times in buffers contained 6.2% Dextran T500 and 6.2% PEG
3350 (Materials and Methods). A. Aliquots of equal volume from each fraction were
resolved in SDS-PAGE and transferred to PVDF membrane for immunoblotting with
various antibodies as probes. Equal amount of protein were also taken fi‘om U3 and L3
fractions for similar immunoblottings. U11 and Ln (n=1, 2 or 3) fractions are the upper and
lower phases after each partitioning. The experimental procedure was explained in the
Materials and Methods. B. The specific activity of G811 and XyG fucosyltransferase

(FUTI) in fractions U3 (white blocks) and L3 (black blocks).

174

A

Equal volume loading

Equal protein loading

 

Antihml) U1 L1

Allis?!

AtELP

AlSliC

)7

12

.~\1.\‘1l'<1l

Dd

GS II activities (pmol-g"-min“)

U2

 

 

 

GSII

175

L2’ U2’ L2 U3 L3’ U3’ L3

U3
L3

 

FUTI

 

U3 L3

“421

.4.
J
...

i.“

" 3500
3000

2500

Location

lti.\—l’\"C

LR

'l‘(1.\

l’.\l

(Vinlgl

XyG FUTI activities (pmol-g"-min")

As a third method to investigate the location of AtCle2, microsome membranes
were isolated from Arabidopsis suspension cultured cells and fractionated via flotation on
a sucrose step gradient (Figure 4.7A). The enrichment of Golgi vesicles and plasma
membrane was evaluated by XyG fucosyltransferase and GS 11 activity assays
respectively (Table 4.1). The specific activity of XyG fucosyltransferase in the Golgi-rich
fraction (fraction 3) had been enriched more than 130 times compared to the starting
membrane material (Table 4.1). Fraction 5 contained highly purified plasma membranes,
showing only 5% the level of specific activity for XyG fucosyltransferase relative to
fraction 3. At the same time the GS 11 activity had almost equal distribution between
fraction 5 (plasma membrane enriched) and fraction 4 (Table 4.1). The purified Golgi
membranes still showed some incorporation of radio labeled UDP-glucose during GS II
activity assay (Table 4.1).

The GSII activity recorded for fraction 3 may not reflect the true GS II enzyme
activity, because when the same reaction was carried out for 15 and 45 minutes, the
radiolabel incorporation did not increase as expected (data not shown). During the same
assay with other fractions, the incorporation of radio-labeled UDP-glucose increased
significantly when assayed for 45 min instead of 15 min. Immunoblotting indicated that
Golgi marker proteins AtSYP61 and AtXTl were almost exclusively found in fraction 3
(Figure 4.7B), indicating that fraction 3 was highly enriched in Golgi vesicles. The
plasma membrane marker protein PMA2 was also detected in fractions 3 and 4, although
the majority of it was detected in the plasma membrane rich fraction (fiaction 5, Figure
4.7B).

Almost equal amount of GS 11 activities were detected in fraction 4 and 5 (Table

176

 

34%
4 O

Homogenate
37-45%

 

 

5 -

Pellet a

2m]

7ml

8m]

8m]

10 ml

5ml

 

 

3 4 5
PMA2 i
AtCleZ iii 3“:
AtSYP61 1‘ ..
AtSEC12 114-1:: mm Mil

Figure 4.7 Fractionation of microsomal vesicles by floatation centrifugation on a

sucrose step gradient. 810 protein fraction isolated from Arabidopsis 5 day-old

suspension cultured cells was adjusted to contain sucrose of around 40% (37-45%) and

laid on top of the 50% sucrose cushion. On top of the sample, buffers containing 34%,

25%, 18% and 9.5% of sucrose were sequentially loaded (A). The gradient was

centrifiiged at 100,000x g for 3 hours and the membranes present at the interfaces

between different sucrose concentrations were collected. Equal amount of proteins from

fraction 3, 4 and 5 were resolved on SDS-PAGE and transferred onto PVDF membranes

for immunoblottings with different antibodies against selected marker proteins (B). Data

shown in this figure is representative of three different experiments.

177

 

Table 4.1: Specific activity of XyG fucosyltransferase and GSII for membrane vesicles
fractions isolated from sucrose step gradient. Gradient was described in Figure 4.7. Assays
were carried out according to Materials and Methods. These data are representative for three

different experiments with similar results.

 

 

FUTase activity GSII activity Protein Volume "
(pmol-g1-min‘1) (pmol-g"-min") (mg/ml) (m1)
810 1552 380 14.67 10
1 0 0 0.11 1.5
2 0 0 0.09 0.2 .-
3 208825 443 0.43 0.4 i
4 30385 1602 3.75 1.1
5 10684 1450 4.07 1.6
Pellet 0 306 5.57 2

 

 

 

 

178

4.1), while the majority of PMA2 proteins were found in fraction 5 by immunoblotting
(Figure 4.7 B). This inconsistency between the two plasma membrane markers was also
observed during the sucrose density linear gradient (Figure 4.5) where GSII activity
peaked twice in different fractions while PMA2 proteins were immuno-localized in
mainly those fractions showing the second peak of GSII activities (Figure 4.5). This

observation may indicate the presence of more than one enzyme that shows activity in the

.-.—‘l

GSII assay.

 

179

DISCUSSION

Three experimental procedures were used to investigate the localization of
AtCle2 proteins. Observations from those experiments led to the conclusion that
AtCle2 proteins are localized in Arabidopsis Golgi membranes. Firstly, AtCleZ was
detected in microsomal membrane fractions different from the plasma membrane marker
protein, PMA2, during linear sucrose density gradient fractionation. The fractions that
contained AtCle2 proteins also showed the highest XyG fucosyltransferase activity and
the presence of AtXTl, the XyG xylosyltransferase. These observations are consistent
with a Golgi localization for AtCle2. Secondly, during an aqueous two phase
partitioning, AtCle2 partitioned into the lower phase which contained highly enriched
XyG fucosyltransferase activity and most of the cellular organelles except the plasma
membrane, while the plasma membrane marker PMA2 partitioned into the upper phases
together with most of the G811 activities. Again the AtXTl co-partitioned with AtCle2
into the lower phase. Thirdly, when the microsomal membranes were fractionated by
sucrose density step gradient floatation centrifugation, AtCle2 was detected in the
fraction rich in Golgi vesicles, as evidenced by marker proteins AtSYP61 and AtXTl and
XyG fucosyltransferase activity, while most of plasma membrane marker PMA2 was
detected in the fraction rich in plasma membranes. When taken individually, none of
above observations could convincingly support a Golgi localization of AtCle2.
However, collectively these three observations provide compelling evidence for a Golgi
localization of AtCle2 proteins.

Electron microscopic (EM) localization studies would provide confirmation for

the conclusions described above. However, the purified antibody against AtCle2 still

180

 

recognized a soluble protein and a smaller membrane protein during immunoblotting
(Figure 4.2). These cross-reactive proteins would cause problems during analysis by EM.

AtCle2’s localization at the Golgi complex is contrary to a previous speculation
of a plasma membrane localization (Doblin et al., 2001) or evidence supporting a ER
localization (Favery et al., 2001). The prediction of plasma membrane localization of
AtCle protein was based on the speculation that NaClel was a tissue-specific cellulose
synthase (Doblin et al., 2001). NaClel was concluded to be a cellulose synthase based
on the fact that Cle proteins contain the D,D,D,QXXRW motif, similar transmembrane
domain structures, and fairly high sequence identities with CesA proteins, and the
observation of specific expression of NaClel in pollen tubes of Nicotiana alata (Doblin
et al., 2001). A single Cle gene (NaClel) and a single CesA gene (NaCesAl) were
detected by RT-PCR reactions in Nicotiana alata pollen tubes. While NaCesAl was
detected to be expressed weakly in many other tissues, it is not detected in the male
gametophyte (pollen or anther) or in vitro cultured pollen tubes (Doblin et al., 2001).
However NaClel was only expressed in male gametophyte tissues after pollen mitosis
and in vitro cultured pollen tubes at fairly high levels (Doblin et al., 2001). The authors
therefore concluded that NaClel was a cellulose synthase based on the assumption that
the high cellulose content in the pollen tubes should have a high expression of cellulose
synthase gene. In this case NaClel seemed to be the only candidate (Doblin et al., 2001).
Our data however, strongly support a Golgi localization of AtCle2, and therefore we
conclude that AtCleZ is not a cellulose synthase.

The localization of AtCle3 at ER came from the observation that the expression

of a AtCle3-GF P fusion protein under the control of constitutive 358 promoter in

181

Nicotiana benthamiana leaf cells led to GP P appearance in the endoplasmic reticulum
(Favery et al., 2001). However, high level expression of an integral membrane protein
directed to the secretory pathway could congest the pathway and cause the protein to
show up in all cellular compartments along the secretory pathway. For example, a GFP
protein fused to the signal anchor sequence of a human galactosyltransferase that was
delivered to the Golgi apparatus showed up mostly in the ER as fusiform bodies and only F
weakly in the Golgi (Hawes et al., 2001). i

It is possible that Cle1, D2 and D3 have different locations and different

 

functions as well, and the conclusions drawn from the localization studies described I
above might all be valid. However the fairly high sequence identity among AtClel , D2
and D3 proteins, from 66% to 88%, suggests that they are very likely to have similar
biochemical functions. Therefore other AtCle proteins may be located on the Golgi
apparatus just as AtCle2 is.

AtCle proteins are therefore not likely cellulose synthases or callose synthases
that are localized at the plasma membranes. All AtCle proteins contain the
D,D,D,QxxRW motif indicative of processive glycosyltransferases (Saxena et al., 1995;
Richmond and Somerville, 2000) and belong to the family 2 of the inverting nucleotide-
diphospho-sugar glycosyltransferases that synthesize repeating B-glycosyl unit structures
(Campbell et al., 1997; Richmond and Somerville, 2001). Therefore AtCle proteins
could be responsible for the backbone synthesis of hemicelluloses such as xyloglucan,
xylan, glucomannan, mannan and galactomannan, or galactan side chains that have B-

(1,4)-linkages on their backbones found on pectin, or AGPs etc.

182

AtCle2 behaved very similarly to the plasma membrane marker PMA2 in the
membrane stripping experiments, showing similar difficulty to be solubilized with
various reagents. This supports the hypothesis of a similar membrane topology as PMA2,
which is predicted to have four transmembrane domains near the N-terminus and six
close to the C-terminus (Morsomme et al., 1998). AtCle2 is predicted to have two
transmembrane domains near the N-terminus and also six close to the C-terminus.
AtCle2 is the first type III transmembrane protein localized in the Golgi complex

(Singer, 1990).

 

The purified antibody against AtCle2 is a valuable molecular tool that will be
useful in future studies. The antigen for this antibody is located near the N-terminus of
the protein. When combined with antibodies against other regions, the topology of
AtCle2 in the Golgi membrane could be examined. This topology study would
eventually tell if the C-terrninal catalytic domain faces the Golgi lumen or the cytosol.
The different topology of AtCle proteins might suggest a very different mechanism of
action. A cytosol-facing AtCle might utilize substrate present in the cytosol, transfer
them across the Golgi membrane, and incorporate them into the product in the Golgi
lumen. This enzyme mechanism is very similar to that of cellulose synthases that localize
at the plasma membrane and incorporate glucose from the cytosol side into cellulose
chains outside of the cell. On the other hand if the catalytic domain of AtCle is facing
the Golgi lumen, AtCle may catalyze reactions similarly to other glycosyltransferases,
utilizing substrate and transferring it into product at the same side of the Golgi

membranes.

183

REFERENCES

Albertsson, P.-A. (1977). "Separation of particles and macromolecules by phase
partition." Endeavour, new series 1 (2): 69-74.

Axelos, M., Curie, C., Mazzolini, L., Bardet, C. and Lescure, B. (1992). "A protocol
for transient gene expression in Arabidopsis thaliana protoplasts isolated from cell-
suspension cultures." Plant Physiol. Biochem. 30 (1): 123-128.

Bar-Peled, M. and Raikhel, N. V. (1997). "Characterization of AtSEC12 and AtSARl ."
Plant Physiol. 114: 315-324.

Baydoun, E. A. H., Abdel, M. R. M., Dani, D., Rizk, S. E. and Brett, C. T. (2001).
"Galactosyl- and fucosyltransferases in etiolated pea epicotyls: Product identification and
sub-cellular localisation." J. Plant Physiol. 158 (2): 145-150.

Benhamou, N. (1995). "Irnmunocytochemistry of plant defense mechanisms induced
upon microbial attack." Micros. Res. Tech. 31: 63-78.

Bolwell, G. P. (1993). "Dynamic aspects of the plant extracellular matrix." Int. Rev.
Cytol. 146: 261-323.

Brown, R. M. J. and Montezinos, D. (1976). "Cellulose microfibrils: visualization of

biosynthetic and oriented complexes in association with the plasma membrane." Proc.
Nat] Acad. Sci. USA 73: 143-147.

Brummell, D. A., Camirand, A. and Maclachlan, G. A. (1990). "Differential
distribution of xyloglucan glycosyltransferases in pea Golgi dictyosomes and secretory
vesicles." J. Cell Sci. 96 (4): 705-710.

Buckeridge, M. S., Vergara, C. E. and Carpita, N. C. (1999). "The mechanism of
synthesis of a mixed-linkage (l-3),(1-4)B-D-Glucan in maize. Evidence for multiple sites
of glucosyl transfer in the synthase complex." Plant Physio]. 120: 1105-1116.

Campbell, J. A., Davies, G. J., Bulone, V. and Henrissat, B. (1997). "A classification
of nucleotide-diphospho-sugar glycosyltransferases based on amino acid sequence
similarities." Biochem. J. 326: 929-939.

Delmer, D. P. and Stone, B. A. (1988). "Biosynthesis of plant cell walls." The
Biochemistry of Plants. J. Priess, New York, Academic Press Inc. 14: 373-421.

Doblin, M. S., Melis, L. D., Newbigin, E., Bacic, A. and Read, S. M. (2001). "Pollen

tubes of Nicotiana alata express two genes from different beta-glucan synthase families."
Plant Physiol. 125: 2040-2052.

184

 

Edwards, M. E., Dickson, C. A., Chengappa, S., Sidebottom, C., Gidley, M. J. and
Reid, J. S. G. (1999). "Molecular characterisation of a membrane-bound
galactosyltransferase of plant cell wall matrix polysaccharide biosynthesis." Plant J. 19:
691-697.

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-fucosyltransferase that catalyzes the last step of cell wall
xyloglucan biosynthesis in pea." J. Biol. Chem. 275 (20): 15082-15089.

Faik, A., Price, N. J., Raikhel, N. V. and Keegstra, K. (2002). "An Arabidopsis gene
encoding an alpha-xylosyltransferase involved in xyloglucan biosynthesis." Proc. Natl
Acad. Sci. USA 99 (11): 7797-7802. 5

F avery, B., Ryan, E., Foreman, J., Linstead, P., Boudonck, K., Steer, M., Shaw, P.
and Dolan, L. (2001). "KOJAK encodes a cellulose synthase-like protein required for
root hair cell morphogenesis in Arabidopsis." Gene. Dev. 15: 79-89.

 

Fredrikson, K., Kjellbom, P. and Larsson, C. (1991). "Isolation and polypeptide
composition of 1,3-beta-glucan synthase from plasma membranes of Brassica oleracea."
Physiol. Plant. 81: 289-294.

Gibeaut, D. M., Hulett, J., Cramer, G. R. and Seemann, J. R. (1997). "Maximal
biomass of Arabidopsis thaliana using a simple, low-maintenance hydroponic method
and favorable environmental conditions." Plant Physiol. 115: 317-319.

Handford, M. G., Baldwin, T. C., Goubet, F., Prime, T. A., Miles, J., Yu, X. and
Dupree, P. (2003). "Localisation and characterisation of cell wall mannan
polysaccharides in Arabidopsis thaliana." Planta 217 (3): 1073-1079.

Harlow, E. and Lane, D. (1988). Antibodies, a laboratory manual. Cold Spring Harbor
Laboratory.

Hawes, C., Saint-Jore, C., Martin, B. and Zheng, H.-Q. (2001). "ER confirmed as the
location of mystery organelles in Arabidopsis plants expressing GFP!" Trends Plant Sci.
6 (6): 245-246.

Horisberger, M. and Rouvet-Vauthey, M. (1985). "Cell wall architecture of the fission
yeast Schizosaccharomyces pombe." Experientia 41: 748-750.

Horisberger, M., Rouvet—Vauthey, M., Richli, U. and Farr, D. A. (1985). "Cell wall

architecture of the halophilic yeast Saccharomyces rouxii. An immunocytochemical
study." Eur. J. Cell Biol. 37: 70-77.

Hotchkiss, A. (1989). "Cellulose biosynthesis in plant cell wall polymers: biogenesis and
biodegradation". N. G. a. P. Lewis, M. G., American Chemical Society: 232-247.

185

Kauss, H. (1987). "Some aspects of calcium-dependent regulation in plant metabolism."
Annu. Rev. Plant Physiol. 38: 47-72.

Kimura, S., Laosinchai, W., Itoh, T., Cui, X., Linder, C. R. and Brown, R. M. J.
(1999). "Immunogold labeling of resette terminal celulose-synthesizing complexes in hte
vascular plant vigna angularis." Plant Cell 11: 2075-2085.

Konishi, T., Nakai, T., Sakai, F. and Hayashi, T. (2001). "Formation of callose from
sucrose in cotton fiber microsomal membranes." J. Wood Sci. 47 : 331-335.

Larsson, C., Widell, S. and Kjellbom, P. (1987). "Preparation of high purify plasma
membranes." Methods Enzymol. 148: 558-568.

Latge, J. P., Cole, G. T., Horisberger, M. and Prevost, M. G. (1986). "Ultrastructure
and chemical composition of the ballistospore wall of C onidiobolus obscurus." Exp.
Mycol. 10: 99-113.

Lu, S. X. and Hrabak, E. M. (2002). "An Arabidopsis calcium-depentend protein kinase
is associated with the endoplasmic reticulum." Plant Physiol. 128: 1008-1021.

Lynch, M. A. and Staehelin, L. A. (1992). "Domain-specific and cell type-specific
localization of two types of cell wall matrix polysaccharides in the clover root tip." J.
Cell Biol. 118 (2): 467-479.

Madson, M., Dunand, C., Li, X., Verma, R., Vanzin, G. F., Caplan, J., Shoue, D. A.,
Carpita, N. C. and Reiter, W.-D. (2003). "The MUR3 gene of Arabidopsis encodes a

xyloglucan galactosyltransferase that is evolutionarily related to animal exostosins."
Plant Cell 15: 1662-1670.

Meikle, P. J., Bonig, 1., Hoogenraad, N. J., Clarke, A. E. and Stone, B. A. (1991).
"The location of (1-3)-beta-glucans in the walls of pollen tubes of Nicotiana alata using a
(1-3)-beta-glucan-specific monoclonal antibody." Planta 185: 1-8.

Mizuta, S. and Brown, R. M. J. (1992). "High resolution analysis of the formation of
cellulose synthesizing complexes in Vaucheria hamata." Protoplasma 166: 187-199.

Moore, P. J., Swords, K. M. 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 (4): 589-602.

Morsomme, P., Dambly, S., Maudoux, O. and Boutry, M. (1998). "Single point
mutations distributed in 10 soluble and membrane regions of the Nicotiana
plumbaginifolia plasma membrane PMA2 H+-ATPase activate the enzyme and modify
the structure of the C-terminal region." J. Biol. Chem. 273: 34837-34842.

186

 

 

Mueller, S. C. and Brown, R. M. J. (1980). "Evidence for an intramembranous

component associated with a cellulose microfibril synthesizing complex in higher plants."
J. Cell Biol. 84: 315-326.

Mueller, S. C., Brown, R. M. J. and Scott, T. K. (1976). "Cellulosic microfibrils:
nascent stages of synthesis in a higher plant cell." Science 194: 949-951.

Northcote, D. H., Davey, R. and Lay, J. (1989). "Use of antisera to localise callose,
xylan and arabinogalactan in the cell plate, primary and secondary walls of plant cells."
Planta 178: 353-366.

Northcote, D. H., Davey, R. and Lay, J. (1989). "Use of Antisera to Localize Callose
Xylan and Arabinogalactan in the Cell-Plate Primary and Secondary Walls of Plant
Cells." Planta 178 (3): 353-366.

Perrin, R., Wilkerson, C. and Keegstra, K. (2001). "Golgi enzymes that synthesize
plant cell wall polysaccharides: finding and evaluating candidates in the genome era."
Plant Mol. Bio]. 47: 115-130.

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 (5422): 1976-1979.

Porchia, A. C., Sorensen, S. O. and Scheller, H. V. (2002). "Arabinoxylan biosynthesis
in wheat. Characterization of arabinosyltransferase activity in Golgi membranes." Plant
Physio]. 130: 432-441.

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

Richmond, T. A. and Somerville, C. R. (2001). "Integrative approaches to determining
Csl function." Plant Mol. Bio]. 47: 131-143.

Rudolph, U., Gross, H. and Schnepf, E. (1989). "Investigation of the turnover of the
putative cellulose-synthesizing particle 'rosettes' within the plasma membrane of F unaria
hygrometrica protonema cells." Protoplasma 148: 57-69.

Saxena, I. M., Brown, R. M. J., Fevre, M., Geremia, R. A. and Henrissat, B. (1995).
"Multidomain architecture of beta-glycosyltransferases: implications for mechanism of
action." J. Bacteria]. 177: 1419-1424.

Sherrier, D. J. and Vandenbosch, K. A. (1994). "Secretion of cell wall polysaccharides
in Vicia root hairs." Plant J. 5 (2): 185-195.

Singer, S. J. (1990). "The structure and insertion of integral proteins in membranes."
Annu. Rev. Cell Biol. 6: 247-296.

187

 

Staehelin, L. A. and Moore, 1. (1995). "The Plant Golgi Apparatus: structure, functional
organization and trafficking mechanisms." Annu. Rev. Plant Physiol. Plant Mol. Biol. 46:
261-288.

. Turner, A., Bacic, A., Harris, P. J. and Read, S. M. (1998). "Membrane fractionation
and enrichment of callose synthase from pollen tubes of Nicotiana alata Link et Otto."
Planta 205: 380-388.

White, A. R., Xin, Y. and Pezeshk, V. (1993). "Xyloglucan glucosyltransferase in Golgi E
membranes from Pisum sativum (pea)." Biochem. J. 294 (1): 231-238.

Widell, S. and Larsson, C. (1990). "A critical evaluation of markers used in plasma
membrane purification." The Plant Plasma Membrane: Structure, Function and Molecular '.
Biology. C. a. M. Larsson, I. M., Berlin, Springer-Verlag: 16-43. E

 

Zaar, K. (1979). "Visualization of pores (export sites) correlated with cellulose
production in the envelope of the green negative bacterium Acetobacter xylinum." J. Cell
Biol. 80: 773-777.

Zhang, G. F. and Staehelin, L. A. (1992). "Functional compartrnentation of the Golgi
apparatus of plant cells: immunocytochemical analysis of high-pressure frozen and
freeze-substituted sycamore maple suspension culture cells." Plant Physiol. 99 (3): 1070-

1083.

188

CHAPTER 5

Generation and Characterization of RNAi Transgenic Plants for AtCle Genes

189

ABSTRACT

Among all AtCsl protein families, AtCle proteins have the highest sequence
identity with cellulose synthases. AtCle proteins are also similar in size to AtCesA
proteins and possess the motifs specific for B-processive glycosyltransferases. In the
previous chapter, AtCle2 proteins were shown to be localized in the Golgi complex and
therefore are likely enzymes involved in the cell wall matrix polysaccharide biosynthesis.
However, the biochemical functions of AtCle proteins are still unknown. Here a reverse
genetic approach was taken to determine the biosynthesis of which wall polysaccharide is
the AtCle proteins involved in. Studies on mutants of single AtCle genes, including
the root hair loss mutant kjk (AtCle3), so far have not yielded a direct link between
AtCle proteins and biosynthesis of a specific cell wall polysaccharide. Therefore a
RNAi approach was used to generate transgenic plant that would possibly have
suppressed expression of multiple AtCle genes. The goal was to define visible
phenotypes and to establish a connection between the function of AtCle proteins and
the presence of a certain polysaccharide. Transgenic plants containing RNAi constructs
targeting AtCle genes showed deformed cotyledons, dwarfed adult plants, and
shortened roots with more branches. However the transcript suppression was only clearly
seen in cotyledons, and there was no clear alterations in sugar composition that could be
associated with transcript suppression. Therefore the exact biochemical functions of

AtCle proteins still could not be determined.

190

 

INTRODUCTION

AtCsl proteins are proposed to be involved in the biosynthesis of glycans that
form the backbones of various hemicellulosic polysaccharides, because of their sequence
similarities with AtCesA proteins (see chapter 3). Among all AtCsl proteins, the AtCle
family showed the highest sequence identity with AtCesA proteins (see chapter 3).
Results from the previous chapter also showed that AtCle2 proteins are localized in the
Golgi complex instead of the plasma membrane and therefore AtCle genes do not likely
encode cellulose synthase or callose synthase (see chapter 4). Due to the chemical
structure similarity shared by the XyG backbone and cellulose, AtCle genes have
become the most likely candidates for encoding the XyG backbone glucan synthase.
Therefore the goal of this chapter is to determine the function of AtCle genes.

The functional studies of AtCsl genes are still in the initial stages. In an effort to
identify genes involved in the synthesis of B-glucans in Nicotiana alata pollen tubes, a
Cle cDNA was isolated as well as a CesA cDNA (Doblin et al., 2001). This NaClel
gene was expressed abundantly in male gametophyte tissues, in mature pollen and in
growing pollen tubes, while the NaCesAl was not expressed in the male gametophyte
(pollen or anthers) at any stage of development (Doblin et al., 2001). Given the existence
of large amounts of cellulose, NaClel was proposed to be a male gametophyte specific
cellulose synthase (Doblin et al., 2001).

Among all cell wall polysaccharides, only cellulose and callose are synthesized at
the plasma membrane and directly incorporated into the wall, while all other
polysaccharides are synthesized in the Golgi complex and then transported to the cell

surface for their incorporation into the wall (Delmer and Stone, 1988). AtCle2, however,

191

was localized at the Golgi instead of the plasma membrane according to studies described
in chapter 4. Therefore a non-cellulosic function was proposed for AtCle2. It seems
reasonable to extend this hypothesis to all members of the Cle family of proteins. While
this hypothesis is contrary to the one of Doblin et a1. (2001) that Cle proteins are
involved in cellulose synthesis, it still does not provide a specific function for them. I am
therefore interested in further detemrining the biochemical functions of AtCle proteins.
At this time activity assays for cellulose synthase and XyG backbone synthase, the two
mostly likely enzymes that AtCle genes may encode, are both unavailable. Therefore I
decided to employ a genetic approach.

Genetics has played a critical role and been very successful in illustrating the
functions of genes involved in cell wall biosynthesis. The most prominent cases are the
characterizations of mutant plants that have led to identification of the cellulose synthase
genes. So far six of the 10 AtCesA genes, including AtCesAl, 3, 4, 6, 7, 8, have been
confirmed to encode cellulose synthase catalytic subunits (Arioli et al., 1998; Taylor et al.,
1999; F agard et al., 2000; Taylor et al., 2000; Scheible et al., 2001; Burn et al., 2002;
Desprez et al., 2002; Zhong et al., 2003). All of the functional confirmations for these
AtCesA genes were accomplished through the characterizations of mutants identified
through forward genetics (AtCesAI: rswl; AtCesA3sixr1; AtCesA6: procuste], ixr2;
AtCesA 7: irx3; AtCesA 8: irx1). These AtCesA mutants all showed decreased cellulose
content in their cell walls, disrupted wall structures, and altered plant morphologies. The
functions of AtCesAl and A3 were also confirmed through the characterizations of

antisense plants generated through reverse genetics (Burn et al., 2002)

192

 

Genetic studies also have been very successful to define the functions of a few
genes involved in the biosynthesis of matrix polysaccharides. Examples include the
identification of the XyG fucosyltransferase gene AtFutl (mur2) (Vanzin et al., 2002),
XyG galactosyltransferase gene AtGalT (mur3) (Madson et al., 2003), a RGII
glucuronyltransferase gene (NpGUTl) (Iwai et al., 2002), and a HGA glycosyltransferase
(qua!) (Bouton et al., 2002). In all cases, careful phenotypic and morphological it
analyses have revealed the absence of a particular sugar linkage from a specific matrix I

polysaccharide, which then led to functional propositions for the specific mutated genes.

 

1n the cases of identification of AtFutl and AtGalT, genetic approaches were used in E
combination with biochemical approaches by expressing the cDNAs in Pichia pastoris to
obtain specific enzyme activities (Perrin et al., 1999; Madson et al., 2003).

However the genetic studies for AtCsl genes have not been as successful. So far,
mutation in only three AtCsl genes yielded specific phenotypes. They are a T-DNA
insertion mutant of AtCslA9, rat4 (Zhu et al., 2003), a transposon insertion line of
AtCslA7 (Goubet et al., 2003), and a AtCle3 mutant kojak (kjk) (Favery et al., 2001;
Wang et al., 2001). The rat4 plants showed less than one fourth of crown gall formation
compared with wild type plants when their roots were infected with Agrobacterium.
tumefaciens A208, as well as decreased lateral roots (Zhu et al., 2003). AtCslA 7 plants
showed recessive embryo lethality and impaired pollen tube growth (Goubet et al., 2003).
lg'k plants showed impaired root hair growth with the root hairs ruptured soon after
initiation, giving rise to a bald phenotype for the roots (F avery et al., 2001). However
none of the mutations has been linked to changes in a specific sugar linkage within a

specific cell wall polysaccharide and therefore the functions of these genes could not be

193

deduced. At the same time mutant plants with defects in other AtCsl genes, including
AtCle4, AtCslGl and AtCslBS, did not show any visible phenotypes or detectable
changes in cell wall sugar composition (Richmond and Somerville, 2001). This could be
due to gene redundancy because the presence of multiple members in most AtCsl gene
families supply redundant functions. Therefore it seems the loss-of-function in multiple
genes in the same AtCsl gene family is necessary to see phenotypic defects. As the effect
of knock-out mutations of AtCsl genes is being determined by Chris Somerville’s group
as part of a 2010 project (personal communication), I decide to pursue the genetic studies
of AtCle genes using the RNAi approach. My goal was to generate transgenic plants
showing suppressed expression of more than one AtCle gene, leading to more easily
detectable morphological phenotypes.

RNA interference (RNAi) is a post transcriptional gene silencing phenomenon
induced by double-stranded RNA (dsRNA) that is found in fungi, plants, and animals,
including mammals (Bosher and Labouesse, 2000). The dsRNA can be a replication
intermediate of many viruses, can be produced by the transcription of inverted repeat
transgenes, or can be synthesized in vitro biochemically and delivered into cells (Voinnet,
2003). The dsRNA molecules are recognized as aberrant RNAs and degraded to 21-26
nucleotide RNA duplexes, called short interference RNAs (SiRNAs), by a dsRNA
specific endonuclease called Dicer (Bernstein et al., 2001; Billy et al., 2001; Grishok et
al., 2001; Ketting et al., 2001). Dicer family proteins usually contain putative RNA
helicase, PAZ, RNase III, and dsRNA binding domains. The RNase 111 domain is thought
to cleave the dsRN A into SiRNAs to mediate the RNAi response. In plants, there are two

distinct classes of siRNAs of 21 nt and 25 nt as found in tobacco, Arabidopsis and wheat

194

 

extracts (Hamilton et al., 2002; Tang et al., 2003; Voinnet, 2003). These two classes of
siRNAs in plants are probably generated by two different Dicer proteins (Tang et al.,
2003; Voinnet, 2003). The siRNAs then are incorporated into a RNA-induced silencing
complex (RISC). This RISC, through base paring, mediates the specific cleavage of
target RNAs at sequences with extensive complementarities to the siRNAs (Hammond et
al., 2000; Hammond et al., 2001; Hutvagner and Zamore, 2002; Martinez et al., 2002).
The large fragments resulted from the endonuclease cleavage of the target mRNAs then
are degraded by the general RNA degradation machineries with conventional
exonucleases (Han and Grierson, 2002).

One of the advantages of RNAi technique is that the siRNA can also induce the
degradation of RNA molecules other than the immediate target as long as these RNA
molecules show sufficient sequence identity with the immediate target RNA (Bosher and
Labouesse, 2000; Parrish et al., 2000). Another advantage, especially for plants, is that
when transformed with a construct whose transcripts form hairpins, a series of
independent lines will be generated with different degrees of phenotypes and different
degrees of mRNA suppression (Helliwell et al., 2002; Stoutjesdijk et al., 2002;
Waterhouse and Helliwell, 2003). This feature of the RN Ai approach will guarantee that
some transgenic plants with partial RNA suppression and partial phenotypes will be
available in case a knockout mutant is lethal.

Six RNAi constructs, including two targeting AtCle genes, and other four
targeting AtCsl genes of other AtCsl subfamilies were generated and introduced into
Arabidopsis plants, together with the vector control construct. It turned out that only the

two constructs targeting AtCle genes showed obvious phenotypes. Therefore transgenic

195

 

plants containing other constructs served as controls, along with the vector alone
construct. The transgenic plants with suppressed AtCle transcripts showed deformed
cotyledons, shortened adult plants and roots, and weakened cell adherence among
cotyledon cells. These phenotypic characterizations suggested that AtCle proteins are
developmentally important and might be involved in cell-cell adherence. However,
biochemical analyses of wall materials from the transgenic plants were not conclusive,
therefore the connection between the function of AtCle genes and the biosynthesis of a

particular wall polysaccharide still could not be established.

196

 

MATERIALS AND METHODS

Arabidopsis transformation and screening

The Columbia ecotype of Arabidopsis was used for all experiments. Arabidopsis
transformation was accomplished using the floral dip method (Clough and Bent, 1998).
A grobacterium GV3101 containing binary plasmid pCAMBIA3300 (Basta resistant)
carrying RNAi constructs were cultured overnight in 5 ml of LB medium with 34 pg/ml
Rifampicin, 25 pg/ml Gentamycin and 25 pg/ml Kanamycin. A large culture of 500 ml
then was inoculated with a 1:100 dilution of the overnight culture and allowed to grow
for 16 to 20 hours until the cultures reached stationary phase. Agrobacterium cells were
collected by centrifugation at 10,000x g at 4°C for 5 minutes, and resuspended in 5.0%
sucrose solution containing 0.05% (i.e. 500p1/L) Silwet L-77 (LEHLE SEEDS, Round
Rock, TX) to give a final OD600 of around 0.8. Arabidopsis plants growing in mesh
covered pots at about 4 weeks old with many opened flowers were dipped in the
A grobacterium solution for 3 to 5 seconds. Pots with transformed Arabidopsis plants
were covered with saran wrap and stored on their side overnight in the dark. Seeds were
collected from transformed plants about 4 weeks later and labeled as the T1 generation.

Arabidopsis seeds were sterilized and grown in MS medium according to
Fitzpatrick and Keegstra (2001). Transgenic seeds were screened on MS medium (plus
1% sucrose) containing 10 pg/ml Basta (Crescent Chemical, Islandia, NY) and 500 pg/ml
C arbenicillin (Invitrogen, Carlsbad, CA). Antibiotic resistant plants were picked at day 4
to 6 and moved to new MS plates for root measurement or other phenotype studies. When
the transgenic seeds were planted directly in soil, screening was achieved after seed

germination by feeding or spraying with water containing 100 pg/ml Basta.

197

 

 

Genomic DNA preparation

Arabidopsis rosette leaves were ground in liquid nitrogen; the powder was
resuspended in C TAB solution (AutoGen, AG00121) and incubated in 65°C for 1 hour.
Genomic DNA was then extracted using the AutoGenprep 240 (Genomics Facility,
Michigan State University). The dried genomic DNA was resuspended in water and

extracted with phenol-chloroform.

RNA preparation

Total RNA was prepared from various tissues using TRIzol Reagent (Invitrogen,
Carlsbad, CA) according to the manufacturer’s instructions and treated with DNase I
(Roche, 776785, Mannheim, Germany) according to the instructions. After determining

the concentration of RNA, all samples were stored at -80°C.

PCR and RT -PCR reactions

All PCR reactions were carried out in a themocycler PTC-200 (MJ Research,
Waltham, MA). One step RT-PCR reactions used the Superscript One-Step RT-PCR with
Platinum Taq (Invitrogen, Carlsbad, CA) according to the instructions. Two step RT-PCR
reactions used the Superscript II RNase H' Reverse Transcriptase (Invitrogen, Carlsbad,
CA) according to instructions, and Taq polymerase (Roche, Indianapolis, IN) for the

following PCR reactions.

Generation of RNAi constructs

198

 

An intermediate cloning vector named pGEM-GUS was engineered from pGEM-
7Zf(+) (Promega, Madison, WI) by inserting between its SmaI and Csp451 sites a 1.1 kb
GUS gene fragment released from vector plasmid pBIlZl (Clontech, Palo Alto, CA).
DNA fragments for each RNAi construct then were inserted into pGEM-GUS at both
ends of the GUS-SmaI-Csp451 fragment in opposite directions utilizing enzyme
recognition sites Xhol, EcoRI or KpnI in the T7 end and BamHI, HindIII or ClaI in the A
SP6 end. All plasmid engineering was carried in E. coli cells DHSa or JM101 (Promega, '-

Madison, WI). The sequences of constructs were confirmed by restriction digestion and

 

sequencing (Genomics Facility, Michigan State University). The finished RNAi '
constructs and the GUS fragment control were released by restriction digestion with XbaI

and Sacl, and inserted into binary vectors pCambia3300 (Basta+), pCambial300

(Hygromycin+), or pBIlZl (Kanamycinl) for different antibiotic resistances. All three

vectors contain 358 promoters for the transcription of the construct gene fragments.

Constructs in binary vectors were introduced into Agrobacterium competent cells

GV3101 (pMP90). Plasmids were isolated from Agrobacterium and transformed back

into DH50t competent cells to confirm the constructs in GV3101 (pMP90) cells. Finally

A grobacterium containing the RNAi constructs and control construct were used to

transform Arabidopsis Columbia plants.

Sugar composition analysis by gas chromatography
Cell wall monosaccharides were analyzed as alditol acetates after acid hydrolysis

(Blakeney et al., 1983; Reiter et al., 1993). Plant tissues were washed twice in lml of

70% ethanol at 70°C, lml of acetone at room temperature for 5 min, and dried. The dried

199

 

samples .were hydrolyzed with 250 pl of 1 M H2S04 at 121°C for 1 hour. 100 pg inositol

was added as control. The resulting sugars were reduced to alditols by adding 100 pl of

9 M ammonium hydroxide and then 1 ml of 2% (w/v) sodium borohydride (NaBH4) in

dimethyl sulfoxide, and incubated at 40°C for 90 min. The remaining borohydride was

destroyed by adding 250 pl of acetic acid to each sample. The alditols were converted

into alditol acetates by adding 250 pl of l-methylimidazole followed by 4 ml of acetic h

anhydride. Finally 8 ml of water was added in each sample to destroy excess acetic

 

anhydride. After the samples cooled to room temperature, 2 ml of dichloromethane
(CHzClz) were added into each tube to partition the alditol acetates. After sitting at 4°C '
for 2 hours, the dichloromethane phase was recovered. For alditol acetates separation, an
Agilent 6890 series GC system fitted with an Agilent DB-225 capillary column (Agilent,
Palo Alto, CA) and a flame ionization detection GC ChemStation software (Agilent) was

-1

used for quantitation. The temperature program used was 1 min at 160°C; 20°C min‘ to

200°C; hold for 5 min; 20°C minT'] to 240°C; hold for 11.5 min.

Microscopy

For light microscopy, tissues were fixed in 2.5% glutaraldehyde and 2.0%
paraforrnaldehyde in 0.1 M phosphate buffer (pH 7.4) overnight and washed with 0.1 M
phosphate buffer (pH 7.4) three times for 20 min each. Samples then were incubated in
0.1 M phosphate buffer (pH 7.4) with 1% osmium tetroxide overnight at 4°C and washed
three times for 20 min each with phosphate buffer. After dehydration samples were

infiltrated in Poly/Bed 812 resin and embedded in silicone molds and polymerized for 2

200

days at 60°C. Samples were then sectioned and stained with different dyes and observed
under light microscope.

For scanning EM, tissues were fixed in 4% glutaraldehyde in 0.1 M phosphate
buffer (pH 7.4) for 3 hours at 4°C and dehydrated. Samples then were dried at critical
point in a Balzers Critical Point Dryer using liquid carbon dioxide as the transitional fluid,
and mounted on aluminum stubs using adhesive tabs. Samples then were coated with 20
nm thick gold in an Emscope Sputter Coater model SC 500 purged with argon gas. The

samples were examined in a J EOL 6400 V SEM (Japan Electron Optics Limited).

 

201

RESULTS

Generation of RNAi constructs

 

It was reported in C. elegans that RNAi constructs of less than 100 bp long still
could induce gene silencing effectively, but the silencing effect increased with longer
constructs (Parrish et al., 2000). At the time when this work began, the shortest RNAi
construct reported to show gene silencing in Arabidopsis was 288 bp long (Chuang and
Meyerowitz, 2000). Therefore, 280 bp was chosen to be the lower limit for the construct
length. On the other hand it was reported that dsRNA produced by RNAi construct could

induce the silencing of related but not identical genes that share significant overall

 

homology and uninterrupted segments of significant length, probably more than 30 bp
(Sharp, 2001). Therefore during the design of RN Ai constructs targeting various AtCsl
genes, I decided to choose construct fragments sharing an identity of 80% or more, with
as many as possible of the other gene members within the same AtCsl family. The
rationale was that this strategy would provide the best chance for these RNAi constructs
to induce the suppression of more than one target gene.

Without any conclusive support for a specific function for any AtCsl genes, I
started with making RNAi constructs targeting genes of each AtCsl family, except for
AtCslA. AtCslA genes do not share more than 81% identity between any two of them. In
addition, more work needs to be done to confirm the re-annotated sequences (data not
shown and chapter 3). Using these criteria, six constructs were generated and designated
as D1235f3, D2312, B123456f3, C58f5, Elfl and G123f3. The first letter in their names
represents the AtCsl family. The numbers after the letter represent the genes within that

family showing more than 80% overall identity with the specific construct. The letter “i”

202

+1_ '1

designates “from”. The last number represents the gene member in that particular AtCsl
family was used as PCR or RT-PCR template for amplification of the RNAi construct.

The construct length, the restriction enzyme recognition sites designed at the end
of primers, the main parameters of PCR or RT-PCR for the construct amplification, and
the primers sequences are all listed in Table 5.1. The primers used for PCR reactions to
confirm the existences of RNAi constructs in transgenic plants are listed in Table 5.2.
The gene specific primers used to test the suppression of target gene transcripts are listed
in Table 5.3.

Six RNAi constructs were generated that should produce hairpin RNAs after
transcription (Figure 5.1). Two of them, D1235f3 and D2312, target AtCle genes, while
the other four, 812345613, C58f5, El fl and G123f3, target genes of other AtCsl families.
An empty vector control was also generated with only a fragment of GUS gene that is
used as the center loop region for all constructs (Figure 5.1). Later on it was found that
the transgenic plants containing construct B123456f3, C58f5, Elfl, or G123f3 did not
show any obvious phenotypes and therefore they also were used as controls.

The 280 bp DNA fragment of construct D123513 was generated by PCR reactions
using AtCle3 as template (Table 5.1). This fragment is derived from the most conserved
region throughout the whole AtCle family and shows more than 80% sequence identity
with the counterpart regions of AtClel , 2, and 5 (Figure 5.2). This DNA fragment
contains uninterrupted identical sequence segments of 24 bp compared with AtClel, D2,
and D4, 26 bp with D5 and 14 bp with D6 (Figure 5.2). Therefore, it was interesting to
see after the construct D1235f3 was introduced into plants which AtCle genes might be

suppressed. The 380 bp DNA fragment of construct D2312 was generated using the

203

 

 

Table 5.1: Design of RNAi constructs targeting different AtCsl genes and primers
used to create them. Each construct was amplified by PCR or RT-PCR reactions twice
with two sets of primers having different restriction enzyme recognition sites at their ends.
The two DNA fragments were then inserted into the cloning intermediate vector plasmid

pGEM-GUS utilizing the enzyme recognition sites introduced by the primers.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Construct Restriction Reaction Primers Primer sequences
(length) sites used
D1235f3 Xhol PCR, 60°C kee546 5’ TC'I‘TGTCIATG'I'ITCCCI‘CGAGAAGCGACCG 3’
(280 bp) EcoRI kee547 5’ CCCATGAATTCATCAAGTGCTCICATGTTCACA 3’
BamHl PCR, 55°C kee558 5’ CGTGAGAAGGATCCGGGTIATG 3’
Clal kee532 5’ ATCAACCCATCGATAGCT CT CAT 3 ’
D2312 Xhol PCR, 55°C kee548 5’ AAGCCACCTCGAGATGAACCAC 3’
(380 bp) Kpnl kee560 5’ ACCGGTACCATAAGCCCATCAACI’ 3’
BamHl PCR, 55°C kee550 5’ ACCT AGTGAGGATCCACT ACAC 3’
Clal kee56l 5’ ACCGGATCGATAAGCCCATCAACT 3’
812345613 X1101 RT-PCR, kee521 5’ GAAGATITCICGAGTACAAAACC 3’
(330 bp) Kpnl 50°C kee523 5’ TTCGTCGGCGGTACCATCATAGAAC 3’
BamHI RT-PCR, kee522 5’ AAGATITCTCGGATCCAAAACCAAA 3’
HindHI 50°C kee524 5’ GTCGGCGTTAGAAGCTTAGAACI‘C’IT 3’
C5815 Kpnl RT—PCR, kee525 5’ CCAAATGGTACCAAAAGGGAGT 3’
(710 bp) X1101 50°C kee527 5’ AGGGAGCT CGAGC’IT CT GGAA 3’
Clal RT-PCR, kee526 5’ TTGCCAAATCGATCCAAAAGGGA 3’
HindlII 45°C kee528 5’ AGGGAGGTAAGC’ITCTGGAACA 3 ’
El fl Xhol PCR, 55°C kee559 5’ GTCTCGAGGTCGTGGTITATTC 3’
(530 bp) Kpnl kee556 5’ AAACAGGTACCGGCAGATAAAG 3’
HindIII PCR, 60°C kee555 5’ GGTCTCAAGCITGTGGTITATTC 3’
Clal kee557 5’ TACAAAATCGATAAACAGGTACAG 3’
612313 Xhol PCR, 55°C kee533 5’ CACAACCCTCGAGACGTGTCAT 3’
(540 bp) EcoRI kee535 5’ CATCTT GAATTCT CAGCI’T CA 3’
HindHI PCR, 50°C kee534 5’ GGCGACACAAGCTTCCACACGTG 3’
Clal kee536 5’ GTACATCATATCGATATTI‘TCAGC 3’

 

204

 

 

Table 5.2: Primers used for PCR reactions to confirm the existence of constructs in

transgenic plants.

 

Primer Specificity Sequence

Kee578 (Cl) Spacer 5’ TGTTACGTCCTGTAGAAACCC 3’
Kee579 (C2) Spacer 5’ CCAGTTCAGTTCGTTGTTCAC 3’
Kee58l (C4) Spacer 5’ AACCACAAACCGTTCTACTIT 3’
Kee582 (bl) B12345613 5’ AAGAGTGTCAGGGTTGATGAC 3’
Kee587 (e1) Elfl 5’ AGATTTGAAAACAATGGGGAT 3’
Kee550 (d2) D2312 5’ ACCTAGTGAGGATCCACTACAC 3’
Kee727 (d3) D123513 5’ CATATACAACTCTCAAGGTTTA 3’

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

205

Table 5.3: Primers used for RT-PCR reactions to confirm the suppression of
transcripts. RT-PCR reactions were carried in two steps using the Oligo-dT primer for the
reverse transcription reactions. All cDNAs were treated with RNase and used as template
for PCR reactions with gene specific primers. “+”: the downstream primer kee625 comes

from the 3’ UTR region of AtCle3.

 

Gene Primers Primer sequences

Cth LM432 5' GGAATCTACTGCGCCTAAGT 3'
LM433 5' GCC TC AGAAAATCAAATGTACT 3'
AtCsIEOl Kee592 5’ TTATGGACACGATTAAGAAGC 3’
Kee593 5’ ACAGGTACAGGCAGATAAAGC 3’
AtCsID2 Kee620 5’ TCTCCTTCACGCTCACGAGT 3’
Kee621 5’ TTTGACAAGAAAGAAATCAACCT 3’
AtCle3+ Kee624 5’ CTCACTI‘AGCCGCGGTTATC 3’
Kee625 5’ GGCCAAAATGTGTCTCCATC 3’
AtCle5 Kee616 5’ CTTCAGCTI'ITGGGTGCTITG 3’
Kee617 5’ AACT'ITCATGCCTTGTACCAC 3’
AtClel Kee1197 5' GCATCACCGTAACCCTAACCC 3'
Keel 198 5' TAAGCGGCATGATGAATAAGCCT 3'
AtCsID4 Keell99 5' TCTTATGATCACCATTTGTITGATI’ 3'
KeelZOO 5' TATTATGTTCACCATTGCTATCAC 3'
AtCsID6 KeelZOl 5' AAGCCGCAAACCGCTTGATITCG 3'
Kee1202 5' CCGTCCTACAAACCGCGAAAAGT 3'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

206

 

 

 

 

 

Figure 5.1: Design of RNAi constructs for AtCsl genes. A. Recombinant plasmid
construction scheme. A 1.1 kb fragment of the GUS gene coding region from vector
plasmid pBIlZl was inserted into plasmid pGEM-7Zf through restriction sites SmaI and
Csp4SI, and used as the spacer loop region for all RNAi constructs. PCR amplified DNA
fragments from various AtCsl genes were inserted into both ends of the spacer fragment
in opposite directions to give rise to RNAi constructs through restriction sites for Xhol,
EcoRI, or KpnI and Clal, HindIII, or BamHI. RNAi constructs were released from
pGEM-7Zf via XbaI and SacI and introduced into binary vectors pBIlZl, pCambia3300,
or pCambia 3100 for Arabidopsis transformation. B. RNAi constructs including six of
them targeting different AtCsl genes and a control construct containing only the spacer
fragment. Open arrows represent the spacer fragment derived from the GUS gene, and the
filled arrows represent DNA fragments amplified from various AtCsl genes. The arrows
point from 5’ end to the 3’ end of all DNA fragments. Restriction sites used for plasmid

construction are indicated.

207

B

Control

D1235f3

D23t2

Elfl

B 123456f3

G123f3

Xhol

C58f5

 

pGEM-7Zf(+)

(2997bp)

 

TN

. pa
Aat 11

Sub]
Xhol
EcoRIk
KpnI

SmaI/
Csp451 ‘

Ch I
Hind III
BamH I

BstX I
Nsi I

t SP6

Ll

 

 

 

 

 

 

 

HingIII BamHI

 

 

 

 

V I V
EcoRI ClaI HindIII
V V
KpnI C 31

 

 

208

Spacer

 

 

 

HindIII
V

 

A J AtCle2

 

 

 

 

 

 

 

 

 

 

 

 

— L AtCle3
AtCle4
q ——l AtCleS
Nucleotide Substitutions (x100) AtCSID6
AtClel
l j l I I l l T I
43.1 40 35 3O 25 20 15 10 5 O
B
D1235f3 D1 D2 D3 D4 D5 D6

 

Identity 81.47% 87.76% 100% 76.22% 83.22% 75.87%
Uninterr 24bpx1 24bpx1 286bpx1 24bpx1 26bpx1 14bpx1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ppteq 14bpx1 23bpx1 14bpx1 20bpxl
'dem'cal 20 bp x 1 18 bp x 1
sequence
match 18 bpxl 17bpx1
17 bp x 1 14 bp x 2

D23f2 D1 D2 D3 D4 D5 D6
Identity 72.21% 100% 85.6% 73.25% 73.91% 74.23%
Uninterru 17bpxl 382bpx1 24bpx1 17bpx1 20bpx2 15bpx1
pted 23bpxl 14bpx2 14bpx3
identical 20 b p x 1
sequence

17 bp x l

14 bp x 1

 

 

 

 

 

 

 

 

 

Figure 5.2: Sequence similarities between RNAi construct D1235f3, D2312 and
AtCle genes. A. Alignment of AtCle cDNA sequences by the Multialign program of
DNAStar (Madison, WI). B. Sequence identities shared by RNAi constructs D1235f3,
D23t2 with the counterpart regions of each AtCle cDNA sequence; and the fiequencies

of uninterrupted identical sequence strings.

209

 

sequence of AtCleZ (Table 5.1). This fragment from AtCle2 contains the whole region
that is contained in the D1235f3 construct plus extra sequence on each end. It shows
more than 80% identity with AtCle3 (Figure 5.2). The longest uninterrupted stretch of
identities that D23t2 DNA fragment showed with its counterparts from AtClel, 4, 5, 6
are 20 bp and shorter (Figure 5.2); therefore plants transformed with D23f2 construct are

unlikely to show mRNA suppression of AtClel, 4, 5 and 6. The DNA fragment for a

 

construct 81234568 was amplified from AtCslB3 at the most conserved region

throughout the AtCslB family and shows more than 80% identity with its counterparts

 

from all other AtCslB genes. The same is true for construct G123f3. The DNA fragment E 1
of C585 is only conserved in AtCleS and C8 and did not show much identity with other
AtCle genes. E1 f1 was amplified from the only member of AtCle family at a region

showed low identity with AtCsl genes of other families (data not shown).

Generation of transgenic plants containing AtCsl RNAi constructs

All six RN Ai constructs that targeted AtCsl genes and the control construct
containing only the spacer segment were introduced into Arabidopsis individually (Figure
5.3). After transformation, plants were allowed to grow under normal conditions and
seeds were collected after about 4 weeks when the plants had matured. These Tl
generation seeds were germinated on MS plates with Basta for selection. The Basta
resistance gene is carried on the transformation vector. Genomic DNA was prepared from
rosette leaves of T1 plants that survived screening on antibiotics; the DNA was used to
confirm the presence of various constructs. The presence of the spacer segment in the

constructs was confirmed by PCR reactions using two specific primers C1 and C2

210

  
   

Agrobacteri:@

fr
1'7

\ A g}

.1

‘\ f.
. 1%
i

 

T2 seeds

 

T1 plants T2 plants

   

Figure 5.3: Transformation and screening of RNAi plants. Arabidopsis plants were

transformed by flower dipping method (Clough and Bent, 1998). T1 seeds were
germinated in MS agar plates with 500 ug/ml Carbenicillin and 10 pg/ml Basta, or in soil
directly with 100 ug/ml Basta. Plants survived the Basta screening were selected after 4-6
days and transferred to new plates for phenotype observation. They were transferred to
soil after another 9 days. T2 seeds were collected from individual Tl plants and screened
the same way as for T1 seeds. A. Flow chart of transformation and screening procedure.
B. Part of a screening plate for T1 plants carrying construt D1235f3 at 10 days old,
showing a couple of antibiotic resistant plants survived and most of other plants died right

after germination.

211

_R

(Figure 5.4). The presence of construct 31234568 was confirmed by two different PCR
reactions using spacer-fragment-specific primer C4 and 81234568 specific primer b1, or
using 81234568 specific primer b1 alone (Figure 5.4). The DNA fragment amplified
with spacer-fragment-specific primers would confirm the presence of the central loop
region of the constructs, while the PCR reaction using just one primer that is B123456f3
specific would not only confirm the presence of this construct but also confirm the A ‘
structure of the construct, showing head to head arrangement of the same DNA fragment

in the construct (Figure 5.4). The presences of RNAi constructs D23t2, D1235f3 and

 

Elfl were all confirmed by PCR reactions using construct specific primers d2, d3, and e1 i
respectively with a spacer-fragment-specific primer C4 (Figure 5.4). More than 25

independent transgenic plants were examined for each RNAi construct. All transgenic

plants that were positive during antibiotic screening showed the presence of a

corresponding construct.

RNAi transgenic plants containing construct D123513 or D23f2 show deformed
cotyledons and reduced adhesion of cotyledon cells

T1 seeds from transgenic Arabidopsis plants transformed with RNAi constructs
were screened on MS medium plates containing Basta. The screening procedure is
summarized in Figure 5.3. At 4 to 6 days old, a cotyledon deformation phenotype was
observed from some T1 seedlings (Figure 5.5). The phenotype was recorded as “strong”,
“intermediate 11”, “intermediate 1”, or “normal” according to the degree of cotyledon

deformation. Collectively about 50% of the T1 transgenic plants containing construct

D1235f3 showed cotyledon deformation of different degrees, and about 21% of the T1

212

 

 

 

 

 

Figure 5.4: Determination of the presence of RNAi constructs in transgenic plants.
A11 T1 transgenic plants that survived antibiotic screening were confirmed to contain the
appropriate constructs by PCR reactions using construct specific primers. At least 25
independent transgenic plants were tested for each construct. Data shown here is results
from only six plants for each construct. RNA samples for F and G are from the salt“? Six
transgenic plants. “M” is DNA molecular size marker, “+” are positive controls using

‘6 9’

construct plasmid DNA as PCR reaction template. are negative controls using RNA
prepared from wild type plants that were grown under similar conditions. A: Positioning
of construct specific primers used for PCR reactions. C1, C2 and C4 are primers specific
for the spacer fragment used as the central loop of RNAi constructs. C1 and C2 are used
for PCR reactions confirming the presence of the spacer fragment. C4 is used in
combination with construct specific primers b1, d2, d3 and e1 for 812345613, D23 f2,
D1235f3 and El f1 respectively. Sequences of all primers are listed in Table 5.2. B: PCR
reactions using primers C1 and C2 for RNAi plants containing the spacer control

construct. C-F: PCR reactions using primer C4 combined with primer d3 (C), d2 (D), 61
(E) and b1 (F) to confirm the presence of construct D1235f3, D2312, Elfl and

812345613 respectively. G: PCR reactions using primer b1 only to confirm the presence

of construct 812345613.

213

 

'ansgenicplait

. . C4 (:2 b1

“ ) 9 ‘\ n‘wa' ~
.1" L4 to Linn; 7;
E b a - «— «—

 

 

 

-§ ‘ '51. i;
.. fit .1. ‘... —.’ ———>

1 . .
r w: 1% ;
A. hint Av ...»-

 

Control

 

D1235f3

D2312

Elfl

8123456f3

. o
‘ 1
11')
. .

812345613

 

214

Control Normal Inter I Inter II Strong

   

 

0123513: 50.8% 19.5% 16.1% 13.6%
DZ3f2= 78.5% 11.4% 6.3% 3.8%

 

Figure 5.5: Cotyledon deformation phenotype of transgenic plants carrying construct
D1235f3 or D2312. Cotyledons were taken from T1 plants at 7 days old growing on MS-
Basta plates. A. Cotyledons from 7-day-old T1 plants containing RNAi construct spacer
control (Control), D1235f3, or D2312 that showed different degrees of cotyledon
deformation (Normal, Inter I, Inter II, and Strong). Numbers listed are the percentages of
T1 plants showing different degrees of cotyledon deformation for D1235f3 and D2312
lines. B. Representative T2 plants of 7 days old on screening plates, showing normal

cotyledons in control plants and strongly deformed cotyledons of D123513 strong lines.

215

 

 

transgenic plants containing construct D2312 showed cotyledon deformations (Figure 5.5).
T1 plants containing construct D123513 not only have more seedlings showing cotyledon
deformation, but also showed a more severe degree of deformation compared to plants
containing D23 12. While 13.6% of T1 plants containing D123513 showed strongly
deformed cotyledons, this number for T1 plants containing construct D2312 is only 3.8%
(Figure 5.5). In contrast, T1 transgenic plants containing the control construct or
constructs targeting AtCsl genes of other families did not show this phenotype (Figure
5.5 and data not shown).

The introduction of all six constructs into Arabidopsis was repeated at least three
more times by different transformation experiments. In each case, the same cotyledon
deformation phenotypes were observed with only plants carrying construct D123513 and

D2312. The percentages of T1 plants showing the phenotype are also similar to what

obtained from the first screening. In one case, the construct D1235f3 was introduced into
Arabidopsis plants in ten separate transformation experiments. The T1 generations from
these 10 plants all showed similar cotyledon deformation phenotype and similar
percentage of offspring that showed the phenotype. At the same time the control
experiment containing the construct with a GUS fragment spacer consistently showed no
cotyledon deformation.

When observed by SEM, the abaxial side of deformed cotyledons showed
distorted epidermal cells protruding from the surface, compared with cotyledons from

either control line or T1 plants showing normal cotyledons (Figure 5.6).

Transverse sections of cotyledons of normal, intermediate and strong deformation

were compared with those from control plants. The number of cell layers were not altered

216

 

 

Control Normal

      

200 um

Inter I Inter II

       

Figure 5.6: Abaxial side of cotyledons from T2 RNAi plants. Cotyledons were selected
frole-day-old plants grown on MS-Basta plates containing the control construct
(Control), or construct D123513 showing different degree of cotyledon deformation
(Normal, Inter I, Inter II and Strong). Samples were fixed in 4% glutaraldehyde,
dehydrated in ethanol series and critical point dried using liquid carbon dioxide as the
transitional fluid. Afier mounted on aluminum stubs using adhesive tabs, samples were

coated with gold (20nm thick) and examined in a J EOL 6400V SEM.

217

 

for cotyledons that are deformed (Figure 5.7). The vascular tissue also seemed unaffected
by the cotyledon phenotype (Figure 5.7). However, while the cells in control and normal
cotyledons did not show any lack of adhesion between each other, the cells in deformed
cotyledons displayed big gaps and were greatly detached from each other (Figure 5.7).

Furthermore, the cells in deformed cotyledons are more rounded compared with cells

 

from control or normal cotyledons (Figure 5.7). There seemed to be a correlation between
the degree of abnormality of cell adhesion and cell shape and the degree of cotyledon

deformation. However, a cause-and-effect relationship between them could not be

 

established, although the distortion of epidermal cells visible in deformed cotyledons

likely resulted from more airspace between cells.

0123513 transgenic plants showing deformed cotyledons have corresponding

suppression of A tC le gene transcripts

The levels of AtCle transcripts were tested by RT-PCR reactions using gene
specific primers (Figure 5.8). The initial efforts to detect the suppression of AtCle
expression used total RNA extracted from cauline leaves from D1235f3 T1 plants that
had been germinated on plates and transferred to soil at 6 days old and grown for another
4 weeks. Among 58 D123513 T1 plants tested, 18 showed strongly deformed cotyledons.
However only a few showed possible weak suppression of transcript levels for AtCle3
(Figure 5.8A). The transcript abundance of a cytochrome C gene was used as a unrelated
control, and that of AtClel was used as a control to show that the expression of a AtCsl

gene with low sequence identity with the RNAi construct was not suppressed (Figure 5.8).

218

 

Figure 5.7: Cells within deformed cotyledons of D1235f3 plants showed less adherence.
Cotyledons of T2 D123513 plants at 12 days old were fixed in 2.5% glutaraldehyde and 2%
paraforrnaldehyde, infiltrated in Poly/Bed 812 resin and embedded in silicone molds, and

sectioned. Sections were observed under light microscope afier Toluidine Blue staining.

219

 

 

 

A 33 35 36373940414246475358 626566676869M89

~

-*’- “CID-9%

 

--fifififim---- ~-
,_.. -mw-“—---g-- .- .

El
Cyto wuocfiq-u-Qip-C-Qu- ..

B 4 6 8 16 27 44 49 73 104 114

 

 

 

 

nsnsnsninninin nn

   
 
 

   

D2". -uouaco-—---—-h

E
E
Figure 5.8: Transcript abundance of AtCle genes in D123513 RNAi plants
examined by two-step RT-PCR reactions. The transcript level of AtCslBl and a
cytochrome C gene were used as controls. All reactions were repeated at least twice. A.
Total RNA was prepared from cauline leaves of 5-week-old D123513 T1 plants. “M”
represents DNA molecular markers. Numbers refer to individual D123513 T1 plants. B.
Total RNA was prepared from pooled cotyledons of 12-day-old T2 siblings showing
similar cotyledon phenotypes on MS plates. Numbers refer to their T1 parent plants. “11”
“i i” and “s ” srepresent pooled T2 siblings showing “normal”, “intermediate” and “strong”
cotyledon deformations. In the T1 generation, plants No. 4, 6 and 8 showed strongly

deformed cotyledons, No. 16, 44 and 49 showed intermediate cotyledon deformation,

while No. 27 and 73 had normal cotyledons. Plant No. 104 and 114 are control lines.

220

 

The expression of the cytochrome C gene and AtCslBl were found to be similar in all Tl
plants tested (data not shown and Figure 5.8).

Because the suppression of AtCle genes was not always found in adult Tl plants
(Figure 5.8A), I reasoned that the molecular phenotype could be absent in tissues that had
normal morphology. Therefore, the expression of AtCle genes was examined in the
cotyledons where the morphological phenotype was present. Total RNA was extracted
from cotyledons of 12-day-old plate-grown T2 plants. Cotyledons from 10 to 15 T2
plants showing similar cotyledon deformation phenotype and coming from the same
parent plant were pooled together. In this case, both AtCle3 and D5 were suppressed in
strongly deformed cotyledons (Figure 5.88). Other AtCle genes did not show any
suppression of their expressions. In summary the suppression of target transcripts could
be detected in deformed cotyledons but not in normal cotyledons or cauline leaves.

The sequence of RNAi construct D123513 was based on AtCle3. It showed the
highest identity of 87.8% with D2 among all AtCle genes (Figure 5.28). However the
messenger suppression of AtCle2 was not detected for all cotyledon-deformation-strong
lines tested (Figure 5.8). On the other hand, AtCleS shows the second highest identity
with D123513 of 83.2% and the maximum length of the uninterrupted sequence match
between them is 26 bp, which is longer than the 24 bp shown between construct D1235f3
and AtCle2 (Figure 5.28). Since AtCleS was clearly suppressed in the strong
cotyledon lines that also had AtCle3 suppression, it seemed like uninterrupted sequence
match is more important than the overall sequence match. This result makes sense

considering the siRNA are usual 21-25 bp long.

221

 

 

DI 235f3 transgenic plants showing deformed cotyledons also produced dwarfed adult
plants and shortened roots with more branching

It was observed that T1 plants containing RNAi construct D123513 have shorter
roots compared to T1 plants containing other constructs or control construct containing
only the spacer fragment (Figure 5.9). When grown side by side on vertical plates, the
average root length of D123513 Tl plants was significantly less than control plants
(Figure 5.9).

When the root length was compared among D123513 T1 plants with different
degrees of cotyledon deformation, it was found that those plants showing strong
cotyledon deformation had the shortest roots and the plants showing normal cotyledons
had the longest roots (Figure 5.10). Thus, reduced root elongation seemed to be
correlated with cotyledon deformation.

Plants with strong cotyledon deformation not only showed root shortening, but

also showed more root branching (Figure 5.11). Among D1235f3 T1 plants having

strongly deformed cotyledons, 54.5% showed root branching. Among D123513 T1 plants

showing normal cotyledons, only 23.9% had root branching. Taken together, 37% of

D123513 T1 plants showed some root branching. The root branching observed in control

lines and other RNAi lines was always less than 10%. However, the presence of this root

branching phenotype did not match exactly the cotyledon deformation or root shortening

phenotypes. The cotyledon normal D123513 T1 plants did not show root shortening

(Figure 5.9, 5.10) compared with control plants, but still showed the root branching

phenotype, although to a lesser degree than the plants with strongly deformed cotyledons

(Figure 5.1 1). Therefore, the root branching seems to be affected more easily by the

222

 

 

   
   
    

 

 

 

’ :O-Control l
L—l—8123413i
204 +C5815 i
E l 1+D2312 1
E 1 1+D123513i
ED 151 1-.—Elfl ;
E * _—+—G123B
O 1 W J
O
a:
101
« l
l
O T I l

 

Day 8 Day 11 Day 14

Figure 5.9: Root length of T1 transgenic plants containing RNAi constructs Control,
812345613, C5815, D2312, D123513, Elfl and G12313. T1 seeds were screened on MS

plates containing 10 ug/ml Basta and 500 ug/ml Carbenicillin. Basta-positive plants were
transferred to new plates at 5 days old and grown vertically. Root length were recorded at
the time points indicated. The numbers are averages of 75 T1 plants for construct D2312,

100 T1 plants for construct D123513, and 50 T1 plants for each other construct.

223

 

        
   

 

 

 

 

 

 

    

 

6 A
.9
g 5
a.
45 43 Inomal
5 l linter
"g 3 Elstrong
:3
Z 2 f
1* lil I l 1
l
0 1 1
(a m '\ C‘ '_ ('1
Rootlength(mm)— — 9 t g _ M w
N N "N” g 05 _
N M a
251
19.9 B
E 20
E
.1:
a}, 15
8
2 10.4
8 10'
at
5 .
01" 7777777 7%
normal inter strong

Figure 5.10: Root length of T1 plants containing construct D123513. 149 Basta resistant
T1 plants were picked from screening plates and transferred to vertical plates at 5 days old.
These plants were grouped as “normal”, “inter” and “strong” according to the degree of
their cotyledon deformation. Root length was measured at age 14 days. A. Numbers of
plants that have a specific root length for plant groups “normal”, “inter” and “strong”. B.
Average root length for D123513 T1 plants showing normal, intermediate and strong
deformed cotyledons.

224

 

 

 

 

Figure 5.11: T1 plants containing construct D123513 showing deformed cotyledons
also showed more branched roots. About 100 positive transgenic D123513 RNAi lines
screened from MS plate with antibiotics were picked and transferred at day 6 to new
plates for vertical growth. Cotyledon deformation and root branching were recorded at
day 14. A. T1 plants containing the control or D1235f3 RNAi construct on vertical MS
plates at 14 days old. 8. Numbers of D123513 T1 plants showing deformed cotyledons
and branching or non-branching roots. Filled bars represent the numbers of plants
showing branched roots, unfilled bars are the numbers of plants show unbranched r0015-
Filled triangles are percentages of plants showing branched roots among the plants in

each category of cotyledon deformation.

225

A

’med tablets
T351181:

nwmm

w

Number of plants

A 54.5%

A 36.4%

Strong Intermediate

226

 

 

 

 

 

' 60%
35
7 50%
r 40%
r 30%
A 23.9%
' 10%
0%
Normal

 

presence of RNAi construct D123513 than other processes such as cotyledon
development or root elongation. The mechanism behind this observation could not be
established from these experiments.

Plant development also was altered by the presence of RNAi construct D1235f3.
The T1 plants showing strong cotyledon deformation were also smaller in overall size
and shorter in plant height (Figure 5.12). At six weeks of age, the D123513 Tl plants
showed a very good correlation between their overall size and the cotyledon deformation
phenotypes recorded at day 6. Plants with more severely deformed cotyledons were
smaller in size at age 6 weeks (Figure 5.12). It was also noted that D123513 T1 plants
with strong cotyledon deformation and small overall size yielded fewer seeds (data not
shown).

T2 plants with strong cotyledon deformation phenotypes also showed dwarfed
adult plants as with the T1 parent plants (Figure 5.13). Twenty T2 plants from a T1 plant
with strong cotyledon deformation were picked up and transferred to soil. Ten of them
showed strong cotyledon deformation and the other ten showed normal cotyledons. At six
weeks, the seedlings with strong cotyledon deformation were much smaller than those
with normal cotyledons (Figure 5.13).

Despite the phenotypes observed for roots and whole seedlings, we have not yet

looked at the suppression of AtCle transcripts in either roots or bolting hypocotyls.

C otyledon deformation is inheritable for 01235f3 transgenic plants but weakens when

passed to the next generation

227

 

 

 

 

Figure 5.12: D123513 T1 plants had deformed cotyledons also yielded dwarfed adult
plants. A. Typical 6-week-old D1235f3 T1 plants with different degrees of cotyledon
deformation. Plants were germinated on MS plates with antibiotics and transferred to soil
at 14 days old. 8. Cotyledon deformation phenotype of plants shown in A. “5”, “ii”, “i”
and “n” represent cotyledon deformation phenotype of strong, intermediate 11,
intermediate I and normal. C. Stem length of mature plants. About 50 Control and 100
D123513 T1 plants were selected from MS screening plates at 14 days old and transferred
to soil. Stem length was measured when plants matured, but before seed collection. Plants

were grouped according to their degree of cotyledon deformation as recorded at 6 days old

on screening plates. Numbers shown are the averages.

228

 

 

0

Stem length (mm)

450 7

400

350

300

250

200

150

100

50

399 393
332
| 266

Control

 

D123513-n

229

D123513-Vii

 

D123513-s

 

 

 

Figure 5.13: Adult T2 plants containing RNAi construct D1235f3 at 6 weeks. All
plants were derived from a single T1 plant D123513-4 that showed strong cotyledon
deformation. Seedlings 1-10 showed strong cotyledon deformation, and seedlings 11-20

showed normal cotyledons when observed at 6 days old on MS plate.

230

 

D123513 T1 plants with different degrees of cotyledon deformation were selfed to
monitor the inheritance of cotyledon deformation phenotype. From a typical T1 screening
experiment for plants containing construct D1235 13, 118 antibiotic resistant T1 plants
were selected. Approximately 50 T2 seeds derived from each of the 12 strong, 14
intermediate ii, 16 intermediate i, and 19 normal cotyledon T1 plants were germinated the
same way as for T1 seeds. When the cotyledons of these T2 plants were examined, all of
the offspring from the parents with strong or intermediate cotyledon deformation showed
phenotype segregation or phenotype disappearance (Table 5.4). Out of the 12 strong T1
lines, T2 progenies from 4 of them showed strong, intermediate, and normal cotyledons
in different siblings (Table 5.4); T2 progenies from 6 other T1 lines showed only
intermediate and nomral cotyledons, while the 2 other T1 lines totally lost the cotyledon
deformation phenotype at the T2 generation (Table 5.4). Similar phenotype weakening
and segregation was also observed for those 20 intermediate cotyledon T1 lines (Table
5.4). At the same time, T1 plants that did not show any cotyledon deformation did not
show any deformation in their progenies (Table 5 .4). Therefore the cotyledon
deformation phenotype can be passed to the next generation together with the antibiotic
markers, but the phenotype weakens and segregates.

In order to determine the number of transgenes inserted in D1235f3 T1 plants and
to find out if the inheritance of cotyledon deformation phenotype follows Mendelian
genetics, the numbers of T2 progenies that are Basta resistant (alive on screening plates)
or sensitive (died right after germination on screening plates) were recorded for
individual T1 parents (Table 5.5). Surprisingly, the ratios of Basta resistant to Basta

sensitive plants were always high among T2 progenies coming from parents with strong

231

 

 

 

 

 

 

 

 

Table 5.4: Summary of the inheritance of cotyledon deformation phenotype from T1
to T2 generation for D123513 RN Ai lines. The cotyledon deformation phenotypes of T2

offspring from 12 strong, 15 intermediate and 10 normal Tl lines were tested.

 

 

T1 Number of Number of T1 plants whose T2 offsprings
cotyledon T1 plants showed cotyledon phenotype as:
phenotype tested 3, i, or n i, or n n

s 12 4 6 2
ii 14 12 2
i 16 8 8
n 19 19

 

 

 

 

232

 

 

 

 

E" " 'm‘-x

Table 5.5: Inheritance and segregation of cotyledon deformation phenoty P6 and

an
Basta resistance marker for D1235f3 RNAi plants. T1 plants were select ad from

. . . . . . ed at 6
antrbrotrc screenrng plate and therr cotyledon deformation phenotype were 1'6 Cord

as for
days old. T2 seeds from each Tl plant were germinated on MS plates the 83m 6 way

1 died
T1 plants. T2 plants with different cotyledon phenotype were counted. Seed 5 tha

short after germination were counted as dead plants (as in Figure 5.38).

 

T1 plants T2 plants

 

 

 

 

 

plant no. cotyledon n i 5 died alive/dead

3 s 10 3 0 2 13/2

4 s 39 20 27 3 86/3

5 s 55 2 O 2 57/2

6 s 13 10 13 0 36/0

8 s 38 O 19 2 57/2

9 s 39 0 O 0 39/0

10 s 45 O 2 45/2

11 s 37 1 O 0 38/0

12 s 4 10 O 1 14/1

13 s 41 3 0 0 44/0

14 s 18 24 0 3 42/3

16 s 10 14 4 14 28/14

18 ii 13 16 O 9 29/9

19 ii 13 5 O 1 18/1

21 ii 41 6 O 11 47/11

22 ii 22 2 0 1 24/1

23 ii 2 22 O 6 24/6

26 ii 18 7 O 1 25/1

nion phenol)?“ 27 ii 13 8 0 9 21/9
28 ii 21 15 O 2 36/2

, (,3: - 29 ii 6 14 O 4 20/4
30 ii 26 O O 12 26/12

“ .1 - 31 ii 23 9 O 9 32/9
32 ii 18 12 O 2 30/2

33 ii 36 4 O 0 40/0

_ 3 34 ii 43 o o 0 43/0
2 I 35 1O 0 6 45/6

, \ - . 39 1 20 5 O 9 25/9
40 l 9 16 0 3 25/3

3, 41 1 17 0 O 9 17/9
42 I 44 0 0 2 44/2

43 1 33 O 0 8 33/8

44 1 20 12 O 13 32l13

45 l 31 O O 8 31/8

46 I 11 21 0 9 32/9

47 I 38 O 0 2 38/2

48 I 55 O 0 3 55/3

49 1 19 22 0 0 41/0

50 I 19 16 0 12 35/12

51 l 18 O O 10 18/10

52 1 7 O O 25 7/25

54 I 24 13 0 0 37/0

60 n 14 0 0 4 14/4

61 n 30 O O 10 30/10

62 n 12 O 0 5 12/5

63 n 17 0 O 3 17/3

64 n 15 O 0 6 15/6

65 n 27 O 0 5 27/5

66 n 31 O O 11 31/11

234

cotyledon deformation (Table 5.5). On the other hand, the ratios of Basta resistant to
sensitive T2 siblings from parents with normal cotyledons were significantly lower and
much closer to the 3:1 ratio expected for a single gene inheritance (Table 5.5). It seemed
like the presence of multiple copies of RNAi construct is necessary to maintain the
cotyledon deformation phenotype, or in other words, to maintain a RNAi effect.

In order to determine if the cotyledon deformation phenotype can be inherited
past two generations, more than 50 independent T1 D123513 plants showing strong
cotyledon deformations were self-fertilized until the T4 generation. Overall the number
of plants with deformed cotyledons decreased from generation to generation but still
followed the rule that only parents with strong cotyledon deformation produced offspring
with the strong phenotype. The offspring from most of the lines with strong cotyledon
deformation at the T1 generation completely lost the cotyledon phenotype by the T4
generation. Only a few plants with strong cotyledon deformation at the T1 generation
kept the same phenotype until the T4 generation. T1 plants 4 and 8, that showed strong
transcripts suppression of AtCle3 and D5 (Figure 5.8), are two that still displayed this
phenotype in the T4 generation.

To see if a phenotypically stable transgenic RNAi line with a single transgene
insert could be identified, T1 seeds from different transformation experiments were
screened again and again. During one of the last screening experiments for D123513 T1
plants, 12 T1 lines showed strongly deformed cotyledons. When these 12 lines were
passed to the T2 generation, one of them, number 94, not only showed perfect 3:1 ratio
between Basta resistant and sensitive siblings, but also showed strongly deformed

cotyledons in all the T2 offspring with Basta resistance. This seemed a rare event

235

 

 

considering the numbers of T1 plants that have been tested. This particular line is still
being tested for the inheritance of the cotyledon deformation phenotype and the transgene
segregation. One possible explanation for this particular T1 line is that a single D1235f3
RNAi construct was inserted into a transcriptional hot spot and therefore the dsRNA
transcript level was high enough to induce the RNAi effect, which in many of the other

lines was induced by the transcription of multiple copies of the inserts in the genome. a

0123513 RNAi transgenic plants did not show clear alteration in cell wall sugar

 

composition

To see if the monosaccharide content of the wall materials from D123513 lines
have been altered, cell wall material was prepared and analyzed for sugar content. The
alditol acetate forms of sugars were examined by gas chromatography. When the wall
materials were prepared from rosette leaves of different T1 D123513 plants, none of the
monosaccharide content was significantly different from controls (Figure 5.14).

Four strong T1 lines, whose offspring at T4 generation still showed some
cotyledon deformation, were selected for cell wall monosaccharide analysis. Young
seedlings of 12 days old from T2 siblings of these 4 lines were pooled together for
preparation of cell wall material. The sugar composition of these cell walls is shown in
Figure 5.15. Although the glucose content of the seedlings with strong cotyledon
deformation was always less than that with normal cotyledons (Figure 5.15A), the
numbers are not statistically significant (Figure 5.158). Besides, the glucose content is
the most difficult to compare due to its widespread presence in various macromolecules

and the contribution from residual starch (Figure 5.15).

236

 

 

 

1
45% 1 ‘ ‘
I Control average ‘
40% J i I n average i
‘ El ii/i average i
35% i J 5 average

 

Rha Fuc Ara Xyl Man Gal Glu

Figure 5.14: Cell wall sugar composition of rosette leaves from T1 D123513 plants
showing different cotyledon deformation. All plants were from the same screening plate.
Samples included 4 Control, 5 cotyledon normal, 8 intermediate, and 15 strong lines. Bars
represent average values for each cotyledon category with standard deviation. Cell wall
preparation, treatment, and measurement of the sugar composition were described in

Materials and Methods.

237

 

 

Figure 5.15: Sugar composition analysis for T2 plants of D123513 RNAi lines. Walls
were prepared from 5 to 10 T2 sibling plants at 12 days old that showed the same strong
(5) or normal (11) cotyledon phenotype. T1 lines 4, 6, 163, and 176 showed strong
cotyledon deformation. Control includes two lines of plants containing spacer loop only
construct. They are processed the same way as for D123513 plants. A. Relative sugar
contents of individual lines. 8. Average sugar contents for seedlings from A that showed

strong (8) or normal (11) cotyledon phenotypes. Standard deviations are included.

238

IV 1111161-

.13.. 3.. . ..;

1‘ an J;-"1'
R...1..~~- r

arr. ..
..a (W.
,...-

. - 1
. u
”max-W

70%
60% ‘

50%

30%

20%

0%1

 

 

 

#1

70% 7 B
60%
’ I control average
500, 1 Inaverage 1
/o 1
F15 average
40% ‘
30% 1
20% 3
1
10% ‘
0%.],fiji ,ijz-l- .
Rha Fuc Ara Xyl Man Gal Glu

239

 

 

 

Due to the limitation on tissue availability, wall material fractionation was not
feasible and therefore I could not determine whether the amount of any particular

polysaccharide was altered by the introduction of the RNAi construct.

 

240

 

 

DISCUSSION

RNAi methodology was used to generate transgenic Arabidopsis plants with
decreased expression of AtCle genes, trying to shed some lights on the biochemical
functions of AtCle proteins. A couple of interesting observations were made for the
RNAi plants that contain construct D123513 and showed suppression of AtCle3 and
AtCleS expressions. These include deformed cotyledons and weakened cell cell
adhesion within cotyledons, shortened roots with more branches, and dwarfed adult
plants. These morphological alterations resulted from the presence of the RNAi construct
indicating that AtCle proteins have an important function during plant growth and
development.

However, no biochemical function could be proposed for the AtCle proteins
because the cell wall sugar composition analysis did not reveal significant difference.
Cell wall monosaccharide composition analysis has been successfully used during mutant
plant characterization to connect the function of a gene with the altered content of a
specific sugar in the wall, and in turn suggesting possible biochemical functions for the
protein encoded by the particular gene (Bouton et al., 2002; Iwai et al., 2002; Vanzin et
al., 2002; Madson et al., 2003). In our case, the monosaccharide composition assay was
made more difficult by the fact that glucose is the sugar most likely to be affected by the
presence of RNAi constructs targeting AtCle genes. AtCle genes have fairly high
degree of identity with AtCesA proteins at 39% to 46% (Table 3.3). They also share very
similar protein structure with common protein motifs and similar transmembrane domain
organizations (Chapter 3). Therefore AtCle proteins might have very similar enzymatic

mechanisms with AtCesA proteins, i.e. using UDP-glucose as substrate and incorporate

241

 

 

 

B-linked glucans. Glucose however has been notoriously difficult to measure among cell
wall monosaccharides because of its widespread presence among different wall
polysaccharides. Glucose could also be obtained during wall material preparation as
contaminant from other molecules, such as starch. One way to circumvent the problem
imposed by analyzing wall glucose content is to analyze the amount of major wall
polysaccharides directly. A decrease of a certain polysaccharide or a certain sugar moiety
within a particular polysaccharide would provide hints for the biochemical functions of
the proteins encoded by the suppressed gene(s). However, limited tissue availability has
prevented this approach because a stable transgenic line with a strong phenotype was not
yet established.

The RNAi methodology was still in its infantry when I started the work described
in this chapter. My data obtained using this technology also revealed some questions
regarding how the RNAi method works in plants, especially the maintenance and
inheritance of the phenotypes induced by RNAi constructs.

Suppressed expression of multiple genes seems necessary to obtain a visible
phenotype when the gene in question is part of a family and other family members supply
redundant functions. This seems to be true for the families within AtCsl super family.
One piece of evidence comes from the observation that single knock-out mutants of
AtCle4, AtCslGl and AtCslBS did not show any visible phenotypes or detectable cell
wall sugar composition alteration (Richmond and Somerville, 2001). Functional gene
redundancy might also play a role when using the RNAi approach to suppress gene
expressions. In my study the tissue deformation phenotype was only observed for the

cotyledons where both AtCle3 and AtCleS were suppressed. For other tissues that did

242

 

 

 

not show deformation phenotype it was likely that the suppression of multiple AtCle
genes did not happen. That was certainly the case for the cauline leaves.

Suppression of AtCle genes other than D3 and D5 was not detected, which
might be the reason that no more profound phenotype was observed for the RNAi plants.
This could be caused by the overall low sequence identities shared by the construct
D123513 with other D members, or because there are not enough sufficiently long
uninterrupted stretches of identity. As indicated in Figure 5.28, AtCleS shares the
highest uninterrupted sequence identity 0126 bp with construct D123513, and none of the

other AtCle members showed uninterrupted sequence identity of more than 24 bp with

 

the construct D123513.

The root hair disappearance phenotype presented by the AtCle3 mutant kojak
was not observed in any RNAi plants, even with the RNAi construct D23513 that is
immediately targeting the same gene. One possible explanation is that the RNAi
approach does not completely suppress the expression of a gene whereas many
conventional mutants are complete nulls. Another possible explanation is that the
promoter used for the RN Ai approach does not provide sufficiently high expression in
root hair cells to express the null phenotype in these cells.

Another possibility for not seeing more severe phenotypes could be that when
more complete suppression of multiple AtCle genes occurred, it led to lethality to the
transgenic plants that died during the screening. Indeed, some of the transgenic plants
containing the RNAi construct D123513 showed deformed true leaves but died on

screening plates shortly afier the appearance of the true leaves.

243

The major phenotype for RNAi plants containing construct D123513, cotyledon
deformation, was identified as a series from weak to strong. The presence of T1 plants
with different degrees of phenotype was also observed in other cases using RNAi
approach and was regarded as one of the unique features and advantages for this
approach. Almost all T1 plants with strong phenotype had multiple transgene insertions
revealed by the fact that the segregation ratios for the antibiotic marker are much higher 7"
than 3: 1. In contrast, T1 plants without a phenotype showed ratios much closer to 3:1. It

seems that multiple copies of transgenes are needed to maintain the phenotype caused by

 

the RN Ai construct, or in fact a concentration threshold of the RNAi construct transcript
has to be reached in order to induce a RN Ai effect. The cotyledon deformation phenotype
was inherited up to the T4 generation but kept weakening and segregating. A possible
explanation for this is that the segregation of inserts diluted the concentration of the
dsRNA transcribed from the RN Ai construct from generation to generation, and therefore
a weaker and weaker RN Ai effect was observed. When the concentration of dsRNA
transcripts becomes below the concentration threshold, the phenotypes caused by the
RNAi effect disappeared. The phenotype drifiing and weakening or even reversion
sometimes for the RN Ai approach has also been noticed by other groups as well (Dr. Jen
Sheen, Dr. Jiankang Zhu, personal communications).

Among those cases where the RNAi approach has been successfully used in
plants to study gene function (Chuang and Meyerowitz, 2000; Smith et al., 2000; Escobar
et al., 2001; Wesley et al., 2001; Klahre et al., 2002; Stoutjesdijk et al., 2002), only two
of them provided more detailed descriptions of the most important features of RNAi

approach in plants, namely the appearance of phenotypes, the suppression of target

244

transcripts and the inheritance of phenotypes (Chuang and Meyerowitz, 2000;

Stoutj esdijk et al., 2002). Chuang and Meyerowitz (2000) used RNAi constructs to
reproduce the knock out phenotypes for four genes that are involved in flower
development. The same phenotypes as with knock out mutants were recorded but in a
series from weak to strong as observed for the cotyledon deformation phenotype in my
case. In situ hybridizations indicated correlated transcripts suppression levels with
phenotypes. However 100% suppression of transcripts was not achieved as seen from the
weak in situ hybridization signals shown in strong lines.

Stoutjesdijk et a1 (2002) used RNAi constructs to successfully suppress the
expression of the F AD2 AlZ-desaturase gene in Arabidopsis seeds. Suppression of target
transcripts was not examined but the oleic desaturation proportion (ODP) controlled
mainly by the enzyme encoded by this gene was examined to monitor the RNAi effect.
This molecular phenotype also covered a gradient from weak to strong in the T1
generation. It was found that a couple of T1 lines with the most severe phenotypes all
contained multiple transgene insertions revealed by DNA blottings. Similar phenomena
was observed in my study as well as described earlier.

In both cases, the phenotype segregation in T2 and possibly later generations was
not addressed in detail. However, at least one “good” line was identified to have just a
single transgene insert and the phenotypes were stably inherited to at least T2 (Chuang
and Meyerowitz, 2000) and T5 (Stoutjesdijk et al., 2002) generations. However, they did
not report how many T1 lines needed to be screened to find a “good” line. It was possible
that those a few “good” T1 lines are rare events that had the transgenes inserted into

transcriptional super active regions of the genome and therefore could produce high

245

 

 

 

concentrations of dsRNAs that in other T1 lines needed multiple transgene insertions to
obtain, to induce the RNAi effect. Toward the end of my studies one such “good” line
was identified after monitoring the inheritance of more than 100 T1 lines with strong
cotyledon deformation phenotype. This particular line showed a perfect 3:1 ratio for the
selection marker and all of its T2 offspring showed the same strong phenotype. This
particular line, once established, could be very useful for analyzing the proportions of cell
wall polysaccharides. If this data could be obtained, it might provide a specific
hypothesis regarding the biochemical functions of AtCle proteins.

The study described in this chapter began when the mechanism of RNAi approach

 

was still not well understood. Consequently, the strategies needed to consistently
suppress gene expression were not well understood. Given the results described in this
chapter and the new knowledge about the RNAi approach, if someone were to start over
again, a couple of modifications should be made to more successfully suppress the
expression of AtCle genes. First, the intron containing RNAi vector, pHANNIBAL
should be used instead of the homemade vector containing a GUS fragment as loop
region. Intron containing RNAi constructs generally produce 90% to 100% of
independent transgenic plants showing the silencing effect, while RNAi constructs with
other DNA fragment as the loop region are only about 60% effective (Smith et al., 2000;
Wesley et al., 2001). Second, the DNA fragment selected from AtCle3 for the RNAi
construct should not only have high overall sequence identity, but also should have long
uninterrupted regions of sequence identity. Third, after identifying the major visible

phenotype for the transgenic plants containing the corresponding RNAi construct, the

246

first priority should be selecting some stable lines with a single transgene insert and put
all efforts of phenotype characterizations on those lines.

Another alternative for pursuing a reverse genetic approach to identify AtCle
functions would be to utilize the traditional knock out mutants. Stable knock out mutant
plants will make detailed cell wall polysaccharide and sugar composition analyses
possible. If gene redundancy becomes a concern, mutant plants with multiple AtCle
genes knocked out can be generated by crossing the single knock out mutants, because

the six AtCle genes are located on five different chromosomes (Figure 3.2).

 

247

REFERENCES

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

Bernstein, E., Candy, A. A., Hammond, S. M. and Hannon, G. J. (2001). "Role for a
bidentate ribonuclease in the initiation step of RNA interference." Nature 409: 363-366.

Billy, E., Brondani, V., Zhang, H., Muller, U. and Filipowicz, W. (2001). "Specific ?
interference with gene expression induced by long, double-stranded RNA in mouse
embryonal teratocarcinoma cell lines." Proc. Natl Acad. Sci. USA 98: 14428-14433.

Blakeney, A. 8., Harris, P. J., Henry, R. J. and Stone, 8. A. (1983). "A simple and
rapid preparation of alditol acetates for monosaccharide analysis." Carbohydr. Res. 113:
291-299.

 

Bosher, J. M. and Labouesse, M. (2000). "RNA interference: genetic wand and genetic
watchdog." Nat. Cell Biol. 2: 831-1336.

Bouton, S., Leboeuf, E., Mouille, G., Leydecker, M.-T., Talbotec, J., Granier, F.,
Lahaye, M., Hofte, H. and Truong, H.-N. (2002). "QUASIMODOI encodes a putative
membrane-bound glycosyltransferase required for normal pectin synthesis and cell

adhesion in Arabidopsis." Plant Cell 14: 2577-2590.

Burn, J., Hocart, C. H., Birch, R. J., Cork, A. C. and Williamson, R. E. (2002).
"Functional analysis of the cellulose synthase genes CesA], CesAZ, CesA3 in
Arabidopsis." Plant Physiol. 129: 797-807.

Chuang, C. H. and Meyerowitz, E. M. (2000). "Specific and heritable genetic
interference by double-stranded RNA in Arabidopsis thaliana." Proc. Natl Acad. Sci.
USA 97: 4985-4990.

Clough, S. J. and Bent, A. F. (1998). "Floral dip: a simplified method for
A grobacterium-mediated transformation of Arabidopsis thaliana." Plant J. 16 (6): 735-
743.

Delmer, D. P. and Stone, B. A. (1988). "Biosynthesis of plant cell walls." The
Biochemistry of Plants. .1. Priess, New York, Academic Press Inc. 14: 373-421.

Desprez, T., Vernhettes, S., Fagard, M., Refregier, G., Desnos, T., Aletti, E., Py, N.,
Pelletier, S. and Hofte, H. (2002). "Resistance against herbicide isoxaben and cellulose

deficiency caused by distinct mutations in same cellulose synthase isoform CESA6."
Plant Physiol. 128: 482-490.

248

Doblin, M. S., Melis, L. D., Newbigin, E., Bacic, A. and Read, 8. M. (2001). "Pollen
tubes of Nicotiana alata express two genes from different beta-glucan synthase families."
Plant Physio]. 125: 2040-2052.

Escobar, M. A., Civerolo, E. L., Summerfelt, K. R. and Dandekar, A. M. (2001).
"RNAi-mediated oncogene silencing confers resistance to crown gall tumorigenesis."
Proc. Natl Acad. Sci. USA 98 (23): 13437-13442.

Fagard, M., Desnos, T., Desprez, T., Goubet, F., Refregier, G., Mouille, G., McCann,
M., Rayon, C., Vernhettes, S. and Hofte, H. (2000). "PROC UST E1 encodes a cellulose
synthase required for normal cell elongation specifically in roots and dark-grown
hypocotyls of Arabidopsis." Plant Cell 12: 2409-2432.

Favery, B., Ryan, E., Foreman, J ., Linstead, P., Boudonck, K., Steer, M., Shaw, P.
and Dolan, L. (2001). "KOJAK encodes a cellulose synthase-like protein required for
root hair cell morphogenesis in Arabidopsis." Gene. Dev. 15: 79-89.

Goubet, F., Misrahi, A., Park, S. K., Zhang, Z., Twell, D. and Dupree, P. (2003).
"AtCSLA7, a cellulose synthase-like putative glycosyltransferase, is important for pollen
tube growth and embryogenesis in Arabidopsis." Plant Physiol. 131: 547-557.

Grishok, A., Pasquinelli, A. E., Conte, D., Li, N., Parrish, S., Ha, I., Baillie, D. L.,
Fire, A., Ruvkun, G. and Mello, C. C. (2001). "Genes and mechanisms related to RNA
interference regulate expression of the small temporal RNAs that control C. elegans

developmental timing." Cell 106 (23-34).

Hamilton, A., Voinnet, 0., Chappell, L. and Baulcombe, D. (2002). "Two classes of
short interfering RNA in RNA silencing." EMBO J. 21: 4671-4679.

Hammond, S. M., Bernstein, E., Beach, D. and Hannon, G. J. (2000). "An RNA-
directed nuclease mediates post-transcriptional gene silencing in Drosophila cells."
Nature 404: 293-296.

Hammond, S. M., Boettcher, S., Candy, A. A., Kobayashi, R. and Hannon, G. J.
(2001). "ArgonauteZ, a link between genetic and biochemical analyses of RNAi." Science
293: 1146-1150.

Han, Y. and Grierson, D. (2002). "Relationship between small antisense RNAs and

aberrant RNAs associated with sense transgene mediated gene silencing in tomato." Plant
J. 29 (4): 509-519.

Helliwell, C. A., Wesley, S. V., Wielopolska, A. J. and Waterhouse, P. M. (2002).

"High throughput vectors for efficient gene silencing in plants." F unct. Plant Biol. 29:
1217-1225.

249

 

Hutvagner, G. and Zamore, P. D. (2002). "A microRNA in a multiple-tumover RNAi
enzyme complex." Science 297: 2056-2060.

lwai, H., Masaoka, N., Ishii, T. and Satoh, S. (2002). "A pectin glucuronyltransferase
gene is essential for intercellular attachment in the plant meristem." Proc. Natl Acad. Sci.
USA 99 (25): 16319-16324.

Ketting, R. F., Fischer, S. E., Bernstein, E., Sijen, T., Hannon, G. J. and Plasterk, R.
H. (2001). "Dicer functions in RNA interference and in synthesis of small RNA involved
in developmental timing in C. elegans." Genes & Dev. 15: 2654-2659.

Klahre, U., Crete, P., Leuenberger, S. A., Iglesias, V. A. and Meins, F. J. (2002).
"High molecular weight RNAs and small interfering RNAs induce systemic

posttranscriptional gene silencing in plants." Proc. Natl Acad. Sci. USA 99(18): 11981-
1 1986.

Madson, M., Dunand, C., Li, X., Verma, R., Vanzin, G. F., Caplan, J., Shoue, D. A.,
Carpita, N. C. and Reiter, W.-D. (2003). "The MUR3 gene of Arabidopsis encodes a

xyloglucan galactosyltransferase that is evolutionarily related to animal exostosins."
Plant Cell 15: 1662-1670.

Martinez, J., Patkaniowska, A., Urlaub, H., Luhrmann, R. and Tuschl, T. (2002).
"Single-stranded antisense siRNAs guide target RNA cleavage in RNAi." Cell 110: 563.

Parrish, S., Fleenor, J., Xu, S., Mello, C. and Fire, A. (2000). "Functional anatomy of a

dsRNA trigger: differential requirement for the two trigger strands in RNA interference."
Mol. Cell 6: 1077-1087.

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 (5422): 1976-1979.

Reiter, W. D., Chapple, C. C. S. and Somerville, C. R. (1993). "Altered growth and
cell walls in a fucose-deficient mutant of Arabidopsis." Science 261: 1032-1035.

Richmond, T. A. and Somerville, C. R. (2001). "Integrative approaches to determining
Csl function." Plant Mol. Biol. 47: 131-143.

Scheible, W.-R., Eshed, R., Richmond, T. A., Delmer, D. and Somerville, C. R.
(2001). "Modifications of cellulose synthase confer resistance to isoxaben and
thiazolidinone herbicides in Arabidopsis Ixr1 mutants." Proc. Natl Acad. Sci. USA 98
( 18): 10079-10084.

Sharp, P. A. (2001). "RNA interference - 2001." Genes & Dev. 15: 485-490.

250

 

Smith, N. A., Singh, S. P., Wang, M. B., Stoutjesdijk, P. A., Green, A. G. and
Waterhouse, P. M. (2000). "Total silencing by intron-spliced hairpin RNAs." Nature
407: 319-320.

Stoutjesdijk, P. A., Singh, S. P., Liu, Q., Hurlstone, C. J., Waterhouse, P. M. and
Green, A. G. (2002). "hpRNA-mediated targeting of the Arabidopsis F AD2 gene gives
highly efficient and stable silencing." Plant Physio]. 129: 1723-1731.

Tang, G., Reinhart, B. J., Bartel, D. P. and Zamore, P. D. (2003). "A biochemical
framework for RNA silencing in plants." Genes & Dev. 17: 49-63.

Taylor, N. G., Laurie, S. and Truner, S. R. (2000). "Multiple cellulose synthase
catalytic subunits are required for cellulose synthesis in Arabidopsis." Plant Cell 12:

2529-2539.

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.

Vanzin, G. F., Madson, M., Carpita, N. C., Raikhel, N. V., Keegstra, K. and Reiter,
W. D. (2002). "The mur2 mutant of Arabidopsis thaliana lacks fucosylated xyloglucan
because of a lesion in fucosyltransferase AtFUTl." Proc. Nat] Acad. Sci. USA 99 (5):
3340-3345.

Voinnet, O. (2003). "RNA silencing bridging the gaps in wheat extracts." Trends Plant
Sci. 8 (7): 307-309.

Wang, X., Cnops, G., Vanderhaeghen, R., Block, S. D., Montagu, M. V. and
Lijsebettens, M. V. (2001). "AtCSLD3, a cellulose synthase-like gene important for root
hair growth in Arabidopsis." Plant Physiol. 126: 575-586.

Waterhouse, P. M. and Helliwell, C. A. (2003). "Exploring plant genomes by RNA-
induced gene silencing." Nat. Rev. Genet. 4: 29-38.

Wesley, S. V., Helliwell, C. A., Smith, N. A., Wang, M. B., Rouse, D. T., Liu, Q.,
Gooding, P. S., Singh, S. P., Abbott, D., Stoutjesdijk, P. A., Robinson, S. P., Gleave,
A. P., Green, A. G. and Waterhouse, P. M. (2001). "Construct design for efficient,
effective and high throughput gene silencing ini plants." Plant J. 27 (6): 581-590.

Zhong, R., Morrison Ill, W. H., Freshour, G. D., Hahn, M. G. and Ye, Z.-H. (2003).
"Expression of a mutant form of cellulose synthase AtCesA7 causes dominant negative
effect on cellulose biosynthesis." Plant Physio]. 132: 786-795.

Zhu, Y., Nam, J., Humara, J. M., Mysore, K. S., Lee, L. Y., Cao, H., Valentine, L.,

Li, J., Kaiser, A. D., Kopecky, A. L., Hwang, H. H., Bhattacharjee, S., Rao, P. K.,
Tzfira, T., Rajagopal, J., Yi, H., Veena, Yadav, B. S., Crane, Y. M., Lin, K., Larcher,

251

 

Y., Gelvin, M. J., Knue, M., Ramos, C., Zhao, X., Davis, S. J., Kim, S. I., Ranjith-
Kumar, C. T., Choi, Y. J., Hallan, V. K., Chattopadhyay, S., Sui, X., Ziemienowicz,
A., Matthysse, A. G., Citovsky, V., Hohn, B. and Gelvin, S. B. (2003). "Identification
of Arabidopsis rat mutants." Plant Physiol. 132 (2): 494-505.

 

252