5mm? . .
a. » . (if m

, . . T§,m@...,ufi.
. , x . . gmwivw a»
.55? , .

at!!! . ‘ V

JEWUJE

..

j. {.172
3 :
{

, , . . y.»
c l
. , . .17....»
‘93, . k? 2%;
A i .. 3...... Fur V ‘ 3 $2 ..
I .hSrHV: .V .3."
:5 Q‘Pvma: . ‘ . . , .
. it; , ‘ .
f. . . , .
. .. A .
, . , . c :.
13.. Rafi.” . .. Qhfia:
g . 3.; .
haw. ...
3:9»; 4

~..

.r..

.4
.

afiflwfim

_ 9%..
JW. {can fig . x ..
. ‘ cu .4 a ., V ,
mm". , V? c. “Wu-”mewfik :1
. “gum...“nhé

a...
.1

av;
...

x\0.
I... .
. 1.9

 

 

 

 

 

 

 

 

 

.5 (5.1 \ .
. u

 

131$?

{at ,2
4’1). I...»

333%. ,1? §§§§1

fimflgfié

THESIS

ZOOI

This is to certify that the

dissertation entitled

ROLE OF THE ARABIDOPSIS CBF FAMILY OF TRANSCRIPTION
FACTORS IN PLANT COLD ACCLIMATION

bresented by

Kirsten Ruth J aglo

has been accepted towards fulfillment
of the requirements for

PhD. degree in Plant Breeding & Genetics-
Crop & Soil Sciences

Major professor

 

 

Date December 44000

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

 

 

LIBRARY
Michigan State
Univeraity

 

 

 

PLACE IN REFURN 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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11/00 cJCIHCJDateDtnpfis-p.“

 

ROLE OF THE ARABIDOPSIS CBF FAMILY OT TRANSCRIPTION FACTORS IN
PLANT COLD ACCLIMATION
By

Kirsten Ruth Jaglo

A DISSERTATION

Submitted to
Michigan State University
in partial fiilfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Crop and Soil Science
Program of Plant Breeding and Genetics

2000

ABSTRACT
ROLE OF THE ARABIDOPSIS CBF FAMILY OF TRANSCRIPTION FACTORS 1N
PLANT COLD ACCLIMATION
By

Kirsten Ruth J aglo

Many plants, including Arabidopsis and canola, increase in freezing tolerance
after exposure to low, nonfreezing temperatures, a process called cold acclimation.
Numerous physiological and biochemical changes are associated with cold acclimation
including changes in gene expression. The COR (cold-regulated) genes are associated
with cold acclimation, and their expression is greatly increased under acclimating
conditions due to multiple copies of a cis-acting DNA regulatory element called the
CRT/DRE. A transcription factor called CBF] (CRT/DRE-binding factor one) was
isolated which can bind to the CRT/DRE sequence and activate transcription in yeast.
Overexpression of CBF] in Arabidopsis resulted in increased COR gene expression and
increased freezing tolerance without a low temperature stimulus. Further research has
shown that CBF] is a member of a small gene family that also includes CBFZ and CBF 3.
RNA accumulation data indicated that all three CBF genes are activated under
acclimating conditions. However, despite repeated attempts using both immunoblot
analysis and irnmunoprecipitation, CBF protein accumulation could not be detected in
wild type or CBF-overexpressing plants. Sequence analysis indicated that a putative CBF
homologue, BnCBF, exists in Brassica napus (canola), a close relative of Arabidopsis. A

time course of BnCBF RNA accumulation showed a similar induction pattern to that seen

with the Arabidopsis CBF genes. Overexpression of the Arabidopsis CBF genes in B.
napus var. Westar, a spring variety of canola, showed increased BN gene (COR gene
homologue) RNA accumulation and increased freezing tolerance under both
nonacclimating and acclimating conditions. Additionally, increases in total soluble sugars
were seen under nonacclimating and acclimating conditions, and increases in free proline
under were seen under acclimating conditions. In summary, the CBF family of
transcription factors appears to play important roles during cold acclimation in

Arabidopsis and its close relative, canola.

To my grandpa:

William George Jaglo

who understood the language of flowers,
and
never
stopped learning.

Acknowledgements

“What lies behind us and what lies before us are tiny matters compared to what
lies within us”

Ralph Waldo Emerson.

I want to thank my advisor, Mike “Ice Man” Thomashow for being an amazing
advisor, for having incredible enthusiasm even when the data were less than encouraging,
and for pushing me to accomplish and learn more than I ever thought (and sometimes
ever wanted to be) possible. I hope that in the end I made up for the fact that I’m a “pain
in the butt”. I would like to thank my committee members Rebecca Grumet, Jim Kelly
and Steven Triezenberg for their advice and constructive comments along way, and for
not permanently scarring me too noticeably during my preliminary exams. A big warm
thank you to all of the current and past members of the Thomashow and Grumet labs for
their support, encouragement, love and absolute insanity. In particular, Sarah Gilmour
who has been an amazing bench partner (although she got more shelf space for HER
solutions. . .), an incredible friend and someone to talk to for all types of advice, Katerina
Papadopoulou who has never allowed me to get away with anything of any sort and who
keeps the Greek mafia alive in Lansing, Sue Hammer and Ann Gustafson who taught me
the joys of race walking and gossiping while working in the hood, Kevin O’Connell and
Eric Stockinger who helped me to learn the basics when I first got to the lab. Thanks to
Susanne Klefi‘ for being a great collaborator and friend and allowing me to join an
amazing project already in progress. A big thanks to the exercising crew, Sue, Katerina,

V

Sarah, Holly and Ann who all helped to make sure that I kept my body moving beyond
running around the lab. Thank you to Art and Marlene for having such great parties, an
incredible house, and being amazing friends with whom I learned the joys of dying

fabrics and creating all sorts of new and fun art. Thank you to my many and various
friends here in Lansing who made life inside and outside the lab more enjoyable: Marty,
Maite, Charlie, Keenan, Kostas, Sam, Judy, Chris, Sarah Chicken, Zakir, Carri, Canadian
Kirsten, Huanying, Wang, Esther, Dawn, Diane, Suzanne, Ann, Deane, Audrey “the other
blonde”, Maria, Pete and Sandi. Thank you to my far and distant friends and family,
Megamunchkin Elliott, Mark “Thamious” Grodzki, Kristie Hirschenberger, mom, dad
and Jas, who could not always literally hold my hand, but were always a telephone call
away and helped to hold me up more times than I can remember. A sad farewell and
thank you to Saren Ottosen who helped me make it to graduate school, but who in the
end was not right for me, so now our paths diverge. Tak for sidst. Lastly, thanks to Phillip
Wharton who came seemingly from nowhere like the miracle of television and has helped
me to revise the way I view the world and myself. Without the support of all of you, I
would not be where I am today, and I will carry you all within me for the rest of my

years.

vi

PREFACE

In Chapter 2, all experiments were conducted by the author of this thesis, except
for the RNA accumulation analyses in Figure 2.2A which were conducted by Daniel
Zarka, and the immunoblot analysis in Figure 2.2B which was conducted by Sarah
Gilmour. Most of the results from this chapter were published in Science. 280: 104-106.

In Chapter 3, all experiments, including immunoblot analysis and
irnrnunoprecipitation experiments were conducted by the author of this thesis. The
constructs encoding for the various portions of GST-labeled CBFl were made by Eric
Stockinger. For use in immunoblot analysis experiments, protein extracts from CBF]-
overexpressing and vector control yeast were kindly donated by Eric Stockinger. For use
in immunoblot analysis and immunoprecipitation experiments, proteins isolated fiom
Arabidopsis nuclei were kindly donated by Charlie Herman, and proteins isolated from
Arabidopsis protoplasts were kindly donated by Yaopan Mao. The transgenic CBF-
overexpressing Arabidopsis lines, G7a—l, E71-1, A30a-l, A38b-7 and the vector control
line B16-l were made by Maite Salazar and Audrey Sebolt.

In Chapter 4, all transgenic canola plants used in experiments were generated by
Susanne Klefl‘. All experiments, including RNA accumulation, immunoblot analysis,
proline analysis, total soluble sugar analysis and electrolyte leakage analysis were
conducted by the author of this thesis. In the salt stress experiments, the first experiment
was conducted by Susanne Kleff, the second experiment was conducted in collaboration

with Susanne Klefi‘ and the third experiment was conducted by the author of this thesis.

vii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. xiii
LIST OF FIGURES ................................................................................. xiv
PREFACE .................................................................................................................. vii
TABLE OF CONTENTS ....................................................................................... viii

1. CHAPTER 1: Effects of Chilling and Freezing Temperatures on Plants and Changes

Associated with Cold Acclimation .................................................................................. l
1.1 INTRODUCTION ............................................................................................. 1
1.2 EFFECTS OF CHILLING TEMPERATURES ON PLANTS ............................ 3
1.3 EFFECTS OF FREEZING TEMPERATURES ON PLANTS ........................... 4
1.4 COLD ACCLIMATION .................................................................................... 6
1.4.1 INTRODUCTION ................................................................................................ 6
1.4.2 INVOLVEMENT OF ABSCISIC ACID (ABA) ............................................................. 6
1.4.3 EFFECTS ON CHLOROPIASTS ............................................................................. 8
1.4.4 CHANGES IN MEMBRANES ................................................................................. 9
1.4.5 CHANGES IN SMALL MOLECULES ...................................................................... 10
1.4.6 CHANGES IN GENE EXPRESSION ....................................................................... 11
1.5 REFERENCES ................................................................................................ 19

2. CHAPTER 2: Overexpression of Arabidopsis CBF] results in increased COR gene

expression and freezing tolerance without a low temperature stimulus ........................... 25

viii

2. 1 INTRODUCTION ........................................................................................... 25

2.2 MATERIALS AND METHODS ..................................................................... 29
2.2.1 PLANTGROWTH ............................................................................................. 29
2. 2. 2 PLANT TRANSFORMATION ................................................................................ 30
2. 2. 3 DNA HYBRIDIZA TION ..................................................................................... 31
2.2.4 RNA HYBRIDIZA TION ...................................................................................... 31
2. 2.5 IMMUNOBLOTANALYSIS .................................................................................. 32
2. 2. 6 ELECTROLYTE LE4KAGEASS4 YS ....................................................................... 33
2. 2. 7 WHOLE PLANT FREEZE TESTS .......................................................................... 34
2. 2.8 SALT STRESS GERMINA TION TESTS .................................................................... 34

2.3 RESULTS ....................................................................................................... 34
2.3.1 ISOLATION OF CBFI-OVEREXPRESSING ARABIDOPSIS PLANTS ............................ 35

2. 3. 2 OVEREXPRESSION 0F CBF] IN ARABIDOPSIS RESULTS IN INCREASED COR GENE
EXPRESSION .............................................................................................................. 35
2. 3. 3 OVEREXPRESSION 0F CBF] RESULTS IN INCREASED FREEZING TOLERANCE AS
DETERMNED BY ELECTROLYTE LEAKAGE ANALYSIS ....................................................... 39
2. 3. 4 OVEREXPRESSION 0F CBFI RESULTS IN INCREASED FREEZING TOLERANCE AS SEEN
BY WHOLE PLANT FREEZE TESTS, BUT NO GROSS PHENO TYPIC CHANGES .......................... 41
2. 3. 5 OVEREXPRESSION 0F CBF] DOES NOT RESULT IN INCREASED GERMINA TION UNDER
OSMOTIC STRESS ........................................................................................................ 43
2. 3. 6 OVEREXPRESSION OF CBF] RESULTS IN TRANSGENE SILENCING IN ALL LINES TESTED
45

2. 3. 7 ANTISENSE CBF! PLANTS DO NOT HA VE REDUCED CBF] OR COR GENE

ix

EXPRESSION .............................................................................................................. 52
2.4 DISCUSSION ................................................................................................. 56
2.5 REFERENCES ................................................................................................ 68

3. CHAPTER 3: Detection of Arabidopsis CBF proteins by immunoblot analysis and

immunoprecipitation ..................................................................................................... 73

3.1 INTRODUCTION ........................................................................................... 73

3.2 MATERIALS AND METHODS: .................................................................... 76
3. 2. I PLANTMATERIAL ........................................................................................... 76
3.2.2 PLANT GROWTH ............................................................................................. 77

3. 2. 3 ISOLATION OF RECOMBINANT CBF] PEPTIDES FROM E. COLI EXTRACTS AND

YEAST ...................................................................................................................... 78
3. 2. 4 SYMHETIC PEPTIDE PRODUCTION AND MANIPULA TION ..................................... 80
3. 2. 5 ANTIBODIES USED .......................................................................................... 81
3. 2. 6 PURIFICATION OFANTT-CBFI SERA ................................................................ 82
3. 2. 7 EXTRACTION 0F PROTEINS FROM PLANTS ......................................................... 84
3. 2. 8 IMMUNOELOTANALYSIS .................................................................................. 86
3. 2.9 MJUNOPRECIPITA TIONS 0F CBF] PROTEINS .................................................. 87

3. 2. 1 0 ISOLATION TOTAL SOL UBLE PROTEINS FR OM 35S-METHIONINE RADIOLABLLED E. COLI
FOR WUNOPRECIPITAIYONS ..................................................................................... 89
3. 2. 1 1 IMMUNOPRECIPITA TIONS WITH 35S—METHIONINE LABELLED PROTEINS FROM
ARABIDOPSIS PLANTS ................................................................................................. 90
3. 2. 12 IMMUNOPRECIPITA TIONS WITH 3 2P-0R THOPHOSPHA TE LABELLED PROTEINS FROM

ARABIDOPSIS PLANTS ................................................................................................. 91

3 .3 RESULTS: ...................................................................................................... 92
3.3.1 RECOMBINANT CBF] IS DEGRADED IN THE PRESENCE OF ARABIDOPSIS PROTEIN
EXTRACTS ................................................................................................................. 92
3. 3. 2 CBF PROTEINSARE NOT DETECTED IN TOTAL SOLUBLE PLANT OR CBF]-
OVEREXPRESSING YEAST PROTEIN EXTRACTS ................................................................. 93
3. 3. 3 PRE-IMMUNE SER UM CONTAINS ANTIBODIES TO NUMEROUS PROTEINS CONTAINED IN
WHOLE PLANTEXIRACTS ............................................................................................ 94
3. 3. 4 MIUNOAFFINITY PURIFICA TION OF ANTI-CBF] ANTISERA DOES NOT ENHANCE
DETECTION OFRECOMBINANT OR NATIVE CBF PROTEINS .............................................. 96
3. 3. 5 ANTIBODY PRODUCTION TO SPECIFIC PEPTIDES FROM THE CBF] PROTEIN ....... 100
3. 3. 6 ALL TESTED PRE-LWUNE SERA FROM RABBITS HA VE ANTIBODIES TO NON-SPECIFIC
PLANT PROTEINS ...................................................................................................... 101
3. 3. 7 ANTISERA RAISED TO SYNTHETIC CBF PEPTIDES RECOGNIZE RECOMBINANT CBF]
PROTEIN, BUT CBF PROTEINSARE NOT DETECTED INPLANT EXTRACTS ........................ 102
3. 3. 8 MUNOAFFINITY PURIFICA TION 0F ANTI- PEPTIDE ANTIBODIES RED UCES
BA CKGRO UND, B UT CBF PROTEINS ARE NOT DETECTED IN PLANT EXTRACTS ................ 1 02
3. 3. 9 RECOMBINANT CBF] MAY BE DETECTED B Y LIWLWOPRECIPITA TION, BUT NOT IN
THE PRESENCE OF PLANT EXTRACTS .......................................................................... 105
3. 3. 1 0 CBF PROTEINS ARE NOT DETECTED BY IMIUNOPRECIPITA TION FROM PLANT
EXTRACTS ............................................................................................................... 108
3. 3. 1 1 3 5S-ME7H10NINE LABELLED RECOMBINANT CBF] CAN BE DETECTED BY
IMMUNOPRECIPITA TION ........................................................................................... 1 10

3. 3. 12 35S-MEIHIONINE LABELLED REC 0MBINA/VT CBF I IS DEGRADED IN THE PRESENCE OF

xi

PLANT EXTRAC T S ..................................................................................................... 113
3.3. 13 355-METHIONINE LABELLED CBF] IS NOT DETECTED BY IIWUNOPRECIPITA T ION OF
PLANT EXTRAC TS ..................................................................................................... I 15

3. 3. I4 CBF] IS NOT DETECTED BY I.MM UNOPRECIPI T AT ION 0F 32P-0RTHOPHOSPHAT E

LABELLED PLANT EXTRACTS ...................................................................................... 11 7
3.4 DISCUSSION ............................................................................................... 1 19
3.5 REFERENCES .............................................................................................. 125

4. CHAPTER 4: Overexpression of Arabidopsis CBF], CBF2 or CBF3 in Brassica

napus var. Westar results in increased BN gene expression and freezing tolerance ...... 129
4.1 INTRODUCTION ......................................................................................... 129
4.2 MATERIALS AND METHODS ................................................................... 132
4.2.1 PLANT GROWTH: .......................................................................................... 133
4. 2. 2 TRANSFORMATION: ...................................................................................... 134
4. 2. 3 IRANSGENIC PLANT SELECTION: .................................................................... 134
4.2.4 CUTIINGS ................................................................................................... 135
4.2.5 SEEDLINGS .................................................................................................. 136
4. 2. 6 RNA HYBRIDIZA TION: .................................................................................. 136
4. 2. 7 IWUNOBLOT ANALYSIS ................................................................................ 138
4. 2. 8 PROLINE ANALYSIS ....................................................................................... 139
4. 2. 9 TOTAL SOLUBLE SUGAR ANALYSIS .................................................................. 139
4.2.10 ELECTROLYTE LEAKAGE ASSA YS ..................................................................... 140
4.2.11 SALT STRESS EXPERIMENTS ............................................................................ 140
4.3 RESULTS ..................................................................................................... 141

xii

4.3.1 THE B. NAP US CBF GENE HAS A SIMILAR INDUCTION PA TTERN TO THE ARABIDOPSIS
CBF GENES ............................................................................................................ 141
4. 3. 2 GENERA TION OF CANOLA PLANTS THA T CONSTIUTIVEL Y EXPRESS ARABIDOPSIS
CBF], CBF2 OR CBF3 ........................................................................................... 143
4. 3. 3 OVEREXPRESSION OF THE ARABIDOPSIS CBF GENES 1N CANOLA RESUTS IN
INCREASED BN28 AND BN115 TRANSCRIPTACCWULATION ....................................... 145
4. 3. 4 OVEREXPRESSION OF THE ARABIDOPSIS CBF GENES IN CANOLA RESUTS IN
INCREASED BN28 PROTEINACCDMULA TION .............................................................. 149
4. 3. 5 OVEREXPRESSION OF THE ARABIDOPSIS CBF GENES IN CANOLA RESULTS IN
INCREASED ACC UMULA TION OF PROLINE AND TOTAL SOL UBLE SUGARS ......................... 15 1
4. 3. 6 OVEREXPRESSION OF ARABIDOPSIS CBF GENES 1N CANOLA RESULTS IN INCREASED
FREEZING TOLERANCE ............................................................................................. 15 4
4. 3. 7 OSMO TIC STRESS TOLERANCE 0F CBF], CBF2 OR CBF3 OVEREXPRESSING CANOLA
LINES ..................................................................................................................... 160

4. 3. 8 OVEREXPRESSION OF CBF], CBF2 OR CBF3 IN CANOLA DOES NOT CA USE GROSS

PHENOTYPIC CHANGES ............................................................................................. 162
4.4 DISCUSSION ............................................................................................... 162
4.5 REFERENCES .............................................................................................. 169
5. Conclusions ......................................................................................................... 1 74

xiii

LIST OF TABLES

Table 2.1. Comparison of Electrolyte leakage 50% (ELSO) values of leaves from wild type
and transgenic Arabidopsis plants. ......................................................................... 42
Table 2.2. Germination of CBFI-overexpressing and wild type plants on NaCl ............. 46
Table 4.1. Accumulation of total soluble sugars and proline in combined CBF], CBF 2
and CBF3-overexpressing canola plants .............................................................. 153
Table 4.2. Comparison of EL50 values of combined CBF], CBF 2 or CBF 3
overexpressing and control plants ........................................................................ 159
Table 4.3. Survival of CBF-overexpressing and control cuttings and plants after salt

stress. .................................................................................................................. 161

xiv

 

LIST OF FIGURES

Figure 2.1 RNA accumulation and southern hybridization of CBFI-overexpressing lines
A6 andB16.......... ................................................................................................. 36

Figure 2.2. Expression of CBF] and COR genes in wild type and transgenic Arabidopsis
plants ..................................................................................................................... 38

Figure 2.3. Electrolyte leakage analysis of wild type and transgenic Arabidopsis plants.
.............................................................................................................................. 40

Figure 2.4. Freezing survival of RLD and CBFI-overexpressing A6 Arabidopsis plants.

.............................................................................................................................. 44
Figure 2.5. Electrolyte leakage analysis and RNA accumulation of wild type and B16

plants. .................................................................................................................... 47
Figure 2.6. Reduced COR15m accumulation is associated with reduced freezing

tolerance, but not the loss of the transgene in A6 T5 Lines ..................................... 48
Figure 2.7. CBF] and COR15 RNA accumulation in CBFl-overexpressing plants ........ 53
Figure 2.8. Electrolyte leakage analysis of wild type and transgenic K16 plants. ........... 54
Figure 2.9. CBF] and COR gene accumulation in antisense-CBFI transgenic lines ....... 55
Figure 3.1. Immunoblot analysis of recombinant and native CBFl ................................ 95
Figure 3.2. Irnmunoblot analysis of recombinant and native CBF 1 ................................ 97
Figure 3.3. . Immunoblot analysis of recombinant CBF 1 peptides ................................. 98
Figure 3.4. Immunoblot analysis of CBF and COR] 5m proteins .................................... 99

XV

Figure 3.5. Irnmunoblot analysis of recombinant and native CBF 1 .............................. 103
Figure 3.6. Irnmunoprecipitation of recombinant CBF 1 protein followed by immunoblot
analysis ............................................................................................................... 107
Figure 3.7. Immunoprecipitation of CBF and COR15 proteins from plant extracts. ..... 109
Figure 3.8. Immunoprecipitation of recombinant 35S-Methionine labelled CBF] from
total E. coli protein extracts. ................................................................................ 112
Figure 3.9. 3SS-methionine labelled recombinant CBF] is degraded in the presence of
plants and protoplast extracts. .............................................................................. 114
Figure 3.10. Immunoprecipitation of 35S-methionine labelled CBF proteins from plant
extracts ................................................................................................................ 116
Figure 3.11. Immunoprecipitation of 32P-orthophosphate labelled CBF proteins from
plant extracts ....................................................................................................... 118
Figure 4.1. Amino acid sequence alignment of Arabidopsis CBF2 and BnCBF fi'om
Brassica napus .................................................................................................... 142
Figure 4.2. Time course of BNCBF and BN115 RNA accumulation ............................ 144
Figure 4.3. Accumulation of Arabidopsis CBF and BN RNA in individual canola cuttings
............................................................................................................................ 146
Figure 4.4. Accumulation of Arabidopsis CBF and EN RNA in pooled canola plants .. 148
Figure 4.5. Accumulation of BN28 protein in individual CBF-overexpressing canola
cuttings and plants ............................................................................................... 150
Figure 4.6. Accumulation of total soluble sugars and proline in control and CBF], CBF 2

or CBF3-overexpressing canola plants ................................................................. 152

Figure 4.7. Electrolyte leakage analysis of wild type and transgenic canola plants and
cuttings ................................................................................................................ 155

Figure 4.8. Photograph of transgenic CBF-overexpressing and control canola plants... 163

xvii

1. CHAPTER 1: Effects of Chilling and Freezing Temperatures on

Plants and Changes Associated with Cold Acclimation

1.1 INTRODUCTION

Environmental stress consistently causes a decrease in maximum crop yield
worldwide. It is estimated that 60% of potential yield is lost annually due to adverse
environmental conditions (Levitt, 1980). One category of adverse environmental
conditions is low and freezing temperatures. Depending on the region of origin, different
species of plants show differing abilities to withstand chilling and freezing temperatures.
For example, plants originating from tropical and sub-tropical climates that have not
evolved the ability to withstand low temperatures, become damaged and may die from
chilling temperatures of 10°C or less (Kratsch and Wise, 2000). In contrast, plants fi'om
temperate regions have evolved the ability to withstand low and even sub-zero
temperatures by cold acclimating, a process by which plants increase in freezing
tolerance after being exposed to low, non-freezing temperatures (Thomashow, 1999).
This difference in the ability to survive low and freezing temperatures has a major impact
on where crops are planted, when crops are planted, and the specific type of crop that is
planted in a given geographical location.

At present, enough food is produced to feed the world (although not everyone is

adequately fed due to complex problems with distribution) (Engelman and LeRoy, 2000).

However, there is reason to be concerned that there may not be enough food in the near
future. The world population, currently at 6 billion, is predicted to reach 8.9 billion by
2050 (Engelman and LeRoy, 2000). Additionally, the amount of arable land per person is
expected to decrease fiom 0.44 hectares to 0.16 - 0.18 hectares by 2025 (Engelman and
LeRoy, 2000). To produce enough food, there will need to be an overall increase in crop
yield and/or an expansion of agricultural production areas. One way to increase both is to
improve the ability of plants to withstand freezing temperatures, both before and afier
cold acclimation. This would increase yield by reducing cold-induced damage, such as
that experienced after a sudden frost, and by allowing higher yielding varieties to be
planted in areas that are currently unsuitable due to inhibitingly low temperatures. For
example, winter varieties of canola (Brassica napus and Brassica rapa) have 40% higher
yield than spring varieties, making them more desirable to grow. However, only spring
varieties are planted in some geographic locations as winter varieties cannot survive the
extreme low temperatures experienced during the winter months (http: //www.canola—
council.orgl). Agricultural expansion could also occur as more fi'eezing tolerant plants
could be sown in areas that are not currently suitable due to temperature restraints.
Although improving freezing tolerance has been a long-term goal of plant
breeders, the most cold tolerant varieties today are only marginally better than those
produced at the turn of the century (Thomashow, 1990). Increasing our understanding of
cold acclimation by focusing on the molecular changes that occur during acclimation may

lead to better strategies in improving freezing tolerance.

1.2 EFFECTS OF CHILLING TEMPERATURES ON PLANTS

Damage due to low temperatures can be separated into two general categories,
chilling damage and freezing damage. Chilling damage occurs in plants from tropical and
sub-tropical regions, at temperatures below ~10° C (Kratsch and Wise, 2000). The extent
to which plants are damaged is highly species specific and depends on the level of
irradiance, the level to which the plant is hydrated at the time of the stress and the
duration of the temperature stress. Despite these variables, there are general trends
observed in chilling sensitive species. The chloroplast is the first organelle in which
ultrastructural chilling damage can be observed. Damage includes swelling of the
thylakoids, reduction in size and number of starch granules, unstacking of the grana, and
depending on the duration of the stress, the eventual disintegration of the chloroplast
(Kratsch and Wise, 2000). As chloroplasts are essential to plant life, even moderate
damage can be lethal.

The molecular basis behind the chilling-induced damage to chloroplasts is
complex. However, the ultimate result is the inhibition of photosynthetic carbon fixation
(Strand et al, 1999), which can result in the death of the plant. In contrast to chilling
tolerant plants, chilling sensitive plants do not recover their photosynthetic capacities
afier a chilling stress (Klimov et al, 1997). At low temperatures, chilling sensitive species
show a reduction in photosynthetic gene expression. This inhibits the ability of a plant to
harvest the light energy entering the chloroplasts and causes production of free oxygen
radicals that damage the plant. Changes in membrane fluidity during low temperature

stress also result in increased damage to chloroplasts (Strand et al, 1999) as the efficiency

3

of electron transport in photosystem II is reduced, resulting in the increased production of

damaging fi'ee oxygen radicals.
1.3 EFFECTS OF FREEZING TEMPERATURES ON PLANTS

Freezing stress can result in protein denaturation, the production of fiee radicals,
and damage to cellular membranes. A primary site of freezing damage in plants is the
cellular membrane system (Steponkus, 1984). As the temperature drops below 0° C, ice
formation occurs in the intercellular spaces as these contain fewer solutes than
intracellular fluid and therefore have a higher freezing point. The chemical potential of
ice is less than that of liquid water, so the intracellular osmotically active water moves
down the water potential gradient toward the lower potential in the intercellular spaces
and leaves the cell until chemical equilibrium is reached (Thomashow, 1999). Plants
frozen at -—10°C generally lose more than 90% of their osmotically active water
(Steponkus and Lynch, 1989).

The damage that occurs to membranes during freezing is due to freeze-induced
dehydration, which mimics damage induced by dehydration and osmotic stress
(Steponkus, 1984; Steponkus et al, 1993b). The primary manifestation of damage caused
by freeze-induced dehydration stress is dependent on the absolute temperature reached,
but includes expansion-induced lysis, lamellar-to hexagonal-II phase transitions and
fi'acture jump lesions (Steponkus et al, 1993 a; Uemura and Steponkus, 1997).

The most common type of freeze-induced damage that occurs in nonacclimated

plants is the formation of hexagonal-II phase transitions. These occur when the plasma

4

 

membrane is brought into close apposition to intracellular organelle membranes during
freeze-induced dehydration. As the membranes are in close proximity to each other and
water and other polar molecules are in limited supply, it is no longer energetically
favorable for the nonpolar fatty acid portion of the membrane to remain in a lipid bilayer.
The lipids can become reoriented and combine with lipids in other membranes into long
cylinders with their polar head groups in an aqueous pore (Steponkus, 1984; Uemura et
al, 1995). When water returns to the cell after the plant is returned to non-freezing
temperatures, the lipid membranes do not reform into independent lipid bilayers and
depending on the amount of membrane rearrangement that has occurred, severe damage
can occur.

Expansion induced lysis can also occur with return of water to the cell.
(Steponkus, 1984). This occurs as endocytotic vesicles bud off from the plasma
membrane during fi'eeze-induced dehydration as the cells shrink in size. After thawing,
the vesicles are not reincorporated into the membrane resulting in the inability of the cells
to accommodate the influx of water (Steponkus and Lynch, 1984).

Free oxygen radicals may also be produced during freezing stress. The radicals
are the result of the reduced ability of chloroplasts to harvest the light energy entering the
cell, due, in part to changes in membrane fluidity during freezing temperatures (McKersie
and Bowley, 1997; Smirnoff, 1998). If many free oxygen radicals are produced, the plant
can experience high levels of damage. Additional damage may result from protein
denaturation or aggregation due to temperature induced conformational changes
(Smirnoff, 1998). Depending on the protein affected, an important enzymatic function

may be lost or changed and damage may ensue.

5

1.4 COLD ACCLIMATION

1.4.1 INTRODUCTION

Cold acclimation is the process by which plants increase in freezing tolerance
after exposure to low non-freezing temperatures. Cold acclimation involves the activation
of a series of physiological and biochemical changes that reduce the amount of damage to
plants caused by both chilling and freezing temperatures. It is associated with changes in
sugar accumulation, lipid composition, isozyrne patterns, total protein content (W eiser,
1970) and changes in protein phosphorylation (Monroy et al, 1993). It is a multi-genic
trait that allows freezing-tolerant plants to survive temperatures that are lethal to the same
plant in the nonacclimated state. However, exactly which of these changes are the result
of a metabolic response to the decrease in temperature, which are induced as a function of
cold acclimation, and how such changes are activated has remained unclear. In the
following sections, I will focus on some of the key changes associated with cold

acclimation that relate to this thesis.

1.4.2 INVOL VEMENT OF ABSCISIC ACID (ABA)

ABA has long been considered a stress hormone. When plants are exposed to cold
or other types of osmotic stress conditions, there is a burst of ABA production

(Thomashow, 1999). Exogenous application of ABA to unstressed plants can result in

6

biochemical changes similar to those seen under conditions of cold/osmotic stress (Hare
et al, 1999). Ifnonstressed plants are first treated with ABA and then exposed to osmotic
stress, such as freezing temperatures, drought or salt stress, they frequently have
increased tolerance to these stresses as compared to untreated nonstressed plants (Hare et
al, 1999). These data suggest that ABA production and signaling are involved in cold
acclimation as well as in acclimation to other osmotic stresses. However, even under
conditions of continual stress, ABA levels eventually decline to those seen in unstressed
plants (Lang and Palva, 1992), whereas freezing tolerance continues to increase for up to
a week (Gilmour et al, 1988).

To investigate the role of ABA in cold acclimation, experiments were conducted
in plants with mutations in either ABA sensing (abil) or in ABA production (abaI)
(Koornneef et al, 1998). While plants with either mutation are more sensitive to freezing
than wild type plants, both types of plants also have compromised health overall, making
it diflicult to determine if the reduction in fi'eezing tolerance is a primary or secondary
effect of aberrant ABA signaling (Gilmour and Thomashow, 1991). Another way to
determine if ABA plays a role in cold acclimation is to investigate induction patterns of
known cold induced genes in the two mutant backgrounds. If cold induced genes have
typical induction patterns under acclimating conditions in the mutant background, then
ABA is clearly not necessary for cold-induced activation of these genes. In fact, this
appears to be the case. The cold induced activation of the COR (cold-regulated) genes
(discussed in detail in section 1.4.6.2) was not impaired in abil or abal plants, whereas
the ABA-induced activation of these genes was lost (Gilmour and Thomashow, 1991).

These data clearly indicate that there are two possible induction pathways of the COR-

7

genes, an ABA-dependent and an ABA-independent pathway. Investigation of other
genes involved in stress induced pathways, such as those involved with the accumulation
of free proline, also show both ABA-dependent and ABA-independent induction pattems

(Hare et al, 1999).

1. 4. 3 EFFECTS ON CHLOROPLASTS

When plants are first shifted to low temperatures, inhibition of photosynthetic
activity occurs in chloroplasts which can result in free oxygen radical production and
damage to the plant (Klimov et al, 1997). Therefore, some of the changes associated with
cold acclimation fimction to protect chloroplasts and prevent permanent inhibition of
photosynthetic carbon fixation (Klimov et al, 1997; Strand et al, 1997; Strand et al,
1999). For example, under acclimating conditions, chloroplasts increase in size (Klimov
et al, 1997) and changes occur in the lipid composition of the inner and outer chloroplast
envelope which reduce the occurrence of hexagonal-H formation and fracture-jump
lesions in rye leaves (Uemura and Steponkus, 1997). To preserve photosynthesis, an
increase in activity in seven of the enzymes associated with the Calvin cycle occurs
(Strand et al. 1999). When Arabidopsis plants are grown under the acclimating condition
of 5° C, a five-fold increase in carbon fixation rates occurs and large pools of soluble
sugars are present in the chloroplast without suppression of photosynthetic gene
expression or metabolism (Strand et al, 1997; Strand et al. 1999).

Cold acclimation also appears to reduce damage to chloroplasts by limiting the

production of free oxygen radicals. In potato, superoxide dismutase isoenzymes are

8

activated under acclimating conditions, presumably to reduce free oxygen radicals
created by the initial inhibition of photosynthesis (Seppanen and Fagerstedt, 2000). Free
radicals are also produced during freezing in winter wheat and cold acclimated plants are
more resistant to the addition of free radicals than nonacclimated plants (Kendall et al,
1989). The importance of this reduction is also seen by the observation that winter
survival in alfalfa plants overexpressing iron superoxide dismutase increases compared to

control plants (McKersie et al, 2000).

1.4.4 CHANGES IN MEMBRANES

Cold acclimation is associated with overall changes in the lipid composition of
cell membranes that result in increased cryostability (U emura et al, 1995; Uemura and
Steponkus, 1997). The exact changes that occur are species and organ specific (Uemura
et al, 1995), as is the absolute amount of the increase in fi'eezing tolerance. In the case of
the plasma membrane, the changes result in the production of exocytotic extrusions under
fi'eeze-induced dehydration as opposed to the more damaging endocyotic vesicles that are
produced if nonacclimated plants are frozen (see 1.3, Steponkus et al, 1988; Uemura et al,
1995). In chloroplast membranes, the lipid changes result in modifications that reduce
their propensity to form freeze-induced hexagonal-II phase transitions (U emura and
Steponkus, 1997). As the plasma and chloroplast membranes are considered the most
vulnerable to fi'eeze-induced dehydration stress, increasing the cryostability of these
membrane systems is critical to increasing overall plant freezing tolerance (Uemura et al,

1995; Uemura and Steponkus, 1997).

1.4.5 CHANGES IN SMALL MOLEC ULES

Upon exposure to low temperatures, drought, or other forms of osmotic stress,
plants accumulate a number of small benign molecules known as compatible solutes
(Smirnoff, 1998). Examples of the solutes produced include sugars, such as sucrose and
glucose, sugar alcohols, low-complexity carbohydrates, sulfonium compounds, and
amino acids, such as proline (Bohnert and Sheveleva, 1998). The functions of these
molecules is predicted to be the maintenance of turgor by lowering the osmotic potential,
protecting macromolecules from denaturation, and protecting and stabilizing membranes
(Smirnoff, 1998; Steponkus, 1984). It has been postulated that the increased
accumulation of sugars and proline could be the result of disturbances in metabolism that
occur under acclimating conditions (Bohnert and Sheveleva, 1998). However, while
increased levels may in part be due to metabolic imbalance, there is also good evidence to
suggest that accumulation increases through cold/osmotic—induced activation of the sugar
and proline production pathways (Hare et al, 1999). Overexpression of a cold-induced
transcription factor called CBF 3 (see 1.4.6.6) results in the increased accumulation of
sucrose and other soluble sugars and increased activation of PSCS, which increases
proline production (Gilmour et al, 2000).

Sugars are thought to firnction as cryoprotectants and have been shown to protect
membranes against freeze-induced damage in vitro (Strauss and Hauser, 1986;
Anchordoguy et al, 1987). Proline also appears to be able to increase freezing tolerance

as seen by the work of Kobayashi and colleagues (1999). By creating Arabidopsis plants

10

that overexpress the proline dehydrogenase gene (ProDH) in the antisense orientation,
transgenic plants with increased proline accumulation were isolated. Compared to control
lines, the transgenic plants showed increased tolerance to freezing and salinity stress,
indicating that proline accumulation is positively correlated with stress tolerance (N anjo
T, et al, 1999). While data on sugars and proline clearly indicate that both play roles in
increasing freezing tolerance, the exact mechanism by which both function to produce

this increase remains to be determined.

1. 4. 6 CHANGES IN GENE EXPRESSION

1. 4. 6. 1 Introduction

In 1970, Weiser (1970) proposed that cold acclimation involved changes in gene
expression. This was first conclusively demonstrated in 1985 in spinach (Guy et al, 1985)
and extended to Arabidopsis in 1988 (Gilmour et al, 1988). Since these discoveries, many
other groups have also found and characterized changes in gene expression in numerous
other species such as canola, (Orr et al, 1992; Weretilnyk et al, 1993) cabbage, (Sieg et
al, 1996), wheat, (Houde et al, 1992), barley (Crosatti et al, 1996; Phillips et al., 1997)
and alfalfa (Wolfraim et al., 1993; Monroy et al., 1993). Some of the changes in gene
expression involve proteins of known activities. One such set of proteins is the antifreeze
proteins, which inhibit ice crystal growth and block small crystals from recrystalizing
into larger crystals (Griflith et al, 1992; Yu and Griffith, 1999). Other proteins include

lipid transferases, alcohol dehydrogenases, translation elongation factors (Nishida and

11

Murata, 1996), phenylalanine ammonia-lyase and chalcone synthase (Levya et al, 1995)

to name a few. Still other changes involve proteins of unknown firnctions.

1.4.6.2 COR genes

In Arabidopsis, some of the best characterized cold-inducible genes are the COR
(cold-regulated) genes also called - LT 1 (low temperature-induced), KIN (cold-inducible),
RD (responsive to desiccation) and ERD (early dehydration-inducible), which are also
induced by drought stress and the application of ABA (Nordin et al, 1991; Wang et al,
1994; Welin et al, 1994). The Arabidopsis COR genes described to date are COR6. 6,
COR15, COR4 7 and COR 78, all of which have homologues located in tandem in the
genome (Thomashow et al, 1997). The COR genes encode proteins that are all highly
hydrophilic, soluble upon boiling in aqueous buffer, and are predicted to form
amphipathic a-helices (Lin and Thomashow, 1992; Thomashow, 1993; Thomashow,
1999). The COR78, COR15 and COR6.6 polypeptides are novel while COR47 has
similarities to LEA type H proteins, also known as dehydrins, which are associated with
desiccation and drought (Dure et al, 1989; Gilmour et al, 1991; Thomashow et al, 1993;
Close, 1997).

COR gene homologues are present in diverse freezing tolerant plant species. In B.
napus, BN28 (Orr et al, 1992) is a homologue of Arabidopsis COR6.6 and BN115
(W eretilnyk et al, 1993) is a homologue of Arabidopsis COR15, in wheat, wcs]20
(Houde et al, 1992), in barley, H VAI (Heino et al, 1990; Hong et al, 1992) and in alfalfa,
caSI5 (Monroy et al, 1993). These data suggest that COR gene homologues exist

12

throughout angiosperms and may play important roles in increasing the freezing/osmotic

stress tolerance of plants.

1.4.6.3 Overexpression of COR] 5a

The COR genes are clearly associated with cold and drought stress. This therefore
raised the question as to whether they actually play roles in increasing freezing tolerance.
To answer this question, Arabidopsis plants that constitutively overexpress COR15a were
created (Artus et al, 1996). Previous experiments had shown that COR15a is imported
into the stroma of the chloroplast and processed into a mature protein designated
COR15am(Artus et al, 1996; Lin and Thomashow, 1992; S. Gilmour and M.
Thomashow, unpublished). Artus and colleagues determined that COR15a-
overexpressing plants have a l-2° C increase in the freezing tolerance of chloroplasts and
isolated leaf protoplasts between the temperature range of —4 and —8°C as compared to
wild type plants. However, when investigating the freezing tolerance of detached leaves
by electrolyte leakage analysis, or by conducting whole plant freeze tests, there were no
differences seen between COR15a-overexpressing plants and wild type plants (Artus et
al, 1996; Jaglo-Ottosen et al, 1998). Fluorescein diacetate staining of the protoplasts, an
indicator of whether or not plasma membranes have retained their semi-permeable nature,
indicated that COR15a overexpression protected the plasma membrane between -—5 to —

8°C (Artus et al, 1996).

13

 

1.4.6.4 Function of COR15a polypeptide

Since overexpression of the COR15a protein results in increased freezing
tolerance, how the protein functions to bring about this increase became of interest. To
answer this question, isolated COR15a-overexpressing protoplasts were frozen and the
effects on the membranes were determined (Steponkus et al, 1998). A deferment in the
production of hexagonal-II phase transitions was observed in protoplasts from COR15a
overexpressing plants (Steponkus et a], 1998). The inner chloroplast membrane may be
the most sensitive to freezing damage (Steponkus et al, 1998). If so, protecting this
membrane, the “weak link” of plant membranes, would result in an overall increase in
freezing tolerance in the plant. This raises the intriguing possibility that the COR15a
polypeptide functions in association with the other COR polypeptides in an additive
manner to protect plant membranes from freezing damage. However, whether this is the

case for the other COR polypeptides remains to be determined.

1.4.6.5 COR gene regulation

COR gene transcript levels increase within about four h of a low temperature
stimulus and remain high for as long as plants are kept at low temperatures (Hajela et a],
1990). This raised the question of how the COR genes are regulated. Promoter analysis of
the COR genes lead to the discovery that they are transcriptionally regulated through a
cold- and drought inducible promoter element, called the CRT/DRE (C-Repeat/Drought

Responsive Element). The DRE element, TACCGACAT, was first identified by

14

Yamaguchi-Shinozaki and Shinozaki (1994) as being sufficient for drought and cold
inducibility. The core of the DRE, CCGAC, called the CRT, is present in multiple copies
in the COR6.6 (Wang et al, 1995), COR15 (Baker et al, 1994) and COR78 (Horvath et a],
1993; Yamaguchi-Shinozaki and Shinozaki, 1993) promoters. Deletion experiments
using the COR15a promoter showed that cold-inducible reporter gene activation was
consistent with the presence of the CRT (Baker et al, 1994). Interestingly, further
research has shown that this sequence is not only found in multiple copies in COR gene
promoters in Arabidopsis, but also in the promoters of some of the cold induced genes in
other species. These include the BN115 gene in Brassica napus (Jiang et al, 1996), and
the wcsIZO gene in wheat (Ouellet et al, 1998). The presence of COR gene homologues
along with the conservation of the cold and drought inducible promoter element, the
CRT/DRE, between diverse species strongly suggests that cold and drought inducible

gene expression is highly conserved.

1.4.6.6 The CBF family of transcription factors

The CRT/DRE sequence is involved in cold- and drought-induced activation of
the COR genes. The next question became, what protein was binding to the CRT/DRE
and activating transcription under acclimating conditions? A major breakthrough in
answering this question was the isolation of CBF], (CRT/DRE binding factor 1)
(Stockinger et al, 1997). CBF] was isolated in a one-hybrid assay using the COR15a
CRT/DRE promoter element as bait. CBF] protein contains a putative nuclear
localization signal, (Raikhel, 1992) an AP2-like DNA binding domain (W eigel, 1995;

15

Ohme-Takagi and Shinshi, 1995) and an acidic C-terminal domain consistent with some
types of transcriptional activators (Hahn, 1993). In vitro assays indicate that CBF] binds
specifically to the CRT/DRE sequence and in vivo assays demonstrate that CBF] can
activate transcription of a reporter gene under the control of the COR15a promoter in
yeast (Stockinger et al, 1997).

Additional research has led to the discovery that CBF] is part of a small gene
family consisting of CBF], CBF2 and CBF3, also called DREBI b, DREBIc and DREBIa
respectively (Gilmour et al, 1998; Liu et al, 1998). The three CBF genes are highly
similar in amino acid sequence (~85%) and are located in tandem about two kb apart on
chromosome 4 (Gilmour et al, 1998). Like CBF], CBF2 and CBF 3 contain AP2-like
DNA binding domains, putative acidic activation domains, and can activate transcription
in yeast (Stockinger et al, 1997; Gilmour et al, 1998; Liu et al, 1998; Shinwari et a],
1998). All three CBF proteins can bind to CRT/DRE sequences contained within the
COR15a, COR15b, and COR 78 promoters, and appear to have the highest binding
aflinity for the COR15b promoter (Gilmour et al, 1998). The reason for this preferential

binding to the COR15b promoter is not currently known.

1.4.6. 7 CBF gene regulation

RNA hybridization analysis has indicated that all three CBF transcripts are
upregulated within 15 min of a low temperature stimulus, peak at two-four h, and remain
at a higher steady state level for up to 24 hours (Gilmour et al, 1998). As seen previously,

COR gene transcripts increase after ~four h suggesting the possibility of a cold induced

16

signal cascade (Gilmour et al, 1998). CBF transcripts are also induced by mechanical
agitation, although only transiently. Whether the CBF transcripts are induced by ABA or
osmotic stress has been difficult to determine accurately due to the low levels of RNA
accumulation, although it appears a slight increase may occur (Gilmour et al, 1998; D.
Zarka and M. Thomashow, unpublished).

Investigation of the promoters of the three CBF genes has shown that none
contain CRT/DRE sequences (Gilmour et al, 1998). Additionally, overexpression of
CBF] did not result in the activation of CBF 3, strongly suggesting that the genes are not
induced by autoregulation (Gilmour et al, 1998). While the promoters of the CBF genes
contain consensus sequences consistent with G-box core and myc recognition sites, they
do not contain known consensus sequences associated with cold activation (Gilmour et
al, 1998). The exact mechanism by which the CBF genes are upregulated under cold-

acclimating conditions remains to be determined.

1.4.6.8 77w SFR6 mutant

One possible clue to how the CBF proteins function to regulate the COR genes
comes fi'om analysis of mutant Arabidopsis plants. Warren and colleagues (1996)
screened for Arabidopsis mutants, Sfi' (sensitive to freezing), that had reduced freezing
tolerance even after cold acclimation. One of the mutants, sfr6, has the interesting
phenotype of an increase in CBF RNA accumulation under acclimating conditions, but
no induction of COR gene expression (Knight et al, 1999). ABA and osmotic stress

induced COR gene expression, as seen by assaying for KIN] (COR6. 6) RNA

17

accumulation, is also lost. Cloning the gene containing this recessive mutation should

give important clues as to how the CBF proteins activate COR gene expression.

1.4.6.9 Changes in calcium

Exposure to low temperature results in a transient increase in cytosolic calcium in
plants. The rise is initiated by calcium influx through the plasmalemma and by release of
calcium from internal stores, such as the vacuole Knight et al, 1996). This change in
calcium has been implicated in the control of cold-responsive gene expression (Knight et
al, 1996; Tahtiharju et al, 1997) and in increasing freezing tolerance in plants (Monroy et
al, 1993; Monroy and Dhindsa, 1995). Specifically, if chickpea plants are exposed to 0.5
mM CaClz, genes that are responsive to low temperature as well as other osmotic stresses
are upregulated (Colorado et al, 1994). Work with alfalfa cell suspension cultures has
shown that the cold inducibility of the cas genes, cas]5 and casl8, is obliterated when
external calcium influxes are blocked with inhibitors, and that the genes are induced at
25° C when calcium ionophores or calcium channel agonists are added (Monroy and
Dhindsa, 1995).

How calcium influxes specifically function to elicit cold-induced gene expression
remains unknown. There is evidence that protein phosphorylation occurs after cold
induced calcium influxes (Monroy et al, 1998). These phosphorylation events could be
activated by calcium dependent protein kinases (CDPKs), or other types of protein
kinases, such as mitogen-activated protein (MAP) kinases. At present, there is evidence

that RNA accumulation of CDPKS in Arabidopsis increases after exposure to low

18

temperatures (Tahtiharju et a], 1997). The sequencing of the Arabidopsis genome has led
Harmon et al (2000) to predict that 40 CDPKs are present in Arabidopsis.

There is evidence that cold- and osmotic stress-inducible MAP kinases exist in
plants. Jonak et a] (1996) have identified a MAP kinase in alfalfa, the activity of which is
activated within 10 min of exposure of plants to low temperatures. Additionally, a MAP
kinase and a MAP kinase kinase kinase have been identified in Arabidopsis, which are
induced by touch, cold and water stress (Mizoguchi et al, 1996). Another kinase
identified in Arabidopsis, ATMEKK] , is induced within 5 min of osmotic (N aCl) stress
and increases the survival of yeast expressing ATMEKK] under high NaCl conditions
(Covic et a], 1999). Future work involving the characterization of these and other cold
and stress-induced kinases should increase our understanding of the upstream signals

involved in cold induction pathways.

1.5 REFERENCES

Anchordoguy, TJ, Ruldolph, AS, Carpenter, JF, Crowe, JH. 1987. Modes of interaction
of cryoprotectants with membrane phospholipids during freezing. Cryobiology. 24: 324-
33 1 .

Artus, NN, Uemura, M, Steponkus, PL, Gilmour, SJ, Lin, C, Thomashow, MF. 1996.
Constitutive expression of the cold-regulated Arabidopsis thaliana COR15a gene affects
both chloroplast and protoplast freezing tolerance. Proc. Natl. Acad Sci. 93: 13404-
13409.

Colorado, P, Rodriguez, A, Nicolas, G, Rodriguez, D. 1994. Abscisic-acid and stress
regulate gene-expression during germination of chickpea seeds - possible role of calcium.
Physiol. Plant. 9]: 461-467.

Covic, L, Silva, NF, Lew, RR. 1999. Functional characterization of ARAKIN

(ATMEKKI): a possible mediator in an osmotic stress response pathway in higher plants.
Biochim. Biophys. Acta. 145 1: 242-254.

19

Crosatti, C, Nevo, E, Stanc, AM, Cattivelli, L. 1996. Genetic analysis of the
accumulation of COR14 proteins in wild (Hordeum spontaneum) and cultivated
(Hordeum vulgare) barley. Theor. Appl. Genet. 93: 975-981.

Engelrnan, R, LeRoy, P. 2000. http://www.cnie.org/pop/conserving/landuse.htm#The
Earth’s Arable Land and Population.

Gilmour, SJ, Hajela, RK, Thomashow, MF. 1988. Cold-acclimation in Arabidopsis-
thaliana. Plant Physiol. 87: 745-750.

Gilmour, SJ, Thomashow, MF. 1991. Cold-acclimation and cold-regulated gene-
expression in ABA mutants of Arabidopsis-thaliana. Plant Mol Bio]. 17: 1233-1240.

Griffith, M, Ala, P, Yang, DSC, Hon, WC, Moffatt, BA. 1992. Antifreeze protein
produced endogenously in winter rye leaves. Plant Physiol. 100: 593-596.

Guy, CL, Niemi, KJ, Brambl, R. 1985. Altered gene-expression during cold-acclimation
of spinach. Proc. Nat]. Acad Sci. USA. 82: 3673-3677.

Hahn, S. 1993. Structure(questionable) and function of acidic transcription activators.
Cell. 72: 481-483.

Hare, PD, Cress, WA, van Staden, J. 1999. Proline synthesis and degradation: a model
system for elucidating stress-related signal transduction. J. Exp. Bot. 50: 413-434.

Harmon, AC, Gribskov, M, Harper, JF. 2000. CDPKs - a kinase for every Ca2+ signal?
Trends in Plant Science. 5: 154-159.

Heino, P, Sandman, G, Lang, V, Nordin, K, Palva, ET. 1990. Abscisic-acid deficiency
prevents development of fi'eezing tolerance in Arabidopsis-thaliana (L) Heynh. Theor.
App]. Genet. 79: 801-806.

Hong, BM, Barg, R Ho, T-HD. 1992. Developmental and organ-specific expression of
an ABA-induced and stress-induced protein in barley. Plant Mol Bio]. 18: 663 -674.

Houde, M, Dhindsa, RS, Sarhan, F. 1992. A molecular marker to select for freezing
tolerance in Gramineae. Mol. Gen. Genet. 234: 43-48.

Jaglo-Ottosen, KR, Gilmour, SJ, Zarka, DG, Schabenberger, O, Thomashow, MF. 1998.
Arabidopsis CBF] overexpression induces COR genes and enhances fi'eezing tolerance.
Science. 280: 104-106.

Jonak, C, Kiegerl, S, Ligterink, W, Barker, PJ, Huskisson, NS Hirt, H. 1996. Stress

signaling in plants: A mitogen-activated protein kinase pathway is activated by cold and
drought. Proc. Nat]. Acad Sci. USA. 93: 11274-11279.

20

Kendall, EJ, McKersie, BD. 1989. Free-radical and fieezing-injury to cell-membranes of
winter-wheat. Physio]. Plant. 76: 86-94.

Klimov, SV, Astakhova, NV, Trunova, TI. 1997. Relationship between plant cold
tolerance, photosynthesis and ultrastructural modifications of cells and chloroplasts. R. J.
ofPlantPhysiol. 44: 759-765.

Knight, H, Trewavas, AJ, Knight, MR. 1996. Calcium signalling in Arabidopsis thaliana
responding to drought and salinity. Plant Cell. 8: 489-503.

Knight, H, Veale, EL, Warren, GJ, Knight, MR. 1999. The sfr6 mutation in arabidopsis
suppresses low-temperature induction of genes dependent on the CRT DRE sequence
motif. Plant Cell. 11: 875-886.

Koornneef, M, Leon-Kloosterziel, KM, Schwartz, SH, Zeevart, JAD. 1998. The genetic
and molecular dissection of abscisic acid biosynthesis and signal transduction in
Arabidopsis. Plant Physio]. Biochem. 36: 83-89.

Kratsch, HA, clever, RR. 2000. The ultrastructure of chilling stress. Plant, Cell and
Environ. 23: 337-350.

Kubacka-Zebalska, M, 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). Plant Science. 148: 59-67.

Lang, V, Palva, ET. 1992. The expression of a RAB-related gene, RABI8, is induced by
abscisic-acid during the cold-acclimation process of Arabidopsis-thaliana (L) Heynh.
Plant Mol. Biol. 20: 951-962.

Levya, A, Jarillo, JA, Salinas, J, Martinez-Zapater, JM. 1995. Low-temperature induces
the accumulation of phenylalanine ammonia-lyase and chalcone synthase messenger-
RNAs of Arabidopsis-thaliana in a light-dependent manner. Plant Physiol. 108: 39-46.

Lin, CT, Thomashow, MF. 1992. DNA-sequence analysis of a complementary-DNA for
cold-regulated Arabidopsis gene C OR15 and characterization of the COR-15 polypeptide.
Plant Physiol. 99: 519-525.

McKersie, BD, Murnaghan, J, Jones, KS, Bowley, SR. 2000. Iron-superoxide dismutase
expression in transgenic alfalfa increases winter survival without a detectable increase in
photosynthetic oxidative stress tolerance. Plant Physiology 122: 1427-143 7.

Mizoguchi, T, Irie, K, Hirayama, T, Hayashida, N, Yamaguchi-Shinozaki, K, et al. 1996.
A gene encoding a mitogen-activated protein kinase kinase kinase is induced
simultaneously with genes for a mitogen-activated protein kinase and an S6 ribosomal
protein kinase by touch, cold, and water stress in Arabidopsis thaliana. Proc. Natl. Acad.
Sci. USA. 93: 765-769.

21

Monroy, AF, Castonguay, Y Laberge, S Sarhan, F Vezina, LP Dhindsa, RS. 1993. A new
cold-induced alfalfa gene is associated with enhanced hardening at subzero temperature.
Plant Physiol. 102: 873-879.

Monroy, AF, Dhindsa, RS. 1995. Low-temperature signal-transduction - induction of
cold acclimation-specific genes of alfalfa by calcium at 25-degrees-C. Plant Cell. 7: 321-
331.

Monroy, AF, Sangwan, V, Dhindsa, RS. 1993. Induction of cold acclimation-specific
alfalfa genes by calcium influx at room-temperature. Plant Physiol. 102: 1227-1235.

Monroy, AF, Sangwan, V, Dhindsa, RS. 1998. Low temperature signal transduction
during cold acclimation: protein phosphatase 2A as an early target for cold-inactivation.
Plant J. 13: 653-660.

Nanjo T, Kobayashi, NT, Yoshiba, M, Kakubari, Y, Yamaguchi-Shinozaki K, Shinozaki
K. 1999. Antisense suppression of proline degradation improves tolerance to fi'eezing and
salinity in Arabidopsis thaliana. Febs Letters. 461: 205-210.

Nishida I, Murata, N. 1996. Chilling sensitivity in plants and cyanobacteria: The crucial
contribution of membrane lipids. Ann. Rev. Plant Physiol. Plant Mol. Biol. 47: 541-568.

Nordin, K, Heino, P, Palva, ED. 1991. Separate signal pathways regulate the expression
of a low-temperature—induced gene in Arabidopsis-thaliana (L) Heynh. Plant Mol. Biol.
16: 1061-107].

Ohme-Takagi, M, Shinshi, C. 1995. Ethylene-inducible DNA-binding proteins that
interact with an ethylene-responsive element. Plant Cell. 7: 173-182.

Orr, W, In, B, White, TC, Robert, LS, Singh, J. 1992. Complementary-DNA sequence of
a low temperature-induced Brassica napus gene with homology to the Arabidopsis-
thaliana kin] gene. Plant Physiol. 98: 1532-1534.

Phillips, JR, Dunn, MA, Hughes, MA. 1997. mRNA stability and localisation of the low-
temperature-responsive barley gene family blt14. Plant Mol. Biol. 33: 1013-1023.

Raikhel, N. 1992. Nuclear targeting in plants. Plant Physiol. 100: 1627-1632.

Savitch, LV, Hamey, T, Huner, NPA 2000. Sucrose metabolism in spring and winter
wheat in response to high irradiance, cold stress and cold acclimation. Physiologia
plantarum. 108: 270-278.

Sieg F, Schroder W, Schmitt JM, Hincha DK. 1996. Purification and characterization of a

cryoprotective protein (cryoprotectin) from the leaves of cold-acclimated cabbage. Plant.
Physiol. 1996, 1 1 1: 215-221.

22

Seppanen, MM, F agerstedt, K. 2000. The role of superoxide dismutase activity in
response to cold acclimation in potato. Physiologia planatarum. 108: 279-285.

Smimofl‘, N. 1998. Plant resistance to environmental stress. Curr. Opion. Plant Biotech.
9: 214-219.

Steponkus, PL. 1984. Role of the plasma-membrane in freezing-injury and cold-
acclimation. Annu. Rev. Plant. Physiol. 35: 543-585.

Steponkus PL, Lynch, DV. 1989. Freeze thaw-induced destabilization of the plasma-
membrane and the effects of cold-acclimation. J. Of Bioener. Biomem. 21: 21-41.

Steponkus, PL, Uemura, M, Balsamo, RA, Arvinte, T, Lynch, DV. 1988. Transformation
of the cryobehavior of rye protoplasts by modification of the plasma-membrane lipid-
composition. Proc. Natl. Acad Sci. USA. 85: 9026-9030.

Steponkus, PL, Uemura, Joseph, RA, Gilmour, SG, Thomashow, MF. 1998. Mode of
action of the COR15a gene on the freezing tolerance of Arabidopsis thaliana. Proc. Natl.
Acad Sci. USA. 95: 14570-14575.

Steponkus, PL, Uemura, M, Webb, MS. 1993a. A contrast of the cryostability of the
plasma membrane of winter rye and spring oat- two species that widely difi‘er in their
freezing tolerance and plasma membrane composition. In Advances in Low-Temperature
Biology, ed. PL Steponkus, 2L 2]]- 312. London: JAAI Press.

Steponkus, PL, Uemura, M, Webb, MS. 1993b. Membrane destabilization during freeze-
induced dehydration. Curr. Topics Plant Physiol. 10: 37-47.

Stockinger, EJ, Gilmour, SJ, Thomashow, MF. 1997. Arabidopsis thaliana CBF]
encodes an AP2 domain-containing transcriptional activator that binds to the C-
repeat/DRE, a cis-acting DNA regulatory element that stimulates transcription in
response to low temperature and water deficit. Proc. Natl. Acad Sci. 94: 1035-1040.

Strand A, Hurry, V, Gustafsson, P, Gardestrorn, P. 1997. Development of Arabidopsis
thaliana leaves at low temperatures releases the suppression of photosynthesis and

photosynthetic gene expression despite the accumulation of soluble carbohydrates. Plant
J. 12: 605-614.

Strand, A, Hurry, V, Henkes, S, huner, N, Gustafsson, P, Gardestrorn, P, Stitt, M. 1999.
Acclimation of Arabidopsis leaves developing at low temperatures. Increasing
cytoplasmic volume accompanies increased activities of enzymes in the Calvin
cycle and in the sucrose-biosynthesis pathway. Plant Physio]. 119: 1387-1397 .

Strauss, G, Hauser, H. 1986. Stabilization of lipid bilayer vesicles by sucrose during
freezing. Proc. Natl. Acad Sci. USA. 83: 2422-2426.

23

Tahtiharju, S, Sangwan, V. Monroy, AF, Dhindsa, RS, Borg, M. 1997. The induction of
KIN genes in cold-acclimating Arabidopsis thaliana. Evidence of a role for calcium.
Planta. 203: 442-447.

Thomashow, MF. 1990. Molecular genetics of cold acclimation in higher plants. Adv.
Genet. 28: 99-131.

Thomashow, MF. 1999. Plant cold acclimation: Freezing tolerance genes and regulatory
mechanisms. Ann. Rev. Plant Physiol. Plant Mol. Biol. 50: 571-599.

Uemura, M, Joseph, RA, Steponkus, PL. 1995. Cold-acclimation of Arabidopsis-thaliana
- efl‘ect on plasma-membrane lipid-composition and freeze-induced lesions. Plant
Physiol. 109: 15-30.

Uemura, M, Steponkus, PL. 1997. Effect of cold acclimation on membrane lipid
composition and freeze-induced membrane destabilization. In Plant Cold Hardiness,
Molecular Biology, Biochemistry and Physiology. New York: Plenum. pp 171-17 9.

Warren, G, McKown, R Marin, A, Teutonico, R. 1996. Isolation of mutations affecting
the development of freezing tolerance in Arabidopsis thaliana (L) Heynh. Plant Physiol.
111: 1011-1019.

Weigel, D. 1995. The APETALAZ domain is related to a novel type of DNA-binding
domain. Plant Cell. 7: 388-389.

Weretilnyk, E, Orr, W, White, TC, Iu, B, Singh, J. 1993. Characterization of 3 related
low-temperature-regulated cDNAs from winter Brassica-napus. Plant Physiol. 101: 17]-
177.

Wiser, CJ. 1970. Cold resistance and injury in woody plants. Science. 169: 1269-1277.
Wolfi'aim, LA, R. Langis, H. Tyson, Dhindsa, RS. 1993. cDNA sequence, expression,
and transcript stability of a cold acclimation-specific gene, cas18, of alfalfa (Medicago-
falcata) cells. Plant Physiol. 101: 1275-1282.

Yu XM, Griflith M. 1999. Antifieeze proteins in winter rye leaves form oligomeric
complexes. Plant Physiol. 119: 1361-1369.

24

2. CHAPTER 2: Overexpression of Arabidopsis CBF] results in
increased COR gene expression and freezing tolerance without a low

temperature stimulus

2.1 INTRODUCTION

The adverse environmental stress of freezing temperatures severely limits crop
yield and geographical area of production (Levitt, 1980; Boyer, 1982). Any increase in
the ability of plants to withstand low and freezing temperatures would therefore be of
great economic importance. One way in which some plant species survive freezing
temperatures is by cold acclimation, the process by which plants increase in freezing
tolerance after exposure to low, nonfreezing temperatures. Understanding the various and
complex processes involved in cold acclimation has long been the focus of research
efforts as a potential means to increase freezing tolerance.

An early hypothesis was that cold acclimation, in addition to being associated
with numerous physiological and biochemical changes, was associated with changes in
gene expression (W eiser, 1970). This hypothesis was first conclusively demonstrated in
spinach, (Guy et al, 1985) then extended in Arabidopsis with the isolation of a family of
genes, called the COR genes (cold-regulated) -also called LTI (low temperature-induced),
KIN (cold-inducible), RD (responsive to desiccation) and ERD (early dehydration-
inducible)- which are greatly upregulated under cold-acclimating and drought conditions

as well as by application of ABA (Hajela et al, 1990; Nordin et al, 1991; Wang et al,

25

1994; Welin et al, 1994). Analysis of the COR genes, which include COR6. 6, COR15,

C 0R4 7 and COR 78, has shown that all four COR genes are actually members of gene
pairs, each of which is located in tandem in the genome (Thomashow et al, 1997). The
proteins encoded by these genes are all highly hydrophilic and have the unusual trait of
being soluble upon boiling in aqueous buffer (Lin et al, 1990; Thomashow, 1993). Amino
acid analysis comparisons indicate that COR78, COR15 and COR6.6 are novel
polypeptides while COR47 has similarities to LEA type H proteins, also known as
dehydrins, which are associated with desiccation and drought (Dure et al, 1989; Gilmour
et al, 1991; Thomashow et a], 1993; Close, 1997; Thomashow, 1999).

Due to their association with cold acclimation and drought conditions, both of
which cause damage to membranes, and their similarities to LEA proteins, it has been
hypothesized that the role of the COR genes is to protect membranes from osmotic stress
damage. Work by Steponkus and colleagues has shown that the primary site of damage to
plants during freezing conditions are the cellular membranes which experience freeze-
induced dehydration (Steponkus et al, 1993; Steponkus, 1984). To identify a functional
role for the COR] 5am protein, which is targeted to the chloroplast, Artus et al. (1996)
generated C OR15a overexpressing plants which were assayed for freezing tolerance.
Compared to wild type plants no increase in freezing tolerance was seen at the whole
plant level, however, an increase in the freezing tolerance of chloroplasts between -4 and
—8° C was observed (Artus et al, 1996). When protoplasts overexpressing COR15a were
investigated under freezing conditions, Steponkus et al (1998) found that there was a
deferment of hexagonal 11 phase transitions to lower temperatures (see section 1.3),

proposed to be due to changes in the intrinsic curvature of the inner membrane of the

26

chloroplast envelope (see section 1.4.6.4). Together, these data indicate a role for COR15
in protecting chloroplast membranes. Additionally, they present the intriguing possibility
that the other COR genes may play similar roles in protecting plant membranes, and that
the combined expression of all of the COR genes together results in the increase in
freezing tolerance seen after cold acclimation. If so, then overexpression of the battery of
COR genes (meaning all genes that are upregulated under acclimating conditions, not just
the identified four COR gene families mentioned above) could result in increased

freezing tolerance, as overexpression of one COR gene alone, COR15, did not.

In order to be able to manipulate the expression of all of the COR genes
simultaneously, a better understanding COR gene regulation during cold acclimation was
required. Promoter analysis led to the discovery that the COR genes are transcriptionally
regulated through a cold- and drought inducible promoter element, called the CRT/DRE
(C-Repeat/Drought Responsive Element). The DRE, TACCGACAT, was first identified
by Yamaguchi-Shinozaki and Shinozaki (1994) as being cold and drought responsive.
The core of the DRE, the CRT, CCGAC, is present in multiple copies within the
promoters of COR6.6 (Wang et al, 1995), COR15 (Baker et aL 1994) and COR 78
(Horvath et al, 1993; Yamaguchi-Shinozaki and Shinozaki, 1993). Deletion experiments
with the COR15a promoter showed reporter gene activation consistent with the CRT
being the cis-acting acting element responsible for the cold-inducible activation (Baker et
al, 1994). This element, however, is not associated with the ABA-induced activation of
the COR genes (Baker et al, 1994; Yamaguchi-Shinozaki and Shinozaki, 1994; Wang et
al, 1995). This observation corroborates the data indicating that ABA-induced activation

of the COR genes occurs through a pathway separate from cold-and drought-induced

27

activation (Gilmour and Thomashow, 1991; Nordin et al, 1991) (see section 1.4.2).

A major breakthrough in understanding the signaling cascade activated by low
temperature stimulus was the isolation of a transcription factor, called CBF] (CRT/DRE
Binding Factor 1) (Stockinger et al, 1997). CBF] was suggested to be a unique gene or a
member of a small gene family the protein product of which binds to the CRT/DRE
element and activates transcription under cold acclimating conditions (Stockinger et al,
1997). The CBF] protein contains an AP2-like DNA binding domain (W eigel, 1995;
Ohme-Takagi and Shinshi, 1995), an acidic C-terminal portion that could act as a
transcription activation domain (Hahn, 1993 ), and a putative nuclear localization domain
(Raikhel, 1992), all of which are consistent with the conclusion that CBF! encodes a
transcription factor. Gel shift experiments showed that CBF] binds specifically to the
CRT/DRE sequences contained within the COR15a promoter in vitro (Stockinger et al,
1997). Additionally, in vivo transcription activation assays showed that CBF] can
activate the transcription of a reporter gene under the control of the COR15a promoter in
yeast (Stockinger et al, 1997). CBF] itself is upregulated under acclimating conditions in
Arabidopsis (Gilmour et al, 1998). However, there do not appear to be any CRT/DRE
sequences in the CBF] promoter, indicating that CBF] is not autoregulated (S. Gilmour
and M. Thomashow, unpublished).

Here I present data indicating that CBF] binds to the CRT/DRE and activates
COR gene transcription in Arabidopsis plants. Transgenic Arabidopsis var. RLD plants
that constitutively overexpress CBF] were generated. The transgenic plants have
increased COR6. 6, COR15, COR4 7 and COR 78 RNA accumulation and increased

COR6.6 and COR15 protein accumulation under nonacclimating conditions as compared

28

to wild type plants. Furthermore, the CBFI-overexpressing plants showed increased
freezing tolerance as determined by electrolyte leakage assays and whole plant freeze
tests. CBF] is therefore a likely activator of COR genes under acclimating conditions.
Additionally the COR genes appear to play direct roles in increasing freezing tolerance as
overexpression of the COR genes results in increased freezing tolerance without a low
temperature stimulus. An attempt was also made to determine if CBFI-overexpressing
plants also have increased tolerance to osmotic stress. However, no differences in
germination frequency were seen between transgenic plants and control plants grown
under conditions of osmotic stress.

Additionally, constitutive overexpression of CBF] was found to result in
transgene silencing. Expression levels of the CBF-transgene decreased as the plants were
self-pollinated and carried to the T3 —T5 generations. The reduction in transgene
expression was closely correlated with a reduction in COR gene expression and a loss of
increased freezing tolerance. In an attempt to determine the effects of eliminating CBF]
expression, plants that constitutively overexpress CBF] in the antisense orientation were
generated. Although RNA in the antisense orientation was present in transgenic plants, no

differences in COR gene expression at either the RNA or proteins levels were detected.

2.2 MATERIALS AND METHODS

2. 2. 1 PLANT GROWTH

Arabidopsis thaliana (Arabidopsis) ecotype RLD plants were grown in pots under

29

100 umol'm'z's'l continuous light for 18-25 days as described (Gilmour et al, 1988). For
experiments involving acclimation, plants were cold acclimated at 4° C under 50
timol'm'z's'l continuous fluorescent illumination for the amount of time indicated. The
types of plants used in experiments include wild type plants, RLD, plants that
constitutively overexpress COR15a, T8 (Artus et al, 1996) and plants that constitutively

overexpress CBF], A6, B16, K16 and 1-11 (described below) in the T2-T5 generations.

2. 2. 2 PLANT TRANSFORMATION

Standard procedures were used for plasmid manipulations (Sambrook et al, 1989).
Specifically, to make transgenic plants, the CBFl-containing AseI-Bgl I] fragment fi'om
pACT-Bgl+ (Stockinger, 1997) was gel purified, end-filled by Klenow (GibcoBRL,
Grand Island, NY) and BamHI linkers were ligated to both ends. The fragment was then
digested with BamI-II and ligated into the BamI-II site of pCIB710 (Rothstein et al, 1987)
which contains the Cauliflower Mosaic Virus (CaMV) 3SS constitutive RNA promoter
and terminator. Plasmids containing both the sense (pKJOl) and antisense (pKJ 02)
orientation of the CBF] gene were isolated and used to make constructs in both
orientations. The chimeric plasmids were individually linearized at the KpnI site and
inserted into the KpnI site of the binary vector pCIBlOg (Ciba—Geigy, Research Triangle
Park, NC), resulting in pKJOla and pKJOZa for the sense and antisense constructs
respectively. The plasmids were transformed into Agrobacterium tumefaciens strain

C58C1 (pMP90) by electroporation. Arabidopsis plants were transformed by the vacuum

30

infiltration procedure (Bechtold, et al, 1993) as modified by van Hoof and Green (1996).
Transgenic plants were selected by survival on 1 x Gamborgs B-5 media (GibcoBRL,
Grand Island, NY) containing 50 ug/ml kanamycin. Homozygous lines of individual
plants that had uniform kanamycin resistance in the T3 generation and high transgene

expression were selected for further study (A6, B16, 1-11, and K16).

2. 2. 3 DNA HYBRIDIZA TION

Total Arabidopsis DNA was isolated as described (Rogers and Bendich, 1988).
Approximately 5-10 ug of total DNA was individually digested with HindIII or EcoRV,
electrophoreased on 1% agarose gels, and transferred to nylon membranes (Micron
Separations, Westboro, MA) following standard protocols (Sambrook et al, 1989). The
high and low stringency conditions used were identical to those used for RNA
accumulation analysis with the exception that the high-stringency wash was conducted at
65° C instead of 50° C (Stockinger et al, 1997). The DNA probe for CBF] (Stockinger et
al, 1997) was gel purified full length CBF] labeled with 32P by random-priming

according to standard procedures (Sambrook et al, 1989).

2. 2. 4 RNA HYBRIDIZA TION

Total RNA was isolated from plant leaves as described (Gilmour et al, 1988).

Northern transfers were prepared using 20 ug of total RNA electrophoreased in 1.5%

31

formaldehyde gels. The resulting membranes were hybridized to 32P labeled double
stranded DNA probes and washed at high stringency (Stockinger et al, 1997). Double
stranded DNA probes for CBF] (Stockinger et al, 1997), COR6. 6, COR15, COR4 7,
COR 78 (Hajela et al, 1990) and EIF 4A (Metz et al, 1992) were gel purified and labeled

with 32P by random-priming according to standard procedures (Sambrook et al, 1989).

2. 2. 5 IWUNOBLOTANAL YSIS

Total soluble protein was extracted by grinding fiozen tissue (about 150 mg) in
400 u] extraction buffer containing 50 mM EDTA (pH 8.0), 50 mM HEPES (pH 7.0),
and 1.5% (wt/vol) polyvinyl-pyrrolidone, after which insoluble material was removed by
centrifugation (13,000 x g for 20 min at 4° C). The protein concentration in the
supernatant was determined using the Bradford dye-binding assay (Bio-Rad, Hercules,
CA). Total soluble protein (100 ug per sample) was fractionated by 10% tricine
SDS/PAGE (Schagger and von Jagow, 1987), and transferred to 0.2 um nitrocellulose
membranes by electroblotting (Towbin et al, 1979) as described (Artus et al, 1996).
COR15 protein was detected using antiserum raised to purified COR15am(Artus et a],
1996) and protein A conjugated to alkaline phosphatase (Sigma) as described (Blake et

al, 1984). No reacting bands were observed with preimmune serum.

2. 2. 6 ELECTROL YTE LEAKAGE ASSA IS

32

Electrolyte leakage assays were conducted as described (Sukumaran and Weiser,
1972; Gilmour et al, 1988) with the following modifications. Five replicates of two to
four detached leaves from nonacclimated or cold-acclimated plants were placed in a test
tube and submerged in a —2° C bath containing water and ethylene glycol in a completely
randomized design for 1 h. Ice crystals were added to nucleate freezing. After an
additional h at —2° C, the samples were cooled in decrements of 1° C each h until —8° C
was reached. Samples (five replicates of each data point) were thawed overnight on ice
and shaken in 3 ml distilled water at room temperature for 3 h. Electrolyte leakage fi'om
leaves was measured with a conductivity meter. The solution was then removed, the
leaves were frozen at -—80° C for 1 to 10 h. The solution was returned to each tube and
shaken for another 3 h to obtain a value for 100% electrolyte leakage.

The temperature at which 50% of electrolytes were leaked (ELso) were
determined by fitting model curves of up to third-order linear polynomials for each
electrolyte leakage test. To ensure unbiased predictions of electrolyte leakage, trends
significantly improving the model fit at the 0.2 probability level were retained. ELso
values were calculated from the fitted models. An unbalanced one-way analysis of
variance (ANOVA), adjusted for the different number of ELso for each type was
determined using SAS PROC GLM [SAS Institute, SAS/ STAT User’s Guide, Version 6

(SAS Institute, Cary, NC, 1989)].

2. 2. 7 WHOLE PLANT FREEZE TESTS

33

Nine-centimeter pots containing ~40 RLD, T8 (COR15a-overexpressing) B16 or
A6 (CBFI-overexpressing) Arabidopsis plants that were either 20 days old and
nonacclimated or 25 days old and 4-day cold-acclimated plants were placed in a
completely randomized design in a —-5° C cold chamber in the dark. After 1 h, ice chips
were added to each pot to nucleate fi'eezing. Plants were removed after 1-3 days, returned
to a grth chamber at 22° C and allowed to recover for 7 days before photographs were

taken.

2. 2. 8 SALT STRESS GERMINA TION TESTS

Wild type RLD plants and two CBFI-overexpressing transgenic lines in the T3
generation, A6 and B16, were germinated on petri plates containing 1 x Gamborgs B-S
media (GibcoBRL,Grand Island, NY) solidified with 1% agarose, or Gamborgs B-S
media, 1% agarose and 75, 100, 125, 150 or 175 mM NaCl (Saleki, 1993).
Approximately 100-200 seeds suspended in a 1 ml 0.1% agarose solution were plated
onto each of the 6 types of plates. All plates were observed for germination/plant growth

for 2 weeks.

2.3 RESULTS

2. 3. I ISOLATION OF CBF l-OVEREXPRESSING ARABIDOPSIS PLANTS

Plants that overexpress CBF] under the control of the 358 constitutive CaMV

34

promoter were created by transforming RLD plants with pKJOla as described (see
section 2.2.2). Transgenic plants were isolated by selecting for kanamycin resistance on
media containing 50 ug/] kanamycin. T2 generation seeds from independent T] plants
were pooled and screened for CBF] RNA accumulation under nonacclimating conditions
and afier four h of cold-acclimation (data not shown). Initial screening identified two
lines that had increased CBF] RNA accumulation under nonacclimating conditions, A6
and 316. Southern hybridization analysis using DNA from plants in the T3 generation
indicated that A6 had T-DNA inserted at a single locus while B16 contained multiple T-

DNA inserts (see Figure 2.1).

2. 3. 2 O VEREXPRESSION OF CBF 1 IN ARABIDOPSIS RESULTS IN INCREASED COR GENE

EXPRESSION

Examination of the A6 and B16 transgenic lines in the T4 generation indicates that
these lines have increased CBFI RNA accumulation under nonacclimating conditions as
compared to wild type plants (Figure 2.2A). To determine if the increased accumulation
of CBFI RNA resulted in increased COR RNA accumulation, northern hybridization
analysis was performed using COR6. 6, COR15, COR4 7 and COR78 as probes (Figure
2.2A). Overexpression of CBFI clearly results in the activation of COR gene

transcription as increased COR6. 6, C OR15, COR4 7 and COR 78 RNA accumulation

35

HindIII EcoRV

 

 

 

 

Figure 2.]. Southern hybridization of CBFI-overexpressing lines
A6 and 816

Southern hybridization of wild type and transgenic DNA. Southern
transfers of RLD, A6 and B16 T3 generation DNA that had been
digested with HindIII and EcoRV were hybridized at high stringency
using the entire CBF] coding region as described in 2.2.4. HindIII cuts
out the CBF] cDNA fi'agment from the binary vector and EcoRV cuts
once within the CBF] cDNA and once within the binary vector. Lanes
are as follows:

1: RLD;

2: B16;

3: A6.

36

occurs in the CBFI-overexpressing lines as compared to wild type under nonacclimating
conditions. A COR15a-overexpressing line, T8 is also included as a control (Artus et al,
1996) (Figure 2.2A). As would be expected, T8 shows an increase in only COR15 RNA
accumulation and does not show increased CBF 1, COR6. 6, COR4 7 or COR 78 RNA
accumulation under nonacclimating conditions. There appears to be a correlation between
CBF! transgene expression and COR gene expression. The A6 line, which has higher
levels of CBF] RNA accumulation than the B 16 line, also has higher levels of COR RNA
accumulation than the B16 line. The COR gene expression levels in the A6 line appear to
approximate those of 3-day cold-acclimated RLD plants. Overexpression of CBF] has no
effect on eIF 4A, (eukaryotic initiation factor 4A) a constitutively expressed gene that is
not involved with cold acclimation (Metz et al, 1992) and is therefore used as a loading
control.

To determine if increased COR RNA results in increased COR protein
accumulation, immunoblot analyses were conducted with antisera raised to COR15
(Figure 2.2B) and COR 6.6 (data not shown). As was observed for RNA accumulation,
the nonacclimated A6 and B 16 lines have increased protein accumulation as compared to
nonacclimated RLD plants. Consistent with the RNA accumulation patterns,
nonacclimated A6 plants have greater levels of COR protein accumulation than
nonacclimated B16 plants. The COR15 protein accumulation in A6 is similar to that of
seven-day cold acclimated RLD plants while the nonacclimated B16 plants have COR15

protein accumulation similar to that of four-day cold-acclimated RLD plants.

37

A Nonaoclimated Cold-acclimated B 0,5 16
'RLD A6 1316 TT'RLD A6 816 T8 ' @5996) 895168?”

CBF1 .‘ t m ‘l
m... o- —
col-2

COH47 . -

- g

Figure 2.2. Expression of CBFI and COR genes in wild type and
transgenic Arabidopsis plants

  

 

A. CBF] and COR gene RNA accumulation. The amount of CBF] and
COR gene RNA accumulation in leaves of 3-day cold acclimated, and
nonacclimated RLD, CBFI-overexpressing A6 or B16 and COR15a-
overexpressing T8 plants was determined as described in 2.2.4. eIF 4A
(eukaryotic initiation factor 43) is included as a loading control..

B. COR] 5m protein accumulation. The amount of COR15m
accumulation in nonacclimated RLD, A6 and B16 leaves (RLDw,
A6w, 316w) and 4- and 7-day cold-acclimated RLD plants (RLDc4d
and RLDc7d respectively) was determined as described in 2.2.5

38

2. 3. 3 OVEREXPRESSION OF CBF] RESUL IS IN INCRE-tSED FREEZING TOLERANCE AS

DETERMINED BY ELEC TROLYT E LEAKAGE ANALYSIS

Once it was established that overexpression of CBF! results in increased COR
gene expression in the absence of a low temperature stimulus, the next step was to
determine whether CBF overexpression also results in increased freezing tolerance. To do
this, electrolyte leakage assays were conducted (see 2.2.6). By freezing leaves to various
sub-zero temperatures then thawing them, the amount of cellular damage that has
occurred due to freeze-induced membrane lesions can be assayed by measuring the
leakage of electrolytes from the cells with a conductivity meter. Figure 2.3 shows the
results of two individual electrolyte leakage tests. Figure 2.3A and B both show the
graphical representation of the percentage of total electrolytes leaked by nonacclimated
RLD, CBFI-overexpressing (A6 and B16), C OR15a-overexpressing (T8) and 7 anle-
dayCold-acclimated RLD plants. Figure 2.3A shows nonacclimated RLD, A6, B16, and
lO-day cold-acclimated RLD plants from 0 to —8° C . Under nonacclimating conditions,
the A6 line has significantly less leakage than RLD plants fiom —4 to —8° C, and the B16
line has significantly less leakage than RLD plants from —5 to —7° C (P< .001). Figure
2.3B shows nonacclimated RLD, A6, T8, and 7-day cold-acclimated RLD plants from 0
to -8° C. While the nonacclimated A6 line has significantly less leakage than the
nonacclimated RLD plants from —3 to —8° C (P<.003) the T8 line is not significantly
different fiom nonacclimated RLD plants at any of the temperatures tested.

In order to combine the results of multiple electrolyte leakage tests, the

39

u

Percent Electrolyte Leakage
8 5' 8 8

 

 

B 110

gm IA6 -———--— —-k . .
.n... *-   E —-
H -- _ r

701

 

 

 

 

 

50+—~-‘ *—-~ I I __ g-
30 -—--~,*

20 JW T"
' I
l: I .

Percent Electrolyte Leakage
8
i
l
i

 

 

 

 

!. T 1
1°- Bil: is I; .- .
0 -2 -3 -4 -5 -6 -7 -8
Temperature in Degrees Centigrade

Figure 2.3. Electrolyte leakage analysis of wild type and
transgenic Arabidopsis plants.

Leaves from nonacclimated RLD (RLDw), ten-day cold-acclimated
RLD (RLDc) and nonacclimated CBFI-overexpressing A6 and B16
or COR15a-overexpressing plants were frozen at the indicated
temperatures and the extent of cellular damage was estimated by
measuring electrolyte leakage as described in 2.2.6. Error bars
indicate standard deviations of the five replicates of each data point.

A. Electrolyte leakage analysis of ten-day cold-acclimated RLD
(RLDc) and nonacclimated A6, B16, or RLD plants (RLDw) as
described above.

B. Electrolyte leakage analysis of ten-day cold-acclimated RLD

(RLDc) and nonacclimated A6, T8 or RLD plants (RLDw) as
described above.

40

temperature at which 50% leakage occurs (EL50) in all electrolyte leakage tests
conducted was estimated (see Table 2.1). These data indicate that the CBFI-
overexpressing transgenic lines A6 and 816 have a statistically significant increase in
freezing tolerance as compared to RLD plants under nonacclimating conditions. In fact,
B16 has a ~1.3° C increase in freezing tolerance and A6 has a ~3.2° C increase in
freezing tolerance compared to RLD plants under nonacclimating conditions. As seen
previously, the COR15a-overexpressing line T8 does not have a significant increase in
freezing tolerance compared to control plants (Artus et a1, 1996). Interestingly, the ELso
of nonacclimated A6 plants is not significantly different from the ELso of 7-10 day cold-
acclimated RLD plants while both 316 and T8 have an ELso that is significantly different
from that of 7-10 day cold-acclimated RLD plants. These data indicate that
overexpression of one COR gene, COR15, does not result in a significant increase in
freezing tolerance but that overexpression of the battery of COR genes, as seen in the A6

and B16 lines, does result in a significant increase in freezing tolerance.

2. 3. 4 OVEREXPRESSION OF CBF] RESULTS IN INCREASED FREEZING TOLERANCE AS SEEN BY

WHOLE PLANT FREEZE TESTS, BUT NO GROSS PHENOTYPIC CHANGES

Electrolyte leakage tests are an excellent way to approximate the freezing tolerance of
plants. Another biologically significant way to determine freezing tolerance is to freeze
whole plants and score for survival. Pots of nonacclimated RLD, T8, A6, and B16 plants
along with S-to 10- day cold-acclimated RLD plants were frozen to —5° C for 1-3 days,

returned to the 22° C grth chamber and allowed to recover for 1-2 weeks. In all nine

41

Table 2.1. Comparison of Electrolyte leakage 50% (ELso) values of

leaves from wild type and transgenic Arabidopsis plants.

 

 

 

 

 

 

 

 

 

EL50 values
RLDw RLDc A6 B16 T8
RLDW -3.9 :r 0.21 P < 0.0001 P < 0.0001 P = 0.0014 P = 0.7406
(8)
RLDC -7.6 i 0.30 P = 0.326] P < 0.0001 P < 0.0001
(4)
A6 -72 :l: 0.25 P < 0.0001 P < 0.0001
(6)
316 -5.2 1 0.27 P = 0.0044
(5)
T8 -3.8 i 0.35
(3)

 

 

 

 

 

EL50 (the temperature at which 50% leakage occurred) values were
calculated and compared by unbalanced one-way AN OVA using SAS
PROC GLM as described in 2.2.6. EL50 values : SE (n) are presented
on the diagonal line for leaves from nonacclimated RLD (RLDw), 7-
10 day cold-acclimated RLD (RLDc) plants, nonacclimated A6, B16
(CBF! -overexpressing), and T8 (C 0R1 5a-overexpressing) plants. P
values for comparisons of the different EL50 values are indicated in the
corresponding boxes.

42

 

replications of this experiment, the survival of plants from the B16 and T8 lines did not
appear significantly different from that of nonacclimated RLD plants. However, the
survival of A6 plants, while somewhat variable, was always greater than that of
nonacclimated RLD plants. An example of a freeze test is shown in Figure 2.4.
Nonacclimated RLD and A6 plants, and 5-day cold-acclimated RLD plants were frozen
to -5° C for 2 days and then allowed to recover for 7 days at 22° C. The nonacclimated
RLD plants did not survive being frozen whereas the A6 plants and the 5-day cold-
acclimated RLD plants did. This clearly indicates that the overexpression of CBF] results
in increased freezing tolerance in the absence of a low temperature stimulus as compared
to wild type plants. There were no gross phenotypical difl‘erences seen between

transgenic A6 and RLD plants (see Figure 2.4).

2. 3.5 OVEREXPRESSION OF CBF] DOES NOT RES ULTIN INCREASED GERMINATION UNDER

OSMOTIC STRESS

As freezing, drought, and osmotic stress are similar stresses in that they all result
in dehydration (Thomashow, 1999), it was possible that CBFI-overexpressing plants also
had increased tolerance to germinate under conditions of osmotic stress. The addition of
NaCl to plant growth media results in osmotic stress. To test if CBFI-overexpressing
plants have an increased ability to germinate under osmotic stress, T3 seeds from the two
CBFI-overexpressing transgenic lines, A6 and B16 were germinated on media containing
0, 75, 100, 125, 150 or 175 mM NaCl as described in section 2.2.8 (see Table 2.2). Afier

two weeks of observation, no differences in germination frequency could be seen

43

 

 

Figure 2.4. Freezing survival of RLD and CBFI-overexpressing A6
Arabidopsis plants.

Nonacclimated (Warm) RLD and A6 plants and 5-day cold acclimated
(Cold) RLD plants were frozen at —5° C for 2 days and then returned
to a growth chamber at 220 C as described in 2.2.7. A photograph of
the plants afier 7 days of recovery at 22° C is shown. This figure is
shown in color.

between the transgenic lines and the RLD control lines. All three types of seed had 80-
100% germination on NaCl concentrations of 125 mM or less and none of the seeds from

RLD or the CBFI-overexpressing lines germinated at NaCl concentrations of 150 mM or

higher.

2. 3. 6 OVEREXPRESSION OF CBF] RESULTS IN TRANSGENE SILENCING IN ALL LINES TESTED

All previous experiments described with the A6 and B 16 transgenic lines were
conducted on the T3 and T4 generations. However, starting in the T4 generation, a
dramatic increase in the amount of electrolytes leaked between 316 plants in the T3 and
T4 generations was observed (see Figure 2.5A). To determine what was causing the
decrease in freezing tolerance between the T3 and T4 generations, CBF] RNA
accumulation in the T3 and T4 generations of the B16 line was compared (see Figure
2.5B). Nonacclimated RLD and 5-day cold-acclimated RLD plants are also included as
controls. There is less CBF! RNA accumulation in the T4 generation than in the T3
generation of the B16 line, and as would be expected, correlatively less COR15 and
COR4 7 RNA accumulation in the T4 generation of B16 plants (see Figure 2.58). A
reduction in transgene expression was also seen with the CBFI-overexpressing line A6 in
the T5 (see Figure 2.6). To understand better the cause of the reduction, individual
families of A6 T5 plants were assayed (see Figure 2.6A). Sixteen individual A6 plants of
the T4 generation were allowed to self-pollinate and T5 seed was collected separately for
each plant. Members of each family of the individual T5 lines were bulked and tested for

CORISm protein accumulation (see Figure 2.6A) due to difficulties detecting CBF]

45

Table 2.2. Germination of CBFI-overexpressing and wild type
plants on NaCl

 

 

 

 

 

Nam C.°m°°“°"0“ 75 100 125 150 175
(1n mM)

RLD + + + - -

A6 + + + - -

316 + + + - -

 

 

 

 

 

 

 

Seeds from RLD, A6 and B16 in the T3 generation were germinated

the various levels of NaCl indicated above and observed for

germination for two weeks as described in 2.2.8. The percentage of
germination is indicated with a “+” or a “-” symbol. All plates that
had 80 to 100% germination are indicated with a “+” and all plates
that had 0% germination are indicated with a “-”.

46

 

 

 

A :33 1?ch 1.----1

90 J,ISBI6T3 k____
80 . .Bl6T4 _,.7L A-
70 i
60
50
40 1 ,
3o
20 «
10

      

Percent Electrolyte Leakage

 

 

 

 

 

 

Temperature in Degrees Centigrade

B RLD B16
C w T4 T3

 

Figure 2.5. Electrolyte leakage analysis and RNA accumulation of
wild type and 816 plants.

A. Freezing tolerance of leaves fi‘om RLD and B16 transgenic
Arabidopsis plants. Leaves fi'om nonacclimated RLD (RLDw), lO-day
cold-acclimated RLD (RLDc), and nonacclimated 316 T3 and B16 T4
generation plants were fiozen at the indicated temperatures and the
extent of cellular damage was estimated by measuring electrolyte
leakage as described in 2.2.6. Error bars indicate standard deviations of
the five replicates of each data point.

B. CBF] and COR RNA accumulation. CBF] and COR gene RNA
accumulation in leaves of five-day cold acclimated RLD plants (C),
nonacclimated RLD (W) and B16 T3 and B16 T4 generation plants
was determined as described in 2.2.4

47

 

Figure 2.6

 

A6 T5 individual families RLD

It 1.: it *
12 34 567 8910111213141516AW

 

 

 

 

 

 

    

 
     

Percent Electrolyte Leakage
Ox
0
L‘
5
U

 

l 2 3 4 5 6 7
Temperature in Degrees Centigrade

A6 T5 lines

RLD A6 T5 line 154 A6 T5 line 3

 

 

 

 

 

  

 

 

 

123 2354 23456

 

”w

 

 

RLD A6 T5 line 15 A6 T5 line 9

 

1234123456123456

 

F . «-

48

J

Figure 2.6 cont.

E A6 T5 lines

 

RLD 1 3 4 5 9 10 14 15 16

 

 

 

reduced freezing tolerance, but not the loss of the transgene in A6 T5
Lines

A. COR] 5m protein accumulation. Amount of COR] 5m protein
accumulation in 6-day cold-acclimated RLD plants (RLDA),
nonacclimated RLD plants (RLDW) and individual families of A6 T5
plants (A6 T5 1-16). The amounts of COR15m were determined as
described in 2.2.5. A6 T5 families that were firrther investigated are
indicated with an .,

B. Freezing tolerance of leaves from RLD and individual families of
transgenic Arabidopsis plants. Leaves from nonacclimated RLD (RLD)
and selected families of A6 T5 lines (A6 #3, A6 #9, A6 #14, A6 #15)
were frozen at the indicated temperatures and the extent of cellular
damage was estimated by measuring electrolyte leakage as described in
2.2.6. Error bars indicate standard deviations of the five replicates of
each data point.

C. COR] 5m protein accumulation. The amount of COR] 5m protein
accumulated in RLD and selected A6 T5 families (#9, #14, #15) after 7-
days of cold-acclimation was determined for 2 individually isolated sets
of protein as described in 2.2.5

D. COR15m protein accumulation. Amount of COR] 5m protein
accumulated in nonacclimated leaves of four individual RLD plants and
six individual plants of selected A6 T5 families, #3, #9, #14 and #15 as
described in 2.2.5

B. Presence of the CBFI-transgene in selected A6 T5 families. Genomic
DNA was isolated from individual families of A6 T5 plants (#1, #3, #4,
#5, #9, #10, #14, #15, #16) and probed for the presence ofthe CBFI-
transgene as described in 2.2.3. The HindIII- CBF I containing fragment
is shown is shown in the upper panel. A portion of the ethidium bromide
stained agarose gel containing genomic DNA to which the CBFI probe
was hybridized is in the panel below

49

RNA accumulation (data not shown). COR] 5m protein accumulation varied greatly
between the families, some of which had little or no expression (family # 2, 6, 7, 8, 9, 15
and 16) while others had intermediate or high levels of expression (family # l, 3, 4, 5, 10,
11, l2, l3, and 14) as compared to nonacclimated RLD plants. Six-day cold acclimated
RLD plants are also included as a control. To determine if there was a correlation
between the low levels of COR] 5m protein accumulation and freezing tolerance,
electrolyte leakage assays were conducted on two lines with low COR] 5m protein
accumulation, #3 and #14, and two lines with high levels of COR] Sm protein
accumulation, #9 and #15 (Figure 2.63). There is a clear correlation between freezing
tolerance and COR] 5m protein accumulation. The families with little or no COR] 5m
protein accumulation, (#3 and #14) have similar electrolyte leakage percentages to
nonacclimated RLD plants while the families with high levels of COR] 5m protein
accumulation (#9 and #15) have lower electrolyte leakage percentages than
nonacclimated RLD plants.

There were four possible reasons for the probable loss of CBFI-overexpression.
1) Co-suppression of the endogenous CBF] gene by the CBF] transgene had occurred
resulting in a total loss of CBF] expression; 2) The individual plants in the lines with low
expression had all independently lost CBFI-induced COR15 expression due to transgene
silencing; 3) The individual plants had variable levels of transgene expression, but the
majority of individual plants had lost expression due to transgene silencing; or 4) due to
mislabeling of seed material, the CBF] transgene was not present in those lines.

To determine which hypothesis was correct, lines with little or no COR] 5m

protein accumulation under nonacclimating conditions were acclimated and assayed for

50

COR] 5m accumulation. Ifco-suppression had occurred, COR] 5m protein accumulation
should not be detected under acclimating conditions, assuming that CBF] was the only
activator of COR gene expression under acclimating conditions (the existence of CBF 2
and CBF 3 was not known at the time the experiment was conducted). There was no
reduction in COR] 5m protein accumulation in the A6 T5 generation transgenic lines #9,
#14 and #15 compared to RLD plants acclimated for seven days suggesting that
cosupression had not occurred (see Figure 2.6C).

To assay for the expression levels of individual plants, six individual plants from
four selected lines, two with high COR15m protein accumulation (#3 and #14), two with
low COR] 5m protein accumulation (#9 and #15), and four individual nonacclimated
RLD plants were assayed for COR15m protein accumulation (see Figure 2.6D). Little or
no COR15m protein accumulation is seen in individual plants fiom the two lines with
low expression (#9 and #15), while variable but significant COR] 5m protein
accumulation is seen in the two lines with high expression (#3 and #14) compared to the
individual nonacclimated RLD plants. These data indicate that the expression level of
COR15m protein was consistent within families, but not across the families.

To ensure that the low level of COR] 5m protein accumulation in lines with low
expression was not due to loss of the CBF] transgene, Southern hybridization analysis
was conducted (Figure 2.6E). The HindIII fragment of the binary plasmid which contains
the CBFI-transgene was present in all nine A6 T5 families tested, and not present in non-
transformed RLD plants. Variations in the intensity of the banding pattern are due to
uneven loading. Combined, these data strongly suggest that loss of COR] 5m protein

accumulation was due to transgene silencing and not due to the loss of the CBF]

51

transgene or cosuppression of all copies of the CBF] gene.

Two additional lines, showing increased CBF] RNA, or COR15 RNA
accumulation under nonacclimating conditions, were isolated, K16 (see Figure 2.7A) and
1-1] (see Figure 2.7B) respectively. In electrolyte leakage tests, T2 plants fiom the K16
line appeared to have equal or less leakage than 5-day cold-acclimated RLD plants
(Figure 2.8A). However, both of these lines exhibited transgene silencing in the T3
generation. Figure 2.8B shows the results of an electrolyte leakage assay on K16 plants in
the T3 generation. There is little difference between K16 plants and nonacclimated RLD

plants indicating that transgene silencing has most likely occurred.

2. 3. 7 ANTISENSE CBF] PLANTS DO NOTHA VE REDUCED CBF] OR COR GENE EXPRESSION

One way to determine the effects of a gene is to delete it from the genome and test for
any changes in phenotype. It was of interest to determine if there were any changes in the
phenotype of plants that had total loss of CBF] gene expression. To test for this, plants
were transformed with the antisense orientation of the CBF I gene under the control of the
constitutive CaMV 358 promoter (see section 2.2.2) as overexpression of a gene in the
antisense orientation can result in the loss of expression of the endogenous gene (van der
Krol et al, 1990). A total of 15 independent transgenic lines in the T2 generation were
screened for CBF! RNA accumulation under nonacclimating conditions, and COR RNA
accumulation under cold-acclimating conditions using double stranded DNA probes.
Figure 2.9A and B are examples of typical data. While there was increased CBF] RNA

accumulation under nonacclimating conditions as determined by using a double stranded

52

RLDW K16 RLDC RLDW l-ll

 

 

CBF] COR15

   

Figure 2.7. CBF] and COR15 RNA accumulation in CBFI-
overexpreesing plants

CBF] or COR15a RNA accumulation in leaves of 5-day cold-
acclirnated RLD (RLDC), nonacclimated (RLDW), K16 and 1-1]
(CBFI-overexpressing) T2 generation plants was determined as
described in 2.2.4.

A. CBFI RNA accumulation. CBF] RNA accumulation in
nonacclimated RLD and T2 generation K16.

B. COR15 RNA accumulation. COR15 RNA accumulation in 5-day
cold acclimated RLD, nonacclimated RLD and T2 generation 1-11.

53

A 110 LA
100 F-ng’ L.,,,,,L_
‘ JIRLDc F

90 1" 51(161‘2‘7 )1
$3 Haney - L -

 

    
   
    

   

 

Percent Electrolyte Leakage

 

Percent Electrolyte Leakage
as
o

 

 

 

Temperature in Degrees Centigrade

Figure 2.8. Electrolyte leakage analysis of wild type and transgenic
K16 plants.

Leaves from nonacclimated RLD, cold-acclimated RLD K16 T2 and T3
generation plants were frozen at the indicated temperatures and the extent
of cellular damage was estimated by measuring electrolyte leakage as
described in 2.2.6. Error bars indicate standard deviations of the five
replicates of each data point.

A. Electrolyte leakage analysis using five-day cold-acclimated RLD
plants (RLDc) and nonacclimated K16 T2 (K16 T2) or RLD plants
(RLDW) as described above.

B. Electrolyte leakage using analysis four-day cold-acclimated RLD
plants (RLDc) and nonacclimated K16 T3 (K16 T3) or RLD plants
(RLDW) as described above.

54

A RLD 2-7 2-11 2-14 2-17 2-20

H. .

CBF] I , .. I. 8 53* .]

 

 

 

 

 

 

RLD 2-12
B N A N A
COR15
C .
123456789101112131415161718
[ TE; " 4...... ' " “It" "‘5 “m3

 

Figure 2.9. CBF] and COR gene accumulation in antisense-CBF]
transgenic lines

CBF] and COR15 RNA accumulation. The amount of CBF] and COR15
RNA accumulation in leaves of nonacclimated (N) and acclimated (A)

RLD and antisense (AS) CBF] -overexpressing lines was determined as
described in 2.2.4.

A. CBF] RNA accumulation. CBF] RNA accumulation in nonacclimated
RLD and 5 independent lines overexpressing AS CBF] in the T2
generation using a double stranded probe. The AS CBF] transcript is
approximately the same size as the endogenous CBF] transcript.

B. COR15 RNA accumulation. COR15 RNA accumulation in
nonacclimated and 7-day cold acclimated RLD and 1 independent line
overexpressing AS CBF] in the T3 generation.

C. COR15m protein accumulation. The amount of COR] 5m Protein
accumulation in nonacclimated RLD (RLDW) and 7-day cold acclimated
RLD (RLDC)and 17 independent lines overexpressing AS CBF] in the T2
generation was determined as described in 2.2.5. Lanes are as described
below:

]: RLDW 5: 2-18.2 9: 2-20.1 13: 2-22.3 17: 2-24.3

2: RLDC 6: 2-18.3 10: 2-20.2 14: 2-22.4 18: 2-24.4

3: 2-17.2 7: 2-19.l 11: 2-22.] 15: 2-22.5

4: 2-18.1 8: 2-19.2 12: 2-22.2 l6: 2-24.1

55

CBF] cDNA probe Figure 2.9A), there were no dramatic differences seen between the
amount of COR15 RNA accumulation between antisense CBF] plants and RLD plants
under acclimating conditions (Figure 2.9B). In fact, there appears to be a slight increase
in COR15 accumulation in antisense-CBF] line 2-12 as compared to control plants (see
Figure 2.9B). The reason for this apparent increase is not known.

An additional 32 T2 generation anti-sense CBF] lines were screened by assaying
for COR] 5m protein accumulation under cold-acclimating condition, 17 of which are
shown in Figure 2.90 Again, no dramatic differences were seen between transgenic and
RLD plants under acclimating conditions. Some of the same lines were again screened in
the T3 generation for COR] 5m protein accumulation and identical results were seen (data
not shown). These data indicate that either the endogenous CBF] had not been
cosuppressed, or that cosuppression of CBF] does not result in a reduction of COR gene

expression under acclimating conditions.

2.4 DISCUSSION

CBF] was previously shown to bind to the CRT/DRE and activate transcription of a
reporter gene in yeast (Stockinger, 1997). Here I present data indicating that CBF 1 can
also bind to the CRT/DRE and activate transcription of the COR genes in plants since
overexpression of CBF] in Arabidopsis resulted in increased COR gene expression under
nonacclimating conditions (see Figure 2.] and Figure 2.2). Interestingly, increased
activation of COR genes resulted in increased freezing tolerance as seen by both

electrolyte leakage tests (see Figure 2.3, Figure 2.5A, Figure 2.6B and Figure 2.8A and

56

Table 2.1) and whole plant freeze tests (see Figure 2.4). These results strengthen the
argument that the COR proteins play additive roles in increasing freezing tolerance
during cold acclimation, as was previously suggested by overexpression of a single COR
protein, COR15a (Artus et al, 1996).

The increase in CBF] RNA accumulation in the absence of a low temperature
stimulus seen in CBF-overexpressing lines (see Figure 2.2) may provide clues as to the
mechanisms of CBF] regulation. If CBF] was regulated by the rapid degradation of RNA
under nonacclimating conditions, then it would expected that ectopic overexpression of
CBF] under a constitutive promoter would result in low levels of CBF] RNA under
nonacclimating conditions and very high levels under cold-acclimating conditions. As
this was not the case, [there is a approximately two fold increase in CBF] RNA under
acclimating conditions as compared to non-acclimating conditions (see Figure 2.2)], it
appears that CBFI is not regulated by RNA degradation. However, the possibility cannot
currently be ruled out that overexpression of CBF] results in an increase CBFI transcript
to the extent that the degradation enzymes are “overwhelmed” under nonacclimating
conditions and cannot degrade all the transcripts.

It is also interesting to note that, while under both nonacclimating and acclimating
conditions, the A6 line appears to have greater amounts of CBFI RNA accumulation than
that of cold acclimated RLD plants, the amount of COR transcript does not appear to be
substantially greater (see Figure 2.2). There are several possible explanations for this
observation. Firstly, the CBFI-transgene RNA may not be efficiently translated. To
determine whether this is correct, CBF 1 protein accumulation levels would need to be

assayed in CBFl-overexpressing plants under both nonacclimating and acclimating

57

conditions and compared to control lines under the same conditions. However, despite
repeated attempts, the CBF] protein was not detected (See Chapter 3; K Jaglo, S.
Gilmour and M. Thomashow, unpublished) making it impossible to definitively answer
this question.

Secondly, the CBF] RNA may be efficiently translated into protein, but the
protein does not bind and/or recruit cofactors as efficiently under nonacclimating
conditions. Thirdly, under nonacclimating conditions, CBF 1 may have a structural
conformation such that it interacts with a repressor or some other molecule that
sequesters it from binding to the CRT/DRE. It is now known that under acclimating
conditions, the CBF] protein denatures and elongates, a rare phenomenon for a
monomeric protein under physiological conditions (Kanaya et al, 1999). This change in
conformation may result in more eflicient binding, or more efficient recruiting of co-
factors under acclimating conditions. We have evidence that CBF] is dependent on the
ADA2, ADA3 and GCNS cofactors for transcription activation in yeast (Stockinger et al,
1997; E. Stockinger, Y. Mao, S. Triezenberg and M. Thomashow, unpublished). If CBF]
is also dependent on these cofactors in Arabidopsis, then this dependence could support
the second hypothesis that the increased abundance of CBFI RNA in the A6 line under
nonacclimating and acclimating conditions does not result in a massive increase in COR
RNA due to inefficient binding or inefficient recruiting of cofactors. Another possibility
is that hypothesized by Kanaya et al (1999): the change in conformation under
acclimating conditions may allow CBF] to be released from a negative repressor. In this
case, overexpression of the CBF] protein would result in a positive change in the ratio

between the CBF] protein and repressor allowing for transcription activation under

58

nonacclimating conditions. Again it would be expected that activation would not be as

eficient under nonacclimating conditions since some protein would still be inactive and
bound to the repressor. However, until we can detect the CBF] protein, we will not be
able to determine which hypothesis, if any, is correct.

The final possibility is that the CBF] protein may be modified under acclimating
conditions. Many transcriptional activators are activated or deactivated by
phosphorylation, methylation or the addition of other small molecules in response to the
correct stimulus (Gallic, 1993; Schwechheirner et al, 1998). The CBF 1 protein contains
sites that are similar to mitogen activated protein (MAP) kinase phosphorylation sites (E.
Stockinger and M. Thomashow, unpublished) suggesting that the protein may be
phosphorylated under either acclimating conditions to activate the protein or under
nonacclimating conditions to deactivate the protein. Given the high level of CBF RNA
accumulation and relatively low level of COR RNA accumulation in A6 plants as
compared to RLD plants, activation by phosphorylation seems more likely. Mitogen
activated protein kinases are known to be activated at low temperatures in alfalfa (J onak
et al, 1996). Conceivably, homologues of these kinases could firnction to phosphorylate
CBF] in Arabidopsis and activate it under acclimating conditions. However, until the
protein can be detected, whether or not CBF 1 is modified by phosphorylation or some
other small molecule will remain unknown.

As stated above, overexpression of CBFI results in increased COR gene
expression under nonacclimating conditions. This correlative increase in both CBF and
COR gene expression, along with the gel shifi data indicating that CBF] can bind to the

CRT/DRE, and the data that CBF] can activate a reporter gene under the control of the

59

COR15a promoter in yeast (Stockinger et al, 1998) are a strong indicator that CBF] is a
regulator of COR gene expression. This is significant as it is the first transcription factor
isolated that acts upstream of proteins known to be associated with cold acclimation, the
COR proteins.

It is now known that CBF] is a member of a small gene family consisting of
CBF 1, CBF 2 and CBF 3 which are approximately 90% similar in amino acid sequence
and all of which to bind to the CRT/DRE and activate transcription of a reporter gene in
yeast (Gilmour et al, 1998). Overexpression of CBF 3 in Arabidopsis results in increased
COR gene expression under both nonacclimating and acclimating conditions supporting
the conclusion that all three CBF proteins are likely regulators of COR gene expression
(Gihnour, 2000; S. Gihnour, M. Salazar and M. Thomashow, unpublished). Interestingly,
CBF3-overexpressing plants have increased levels of proline and total soluble sugars
compared to control lines, as well as increased transcript levels of P5CS which is
involved in the production of free proline (Gilmour et al, 2000). CBF3-overexpressing
plants also have changes in membrane lipids which are associated with some of the
changes that occur during cold acclimation. Combined, these data indicate that CBF3
activates multiple pathways involved with cold acclimation and suggest that the CBF
genes may play critical roles in inducing the many changes in gene expression associated
with cold acclimation.

However, to determine definitively if all three CBF proteins are regulators of
COR gene expression, all three CBF genes would need to be deleted and COR gene
expression monitored. Ifdeletion of all three CBF genes resulted in a loss of COR gene

expression, this would be proof that the CBF proteins are the activators of COR gene

60

expression under acclimating conditions. By adding back each CBF individually,
differences in activation between the CBF proteins could be investigated. However, it is
possible that deletion of all three CBF genes may not have an effect on COR gene
expression. This could either be an indication that the CBF proteins are not the natural
regulators of COR genes, or that there is redundancy in COR gene activation. The second
possibility would not be unexpected as the ABA induced activation of the COR genes is
independent of cold and drought induced activation (Baker et a], 1994; Yamaguchi-
Shinozaki and Shinozaki, 1994; Wang et al, 1995) indicating that several independent
pathways may converge in the induction of the COR genes. The exact role of each CBF
and differences in the genes they activate will remain unknown until triple CBF knockout
lines are isolated.

While it has long been hypothesized that the COR genes play direct roles in
increasing fi'eezing tolerance, direct evidence has been lacking. Previous work has shown
that overexpression of COR15a results in increased freezing tolerance of chloroplasts, but
not of whole leaves (Artus et al, 1996, Figure 2.3B). The slight but significant increase in
fi'eezing tolerance brought about by overexpression of one COR protein raised the
intriguing possibility that overexpression of the battery of COR proteins would increase
fieezing tolerance more dramatically. Here I present evidence that the COR genes play a
role in increasing freezing tolerance since overexpression of the battery of COR genes
results in increased freezing tolerance without a low temperature stimulus (see Figure 2.3,
Figure 2.4, Figure 2.5A, Figure 2.6B, Figure 2.8A, and Table 2.1). These data are
significant as they are strong evidence that overexpression of all the CBFl-induced COR

genes results in increased fieezing tolerance that can be observed at the level of whole

61

leaves, and for the A6 line, whole plants.

There appears to be a positive correlation between increased fi'eezing tolerance
and COR gene expression. The A6 line, which has higher levels of COR gene expression
than the B16 line in the T4 generation (see Figure 2.2), also has a greater increase in
freezing tolerance than the B 16 line. As determined by ELso values in electrolyte leakage
assays, the A6 line has a ~3 .2° C increase in freezing tolerance while the B16 line only
has a ~1.3°C increase in freezing tolerance (see Table 2.1). Furthermore, the ELso value
of the A6 line was not significantly different from 7-10 day cold-acclimated RLD plants
(see Table 2.1). Perhaps the A6 line was the only line to consistently survive whole plant
freeze tests because it exhibits a similar increase in freezing tolerance to acclimated wild
type plants (see Figure 2.4). When 5-10 day cold-acclimated RLD plants, nonacclimated
CBFI-overexpressing A6, B 16, K16 and COR15a-overexpressing T8 plants were frozen
to -5° for varying amounts of time, only cold-acclimated RLD plants and plants in the A6
line consistently showed increased survival. The similarity in COR gene expression levels
of the A6 line and cold-acclimated RLD plants and the ability of both types of plants to
survive freezing are a good indication that COR protein accumulation is a critical factor
involved in increasing freezing tolerance during cold acclimation.

Despite the increase in freezing tolerance as seen by electrolyte leakage assays,
CBFI-overexpressing plants from the A6, and B16 lines in the T3 generation did not
appear to have increased germination under osmotic stress. Due to the similarities
between drought, cold and osmotic stress, (Thomashow, 1999) it was possible that CBFI-
overexpressing plants might have increased tolerance to osmotic stress. I tested this

hypothesis by doing germination assays on media containing differing levels of sodium

62

chloride. However, no differences were seen between the CBF-overexpressing lines and
the control RLD plants as none of the plants germinated in sodium chloride levels of
ISOmM or higher. There are several possibilities as to why there were no differences seen
between RLD plants and the CBFI-overexpressing lines. The experiment should be
repeated with SmM differences in salt concentrations from 125 — 150 mM to see if there
are small differences between CBFI-overexpressing and control RLD plants in the ability
to germinate on sodium chloride. Additionally, there are different genes involved
between tolerance to germinate under osmotic stress and ability to grow in osmotically
stressed conditions (Saleki, 1993). In our case, it could be that CBFI-overexpressing
plants do not have an increased ability to germinate under osmotic stress, but have
increased ability to grow under osmotic stress, as whether the COR genes are expressed
in seeds is not known. To test this hypothesis, control and CBFI-overexpressing plants
should be germinated under nonstressed conditions, then transferred to plates containing
NaCl or mannitol, to test for differences in ability to grow under osmotic stress.

While overexpression of CBF] initially led to increased COR gene accumulation,
several generations of self-pollination led to a reduction in COR gene expression in
CBFI-overexpressing plants. The B16 plants in the T4 generation did not appear to have
as dramatic an increase in freezing tolerance as B16 plants in the T3 generation (see
Figure 2.5). This could be seen by both electrolyte leakage assays and RNA accumulation
of CBF] and the COR genes. Similar patterns of reduced expression were noticed later
with the A6 line in the T5 generation (see Figure 2.6) and the K16 line in the T3
generation (see Figure 2.8).

To determine what was causing the decrease in COR gene expression, individual

63

families of A6 T5 lines were isolated and checked for COR15m protein accumulation.
Some of the lines had high levels of COR] 5m protein accumulation while others had

little or no COR15m protein accumulation as compared to RLD plants under
nonacclimating conditions (see Figure 2.6A). As the loss of COR] 5m protein
accumulation could be an indication of either CBF] co-suppression, CBF] transgene
silencing, or the physical loss of the CBF] transgene, experiments were conducted to
determine which hypothesis was correct. Control RLD plants and the individual A6 T5
generation lines were cold-acclimated for seven days and COR] 5m protein accumulation
was investigated (see Figure 2.6C). As COR] 5m protein accumulation did not appear to
be significantly different between RLD plants and any of the individual families of the

A6 T5 generation lines, I concluded that the endogenous copy of the CBF] gene was still
firnctional and that co-suppression had not occurred (Figure 2.6C). However, the
possibility that the COR gene expression detected was due to redundancy of activation by
the other two CBF genes (of which I was not aware at the time) cannot currently be ruled
out.

Electrolyte leakage assays on four selected A6 TS lines indicated a positive
correlation between COR gene expression and freezing tolerance (see Figure 2.6B). The
lines with little or no COR15m protein accumulation (#15 and #9) had less of an increase
in freezing tolerance as compared to lines with high levels of expression (#14 and #3). To
determine if the amount of COR] 5m protein accumulation was identical in all the plants
from each individual A6 T5 family, individual plants from the lines with low and high
levels of COR] 5m protein accumulation were assayed for COR] 5m protein accumulation

(see Figure 2.6D). Relatively consistent patterns of COR] 5m protein accumulation were

64

seen. In the two lines with high levels of expression, (#14 and #3 ), the majority of the six
individual plants investigated had greater levels of COR] 5m protein accumulation than
the control RLD plants under nonacclimating conditions. In the two lines with low levels
of expression (#15 and #9), virtually no COR15m protein accumulation was detected.
These data indicate that individual plants within A6 T5 families were experiencing loss
of COR] 5m protein accumulation presumably as a result of loss of CBFI-transgene
expression, although to know definitively, more individual plants fiom each family
would need to be tested for COR] 5m accumulation. The families with high expression
had fewer individuals with low expression (lines #14 and #3) whereas families with low
levels of expression appeared to have all lost CBFI-transgene expression (lines #15 and
#9).

To ensure that the loss of expression seen in some of the individual A6 T5 families
was not due to loss of the transgene, Southern hybridization analysis was conducted (see
Figure 2.6E). The Hind III —CBF1 containing fragment was present in all of the A6 T5
families tested, and not in the RLD plants, regardless of the amount of COR] 5m protein
accumulation. These data clearly indicate that the loss of expression was not due to the
loss of the transgene, but was most likely due to transgene silencing.

Overexpression of a given gene in the antisense orientation can result in
the loss of expression of the endogenous gene (van der Krol et al, 1990). Based on this
knowledge, I transformed plants with antisense 3SS:CBF1, pKJOZa (see section 2.2.2) in
the hopes of reducing or eliminating CBF] gene expression. A total of 47 independent
lines were investigated directly and indirectly for indications that endogenous CBF] gene

expression had been reduced or lost. In the T2 generation, CBF] RNA accumulation

65

under nonacclimating conditions was assayed in the transgenic lines to determine if the
transgene was being expressed (Figure 2.9A). As seen inFigure 2.9A, some of the
antisense-CBF] lines have an increase in CBF] RNA accumulation indicating that the
transgene was expressed. In the T3 generation, COR15 RNA accumulation was
investigated under both nonacclimating and acclimating conditions (Figure 2.9B), but no
reduction in COR15 RNA was seen in the antisense-CBF] line. Lastly, 32 lines were
screened for COR15m protein accumulation under cold-acclimating conditions (Figure
2.9C). No dramatic differences were seen between the control and transgenic plants. It is
important to note that to check for expression of antisense-CBFI RNA, total plant RNA
was probed with a double stranded CBF] DNA probe, which hybridizes to both sense
and antisense CBF] RNA. Given that no increase in CBF] RNA was seen in control
plants, it was assumed that the increase in CBF] RNA accumulation was due to antisense
CBF] RNA. The possibility that the increase in RNA could have been due to an increase
in sense-CBF], therefore, cannot be ruled out.

There are several possible reasons why a decrease in COR gene expression was not
observed in the antisense lines under acclimating conditions. Firstly, CBF] is part of a
small gene family consisting of three members (Gilmour et al, 1998). While the genes are
90% similar at the amino acid level, they are only approximately 70% identical in nucleic
acid sequence which may be enough difference that the antisense-CBF] construct did not
effect the expression of the endogenous CBF 2 and CBF 3 genes. It is entirely possible that
the antisense plants did cause a reduction in CBF] RNA accumulation, but did not efl‘ect
CBF2 or CBF3 RNA accumulation. With two other copies of unaffected CBF genes, a

reduction in COR gene expression may not occur, as the expression of CBF 2 and CBF 3

66

may have compensated for the reduction or loss of CBF] expression. Secondly, it is
possible that the antisense-CBF] construct, pKJOZa, resulted in total loss of all three
CBF genes which was lethal, thus making it impossible to isolate knockout plants.
Thirdly, it is possible that CBF] knockout plants could be isolated using the described
construct, but not enough lines were screened to isolate these lines. Lastly, it is possible
that the antisense-CBF] may not have effected the expression of the endogenous CBF]
gene. Waterhouse et al, (1998) have done elegant experiments indicating that
overexpression of the antisense orientation of a gene does not always effect gene
expression. However, expression of a short sequence of the sense orientation of the gene
connected to the antisense orientation results in significant reductions of endogenous
gene expression. It is possible that if such constructs were transformed into Arabidopsis
plants, CBF gene expression could be reduced or eliminated. If so, studies to determine
the COR gene expression and freezing tolerance of these lines could aid in our
understanding of the roles of the CBF genes and the COR genes in cold-acclimation.

In summary, it has been determined that CBF] is an activator of the COR genes
and is therefore likely to be one of the early activators involved in cold acclimation.
While it has not been determined with absolute certainty that CBF] is one of the
transcription factor(s) required for activation of the COR genes under acclimating
conditions, overexpression of CBF] does result in increased COR gene expression
without a low temperature stimulus. Additionally, overexpression of CBF] and the COR
genes results in increased freezing tolerance as seen by electrolyte leakage assays and
whole plant fi'eeze tests. These data clearly indicate that CBFI-induced genes play a role

in increasing freezing tolerance and give more direct evidence to the theory that the COR

67

genes are involved in increasing freezing tolerance. However, until the efi‘ects of
blocking all COR gene expression are known, direct evidence of their specific roles will
still be lacking.

The potential long term applications of the effects of CBFI-overexpression are
exciting. Traditional plant breeders have long sought to increase freezing tolerance as, for
example, the most freezing tolerant wheat varieties today are essentially identical to those
from the turn of the century (Thomashow, 1999). If the CBF-induced signaling cascade is
present in other plant species, and searches of GenBank (http: //www.ncbi.nim.nih. gov:
80/BLAST) show sequence similarities in numerous other plants species (see also
Chapter 4), then manipulation of this family of transcription factors could be the key to

increasing freezing tolerance in agronomically important species.

2. 5 REFERENCES

Artus, NN, Uemura, M, Steponkus, PL, Gilmour, SJ, Lin, C, Thomashow, MF. 1996.
Constitutive expression of the cold-regulated Arabidopsis thaliana COR15a gene affects
both chloroplast and protoplast freezing tolerance. Proc. Natl. Acad. Sci. 93: 13404-
13409.

Baker, SS, Wilhelm, KS, Thomashow, MF. 1994. The 5’ region of Arabidopsis thaliana
COR15a has cis-acting elements that confer cold-, drought- and ABA-regulated gene
expression. Plant Mol Biol. 24: 701-713.

Bechtold, N, Ellis, G, Pelletier, CR. In-planta agrobacterium-mediated gene-transfer by
infiltration of adult Arabidopsis-thaliana plants 1993. Acad Sci. Ser. 111 Life Sci. 316:
1194-1203.

68

Blake, MS, Johnston, KH, Russell-Jones, GT, Gotschlich, EC. 1984. A rapid, sensitive
method for detection of alkaline-phosphatase conjugated anti-antibody on western blots.
Anal. Biochem. 136: 175-179.

Boyer, J S. 1982. Plant productivity and environment. Science. 218: 443-448.

Close, TJ. 1997 . Dehydrins: A commonality in the response of plants to dehydration and
low temperature. Physiol. Plant. 100: 291-296.

Dure 111, L, Crouch, M, Harada, J, Ho, THI), Mundy, J. et al. 1989. Common amino-acid
sequence domains among the LEA proteins of higher-plants Plant Mol Biol. 12: 475—486.

Dure 111, L. 1993. Plant Responses to Cellular Dehydration During Environmental
Stress, ed. By T] Close, EA Bray, Am. Soc. of Plant Physiol. Rockville: 91-103.

Gallie, DR. 1993. Posttranscriptional regulation of gene-expression in plants Annu. Rev.
Plant Physiol. Plant Mol. Bio]. 44: 77- 1 05.

Gihnour, SJ, Artus, NN, Thomashow, MF. 1991. cDNA sequence-analysis and
expression of 2 cold-regulated genes of Arabidopsis-thaliana. Plant Mol Biol. 18: 13-21.

Gilmour, SJ, Hajela, RK, Thomashow, MF . 1988. Cold-acclimation in Arabidopsis-
thaliana. Plant Physiol. 87: 745-750.

Gilmour, SJ, Sebolt, AM, Salazar, MP, Everard, JD, Thomashow, MF. 2000.
Overexpression of the Arabidopsis CBF 3 transcriptional activator mimics multiple
biochemical changes associated with cold acclimation. Plant Physiol. In press.

Gilmour, SJ, Thomashow, MF. 1991. Cold-acclimation and cold-regulated gene-
expression in ABA mutants of Arabidopsis-thaliana. Plant Mol Biol. 17 : 1233-1240.

Gilmour, SJ, Zarka, DG, Stockinger, EJ, Salazar, MP, Houghton, JM, Thomashow, MF.
1998. Low temperature regulation of the Arabidopsis CBF family of AP2 transcriptional
activators as an early step in cold-induced COR gene expression. Plant J. 16: 433-442.

Guy, CL, Niemi, KJ, Brambl, R. 1985. Altered gene-expression during cold-acclimation
of spinach. Proc. Natl. Acad. Sci. USA. 82: 3673-3677.

Hajela, RK, Horvath, DP, Gihnour, SJ, Thomashow, MF. 1990. Molecular-cloning and
expression of COR (cold-regulated) genes in Arabidopsis-thaliana. Plant Physiol. 93:
1246-1252.

Hahn, S. 1993. Structure(questionable) and function of acidic transcription activators.
Cell. 72: 481-483.

69

Jonak, C, Kiegerl, S, Ligterink, W, Barker, PJ, Huskisson, NS Hirt, H. 1996. Stress
signaling in plants: A mitogen-activated protein kinase pathway is activated by cold and
drought. Proc. Nat]. Acad. Sci. USA. 93: 11274-11279.

Kanaya, E, Nakajima, N, Morikawa, K, Okada, K, Shirnura, Y. 1999. Characterization of
the transcriptional activator CBFI from Arabidopsis thaliana - Evidence for cold
denaturation in regions outside of the DNA binding domain. J. Biol. Chem. 274: 16068-
16076.

Levitt, J. 1980. Responses of Plants to Environmental Stresses. Chilling, fieezing, and
high temperature stresses, Ed 2. Academic Press, New York.

Lin C, Guo, WW, Everson, E, Thomashow, MF. 1990. Cold-acclimation in Arabidopsis
and wheat - a response associated with expression of related genes encoding boiling-
stable polypeptides. Plant Physiol. 94: 1078-1083.

Liu Q, Kasuga, M, Sakuma, Y, Abe, H, Miura S, et al. 1998. Two transcription factors,
DREBI and DREBZ, with an EREBP/AP2 DNA binding domain separate two cellular
signal transduction pathways in drought- and low-temperature-responsive gene
expression, respectively, in Arabidopsis. Plant Cell. 1391-13406.

Metz, AM, Tirnmer, RT, Browning, KS. 1992. Sequences for 2 cDNAs encoding
Arabidopsis-thaliana eukaryotic protein-synthesis initiation factor-4a. Gene. 120: 313—
314.

Nordin, K, Heino, P, Palva, ED. 1991. Separate signal pathways regulate the expression
of a low-temperature-induced gene in Arabidopsis-thaliana (L) Heynh. Plant Mol. Biol.
16: 1061-1071.

Nordin, K, Vahaloa, T, Palva, ET. 1993. Differential expression of 2 related, low-
temperature-induced genes in Arabidopsis-thaliana (L) Heynh. Plant Mol Bio]. 21: 641-
653.

Ohme-Takagi, M, Shinshi, C. 1995. Ethylene-inducible DNA-binding proteins that
interact with an ethylene-responsive element. Plant Cell. 7: 173-182.

Raikhel, N. 1992. Nuclear targeting in plants. Plant Physiol. 100: 1627-1632.

Rogers, SO, Bendich, AJ. 1988. Plant Molecular Biology Mannua] (Gelvin, SB and
Schiperoot, RA, eds). Dordecht: Kluwer Academic Publishers, pp. A6: 1-10.

Rothstein, SJ, Kahners, KN, Lotstein, RJ, Carozzi, NB, Jayne, SM, Rice, DA. 1987.

Synthesis and secretion of wheat alpha-amylase in Saccharomyces-cerevisiae. Gene. 53:
153-161.

70

Saleki R, Young PG, Le Febvre DD. 1993. Mutants of Arabidopsis thaliana capable of
germination under saline conditions. Plant Phys. 101: 83 9-845.

Sambrook, J, Fritsch, EF, Maniatis, T. 1989. Molecular Cloning: A Laboratory Manual
(Cold Spring Harbor Lab. Press, Plainview, NY), 2“d Ed.

Schagger, H and von Jagow, G. 1987. Tricine sodium dodecyl-sulfate polyacrylanride-gel
electrophoresis for the separation of proteins in the range from l-kDa to 100-kDa. Anal.
Biochem. 166: 368-379.

Schwechheirner, C, Zourelidou, M, Bevan, MW. 1998. Plant transcription factor studies.
Annu. Rev. Plant Physiol. Plant Mo]. Biol. 49: 127-150.

Steponkus, PL. 1984. Role of the plasma-membrane in freezing-injury and cold-
acclirnation. Annu. Rev. Plant. Physiol. 35: 543-585.

Steponkus, PL, Uemura, Joseph, RA, Gilmour, SG, Thomashow, MF. 1998. Mode of
action of the COR15a gene on the freezing tolerance of Arabidopsis thaliana. Proc. Natl.
Acad Sci. USA. 95: 14570-14575.

Steponkus, PL, Uemura, M, Webb, MS. 1993. Membrane destabilization during freeze-
induced dehydration. Curr. Topics Plant Physiol. 10: 37-47.

Stockinger, EJ, Gilmour, SJ, Thomashow, MF. 1997. Arabidopsis thaliana CBF]
encodes an AP2 domain-containing transcriptional activator that binds to the C-
repeat/DRE, a cis-acting DNA regulatory element that stimulates transcription in
response to low temperature and water deficit. Proc. Natl. Acad Sci. 94: 1035-1040.

Sukumaran, NP, Weiser, CJ. 1972. An excised leaflet test for evaluating potato frost
tolerance. HortScience. 7: 467-468.

Thomashow, MF. 1993. Plant Responses to Cellular Dehydration During Environmental
Stress, ed. By TJ Close, EA Bray, Am. Soc. of Plant Physiol. Rockville: 137-143.

Thomashow, MF. 1999. Plant cold acclimation: Freezing tolerance genes and regulatory
mechanisms. Ann. Rev. Plant Physiol. Plant Mol. Biol. 50: 571-599.

Thomashow, MF, Gilmour, SJ, Lin, C. 1993. Advances in Plant Cold Hardiness, ed. By
PH Li and L Christersson. CRC Press, Boca Raton: 31-44.

Thomashow, MF, Stockinger, EJ, Jaglo-Ottosen, KR, Gihnour, SJ, Zarka, DG. 1997.
Function and regulation of Arabidopsis thaliana COR (cold-regulated) genes. Acta
Physiol. Plant. 19: 497-504.

Towbin, H, Staehlelin, T, Gordon, J. 1979. Electrophoretic transfer of proteins from
polyacrylamide gels to nitrocellulose sheets - procedure and some applications. Proc.

71

Natl. Acad Sci. 76: 4350-4354.

van der Krol AR Mur LA, Delange P, Gerats AGM, Mol JNM, Stuitje AR. 1990.
Antisense chalcone synthase genes in petunia - visualization of variable transgene
expression. Mol. Gen. Genet. 220: 204-212.

Van Hoof, A, Green, PJ. 1996. Premature nonsense codons decrease the stability of
phytohemagglutinin mRNA in a position-dependent manner. Plant J. 10: 415-417

Wang, H, Datla, R, Georges, F, Loewen, M, Cutler, AJ. 1995. Promoters from kin] and
COR6. 6, 2 homologous Arabidopsis-thaliana genes - transcriptional regulation and gene—
expression induced by low-temperature, ABA, osmoticum and dehydration. Plant Mol
Biol. 28: 606-615.

Waterhouse, PM, Graham, HW, Wang, MB. 1998. Virus resistance and gene silencing in
plants can be induced by simultaneous expression of sense and antisense RNA. Proc.
Natl. Acad Sci. USA. 95: 13959-13964.

Weigel, D. 1995. The APETALAZ domain is related to a novel type of DNA-binding
domain. Plant Cell. 7: 388-389.

Wiser, CJ. 1970. Cold resistance and injury in woody plants. Science. 169: 1269-1277.
Welin, BV, Olson, A, Palva, ET. 1995. Structure and organization of 2 closely-related
low-temperature-induced DHN/LEA/RAB-like genes in Arabidopsis-thaliana L Heynh.
Plant MolBiol. 26: 131-144.

Yamaguchi-Shinozaki, K and Shinozaki K. 1994. A novel cis-acting element in an

Arabidopsis gene is involved in responsiveness to drought, low-temperature, or high-salt
stress. Plant Cell 6: 251-264.

72

3. CHAPTER 3: Detection of Arabidopsis CBF proteins by immunoblot

analysis and immunoprecipitation

3.1 INTRODUCTION

There are many ways in which transcription factors can be regulated in a
eukaryotic cell. These include transcriptionally, post-transcriptionally, translationally,
post-translationally through modification, or combinations of the above (Gallic, 1993;
Schwechheimer et al, 1998). Additionally, transcription factors can be regulated by their .
physical location in the cell. For example, some transcription factors are sequestered in
the cytoplasm, then translocated to the nucleus when the correct stimulus is detected
(Gallic, 1993; Schwechheimer ct a], 1998).

Changes in gene expression are associated with cold acclimation, the process by
which plants increase in freezing tolerance after exposure to low, nonfreezing
temperatures, as was first demonstrated in spinach (Guy ct al, 1985). Genes associated
with cold acclimation, called COR (cold-regulated) -also called LTI (low temperature-
induced), KIN (cold-inducible), RD (responsive to desiccation) and ERD (early
dehydration-inducible), were later isolated in Arabidopsis. The COR genes, which
include COR6. 6, COR15, COR4 7 and COR 78, are greatly upregulated under cold-
acclimating and drought conditions as well as by application of ABA (Hajela et al, 1990;

Nordin et al, 1991; Wang et a], 1994; Welin ct al, 1994).

73

A major breakthrough in understanding how COR genes are regulated by low
temperature was the isolation of CBF 1, CBF2 and CBF 3 (CRT/DRE Binding Factor) also
called DREBI b, DREBIc and DREBIa respectively (Dehydration Responsive Element
Binding factor). The CBF/DREBs encode for a family of transcription factors which bind
to the CRT/DRE (C-Repeat/Dehydration Responsive Element) present in COR gene
promoters and activate transcription under cold-acclimating conditions (Stockinger ct a1,
1997; Gilmour et al, 1998; Liu et al, 1998; Shinwari ct a1, 1998). The three genes, CBF 1,
CBF2 and CBF 3 are highly similar in the amino acid sequence (~85%), contain AP2-like
DNA binding domains, putative acidic activation domains, and have been shown to
activate transcription in yeast (Stockinger ct al, 1997; Gilmour ct al, 1998). Constitutive
overexpression of CBF 1, CBF 2 or CBF 3 in Arabidopsis results in constitutive COR gene
expression and increased freezing tolerance under non-acclimating conditions (J aglo-
Ottosen et a], 1998; Liu ct al, 1998; Gilmour et al, 2000; S. Gihnour, M. Salazar, A.
Sebolt, and M. Thomashow, unpublished).

The three CBF genes are regulated at the level of RNA accumulation. RNA levels
of all three CBF genes increase within 15 min of a low temperature stimulus, peak at two
to four hours and remain at an elevated state for as long as the plants are under
acclimating conditions (Gilmour ct al, 1998). While these data indicate that the CBF
genes are transcriptionally and/or post-transcriptionally regulated, CBF proteins may also
be translationally regulated or post-translationally modified. Analysis of the amino acid
sequence of the CBF2 protein shows that there are 17 predicted potential phosphorylation
sites; 9 on serinc residues, 6 on threonine residues and 2 on tyrosine residues (http:

//www.cbs.dtu.dldservices/thPhos/-; Blom ct al, 1999). Additionally, one of the sites

74

resembles the recognition site for MAP kinases (E. Stockinger, M. Thomashow,
unpublished). This raises the intriguing possibility that the CBF proteins could be post-
translationally modified through phosphorylation.

Phosphorylation events can modify the activity transcription factor at three levels:
import into the nucleus, enhancement or repression of DNA binding activity, and
enhancement or repression of the activation potential (Hunter and Karin, 1992;
Schwechheimer ct al, 1998). There are data that are consistent with the possibility of a
CBF-phosphorylation event under cold acclimating conditions. A low-temperature
induced mitogen activated kinase was recently isolated from alfalfa (J onak et al, 1996)
which is activated within 10 min of exposure of plants to low temperatures. Ifa
homologous kinase is present in Arabidopsis, then it is possible that the CBF proteins
could be activated by a phosphorylation event under acclimating conditions.
Alternatively, an upstream factor could be modified which would then activate the CBF
proteins (Thomashow, 1999).

To date, the only data available on the regulation of the CBF genes are RNA
accumulation data. To firlly understand how the genes are regulated, information on the
level to which the proteins are accumulated, and if the proteins are post-translationally
modified is critical. I attempted to determine the amount of CBF protein accumulation in
transgenic CBFI- CBF2- or CBF3-overexpressing plants and wild type Arabidopsis
plants. This was done not only to determine the levels of CBF protein accumulation, but
also to determine the location of the protein in the cell and whether the protein was post-
translationally modified under nonacclimating or acclimating conditions. If a

modification event was detected, the type of modification and whether the modification

75

worked to activate or repress the firnction of transcription activation of the COR genes
would then be investigated. Additionally, determining the amount of protein accumulated
and the location of the CBF proteins in transgenic plants as compared to wild-type
nonacclimated and acclimated plants was of interest. The amount of CBF] RNA
accumulated in the nonacclimated A6 line appears to be greater than that of four-day
cold-acclimated control plants, whereas the amount of COR gene RNA that is
accumulated appears equal in both types of plants. Determining the amount of CBF 1
protein that is accumulated in both types of plants may give an indication as to why the
COR gene RNA accumulation is not directly reflective of the amount of CBF! RNA

accumulated.

3.2 MATERIALS AND METHODS:

3.2.1 PLANTMATERIAL

Types of plant material used in this chapter are as follows:

 

Plant material Nomenclalure

Wild type Arabidopsis plants: ecotype RLD or WS

CBFI-overexpressing plants: A6 (described in 2.2.2)
G7a-1 *

CBF2-overexpressing plants: E71-1 *

CBF3-overexpressing plants: A30a-1 *

76

A38b-7 *

Vector control plants: B16-1 **

For the CBF-overexpressing lines, C“) the coding region of each specific CBF cDNA was
cloned into the binary expression vector pGA643 (An, 1987) and plants were transformed
by the floral dip procedure as described (Gilmour et al, 2000; M. Salazar, A. Sebolt, S.
Gilmour M. Thomashow, unpublished). For the vector control plants C”) the pGA643

vector alone was transformed as described above.

3. 2. 2 PIANT GROWTH

3. 2. 2. 1 Plant growth (no radiolabelling)

Arabidopsis plants were grown in pots under 100 umol m'zs'l continuous light for
18-25 days as described (Gilmour ct a], 1988; see 2.2.1). For experiments involving cold-
acclimation, plants were cold acclimated at 4° C under 50 mo] m'zs'l continuous

fluorescent illumination for various amounts of time as indicated (Gilmour et al, 1988;

see 2.2.1).

3.2.2.2 Plant growth for 35S-methionine radiolabelling

Arabidopsis plants were grown on petri plates or in magenta boxes containing 1 x

Gamborgs B-S media (GibcoBRL,Grand Island, NY) as recommended by the

77

manufacturer and solidified with 1% agarose. Plants were placed under 30-60 umoh’n'zs'l
florescent illumination in a 16/8 hr light/dark cycle. When petri plates were used, plants
were grown in a cluster in the center of the plate and a total of ~1 5-50 19-day old plants
were analyzed under both nonacclimating and acclimating conditions. When magenta
boxes were used, l9-day old plants were thinned from ~15 to 4-8 individuals before
adding the 35 S-methionine as described in 3.2.7.2. For experiments involving cold-

acclimation, plants were put in the 4° C cold room as described in 3.3.3.] for a total of 24

h

3. 2. 2. 3 Plant growth for 32P-orthophosphate radiolabelling

Arabidopsis plants were grown in magenta boxes containing 1 x Gamborgs B-5
media (GibcoBRL,Grand Island, NY) as recommended by the manufacturer solidified,
with 1% agarose and grown under the light conditions described above (see 3.2.2.2). For
use in experiments, 14-22 day old plants were thinned from ~15 to 4-8 individuals before
the addition of 32P-orthophosphatc as described in 3.2.7.3. For the experiments involving
cold-acclimation, plants were put in the 4° C cold room as described in 3.3.3.] for a total

of6h

3. 2. 3 ISOLATION OF RECOMBINANT CBF] PEPTIDES FROM E. COLI EXTRACTS AND YEAST

3. 2. 3. I Overexpression in E. coli

78

Full length CBF! (amino-acids 1-213), the N-terrninal portion of CBF] (amino
acids 1-115) and the C-tcrrninal portion of CBF] (amino acids 116-213) were cloned into
the pGEX expression vectors (Pharmacia Biotechnology) under the control of the tac
promoter, which results in the production of GST tagged polypeptides (E. Stockinger, M.
Thomashow, unpublished). The resulting plasmids, pEJ S3 69, pEJ S3 70 and pEJS371
respectively, were transformed into BL21 E. coli cells (E. Stockinger, M. Thomashow,
unpublished). Translation of the recombinant CBF] polypeptides was induced with
isopropyl thiogalacto-pyranoside (IPTG) as recommended by the supplier (Pharmacia
Biotechnology) after which cells were lysed using a French press (Spectronic
Instruments, Rochester, NY). Total soluble proteins, which included 8 M urea solublizcd
CBFl-containing inclusion bodies from the pellet, were incubated with glutathione-
agarose beads (Sigma, St. Louis, M0) to purify the GST-labelled CBF] polypeptides as
described (Ausubcl et al, 1987). In some cases, the GST tag was removed by cleaving

with thrombin (Ausubcl et al, 1987).

3.2.3.2 Overexpression in yeast

Yeast cells that overexpress CBF 1 were created by placing the coding region of
the CBF] cDNA into the pDB20.1 expression vector (kindly provided by S. Triezenberg)
under the control of the yeast ADC] promoter (Stockinger et al, 1997). Total protein
extracts (kindly provided by E. Stockinger) were isolated following the methods of Rose
and Botstein (1983). This involves growing cells to A600 = 0.8- 1.0, chilling them on ice,

briefly centrifuging the cells, suspending the cells in breaking buffer (100 mM Tris-HCl

79

(pH 8), 1 mM dithiothreitol and 20% glycerol) adding 0.45- 0.5-mm glass beads (Sigma)

vortexing the cells, then centrifuging for 15 s and isolating the supernatant.

3. 2. 4 SYNTHETIC PEPTIDE PRODUC TIONAND Ii/IANIPULAT ION

In order to generate antibodies that specifically detect CBF], the C-terrninal
sequence CWNHNYDGEGDGDV from the CBF] protein was used as an epitope. This
portion of the protein is divergent from the CBF2 and CBF3 proteins (Gilmour et al,
1998). To generate antibodies that were expected to detect all three CBF proteins, we
selected an epitope, the N-tenninal sequence, FSEMFGSDYEC, that is conserved in all
three CBF proteins (Gilmour et al, 1998). A cysteine residue was added to the end of
each peptide to link the peptide to a carrier protein. Peptides were synthesized at the Keck
Foundation Biotechnology Resources Laboratory on a Protein Technologies Symphony
Multiple Peptide Synthesizer at the request of the Biopolymers Facilities at the Howard
Hughes Medical Institute (Harvard Medical School).

Short peptides and other molecules smaller than 3000- 5000 Daltons, are not large
enough molecules to firnction as irnmunogens (Harlow and Lane, 1988). Therefore, to
elicit an immune response, 2 mg of peptides were conjugated to maleirnide-activated
keyhole limpet hemocyanin (KLH) as described by the manufacturer (Pierce, Rockford,
IL) and checked for efficient conjugation by using Ellman’s reagent (Pierce, Rockford,
IL). As an adjuvant to stimulate the immunoresponse, 19211] of TiterMax (prepared as

recommended by the manufacturer) (Cthx Corporation, Norcross, GA), was mixed with

80

288 u] of the peptides conjugated to KLH (containing approximately 0.5-1 mg of
peptide) immediately prior to injection. Two rabbits for each peptide (#58 and #62 for the
N-terrninal, #63 and #64 for the C-terrninal), were injected with ~1 ml of the peptide-
KLH/TiterMax mix by a professional ULAR employee in accordance with the ULAR
guidelines (animal use form approval number 03/96-021-00). Rabbits were bled
approximately every four weeks and boosted with KLH-peptide when the titer of the
antisera dropped. Serum was obtained by centrifugation of the clotted blood to remove

red blood cells.

3. 2. 5 ANTIBODIES USED

Three types of anti-CBF] antisera raised against the three following epitopes

along with anti-COR15a antibodies were used in this chapter,:

Epitope Used/Abbreviated name Experimental detail_s

 

 

Recombinant firll length CBF 1 (F 1, F2) Rabbit #4 and #5, bleed 110 days (E.
Stockinger, unpublished)

Peptide of N-tenninus of CBF] (N 1, N2) Rabbit #58 and #62, bleed #4 unless
otherwise indicated

Peptide of C-terrninus of CBF 1 (C1, C2) Rabbit #63 and #64, bleeds #4 unless
otherwise indicated

COR15a(COR15m) Rabbit Larry (Artus et al, 1996)

Sera containing antibodies to firll length CBF] was raised to recombinant histidine—
tagged CBF] (Stockinger, et al, 1997 ; E. Stockinger, M. Thomashow, unpublished). Sera

81

to the CBF] N- and C-terrninal peptides were raised as described in 3.2.4. Sera to the

COR15a protein was raised to recombinant COR15a polypeptides (Artus, 1996).

3. 2. 6 PURIFICATION OF ANTI-CBF] SERA

3. 2. 6. 1 Introduction

Sera containing anti-CBF] antibodies fi'om F 1, F2, N1, N2, C1 and C2 (see 3.2.5)
were purified in a variety of ways. These included IgG purification, purification against
firll length recombinant CBF] (see section 3.2.3, expression vector pEJ S369),
purification against portions of CBF 1 (see section 3.2.3, expression vectors pEJ S3 70 and

pEJ S371) and purification against synthesized peptides (see section 3.2.4).

3. 2. 6.2 Purification of IgG

Protein A, a cell wall protein of S. aureus, contains four potential binding sites for
antibodies (Harlow and Lane, 1988). Therefore, it is an ideal molecule with which to
purify the IgG fraction from other proteins and molecules contained in crude rabbit sera.
To purify the IgG portion of F 1, F 2, C1, C2, N1 and N2 (see 3.2.5), a standard protocol
based on adsorption of the IgG fraction to Protein-A sepharose beads (Sigma) was

followed (#18. 12, Sambrook et al, 1989).

82

3. 2. 6. 3 Immunoaflinity purification to recombinant CBF peptides

For purification of antibodies C1 and C2 (see 3.2.5) to full length CBF] protein
(see section 3.2.3, expression vector pEJ S369) or the N- and C-terminal portions of CBF]
(see section 3.2.3, expression vectors pEJSB7O and pEJS37 1 respectively), a standard
protocol using recognition of the GST-tag (Pharmacia Biotechnology) was followed
(Bar-Peled and Raikhel, 1996). Briefly, total E. coli protein lysates containing
recombinant CBF peptides were isolated as described (3.2.3), then incubated with
glutathione-agarose beads, (Sigma, St. Louis, MO) which bind the GST-labelled CBF]
polypeptides. The bound GST-CBF peptides are cross-linked to the glutathione-agarose

beads and used to immunoaffinity purify anti-CBF antibodies from crude rabbit sera.

3. 2. 6. 4 Immunoaflinity purification to synthetic peptides

To purify the anti-CBF 1 antibodies contained in N1, N2, C1 and C2, raised
against synthetic peptides (see 3.2.5), a standard protocol “purification of antibodies
against your specific peptide using Sulfolink Gel” was followed according to the
manufacturers instructions (Pierce, Rockford, Il). Conjugation was checked with
Ellman’s reagent as recommended by the manufacturer (Pierce, Rockford, 11). This
protocol involves conjugating the synthetic peptides to a matrix, then using the bound

peptides to immunoaffinity purify the antibodies of interest.

83

3. 2. 7 EXTRACTION OF PROTEINS FROM PLANTS

3. 2. 7. 1 Extraction of unlabelled proteins

Plants were grown as described in 3.2.2. 1. For immunoblot analysis, total soluble
protein was extracted from by grinding frozen leaf tissue (about 150 mg) into 300-400ul
of one of the following extraction buffers:

1) 3x loading buffer for tricine SDS- PAGE (see section 2.2.5; Schagger and von

Jagow, 1987) (30% glycerol, 6% SDS and 3 x upper tricine buffer); or
2) “buffer A” containing 50 mM Tris-HCI (pH 8.0), 5% glycerol, 100 mM KC1,
1.5% (wt/vol) polyvinyl-pyrrolidone, and the following cocktail of protease
inhibitors: lmM PMSF, lmM benzamide, lmM behzamidine'HCl, 5 mM E-
Amino-n-caprioic acid, 10 mM EGTA, lug/mg antipain, 1 ug/mg leupcptin 0.1
mg/ml pepstatin (Sigma) and 5 mM DTT.
Insoluble material was removed by ccntrifirgation at 3,000 x g for 20 nrin at 4° C, and the

remaining protein in the supernatant was quantitated by the Bradford dye-binding assay

(Bio-Rad, Hercules, CA).

3. 2. 7. 2 Extraction of 35S-methionine labelled proteins

Plants were grown on media as described in 3.2.2.2. Labelling was conducted by

adding 0.5 mCi of 3’SS-methionine (44.5 111) suspended in 155.5 111 0.2% tween and

pipetting the 200 u] of solution onto the leaves of the plants, after which plates were
wrapped with parafilm. For nonacclimating conditions, one set of each type of plant was
placed in the radioactive-use hood under low light conditions (~15 umolm’zs") for 24 h at
~21° C. For cold-acclimating conditions, the second set of plants were placed at 4° C as
described in 3.2.2.2 for 24 h. After the incubation, all plant material, including root
material, was removed from plates using forceps. Plants were rinsed with deionized water
and instantly flown in liquid nitrogen in microcentrifuge tubes. Tissue was ground in 200
u] extraction buffer consisting of: 50 mM Tris-HCl pH 8.0, 100 mM KC], 2 mM MgC12,
2 mM EDTA, 5% glycerol, 0.5% DOC, 5 mM DTT, ~1.5% (wt/vol) polyvinyl-
pyrrolidone (Mao, unpublished) the CompleteTM Mini EDTA-free protease inhibitor
cocktail tablet (Bochringer-Mannheim GmbH, Mannheim, Germany) lmM PMSF, 0.]
mg/ml pepstatin, 5 mM DTT and 5% B-mercapto-ethanol (Sigma). The insoluble
material was removed by centrifirgation at 13,000 x g for 20 min at 4° C. The supernatant
was quantified for radioisotope incorporation using standard trichloroacetic acid (TCA)

procedures (Sambrook et al, 1989) and stored at -80° C until use.

3. 2. 7. 3 Extraction of 3ZP-orthophosphate labelled proteins

Plants were grown on media as described in 3.2.3.3. Labelling was conducted by
pipetting 0.5 mCi of 32P-orthophosphate (100 pl) suspended in 200 u] 0.2% tween onto
the leaves of plants, ~75 u] of radioactive material per plant, after which the magenta

boxes were wrapped with parafilm. For nonacclimating conditions, one set of each type

85

of plant was placed in the radioactive-use hood under low light conditions (~15 umolm’
2s") for 6 h at ~21° C. For cold-acclimating conditions second set of plants were placed
at 4° C as described in 3.2.2.2. After the 6 h incubation, all plant material, including root
material, was then removed from plates using forceps and extracted as described in

3.2.7 .2 with the modification that 15mM B-glycerophosphatc and 5 mM NaF were added
as phosphoinhibitors. The supernatant was quantified for radioisotope incorporation as

described above (see 3.2.7.2) and stored at —80° C until use.

3. 2.8 IWUNOBLOT ANALYSIS

3.2.8.1 Analysis without immunoprecipitations

To analyze CBF protein content in whole plant extracts, immunoblot analyses

were conducted as described in 2.2.5 with the following modifications. For recognition of

the CBF 1 protein, several forms of the three types of antibodies were used as indicated in

the Figures. The specific anti-CBF 1 antibody used and the level of purification are as

 

follows:
Type of antibody Level of purification
F1 and F2: Unpurified
IgG purified (see 3.2.6.2)
Immunoaffrnity purified to recombinant peptides (see
3.2.6.3)
C1, C2, F1, and F2 Unpurified
IgG purified (see 3.2.6.2)

Immunoafiinity purified to synthetic peptides (see 3.2.6.4)

86

Three types of secondary antibody, to detect the rabbit anti-CBF] antibodies,
were used as indicated in the Figures (Sigma, St. Louis, MO):

1) unpurified anti-rabbit immunoglobulin peroxidase conjugate; or

2) anti-rabbit IgG (whole molecule) peroxidase conjugate; or

3) monoclonal anti-rabbit immunoglobulins peroxidase conjugate.

Proteins were visualized using the ECL system (Amersham Buckinghamshire, UK).

3. 2. 8. 2 Analysis after immunoprecipitations

For detection of protein accumulation after immunoprecipitations (see section
3.2.9) the protein-A agarose beads (to which the antibodies and protein of interest were
bound) were washed, isolated and heated in 1 x extraction buffer (see sections 3.2.7.2 and
3.2.7.3) containing 5% SDS and 5% B-mercapto-ethanol (Laemmli, 1970). The
supernatant was removed from the and loaded on 10% Laemmli gels (1970), transferred

to nitrocellulose (see 2.2.5) and visualized by ECL (see 3.2.8.1).
3.2.9 IWUNOPRECIPITAT IONS OF CBF I PROTEINS

Immunoprecipitations of nonradiolabelled plant extracts and 35 S-methioninc
labelled E. coli extracts were conducted using one of two following protocols: the

modified Hondred (see 3.2.9.1) or the Mittlera (see 3.2.9.2).

87

3. 2. 9. 1 Modified Hondred protocol

Following the modified Hondred protocol (Hondred et al, 1987, S. Gilmour,
unpublished) plant extracts (see section 3.2.2.], extraction buffer #2) or E. coli extracts
(see section 3.2.10) were pre-incubated with rabbit serum-agarose beads and protein-A
agarose beads (Sigma) for 1 h at 4° C with shaking to remove proteins that bind non-
specifically to the beads. Specifically, 10 u] of each type of bead were added to 200111 of
protein lysate. An equal volume of TNET/SDS (SOmM Tris-HCl (pH 7.5), 150 mM
NaCl, 1 mM EDTA, 2% Triton-X 100 and 0.2% SDS), the protease inhibitor cocktail
(see section 3.2.2.], extraction buffer #2), and an equal volume of tricine loading buffer
(Schagger and von Jagow, 1987) and were added. To this solution, 10 u] of crude or IgG
purified (described in 3.2.8.1) immune serum, or pre-immune serum was added. The
solution was shaken for 15 rrrin on ice after which 10 u] of protein-A agarose beads
(Sigma) were added and the solution was shaken for an additional 15 min on ice. The
supernatant was removed, the beads were washed three times with urea wash (2 M urea,
1% Triton-X 100, and 10 mM Tris‘HCl pH 7.5), then one time with THE; (pH 7.5).
Beads were heated to 100° C in 40 ul 1 x loading buffer with 5% B-mercapto-ethanol for

5 min and electrophoreased on a 10% tricine gel and visualized by immunoblot analysis

(see section 3.2.8.2).

3. 2. 9.2 Mittlera et a] protocol

88

Following the protocol of Mittlera et a] (1998), tissue was extracted in ice cold 50 mM
Tris‘HCl (pH 8.0), 150 mM NaCl and 1% Triton-X 100 plus the cocktail of protease
inhibitors (see section 3.2.2. 1, protease inhibitors in extraction buffer #2). Three volumes
of 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% Nonidet P-40, 1 mM‘ EDTA and
0.25% gelatin were added to the sample. The samples were centrifuged for five min at
10,000 x g at 4° C, the supernatant was quantitated by Bradford dye-binding assay (Bio-
Rad, Hercules, CA) before use. Irnmunoprecipitation experiments were conducted by
incubating protein extracts with 5 or 10.1] of pre- or immune serum as described in the
Figures, shaking the solution for three h on ice, after which 40 pl of protein-A beads were
added and incubated for an additional 30 rrrin on ice. The beads were washed three times
with wash buffer (50 mM Tris-HCl (pH 7 .5), 150mM NaCl, 0.1% Nonidet p-40 (NP-40)
and 0.25% gelatin) resuspended in 40 u] 1 x loading buffer with 5% B-mercapto-ethanol,
heated to 100° C in for 5 min and electrophoreased as described above (see section

3.2.9.1).

3. 2. 1 0 ISOLA TION TOTAL SOL UBLE PROTEINS FROM 3 5S-METHION1NE RADIOIABLLED E. COLI

FOR M4 UNOPREC IPI TA TIONS

E. coli cells (BL21) that overexpress GST-CBF] (see section 3.2.3, vector
pEJ S369) were grown and labelled as described (Harlow and Lane, 1988 p. 442, 458).
Briefly, a single colony of pEJ S3 69 overexpressing E. coli was inoculated into four ml of
LB growth media and grown overnight with shaking. To obtain high-specific-activity

89

labelling, the cells were diluted 1/ 10 into M9 medium, (Harlow and Lane, 1988) and
allowed to grow until an ODsoo of 0.4 was reached, after which 0.1 mM IPTG and 100
uCi of 35 S-methionine were added. Cells were grown for another four h, placed on ice for
30 nrin after which cells were centrifuged for five min at 5,000 x g, washed with
phosphate buffered saline (PBS) then resuspended in 10 cell volumes 50 mM glucose, 10
mM EDTA, 25 mM Tris‘HCl (pH 8.0) 4 mg/ml lysozyme and the protease inhibitor
cocktail (see section 3.2.2.] buffer #2). Cells were incubated at ~21° C for five min,
placed on ice, centrifirged for ten nrin at 10,000 x g at 4° C and resuspended in lysis
buffer (150 mM NaCl, 1% NP-40, 0.5% DOC, 0.1% SDS and 50 mM Tris‘HCl, pH 8.0)
and placed on ice for 30 min. The cells were then centrifuged for 10 min at 10,000 x g at
4° C afier which the lysatc was removed to a new tube. The pellet was re-suspended in 8
M urea and 0.1 M Tris‘HCl (pH 8.0), centrifirged briefly and the supernatant added to the
retained supernatant, while ensuring that the total urea concentration of the protein buffer
remained <1 M. The proteins were aliquoted into microcentrifirge tubes and frozen at

-20° C until use. Immunoprecipitations were conducted as described in 3.2.9.].

3. 2. 1 1 IMMUNOPRECIPITA TIONS WITH 35S-METHION1NE LABELLED PROTEINS FROM

ARABIDOPSIS PLANTS

For immunoprecipitations conducted with 35 S-methionine labelled proteins from
Arabidopsis plants, plants were grown as described in 3.2.2.2. Plants were 3‘5 S-methionine
labelled, total soluble proteins were isolated, and the level of radioisotope incorporation
was determined as described in 3.2.7.2. Protein extracts with equal counts of radioactivity

90

(2,940,000 counts) were used in all immunoprecipitation experiments.
Irnmunoprecipitation of CBF 1 was conducted as follows. Protein extracts (isolated as
described in 3.2.7.2) were pre-incubated over night on ice with 30 u] of each of protein-A
agarose beads and rabbit serum-agarose beads (Sigma). The supernatant was removed,
and 20 u] of pre- or immune serum from N2 (see section 3.2.5) was added and the
solution was shaken on ice for 3 h after which 45 u] of protein-A beads (Sigma) were
added and solution was shaken for another h. After completing the incubations, the
solutions were centrifuged briefly, the supematants removed and the beads were washed
three times with 750 u] extraction buffer (see section 3.2.7.2) for two to five nrin with
shaking. The beads were then resuspended in 40 u] 1 x Laemmli loading bufi‘er (1970),
heated to 100° C for five min in the presence of 5% B-mercapto-ethanol and loaded onto
10% Laemmli gels (1970). Gels were dried on a model 583 gel drier (BioRad) overnight
and fluorography was performed with BioMax MS scientific imaging film and a BioMax

MS intensifying screen (Kodak, Rochester, NY).

3. 2. 12 [MM UNOPREC 1P1 TA TIONS WITH 3Z’IJ—ORTHOPHOSPHA TE LABELLED PROTEINS FROM

ARABIDOPSIS PLANTS

For immunoprecipitations with 32P-orthophosphate labelled proteins from
Arabidopsis plants, plants were grown as described 3.2.2.3. Plants were 32P-

orthophosphate labelled, total soluble proteins were isolated, and the level of radioisotope

91

incorporation was determined as described in 3.2.7.3. Protein extracts with equal counts
of radioactivity (140,000 counts) were used in all immunoprecipitation experiments.
Protein extracts (see 3.2.2.3) were first pre-incubated overnight with 30 ul of protein-A
agarose beads and 30 u] of rabbit serum-agarose beads (Sigma). The supernatant was
removed and incubated with 20 u] of pre- or immune serum from C2 (see section 3.2.5)
and 30 ul of protein-A beads for two nights. Beads were washed and loaded onto
Laemnrli gels, and dried as described in 3.2.1 1. Fluorography was performed with

Hyperfilm MP (Amersham Pharrnacia Biotech UK Limited, Buckinghamshire, UK).

3.3 RESULTS:

3. 3. 1 RECOMBINANT CBF] IS DEGRADED IN THE PRESENCE OFARABIDOPSIS PROTEIN

EXTRAC TS

As CBF proteins had not been detected by immunoblot analysis in previous
experiments (S. Gilmour, M. Salazar, and K. Jaglo-Ottosen, M. Thomashow, unpublished
results), I first wanted to determine if the buffer used to extract soluble plant proteins was
permitting the degradation of the CBF proteins. To answer this question, 100 ng
recombinant CBF] protein (see section 3.2.3) was added to leaf tissue from
nonacclimated RLD plants and total soluble proteins were extracted (see 3.2.7.1). To
determine if the specific buffer used and/or the addition of protease inhibitors had any

effect on CBF degradation, two different types of buffer (see 3.2.7.] #1 and #2) and

92

various protease inhibitors were used. Specifically, both buffers were used to extract
proteins in the absence of protease inhibitors, or one of each of the following protease
inhibitors were added to “buffer A” (see 3.2.7.] #2) before extraction: lug/mg antipain,
lmM PMSF, 1 ug/mg leupcptin or 0.1 mg/ml pepstatinA (Sigma) afier which
immunoblot analysis was conducted using sera from F 1 (data not shown). The
recombinant CBF] protein was degraded if no protease inhibitors were added to the
extraction buffer resulting in the absence of a band where the CBF 1 protein is predicted
to run and all protease inhibitors appeared equally effective at blocking total degradation
of CBF]. Therefore, either the described cocktail of protease inhibitors were added to all
extraction buffers (see section 3.2.7.] #2), or a CompleteTM Mini EDTA-free protease
inhibitor cocktail tablet was added to all extraction buffers as described by the

manufacturer (Boehringer-Mannheim GmbH, Mannheim, Germany).

3. 3. 2 CBF PROTEINSARE NOT DETECTED IN TOTAL SOLUBLE PLANT OR CBFI-

OVEREXPRESSING YEAST PROTEIN EXT RAC TS

Detecting CBF] protein in total soluble protein extracts from leaf tissue of wild
type or CBFI-ovcrexpressing plants using crude rabbit sera proved difi'rcult due to
complex banding patterns (see Figure 3.1A). These are presumably due, in part, to the
cross reactivity of proteins contained within crude anti-CBF 1 rabbit sera to proteins
present in total soluble plant extracts. IgG purification of rabbit serum used in
immunoblot analysis can result in decreased background (Harlow and Lane, 1988). To

determine if isolation of the IgG fraction of F2 (see section 3.2.5) decreased

93

“background” banding, the IgG fraction was purified from serum (see section 3.2.6.2).
CBF] protein accumulation was then determined with recombinant CBF] protein, protein
extracts from CBFI-expressing yeast, vector only-expressing yeast (Stockinger et al,
1997, see 3.2.3.2), nonacclimated RLD and A6 plants (see section 2.2.2, see 3.2.7.] #2
for extraction buffer) and 5- and 8- day cold-acclimated RLD plants (see Figure 3.1).
While 50 ng of recombinant CBF] protein was detectable with both types of sera (the
CBF] protein is a doublet of ~ 29 KDa), CBF] could not be detected in CBFI-
overexpressing yeast or in plant extracts over “background” patterns. As CBF] RNA
accumulation increases within 15 min of touch stimulation (Gihnour et al, 1998) I wanted
to determine if moving the plants fi'om growth chambers to the lab before harvesting
proteins had an effect on the ability to detect the CBF 1 protein. However, the inability to
detect CBF proteins was independent of where plant tissue was harvested as CBF
proteins could not be detected regardless of whether plants were harvested in grth
chambers or in the lab (data not shown).

3. 3. 3 PRE-M/IUNE SERUIW CONTAINS ANTIBODIES TO NUMEROUS PROTEINS CONTAINED IN

WHOLE PLANTEXTRACTS

To determine specifically what was causing the complex banding patterns seen in
Figure 3.], several control experiments were conducted. Irnmunoblot analysis was
conducted using either pre-irnmune serum, no primary antibody or immune serum fi'om
F2 (see section 3.2.5) against 50 ng recombinant CBF], total soluble proteins from
nonacclimated RLD and CBFI-overexpressing A6 plants (see sections 2.2.2 and

3.2.7.1#2), and extracts from CBFI-overexpressing yeast cells (see 3.2.3.2) (Figure 3.2).

94

12345678

 

A
_¢--——
-‘
29kDa> s“ g; 3‘
"e . If”.
t.

 

 

 

 

 

Figure 3.1. Immunoblot analysis of recombinant and native CBFl

Irnmunoblot analyses (see 3.2.8.1) were conducted using both IgG purified
(1B, see 3.2.6.2) and crude F2 antisera (1A, see 3.2.5) against recombinant
CBF] protein (see 3.2.3.1) and soluble proteins from plant (see 3.2.] for
plant types and 3.2.7.] #2 for protein extraction) and yeast extracts (see
3.2.3.2). In both immunoblots, the secondary antibody was unpurified (see
3.2.8.] #1) used at a dilution of 1:5,000. The images are from a 5 minute
exposure to x-ray film. The 29 kDa marker, the predicted size of CBF], is
indicated by the arrow.

Lanes are as follows:

5 ng CBF] recombinant protein;

50 ng CBF] recombinant protein

100 pg total yeast protein extract (vector-only);

100 ug total yeast protein extract (CBF I -expressing);

100 ug protein extract from nonacclimated RLD plants;

100 pg protein extract from 5-day cold-acclimated RLD plants;

100 ug protein extract from 8-day cold-acclimated RLD plants;

100 pg protein extract fiom nonacclimated A6 plants.

99$9‘H‘fft’l‘???

A. Irnmunoblot analysis using crude F2 antiserum diluted 1:50.
B. Irnmunoblot analysis using the IgG fraction purified from F2 antiserum
diluted 1:500.

95

The immunoblot analyses indicated cross reactivity to yeast and plant proteins in the pre-
immune serum of F2. Additionally, faint bands were seen when no primary antibody is
used. This indicates that the secondary antibody alone has cross reactivity to plant

proteins. Both of these factors are likely to increase background problems.

3. 3. 4 IWUNOAFFINI TY PURIFICATION OF ANTI-CBF I AN TISERA DOES NOT ENHANCE

DETECTION OF RECOMBINANT OR NA TIVE CBF PROTEINS

Purification of the IgG portion of anti-CBF 1 antiserum did result in a decrease in
“background” banding, but did not allow for clear visualization of the CBF proteins
(compare Figure 3.1A and B). Therefore sera fi'om F 1 and F2 (see section 3.2.5) were
purified to full-length recombinant CBF] peptides (see section 3.2.6.3) in an attempt to
eliminate antibodies to non-CBF proteins contained within the antiserum. Immunoblot
analysis using pre-immune serunr, crude immune serum, and sera purified to recognize
CBF peptides indicates that the purification process did not seem to effect recognition of
the three CBF peptides (see Figure 3.3; data for F1 not shown). Given that serum purified
to the C-terrninal portion of CBF I recognized the N-terrrrinal portion of CBF] (see
Figure 3.3, “CBF l-C” section, lane 2), it is unclear that the purification process worked.
However, given that there appear to be very few antibodies to the C-terminal portion of
CBF] in crude immune serum (see Figure 3.3, “1mm” section, lane 1), purification of
these antibodies would not be possible. Additionally, there appear to be antibodies
contained within the pre-immune serum that recognize the CBF] protein. When

immunoblot analyses were conducted against total soluble proteins from 4-day cold-

96

No 10 antibody Pre—immune Immune

123412341234

 

29kDa >1

 

 

 

Figure 3.2. Immunoblot analysis of recombinant and native CBFl

Irnmunoblot analyses (see 3.2.8.1) were conducted using either no
primary antibody (No 1° antibody), pre-immune serum (Pre-immune) or
immune serum (Immune) from F2 (see 3.2.5) against recombinant CBFl
protein (see 3.2.3.1) and total soluble proteins fi‘om plant (see 3.2.1 for
plant types and 3.2.7.1 #2 for protein extraction) and yeast extracts (see
3.2.3.2). Both pre- and immune sera were diluted 1:50. Images shown are
the result of a 2 minute exposure to x-ray film for the pre- and immune
sera and a 10 minute exposure for the No 1° antibody blot. In all blots,
the secondary antibody was unpurified (see 3.2.8.1 #1) used at a dilution
of 12,500. The 29 kDa marker, the predicted size of CBF proteins, is
indicated by the arrow.

Lanes are as follows:

1: 50 ng CBFl recombinant protein;

2: 100 pg protein extract from nonacclimated RLD plants;

3: 100 ug protein extract from nonacclimated A6 plants;

4: 25 ug protein extract from CBF 1-overexpressing yeast.

97

Pre-imm Imm CBF l CBF 1 -N CBFl -C
12341234123412341234

 

 

       

 

 

Figure 3.3. Immunoblot analysis of recombinant CBFl peptides

Irnmunoblot analyses were conducted (see 3.2.8.1) with pro-immune
(Pre-imm), crude immune (1mm), and recombinant full length
(CBF1), N-terminal(CBF1-N), and C-terrninal(CBF1-C) CBF]
peptide purified (see 3.2.6.3) serum from F2 (see 3.2.5).

The dilutions of the primary antibodies were as follows:
pre-immune: 1:1000;

crude immune: 1:1000;

CBF] purified: 1:1000;

CBF] -N-terminal purified: 1 : 100;

CBFl -C-termina1 purified 1:200.

In all blots, the secondary antibody was unpurified (see 3.2.8.1 #1)
used at a dilution 1:5,000. All of the blots shown are the result of a 3
minute exposure except for the immune serum which is from a 30
second exposure. All recombinant protein extracts were thrombin
cleaved. The 29 kDa, predicted size of CBFl, the approximate size
of the CBFl-N and C terminal peptides and the the 14.4 kDa markers
are indicated with arrows.

Lanes are as follows:

I: 1 pg C-terminal portion of recombinant CBFl;

2: 1 pg N-terminal portion of recombinant CBFl;

3: 1 pg fiill length recombinant CBF 1;

4: 1 pg GST (Sigma)

98

A Pre-Imm 1mm IgG CBF] CBF-N
1 2 3 1 2 3 1 2 3 1 2 1 2 3
60' I g

. s 9 9,,
.

,
‘ l
. ' I, l .
' 1 ' l
.r‘ . - V t
..'. t - ,.. .

 

 

   

~29 kDa! . .

 

 

 

 

 

 

 

Figure 3.4. Immunoblot analysis of CBF and CORISm proteins

Lanes are as indicated:

1: Nonacclimated RLD

2: 4-day cold-acclimated RLD
3: Nonacclimated A6 (see 3.2.1)

A. Irnmunoblot analysis of CBF]. Immunoblot analyses (see 3.2.8.1) were
conducting using pre-immune (Pre-Immune), crude immune (1mm), IgG
purified (IgG) (see 3.2.6.2) and recombinant full length (CBF1), and N-
terminal(CBF1-N) CBFl peptide purified (see 3.2.6.3) serum from F2
(see 3.2.5). The dilutions of the primary antibodies are as follows:
Pre-immune: 1:5000;

Immune: 125000;

IgG: 1:3000;

CBF]: 1:1000;

CBF l -C-terminal purified 1 :200;

CBFl-N-terrninal purified: 1:300.

In all blots, the secondary antibody was unpurified (see 3.2.8.1 #1) used at
a dilution 1:5,000. The 33 kDa marker, and the approximate position of 29
kDa, predicted size of CBFl, are indicated with arrows.

B. Immunoblot analysis of COR] 5. Immunoblot analysis was conducted
as described (3.2.8.1) using antiserum raised to COR15 (see 3.2.5) diluted
1:2000. The secondary antibody was unpurified (see 3.2.8.1 #1) used at a
dilution 1:5,000.

All of the blots shown are the result of a 30 second exposure to x-ray film.

99

acclimated RLD plants and nonacclimated RLD and A6 plants and using pre-immune
serum, crude immune sera, IgG purified serum(see 3.2.6.2) and recombinant CBFI-
peptide purified sera (see 3.2.6.3) no differences in banding patterns were detected (see
Figure 3.4A). As a control, immunoblot analysis was conducted using anti-COR15a
antibodies against the same protein extracts used to detect CBF] (see Figure 3.43). The
data clearly show accumulation of the abundant COR] 5m protein. This indicates that the
inability to detect CBF proteins is not due to technical difficulties in conducting the
immunoblot analysis. Collectively, these results indicate that the purification process was
ineffective at allowing for detection of the endogenous plant CBF proteins, possibly due

to the fact that purification process was unsuccessfial.

3.3.5 ANTIBODY PRODUCTION TO SPECIFIC PEPTIDES FROM THE CBF] PROTEIN

Despite the repeated efforts in using sera F1 and F2 (see section 3.2.5), I was
unable to detect CBF proteins in whole leaf plant extracts (Jaglo-Ottosen et al, 1998;
Figures 1 and 2). This could be due to low levels of CBF 1 protein accumulation, the anti-
CBFl antibodies having a weak affinity for the CBF 1 protein, or a combination of both.
However, given that 50 ng of recombinant CBFl protein was detected, it seems more
likely that CBF proteins do not accumulate to high levels in plants. Using antibodies
against full length CBF 1 also caused concerns. Depending on which part of the protein
formed the epitope for the immunoresponse, it was possible that other AP2-domain
containing proteins could be detected by anti-CBF 1 antibodies. Additionally, since a

large portion of the amino acid sequence is highly conserved between CBF 1, CBF 2 and

100

CBF3, it would be impossible to distinguish between the three proteins with anti-CBF]
antibodies. To avoid these problems, I decided to make synthetic peptides of small
portions of the CBFl protein to use as epitopes for raising new anti-CBF antibodies in
rabbits. All portions of the CBF 1 protein were initially checked for hydrophilicity (Kyte-
Doolittle), surface probability (Emini) and antigenicity (Jameson-Wolf), using Protean in
DNAStar (Madison, WT) to determine which portion of the protein would be best for the
creation of peptides (Grant, 1972). To be able to detect the CBF] protein specifically, a
C-terminal sequence which is divergent between the three CBF proteins (Gilmour et al,
1998) and should only detect CBF 1 protein was used. To generate antibodies to detect all
three CBF proteins, an N-terminal epitope, that is conserved in all three CBF proteins

(Gilmour et al, 1998) was selected. Antibodies were made as described in 3 .2.4.

3. 3. 6 ALL TESTED PRE-[A/flt/IUNE SERA FROM RABBITS HAVE ANTIBODIES TO NON-SPECIFIC

PIANT PROTEINS

In order to avoid high levels of cross reactivity to non-CBF plant proteins in the
pre-immune serum, the pre-immune sera from prospective anti-CBFI antibody producing
rabbits were screened by immunoblot analysis. Of six rabbits tested, all had antibodies to
plant proteins present in total soluble protein extracts fiom nonacclimated and acclimated
leaves (data not shown). However, as the banding pattern indicated that there were no
distinctive bands at ~29 kDa, the predicted size of the CBF proteins, the four rabbits that
gave pre-immune serum that was least reactive to total soluble leaf extracts were selected.

For the N-terminal peptides #58 and #62 were selected, for the C-terminal peptides, #63

101

and #64 were selected (see sections 3.2.4, and 3.2.5).

3. 3. 7 ANTISERA RAISED TO SYNTHETIC CBF PEPTIDES RECOGNIZE RECOMBINANT CBF]

PROTEIN, B UT CBF PRO TEINS ARE NO T DETECTED IN PLANT EXTRACTS

All bleeds were titered to determine which dilutions gave maximum visualization
of CBF proteins and minimum background patterns (data not shown). Despite the fact
that immune serum from all four rabbits contained anti-CBF] antibodies that could detect
recombinant CBF] protein, immunoblot analysis with total soluble plant extracts gave
complex “background” banding patterns as seen with F1 and F2 (see 3.2.5) (see Figure
3.5A, C1 and N1). The complex “background” banding patterns resulted in the inability
to definitively detect CBF proteins in total soluble leaf extracts from cold-acclimated

RLD and nonacclimated RLD and A6 (see section 2.2.2) plants.

3. 3. 8 [W UNOAFFINI TY PURIFICATION OF ANTI- PEPTIDE ANTIBODIES REDUC ES

BA CKGRO UND, B UT CBF PROTEINS ARE NOT DETECTED IN PLANT EXTRACTS

As the crude sera from anti-CBF peptide antibodies resulted in complex banding
patterns in immunoblot analyses using plant extracts, antibodies were immunoaffinity
purified to synthetic peptides (see section 3.2.6.4). No differences were seen between
purified and crude sera in immunoblot analyses using total soluble plant extracts, raising
the possibility that the purification process was unsuccessful (see Figure 3 .SA). There are

many possible reasons for the complex banding pattern. One possibility was that it was

102

 

l

2 3
29kDa> '. ‘

w
’W

 

 

 

3...;

 

NoJ° Ab N1: l/10.000
2° Ab :1/5,000 2° Ab :1/5,000 2° Ab :1/50.000
, ‘ 3 4 5 1 2 3 4 5

    

 

 

 

 

 

Figure 3.5. Immunoblot analysis of recombinant and native CBF]

A. Immunoblot analysis with crude and purified antisera against total
soluble plant extracts. Irnmunoblot analyses were conducted as described
(3.2.8.1) using crude N1 (N1) or C1 (C1) (see 3.2.5) and synthetic peptide
purified (see 3.2.6.4) N1 and N2 combined (N-pur) or Cl and C2
combined (C-pur). All primary sera were diluted 1:100,000, the secondary
antibody was IgG purified (see 3.2.8.1 #2) diluted 1:5,000. Lanes are as
follows:

1: 3-day cold-acclimated A6 (see 3.2.1);

2: 3-day cold-acclimated RLD (see 3.2.1);

3: nonacclimated RLD (see 3.2.1).

The 29 kDa marker, the predicted size of the CBF proteins, is indicated
with an arrow.

B. Immunoblot analysis of recombinant and native CBFl protein.
Irnmunoblot analyses were conducted as described (3.2.8. 1) using no
primary antibody (No 1° Ab) or N1 (see 3.2.5) diluted as described. The
secondary antibody was IgG purified (see 3.2.8.1 #2) and diluted as
indicated. Lanes are as follows:

1: 100 pg nonacclimated RLD; 2: 100 pg 3-day cold-acclimated RLD;
3: 300 ng CBFl; 4: 30 ng CBFl;

5: 3 ng CBFl.

The 29 kDa marker, the predicted size of the CBF proteins, is indicated
with an arrow.

103

again due to cross-reactivity of the secondary antibody with proteins in total soluble
protein extracts from leaf tissue. To test this hypothesis, immunoblot analyses were
conducted in the absence of a primary antibody (see Figure 3.5B, “no 1° antibody”
section lanes 1 and 2). There is an interaction between the secondary antibody and
nonspecific proteins in total soluble leaf extracts. This interaction was seen regardless of
whether unpurified goat anti-rabbit antibodies, IgG purified goat anti-rabbit antibodies, or
monoclonal goat anti-rabbit antibodies (see section 3.2.8.1) were used in immunoblot
analyses (data not shown).

One way to reduce or eliminate the interaction of the secondary antibody with
nonspecific plant proteins is to change the initial blocking buffer used. Therefore, I
experimented with difl‘erent blocking agents to see if any of them resulted in a reduction
in background banding. Bovine Serum Albumin, (BSA) (5%) resulted in less background
than 5% non-fat powdered milk, or 10% rabbit serum, therefore 5% BSA was used in all
the following immunoblot analyses (data not shown).

Another potential way to reduce complex banding patterns was to remove the
non-specific plant proteins with which the secondary antibody interacts. To do this, I
conducted immunoblot analyses using total soluble proteins isolated from roots,
protoplasts (kindly donated by Yaopan Mao) and from nuclei (kindly donated by Charlie
Herman). The background banding patterns with the three alternative protein sources
were different than those seen in total soluble leaf extracts, but CBF proteins could still
not definitively be detected (data not shown). Additionally, in the case of the protoplast
extracts, the same complex banding patterns were visible with the pre- and immune

serum, indicating that nonspecific interactions were still occurring.

104

Using the maximum possible dilution of the primary and secondary antibodies is
another way to reduce non-specific banding patterns. Therefore, I wanted to determine
the maximum possible dilution of antibodies where recombinant CBFl protein could still
be visualized. Additionally, I wanted to determine if nonspecific interactions occurred in
immunoblot analyses with total soluble plant extracts using identical dilutions (Figure
3.5B). All three amounts of recombinant CBF 1 protein, 300, 30 and 3 ng, (lanes 3-5) can
clearly be seen in the 2° Ab: 1/5,000 blot. However, under identical conditions,
immunoblot analysis using total soluble plant extracts resulted in very high levels of
background banding patterns (lanes 1 and 2). This level of nonspecific binding would
undoubtedly mask detection of CBF proteins. When the secondary antibody is used at a
dilution of 1: 50,000 (2° Ab: 1/S0,000) there are no background banding patterns seen in
immunoblot analyses against total soluble plant proteins (lanes 1 and 2). However, at this
dilution, 300 ng of CBF recombinant protein, an amount which is predicted to be several
fold greater than that found in 100 pg total soluble leaf extracts, is only just detectable
(lanes 3-5). Therefore, these data indicate that CBF proteins in plant extracts probably
cannot be detected with the present antibodies using immunoblot analysis against total
soluble plant proteins due to nonspecific banding patterns which mask visualization of

the CBF proteins.

3.3.9 RECOMBINANT CBF] Il/[AYBE DETECTED BYIAWUNOPRECIPITATION, BUTNOTIN THE

PRESENCE OF PIANT EXTRAC T S

Detection of the CBF proteins by immunoblot analysis proved to be problematic

105

(see Figure 3.5B). Therefore, I decided to try detecting the CBF proteins by
immunoprecipitation. By enriching for the CBF proteins and reducing the amount of
other plant proteins, it was anticipated that the CBF proteins could be detected. For the
first experiment, I wanted to determine if the CBFl recombinant protein could be
immunoprecipitated using N1, N2, C1 and C2 (see section 3.2.5). Previous work had
indicated that the CBF proteins could adhere to microcentrifuge tubes and cause
background problems (S. Gilmour, E. Stockinger, M. Thomashow, unpublished).
Therefore, non-stick surface microcentrifuge tubes were used for all immunoprecipitation
experiments (VWR Scientific Products, W. Chester, PA). Figure 3.6A shows the results
of an immunoprecipitation experiment performed using 300 ng recombinant CBFl
protein as described (see section 3.2.9.1) with the exception that the proteins were not
pre-cleared with rabbit-serum agarose beads (Sigma) before the experiment was
conducted. The CBF 1 protein is not immunoprecipitated without the addition of protein-
A linked agarose beads (lanes 2-4) (Sigma) indicating that the CBF] protein is not
adhering to the microcentrifiige tubes. Additionally, the CBF] protein is not
immunoprecipitated if water instead of serum is added (lane 5). The pre-immune sera
from N1 (lane 6), N2 (lane 9), C1 (lane 11) precipitated some of the CBFl recombinant
protein, but immune sera from N2 (lane 10) and C2 (lane 14) gave the best recovery of
the recombinant protein. IgG purification of C1 (lane 8)(see section 3.2.6.2) did not seem
to have any effect on the amount of CBF] protein immunoprecipitated (compare to lane
7, unpurified C1). While the sample containing 300 ng recombinant protein loaded on a
tricine gel (Schagger and von Jagow, 1987) appears as a doublet (see Figure 3.6, lane 1),

all of the immunoprecipitated proteins appear to result in one band of a lower molecular

106

A B
234567891011121314, ‘23

l
.r, r
.‘ .
mt m... - «I.

Figure 3.6. Immunoprecipitation of recombinant CBF] protein
followed by immunoblot analysls

A. Immunoprecipitation of recombinant CBFl protein.
Immunoprecipitations were conducted as described (see 3.2.9.1), with the
exception that protein extracts were not precleared, using N1, IgG purified
N1 (3.2.6.2) N2, C1, and C2 (see 3.2.5). 300 ng of recombinant CBF 1
protein (see 3.2.3) 10 pl pre- or immune sera and 10 p1 of protein-A beads
were used in each experiment unless otherwise indicated. Irnmunoblot
analyses were conducted as described (3.2.8.1) using N1 (see 3.2.5) diluted
1/5,000. The secondary antibody was IgG purified (see 3.2.8.1 #2) used at
a dilution of 1/100,000. The 29 kDa size marker, the predicted size of
CBF 1 , is indicated with an arrow. Lanes are as follows:

1: 300 ng recombinant CBF] protein; 2: water (no serum), no beads;
3: N1 pre-immune serum, no beads; 4: N1, no beads;

5: water (no serum); 6: N1 pre-immune serum;

7: N1; 8: N1 IGG purified

9: N2 pre-immune serum; 10: N2;

11:C1pre-immune serum; 12: C1;

13: C2 pre-immune serum; 14: C2.

B. Immunoprecipitation of recombinant CBFl protein in the presence of 7-
day cold-acclimated RLD plant extracts. Immunoprecipitations were
conducted as described (3.2.9.2) using 10 p1 N1 (see 3.2.5) and 10 pl of
protein-A beads. The immunoblot analysis was conducted with C2 diluted
1/ 1,000, the secondary antibody was IgG purified (see 3.2.8.1 #2) diluted
l/30,000. The 29 kDa size marker, the predicted size of CBF 1, is indicated
with an arrow. Lanes are as follows:

1: 3 pg recombinant CBF 1;

2: 3 pg CBF] incubated with 3 mg 7-day cold acclimated protein extracts;
3: 3 mg 7-day cold acclimated RLD protein extract.

107

weight. The reason for this difference is not known.

Recombinant CBFl protein is degraded when added to total soluble plant extracts
in the absence of protease inhibitors. Therefore, I wanted to determine if recombinant
CBF] could be immunoprecipitated in the presence of total soluble plant extracts with
added protease inhibitors (see Figure 3.6B, and section 3.2.7.1 #2 for protease inhibitors
and section 3.2.9.2 for the immunoprecipitation protocol). As a control, 3 pg CBFl
recombinant protein can clearly be detected (see lane 1). However, no CBFl protein can
be detected after immunoprecipitation of 3 pg CBFl recombinant protein in the presence
of 3 mg total soluble plant extracts from 7-day cold acclimated RLD plants using C1 (see
lane 2). Additionally, no CBF proteins were detected after immunoprecipitation using 3

mg total soluble plant extracts from 7-day cold acclimated RLD plants (see lane 3).

3. 3. 10 CBF PROTEINS ARE NOT DETECTED BY [MA/{UNOPRECIPITAT ION FROM PLANT EH RAC T S

As immunoprecipitation of recombinant CBF 1 in the absence of total soluble
plant extracts appeared successfill (see Figure 3.6A) I attempted to immunoprecipitate
native CBF proteins from total soluble plant extracts as described (section 3.2.9.1, see
Figure 3.7A). In this case, all protein extracts were pre-cleared with rabbit serum agarose
(Sigma) to reduce non-specific binding. Surprisingly, in the lanes resulting from pre-
clearing the total soluble plant proteins with rabbit-serum agarose beads (Sigma) ~29 kDa
bands, the predicted size of the CBF proteins, appear to be present (see lanes 2-4. The
white area is due to air bubbles). However, no distinct bands are present in any of the

lanes where immunoprecipitations were conducted against 4-day cold acclimated RLD

108

A 12345678910111213

 

 

29kDa> ' - Tak
5p] 1&1]
B 1 2 3 1 2 3 1 4

 

 

”2‘1

Figure 3.7. Immunoprecipitation of CBF and COR15 proteins from
plant extracts.

A. Immunoprecipitation of CBF proteins from total plant extracts.
Immunoprecipitations were conducted as described (see 3.2.9.1) using 10
pl of pre- or immune sera from N2 or C2 (see 3.2.5) as indicated and 10 pl
of protein-A beads against 2 mg total soluble plant extracts. Plant types
were nonacclimated (NA) and four-day cold acclimated RLD and A6
(CBFl-overexpressing) plants (see 3.2.1). Irnmunoblot analyses were
conducted as described (see 3.2.8.1) using N1 diluted at l/5,000 and IgG
purified secondary antibody (see 3.2.8.] #2) diluted 1;100,000. The 29
kDa size marker, the predicted size of CBF], is indicated with an arrow.
The lanes are as follows:

1: 300 ng recombinant CBF 1 protein; 2: rabbit-serum agarose beads

(RSA), NA RLD;
3: RSA, 4-day CA RLD 4: RSA beads, CA A6;
5: water (no serum), NA RLD; 6: C2 pre-immune, NA RLD;
7: C2, NA RLD; 8: C2, pre—immune, CA RLD;
9: C2, CA RLD; 10: water (no serum), CA A6;
11: C2 pre-immune, CA A6; 12: C2, CA A6 extracts;

13: Ca bleed #4, CA A6 extracts.

B. Immunoprecipitation of recombinant COR15. Immunoprecipitations
were conducted as described (see 3.2.9.2) using 5 or 10 pl pre- or immune
serum (see 3.2.5) as indicated and 10 pl protein-A beads against 100 ng
recombinant COR15a. Immunoblot analyses were conducted with COR15
diluted 1/10,000, secondary antibodies were IgG purified (see 3.2.8.] #2)
and diluted 1/10,000. The lanes are as follows:

1: 100 ng COR] Sam; 2: COR15 pre—immune;
3: COR15 immune; 4:COR 15, 100 ng CORISam in] mg 7-day
CA RLD.

109

(lane 9) and A6 plants (lane 12) using serum from C2 (see 3.2.5). One
immunoprecipitation was conducted using serum from N2 (see 3.2.5) against 4-day cold
acclimated A6 plants (lane 13). No distinct bands are observed at ~29 kDa, the predicted
size of the CBF proteins. The experiment was repeated several times in order to
investigate why the CBF proteins were not seen. Changing the volume in which the
immunoprecipitation was conducted (from 100-400 pl) or the type (C1, C2, N] or N2
(see 3.2.5), unpurified or purified (see 3.2.6.2 and 3.2.6.4) or amount (diluted 1/2,000 to
1/3 0,000) of primary antibody or the type (crude, IgG purified, and monoclonal, see
3.2.8.1) or amount (diluted 1/5,000 to 1/ 100,000) of secondary antibody used in
immunoblot analyses did not result in the ability to detect native CBF proteins (data not
shown). To check if the failure to immunoprecipitate CBF 1 was due to technical
difiiculties, COR] 5m was immunoprecipitated (see Figure 3.7B, section 3.2.9.2). Lane 1
under all headings is 100ng COR15a protein loaded directly onto the gel as a positive
control. Immunoprecipitation of recombinant and native CORISm is clearly possible
under the given conditions (see lanes 3 and 4 under both headings), indicating that it is
not due to technical difficulties that CBF proteins are not detected. Additionally, it
appears as though 10 p1 of anti-CORI 5a serum results in a better recovery of COR15a
than using 5 pl (see “10pl” lanes 3 and 4 as opposed to “Spl” lane 3). However, it is
important to note that the recovery percentage of the COR15a protein in both cases is low
~10-20%. If CBF proteins are recovered at an equally low percentage level, recovery

amounts may be below the limits of protein detection using irmnunoblot analysis.

3. 3. I I 355-METHIONINE LABELLED RECOMBINANT CBF] CAN BE DETECTED BY

110

1AM! UNOPRE C [P] TA TI ON

To increase the level of detection of the CBF proteins, I attempted to radiolabel
the proteins with 35 S-methionine. This eliminates the need to conduct immunoblot
analysis and was expected to allow the detection very low levels of protein, as few as
103- 105 molecules per cell (Harlow and Lane, 1988). Immunoprecipitations were
conducted with proteins isolated from CBFl-overexpressing E. coli (section 3.2.3,
construct pEJ S3 69) grown in the presence of 100 pCi of 35S (section 3.2.10) (see Figure
3.8). Antisera from N2, C2 and F1 (see section 3.2.5) appeared able to immunoprecipitate
E. coli-expressed CBF 1 (Figure 3.8, see lanes 6, 8 and 10 under the “CBFl” heading) .
However, as was seen before, it appears that the pre-immune sera from N2, C2 and F1
also immunoprecipitate CBF] (Figure 3.8, see lanes 5, 7 and 9 under the “CBFl”
heading), with F] pre-immune resulting in almost identical amounts of protein to that
seen in immune sera (Figure 3.8, compare lanes 9 and 10 under the “CBF 1” heading).
Alternatively, the increase in abundance of the putative CBF protein after
immunoprecipitation could be due to nonspecific binding of the protein to the protein-A
beads (Sigma).

The putative CBF] protein immunoprecipitated by F1, N2 and C2 has a
significantly higher molecular weight than the predicted weight of CBF], ~29 kDa. This
is presumably due to the presence of a 27.5 kDa GST tag (section 3.2.3, construct
pEJ S3 69). To ensure that the immunoprecipitated band was recombinant GST-tagged
CBF], total E. coli protein extracts were cleaved with thrombin to remove the GST tag

(see Figure 3.8, lane 3 vs. 1). In the thrombin-cleaved GST-CBF] extracts (Figure 3.7,

111

Vector CBF]

2 56789105678910

 

Illw

4
:1 ‘2! 1'!—

f
1

t

 

 

 

 

t

 

 

Figure 3.8. Immunoprecipitation of recombinant 358-Methionine
labelled CBF] from total E. coli protein extracts.

Proteins were isolated fiom E. coli radiolabelled with 35S-methionine
(see 3.2.10). Immunoprecipitations were conducted (see 3.2.9.1) using
20 pl E. coli protein extract transformed with vector (Vector) or CBF]
(CBFl) and 20 p1 pre- or immune serum fiom F1, N2 or C2 (see 3.2.5)
as indicated and and 30 pl protein-A beads. The GST-tagged CBF]
position and the 29 kDa marker are indicated with arrows. The band in
lane 2 is the 27.5 kDa GST protein Lanes are as indicated:

1: 2 pl extract fiom E. coli transformed with CBF] ;

2: 2 pl extract from E. coli transformed with vector;

3: 5 pl thrombin cleaved extract fiom E. coli transformed with CBF] ;
4: 5 pl thrombin cleaved extract from E. coli transformed with vector;
5: N2 pre-immune;

6: N2;

7: C2 pre-immune;

8: C2;

9: F1 pre-irnmune;

10: F1.

The images for the two panels on the left are from a 9 day exposure to

x-ray fihn at room temperature, images for the two panels on the right
are from an over night exposure at room temperature.

112

lane 3) two bands of ~27 and 29 kDa are present, as would be predicted for the
GST and CBF 1 proteins respectively. Additionally, immunoblot analyses were conducted
on GST-CBF 1 protein extracts before and alter cleaving with thrombin. The bands
detected by immunoblot analysis were located in identical positions to those predicted to
be the two forms of 3 5 S -labelled CBFl (data not shown). Collectively, these data indicate
that sera from F 1, N2 and C2 can immunoprecipitate 35S-methionine labelled

recombinant CBF 1 protein from total E. coli protein lysates.

3. 3. 12 3 5S-METIIIONINE [ABELLED RECOMBINANT CBF] IS DEGRADED IN THE PRESENCE OF

PLANT EXTRAC TS

The ultimate goal of immunoprecipitation experiments with anti-CBF antibodies
was to detect the native CBF proteins in wild type and CBF-overexpressing plants.
Recombinant CBF] protein appeared to be immunoprecipitated from total E. coli lysates
(see Figure 3.8, see lanes 6, 8 and 10 under the “CBF 1” heading). Therefore, the next
step was to determine if the recombinant 35’S -labelled CBF 1 protein could be
immunoprecipitated in a solution containing plant proteins. Specifically, I wanted to
determine if 3 5 S -labelled recombinant CBF 1 was degraded in the presence of plant
extracts as an indication that native CBF proteins would also be degraded. To test this,
Arabidopsis plants were grown on media (see section 3 2.2.2 with the exception that
plants themselves were not radiolabelled) and total soluble proteins were extracted from

non-acclimated RLD and A6 plants as described (section 3.2.9.2). A total of 5 pl of 35S —

113

29 kDa>

 

Figure 3.9. 35S-methionine labelled recombinant CBFl is degraded in
the presence of plants and protoplast extracts.

Incubation of recombinant 35S-methionine labelled CBF l in plant
extracts. Proteins were isolated from CBF] -overexpressing E. coli
radiolabelled with 35S-methionine, thrombin cleaved and incubated with
plant extracts (see 3.2.7.] #2) or protoplast extracts (see 3.2.7.2 for
buffer). A total of 5 pl of E. coli extracts were incubated with plant
extracts, protoplast extracts or water on ice for a total of 3 h. The 29 kDa
size marker, the predicted size of CBF], is indicated with an arrow.
The lanes are as indicated:
1: 5 pl extract from E. coli transformed with CBF] in 30 pl water;
2: 5 pl extract from E. coli transformed with CBFI in 25 p1 plant
extracts;
3: 5 pl extract from E. coli transformed with CBF] in 20 p1
protoplast extracts.

The image is from an 11 day exposure to x-ray film at room temperature.

114

labelled, thrombin cleaved E. coli extracts were added to 30 pl of water, 25 pl total

soluble plant proteins, or 20 pl total soluble protoplast proteins (kindly donated by
Yaopan Mao, for extraction bufi°er see 3.2.7 .2) (see Figure 3.9). Recombinant thrombin
cleaved CBF] protein is degraded in the presence of total soluble plant extracts and
protoplast extracts (Figure 3.8, see lane 2 and 3 compared to lane 1). There appeared to
be slightly more CBF] protein present in the presence of protoplast extracts as compared
to the plant extracts (lane 3) indicating that perhaps the buffer used to extract protoplasts
resulted in less degradation than the other immunoprecipitation bufl‘ers used. When a
second set of incubation experiments were conducted with recombinant CBF 3, the
protein was not degraded in the presence of protoplast extracts (data not shown).
However, when immunoprecipitations were conducted with 20 pl 3 5 S —labelled, CBF]-
overexpressing E. coli lysates spiked with 75 pl total soluble plant extracts, protoplast

extracts or extraction buffer alone little or no CBF protein was detected (data not shown).

3. 3. I 3 3jig-METTIIONINE LABELLED CBF] IS NOT DETECTED BY WUNOPRECIPITATION OF

PLANT EXTRAC TS

Once the most usefiil extraction bufi‘er was determined, an attempt was made to
immunoprecipitate CBF] from total soluble plant extracts. Plants were grown (see
3.2.2.2), labelled, and extracted as described (see 3.2.7.2). As seen in Figure 3.10, very
faint but complex banding patterns were present in all lanes. No differences were

detected between banding patterns generated by using pre-immune or immune sera (for

115

123456789101112

4“

   

29 kDa>

 

 

 

Figure 3.10. Immunoprecipitation of 35S-methionine labelled CBF
proteins from plant extracts

Immunoprecipitations were conducted using 3sS-methionine labelled
plant tissue as described (see 3.2.11 #1 and #2). The plants types were
nonacclimated (NA) and 24-h cold-acclimated (CA) A30a-l, A38b-7
(CBF3-overexpressing), and B16-1 (vector) (see 3.2.1).
Immunoprecipitation of CBF] was conducted with 20 pl of pre (pre)-
or immune sera fi'om N] (N 1) (see 3.2.5) and 45 pl of protein-A beads.
The 29 kDa size marker, the predicted size of CBF], is indicated with
an arrow. The image shown is from a 4 hour exposure using an
intensifying screen. The lanes are as follows:

1: N], CA B16-l; 2: pre, CA B16-l;
3: N1,NA B16-1; 4: pre, NA B16-l;
5: N], CA A38b-7; 6: pre, CA A38b-7;
7: N1, NA A38b-7; 8: pre, NA A38b-7;
9: N], CA A30a-1; 10: pre, CA A30a-l;
ll: N1,NA A30a-1; 12: pre, NA A30a-l.

116

example, compare lane 2 to lane 1, or lane 4 to lane 3). Additionally, the type of tissue
used, (compare lanes 3, 7 and 11) or the state of acclimation (for example, compare lanes
1 and 3) did not seem to have an effect on the banding patterns. This experiment was
repeated numerous times with similar results (data not shown). Overall, CBF proteins
could not be clearly detected. The failure to detect the proteins was not due to the
extraction buffer used as recombinant CBFl protein was successfiilly
immunoprecipitated in the protoplast bufi‘er (data not shown). Immunoprecipitation
experiments were also conducted using 35 S-methionine labelled Arabidopsis protoplasts
(kindly donated by A. Sanderfoot). However, despite repeated attempts, incorporation of
35 S-methionine into total plant proteins remained low, possibly due to bacterial infection
of the plants in tissue culture. In both cases, immunoprecipitations resulted in no visible

banding patterns (data not shown).

3. 3. I4 CBF] IS NOT DETECTED BY [Ail/I UNOPREC [PITA TION OF 32P-ORTHOPHOSPHAT E

LABELLED PLANT EXTRAC T S

Immunoprecipitations using 35S-methionine labelled plants did not result in the
detection of CBF] proteins (see Figure 3.10A), possibly due to the low intensity of the
3 5 S signal. 32P-orthophosphate produces a signal of higher intensity on x-ray film than
35S-methionine, making it more desirable as a label (Harlow and Lane, 1988). Analysis of
the CBF protein sequences indicated that they contain a total of 17 potential
phosphorylation sites, including one putative MAP kinase site (Blom et al, 1999; E.

Stockinger, M. Thomashow, unpublished). Therefore, plants were labelled with 32P-

117

#12 34 56 78 910111213141516,

29kDa>[

 

Figure 3.11. Immunoprecipitation of 32P-orthophosphate labelled
CBF proteins from plant extracts

Immunoprecipitation of CBF proteins from 32P-orthophosphate labelled
total soluble plant extracts. Immunoprecipitations were conducted as
described (see 3.2.12) using 20 pl of pre (pre)- or immune sera fi'om
N2 (N2) (see 3.2.5) and 30 pl of protein-A beads. The plants types
were nonacclimated (NA) and 6-h cold-acclimated (CA) WS, A30a-l
(CBF3-overexpressing), E7l-l (CBF2-overexpressing), and G7a-l
(CBFl-overexpressing) (see 3.2.1). The 29 kDa size marker, the
predicted size of CBF l, is indicated with an arrow. The image shown is
from a 21 day exposure using 2 intensifying screens. The lanes are as
follows:

1: pre, NA WS; 2: N2, NA WS;

3: pre, CA WS; 4: N2, CA WS;

5: pre, NA A30a-1; 6: N2, NA A30a-1;
7: pre, CA A30a-l; 8: N2, CA A30a—l;
9: pre, NA E71-l; 10: N2, NA E71-1;
1]: pre, CA E7l-l; 12:N2, CA E7l-1;
13: pre, NA G7a-l; 14: N2, NA G7a-l;
15: pre, CA G7a-1; 16: N2, CA G7a-1.

118

orthophosphate (see 3.2.12). No distinct bands were seen at ~29 kDa, the predicted
molecular weight of the CBF proteins even after exposure to x-ray film with two
intensifying screens for 21 days (see Figure 3.11). However, very faint bands are present
in the lanes containing immunoprecipitations from nonacclimated WS (lane 2), A30a-1
(lane 6) and G7a-1 (lane 14) (see 3.2.1 for plant types). To confirm that the putative
bands were indicative of real phosphorylation events, the experiment was repeated using
the same types of tissue and the same protocols (data not shown). However, no distinct
bands were seen and the CBF proteins could not be clearly visualized under any
conditions. The failure to visualize the CBF proteins was not due to poor incorporation of
32P-orthophosphate into total soluble proteins as visualization of 4 pl of radiolabelled
protein showed distinct banding patterns indicating that the 32P-orthophosphate had been

incorporated into plant proteins (data not shown).

3 .4 DISCUSSION

The regulation of a transcription factor can give many clues as to the regulation of
the entire signal-transduction pathway of which it is a part. In the cold-acclimation signal
transduction pathway, it had previously been determined that the CBF proteins are likely
activators of the COR genes (Jaglo-Ottosen et al, 1998; Gilmour et al, 1998). Therefore,
the CBF proteins appear to play key roles in initiating the changes in gene expression,
such as activating the COR genes. Increasing our knowledge as to the level of CBF
protein accumulation under nonacclimating and acclimating conditions, and whether the

proteins are regulated by modification would increase understanding of the regulation of

119

the entire cold-induced signal transduction pathway.

Transcription factors, such as the CBF genes, can be regulated at many steps.
They can be transcriptionally regulated or post-transcriptionally regulated, the
polypeptide can be translationally regulated and post-translationally regulated through
modification (Gallie, 1993; Schwechheimer et al, 1998). Transcription factors can also be
regulated by their physical location in the cell (Gallie, 1993; Schwechheimer et al, 1998)
or regulation can occur through a combination of some or all of the above. RNA analysis
has indicated that CBF RNA accumulation increases within 15 min of a low temperature
stimulus, peaks at 2-4 h of acclimation, and remains at a higher steady state level for as
long as plants are under acclimating conditions (Gilmour et al, 1998). These data would
indicate that the CBF genes are regulated in part, either transcriptionally, or post-
transcriptionally by RNA stability. However, regulation through increased RNA
accumulation under acclimating conditions does not rule out the possibility that the
production and/or activity of the proteins are also translationally or post-translationally
regulated under acclimating conditions as well.

One way to determine if the CBF proteins are translationally or post-
translationally regulated is to examine the amount of protein and the location of the
protein under both nonacclimating and acclimating conditions. Any changes in amounts
or location of the CBF proteins would indicate if and how the proteins are regulated.
Additionally, it would be useful to determine whether overexpression of CBF], CBF2 or
CBF 3 resulted in changes in the amount of protein accumulation and/or the location of
the protein under either condition. To answer these questions, the ability to recognize and

detect the three CBF proteins was required. To this end, three types of anti-CBF]

120

antibodies were generated (see 3.2.5): those against full length recombinant CBF] (Fl
and F2, E. Stockinger, M. Thomashow, unpublished) those against an N-terminal peptide
of CBF] (N l and N2) and those to a C-terminal peptide of CBF 1 (C1 and C2). Peptides
were selected as antigens to generate antibodies which should either detect CBF 1 alone
(N—terminal peptide) or all three CBF proteins (C-terminal peptide). The ability to
distinguish between the CBF] protein and all three CBF proteins combined would allow
for detection of any difierences in location or protein accumulation between CBF 1 and
the other CBF proteins.

Sequence analysis of the CBF2 protein indicates the presence of 17 potential
phosphorylation sites; 9 on serine residues, 6 on threonine residues and 2 on tyrosine
residues (http: //www.cbs.dtu.dk/sewices/NetPhos/-; Blom et al, 1999). Additionally, one
of the sites, which is conserved with all three CBF proteins is a potential map kinase
phosphorylation site (B. Stockinger, M. Thomashow, unpublished). This raised the
possibility that the CBF proteins were post-translationally modified. Therefore, I wanted
to determine if the CBF proteins are modified by a phosphorylation event under
nonacclimating or acclimating conditions. Additionally, by investigating the
phosphorylation state of CBF proteins under both conditions in transgenic plants, the
mechanism by which overexpression of the CBF proteins activates the COR genes
without a low temperature stimulus could be better understood.

Despite numerous and repeated attempts to detect the native CBF proteins in
plants, the proteins were never definitively detected. There are many possible reasons as
to the cause of the difficulties. The first reason could be that the anti-CBF] antibodies

were not effective at detecting the CBF proteins. This does not seem likely as all three

121

types of antisera were able to detect the recombinant CBF] and CBF 3 proteins in
immunoblot analyses (see Figure 3.1, Figure 3.4).

One important and confounding factor associated with the creation of anti-CBF
antibodies was with the pre-immune sera from all rabbits used. The pre-immune sera
contained antibodies to many proteins in total soluble plant extracts (see Figure 3.2B) and
also appeared to specifically recognize the CBF] protein (see Figure 3.3 and Figure
3.4B). The recognition of CBF 1 could very well be due to the AP2 DNA binding domain
portion of the protein given that there are predicted to be ~90 AP2 domain containing
proteins in the Arabidopsis genome alone (Riechmann and Meyerowitz, 1998). Ifother
plant species, particularly those used to produce rabbit feed, also contain numerous AP2
DNA binding domain proteins, then it is very likely that rabbits are exposed to, and
produce antibodies to, AP2 DNA binding domain proteins. Given that pre-immune serum
is used as a negative control for immune serum, the presence of antibodies in the pre-
immune serum that recognize CBF proteins could confound the interpretation of data.
Additionally, in immunoprecipitation experiments, total soluble plant extracts are pre-
cleared by incubating them with rabbit-serum bound to agarose beads (Sigma). Ifthe
rabbit serum contained antibodies that recognized the AP2 domain portion of the CBF
proteins, it is possible that some or all of the native CBF proteins may have bound to
these beads reducing the possibility of detecting the CBF proteins afier
immunoprecipitation.

One way to avoid the problem of anti-CBF or anti-AP2 domain antibodies in pre-
irnmune serum would be to produce different types of antibodies. Monoclonal antibodies

to the CBF proteins or CBF-specific peptides would be one way to eliminate antibodies

122

to other plant proteins. It is due to the high cost and technical difficulty involved with
their production (Harlow and Lane, 1988) that monoclonal antibodies were not initially
created. Another way to minimize the amount of non-specific plant protein antibodies in
pre-immune serum would be to generate antibodies in another species, such as chickens.
As chickens consume a different type of feed and have beaks instead of fleshy mouths
like rabbits, they would presumably be exposed to far fewer plant proteins, and therefore
make far fewer antibodies to plant proteins than rabbits.

Another possible difficulty in detecting the protein could be due to the fact that
the protein is rapidly degraded. Regulation of transcription factors through conditional or
constitutive degradation has been observed previously (Varshavsky, 1997). There is also
evidence to support the idea that the CBF proteins are rapidly degraded. In all
experiments where recombinant CBF] was exposed to total soluble plant extracts, the
protein was degraded, either partially or totally (see Figure 3.6B and Figure 3.9). This
degradation occurred regardless of which extraction buffer was used, or if a complete
mixture of protease inhibitors was added. Finally, the tight regulation of the COR genes
could be an indication that the CBF proteins are rapidly degraded. If plants are removed
from acclimating conditions and returned to warm temperatures, COR RNA
accumulation decreases rapidly, and is virtually undetectable within eight h after a
temperature shifi (Hajela et al, 1990). This rapid decrease in RNA accumulation could be
the result of the rapid degradation of the CBF proteins under nonacclimating conditions.
However, until the protein is detected, whether or not it is rapidly degraded will remain
unknown.

One interesting outcome of the 32P-orthophosphate labelling experiments is the

123

possible indication that the CBF proteins may be phosphorylated under nonacclimating
conditions (see Figure 3.1] lanes 2, 6 and 14). While the result was not repeatable, a faint
signal was seen under nonacclimating conditions. If so, this would suggest that the CBF
proteins are inactivated under nonacclimating conditions by a phosphorylation event.
Inactivation of a transcription factor by phosphorylation is not uncommon. The
phosphorylation event can prevent the import of the factor to the nucleus, as is seen with
SW15, the binding of the factor to DNA, as is seen with Oct] and Myogenin, or may
prevent transactivation as is seen with ADR] (Hunter and Karin, 1992). These data seem
contrary to the evidence that protein kinases are activated under acclimating conditions in
alfalfa (Jonak, 1996), if the data can be taken as an indication that homologous kinases
are activated in Arabidopsis under acclimating conditions. However, it is possible that
these kinases fiinction to activate some other protein upstream of the CBF proteins, or
proteins in another cold-induced pathway. One possibility is that under acclimating
conditions, a cold-induced kinase functions to activate a phosphatase which then activates
the CBF proteins by de-phosphorylation. However, until the experiment can be repeated,
it will remain unknown as to whether or not the CBF proteins are actually phosphorylated
under nonacclimating conditions.

While I was never able to detect the CBF proteins in plant extracts, I believe the
effort to do so was worthwhile. In order to understand fully how the CBF genes function
in the cold acclimation signal transduction pathway, the quantity and location of the
proteins must be determined. Additionally, it would be of great importance and interest to
determine if the proteins are regulated by modification under either nonacclimating or

acclimating conditions. Without this information, complete understanding of the cold

124

acclimation pathway will not be possible. Despite the fact that these protein-related
questions could not be definitively answered, they remain interesting questions and

hopefully through some future effort, the answers will be resolved.

3 .5 REFERENCES

An, G. 1987. Binary TI-vectors for plant transformation and promoter analysis. Methods
in Enzymology. 153: 292-303.

Artus, NN, Uemura, M, Steponkus, PL, Gilmour, SJ, Lin, C, Thomashow, MF. 1996.
Constitutive expression of the cold-regulated Arabidopsis thaliana COR15a gene affects
both chloroplast and protoplast freezing tolerance. Proc. Natl. Acad Sci. USA. 93: 13404-
13409.

Ausubel, FM, Brent, R. Kingston, RE, Moore, DD, Seidman, JG, Smith, JA, Struhl, K.
1987. Current Protocols in Molecular Biology. John Wiley and Sons, NY.

Baker, SS, Wilhelm, KS, Thomashow, MF. 1994. The 5’ region of Arabidopsis thaliana
cor15a has cis-acting elements that confer cold-, drought- and ABA-regulated gene
expression. Plant Mol Biol. 24: 701-713.

Bar-Peled, M and Raikhel, NV. 1996. A method for isolation and purification of specific
antibodies to a protein fused to the GST. Anal Biochem. 241: 140-142.

Blom, N, Gammeltoft, S, Brunak, S. 1999. Sequence and structure-based prediction of
eukaryotic protein phosphorylation sites. J. Mo]. Bio]. 294: 1351-1362

Gallic, DR. 1993. Posttranscriptional regulation of gene-expression in plants. Annu. Rev.
Plant Physio]. Plant Mo]. Biol. 44: 77-105.

Gilmour, SJ, Hajela, RK, Thomashow, MF. 1988. Cold-acclimation in Arabidopsis-
thaliana. Plant Physiol. 87: 745-750.

Gilmour, SJ, Sebolt, AM, Salazar, MP, Everard, JD, Thomashow, MF. 2000.
Overexpression of the Arabidopsis CBF 3 transcriptional activator mimics multiple
biochemical changes associated with cold acclimation. Plant Physiol. In press

Gihnour, SJ, Zarka, DG, Stockinger, EJ, Salazar, MP, Houghton, JM, Thomashow, MF.
1998. Low temperature regulation of the Arabidopsis CBF family of AP2 transcriptional
activators as an early step in cold-induced COR gene expression. Plant J. 16: 433-442.

125

Grant, GA. 1972. Syntheic Peptides, a user’s guide. W.H. Freeman and Co, NY.

Guy, CL, Niemi, KJ, Brambl, R. 1985. Altered gene-expression during cold-acclimation
of spinach. Proc. Natl. Acad Sci. USA. 82: 3673-3677.

Hajela, RK, Horvath, DP, Gilmour, SJ, Thomashow, MF. 1990. Molecular-cloning and
expression of COR (cold-regulated) genes in Arabidopsis-thaliana. Plant Physiol. 93:
1246-1252.

Harlow E, Lane, D. 1988. Antibodies, a laboratory mannual. Cold Spring Harbor
Laboratory.

Hondred, D, Wadle, DM, Titus, DE, Becker, WM. 1987. Light-stimulated accumulation
of the peroxisomal enzymes hydroxypyruvate reductase and serine glyoxylate

aminotransferase and their translatable messenger-RN As in cotyledons of cucumber
seedlings. Plant Mo]. Biol. 9: 259-275.

Horvath, DP, McLamey, BK, Thomashow, MF. 1993. Regulation of Arabidopsis-
thaliana L (I-Ieyn) COR78 in response to low-temperature. Plant Physiol. 103: 1047-
1053.

Hunter, T, Karin, M. 1992. The regulation of transcription by phosphorylation. Cell. 70:
375-3 78.

Jaglo-Ottosen, KR, Gilmour, SJ, Zarka, DG, Schabenberger, O, Thomashow, MF. 1998.
Arabidopsis CBF] overexpression induces COR genes and enhances freezing tolerance.
Science. 280: 104-106.

Jonak, C, Kiegerl, S, Ligterink, W, Barker, PJ, Huskisson, NS Hirt, H. 1996. Stress
signaling in plants: A mitogen-activated protein kinase pathway is activated by cold and
drought. Proc. Nat]. Acad. Sci. USA. 93: 11274-11279.

Laemmli, UK. 1970. Clevage of structural proteins during the assembly of the head of
bacteriophage T4. Nature. 227: 680-685.

Lin, C, Thomashow, MF. 1992. DNA-sequence analysis of a complementary-DNA for
cold-regulated Arabidopsis gene COR15 and characterization of the COR-15 polypeptide
Plant Physiol. 99: 519-525.

Liu Q, Kasuga, M, Sakuma, Y, Abe, H, Miura S, et al. 1998. Two transcription factors,
DREBI and DREBZ, with an EREBP/AP2 DNA binding domain separate two cellular
signal transduction pathways in drought- and low-temperature-responsive gene
expression, respectively, in Arabidopsis. Plant Cell. 1391-13406.

Mittlera, K Fenga, X, Cohena, M. 1991. Plant Cell. 10: 461-474.
126

Nordin, K, Heino, P, Palva, ED. 1991. Separate signal pathways regulate the expression
of a low-temperature-induced gene in Arabidopsis-thaliana (L) Heynh. Plant Mol. Biol.
16: 1061-1071.

Riechmann, JL, Meyerowitz, EM. 1998. The AP2/EREBP family of plant transcription
factors. Biol. Chem. 379: 633-646.

Rose M, Botstein, D. 1983. Construction and use of gene fusions to lacZ (beta-
galactosidase) that are expressed in yeast. Methods Enzymol. 101: 167-180.

Sambrook, J, F ritsch, EF, Maniatis, T. 1989. Molecular Cloning: A Laboratory Manual
(Cold Spring Harbor Lab. Press, Plainview, NY), 2nd Ed.

Schagger, H and von Jagow, G. 1987. Tricine sodium dodecyl-sulfate polyacrylamide-gel
electrophoresis for the separation of proteins in the range from l-kDa to 100-kDa.Anal.
Biochem. 166: 368-3 79.

Schwechheimer, C, Zourelidou, M, Bevan, MW. 1998. Plant transcription factor studies.
Annu. Rev. Plant Physiol. Plant Mo]. Bio]. 49: 127-150.

Shinwari, ZK, Nakashima K, Miura, S, Kasuga, M, Seki, M, et al. 1998. An Arabidopsis
gene family encoding DRE/CRT binding proteins involved in low-temperature-
responsive gene expression. Biochem. Biophys. Res. Commun. 250: 161-170.

Stockinger, EJ, Gilmour, SJ, Thomashow, MF. 1997. Arabidopsis thaliana CBF]
encodes an AP2 domain-containing transcriptional activator that binds to the C-
repeat/DRE, a cis-acting DNA regulatory element that stimulates transcription in
response to low temperature and water deficit. Proc. Natl. Acad Sci. 94: 1035-1040.

Thomashow, MF. 1999. Plant cold acclimation: Freezing tolerance genes and regulatory
mechanisms. Ann. Rev. Plant Physiol. Plant Mol. Biol. 50: 571-599.

Varshavsky, A. 1997. The ubiquitin system. Trends Biol. Sci. 22: 383-3 87.

Wang, H, Datla, R, Georges, F, Loewen, M, Cutler, AJ. 1995. Promoters from kin] and
COR6. 6, 2 homologous Arabidopsis-thaliana genes - transcriptional regulation and gene—
expression induced by low-temperature, ABA, osmoticum and dehydration. Plant Mol
Biol. 28: 606-615.

Welin, BV, Olson, A, Palva, ET. 1995. Structure and organization of 2 closely-related
low-temperature-induced DHN/LEA/RAB-like genes in Arabidopsis-thaliana L. Heynh
Plant Mol Biol. 26: 131-144.

Yamaguchi-Shinozaki, K, and Shinozaki K. 1993. The plant hormone abscisic-acid
mediates the drought-induced expression but not the seed—specific expression of RD22, a

127

gene responsive to dehydration stress in Arabidopsis-thaliana. Mol. Gen. Genet. 236:
33 1-3 40.

Yamaguchi-Shinozaki, K, and Shinozaki K. 1994. A novel cis-acting element in an

Arabidopsis gene is involved in responsiveness to drought, low-temperature, or high-salt
stress. Plant Cell 6: 251-264.

128

4. CHAPTER 4: Overexpression of Arabidopsis CBFI, CBF2 or CBF3
in Brassica napus var. Westar results in increased BN gene expression

and freezing tolerance

4. 1 INTRODUCTION

Freezing temperatures, drought and other environmental stresses limit the
geographical areas of crop production and are estimated to cause up to a 60% reduction
in maximum crop yield annually (Levitt, 1980). Some plants withstand fieezing
temperatures by cold acclimating, the process by which plants increase in fi'eezing
tolerance after exposure to low, non-freezing temperatures. In 1970, Weiser proposed
that, in addition to being associated with numerous physiological changes, such as
increases in proline levels and total soluble sugars, cold acclimation is associated with
changes in gene expression (W eiser, 1970). The identification of the COR (cold-
regulated) genes in Arabidopsis -also called L77 (low temperature-induced), KIN (cold-
inducible), RD (responsive to desiccation) and ERD (early dehydration-inducible)-, a
small family of novel, highly hydrophilic polypeptides that are induced during both cold
acclimation and drought stress, was direct evidence of such changes (Nordin et al, 1991;
Wang et al, 1994; Welin et al, 1994), and indicated the presence of a cold and drought
induced signal transduction pathway. Further research has shown that COR gene

homologues are present in diverse freezing tolerant plant species, such as the BN28 (Orr

129

et al, 1992) and BN115 (Weretilnyk et al, 1993) genes in canola, the wcs120 gene in
wheat (Houde et al, 1992), the H VA] gene in barley (Heino et al, 1990, Hong et al, 1992)
and the casIS gene in alfalfa (Monroy et al, 1993).

In Arabidopsis, the COR genes are transcriptionally regulated through a cold- and
drought inducible promoter element, called the CRT/DRE (C-Repeat/Dehydration
Responsive Element). The DRE element, TACCGACAT, was initially identified by
Yamaguchi-Shinozaki and Shinozaki (1994) as being responsive to low temperature and
dehydration. Later work showed that the core of the DRE, the CRT, CCGAC, was
present in multiple copies in the promoters of COR6.6 (Wang et al, 1995), COR15 (Baker
et al, 1994) and COR78 (Horvath et al, 1993, Yamaguchi-Shinozaki and Shinozaki,
1993). Deletion analysis with the COR15a promoter showed that cold-induced activation
of a reporter gene occurred when the CRT sequence was present in the promoter (Baker
et al, 1994). Interestingly, this sequence is not only found in multiple copies in COR gene
promoters in Arabidopsis, but also in the promoters of some of the cold induced genes in
other species, such as the BN115 gene in canola (Jiang et al, 1996), and the wcsIZO gene
in wheat (Ouellet et a], 1998), giving rise to the hypothesis that cold and drought
inducible gene expression may be highly conserved among different plant species.

A major breakthrough in understanding how plants sense and respond to low
temperatures was the isolation of the CBF/DREB (CRT/DRE Binding F actors/Drought
Response Element Binding factors) family of transcription factors, which bind to the
CRT/DRE element and activate transcription under cold acclimating conditions
(Stockinger et al, 1997; Gilmour et al, 1998; Liu et al, 1998; Shinwari et al, 1998). The

three genes, CBF], CBF 2 and CBF 3 have highly similar amino acid sequences (~85%)

130

contain AP2-like DNA binding domains, acidic activation domains, and have been shown
to activate transcription in yeast (Stockinger et al, 1997; Gilmour et al, 1998).
Constitutive overexpression of CBF], CBF2, or CBF 3 in Arabidopsis results in
constitutive COR gene expression and increased freezing tolerance under both
acclimating and non-acclimating conditions (Jaglo-Ottosen et al, 1998; Liu et al, 1998;
Gilmour et al, 2000; S. Gilmour and M. Thomashow, unpublished). The GenBank
database contains several expressed sequence tags (ESTs) of sequences with a high level
of similarity to the CBF/DREB family from diverse plant species such as canola, rice and
tomato. This presents the possibility that homologous cold acclimation pathways exist in
diverse plant species and creates the possibility of increasing freezing tolerance through
genetic engineering.

Here, results are presented indicating that the CBF-activated signal transduction
pathway is conserved in canola. An Expressed Sequence Tag (EST) with high similarity
to the CBF genes in B. napus is present in GenBank. My results indicate that induction
pattern for this putative CBF homologue from B. napus var. Westar, (BnCBF), is highly
similar to that of the Arabidopsis CBF genes; transcript accumulation occurs within 15
min of a low-temperature stimulus and remains at a higher steady state level for at least
24 h of cold treatment. The increased BnCBF RNA levels are followed by increased
expression of COR gene homologues (called BN genes) indicating the possibility of a
CBF-induced signaling cascade similar to that seen in Arabidopsis (Thomashow, 1999).
Additionally, constitutive overexpression of Arabidopsis CBF], CBF 2 or CBF 3 in canola
results in increased BN gene expression and increased freezing tolerance as compared to

control plants under both non-acclimating and acclimating conditions. This increase was

131

seen in both cuttings and seedlings of transgenic canola plants. Additionally, as compared
to control plants, CBF-overexpressing canola plants show increases in total soluble
sugars under both nonacclimating and acclimating conditions, but significant increases in
proline levels only under acclimating conditions. These data indicate that overexpression
of the Arabidopsis CBF genes not only activates the BN genes, but also activates other
known cold-acclimation inducible pathways such as those involved in increasing proline
and total soluble sugars (see 1.4.5). These increases in proline and soluble sugars have
also been seen in CBF3-overexpressing Arabidopsis plants and are associated with
increases in freezing tolerance (Gilmour et al, 2000). I also investigated whether
constitutive CBF-overexpression resulted in increased salt tolerance, as salt, drought and
freezing stress all result in similar damage to plants (Thomashow, 1999; see 1.3).
Additionally, overexpression of CBF 3 in Arabidopsis was found to result in increased
freezing tolerance as well as increased tolerance to salt (Liu et al, 1998). However, I was
not able to conclusively determine whether overexpression of the CBF family of
transcription factor resulted in an increase in salt tolerance. Here I propose that the CBF-
signal transduction pathway is conserved from Arabidopsis to its close relative, canola,
and that overexpression of Arabidopsis CBF genes can increase freezing tolerance in the

agronomically important crop, canola.

4.2 MATERIALS AND METHODS

132

4. 2. I PLANT GROWTH:

Brassica napus var. Westar seeds, a spring variety of canola, were planted in 10.5
x 10.5 or 12.5 x 12.5 cm pots containing Baccto Planting Mix (Michigan Peat, Houston,
TX). One to five plants per pot were germinated in controlled environment chambers at
20-22° C under continuous cool white fluorescent illumination of 100-150 pmolrn'zs’l
light intensity as described (Gilmour et al, 1988). After ~2 weeks, plants were screened
for the presence of the transgene (see 4.2.3 for details) and thinned to one plant per pot.
Plants or cuttings were fertilized every 1-2 weeks with Peters 20-20-20, diluted as
indicated by the manufacturer (Scotts, Chicago, IL). To induce cold acclimation, 4-6
week old plants were placed at 4° C under continuous fluorescent illumination of ~50
pmolm'zs'l for three weeks, or 4-6 week old cuttings were placed under the same
conditions for two weeks. For the time course experiments, a total of 18 individual 4-6
week old wild-type plants were used. Tissue was collected from two individual plants at
room temperature after which the remaining 16 individual plants were placed at 4° C
under continuous fluorescent illumination of ~50 pmolm'zs'l, and tissue was collected
from two plants at each of the time points indicated. To analyze proline and sugar levels,
plants were grown as above, with the exception that up to three plants were grown per

pot.

133

4. 2. 2 TRANSFORMATION:

Canola plants that constitutively overexpress CBF], CBF2, or CBF 3 were created
by Susanne Kleff. Briefly, the cDNA sequence fiom each of the CBF genes were cloned
into the pGA643 expression vector, which contains the NPTH reporter gene, under the
control of the constitutive CaMV 35S promoter (An, 1987). Agrobacterium tumefaciens
strain GV3101 was transformed with these plasmids by electroporation for use in
cotyledonary petiole transformation (Maloney et al, 1989) into canola plants (kindly
donated by W. Keller, Natl. Research Council, Canada and M. Moloney, University of
Calgary, Canada).

Regenerated plants were analyzed for the presence of the T-DNA by detecting
expression of the NPTII gene (encoding kanamycin resistance) using the NPTH ELISA
kit (5 Prime — 3 Prime, Inc. Boulder, CO). Regenerated shoots which tested positive for
the presence of the T-DNA were further analyzed for the presence of the CBF transcript
and the expression of the cold-regulated genes, BN115 and BN28. These To plants were

self-pollinated, and the T1 generation seeds were collected and used for further studies.

4. 2.3 TRANSGENIC PLANT SELECTION:

As plants in the T1 generation are not homozygous for the T-DNA, all plants were
tested for either the expression of the NPTH gene using the NPTII ELISA kit (5 Prime —
3 Prime, Inc. Boulder, CO) or for the presence of the NPTII gene using Polymerase

Chain Reaction (PCR) before use in experiments. Primers were designed using the NPTII

134

gene sequence found in GenBank and were synthesized at the Michigan State University
Macromolecular and Structure Facility (5’: TGGAGAGGCTATTCGGCTA, 3’:
CACCATGATATTCGGCAAG). PCR was carried out in 25 pl reactions containing ~40
pM of each primer, ~10 ng of genomic DNA, {isolated using the WizardR genomic DNA
purification kit (Promega, Madison, WI)}, SOpM dNTPs and 5% DMSO using a
RoboCycler Gradient 96 Temperature Cycler (Stratagene, La Jolla, CA). Conditions
were: 5 min at 94°C, then 30 cycles of: 1 min at 94°C, 1 min at 62°C, and 2 min at 72°C,
followed by an additional 5 min at 72°C. The entire reaction mixture was combined with
DNA loading buffer (Sambrook et al, 1989) and 0.1% ethidium bromide, visualized on
1% TBE agarose gels and checked for the presence of a ~600 bp band diagnostic of the

NPTH gene.

4. 2. 4 CUTTINGS

For the experiments with cuttings, multiple plants from independent CBF-
overexpressing T1 lines were analyzed for high levels of BN28 protein expression.
Selection resulted in eight CBF] lines (#9, 10, 11, 21, 26, 47, 55, 97), seven CBF2 lines
(#40, 45, 53, 54, 65, 101, 113) and seven CBF3 lines (#25, 87, 108, 120, 129, 130, 145).
Three independent lines for each of the CBF], CBF 2, and CBF 3 constructs were selected
and used to make cuttings: #9, 10 and 26 for CBF]; #45, 53 and 65 for CBF2; and, #25,
87 and 145 for CBF 3 . Additionally, cuttings were made of two independent vector lines,

#23 and 161, as controls. To make the cuttings, leaves with petioles and some meristem

135

 

 

tissue were excised from plants with a razor blade, dipped into Bontone rooting powder
(Bonide Products, Inc, Yorkville, NY) and placed in 8.5 x 8.5 cm pots containing moist
soil. The pots were covered with plastic wrap, and set in controlled grth chambers
under the conditions described above (see 4.2.1). After ~4 days the plastic wrap was slit
and after another two days, it was completely removed. Cuttings were allowed to grow

for 4-6 weeks or until new leaf tissue formed before being used in experiments.

4. 2. 5 SEEDLINGS

Seedlings from the three independent T1 lines generated by transformation of the
CBF], CBF 2 or CBF 3 constructs described above (see 4.2.4) were analyzed for further
study. Additionally, five independent vector lines (23, 74, 161, 163 and 165), one line
that no longer contained a copy of the vector, (28) and non-transformed wild type plants
were used as controls. As above, all transgenic CBF-overexpressing, or vector control
lines were analyzed for the presence of the NPTII gene either by ELISA or PCR before

use in experiments (see 4.2.3).

4. 2. 6 RNA HYBRIDIZA TION:

4. 2. 6. 1 RNA isolation, northern transfer and hybridization

RNA was isolated using TRIZOL reagent (GibcoBRL, Grand Island, NY) as

recommend by the manufacturer, except that ~200 mg canola tissue was ground to

136

powder with liquid N2 prior to the addition of TRIZOL reagent. Northern transfers were
prepared and hybridized as described (Stockinger et al, 1997). A total oflO pg of total
RNA was used for hybridization with the CBF] and BnCBF probes and 5 pg total RNA

was used when BN115 or BN28 were used as hybridization probes.
4.2.6.2 Generation of DNA probes

The hybridization probe for detecting all three Arabidopsis CBF transcripts was
on the XhoI/XbaI fragment from pSJG6 (S. Gihnour, M. Thomashow, unpublished)
which contains the full-length Arabidopsis CBF] coding sequence. Due to the high
similarity of the three CBF genes, the CBF] probe hybridizes to transcript fi'om all three
CBF genes and was used to detect all three CBF transcripts. Plasmids containing BN28
(Orr et al, 1992) and BN115 (W eretilnyk et al, 1993) were kindly donated by Jas Singh.
For hybridization probes, pBN28 and pBN115 were individually digested with EcoRI and
the resulting fiill-length sequences were band isolated as described (Sambrook et al,
1989). The probe for BnCBF (the Brassica napus CBF homologue) was made by PCR
amplification of the sequence from genomic B. napus var. Westar genomic DNA isolated
using the WizardR genomic DNA purification kit (Promega, Madison, WI). Primers for
the BnCBF gene (GenBank accession #AF 084185) were created using the Primer Select
program in DNAStar (Madison, WI) and were synthesized at the Michigan State
University Macromolecular and Structure Facility (5’: GGTTACGTTAGGCGGAGAGT,

3’: GGACGGCGGCGGCAAAAG). PCR was carried out in 25 pl reactions containing

137

~40 pM of each primer, ~10 ng of genomic DNA, SOpM dNTPs and 5% DMSO using a
RoboCycler Gradient 96 Temperature Cycler (Stratagene, La Jolla, CA). Conditions
were: 2 min at 94°C then 30 cycles of: l min at 94°C, 1 min at 53°C, 1 min at 72°C’
followed by an additional 5 min at 72°C. Total reactions were mixed with DNA loading
buffer (Sambrook et al, 1989) and 0.1% ethidium bromide, visualized on 1% agarose
TBE gels after which the fragment of interest was band isolated and purified as described

(Sambrook et al, 1989).

4. 2. 6. 3 Membrane washing and stripping

Membranes were washed as described (Stockinger, 1997). When required,
membranes were stripped by adding boiling SDS buffer (0.1x SSPE and 0.5% SDS) and

shaking at 80°C until no fiirther radioactivity was detected on the membrane.

4. 2. 7 IWUNOBLOT ANALYSIS

Total protein was extracted by grinding frozen tissue (about 300 mg) in 300 pl
extraction buffer containing 50 mM Tris-HCl (pH 8.0), 5% glycerol, 100 mM KC1, 1.5%
(wt/vol) polyvinyl-pyrrolidone, after which insoluble material was removed by
centrifugation at 13,000 x g for 20 min at 4° C. The protein concentration in the

supernatant was determined using the Bradford dye-binding assay (Bio-Rad, Hercules,

138

CA) anle0 pg of total soluble protein per sample was fractionated by 10% tricine
SDS/PAGE (Schagger and von Jagow, 1987), and transferred to 0.1 pm nitrocellulose
membranes by electroblotting (Towbin et al, 1979) as described (Artus et a], 1996).
BN28 protein was detected using antiserum kindly given by A. J ohnson-Flanagan
(diluted 1: 5000) (Boothe et al, 1995) and visualized using the ECL system (Amersham

Buckinghamshire, UK).

4. 2. 8 PROLINE ANALYSIS

Proline was isolated using standard protocols as modified by Gilmour (Gilmour et
al, 2000). Briefly, 50 mg of freeze-dried tissue harvested from 3-16 plants of each line
were suspended in 5 ml of water at 80°C for 15 min. Samples were shaken for 1 h at
room temperature then allowed to stand over night at 4°C. Samples were filtered through
glass wool, after which proline levels in three replicates of each sample were measured
by an acid nyhydrin assay. An unbalanced analysis of variance (ANOVA) of the samples
was done using SAS PROC GLM [SAS Institute, SAS/STAT User’s Guide, Version 6

(SAS Institute, Cary, NC, 1989)].

4.2.9 TOTAL SOLUBLE SUGAR ANALYSIS

Total soluble sugars were isolated using standard protocols as modified by

Gilmour (Gilmour et al, 2000). Briefly, 50 mg of freeze-dried tissue harvested from 3-16

139

 

plants of each line were suspended in 5 ml of 80% ethanol at 80°C for 15 min. Samples
were shaken for 1 h at room temperature then allowed to stand over night at 4°C.Three
replicates of each sample were measured by a phenol sulfuric acid assay. An unbalanced

ANOVA of the samples was conducted as described above (see 4.2.8).

4. 2. 10 ELECTROLYTE LEAKAGE ASSA YS

Electrolyte leakage freeze tests were conducted as described (Gilmour et al, 2000;
see 2.2.6) with the modification that one individual plant or cutting was selected as a
representative of a given transgenic line. Tissue was removed from all healthy looking
leaves of a cutting or the smallest two leaves of a seedling using a 6-mm paper punch and
three to four punches were used in each of the three replicate samples for each
temperature point. Data shown is either that of individual cuttings or plants or the
combined data of individual cuttings or plants. For the nonacclimated and acclimated
seedling data, an unbalanced AN OVA of the combined leakage values at each given
temperature was performed using SAS PROC GLM [SAS Institute, SAS/STAT User’s
Guide, Version 6 (SAS Institute, Cary, NC, 1989)]. The temperature at which 50% of

electrolytes were leaked (ELso) were determined as described in 2.2.6.

4. 2. 1 1 SALT STRESS EXPERIMENTS

In the first experiment, water was withheld from cuttings for 1 day prior to the

addition of NaCl. Cuttings were placed in 150 mM NaCl for 7 days, rinsed, returned to

140

water and allowed to recover for four weeks before photographs were taken (Experiment
conducted by S. Klefl). Seedlings were salt stress using one of two similar protocols. In
the first protocol, seedlings were flushed with 1 l of 150 mM NaCl, then placed in trays
containing 150 mM NaCl for six days. They were then flushed with 1 1 200 mM NaCl
and placed in trays containing 200 mM NaCl for another seven days. Plants were then
flushed with l l deionized water, placed in water containing Peters 20—20—20 as
recommended by the manufacturer (Scotts, Chicago, IL) and allowed to recover for four
weeks before photographs were taken. In the second protocol, seedlings were flushed
with 1 l 200 mM NaCl and placed in trays containing 200 mM NaCl for ten days. Plants
were then flushed with 1 l deionized water, placed in water containing Peters 20-20-20 as
recommended by the manufacturer (Scotts, Chicago, IL) and allowed to recover for four

weeks before photographs were taken.

4.3 RESULTS

4.3.1 THE B. NAPUS CBF GENE HASA SIMIIAR INDUCTION PATTERN TO THE ARABIDOPSIS

CBFGDwm

Amino acid sequences from the CBF genes were used to search the GenBank
database for other sequences with high levels (greater than 50%) of similarity, . The
sequence for an EST from B. napus encoding for a protein with high similarity to the

CBF proteins was found (see Figure 4.1). To determine if this putative CBF homologue

141

CBF2
BnCBF

CBF2
BnCBF

CBF2
BnCBF

CBF2
BnCBF

CBF2
BnCBF

CBF2
BnCBF

10 20 30 40
l l l l

MNS F SAFS EMFGSDYES PVSSGGDYS PK LATSC PKKPAGR
MTS FSTFSEMLGSEYESPVTLGG EYCPKLAASCPKKPAGR

SJO 610 710 8'0

KKFRETRHPIYRGVRQRNSGKWVCELREPNKKTRIWLGTF
KKFRETRH PVYRGVRL RNSGKWVCEVREPNKKS RIWLGT F

99 100 110 120
QTAEMAARAHDVAAIALRGRSACLNFADSAWRLRIPESTC
LTAEIAARAHDVAAIALRGKSACLNFADSAWRLRIPETTC

130 140 150 160

1 1 LJ
AKEIQKAAAEAALNFQDEMCHMTTDAHGLDMEETLVEAIY
PKEIQKAAAEAALAFQAEINNIllD‘HGLDMEETIVEAIF

170 1%) 190 290
TPEQSQDAFYMDEEAMLGMSSLLDNMAEGMLLPSPSVQWN
T- EENNDVFYMDEESMLEMPALLASMAEGMLLPPPSVH FG

210
YN FDVEGDDDVS LWSY
H NYD FDGDADVS LWSY

 

Figure 4.1. Amino acid sequence alignment of Arabidopsis CBF2
and BnCBF from Brassica napus

Alignment of the amino acid sequences from Arabidopsis CBF2 and
BNCBF, the putative CBF homologue in canola The amino acid
sequences encoded for by CBF2 and BNCBF were aligned using
Megalign (DNAstar Inc, Madison, WI). All divergent amino acids
between the two sequences are indicated with orange. This image
contains color.

142

from canola (BnCBF) was also cold induced, a time course of BnCBF RNA accumulation
at 4°C was performed (Figure 4.2). Tissue was harvested from wild type canola plants
under nonacclimating conditions and immediately frozen in liquid nitrogen after which

the remaining plants were then put into acclimating conditions and tissue was harvested
after 0.25, 0.5, 1, 2, 4, 12 and 24 h at 4°C (see 4.2.1). Total RNA was extracted and
BnCBF RNA levels were analyzed by northern hybridization analysis (see 4.2.6). BnCBF
RNA accumulation increased after 0.5 h at 4°C, reached a maximum level between two
to four h at 4°C, and returned to a lower steady state level for up to 24 h at 4°C when the
experiment was discontinued. To determine if the COR gene homologue, BN115, also
had increased RNA accumulation under the same conditions, northern membranes were
hybridized with a BN115 probe (see 4.2.6). BN115 RNA accumulation began to increase
after ~4 h and continued to increase until the experiment was discontinued after 24 h. The
BN115 RNA increased ~ 3.5 h after the BnCBF RNA suggesting the presence of a
BnCBF-induced signaling cascade in canola that is highly similar to that seen with the

CBF and COR genes in Arabidopsis (Thomashow, 1999).

4. 3 .2 GENERATION 0F CANOLA PLANTS THA T CONSTIUTI VELY EXPRESS ARABIDOPSIS CBF 1,

CBF2 0R CBF3

Overexpression of CBF], CBF 2 or CBF 3 in Arabidopsis results in constitutive
COR gene expression and increased freezing tolerance under both nonacclimating and
acclimating conditions (Jaglo-Ottosen et al, 1998; Liu et al, 1998; Gilmour et al, 2000; S.

Gilmour and M. Thomashow, unpublished). To detemiine if a similar increase in freezing

143

Hoursat4°C
0 1 2 4 12 24

 

 

.25 .5
BNCBF . *.

BN115

 

 

 

   

Loading control

Figure 4.2. Time course of BNCBF and BN115 RNA
accumulation

BNCBF and BN1 15 RNA accumulation. Total RNA was isolated
from leaves of wild type canola plants acclimated at 4° C for the
times indicated above the lanes (see 4.2.] and 4.2.6.1). BNCBF and
BN115 transcripts were detected as described (see 4.2.6). The loading
control panel is the ethidium bromide stained agarose gel containing
RNA to which the BNCBF and BN115 probes were hybridized in the
panels above.

tolerance could be produced in an agronomically important species, canola, B.
napus var. Westar plants that constitutively express the Arabidopsis CBF], CBF2 or
CBF 3 genes were generated (see 4.2.2). Specifically, a cDNA copy of each individual
Arabidopsis CBF was cloned into the binary vector pGA643 (An, 1987) under the control
of the CaMV3SS promoter and transformed into B. napus var. Westar plants using a
cotyledonary petiole transformation method (see 4.2.2). In T1 plants resulting from self-
fertilized To plants, immunoblot analysis indicated that nine lines, three lines for each
CBF, had high levels of BN28 protein accumulation. These lines were selected for further
analysis: #9, 10 and 26 for CBF], #45, 53 and 65 for CBF2 and #25, 87 and 145 for
CBF 3 (data not shown). Additionally, five vector lines, which contained the pGA643
vector, #23, 74 and 161, 163 and 165, one transformed line that no longer contained a
copy of the vector, #28, and non-transformed wild type plants were used for negative
controls. As plants in the T1 generation are not homozygous, all plants except for line #28
and wild type plants were tested for the presence of the T-DNA by assaying for the

NPTII gene before use in experiments (see 4.2.3).

4. 3. 3 OVEREXPRESSION OF THE ARABIDOPSIS CBF GENES IN CANOLA RESUT S IN INCREASED

BN28 AND BN1 15 TRANSCRIPT ACC UMULA TION

The Arabidopsis CBF family of transcription factors can activate transcription of
the COR genes in Arabidopsis (Jaglo-Ottosen et al, 1998; Gilmour et a1, 1998; Liu et al,
1998; Gilmour et al, 2000). To determine if overexpression of the Arabidopsis CBF

genes was able to initiate transcription of the COR gene homologue BN28 in canola,

145

Vector CBF] CBF 2 CBF 3
23 161 9 10 26 45 53 65 25 87 145
NANANANANANANANANANANA

_. -— a 092 0

Loading MUWOO.DCDO“-

 

 

 

  
   
 

Figure 4.3. Accumulation of Arabidopsis CBF and BN RNA in
individual canola cuttings

CBF and BN28 RNA accumulation in individual nonacclimated (N)
and two-week cold acclimated (A) cuttings. Total RNA was isolated
fi'om the transgenic cuttings indicated above the lanes as described
(see 4.2.1 and 4.2.6.1). CBF and BN28 transcripts were detected as
described (see 4.2.6). The loading control panel is the ethidium
bromide stained agarose gel containing RNA to which the CBF and
BN28 probes were hybridized in the panels above.

146

individual cuttings of the nine CBF-overexpressing lines were probed for
accumulation of CBF and BN28 transcripts under both acclimating and nonacclimating
conditions (section 4.2.6, see Figure 4.3). The cDNA for CBF] was used as the
hybridization probe to detect all three Arabidopsis CBF transcripts (see 4.2.6.2) but not
the BnCBF transcript as the CBFI probe did not hybridize with BnCBF RNA to a
significant extent under conditions used (S. Kleff, K. J aglo and M. Thomashow,
unpublished results). The levels of CBF RNA accumulation under nonacclimating and
acclimating conditions did not seem to correlate well (i.e., in the CBFI-overexpressing
transgenic line 10, RNA levels appear to increase under acclimating conditions; in the
CBFI-overexpressing transgenic line 9, levels appear to be equal under both conditions;
and in the CBF2-overexpressing transgenic line 45, RNA levels appear to decrease under
acclimating conditions) indicating that individual cuttings may not have consistent CBF-
transgene expression levels. To determine if the changes in CBF RNA levels isolated
from the same individual plants were more consistent to levels seen in Arabidopsis, lower
in the warm and higher in the cold, RNA was isolated from the same individual plants
under both sets of conditions. Tissue from three to twelve individual plants from six
transgenic and two control lines was isolated under nonacclimating and acclimating
conditions and total RNA samples were probed for CBF], BN28 and BN115 RNA
accumulation (Figure 4.4). CBF transcript levels are increased in both nonacclimated and
acclimated transgenic plants compared to control plants under identical conditions. BN28
and BN115 transcripts also accumulate to a greater extent in the CBF-overexpressing
lines under nonacclimating conditions as compared to the control lines. However, in

contrast to RNA accumulation under nonacclimating conditions, the amount of BN1 15

147

Vector CBF] CBF2 CBF 3

28 74 10 26 45 65 25 87
NANANANANANAN ANA

 

 

CBFl
Loading

BN28
Loading

BN115

_m_-

L.—
Loading ---.--..--------

 

Figure 4.4. Accumulation of Arabidopsis CBF and RN RNA in
pooled canola plants

CBF, BN28 and BN115 RNA accumulation in pooled nonacclimated
(N) and three-week cold acclimated (A) plants. Total RNA was
isolated fi'om the pooled plants indicated above the lanes as
described (see 4.2.] and 4.2.6.1). The total number of pooled plants
used to extract RNA was: 9 of #28, 7 of #74, 9 of #10, 9 of #26, 3 of
#45, 7 of #65, 7 of #25 8 of #87. BNCBF, BN28 and BN115
transcripts were detected as described (see 4.2.6). The loading
control panels are the ethidium bromide stained agarose gels
containing RNA to which each probe was hybridized in the panel
above.

148

and BN28 transcript in acclimated transgenic plants does not appear to be greatly
increased over that of acclimated control lines. This could indicate that maximum RNA

accumulation is reached in the CBF-overexpressing lines under acclimating conditions.

4. 3. 4 OVEREXPRESSION OF THE ARABIDOPSIS CBF GENES IN CANOLA RESUT S [N INCREASED

BN28 PROTEIN AC C UMULA TION

Overexpression of the Arabidopsis CBF genes results in increased BN28 and
BN115 RNA accumulation (see Figure 4.3 and Figure 4.4). To determine if the levels of
BN28 transcript accumulation are representative of BN28 protein accumulation, protein
extracts from individual nonacclimated and cold-acclimated CBF-overexpressing and
control cuttings under were analyzed by immunoblot analysis (see 4.2.7) (Figure 4.5A).
Under nonacclimating conditions, no protein accumulation is seen in extracts isolated
from control cuttings. However, BN28 protein accumulation is observed in extracts from
nonacclimated CBF-overexpressing lines, approximately equivalent to that of two-week
cold-acclimated control cuttings, although there is variation between the lines. In extracts
isolated from two-week cold-acclimated cuttings, there are increased levels of BN28
protein in transgenic cuttings as compared to the controls. A very similar pattern is seen
when conducting the same experiments on nonacclimated and three-week cold-
acclimated seed-grown plants (Figure 4.5B) indicating that overexpression of the
Arabidopsis CBF genes results in increased BN28 protein accumulation under both

nonacclimating and acclimating conditions, in both types of plant tissue.

149

Control CBF] CBF2 CBF3
23 161 9 10 26 45 53 65 25 87 145

 

         
  
   

Nonacclimated
7 a .

Cold-Acclimated

 

 

 

Control CBFI ( {BEZ ( 13E:

 

Figure 4.5. Accumulation of BN28 protein in individual CBF-
overexpressing canola cuttings and plants

Total soluble protein was extracted from individual nonacclimated and
cold-acclimated cuttings and seedlings grown as described (see 4.2.]
and 4.2.7). The amount of BN28 protein accumulation was determined
by immunoblot analysis using 100 pg total soluble protein extracts and
antiserum raised against the BN28 polypeptide (Boothe et al, 1995) as
described (see 4.2.7).

A BN28 protein accumulation in individual nonacclimated and two-
week cold acclimated cuttings.

B. BN28 protein accumulation in individual nonacclimated and three-
week cold acclimated seed grown plants.

150

4. 3. 5 0 VEREXPRESSION OF THE ARABIDOPSIS CBF GENES IN CANOLA RESULTS IN INCREASED

AC C UM ULA T ION OF PROLINE AND TOTAL SOLUBLE S UGARS

Overexpression of CBF 3 in Arabidopsis results in increased accumulation
of proline and total soluble sugars under both nonacclimating and acclimating conditions
(Gilmour et al, 2000). To determine if there are similar increases in the accumulation of
proline and soluble sugars in CBF 1- CBF2-or CBF3-overexpressing canola plants, both
types of small molecules were assayed. Figure 4.6 shows the graphical representation of
the total amounts of free proline (upper graphs) and soluble sugars (lower graphs) under
both nonacclimating (graphs on lefi) and acclimating (graphs on right) conditions in
pg/mg dry tissue. None of the CBF-overexpressing lines have significant increases in
proline as compared to control lines under nonacclimating conditions, while all three
types of overexpressing lines have significant increases under acclimating conditions
where p<0.001 (see 4.2.8). In regards to the amounts of total soluble sugars, the CBF-
overexpressing lines all have different p values compared to the control lines under
nonacclimating conditions. For CBFI, p=0.056, for CBF 2, p=0.018 and for CBF 3,
p=0.09. Under acclimating conditions, all three CBF-overexpressing lines have
significant increases in the amount of total soluble sugars as compared to control lines
where p<0.001 (see 4.2.9). Table 4.1A shows proline accumulation in pg/mg of dry
tissue from all CBF-overexpressing lines combined and all vector lines combined under
both nonacclimating and acclimating conditions. Under nonacclimating conditions, the

proline levels of transgenic plants are not significantly higher than those of control plants.

151

Nonacclimated 35 3-Week Acclimated

 

 

 

  
  

 

 

 

  

 

 

 

 

16*rrr‘*“*”**‘ 307- =7-
14 , 4' - E ,
12 E, _
s g \
6 -
4

.Control 2

ECBFI 0

ICE” 65 so

ICBF3 9(5) 7 70 7- \V
29 * 60 * A
40 50 fl .

pg/mg g8 40 /=\

25 ' 30
i5) 20 A
150’ 10
0 15!.

0
Total soluble sugars

Figu re 4.6. Accumulation of total soluble sugars and proline in control
and CBF], CBF2 or CBF3-overexpressing canola plants

Pooled individuals of CBF 1, CBF 2 and CBF3-overexpressing plants were
used for both nonacclimated and 3-week cold-acclimated samples.
Replicate samples fiom the same plant tissues were used for both the sugar
(4.2.9) and proline (4.2.8) extractions. CBF], CBF2 and CBF3 designate
the combined results of CBFI- (9 of #10, 9 of #26); CBF2—(3 of #45, 7 of
#65); and, CBF3- (7 of #25 8 of #87) overexpressing plants respectively.
Control designates combined results from control plants (12 of WT, 9 of
#28, 7 of #74). The amount of sugars or proline in pg/mg dry tissue are
shown graphically. Error bars indicate standard error. The amount of free
proline in CBF-overexpressing plants was not significantly different from
control plants under nonacclimating conditions, whereas under
acclimating conditions, all were significantly different at p<.001. The
amount of total soluble sugars in CBF-overexpressing plants as compared
to control plants had P values of .056 CBF], .018 for CBF2 and .09 CBF3,
whereas all the acclimated samples are significantly different at p<.001.

152

Table 4.1. Accumulation of total soluble sugars and proline in
combined CBF], CBF 2 and CBF3-overexpressing canola plants

 

 

 

 

 

 

 

 

A NonAcc Acc
CONT CBF CONT CBF
CONT 2.1 i 1.8 P=.1562 P =. 0012 P<.0001
NonAcc
CBF 5.5 i 1.5 P=. 0277 P<.0001
CONT 10.2 i 1.5 P<.0001
Acc
CBF 23.6 i 1.3
B NonAcc Acc
CONT CBF CONT CBF
CONT 19.6 i 3.3 P=. 0089 P <. 0001 P<.0001
NonAcc
CBF 30.7 :1: 2.4 P<.0001 P<.0001
CONT 51.3 i 3.5 P<.0001
Acc
CBF 72.4 i 2.3

 

 

Pooled individuals of CBF], CBF2 and CBF3-overexpressing plants were
used for both nonacclimated and acclimated samples. Replicate samples
from the same plant tissues were used for both the sugar and proline
extractions. CBF designates combined results of all CBF] , CBF 2 and
CBF3 over expressing plants, (9 of #10, 9 of #26, 3 of #45, 7 of #65, 7 of
#25 8 of #87). CONT designates combined results from all control plants
(12 of WT, 9 of #28, 7 of #74). NonAcc designates nonacclimated and
Acc designates three-week cold-acclimated. The amount of sugars or
proline in pg /mg dry tissue i standard error are represented on the
diagonal in bold. P values for the comparisons of sugar or proline levels
are indicated in the intersecting cells in italics.

A. Amount of proline in combined CBF-overexpressing and control
plants.

B. Amount of total soluble sugars in combined CBF-overexpressing and
control plants.

153

 

 

However, under acclimating conditions, there is a statistically significant increase in
proline levels in transgenic plants as compared to control plants. Table 4.1B shows the
total soluble sugar accumulation in pg/mg dry tissue from all CBF-overexpressing lines
combined and all vector lines combined under both nonacclimating and acclimating
conditions. The levels of total sugars are significantly higher in combined transgenic

plants as compared to control plants under both temperature regimes.

4. 3. 6 OVEREXPRESSION OF ARABIDOPSIS CBF GENES IN CANOLA RESULTS IN INCREASED

FREEZING TOLER4NC E

Overexpression of the three CBF genes in Arabidopsis results in increased
freezing tolerance under both nonacclimating and acclimating conditions as determined
by electrolyte leakage analysis (Jaglo-Ottosen et al, 1998; Liu et al, 1998; Gilmour et al,
2000; S. Gilmour and M. Thomashow, unpublished). To determine if overexpression of
the Arabidopsis CBF genes in canola plants resulted in increased freezing tolerance,
electrolyte leakage tests were conducted on both cuttings and seedlings, as seen in Figure
4.7 A-E. Individual cuttings were used as representatives of transgenic lines and tissue
was taken from all healthy looking leaves as described (see 4.2.10). Data shown are fiom
one representative electrolyte leakage test using tissue from nonacclimated cuttings
(Figure 4.7A) and two-week cold-acclimated cuttings (Figure 4.7B). Under both
nonacclimating (Figure 4.7A) and two-week cold-acclimating conditions (Figure 4.7B),
CBFI, CBF 2 and CBF3 -overexpressing cuttings show a dramatic increase in freezing
tolerance. For example, in Figure 4.7A, comparison of the amount of electrolytes leaked

154

Figure 4.7. Electrolyte leakage analysis of wild type and transgenic
canola plants and cuttings

Leaves from nonacclimated and cold acclimated (see 4.2.1) cuttings
(4.2.4) and seedlings (4.2.5), were frozen to the temperatures indicated
and the extent of cellular damage was estimated by measuring
electrolyte leakage as described (see 4.2. 10). Error bars indicate the
standard deviations of the three replicates of each data point unless
otherwise indicated.

A. Electrolyte leakage analysis using individual nonacclimated
cuttings of CBF] (#10), CBF 2 (#53) or CBF 3 (#145) overexpressing
plants and two vector lines (#23 and #161). Color in this image.

B. Electrolyte leakage analysis using individual two-week cold-
acclimated cuttings of CBF] (#10), CBF 2 (#53) or CBF3 (#145)
overexpressing plants and one vector line (#161). Color in this image.

C. Electrolyte leakage analysis using individual nonacclimated CBF]
(#26), CBF2 (#65) or CBF 3 (#87) overexpressing plants, two vector
lines (#161 and #165) and WT plants. Color in this image.

D. Electrolyte leakage analysis using individual three-week cold-
acclimated CBF] (#10) or CBF2 (#53) overexpressing plants, two
vector lines (#23 and #161) and WT plants. Color in this image.

E. Electrolyte leakage analysis using combined data from all
individual nonacclimated CBF 1, CBF 2 or CBF 3 overexpressing plants
(23 total) or control plants (10 total). Error bars indicate the standard
deviation of all the replicates used at each data point. Except for 0° C,
all temperatures are significantly different at p<.01 as determined by
unbalanced one-way AN OVA (see 4.2.10).

F. Electrolyte leakage analysis using combined data from all individual
three-week cold-acclimated CBF], CBF 2 or CBF 3 overexpressing
plants (12 total) or control plants (8 total). Error bars indicate the
standard deviation of all the replicates used at each data point. Except
for 0° C, and —3° C, all temperatures are significantly different at p<.01
as determined by unbalanced one-way ANOVA (see 4.2. 10).

155

Figure 4.7 cont

A. Individual nonacclimated cuttings

' i537 (CBF2I1

l
l
l

I 10 (CBFl) 1
n145(CBF31
B161 (Vect) 1

. '23 (V00!) ,1

 

 

onus—.34 «:02».—

8. Individual cold-acclimated cuttings

 

 

-9 -10 -11

-8

C. Individual nonacclimated plants

BF3)

 

1 I87 (c
. I65 (CBF2)

. u\\\\\\\\\\\\\\\\

1,

I 26 (CBFl)

165 (Vect)

1
1
1
l

.91

 

99310..— 2.3th

\\\\\\\\\\\\\\\\\\\\\\\\

Temperature in Degrees Celsius

156

Figure 4.7 cont.
D. Individual cold-acclimated plants

120

—n—-

o:—

o9
1

Percent Leakage

E. Combined nonacclimated plants

110 1'

100' '
90 1*
80 :-
70 ..
60
50
40
30
20
10
0

.1

-2 -3 -4
F. Combined cold-acclimated plants

110‘—
100*

Percent Leakage

\IOOO
COO

O
O

50L
40:

1,7 ,7 L,
301 -
101.i
o L.§,gfi , - ’r 1 >
0 -3 -5 -7 -9 -11 -13 -15

Temperature in Degrees Celsius

Percent Leakage

  

 

157

 

   
  

'IWT

12116.1

1ECBF

 

 

1 E CBF Ave
1 I Control

1.7537(CBFT2)T‘
, 1- 10 (CBF1)‘

I23 (Vect) 1

9°31

_7

V6.

1 I Control 1

 

 

from the three CBF-overexpressing lines to that of the control lines at —5° C
shows that the transgenic lines have ~42% leakage while the control lines have ~95%
leakage. A similar comparison with two-week acclimated cuttings (7B) at—6° C shows
~20-40% leakage for transgenic plants and 65% leakage for the control line. These data
indicate that significantly more membrane damage has occurred in the control lines than
in the transgenic lines at the given temperatures under both nonacclimating and
acclimating conditions. Figure 4.7 C shows one representative electrolyte leakage test
using tissue from nonacclimated plants and Figure 4.7 D shows a representative
electrolyte leakage test of three-week cold-acclimated plants. As seen with cuttings, there
is a dramatic increase in fieezing tolerance seen in CBF], CBF2 and CBF3-
overexpressing plants under nonacclimating (7C) and three-week cold-acclimating (7D)
conditions. To test whether overexpression of the CBF genes results in a statistically
significant increase in freezing tolerance, Figure 4.7 E shows the combined data of all
individual nonacclimated CBF-overexpressing plants (23 total) and all nonacclimated
control plants (13). Figure 4.7 F shows the combined data for all CBF-overexpressing
plants (12) and control lines (8) after three-weeks of cold-acclimation. Unbalanced one
way analysis of variance (ANOVA) indicates that there is a statistically significant
difference in electrolyte leakage over all temperatures tested except for at 0° C for
nonacclimated plants, and a statistically significant difference in electrolyte leakage over
all temperatures tested except at 0° C and —3° for three-week cold-acclimated plants
where t< 0.01 (see 4.2.10). To quantitate the increase in freezing tolerance, ELso values
(the temperature at which 50% leakage is reached) were determined for CBF-

overexpressing plants as described (see 4.2. 10) and are shown in Table 4.2. When

158

Table 4.2. Comparison of BL,» values of combined CBF], CBF2 or
CBF3 overexpressing and control plants

CONT

NonAcc
CBF

CONT

Acc
CBF

 

 

 

 

NonAcc Acc

CONT CBF CONT CBF

-2.1 :h .34 P<.0001 P <.0001 P<.0001
(10)
-4.7 :1: .40 P<.0001 P<.0001
(23)
-8.1 :1: .42 P<.0001
(8)
-l2.7 :1: .52
(12)

EL50 values were calculated using combined data from all individual
nonacclimated and cold-acclimated CBF], CBF 2 or CBF 3
overexpressing plants and compared ANOVA (see 4.2.10). ELSO i
standard error are represented on the diagonal in bold. P values are in
italics and indicated in the intersecting cells, the number in the
parenthesis indicates the number of individual plants combined.
CONT designates combined control plants. CBF designates all
combined CBF], CBF2 or CBF 3 overexpressing plants. NonAcc
designates nonacclimated and Acc designates three-week cold-

acclimated.

159

 

comparing ELso values by AN OVA, all leakage amounts are significantly different from
each other at p< .0001. Combined, these data clearly indicate that overexpression of
Arabidopsis CBF], CBF2 or CBF 3 in canola results in a significant increase in freezing

tolerance.

4. 3. 7 OSMOTIC STRESS TOLERANCE 0F CBF], CBF2 OR CBF3 OVEREXPRESSING CANOLA

LINES

Freezing, drought, and osmotic stress all result in cellular dehydration and cause
damage to membranes in plants (Thomashow, 1999; see 1.3). Additionally, Liu et al
(1998) found that overexpression of CBF 3 in Arabidopsis resulted in increased freezing,
drought and osmotic stress tolerance. As overexpression of the Arabidopsis CBF genes in
canola results in increased freezing tolerance (see 4.3.6), it was of interest to determine if
it also resulted in increased tolerance to salt. Plants and cuttings were exposed to 150
mM-200mM NaCl for 7 to 13 days, returned to deionized water and observed for
recovery as described (see 4.2.11). While CBF-overexpressing plants and cuttings had
increased survival after salt stress compared to control plants and cuttings in two
experiments (Table 4.3) no differences in survival between transgenic and control plants
were seen in the third experiment (Table 4.3). Further experiments need to be conducted
to determine if overexpression of the Arabidopsis CBF genes in canola results in

increased tolerance to salt.

160

Table 4.3. Survival of CBF-overexpressing and control cuttings and
plants after salt stress.

Experiment number
1 2 3

 

 

CBF] 2/4 3/7 1/5
CBF2 4/7 4/5 2/2

CBF 3 5/6 4/6 4/6

Total CBF 11/17 11/18 7/13

 

control 0/3 0/7 8/14

Survival of CBF-overexpressing plants and cuttings after salt stress.
Cuttings (experiment 1) and seedlings (experiments 2 and 3) were
subjected to 150 mM to 200mM NaCl for 7-13 days as described (see
4.2.11). Recovery after salt stress was determined approximately 1
month after the removal of the salt stress. Cuttings and seedlings were
considered “recovered” if tissue remained green and growth continued
in the month following the removal of the osmotic stress. The numbers
shown in the table are the number of individuals that recovered over
the total number of individuals salt stressed.

CBF] , CBF 2 and CBF3-overexpressing plants and cuttings used are
those described in 4.2.4 and 4.2.5. Control lines include all lines in
4.2.4 and 4.2.5 and wild type plants.

161

4. 3. 8 OVEREXPRESSION 0F CBF 1, CBF2 OR CBF3 IN CANOLA DOES NOT CA USE GROSS

PHENO TYPIC CHANGES

Overexpression of CBF], CBF 2 or CBF 3 can cause a dwarf phenotype in CBF-
overexpressing Arabidopsis plants with high transgene expression levels (Liu et al, 1998;
Gilmour et al, 2000; A. Sebolt, S. Gilmour, M. Thomashow, unpublished). To determine
if overexpression of the Arabidopsis CBF genes had an efl‘ect on the phenotype of canola
plants, plants were grown to 27 days of age and photographs of representative plants
were taken (Figure 4.8). There were no gross phenotypic differences observed between
transgenic and control plants under the described growing conditions (see 4.2.1) with the
exception that some transgenic plants appear to be a darker shade of bluish-green than
control plants. The basis for this difference is not known. Additionally, when transgenic
plants were grown from seeds to maturity, no dramatic changes in the overall life cycle of
plants or time to flowering were observed. It should be noted that plants were grown in
grth chambers under 24 h of artificial light (see 4.2.1). These conditions may not result

in phenotypes representative of those seen by plants grown in the field.

4.4 DISCUSSION

The recent sequencing of the Arabidopsis genome suggests that there are many
AP2 domain-containing proteins. Estimates indicate that there may be 90 AP2 domain-
containing proteins in all (Riechmann and Meyerowitz, 1998). Sequencing efforts in

other plant species have shown that AP2 domain containing proteins are conserved in

162

161(kanor) 9((13F1)

145((13F3)

 

Figure 4.8. Photograph of transgenic CBF-overexpressing and
control canola plants

One representative plant grown as described (see 4.2.1) of each CBF-
overexpressing line is shown at 27 days of age: #9 for CBF 1, #53 for
CBF 2 and #145 for CBF3. As a comparison, the vector control line
#161 is shown. No gross phenotypic differences were observed under
the growing conditions used between CBF-overexpressing plants and
the vector lines except that CBF-overexpressing plants sometimes
appeared to be a slightly darker shade of bluish green than control
plants. This image is in color.

163

angiosperms and have been identified in both monocots and dicots (Riechmann and
Meyerowitz, 1998). In fact, when searching the entire database, sequences significantly
similar to the Arabidopsis CBFI AP2 domain were found, including sequences in Oryza
sativa, Prunus armeniaca, Catharanthus roseus, Solanum tuberosum, Nicotiana
sylvestris, Stylosanthes hamata and others. Presumably more exist and will continue to be
found as sequencing efforts advance in other plant species. AP2 domains are DNA-
binding domains first identified in the APETALA2 protein that, to date, have only been
found in plants (Riechmann and Meyerowitz, 1998). While not all of the large family of
AP2 domain-containing proteins have been assigned functions, many, if not all, of the
characterized proteins appear to play regulatory roles: APETALA2 in flower
development (Jofuku et al, 1994); TINY in plant cell size (Wilson et al, 1996); atERFs in
ethylene signaling in Arabidopsis (Hao et al, 1998; Fujimoto et al, 2000); EREBS in
ethylene signaling in tobacco (Ohme-Tagaki and Shinshi, 1995); and the CBF genes in
activating cold regulated gene induction (Stockinger et al, 1997; Liu et al, 1998; Jaglo-
Ottosen et al, 1998; Gilmour et al, 1998; Medina et al, 1999; Gilmour et al, 2000) to
name a few.

When generating a dendrogram of all AP2 domain-containing proteins in
GenBank using Clustalx, a subset of proteins were found to cluster closely with the CBF-
cold induced proteins in Arabidopsis indicating the possibility of functional similarity (T.
Wagner, K. Amundsen, M. Thomashow, unpublished). Indeed, when the cold-
acclimation induction pattern of the putative CBF homologues in canola, wheat and rye
were investigated, a similar pattern to that seen in Arabidopsis was observed: little or no

RNA is present under nonacclimating conditions, after ~30 min RNA accumulation

164

increases and remains at a higher steady state level for a least 24 h of acclimation (see
4.3.]; K. Amundsen, M. Thomashow, unpublished). If the COR gene homologues, BN28
in canola and COR39 in wheat and rye, are then used to probe the same RNA samples, a
delayed increase in RNA accumulation is seen after ~ 4hrs, strongly indicating that the
same cold-acclimation pathway present in Arabidopsis may be present in agronomically
important crop species (K. Amundsen, M. Thomashow, unpublished). This potential
conservation of induction pathways in diverse plant species is exciting for two reasons: 1)
it suggests that the CBF pathway may have evolved before the crop species diverged,
and; 2) it suggests the possibility that overexpression of the CBF genes and CBF-
homologues in different crops species may be a viable way to increase freezing tolerance
of agronomic crops.

To a limited extent, the CBF genes from closely related species appear to be
interchangeable. Overexpression of the Arabidopsis CBF genes in canola resulted in
increased expression of the BN28 and BN115 genes (Figure 4.3, Figure 4.4 and Figure
4.5), a statistically significant increase in the accumulation of soluble sugars under both
nonacclimating and acclimating conditions (Figure 4.6 and Table 4.1A), and a
statistically significant increase proline levels under cold acclimating conditions (Figure
4.6 and Table 4.1B). Additionally, when assayed for increased freezing tolerance by
electrolyte leakage assays, a statistically significant increase in freezing tolerance was
seen under both nonacclimating and acclimating conditions, as determined by comparing
the temperature at which 50% leakage occurs, ELso (Table 4.2). The increase in freezing
tolerance was also seen when comparing leakage of all CBF-overexpressing plants to

control plants over all temperatures tested except for at 0° C for nonacclimated plants,

165

and at 0° C and —3° C for acclimated plants (see Figure 4.7E and 4.7F).

Despite the fact that the CBF genes were under the control of the constitutive 35S
promoter, there appears to be a “super induction” of the downstream genes in the CBF
regulon under acclimation conditions. This induction was seen by the increase in BN28
protein accumulation in acclimated CBF-overexpressing plants as compared to control
plants (Figure 4.5), and a statistically significant increase in proline accumulation seen
only under acclimating conditions (Figure 4.6 and Table 4.1B). This superinduction could
be due to the cumulative effect of the induction of the endogenous BnCBF genes in
addition to the Arabidopsis CBF genes under the control of the CaMV 3SS promoter,
increased binding or activation of CBF proteins under acclimating conditions (Kanaya et
al, 1999). Alternatively, it could be due to increased RNA stability of the BN genes
and/or the genes involved in proline accumulation such as P5CS (Hare et al, 1999) under
acclimating conditions. This increase in gene expression also appears to result in an
additional increase in freezing tolerance as there is a ~2.5° C increase in the ELso of
nonacclimated transgenic plants compared to control plants whereas there is a ~4.5° C
increase in the ELso of acclimated transgenic plants compared to control plants. This
could again be a cumulative effect of the transgene-induced BN gene expression in
addition to the endogenous BnCBF induced BN gene expression. The statistically
significant increase in proline levels under acclimating conditions could also be due to
the “super induction” or, alternatively, there could be a significant increase in proline
under nonacclimating conditions, but larger sample sizes and/or homozygous lines would
need to be used to demonstrate the difference.

Interestingly, a high correlation has been seen between electrolyte leakage tests

166

 

and winter survival in difl’erent varieties of winter and spring canola under acclimating
conditions (Teutonico et al, 1993; Song and Copeland, 1994). If this correlation is
applicable to acclimated CBF-overexpressing transgenic plants, then I could anticipate
that field trials will show an increase in freezing tolerance/winter survival of CBF-
overexpressing plants as compared to control plants. However, this difference remains to
be tested.

Despite the fact that freezing tolerance, drought tolerance and salt tolerance are
interrelated (Thomashow, 1999), and CBF3-overexpression in Arabidopsis resulted in
increased fieezing, drought and osmotic stress tolerance (Liu et al, 1998) I could not
detect a significant increase in salt tolerance between CBF-overexpressing and control
plants (Table 4.3). While there was a significant increase in survival after salt stress
observed in the first two experiments, the difference was not observable in the third
experiment. Whether this was due to the differences in how the experiments were
conducted, the differences in expression in the individual plants used in each experiment,
or some other factor remains to be determined. By doing large scale experiments with
larger sample sizes or perhaps even field trials with homozygous lines, it should be
possible to determine definitively whether or not overexpression of the Arabidopsis CBF
genes in canola results in increased salt tolerance under nonacclimating conditions.

Overexpression of Arabidopsis CBF 1, CBF 2 or CBF 3 in canola plants did not

result in gross phenotypic changes in the investigated lines (Figure 4.8). Overexpression
of CBF] in Arabidopsis ecotype RLD also did not result in gross phenotypic changes in
the lines investigated (J aglo-Ottosen et al, 1998). However, overexpression of CBF 1,

CBF2 or CBF 3 in the WS ecotype (Gilmour et al, 2000; A. Sebolt, S. Gilmour, M.

167

Thomashow, unpublished) and overexpression of CBF 3 in the Columbia ecotype (Liu et
al, 1998) resulted in a dwarf phenotype and delayed growth in some homozygous lines.
While it appears that there is a correlation between levels of transgene expression and
the slow growth/dwarf phenotype (Liu et al, 1998; S. Gilmour, M. Thomashow,
unpublished) the ultimate cause of these phenotypic changes remains unknown. [fit is
due to high levels of gene expression, then perhaps none of the recovered CBF-
overexpressing canola lines have an expression level high enough to induce phenotypic
changes. Alternatively, the phenotypic changes could be species and/or ecotype specific,
or may only occur if the endogenous CBF genes are overexpressed, not if CBF genes
fi'om another species are overexpressed. Finally, the conditions in which the plants are
grown could have an effect on whether phenotypic changes are seen. The exact reason(s)
for the dramatic change in phenotype seen in the Arabidopsis plants, however, remain to
be determined.
Given that CBF-like genes appear to be present in a variety of plant species,
including chilling sensitive plants, such as tomato and rice (http:
//www.ncbi.nim.nih. gov: 80/BLAST; Altschul et al, 1997) it will be interesting to
determine if the CBF homologues in those species are also cold regulated. Data from
Tanksley’s lab indicate that when looking at complex quantitative traits, it is often the
regulation of the critical transcription factor(s) that change, and not the presence or
absence of the transcription factor itself that cause the differences in phenotype (F rary et
al, 2000). If this is correct for the CBF genes, then chilling sensitive plants (plants that do
not acclimate well) may still contain the CBF transcription factors, as is supported by

sequence data, but the regulation of the gene(s) has been changed or lost. Ifchilling and

168

freezing sensitive plants do contain CBF and COR gene homologues, then genetic
engineering by overexpressing the CBF genes has the potential of being highly successful
in these species. By combining traditional plant breeding and genetic engineering, the

long term goal of increasing freezing tolerance may finally be realized.

4. 5 REFERENCES

Altschul, SF, Madden, TL, Schaffer, AA, Zhang, J, Zhang, Z, Miller, W, Lipman, DJ.
1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search
programs. Nucleic Acids Res. 25: 3389-3402.

An, G. 1987. Binary TI-vectors for plant transformation and promoter analysis. Methods
in Enzymology. 153: 292-303.

Artus, NN, Uemura, M, Steponkus, PL, Gilmour, SJ, Lin, C, Thomashow, MF. 1996.
Constitutive expression of the cold-regulated Arabidopsis thaliana COR15a gene affects
both chloroplast and protoplast freezing tolerance. Proc. Natl. Acad Sci. USA. 93: 13404-
13409.

Baker, SS, Wilhelm, KS, Thomashow, MF. 1994. The 5’ region of Arabidopsis thaliana
COR15a has cis-acting elements that confer cold-, drought- and ABA-regulated gene
expression. Plant Mol Bio]. 24: 701-713.

Boothe, JG, Debeus, MD, Johnson-Flanagan, AM. 1995. Expression of a low-
temperature-induced protein in Brassica-napus. Plant Physiol. 108: 795-803.

Frary, A, Nesbitt, TC, Frary, A. Grandillo, S, van der Knapp, E et al. 2000. Fw2.2: A
quantitative trait locus key to the evolution of tomato fruit size. Science 289: 85-88.

Fujimoto, SY, Ohta, M, Usui, A, Shinshi, H, Ohme-Takagi. 2000. Arabidopsis ethylene-
responsive element binding factors act as transcriptional activators or repressors of GCC
box-mediated gene expression. Plant Cell, 12: 393-404.

Gilmour, SJ, Hajela, RK, Thomashow, MF. 1988. Cold-acclimation in Arabidopsis-
thaliana. Plant Physio]. 87: 745-750.

Gihnour, SJ, Sebolt, AM, Salazar, MP, Everard, JD, Thomashow, MF. 2000.
Overexpression of the Arabidopsis CBF 3 transcriptional activator mimics multiple
biochemical changes associated with cold acclimation. Plant Physiol. In press.

169

 

Gilmour, SJ, Zarka, DG, Stockinger, EJ, Salazar, MP, Houghton, JM, Thomashow, MF.
1998. Low temperature regulation of the Arabidopsis CBF family of AP2 transcriptional
activators as an early step in cold-induced COR gene expression. Plant J. 16: 433-442.

Guy, CL, Niemi, KJ, Brambl, R 1985. Altered gene-expression during cold-acclimation
of spinach. Proc. Nat]. Acad Sci. USA. 82: 3673-3677.

Hao, DY; Ohme-Takagi, M; Sarai, A. 1998. Unique mode of GCC box recognition by the
DNA-binding domain of ethylene-responsive element-binding factor (ERF domain) in
plant. J Biol. Chem. 273: 26857-26861 .

Hare, PD, Cress, WA, van Staden, J. 1999. Proline synthesis and degradation: a model
system for elucidating stress-related signal transduction. J. Exp. Bot. 50: 413-434.

Heino, P, Sandman, G, Lang, V, Nordin, K, Palva, ET. 1990. Abscisic-acid deficiency
prevents development of freezing tolerance in Arabidopsis-thaliana (L) Heynh. Iheor.
Appl. Genet. 79: 801-806.

Hong, BM, Barg, R, Ho, T-HD. 1992. Developmental and organ-specific expression of
an ABA-induced and stress-induced protein in barley. Plant Mol Biol. 18: 663-674.

Horvath, DP, McLamey, BK, Thomashow, MF. 1993. Regulation of Arabidopsis-
thaliana L (Heyn) COR 78 in response to low-temperature. Plant Physiol. 103: 1047-
1053.

Houde, M, Dhindsa, RS, Sarhan, F. 1992. A molecular marker to select for freezing
tolerance in Gramineae. Mol. Gen. Genet. 234: 43-48.

Jaglo-Ottosen, KR, Gilmour, SJ, Zarka, DG, Schabenberger, O, Thomashow, MF. 1998.
Arabidopsis CBF 1 overexpression induces COR genes and enhances fi'eezing tolerance.
Science. 280: 104-106.

Jiang, C, Iu, B, Singh, J. 1996. Requirement of a CCGAC cis-acting element for cold
induction of the BN115 gene from winter Brassica napus. Plant Mol Bio]. 30: 679-684.

Jofuku, KD, den Boer, BG, Van Montagu, M, Okamuro, JK. 1994. Control of
Arabidopsis flower and seed development by the homeotic gene APE TALAZ. Plant Cell.
6: 1211-1225.

Kanaya, E, Nakajima, N, Morikawa, K, Okada, K, Shimura, Y. 1999. Characterization of
the transcriptional activator CBF] fiom Arabidopsis thaliana - Evidence for cold
denaturation in regions outside of the DNA binding domain. J. Biol. Chem. 274: 16068-
16076.

Levitt, J. 1980. Responses of Plants to Environmental Stresses. Chilling, freezing, and
1 7O

 

high temperature stresses, Ed 2. Academic Press, New York.

Liu Q, Kasuga, M, Sakuma, Y, Abe, H, Miura S, et al. 1998. Two transcription factors,
DREBI and DREBZ, with an EREBP/AP2 DNA binding domain separate two cellular
signal transduction pathways in drought- and low-temperature-responsive gene
expression, respectively, in Arabidopsis Plant Cell. 1391-13406.

Medina, J, Bargues, M, Terol, J, Perez-Alonso, M, Salinas, J .1999. The Arabidopsis
CBF gene family is composed of three genes encoding AP2 domain-containing proteins

whose expression is regulated by low temperature but not by abscisic acid or dehydration.
Plant Physiol. 1 19: 463-469.

Moloney, MM, Walker, JM, Sharma, KK. l989.I-Iigh-efliciency transformation of
Brassica-napus using agrobacterium vectors. Plant Cell Rep. 8, 238-242.

Monroy, AF, Castonguay, Y Laberge, S Sarhan, F Vezina, LP Dhindsa, RS. 1993. A new
cold-induced alfalfa gene is associated with enhanced hardening at subzero temperature.
Plant Physiol. 102: 873-879.

Nordin, K, Vahaloa, T, Palva, ET. 1993. Differential expression of 2 related, low-
temperature-induced genes in Arabidopsis-thaliana (L) Heynh. Plant Mol Biol. 21: 641-
653.

Ohme-Takagi, M, Shinshi, C. 1995. Ethylene-inducible DNA-binding proteins that
interact with an ethylene-responsive element. Plant Cell. 7: 173-182.

Orr, W, Iu, B, White, TC, Robert, LS, Singh, J. 1992. Complementary-DNA sequence of
a low temperature-induced Brassica napus gene with homology to the Arabidopsis-
thaliana kin] gene. Plant Physio]. 98: 1532-1534.

Ouellet, F, Vazquez-Tello, A, Sarhan, F. 1998. The wheat wcs120 promoter is cold-
inducible in both monocotyledonous and dicotyledonous species FEBS Lett. 423: 324-
328.

Riechmann, JL, Meyerowitz, EM. 1998. The AP2/EREBP family of plant transcription
factors. Biol. Chem. 379: 633-646.

Sambrook, J, F ritsch, EF, Maniatis, T. 1989. Molecular Cloning: A Laboratory Manual
(Cold Spring Harbor Lab. Press, Plainview, NY), 2'”d Ed.

Schagger, H and von Jagow, G. 1987 . Tricine sodium dodecyl-sulfate polyacrylamide-gel
electrophoresis for the separation of proteins in the range from l-kDa to 100-kDa. Anal.
Biochem. 166: 368-3 79.

Shinwari, ZK Nakashima K, Miura, S, Kasuga, M, Seki, M, et al. 1998. An Arabidopsis
gene family encoding DRE/CRT binding proteins involved in low-temperature-

171

responsive gene expression. Biochem. Biophys. Res. Commun. 250: 161-170.

Song, MT, Copeland, L0. 1994. Effect of cold-acclimation on freezing tolerance in
canola (Brassica napus L.). Korean J. Breed 27: 117-123/

Stockinger, EJ, Gilmour, SJ, Thomashow, MF. 1997. Arabidopsis thaliana CBF]
encodes an AP2 domain-containing transcriptional activator that binds to the C-
repeat/DRE, a cis-acting DNA regulatory element that stimulates transcription in
response to low temperature and water deficit. Proc. Natl. Acad Sci. 94: 1035-1040.

Teutonico, RA, Palta, JP, Osborn, TC. 1993. The study of low-temperature responses in
the oilseed Brassica-rapa using molecular markers. Crop Sci. 33: 103-197.

Thomashow, MF. 1999. Plant cold acclimation: Freezing tolerance genes and regulatory
mechanisms. Ann. Rev. Plant Physiol. Plant Mol. Bio]. 50: 571-599

Towbin, H, Staehlelin, T, Gordon, J. 1979. Electrophoretic transfer of proteins fiom
polyacrylamide gels to nitrocellulose sheets - procedure and some applications. Proc.
Nat]. Acad Sci. 76: 4350-43 54.

Wang, H, Datla, R, Georges, F, Loewen, M, Cutler, AJ. 1995. Promoters from kin] and
C 0R6. 6, 2 homologous Arabidopsis-thaliana genes - transcriptional regulation and gene-

expression induced by low-temperature, ABA, osmoticum and dehydration. Plant Mo]
Biol. 28: 606-615.

Wiser, CJ. 1970. Cold resistance and injury in woody plants. Science. 169: 1269—1277.

Welin, BV, Olson, A, Palva, ET. 1995. Structure and organization of 2 closely-related
low-temperature-induced DHN/LEA/RAB-like genes in Arabidopsis-thaliana L Heynh
Plant Mo] Biol. 26: 131-144.

Weretilnyk, E, Orr, W, White, TC, Iu, B, Singh, J. 1993. Characterization of 3 related
low-temperature-regulated cDNAs from winter Brassica-napus. Plant Physio]. 10]: 171-
177.

Wilson, K, Long, D, Swinbume, J, Coupland, G. 1996. A dissociation insertion causes a

semidominant mutation that increases expression of TINY, an arabidopsis gene related to
APETALA2. Plant Cell. 8: 659-671.

Yamaguchi-Shinozaki, K, and Shinozaki K. 1993. The plant hormone abscisic-acid
mediates the drought-induced expression but not the seed-specific expression of RD22, a

gene responsive to dehydration stress in Arabidopsis-thaliana. Mol. Gen. Genet. 236:
33 1-3 40.

Yamaguchi-Shinozaki, K, and Shinozaki K. 1994. A novel cis-acting element in an
Arabidopsis gene is involved in responsiveness to drought, low-temperature, or high-salt

172

 

. W.H.-71““ ‘F‘I

stress. Plant Cell 6: 251-264.

173

 

5. CONCLUSIONS

This thesis reports on the roles of the Arabidopsis CBF family of transcription
factors in cold acclimation. When CBF] was first identified by Stockinger et al (1997), it
was unclear if the protein encoded by the gene played a role in the activation of COR
genes in plants. Additionally, it was unknown if all the characterized COR genes would
be activated, whether other uncharacterized COR genes would be activated, and if the
activation of the battery of COR genes would increase fi'eezing tolerance without a low
temperature stimulus. The work done in Chapter 2 clearly shows that the CBF]
transcription factor can activate transcription of all of the known COR genes without a
low temperature stimulus and that this activation results in an increase in freezing
tolerance which was seen by electrolyte leakage analysis and whole plant freeze tests.

While the data presented in Chapter 2 show that overexpression of CBF] results
in increased CBF] RNA accumulation, they did not address whether there was also
increased accumulation of the CBF 1 protein. Additionally, all of the regulation data on
the CBF family of transcription factors to date is based on RNA accumulation of the
genes under varying conditions. To truly understand how the gene is regulated,
information as to the amount of CBF protein that is accumulated under nonacclimating
and acclimating conditions is critical. While RNA accumulation data may give an
indication as to the amount of protein that can be produced, it does not necessary reflect
the actual amount of protein produced, or the activity of the protein. To answer these

questions, the amount of protein that is accumulated, the location of the protein under

174

differing environmental conditions, and whether the protein is modified to regulate its
activity must be determined. Despite repeated efl’orts to detect the CBF proteins in plant
tissue, I was never successful in this endeavor.

Given that overexpression of each of the three CBF genes in Arabidopsis results
in increased fieezing tolerance, I wanted to address the question if overexpression of the
three CBF genes in an agronomically important species also results in increased freezing
tolerance. To this end, Susanne Klefl’ generated transgenic Brassica napus var. Westar
plants, a close relative of Arabidopsis, that constitutively overexpress CBF], CBF 2 or
CBF 3 . Chapter 4 shows that the Arabidopsis CBF proteins can activate expression of the
Brassica COR gene homologues, the BN genes, and also results in increased freezing
tolerance. This increase in fieezing tolerance was seen under both nonacclimating and
acclimating conditions, with a more dramatic increase seen under acclimating conditions.
A definitive increase in salt tolerance was not detected, although two out of three
experiments showed that CBF-overexpressing canola plants had increased survival after
salt stress as compared to control plants.

Overall, the data indicate that the CBF family of transcription factors play
important roles in cold acclimation. Overexpression of CBF] can activate the COR genes
and increase fieezing tolerance without a low temperature stimulus in Arabidopsis, and
overexpression of all three of the CBF genes in canola results in an increase in freezing
tolerance under both nonacclimating and acclimating conditions. While these data are
informative, there are still many interesting questions that remain to be asked.

One key question is: to what level do the CBF proteins accumulate under

acclimating and nonacclimating conditions? Where are the CBF proteins located under

175

 

nonacclimating and acclimating conditions? Are the proteins sequestered fi'om the
nucleus under nonacclimating conditions, and if so, is a modification event involved?
How does constitutive overexpression of the genes effect the accumulation of the protein
in transgenic plants? While I was not able to successfully answer these questions, this
does not mean that the questions cannot be answered, nor does it mean that future efforts
should not be directed towards answering these questions. Due to the many problems
associated with the detection of the protein using polyclonal antibodies, I believe that
future efforts should involve the use of monoclonal antibodies. This change in antiserum
should greatly reduce problems with cross hybridization to other plant proteins and
should increase detection of the CBF proteins. Additionally, creating transgenic plants
with tagged CBF proteins may aid in the detection of the protein. Ifthe protein is rapidly
degraded, and the data in Chapter 3 support this hypothesis, then adding a large tag to the
protein may slow or stop the degradation process. While this would not aid in detecting
native CBF proteins, it could elucidate the location of the chimeric protein and whether
the protein is modified, both of which are important clues as to the regulation of the
native proteins.

Another interesting question that remains to be answered is how many genes are
activated by the CBF proteins? All of the characterized COR genes are activated by
overexpression of the CBF genes, but how many more as of yet uncharacterized genes
with CRT/DRE sequences in their promoters are also activated? There is now good
evidence that overexpression of CBF 3 in Arabidopsis results in activation of P5CS, a
gene involved in proline production, and the increase in total soluble sugars in CBF-

overexpressing canola plants would indicate that changes in the metabolism of sugars are

176

also regulated by the CBF proteins. However, which of these changes in gene expression
are direct effects of CBF-overexpression and which are indirect effects is not currently
known.

One way to address these questions would be to analyze the total changes in RNA
accumulation in CBF-overexpressing and control plants under various conditions. This
could be done using microarray analysis. By looking at the profiles of the changes in
gene expression in the CBF-overexpressing plants as compared to control plants, the
number of genes activated and the level of activation induced by CBF expression could
be determined. By comparing the genes activated in CBF-overexpressing plants to both
nonacclimated and acclimated control plants, the number of genes directly and indirectly
activated by CBF proteins, and the percentage of those genes to the total number of genes
activated by cold acclimation could be determined. After establishing which changes in
gene expression that occur in CBF-overexpressing plants reflect the changes that occur
during cold acclimation, a time-course of the genes activated by cold acclimation in
control plants could be conducted. The cascade of gene activation in control plants could
then be compared to the genes activated in CBF-overexpressing plants. Ifthe
overexpressing plants have increased activation of both “early” induced and “late”
induced genes, this could give an indication that activation of the “late” genes is an
indirect effect of CBF-overexpression.

There are also more questions to be addressed in the CBF-overexpressing canola
lines. While the data in Chapter 4 indicate that the transgenic canola plants have
increased fi'eezing tolerance under controlled lab conditions, they do not address how

well the plants survive in the field, or if overexpression of the CBF genes results in

177

deleterious effects in terms of yield and oil quality. To fully characterize these lines,
rigorous field testing needs to be conducted. The overall phenotype of the plants, the
tolerance to osmotic and freezing stress as well as other types of stress, the oil quality and
the yield of CBF-overexpressing lines all need to be determined before it can be
established whether or not overexpression of the CBF genes is a viable solution to
increasing the freezing tolerance of canola plants.

The last set of questions that are only briefly addressed in this thesis are what is
the extent of conservation of the CBF signal transduction pathway? How may different
types of plant species contain CBF homologues? If homologues are contained in chilling
sensitive species, and preliminary data indicate that they do, then how are the CBF genes
regulated in these species? Do COR gene homologues exist in chilling tolerant species, if
so, do they contain CRT/DRE sequences in their promoters? By continuing to search for
and characterize CBF homologues in diverse plant species, the answers to these questions
should become apparent. Ifthere is a high level of conservation in diverse plant species,
then manipulation of the expression of the CBF family of transcription factors may be a

viable means of increasing freezing tolerance in many important agronomic crops.

178

   

11111111:1111