wesus
Z

MICHIGAN STATE UNIVERSITY LIBRARIES

1 Wily/Ill!liljfl/lflljlflfllflfllflfl/U/I

 

n

This is to certify that the

dissertation entitled

STUDIES ON THE REGULATION AND PROTEIN PRODUCTS OF THE COR15
GENE FAMILY IN ARABIDOPSIS THALIANA

presented by
Kathy Suzanne Wilhelm

has been accepted towards fulfillment
of the requirements for

Genetics

Ph ’ D degree in

    

Maj r r p ofessor

   

Date July 29, 1996

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

 

 

—"" fiwm

-——u—.-
m.

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove We checkout from your record.
TO AVOID FINES return on or bdore date due.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Ie An Afflnnetive Action/Equal Opportunlty lnetltulon
W

”9.1

 

STUDIES ON THE REGULATION AND PROTEIN PRODUCTS OF THE CORIS
GENE FAMILY IN ARABIDOPSIS THALIANA

By
Kathy Suzanne Wilhelm

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Program in Genetics

1996

ABSTRACT
STUDIES ON THE REGULATION AND PROTEIN PRODUCTS or THE cams
GENE FAMILY IN ARABIDOPSIS THALIANA
By

Kathy Wilhelm

During cold acclimation in Arabidopsis fllaliaua, several families of genes are
induced, but the mechanism(s) by which they are induced remains unknown. A
genomic clone containing the COR15 gene family was isolated for the purpose of
examining the cold-regulation of its promoter. This clone was found to contain
two tandem members of the family, COR15a and COR15b. Their predicted
coding regions are 82% identical and both are transcriptionally regulated by low
temperature. A structural model of the protein encoded by COR15b is proposed.
The promoter of CORISa fused to GUS was then used as a screenable marker in
an attempt to find signal transduction mutants aberrant in the regulation of
CORISa. The only mutants found showed aberrant regulation of the transgene,
but normal induction of the endogenous gene, indicating that the mutations

were not in the desired pathway. Ways to improve the method are discussed.

.

An examination of the temperature induction profiles of four COR gene families
showed that all are induced between 19 and 16°C, and that message levels
gradually increase as the temperature is lowered and level off at about 8°C. For
CORISa, this profile appears to be promoter-based, since message levels of a
CORISa/ GUS promoter fusion gave the same profile as the endogenous gene.
The temperature induction profile was the same whether the temperature was
progressivelylowered twoorthreedegreesatatimeorwasshiftedbyasmuch
as 18°C at once, suggesting that the signal involves the temperature per se and
not the change in temperature. Also, mutants deficient in polyunsaturated fatty
acids in the chloroplast or plasma membrane show the same induction profile as
wildtype plants. This fails to support the idea that an alteration in membrane

fluidity may be involved in the induction of COR gene families.

iv

DEDICATION

To the One around Whom my life revolves,
without Whom not a word could have been written,
the King of Kings and Lord of Lords.
Mayallhonorand praisebeHisinthisandineverythingIdo.

ACKNOWLEDGEMENTS

Thanks are due to my professor, Dr. Mike Thomashow, for providing a
stimulating, amicable research environment and for offering insight and
guidancethroughoutthisproject. Ialsoappreciatetheguidanceandchallengel
have received from my guidance committee, Dr. Pam Green, Dr. Ken Poff and
Dr. Steve Triezenberg. I offer many thanks to Dr. Leslie Kuhn for her infectious
enthusiasm and for the expertise in protein modelling that made the model of
COR‘lem possible, and to Pappan (Kailla Padmanabhan) for his gracious
assistance in the operation of unix-based computers. I am also grabful to
Michelle Marshall and Bob Gifford who were a tremendous help in screening
thousands of mutagenized plants. Thanks are due to Dave Horvath, for showing
me the ropes when I entered the laboratory, to Stokes Baker for working with me
in the analysis of CORISa and for providing the transgenic plants I used in the
mutagenesis experiment, to Leonard Bloksberg for his assistance in analyzing
the plants transgenic for COR15b promoter constructs and to Sarah Gilmour for
her editorial skills and general knowhow around the lab. I am grateful to my
colleagues Eric Stockinger, Kevin O'Connell, Chentao Lin, Todd Cotter, Dan
Zarka, Nancy Artus, Weiwen Guo, Brett McLarney, Deane Lehman, Ann

Gustafson, Beth Seymour, Kirsten Jaglo-Ottosen, to the many students (graduate

vi
and undergraduate) who have worked in Dr. Thomashow’s laboratory, and to

fire members of Dr. Rebecca Grumet’ s laboratory for their discussions, assistance
and general friendliness. Finally, I wish to firank my Godbrofirer, Ralph

Benedetto, for his support, encouragement and editorial assistance.

vii

TABLE OF CONTENTS

Page
LIST OF FIGURES
CHAPTER 1: Literature Review
Introduction 1
Cold -inducible genes 10
Low temperature signal transduction 17
Mutagenesis Strategies 21
Librature Gibd 23
CHAPTER 2: Screen for mutations in fire signal transduction pafirway
responsible for induction of CORISa
Summary 34
Introduction 35
Results 38
Discussion 55
Materials and Mefirods 62

Literature Cited 70

viii
CHAPTER 3: Analysis of a genomic clone containing COR15a and COR15b

Summary

Mabrials and Mefirods

Librature Cited

CHAPTER 4: Examination of fire low temperature induction profiles of COR

gene families

Summary
Introduction

Results

Materials and Methods

Literature cited

APPENDD(

74

103

116

125

131

132

134

139

145

149

152

LIST OF FIGURES

Figure 2.1 Screen for signal transduction mutants.

Figure 2.2 Test for homozygosity of inserts in line 1,6 using
genomic Soufirerns.

Figure 2.3 Putative organization of inserts in line 1,6.
Figure 2.4 MUG assay of putative mutants.

Figure 2.5 Variability of MUG assay.

Figure 2.6 Expression of GUS and of COR15 mRNA in putative mutants.

Figure 2.7 Soufirern analysis of putative mutants
screened at low temperature.

Figure 2.8 Irrducibility of putative mutant line 6-8#4.

Figure 2.9 Coeegregation of fire mutant phenotype wifir fire locus
firought to contain two linked copies of fire reporter construct.

Figure 3.1 Sequence of 4.2 kb of genomic DNA containing COR15a and
COR15b.

Figure 3.2 Alignment of COR15a and COR15b.

Figure 3.3 Predictions of secondary structure and helicity of COR15b.

41

45

47

49

52

81

x
Figure 3.4 CORISbnr modelled as an amphipathic helix.

Figure 3.5 Helical net model of COR15bm.
Figure 3.6 Superimposition of firree Class A beta lactamases.

Figure 3.7 Model of fire proposed structure of COR15bm.

Figure3.8 RegionofbetalactamasewithsimilaritytoCORISamand
~ COR15bm.

Figure 3.9 Comparison of fire regulation of COR15a and COR15b.
Figure 3.10 Comparison of fire promoters of COR15a and COR15b.
Figure 3.11 Regulation of COR15b promoter constructs.

Figure 3.12 Genomic organization of COR gene families.

Figure 4.1 Two methods of lowering fire temperature.

Figure 4.2 First temperature shift experiment.

Figure 4.3 Second temperature shift experiment.

95

100

102

114

135

136

138

Chapter 1:

Literature Review

Introduction

In late summer orwa autumn in northernclimes, fire cry goes forth “There’s
going to be a frost tonight! Cover your tomatoes!" No one worries about
protecting their carrots or columbirre. These plank and many ofirers survive not
only fire light frost, but even fire subzero weafirer of deep winter wifirout
intervention. What is fire difference? Why do snowdrops and crocuses come up
through fire snow and bloom successfully, while cherry blossoms are sterilized

by an ill-timed frost?

While fire emphasis of this work is on an examination of gene regulation in
Arabidopsis We in response to low temperature, it is worfirwhile to place firis
wifirin fire larger context of what happens in plank exposed to low temperature.
The questions presented can only be partially answered. A brief examination of
fire literature quickly reveals fireir complexity. The response of a given plant to

low temperature depends on the organ, fire stage of growth, fire amount of light,

2
the humidity, fire hmperature, fire time spent at low temperature (Raison and

Orr, 1990), the rate of cooling (Minorsky, 1%9) fire plant’ s climate of origin and

wlrefirerfireplanthasbeenhardened.

With respect to low temperature, plank fall into two major classes first
correspond to their climate of origin. Those from tropical or subtropical
climates, like tomatoes and cucurbik, are chilling-sensitive. There is substantial
variability in the susceptibility of chilling-sensitive plank, but when
temperatures drop below a firreshold level (about 10°C for most chilling-
sensitive species), firey suffer damage ranging from loss of vigor to deafir. Of fire
many physiological changes that occur at low temperature, fire main causes of
injury in chilling-sensitive plank appear to be dehydration from a drop in root
pressure (Mirrorsky, 1985) and fire failure of stomata to close (Patterson and
Reid, 1990), the production of toxic oxygen species due to impaired
photosynthesis and respiration (Purvis and Shewfelt, 1993), and fire increased
permeability of cellular membranes (Murata, 1990). Which low temperature
response(s) is primary and which is secondary is not yet completely clear,
alfirough Lyons (1973) and Raison (1973) propose that a low temperature-
induced change in the physical state of membranes in chilling-sensitive plank
leads to fire other symptorrrs observed. Mirrorsky (1%5) suggests that increases
in cytosolic calcium levels might be responsible for fire symptoms of chilling

injury. Subsequent studies with transgenic Nioofim plumbsgirrifillia, a chilling-

3
sensitiveplant, haveshownfiratcoldshocktoOor5°Cdoescauseatransient

increase in the intracellular calcium concentration (Knight et al., 1991).
However, a shock to 10°C, a temperature at which many chilling sensitive plank
are susceptible to growfir inhibition or damage, did not cause an increase in

intracellular calcium in firese experiments (Knight et. al., 1991).

Itisalsoknownfiratmsnydufling-semifivephnkcanbecomemoreresiskntto
chilling bmperatures if firey are chill- or drought-hardened, and first most show

delayed injury when chilled at 100% relative humidity (Wilson, 1979).

Plank of Emperate or arctic origin are chilling-tolerant, alfirough certain organs
or stages of life may be chilling-sensitive (Bramlage and Meir, 19%). Being
poikilofirermic organisms, firey have litfie opportunity to avoid low kmperature,
so resistant plank must either tolerate or be insensitive to its numerous effeck.
With respect to the water skess associated wifir chilling injury, Marklrsrt (1%6)
has shown first when detopped roots of bean (sensitive) and spinach (resistant)
wa'e chilled, fire bean lost 90% of ik root conductance while fire spinach lost
80%. Wifirin eight hours, fire bean had only recovered to 30% of ik original
capacity, while fire spirrsch was up to 70%. Thus, fire most severe water stress in
spinach was transient. Long-term water stress was avoided. Resistant plants
may also avoid fire damage caused by toxic oxygen species. Purvis and Shewfelt
(1993) report first ”cold-resistant cultivars and chilling-resistant tissues generally

develop a greata' potential for respiratory electron flux firrough fire alternative

4
pathway firan do cold sensitive cultivars and tissues." Electrons directed

firrough fire alternative pathway are not available to create toxic oxygen species.
Finally, differences in fire composition of cellular membranes may allow
resistant plank to avoid membrane permeability. A high proportion of
W Wylslycml (PG) and wfloquinovo-yldiacylslywol (SQDG).
glycerolipidswifirphasetransitiontwperahues above30°C,iswellcorrelated

wifir chilling-sensitivity (Murata and Nishida, 1990).

Work wifir Aratidopsis mutank deficient in membrane fatty acid lipid
unsaturation has also suggested first membrane composition affects chilling
tolerance. Wildtype Arabidopsis is not injured by low temperature, slfirough ik
growth rah declines considerably and it has a higher content of chlorophyll
whmr grown at 5°C (Hugly and Somerville, 1992). While several fad mutank are
morphologically indistinguishable from wildtype plank at eifirer normal growfir
kmperaturesorunderchilling stress,fndB W5) andfldC W6) grownat5°C
have a reduced growth rate and chlorophyll content compared to wildtype
Arabidopsis grown at 5°C. leaves first develop at 5°C in firese two mutank are
chlorotic (Hugly and Somerville, 1992). A firird fird mutant, fnd2-2, develops
necrotic lesions and eventually dies if maintained at 6°C for longer than ten days
(Miquel et al., 1993).

5
Whilebemoleculermecharusmsbelundfiredifferemesbetweendulling

sensitive and chilling-resistant plants are still under investigation, Murata and
colleagues have shown that a glycerol-B-phosphate acyltransferase isolated from
peaandspimchwhicharechfllingresistantpreferentiaflyacylatedfire
unsaturated fatty acid 18:1 to fire sat-1 position of glycerol 3-phosphab. The
same enzyme isolated from chilling-sensitive squash hardly discriminated
between unsaturated 18:1 and saturated 16:0 (Murata and Nishida, 1990). They
firen fiansformed tobacco with the glycerol-S-phosphate acyltransferase gene
isolated from squash, a more chilling-sensitive plant than tobacco, or Arebidopsis,
which is chilling tolerant, and found firat fire tobacco plants trursformed wifir
firesquashenzymehad lowerlevelsofcis-unsaturated fattyacidsinPGand
were more chilling-sensitive than wildtype tobacco. The opposib was true of
tobacco transformed wifir fire Arabidopsis gene (Murata et al., 1992). Thus it has
beenshown directlyand invivo firattheselectivityofa singleenzymecanaffect
membrane composition and alter chilling tolerance. It remains to be learned

whyandlrowfirisisso.

It has also been noted that tobacco plants overexpressing chloroplastic Cu/ Zn
superoxide dismutase (SOD) from pea are less sensitive to chilling injury than
wildtype tobacco (Gupta et al., 1993). Whether SOD from chilling-resistant pea

is superior to tobacco isoforms under low temperature conditions or whether fire

6
simpleoveeexpressionofSODcausedfiusleeaeningofsuscepfibilityisnot
known.

Somechillirrg-resistantspeciesarealsofreezing-tolerant, manyoffiremableto
increauinfreezingtoleranceinresponsetolowmnfreezingtemperahrres.
Planbdamagedduetofreezingtypicallyappearflaccidandwabbsoaked,
suggeefingfiatfreezinginjuryismainlyduetofirecompromiseofcell
membrane integrity. Dependingonfirenatureoffirefreezeandonfire
properties of fire cells involved, firis may be manifested as intracellular ice
foreration; as expansion-induced lysis, which is caused by fire inability of fire
thawing cell to properly rehydrate; or as fire loss of osmotic responsiveness
(W, 1984).

Kendall et al. (1%9) suggest first membrane disruption may be due to free
radicals produced during freezing, since firey could detect free radical
production during fire freezing of winter wheat, and acclimated plants were
more resistant to applied free radicals as well as to freezing firan nonacclimated
plants. Also, the constitutive expression of an extra superoxide dismutase in
alfalfa resulted in one transgenic line first had bofir greater resistance to a free
radical-producing herbicide and enhanced freezing tolerance (McKersie et al.,

1993).

7
Intracellular ice formation, whether it be due to intracellular nucleation or to

penehafionoffirecellbyanexternalicecrystaLisgeneraHybelievedtobe
instanfiylefiralunlessfiaeratesoffreeu’ngandfinwingareupidenoughfiut
the icecrystals arevery fine (Sakai and Sugawara, 1978). The intact plasma
membraneisthoughttobeaneffecfivebaniertoseedingbyexfiacellularice
(Charnba'sand Hale, 1932), and firework obeponkusarrd Dowgert(1983)
urggeebfirstinfiacdlularseedingmaybearesultofandterafionoffireplasma
membranesincemeclunicalfaflureoffireplasmalemmacanbeobservedpfior

to intracellular ice formation.

More commonly, ice formation is extracellular. The solub concentration of fire
apoplast is lower firan first of fire cytoplasm, wifir fire result being first fire
cytoplasm has a greater freezing point depression (Guy, 1990). Once fire
exfiacellular water begins to freeze, however, fire water pohntial outside fire cell
decreases. Ice has a lower water potential firan water at fire same hmperature
and solutes excluded from fire growing ice crystal lower fire apoplastic want
pohmtial furfirer, so water diffuses out of fire cell (Thomashow, 1994). Thus fire
cell suffers from dehydration as well as from fire mperam itself. Plants
frozen to -10°C lose more firan 90% of fireir osmotically active water (pronkus
and Lunch, 1%9). When fire cells are rehydrsted, water moves back into fire cell,
and if fire cells are unable to accommodab fire influx, “expansion-induced lysis"

is the result (Steponkus, 1984).

8
louofosmoficresponsivenessoccunafternomcchmatedpmtophsharefrozen

bdow-S‘Qbutcanbemimickedbyosmoficdehydrationinfireabsenceofice
formafioanhislossisassociabdwifirclungesinfireulfiasfi'ucmreoffire
plasma membrane, including lamellar to hexagonal-II phase transitions
(Steponkus and Lynch, 1989), in which lipids are reoriented. Insbad of forming
abilayer,thelipidsarearrangedinlongcylinderswifirfireirpolarheadgroups

in an aqueous core (Steporrkus, 1984).

An increase in freezing tolerance, firerefore, acquired in a process commonly
referred to as “cold acclimation," must prepare fire cell for avoidance or
tolerance of dehydration and rapid rehydration, fire formation of intracellular
keandfirelouofosmoticresponsiveness. Thisbeirrgfirecase,itmakessense
“drought-hardening, whichcancauseanincreaseinchillingtoleranceof
clunkrgaensifiveplanbmanahobfingaboutmmcmaseinfieeeingtoleramein
chilling-resistant cabbage (Cox and Levitt, 1976), spinach (Guy et al., 1992),

wheat and rye (Siminovitch and Cloutier, 1983).

Ithasbeenseenfiratprotoplasts isolated fromcold acclimated plantsexperierrce
intracellular ice formation at lower temperatures firsn firose from nonacclimated
plants (Dowgert and Steponkus, 1983). It has also been observed first
protoplasts isolated from cold acclimated rye, unlike those from norracclimated

rye, were not susceptible to expansion-induced lysis after having been frozen to

4 9
6C (Steponkus, 1%). Finally, protoplasts isolated from acclimated rye become

oemotically unrespomive at much lower temperatures firan firoee from
nonscdimatedrye,andevenfiren,fireyundergodifferentmembranephase
W firan do their nonacclimated counterparts (Fujikawa and Steponkus,
1990). This difference in membrane phase transitions during freeflng has been
confirmedinshrdiesushrgleafsecfiomhkenfromacclimatedormnacdimabd
rye(WebbandSteponkus,1993). Thuafireavoidanceofintracellularice
formationandoflossofosmoficresponsivenessandfiretoleranceofrapid

rehydration appear to be operative in frozen acclimated plants.

The physical properties of fire tolerance of fire dehydration and rehydration in
acclimatedplantshavebeenclarifiedbySteponkusandcolleagues (pronkus,
1984). They dwonstrated firat protoplasts isolated from nonacclimated rye form
endocytoticvesicleswhichbud offfrom fireplasma membrarreasfirecellsshrink
during freeze-induced dehydration. Upon firawing, rehydration results in
intolerable osmotic pressure because fire vesicles are not reincorpor'ated into fire
membrane and fire protoplasts burst. Protoplasts from acclimated rye, however,
form exocytotic extrusions as firey dehydrate. These extrusions remain in
association wifir fire membrane so first when fire cells regain water, fire
extrusions are reirrcorporated into fire membrane and fire cells do not lyse. It has
also been shown first a change in lipid composition is sufficient to favor eifirer

endocytotic vesiculstion or exocytotic extrusions (Steponkus and Lynch, 1%9).

10
Indeed, the membrane composition of Arsbr’dapsr’s is such first even protoplask

isolated from nonacclimated plank are resistant to expansion-induced lysis
(Umera at al., 1995). erefirer whole plant cells behave in firis manner remains

tobeseen.

Theproceesofcold acclimation, bywhichfiresechangesarebroughtabout,
involves many physiological alterations. The lipid composition of cellular
membranes changes (Lynch and Steponkus, 1987, Umera et al., 1995), resulting
inthedifferencesinmembraneresponsetofreezingalreadydescribed. There
areincreasedlevebofsugam,solublepmteim,pmhneandofirerorganicadds
(Sakai and Larcher, 1%7). These may be involved in ameliorating fire effeck of
freeze-induced dehydration and/ or in reducing ik severity. New isozymes are
formed (Guy, 1990) and changes infireexpressionofa modeetnumberofgenes

are seen (Thomashow, 1993).

Cold-inducible genes

While fire overall pattern of gene expression in plank exposed to low
temperature does not drastically change (Gilmour et al., 1988), still, quite a
number of COR (cold regulated) genes have been identified from numerous
plank. For fire "sake of brevity, and because fire plant examined in firis work is

resistant to chilling damage, only genes induced in chilling-tolerant plank will

11
be considered here. The functions of some low temperature-inducible genes

have been demonstrated experimentally or inferred based on sequence

comparisons wifir genes of known identity.

'l'lreusefulnessofsomeoffiregenesfoundtobeinduced atlow temperaturesis
readily explairrsble. Increased levels of sucrose synfiretase in wheat (Newsted et
al., 1991, Marans et al., 1990 and Crespi et al., 1991) and of phosphoerrolpyruvte
carboxykinase in rapeseed (Saez-Vasquez et al., 1995) may be involved in the
increase in soluble sugar content firought to ameliorate fire dehydration stress
imposed on frozen cells. The accumulation of extensin mRNA in acclimated pea
seedlings is firought to lead to a buildup of fire extensin probin, which may
protect cell walls from collapse under fire extreme dehydration pressures caused
by freezing (Weiser et al., 1990). This protein contains a large portion of fire
hydroxyproline in fire cell, and levels of hydroxyprolirre in cell walls increased
during fire same period first mRNA levels of extensin increased, suggesting first
extensin was being irrcorporsted into cell walls and enhancing fireir rigidity
(Weiser et al., 1990). Antifreeze proteins found at increased levels in acclimated
winter rye could help prevent recrystallizsh’on of extracellular ice, during which
crystals large enough to cause physical damage to tissues and cells could form
(Hon et al., 1995). Lipid transfer proteins induced in acclimahd barley could
play a role in fire alteration of lipid content seen in cell membranes at low

temperature (Hughes et al., 1992, White et al., 1994).

12
Wifirrespecttogareralmetsbolism, heatshockproteinsandhestshockcognate

proteinsfoundirracclimated Brassics napus(Kr-ishnsetal., 1995)orspinach
(NevenetaL, 1992, LietaL, 1994, Andersonetal, 1994) maybeinvolved in
prokcfingpsoteinsfiomlowtemperahrredenahrrafionorinhelpingpmteimto
refold. Alcohol delrydrogenase induced at low hmperature is firought to play a
roleinfireshifttoanaerobicmetsbolism firstoccurswhenrespirationin
inhibited (Jarillo et al., 1993). Phenylalanine ammonia-lyase and chalcorre
synfirase, enzymes involved in fire production of snfirocyarrins, are also induced
by exposure to low temperature (Levya et al., 1995). Anfirocyanins are believed
toactaslight-screening pigmenk (LevyaetaL, 1995)andmay lrelpdecreasefire

amount of toxic oxygen species produced by cold-impaired photosynthesis.

The induction of genes first'could be involved in fire regulation of translation or
in signal trsrrsduction pathways active at low temperature was also observed. In
barley, a translation elongation factor 1a is inducible in cold-treahd plank
(Dunn et al., 1993). In addition, several protein kinases, including a novel
protein kirrase in wheat (Holspps and Walker-Simmons, 1995), two calcium-
dependent protein kinases in alfalfa (Monroy and Dhindsa, 1995), a mitogen-
activated proteinkinssekinasekinaseinArsbr’dopsis (MizoguchietaL, 1996) and
twogeneswithhighsequerwesimflsfitytofibosomabpmteinfikinasafikoin
Arsbidopsis (Mizoguchi et al., 1995), are inducible by low temperature. Finally,

two genes showing high identity wifir 14-3-3 proteins show low Wperature

13
inducibility in Arsbidopsis Oarillo et aL, 1994). Thought to be involved in fire

reguhfionofproteinldmsa,14-3—3pmteimarehkelytobeinvolvedinsignal
transduction. Onememberoffirefamilyhasbeenfoundtobepsrtoffireprotein
complexfirstbindstofireg—boxpromoterelementinArebidopsisauetaL, 1992).
Thissupporkfireideafiut14-3—3proteinsmayhsveamleingeneregrfiafionin
plank.

The purpose of a few other identified cold-responsive genes in low mperamm
acclimationormetabolism islessobvious. InArabidapsis, messagelevelsofone
member of fire B-tubulin gene family increase while levels of four ofirer
members of the family decrease and still ofirers remain fire same (Chu et al.,
1998). Low temperature does affect fire integrity of microtubules (Pihakaski-
Maunsbach and Puhsksinen, 1995), but fireir role in cold tolerance is still
unknown. Also in Arabidopsis, Williams et al. (1994) have found a potential firiol
prokaseisinduced infirecold.Theysuggestfiratitmsybeinvolvedinfire
degradation of proteins denatured by low temperature, in fire degradation of
storage proteins to alter fire osmotic potential of fire cell or in fire stress-induced
proteolytic activation of enzymes. Anofirer low temperature-inducible gene
which may assist in osmotic adjustment is fire 70 kd subunit of torroplast

ATPase, which has been isolated from winter Brassica napas (Orr et al., 1995).

. 14
Inaddifiontofiresegenesofhrownorpredictedidenfity,firerearemanywifir

litfieornosequenceidenfitytogenesofknownfuncfionTheseirrclude HVAl
(Sutton et al., 1992), blt101, 111th15, blt63, blt49, M410, blt14, blt411, blt801 (Dunn
et al., 1994), COR14 (Crosatti et al., 1995), Dhnl and Dhn2 (van Zee et al., 1995)
from barley, m39/W0120 from wheat (Guo et al., 1992, Houde et al., 1992), and
possibly pBGA12, pBGA56, pBGABS and pBGA25 from bromegrsss (Lee and
Chen, 1993). Inlegumirrousplank, ELIPfromgreenpea(Adamskaand
Kloppskch, 1994), GAB-8 and GAB-9 from chick pea (Colorado .et al., 1994),
0‘15 (Monroy et al., 1993a), c018 (Wolfraim et al., 1993) and arse CIC
(Castongusy et al., 1994) from alfalfa have been found to be responsive to low
Emperature. In spinach, fire gene or genes encoding CAP85 show cold-
irrducibility (Neven et al., 1993). Low temperature-responsive genes in crucifers
include BN19, BN26, BN115 (Weretilnyk et al., 1993), BnC24A, BnQ4B (Saez-
Vasquez et al., 1993), BN28 (Boofire et al., 1995) and big-26 (Stroehr et al., 1995)
from W rapes, and rsb18 (Lang et al., 1992), 11130 (Welin et al., 1994), Rial
(Kurkela and Franck, 1990), COR6.6/kin2 (Gilmour et al., 1992, Kurkela and
Borg-Franck, 1992), COR15a (Lin and Thomashow, 1992), COR15b (W ilhelm and
Thomashow, 1993), COR47 (Gilmour et al., 1992), 111'45/ “£29 (Welin et al., 1994,
Welin et al., 1995), COR78/ rd29A/ 11178 (Horvafir et al., 1993, Yamaguchi-
Shinozaki and Shinozski, 1993, Nordin et al., 1993) and rd298/ “1'65 (Yamaguchi—
ShinozakiandShinozaki, 1993, Nordirretal, 19GB) fromArsbidopsr’s. Manyof

fire genes listed show sequence similarity to dehydrins or late embryogenesis

. 15 .
abundant genes, which are prevalent in plank during conditions of water stress,

butfirefunctiensoffiresegenesarestillamatterofspeculation.

Work done with fire COR gene families of Arabidopsis has shown first fire known
members of families COR6.6 (16111 and COR6.6), COR15 (COR15a and COR15b),
COR47 (COR47 and 1545/ lti29 ) and COR78 (COR78/ rdZQA/ 11178 and

r4298 / 1565) are all inducible by low temperature (Thomashow, 1994). The
mRNA levels offire genefsmilies remain high as long as plank are keptatlow
tempa'ature, declining to control levels wifirin eight hours of deacclimation
(HajelsetaL, 1990). Atleastonememberofeachfamilyisirrducedbydrought
stressand bytheexogenous applicationofsbscisicacid (ABA), butnone
respond to heat shock. All are very hydrophilic and remain soluble upon

Of these Arsbr’dopsis COR genes, fire most is known about COR15a. Promoter
deletion studies have shown first its expression is regulated at fire promobr level
by low hmperature, drought and ABA and suggest first fire promoter elemenk
responsible for firis regulation are located between -305 and -78, relative to fire
start of transcription (Baker et al., 1994). Indeed, an element found in firis
region, CCGAC, contains fire-fivecorebasesoffire DRE, anelementfrom fire
CORN/r429A/lti78 promoter identified 'as being sufficient for cold and drought

inducibility (Y smaguchi-Shirrozski and Shinozaki, 1994). Expression of COR15a

16
at low temperature is found in most tissues of the plant, except for fire roots and

ovaries. There is also constitutive expression of a COR15a / GUS promoter fusion

constr-uctinanfirers(Bakeretsl.,1994).

11» protein encoded by coarse, comsll, is targeted to fire chloroplast. This
mightbeexpecbdfromikexpressionpattern, sinceitismairrlyfoundingreen
tissues. Infireprocessofimportintofirechloroplast. COR15aisclesvedfromik
original molecular weight of 14.7 kd to fire 9.4 kd mature form, COR15sm (Lin
and Thomashow, 1992). The mature form of fire protein is acidic, wifir a pl of 4.6
(Gilmour et al., 1996). The function of COR15sm wifirin fire chloroplast is not
yet known, but it has recenfiy been shown first transgenic plank carrying'fire
COR15a coding region driven by a constitutive promoter, which express COR15a
at normal growfir temperatures, show increased freezing tolerance. The
effkierrcyofplwtosystemnwsslesssensifivetofreezingumessuredbyfire
Fv/ Fm ratio, firsn in non-transgenic controls or in transgenic plank containing
an unrelated coding region, and protoplsst survival was enhanced (Artus et al.,
in preparation). The mechaniSm for firis increased freezing tolerance has yet to

be elucidated.

17
Low temperature signal transduction

Among fire many questions yet to be answered is first regarding fire mechanism
by which cold-regulated genes are induced. Nor is it clear first any or all cold-
regulated genes are induced by fire same signal. Msntyls et al. (1995) have
shown first in mutank deficient in or insensitive to ABA, RABlB (fire probin
product of reb18) is no longer cold-inducible, while LTUB (fire protein encoded
by “178 / COR78) is responsive to low temperatrrre, suggesting first fire low
temperature induction of RABIB requires fire presence of ABA, while first of

LTIYB does not.

Given fire many changes first occur wifirin fire cell upon exposure to low
bmpersture, firere are many possible signals. A signal cascade could be
initiated by a change in water potential, by an increase in fire amount of reactive
oxygen species wifirin fire cell, by fire conformational change of a cold-sensitive
protein, by an alteration in fire properfies of cellular membranes or by some
other cellular consequence of low temperature. The signal could be as simple as
a single protein first is bofir sensor and transcriptional activator, as is true of
sterol regulatory element-binding protein 1 (SREBP-I) in mammals (Wang et al.,
1994) or as complex as s multi-component, branched cascade. Membrane
fluidity appears to be irrstrumentsl in fire induction of fire low temperature-

responsive gerre deal from fire cyanobacterium Synedrocyslis PCC6803. This gene

18
kresponsivetoachangemtemperatureofgreaterfimnfivedegreesCekiusaps

et al., 1993), but is also responsive to palladium-catalyzed hydrogenation of fire
plasma membrane (Vigh et al., 1993). Whefirer membrane fluidity or any
mmbrarre—locslizedlowkemperatureresponsealtersfireexpressionofCOR

genesinplsnkisstillurrknown.

Otherfactorsfiratmsyplsyaroleinsignslksnsducfionincludefirelevelsof
ABA,fireinkacellularcslciumconcenkafionsndfirephosphorylstionskteof
prodns. ABAhasbeenimplicatedinfireprocessofacc’limafionbecauseABA
levels rise in response to low temperature in potato (ChenetaL, 1983), wirrkr
wheat (lek and Dorffling, 1%5) and spinach (Guy and Haskell, 1M) and
becauseexogenousapplicsfionofABAcancausemcreaudfieedngtolerancem
potato (Chen et al., 1979), alfalfa (Mohspstra et al. 1989), W (Lang at al.,
1988), and irrcell suspension cultures ofwinter wheat, winterryeand
bromegrass (Chen and Gusts, 1983), while some of fire ABA mutants of
Arsbidopasareimpairedmfreezingtoleranceflieimetsl,1990,6flmourand
Thomsshow,1991). AswasmentiorredearlienABAappearsbbeneededfor
the cold-regulation of RABIB, but not LTI78 (Msntyls et al., 1995). Thus, it
seemstoberequiredforsome,butnotall,offirechsngesfirstoccurduringlow

Empersture.

19
Calcium has been implicated in low temperature metabolism in a number of

ways. In onion, tension-dependent activity of calcium-selective cation co-
chanrrelsincreasedasfiretemperaturewaslowered from25‘Cto6°C(Dingand
Pickard, 19%). Akanaierrtincreaseincytosoliccalciumlevelswasseenin
tobacco (Nicotine plantagimfirlia) (Knight et al., 1991) and in Arabidopeis (Knight
etal., 1996) irrresponsetocold shock. Intobacco, firecold shockwmperature
hadtobe5°Corlowerforfireincreaseincalcium levelstobedebctsbleafiriglrt
et al., 1991), while Arsbr’dopsis was only tested using ice water for fire cold shock
(Knight et al., 1996). A calcium influx was also seen in alfalfa below 15°C
(Monroy and Dhindsa, 1995). That calcium levels may actually be involved in
fire almration of gene expression is suggested by work done in Chick pea and
alfalfa (Colorado et al., 1994, Monroy and Dhindsa, 1995). In chick pea, genes
first are responsive to low temperature, heat shock, ABA and fire osmotic stress
applied by NaCl or polyefirylene glycol (PEG) were also up-regulabd in fire
presence of 0.5 mM CsCl: (Colorado et al., 1994). More extensive experiments
performed wifir alfalfa cell suspension cultures examined fire irrducibility of
€615 and as“, which are inducible by low temperature, but not byABA, lrest
shock, water stress (imposed by a solution of PEG-6000) or wounding
(Mohapstra et al., 1%9). Inhibitors which blocked fire influx of external calcium
also inhibited fire induction of fire as genes at low temperature, while fire
addition of a calcium ionophore or a calcium channel agonist resulted in as gene

induction at 25°C (Monroy and Dhindsa, 1995).

20
Cascades of plrosphorylation and dephosphorylstion reactions are a common

Memdgnflkansducfionmuumaksndambecommgcommonmfinphnt
literature as pathways are being unraveled (nghuram and Sopory, 1995, Ecker,
1995, Zhou et al., 1995). The role of protein phosphorylation in low temperature
generegulstionisinferredfromfinlowtemperatureinducibilityoffin
numerousldnasesalreadymentionedaswefiasfromcold-inducedchangesin

phosphorylstion patterns seen in alfalfa (Monroy et al., 1993b).

Two principle approaches have been used in studying fire molecular biology of
signal transduction pafirways. One can study a promoter and ik elemenk,
determine what ack upon firose promoter elemenk, what interacts wifir first
factor and so on backward firrough fire psfirway. Alternatively, mutations can
be induced and plank that show aberrant expression of fire gene or genes of
interest can be characterized. This approach can reveal members of fin pafirway
in any order. The first approach in studying fire pafirway regulating COR15a
has already borne some fruit. A promoter element necessary and sufficient for
low temperature induction of COR78, fire DRE or dehydration-responsive
element, has been defirnd (Yamaguchi-Shinozaki and Shinozaki, 1994) and an
element containing fire core CCGAC of fire DRE is found in fire region believed
to be responsible for cold-regulation of COR15a (Baker et al., 1994). In Brassica
napus this same DRE-like element, CCGAC, has been shown by transient

expression studies to be required for cold induction of BN115 (Iiang et al., 1996).

21
Furfirermorenproteincapableofbindhgtofidselementhasbeenkohtedfiom

W(Stsckingeretal.,inprepsration). Whefirerfirisprokeinisactually
involved in fire activation of COR15a or any of fin ofirer COR genes in vivo

awaits furfirer study.

Mutagenesis Strategies

When using mutank to examine low temperature signal transduction in a plant
first can cold acclimate, firere are two main skategies. Plank can be
mutsgenized and screened or selected looking for increased freezing tolerance
wifirout acclinration or for frost-sensitivity despite acclimation or firey can be
screened or selected more direcfiy for fire induction or lack of induction of COR
genes. Thefirstmefirodhasfinadvantageofareadilyobservablephenotype,
but may result in mutank in a number of processes unrelated or tangentially

relsnd to signal transduction.

Using gene expression as fin phenotype of interest requires more setup, and
could still yield mutank not direcfiy involved in fin pafirway, but is a more
direct approach to fire question. In order to identify mutank in fire regulation of
a gene whose expression has no obvious morphological phenotype, fin promoter

Offiregeneofinterestisfusedtofirecodingregionofareportergeneandplank

22
carryingfirisconskuctaremukgenizedandscreerndorselectedbasedonfin

expressionoffirereporter.

Variafionsoffiisapproachhavebeenusedseveralfimeswifirvarying resulk.
Takahashietal. (1992)fusedfinpromoteroffinH$P18.2 genetofingern
encoding bets-glucuronidase (GUS) and screened using a 4-methyl umbelliferyl
glucurorride (MUG) assay. Two of fin mutank isolated showed a large
reduction in GUS activity, but only a small reduction in the endogmus
transcript, while fin mutation in fin firird acted as a dominant trans-suppresser
of fire introduced gern wifirout affecting fire endogenous gene. When Brusslan
et al. (1993) fused fire cab140 promoter to m2, fire protein product ofwhich
converk nontoxic naphfiralene acetsmide'l into toxic naphfiralern acetic acid, fireir
selectionresulted inamutantwhichhad reduced levelsofbofirfinendogenous
and fin introduced gene, but not of any ofirer phytochrome-regulated genes, nor
could fire phenotype be genetically separated from fin T—DNA insert. From firis
firey infer first fire lowered levels of mRNA are a result of co-suppression caused

by a mutation in fire introduced gene.

Susek et al. (1993) employed two reporters in fireir construct, putting the CAB3
promoter in front of bofir fire GUS gene and fire gene for hygromycin resistance.
They selected for hygromycin resistsrrce and firen checked fire putative mutank

for GUS activity to make sure fire mutation wasn’t in fire transgene. This

. , 23
resultedinfinisolationofatleastfiueegenesnecessaryforcouplingfin

expmsion of some nuclear genes to fire furrctionsl state of fire chloroplast. The
Donglaborator-yusedfinpromoteroffinArabidopsis 8-1,3-glucuronidasefused
tofinGUScodingregionandfireirscreengavefirembofiramutantfiratis
norrresponsive to inducers of syskmic acquired resistance (SAR) (Cao et al.,

1994) and one first leads to constitutive expression of SAR (Bowling et al., 1994).

While mutank were recovered wifir a screen and wifir fin use of only one
reporter, fin development of a selection would allow a greater number of
mutanktobetesmdwifirlesstimeandenergy,andfinuseofasecondreporter
doesfacilitatefinidentificafionoffalsepositives. Givenfirkinformafionit
should be feasible to design and implement a mutsgenesis approach for isolating
genes involved in fire signal trandsuction pafirwsy responsible for fin induction
of cold-regulated genes.

Literature Cited

Adamska I, Kloppstech K (1994) Low temperature increases fire abundance of
early light-inducible transcript under light stress conditions. The Journ. of Biol.
Chem. 269:30221-30226

Anderson W, Li Q-B, Haskell DW, Guy CL (1994) Structural organization of fire
spinach endoplasmic reticulum-luminsl 70-kilodalton heat-shock cognate gene
srrd expression of 70-kilodalton heat-shock genes during cold acclimation. Plant
Physiol. 104: 1359-1370

24

Baker $, Wilhelm KS, Thomashow MF (1994) The 5’-region of Arabidopsr’s
We cor15s has en's-acting elemenk first confer cold-, drought- and ABA-
regulated gene expression. Plant Mol. Biol. 24: 701-713

Bowling SA, Guo A, Cso H, Gordon AS, Klessing DF, Dong X (1994) A mutation
in Arsbidopsis first leads to constitutive expression of systemic acquired
resistance. The Plant Cell 6:1845-1857

BramlageW],MeirS(1990)Chillingirrjuryofcropsoftemperanorigin. Irr,CY
Wang, ed.,('.‘lrillinginjuryofhorticulturslcrops. CRC Press, Inc., BocaRaton,
Florida,pp37-49

BrusslanIA, Karlin-NeumannGA, Huang L, Tobin E (19%) AnArabidapsr’s
mutant wifir a reduced level of 01111140 RNA is a result of cosuppression. The
Plant Cell 5:667-677

CaoILBowlingSA,GordonAS, DongX(1994) Clrsracterizationofan
. ' ' mutant first is nonresponsive to inducers of systemic acquired
resistance. The Plant Cell 6:1583—1592

Castonguay Y, Laberge S, Nadeau P, Vezins LP (1994) A cold-induced gern from
Medicago satius encodes a bimodular protein similar to developmentally
regulated proteins. Plant Mol. Biol. 24:799-804

Chen TI-IH, Gavinlertvatana P, Li PH (1979) Cold acclimation of snm-cultured
plank and leaf callus of Sdauum species. Bot. Gaz. 140:142-147

ChtmTHH, Gusts LV (1%3) abscisic acid-induced freezingresistarrceincultured
plant cells. Plant Physiol. 73:71-75

Clan H-H, Li PH, Brenner ML (1983) Involvement of abscisic acid in potato cold
acchmation. Plant Physiol. 71: 362-365

Chu B, Sunstad DP, Carter N (1993) Alteration of B—tubulin gene expression
during low-temperature exposure in leaves of Arabidopsis Hraliana. Plant Physiol.
103:371-377

Colorado P, Rodriguez A, Nicolas c, Rodriguez D (1994) Abscisic acid and stress
regulate gene expression during germination of chick pea seeds. Possible role of
calcium. Physiol. Plant. 912461-467

Cox W, Levifi] (1976) Interrelations between environmental factors and freezing
tasistance of cabbage leaves. Plant Physiol. 57:553-555

25

Crespi MD, Zabaleta E], Pontis HG, Salerno GL (1991) Sucrose synfiretase
expression during cold-acclimation in wheat. Plant Physiol. 96:887-891

Crosatti C, Soncici C, Stanca AM, Cattiveli L (1995) The accumulation of a cold-
regulahd chloroplastic protein is light-dependent. Plants 196:458-463

Di!“ JP, Pickud BC (1993) Modulation of mechanosensitive calcium-selective
cation channels by temperature. The Plant Iourn. 3:713-720

Dowgert MP, Steponkus PL (1983) Effect ofcold acclimation on intracellular ice
formation in isolated protoplasts. Plant Physiol. 72:978-%8

Dunn MA, Goddard NI, Zhang L, Pearce RS, Hughes MA (1994) Low-
temperature-responsive barley genes have different control mechanisms. Plant
Mol. Biol. 24:879-888 ‘

Dunn MA, Morris A, lack PL, Hughes MA (1993) A low-temperature-responsive
translation elongation factor la from barley (Harkm vulgar: L). Plant Mol.
Biol. 23:221-225

Faker IR (1995) The efirylene signal transduction pathway in plants. Science
268:667-675

Gilmour 5], Lin C, Thomashow, MF (1996) Purification and properties of

Arabidopsis W COR (gold-regulated) gene polypeptides COR15arn and
C0865 expressed in Esdreridria 001i. Plant Physiol. 111:293-299

Gilmour SI, Hajela RK, Thomashow MF (1988) Cold acclimation in Arabidopsis
dualism. Plant Physiol. 87:745-750

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

Guo W, Ward RW, Thomashow MF (1992) Characterization of a cold-regulated
wheat gene related to Arabidopsr's can! 7. Plant Physiol. 100:915-922

Guy C (1990) Cold acclimation and freezing stress tolerance: role of protein
metabolism. Annu. Rev. Plant Physiol. Plant Mol. Biol. 41:187-23

Guy CL, Haskell D (1%8) Detection of polypeptides associated wifir file cold
acclimation process in spinach. Electrophoresis 9:787-796

Guy CI. Haskell D, Neven L, Klein P, Smelser C (1992) Hydration-stable
responsive protein link cold and drought stress in spinach. Planta 188:265-270

26

Hajela RK, Horvafil DP, Gilmour SI, Thomashow MF (1990) Molecular cloning
and expression of car (Cold-regulated) genes in Arabidopsis firalr'ana. Plant Pysiol.
93:1246-1252

Holappa LD, Walker-Simmons MK (1995) The wheat abscisic acid-responsive
protein kinase mRNA, PKABAI, is up-regulated by dehydration, cold
temperature, and osmotic stress. Plant Physiol. 108:1203-1210

Hon w-c, Griffith M, Mlynarz A, Kwok yc, Yang DSC (1995) Antifreeze
prohins ln winter rye are similar to pafirogenesis-related proteins. Plant
Physiol. 109-379-889

Horvsfir DP, McLarney BK, Thomashow m (1993) Regulation of Arabidapsis
Minna L. (Heyn) 00178 in response to low bmperature. Plant Physiol. 103:1047-
1053

Houde MI, Laliberte I-F, Rassart E, Dhindsa RS, Sarhan F (1992) Cloning,
characterization, and expression of a cDNA encoding a 50-kilodalton protein
specifically induced by cold acclimation in wheat. Plant Physiol. 99:1381-1387

Hughes MA, Dunn MA, Pearce RS, White AI, Zhang L (1992) An abscisic-acid-
responsive, low bmperature barley gene has homology wifit a maize
phospholipid transfer protein. Plant, Cell and Environ. 15:861-865

Hugly S, Somerville C (1992) A role for membrane lipid polyunsaturation in
chloroplast biogenesis at low temperature. Plant. Physiol. 99:197-202

Iarillo IA, Capel I, Leyva A, Martinez-Zapata IM, Salinas I (1994) Two related
low-temperatlrre-inducible genes of Arabidopsis encode proteins showing high

homology to 14-3-3 proteins, a family of putative kinase regulators. Plant Mol.

Biol. 25: 693-704

Iarillo IA, Levya A, Salinas I, Martinez-Zapata IM (1993) Low temperature
induces the accumulation of alcohol dehydrogenase mRNA in Arabidopsis
firaliaua, a chilling-tolerant plant. Plant Physiol. 101 :833-837

Iiang C, In B, Singh I (1996) Requirement of a CCGAC orb-acting element for cold
induction of the BN115 gene from winter Brassica napus. Plant Mol. Biol. 30:679-
534 .

Kendall BI, McKersie BD (1989) Free radical and freezing injury to cell
membranes of winter wheat. Physiol. Plant. 76:86-94

27

Knight MR, Campbell AK, Smith SM, Trewavas A] (1991) Transgenic plant
aequorin reports the effects of touch and cold-shock and elicitors on cytoplasmic
calcium. Nature 352:524-526

Knight H, Trewavas AI, Knight MR (1996) Cold calcium signaling in
Arabidopsis involves two cellular pools and a change in calcium signature after
acclimation. The Plant Cell 8:489-503

Krhhna P, Sacco M, Cherutti IF, Hill S (1995) Cold-induced accumulation of
hsp90 transcripts in Russia: nspus. Plant Physiol. 107:915-923

Kurkela S, Borg-Franck M (1992) Structureandexpressionofla'nz, oneoftwo
cold-andABA-induced gqresofArsbidmisduslism. PlantMol. Biol. 19:689-692

Kurkela S, Franck M (1990) Cloning and characterization of a cold- and ABA-
inducible Arabidopsis gene. Plant Mol. Biol. 15:137-144

LdkLDorfflingK(1%5)Hardening,abscisicacid,prolineandfreezing
resistance in two winhr wheat varieties. Physiol. Plant. 63:287-292

Lang V, Heine P, Palva ET (1%9) Low temperature acclimation and treament
wifiexogenous abscisic acid induce common polypepfidesinArsbidapsis dualism
(L) Heynlr. Theor. Appl. Genet. 77:729-734

Lang V, Palva ET (1992) The expression of rah-related gene, rsb18, is induced by
abscisic acid during the cold acclirnation process of Arsbidopsis dualism (L)
Heynh. Plant Mol. Biol. 20:951-962

Lee SP, Chen THH (1993) Molecular cloning of abscisic acid-responsive mRNAs
expressed during fire induction of freezing tolerance in bromegrass (Browns
islands Leyss) suspension culture. Plant Physiol. 101:1089-1096 -

Levya A, Iarillo IA, Salinas I, Martinez-Zapater IM (1995) Low temperature
induces the accumulation of phenylalanine mnwuis-lysse and duslcoue syndrsse
mRNAs of Arsbidopsis duslism' in a light-dependent manner. Plant Physiol.
108:39-46

Li Q-B, Anderson IV, Guy CL (1994) A cDNA clone encoding a spinach 70-
kilodalton heat—shock cognate. Plant Physiol. 105:457-458

Lin C, Thomashow MF (1992) DNA sequence analysis of a complementary DNA
for cold- -regulated Arsbidopsis gene curls and characterization of fire COR15
polypeptide. Plant Physiol. 99:519-525

28

Los D, Horvatlr I, Vigh L, Murata N (1993) The temperature-dependent
expressionofthedsssturasegenedrsAinSysedussystisPCCQOS. FEBS
LE'l'I'ERS318:57-60

LuG,deLisleAI,deVettenNC,FerlRI(1992)Brainproteinsirr_plants: An
Arsbidapsis homolog b neurotransmitter pathway activators is part of a'DNA
binding complex. PNAS 89:11490-11494

Lynch DV, Steponkus PL (1987) Plasma membrane lipid alterations associated
with cold acclimation ofwinter rye seedlings (Secsle meals L. cv Puma). Plant
Physiol. 83: 762-767

Markhsr'tlllAl-l(1%6)Chillinginjury:areviewofpossiblecauses. Hortscience
21: 1329-1333

Msntyls E, Lang V, Palva ET (1995) Role of abscisic acid in drought-induced
freezing tolerance, cold acclimation, and accumulation of LTI78 and RABIB
proteins in Arabidopsis dialism. Plant Physiol. 107:141-148

Marans C, Garcia-Olmedo F, Carbonero P (1990) Differential expression of two
types of sucrose synfiretase-errcoding genes in wheat in response to anaerobiosis,
cold shock and light. Gene 88:167-172

McKersieBD,CharY, deBeusM, BowleySR, BowlerC, InzeD, D'HalluinK.
BottermanI (1993) Superoxide dismutsseenharrces toleranceoffreezing stressin
bansgerric alfalfa (Medicago sativa 1...). Plant Physiol. 103:1155-1163

Minorsky PV (1%5) An heuristic hypothesis ofchilling injury in plants: a role
for calcium as fire primary physiological transducer of injury. Plant, Cell and
Environ. 8:75-94

Minorsky W (1989) Temperature sensing by plants: a review and hypofiresis.
Plant, Cell and Environ. 12:119-135

Miquel M, Iames D, Dooner H, Browse] (1993) Arabidopsis requires
polyunsaturated lipids for low-temperature survival. PNAS 90:6213-6212

Mizoguchi T, Hayaslrida N, Yamaguchi-Shinozaki K. Kamada H, Shinozaki K
(1995) Two genes first encode ribosomal-protein S6 kinase homologs are induced
by cold or salinity stress in Arabidopsis dualism. FEBS Lett. 358:199-204

29
Mizoguchi T, lrie K Hirayama T, Hayashida N, Yamaguchi-Shinozaki K
Mstsumote K Shinozski K (1996) A gene encoding a mitogen-activamd protein
kinsse kinsse kinsse is induced simultaneously with genes for a mitogen-
activated kinsseand anS6 ribosomal proteinkinaseby touch, cold andwater
stress in Arabidapsis dualism. PNAS 93:765-769

MohspatraSS,WolfraiinL, Poole RI, Dlrirrdsa R80”) Molecularcloningand
relationship to freedng tolerance of cold-acclimation-specific genes of alfalfa.
Plant Physiol. 89:375-380 .

Monroy AF, Castonguay Y, Laberge S, Ssrhan F, Vezina LP, Dhindsa RS (1993a)
A new cold-induced alfalfa gene is associated wifir errhsnced hardening at
subzero mpaahin. Plant Physiol. 102:873-879

Monroy AF, Ssrhan F, Dhindsa RS (1996b) Cold-induced changesin freezing
tolerance, protein phosphorylation, and gene expression. Plant Physiol.
102:1227-1235

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

Moynihan MR, Ordenfiich A, Raskin I (1995) Chilling-induced lrest evolution in
plants. Plant Physiol. 108:995-999 ‘

MurataT(1990)Relationofchillingstresstomembranepermeability. ln,CY
Wang, ed., Chilling injury of horticultural crops. CRC Press, Irrc., Boca Raton,
Florida,pp201-209

Murats N, Nishida 1(1990) Lipids in relation to chilling sensitivity in plants. In,
CY Wang, ed., Chilling injury of horticultural crops. CRC Press, Inc., Boca
Ratorr, Florida; pp 181-199

Murata N, lshizski-Nishizawa o, Higashi s, Hayashi H, Tasska Y, Nishida l
(1992) Genetically engineered alteration in fire chilling sensitivity of plants.
Nature 356:710-713

Neven LG Haskell DW, Guy CL, Denslow N, Klein PA, Green LG, Silverman A
(1992) Association of 70-kilodalton heat-shock cognate proteins wifir acclimation
to cold. Plant Physiol. 99:1362-1369 -

Neven LG, Haskell DW, Hofig A, u Q—B, Guy CL (1993) Characterization of a
spinachgene responsive to low temperature and water stress. Plant Mol. Biol.
21:291-305 .

30

Newsted WI, Chibbar RN, Georges F (1991) Effect of low temperature stress on
fire expression of sucrose synfiretase in spring and winbr wlrest plants.
Development of a monoclonal antibody against wheat germ sucrose synfiretase.
Biochem. Cell Biol. 69:36-41

Nordin K Vahsla T, Palva ET (1993) Differential expression of two related, low-
temperature-induced genes inArsbidapsis dualism (1..) Heynh. PlantMol. Biol.
21:641-653

Orr W, White TC, In B, Robert L, Singh] (1995) Characterization of a low-
temperature—induced cDNA from winter Brassics mpus encoding fire 70kDa
subunit of tonoplast ATPase. Plant Mol. Biol. 28:943-948

Patterson BD, Reid MS (1990) Genetic and environmental influences on fire

expressionofchillinginjury. In,CYWang,ed., Chillinginjuryofhorticultural
crops. CRCPress,Inc., BocaRatorrFlorida,pp87-112

Pihakaski-Maunsbach K Puhakainen T (1995) Effect of cold exposure on cortical
microtubules of rye (Secslc cereals) as observed by immunocytoclremistry.
Physiol. Plant. 93:563-571

Purvis AC, Shewfelt RI. (1993) Does fire alternate pathway ameliorab chilling
injury in sensitive plant fissues? Phys. Plant. 88:712-718

Raghuram N, Sopory SK (1995) Evidence for some common signal transduction
events for opposite regulation of nitrate reductase and phytochrome-l gene
expression by light. Plant Mol. Biol. 29:25-35

RaisonIKOrrGR(1990)Pmposalsforabetterunderstsndingoffiremolecular
basisofchillinginjury. In,CYWang,ed., Chillinginjuryofhorticulturalcrops.
CRC Press, Inc., Boca Raton, Florida, pp 145-164

Saw-Vasquez I, Raynsl M, Delsney M (1995) A rapeseed cold-inducible
transcript encodes a phosphoenolpyruvate carboxykinase. Plant Physiol.
109:611-618

Sakai A, Larcher W (1987) Frost survival of plants: response and adaptation to
freezing stress. Spring-Verlag, Berlin, Heidelberg New York, London, Paris,

Tokyo

Sakai A, Sugawara Y (1978) Survival of plant germplasm in liquid nitrogen. In,
PH Li, A Sakai, eds., Plant cold hardiness and freezing sfi'ess: mechanisms and
crop implications. Academic Press, New York, San Francisco, London, pp 345—
359

31

Simirrovitch D, Ooutier Y (1993) Drought and freezing tolerance and adaptation
in plant: Some evidence for near equivalerrces. Cryobiology 20:487-503

pronkusPLOQQRoleoffireplasmamembraneinfreezinginjuryandcold
acclunstion. AnrrRev.PlsntPhysiol.35:543-584

Shponkus PL, Lynch DV (1W9) Freeze/ thaw-induced destabilization of fire
membrane and effects of cold acclimation. I. of Bioenergetics and
Biomembranes 21 :21-41

Susek RE, Ausubel FM, OroryI (1993) Signal transduction mutants of Arabidopsis
urrcouple nuclear CAB and RBCS gene expression from chloroplast
development. Cell 74:787-799

Sutton F, Ding X, Kenefick DG (1992) Group 3 LEA gene HVAl regulation by
cold acclimation and deacclimation in two barley cultivars wifir varying freeze
resistance. Plant Physiol. 99:338-340

Takahashi T, Naito S, Komeda Y (1992) The Arabidopsis HSP18.2 promoter/ GUS
gene fusion in transgenic Arabidopsis plants: a powerful tool for fire isolation of
regulatory mutants of fire heat-shock response. The Plant Iourn. 2:751-761

Thomashow MF (1993) Genes induced during cold acclimation in higher plants.
In, PL Steponkus, ed., Advances in low. temperature biology, voL 2. IAI Press
Ltd., London, pp 183-210

Thomashow MF (1994) Arabidapsis as a model for studying mechanisms of plant
cold tolerance. In, EM Meyerowitz, CR Somerville, eds., Arabidopsis. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, New York, pp 807-834

Uemura M, Joseph RA, Steponkus PL (1995) Cold acclimation of Arsbidopsis
dualisius: Effect of plasma membrane lipid composition and freeze-induced
lesions. Plant Physiol. 109:15-30 -

van Zee K Oren FQ, Hayes PM, Close TI, Oren THH (1995) Cold-specific
induction of a dehydrin gene family member in barley. Plant Physiol. 108: 1233-
1239

Vigh L, Los DA, Horvafir I, Murata N (1993) The primary signal in fire biological
perception of temperature: Pd-catalyzed hydrogenation of membrane lipids
stimulated fire expression of the desA gene in Syneduocystis PCC 6803. PNAS
”9090-9094

32

Wang X, Sato R, Brown MS, Hus X, GoldsteinIL(1994)SREBP-1, a membrane-
bound transcription factor released by sterol-regulated proteolysis. Cell 77:53-62

Webb MS, Steponkus PL (1993) Freeze-induced membrane ultrastructural
alterations in rye (Seoul: cereals) leaves. Plant Physiol. 101:955-963

Weiser RL, Wallner SI, Waddell IW (1990) Cell wall and extensin changes during
cold acclimation of pea seedlings. Plant Physiol. 93:1021-1026

Welin BV, Olson A, Nylander M, Palva ET (1994) Characterization and
differential expression of tflinfles/rab-like genes during cold acclimation and
drought stress in Arabidopsis dualism. Plant Mol. Biol. 26:131-144

Welin BV, Olson A, Palva ET (1995) Structure and organization of two closely
related low-temperatureinduced dlm/leq/rab-like genes in Arabidopsis dualism L.
Heyflr. Plant Mol. Biol. 29:391-395

Whik AI, Dunn MA, Brown K Hughes MA (1994) Comparative analysis of
genomic sequence and expression of a lipid transfer protein gene family in
winhr barley. Iourn. of Exp. Bot. 45:1885:1892

Wilhelm KS, Thomashow MF (1993) Arabidopsis dualism COR15b, an apparent
homologue of COR15a, is strongly responsive to cold and ABA, but not drought.
Plant Mol. Biol. 23:1073-1077

WifliamsLBulmanM,HutfiyA,PhillipsA,NeillS(1994)Oraracbrizafionofa
cDNA from Arabidopsis dualism encoding a potential tlriol protease whose
expressionisinducedindependenfiybywiltingandabscisic acid. Plant Mol.
“25:259-270

Wilson IM (1979) Drought resistance as related to low temperature stress. In, IM
Lyons, D Graham, IK Raison, eds., Low temperature stress in crop plants: fire
role of fire membrane. Academic Press, New York, London, Sydney, Toronto,
San Francisco, pp 47-65

Wolfraim LA, lanai: R, Tyson H, Dhindsa RS (1993) cDNA sequence,
expression and transcript stability of a cold acclimation-specific gene, cas18, of
alfalfa (Mediagofilcata) cells. Plant Physiol. 101:1275-1282

YamaguchiShinozski K Shinozaki K (1993) Oraracterization 'of fire expression
of a desiccation-responsive rd29 gene of Arabidopsis dualism and fire analysis of
its promoter in transgenic plank. Mol. Gen. Genet. 236:331-340

33

Yanr’aguchi-Slrinozaki K, Slrinouki 1((1994) A novel en's-acting element m an
Arabidopsis gene is involved in responsiveness to drought, low-temperature, or
high-salt stress. The Plant Cell 6:251-264

Zhou J, Lolr Y-T, Bressan RA, Martin GB (1995) The tomato gene PtiI encodes a
serine/firreonine kinase first is phosphorylated by Pto and is‘involved in fire
hypersensitive response. Cell 83925-935

Chapter 2:

Screen for mutations in the signal transduction pathway
responsible for induction of COR15a

Summary

In preparation for gene regulation studies at fire promoter level, a genomic clone
containing COR15a, a gene inducible by low temperature, drought and abscisic
acid (ABA) and encoding a chloroplast-targeted protein, was isolated.
Approximately one kilobase of fire COR15a promoter was firen taken from this
clone and fused to fire coding region of beta-glucuronidase (GUS) in order to
find elements of fire signal transduction pathway responsible for fire induction of
COR15a. Transgenic plants carrying firis construct were mutagenized and
screened for fire altered expression of GUS activity in warm-grown or cold-
treated plants. Plants detectably altered in GUS expression were found in bofir
screens. Some of these plants do not contain fire transgene, while others appear
to have a mutation that only affects fire transgene. None show altered
expression of endogenous COR15a, and hence do not carry mutations in fire
signal transduction pathway responsible for fire induction of COR15a. Means of

improving fire mutagenesis strategy are discussed.

34

35

Introduction

Plants, being stationary poikilotherms, require first a series of physiological
changes occur when fire temperature becomes too low for optimal growth. Of
those plants that are able to wifirstand chilling temperatures, some are also able
toaclfievesnhwreasedleveloffreezingtoleranceafterexposuretolowmn—
freezing temperatures in a process known as cold acclimation (Levitt, 1%0).
Among fire many metabolic alterations first occur during firis process are

changes in gene expression (Thomashow, 1994).

In Arabidspsis dualism, a plant first is able to cold acclimate (Gilmour et al., 1%8),
fire mRNA levels of a number of COR (cold-regulated) genes have been found to
increase dramatically during low temperature treatment (Thomashow, 1993). A
subset of fire genes induced at low temperature, including members of gene
families COR78, COR47, COR15 and COR6.6 have a number of properties in
common. Themessagelevelsoffiresegenesremainhighaslongasplantsare
maintained at low temperature, but decrease rapidly when fire plants are
returned to normal growth temperatures (Thomashow, 1994). They are also
induced by water stress and / or by fire exogenous application of abscisic acid
(ABA), but not by heat shock. Finally, fire proteins encoded by firese genes
remainsolubleuponboflingapropertyfintisfiroughttobeaconsequenceof

m high hydrophilicity (Hajela et al., 1990).

36
ThemeclunismbywhichCORgenesarereguhtedisofintaestforpracficslas
wellaspurelyacademicressons. Aprogramusinggeneticengineeringtoslter
plantfsesm‘ngtoleraneemusttakeintoaccountfirereguhfionoffiregenes
involvedaswellasfireproperh’esoffiregenesfiremselves. Atfire-startoffiris
worknofiringwas known aboutcis-orb’sss-actingelements involvedinfire
krduction of any number of fire four COR gene families. When experiments
analyzing fire regulation offire COR genes were initiated, nuclear run-on assays
indicated first only fire COR15 gene family was regulated primarily at fire
transcriptional level (Hajela et al., 1990). Thus, it was chosen for promoter

analysis experiments.

A genomic clone containing COR15a was isolated and promoter deletion
analysis was performed. In firis analysis, fire COR15a promoter was found to be
inducible by low temperature, drought and exogenously applied ABA (Baker et
al., 1994). A mutagenesis approach designed to identify proteins involved in a
signal transduction pathway first activates transcription of COR15a was also
undertaken. Because firere is no easily detectable phenotype specific for fire
induction of COR15a, one of fire constructs used in fire promoter deletion
analysis, which contained approximately one kilobase of fire COR15a promoter
fused to fire GUS coding region, was used as a screenable marker. The
expression of GUS from this construct should mimic expression of fire

endogenous COR15a gene in transgenic plants.

37
Mutational analyses of reporter gern fusions of firis type have been successfully

usedtoexaminefinsigmhrespomibleforfinhrducfionofgenesresponsiveto
heat-shock (Takahsshi et al., 1992), to systemic acquired resistance (Bowling et
al., 1994 and Cao et al., 1994), and to chloroplast development (Susek et al., 1993)
alfiroughfinidentitiesoffinmutated geneshaveyettobereported. Ascreen
using fire aabliO promoter fused to #1152 yielded, unexpectedly, a cosuppressed

mutant (Brusslan et al. 1993).

It was hoped first mutants obtairnd could be used to denrmirn wlnfirer all fire
COR genes were controlled by a common pafirway, to see if mutants altered in
fireiu'resposuetooneoffininducingstimuliwouldalsobealteredinfinir
responsetofireothers, toestimatehowmanystepswereinfinpafirwayand
ultimately to clone the genes involved. In fire end, however, fire experiment
only suggesnd ways to improve fire screen, as no plants wifir mutations in fire
signal transduction pafirway involved in fire induction of COR15a were '

Results

Isolation of a genomic clone encoding COR15a

A genomic library of Arsbidopsis DNA in arms was screernd using pHH67
(Hajela, 1990), a cDNA corresponding to COR15a, as a probe. A phage
containingsninsertofapproximately 12kilobases ofDNA, wifirfireS’ end of
COR15a located approximately 3 kilobases from one end, was isolated. The

insert DNA from firis phage was subclorred into pUC19 and was furfirer

manipulated in first form.

Deletion studies (Baker et al., 1994) demonstrated first the COR15a promoter is
responsive to low temperature, drought and exogenously applied abscisic acid

(ABA).

Mntagenesis strategy

Because finre is no recognized plnnotype for fire induction of COR15a, a system
wasdevisedtofacilitateascreenformutants firstwerealteredinfiresignal
transduction pathway responsible for fin induction of firis gene (Figure 7.1).
The promonr of COR15a fused to fire coding region of GUS was introduced into
plank for use as a screenable marker. Transgenic plants first expressed fin

reporter gern, GUS, at a cold temperature (4°C), but not at normal growfir

 

ass 4»
ms, my“
plant
transformation

\GUS expression
EMS . (fluorescence)

mutagenesis
& Screen

 

warm GUS expression
/ (fluorescence)
N‘

Mun!“ no GUS

Figure 2.1 Screen for sigrrsl transduction mutants. A construct consisting of fire
promoter of COR15a, fire coding region of GUS and a NOS terminator was
introduced into Arabidopsis. A transgenic lirn first strongly expressed GUS at
low nmperature but not at normal growfir temperatures and was homozygous
forfiretransgenic insertwas EMS mutagenized and screened usinga MUG
assay. Plants kept at normal growfir temperatures were firen screened for GUS
expression and cold-treated plants were tested for the lack of GUS expression.
Contrary to depiction, plants were screened after about firree weeks of growfir,
at which point they were not yet flowering.

40
nmperature (22°C), were EMS (ethyl mefirsrn sulfonste) mutagenized and
screened at 22°C, looking for those wifir elevated expression, or at 4°C, looking
for decreased expression.

Alfiroughcoldinductionisfinresponseofinterest, alterationoffinexpression
offirisprornotercouldalsobeduetoamutationinfiresignsl(s)initiatedby

drought or ABA, since fire COR15a promoter is also responsive to firese stresses.

Prepmtioueftransgenicreporterlines,mutagenesisandscreening

A line of transgenic plants containing approximately orn kilobase of fin COR15a
promonr (fin 800/ +78 fragment described by Baker et al., 1994) fused to fire
GUS coding region of p31 101.2 (Jefferson et al., 1987) was selected because of its
strongcold-inducibleexpressionofGUSastested byhistochemical analysis. It
hasbeenslrown firatfirisconstructis alsoirrduciblebydroughtandby
exogenously applied ABA (Baker et al., 1994). The R3 generation (fin firird
generation after regeneration of fin transgenic plants) was tested for
homozygosity first on Kansmycin plates and firen by genomic Soufinrns done
on eleven individual plants (Figure 2.2). Furfinr genomic Soufirern analysis of
firisline(referredtoasl,6) suggests firatfirerearefirreeinsertsattwolociand
first firetwolinkedinsertsareinverted wifirrespecttoeachofirer (Figure2.3). It
was hoped first having more firan orn insert would prevent fin isolation of cis

knockout mutations in fin transgene, since it should be less likely for multiple

41

SnaBI BamHI
123456789101 1234567 89191.]

 

       

Figure 2.2 Test for homozygosity of inserts in line 1,6, using genomic Southems. A line
of transgenic plants containing a -900 to +78 COR15a promoter fragment fused to the
GUS coding region was tested for homozygosity of the insert(s). Preparations of
genomic DNA from 11 individual R3 plants of transgenic line 1,6 were digested with Sna
BI, which cuts once within the inserted DNA, and with Bam HI, which does not cut
within the insert, to determine if fire previous generation (R2) of line 1,6 was
homozygous for the transgenic insert.

42

 

 

Book]
In ECRI
nptII idA
" H GUS Kan
E H P S E H P S
W 3

 

 

 

 

 

 

 

 

Figure 2.3 Putative organization of inserts in line 1,6. Genomic DNA isolated
from transgenic line 1,6 was digested with enzymes EcoRI (E), HindIII (H), PstI
(P) and SacI (S) and probed for the GUS (uidA) or Kan coding regions. The
autoradiograms are shown to the right and the putative genomic arrangement is
diagrammed at the left. The three bands in the SacI digest suggest that there are
three copies of the insert, and the GUS band shared between the Hind III and Pst
I digests could be due to two copies of the transgene inserted tail to tail. The
insert thought to be separate from the two linked inserts is shown four times.
once for each restriction enzyme. All four enzymes are shown at once on the
diagram of the two adjoining inserts.

43

insuta to he mutated at once. However, it also provides more targets for cis-
acting mutations that might increase expression in the warm. Genetic analysis of
a cross between this line and nontransgenic Arabidopsis (ecotype Columbia) gave
ratiosconsishntwiththepreaenceoftwoGUS—expressingloci“ noninducible
plantsoutoflOOhatedinIonpulationaé),althoughonlyonelocusconferred
Kanamycin resistance (120 resistant to 41 sensitive, 83 resistant to 26 sensitive
andélresistanttofisensitive, inF2populationsa2, a3and a6respectively).
Thissuggesbthatfineimerfls) atoneoftheh'ansgeniclocinolongerhasa

functional gene for kanamycin resistance.

Seeds of line 1,6 were then mutagenized with EMS and planted in 12 flats.

Insect damage (principally due to thrips and aphids) severely limited the M2
(second mutantgeneration) seed productionofthisand ofthesubsequentround
ofrnntagelwsis. M2seedwasharvestedinbulkfromeachflat, rendtingian
poolsofseed. Anestimated 6,560seedswereplanted onplatescontaining DAP
(2,6-diaminopurine) to test the effectiveness of mutagenesis. Plants homozygous
for recessivemutationsirittieaptgeneareableto growinthepresenceofDAP,
sothenumberofresistantplants providesameasm'eoftheeffectivenessof
mutagenesis (Moffatt and Somerville, 1%8). Four resistant plants were
observed, dmonstrating that the phenotype of a single recessive mutation was

detectable in about one out of 1,600 plants. Later, more seed was mutagenized

44

inthesamemannerandumorepoolsweregenerated. TheDAPtestwasnot

performed on this mutagenized seed.

Seed from each pool was planted in pots that were divided into quarters.

Leaves of individual plants from each quadrant were clipped and placed into a

2 mM methylumbelliferyl glucuronide (MUG) solution in microtiter plate wells
(Figure 2.4). Quadrant that contained a leaf of mutant phenotype (GUS
expression at normal growth temperatures or a lack of GUS expression 4°C) were
then rescreened and each plant was marked with a numbered toothpick. About
halfoffisepohofplanhwerescreenedmfliecoldmomandflieoflierhalfmflie

growth chamber at standard growth conditions.

Putative mutants recovered ,

In due first round of screening, approximately 26 strong and 54 weak putative
mutants inthewarm screenand 36 strong and 67weak putativemutants inthe
cold screen were selecbd and allowed to set seed. The next generation (M3) of
the strong putative mutants was tested. Six of those selected in the warm screen
(Wlw, Hflw, 6-8#10w, 6-8#15w, 18#6w and 13-7#1w) and one of those
selected in the cold screen (3-264c) again showed the mutant phenotype. The
weak putative mutants were also tested, but none of those isolated in the warm
screen displayed the mutant phenotype in the second generation. The twelve

putative mutants from the cold screen that still showed a weak mutant

45

1,9 36—]#6C 38-1#6C 3—2#4C 20-8#3C l3—7#1w18#6w 4-11#3w 6—8#4w
"'er ”IH‘QL C"? 9.49

_ _ ‘ x'QQtQOQO

W , . @9’090 k4“

    

>8

":6
3%

56
' @o
toes-J

6.» I
so on
’6' -

;Q;Q‘! 1

cos
we;
eras

3
3'
8

Figure 2.4 MUG assay of putative mutants. Two columns of microtiter
plate wells are shown for each line of plants, with nonmutagenized 1,6
serving as the control. Leaves in the wells in the top half of the plate were
clipped from warm-grown plants (W, 22°C), while those in the bottom half
of the plate were taken from cold-grown plants (C, 25°C).

46
phenotypeweunotsmdiedfurflwrbecauufiievafiabihtyofflseirplmwtype

made confirmation and characterization difficult. Initial testing of the non-
mutagenized reporbr transgenic had made it appear that the warm and cold
response of the MUG assay was clear, but substantial variability was seen during
screening, evenmfiuconhohmakingitdifficulttoidentifyh'uemutants. The
reasonforthisvariabilityisnotknown, althoughithasbeenseenthatdifferent
leaves from the same plant can give different levels of expression of the reporter
gene (Figure 2.5). The number of false positives was decreased in later screening
by renting the putative mutants a week or two after the initial screen, although

the improvement was not quanh’fied.

A second round of screening turned up three putative mutants that expressed
GUS at normal growth temperatures and three that did not express GUS at low
hmperature. Ofthese, two ofthose thatexpress GUS atnormal growth
temperatures (6-10#14w and 4-11#3w) and one of those that does not express
GUS at low temperature (20-8#3c) were still positive in the next generation. Two
more plants that did not express GUS at low hmperature (36-1c and 38-1c) were
confirmed from yet a third round of screening. In all, approximately 13,000
plants were screened. Approximately 39 lines of putative mutants that do not
express GUS at low temperature and 9 putative mutants that express GUS at

normal growth temperatures remain to be tested at the M3 generation.

1,6 l 6 xl8#6w F2a6

 

W C 1 2 3 4

 

Figure 2.5 Variability of MUG assay. Four leaves from each of 12 warm-grown plants
from the cross 1,6x18#6w (F2 generation, population a6) were screened for MUG
activity. The identity of these plants is included for the sake of completeness, but is not
important for the observation that different leaves from the same plant give different
results in the MUG assay. Leaves from the twelve plants are shown in twelve rows, with
the leaf number being indicated at the top of the microtiter plate. Leaf number does not
correspond to the position of the leaf on the plant. Columns of nonmutagenized 1,6
grown at standard temperatures (W) and at 25°C (C) are included for comparison.

48
Inordertodetermmewhefiierhneswluchsfiflslwweddiemutantphenotypem

the M3 generation were affected in a signal transduction pathway responsible

for inducing COR15a, northern blots of each were probed wifli GUS and COR15.

In all cases, the endogenous COR15 message of the plants selecbd as mutants
was indistinguishable from that of the nonmutagenized transgenic control
(Figure 2.6), except for line 13-7#1w, which was later found to contain a CaMV
358 promoter fragment (data not shown), and thus was not derived from the
original transgenic line. Line 13-7#1w was not studied further. In the figure
shown, it appears as though the endogenous COR15 message may be elevated
over background for line 18#6w, but this was not reproducible. However, the
GUS message was elevabd in plants selected as expressing GUS at normal
growth temperatures (4-11#3w, 6-8#1w, mm, 6-8#10w, s-ss15w, s-iosuw,
13-7#1w and 18#6w) and was not detectable in plants chosen as not expressing
GUS at low bumerature (3-2#4c, 20-8#3c, 36-1c and 38-1c). In genomic Southern
analysis, theilatter were found to not contain the reporter (Figure 2.7), so they
wa'enotstudied further. Itseems unlikely thatalltlueeinsertsofthetransgene
would have been deleted by EMS mutagenesis, suggesting that these may be
contaminants. Indeed, a subsequent genomic Southern of line 38-1c showed
that it contained an extra copy of the COR15a coding region indistinguishable

from a CaMV 355/ COR15a sense transgenic used in the lab (data not shown).

49

£33 33
3 LOHSCOH 3
VHH4t©=fl==tt H
=tt:=tt:=tt:O#\—1[\ 2s:
°9°?°?"‘oo".‘c¢'a @

C W\D\O\D\éF-t<l‘r-IW
GUS Okhfi .

        

38-1#6c
36-1#6c

(X
is
«58

:3 Mt

Figure 2.6 Expression of GUS and of COR15 mRN A in putative mutants.
Samples of warm- (W) and cold- (C) grown nonmutagenized 1,6 are included as
controls on each blot. Lines chosen as having increased GUS activity at 22°C
(normal growth temperature) were grown at 22°C for this experiment. These are
shown at the top of the page. Lines which did not have GUS activity at low
temperatures (25°C) were grown at 25°C for this experiment and are shown at

the bottom.

50

 

 

sééé

Bofiiz”

M x-Td’am «3%

COR15*"§."' s *
GUS .

 

 

 

Figure 2.7 Southern analysis of putative mutants screened at low temperature.
Duplicate blots of EcoRI-cut genomic DNA were probed for COR15 and for GUS
sequences. RLD is an untransformed control, which should only contain COR15
sequences; 1,6 is the nonmutagenized transgenic control, which should contain
GUS as well as COR15 sequences; and the other four lines are putative mutants
that were chosen for their lack of GUS expression at low temperature. These
were expected to contain both GUS and COR15 sequences. Both endogenous
copies of COR15 are contained within a single EcoRI fragment, while the
introduced copies of the transgene contain GUS on an EcoRI fragment internal to
the insert (see Figure 2.3).

51
ItwasalsonobdfiratfireGUSmessageproducedbylineflflw (selectedas
expressing GUS at normal growth temperatures) was still responsive to low
temperature, drought and ABA treatment, and that the signal was lowered by
incubation in water, as is true with fire nonmutagenized control (Figure 2.8).
Line mu may be slightly lea. responsive to drought firan 1,6. Lines 6-8#1w,
m10w, 68015w and 6—10#14w were not included in this or subsequent
analysesbecausefireyarefromfiresamepoolofMZseedaslineb-Bflwandare
likely to be siblings ofit. There has been no detected difference among these

five putative mutants.

Genetic analysis of putative mutants

Because the mutations did not appear to be in the signal transduction pathway,
and because cis-acting elements in the COR15a promoter are only partially
understood, it was of interest to determine whether any of the mutations that
caused elevated levels ofGUS message in warm-grown plants were in fire
introduced pita. lfso,sndifthemutationwasnotintheGUScodingregion,
isolation and examination of the altered genes might give some insight into fire
function of COR15a promoter elements. If not, firen fire mutation would have to
be in a gene that affected the introduced gene and not endogenous COR15,
perhaps influencing the stability of fire GUS message. Should such be fire case,
more care would have to be taken to avoid or quickly detect such mutations in

subsequent mutagenesis attempts.

52

_6-8__#4w 66-8_#4w
WSI6ADWSAD1CC

COR15 .

 

Figure 2.8 lnducibility of putative mutant line 6—8#4w. Plants of
nonmutagenized 1,6 and putative mutant line 6-8#4w were grown under normal
conditions (22°C, W), soaked in water (S), soaked in 100 uM ABA (A), drought-
stressed (D) or cold-stressed (25°C, C). See materials and methods for details.
The RNA was probed for GUS and for COR15 messages.

53
Ihefh'stsbpinfiusunlysiswutodeterminewhefirerfiremutantphenotype
congregatedwifirfireinuoducedgermandifso,wifirwhichlocus. Lines
4.11m», 18M and straw were crossed to fire non-mutagenized reporter line
and to nonfiansgenic Arabidapsr's, ecotype Columbia. The F1 generation of each
cross demonstrated fire mutant GUS phenotype, and in the F2 generation, each
putativenrutantsegregatedataboutiibrighttol dark, indicatingfiratthe
mutation was dominant. Soufirerns were done on individual bright plants of
crosses between Columbia and lines 18%w, 4-11#3w and 6-8#4w, and in each
case, when there was only one locus represented, it was fire one expected to
contain fire two linked transgenes (data shown for 4-11#3w in Figure 2.9). The
elevenlinesshownshouldbeenoughfor95%confidenceoflinkage(see

MaterialsandMethodsfordetails).

This co-segregation of fire mutant phenotype wifir fire transgerres was consistent
wifirfireideafiratfiremutationiseifirerinfireintroduced geneorislinkedtoit,
alfiroughnofiringcanbesaidregardinghowtightfirelinkageis. Italso
supported fire conclusion firat fire mutation is not in fire signal transduction
pathwayand furthersuggested firatitisalsonotinanunlinkedgenefirataffects

GUS stability.

Preliminary attempts to isolate fire locus believed to contain fire mutation were

unsuccessful. Assuming first this locus does contain two copies of fire transgene,

54

RLDx4-ll#3w
1,612 3 45 6 7 3 91011
individual insert ”mi-““9 “fir-- an , H

.5

lmked inserts “*m y - -

 
     

 

Figure 2.9 Cosegregation of the mutant phenotype with the locus thought to
contain two linked copies of the reporter construct. Wildtype Arabidopsis,
ecotype RLD, was crossed as the female parent to putative mutant line 4-11#3w.
Genomic Southerns were performed on PstI-cut DNA isolated from 11 individual
F2 plants which had shown the mutant phenotype by the MUG assay. This
digest (see Figure 2.3) separates two loci containing the transgene, with the
upper band being that thought to contain a single insert and the lower band
thought to contain two linked copies of the transgene. The blots were probed for
the GUS coding region. Nonmutagenized line 1,6 is shown at left as the control.

55
bohwoflduedbbesequwedandfirepossibfiitydoessfillerdstfiutfire

mutationisinfireGUScodingregion. Forfiusereasonthwasdetermined firat

resourcescouldbemoreeffecfivelyinvestedinlessriskyprojects.

Testforpolyploidy
Onepossibleexphnafionforfirelackofreeessivemuhfioruwasfiutfireoflginal
transmic line may have become tetraploid during fire tissue culture stage
(Negrufiu et al., 1975) of fire transformation process, firus making recessive
mutations unobservable. However, if firis were the case, the crosses of 1,6 to
nonh'ansgenicArabidopsisshouldhavepr-oduced abortedseedinfireFZ
gsnerafionbecsuseoffirehigh'amount ofnondisjunction expected whens
waploidsnddiploidarecrossed.11reF2seedoffiriscrosswsswfidtypein
appearance and germinated normally, demonstrating first fire fiansgenic was not

biraploid (data not shown).
Discussion

In fire examination of COR gene regulation, a genomic clone was needed so firat
promoter analyses could be conducted. A cDNA corresponding to fire coding
region of COR15a was used to isolate from a genomic library a clone first

contained about 3kb of sequence 5’ of COR15a and 8 to With 3’. Deletion

56
analysis of fire COR15a promoter found on firis clone demonstrated first firis

promoter is inducible by low temperature, drought and ABA (Baker et al., 1994).

hrsnattempttaisolabmutantsalteredinfiresignalfiamducfionpafirway
responsible for fire induction of low temperaturevresponsive genes, plants
containingfireCORlSapromohrfrmd tofireGUScodingregionwereselected,
EMSmutagenized and screened for altered GUS expression. Approximately
6,000 plants were screened at normal growth temperatures and anofirer 6,000 at
4°C and a large number of putative mutants was isolated. While these numbers
are by no means exhaustive, growing mutant seed on DAP had already shown
first one recessive mutation in fire apt gene could be found in about 1,600 plants.
lnofirersysterns, ascreenlookingformutantsinfiresignal transduction
pathway for heat-shock inducible genes yielded about one mutant per 5,000
plants‘screened (l‘skshashietal., 1992), and inofirerscreensdoneusingreporter
constructs, one mutation was found in 2,000 or fewer plants screened (Cso et al.,
1994, Bowling et al., 1994 and Brusslsn et al., 1993). Practically, no more plants
were screened because fire available resources went into characterizing fire

mutants isolated, which turned out to be false positives.

Of firose screened, only 8 (9.6%) of fire initial putative mutants from fire warm
screen and 4 (3.7%) of fire irritisl putative mutants from fire cold screen still

looked mutant in fire next generation, indicating first fire screen allowed for a

57
large percentage of false positives. The substantial variability seen during
meningevesrinfirecontrols,maybeduetofirestabilityoffireGUSprotein
mrmcoupied wifirasensitivityoffireCORlSapromoter-toleafto
leafvaristioninmicroenvironmenh. Testingasecondleafafhrfireplanthads
chansetogrowforanofirerweekortwodideliminsteanumberoffalse

positives more quickly firan waiting to test fire next generation.

(X fire putative mutants first-came through secondary screening. none were
found to contain a mutation in fire signal transduction pafirway responsible for
the induction of COR15a. Putative mutants screened at low temperature had
eifirarsuffereddelefionofallfirreecopiesofthefisnsgeneor,aswas
demonsfi'aHyfi-uervifirhrre38-lc,werecontaminanh,wlulefiwsesae«redat
normal growth tempera-m only exhibited fire mutant W with regard
to the transgene, and hence do not affect fire signal transduction pafirway of

interest

Sinoethe mutantphenotypeoffirelineswhichexpras GUS atnormalgrowfir
hmperatures cosegregstes wifir fire locus firought to contain fire two linked
fiamgenes,firemutafionmustbeeifirerinoneoffireimerborinalinkedgene
that affects fire expression of GUS, but not of COR15 message. The former seems
more likely, especially in line 6-8#4w, which shows fire strongest phenotype and

is still responsive to low temperature, drought, ABA and to suppression by

58
submsrgeuee in water. Lines raw and 4-11#3w are still inducible by low

Wflahnotshownxbutfireresulbfromfirewateraoaktestwereless
conchssivebecausefireGUSexpressionoffiresehnesissoweakatfirenormsl
growth temperature. This alteration in expression wifirout loss of inducibility
suggest firatthe mutation may be in fire introduced COR15a promoter. Cloning
firistransgenefromanyofthefirreelinesknowntocarryindependent
mutations, but particululy from line “kw, and determining what fire
mutation is might give insight into fire function of cis-acting promoter elements.
However,ithascurrenfiybeendeterminedfiratdirectexpefimentsfusing _
promoterelemenhtoreportergenesareamoreefficientwaytoshrdyds—acfing
slements,psrticularlybecausefirereisnowayofknowingwhichoffiretwo
insartsfioughttoresideatfirismutantlocusmightberesponsibleforfire
phanotypeandbecauseitissfiflpossiblefiratfiremutafioniswififinfirecoding

regionofGUSorisonlylinkedtofirelocus.

There are a number of measures first could be taken in setting up a new attempt
to find plants mutant in fire signal transduction pafirway responsible for
inducingCORlSainordertoimprovefirechancesoffindingfiredesired
mutants. To reduce fire amount of contamination, flowering plants should be
kept in a reparate chamber from firose newly planted and, where parade, firere
should be an easily observable phenotype associated wifir fire transgenic line

used for mutagenesis which is unrelated to expression of COR15a or fire

59
transgenesofiratcontaminantscanberecogruredanddiscardedwifira

minimum amount of effort.

lnordsrtomakemutationsinfireh-ansgenemorereadilyidenfifiable, fire
origindreporbrcoasfiuctcouldbemdedgnedbhwludeasecondnporteralso
fussd to fire COR15a promoter. This way, anofirer relatively rapid test could be
done on each putative mutant to see if fire second reporter also demonstrated fire
mutant phenotype rather firsn waiting for a generation or two to get enough
materialforRNAisolation. Sinoemutations firstsppeartobeinfirelocus
containingfirelinked insertswerefound instleastoneoutof2,000 plants
screened (counting fire five mutants isolated from mutant pool 6 as a single
mutation event), it might also be wise to select a transgenic line containing only
onepesdictableinsert, firusreducingfirecomplexityoffiresysbm. Thepresenee
ofthesecondreporbrwould obviatefirensedfora secondinserttoweed out

null (is-mutations.

lnlightoffireobserved vsrisbilityofGUS expression, it mightalsobewiseto
usefirehistochemkslstainonseveralpot—gmwnphnhoffirefi'amgeniclines
proposedforuseinmutagenesistoseeifsomelinesshowmorevariabilityfrom

leaftoleaffiranothers.

60
Experimentsinprogresstoteaseapartfirefuncfionsofvariousa's-acting
elements may make it possible to design a construct first is only inducible by low
temperatrue, musmakingitmorelikely firstthescreenorselectionwillyield
mutations related to low temperature induction and not to water relations. The
prom for COR15b may in fact only be responsive to low temperature (see
Chapter 3).

Finally,fireuseofaselectablerafirerfiranascreenablereporérshould make
idenfificationofmutantseasierandshouldmskefiretestingofamuchlarger
numberofplantspossible. Resistancetosnherbicideorantifioticnuchas
BASTA (Douche: et al., 1993), ksnsmycin (V alvekens et al., 1988) or hygromycin
(Susek et al., 1993) could be used to select for plants first allow expression from
fire COR15a promoter under warm conditions. and genes encoding proteins first
convert a nontoxic substrate to a toxic substsrrce, such as firose for
indoleacetsmide hydrolase (Klee et al., 1987), nitrate reductase (Nussaume et al.,
1991), or cytosine desminsse (Perera star, 1993) could be used to select plants
first do not have expression from ,fire COR15a promoter at low temperature. For
this low temperature screen, it would be advantageous to bat fire induction of
fire non-mutagenised reporter construct at varying temperatures, since fire plants
grow very slowly at 2°C and sensitivity to toxic compounds is ofbn seen best
aftergrowfirhasoccurredinfireirpresence. Ithasbeenshown(seeChapter4)

first COR15a expression is increased above first in warm-grown controls even at

Fl

PK

(Or

1110

61

19C, and firstitsexpressioni'satorrrearits maximum “8°C, so itshould be
possibleto find a temperature atwhichexpressionis strong, and growfirisnot
prohibitively slow.

It is possible first no mutants were found because fire expression of COR15a is
necessaryduringsomestageoffirelifecycleofArsbidapsis. Thisviewis
consistentwifirfire finding firststaining was found inanfirers ofwarm-grown
plants containing fire same construct as 1,6 (Baker et al., 1994). It was hoped first
fire use of EMS, rather firan a mutagen first caused larger deletions, might allow
some leaky mutations first would show alteration of expression without being
lethal. Unfortunately, fire variability of fire screen led to focus on plants wifir
strong phenotypes because fire weaker ones were simply too hard to follow.
Thus, steps taken to reduce fire variability of fire screen or selection take on

added importance.

Furfirer work could be directed toward rescreening fire 39 lines of putative down
in fire cold mutants and 9 putative up in fire warm mutants first remain to be
muted at fire M3 generation. This should, however, be secondary to fire
preparation of new reporter constructs for use in starting over because a new
construct enabling a selection and fire elimination of cis mutations should allow a

more dficient identification of relevant mutations.

 

 

 

..._-.._— _- _.

62
Materials and Methods

Library saeeuirrg

The A. Minis (ecotype Columbia) genomic library used to screen for COR15a
genomic clones was a giftfrom Harry Klee(UniversityofFlorida). Itwas
preparedbyligatingaSauSAIpartisldigestofgenomic DNAintotheBmHI
site of lambda EMBL3 (Ssmbrook et al., 1989). The library was screened wifir
pill-[67 (Hsjels et al., 1990),. a cDNA representing COR15a, using standard
methods (Sambrook et al., 1%9). One of fire recombinant phage first hybridized
with pill-167, L67g, was characterized furfirer. The Arabidapsr's insert DNA was
cut out as a Sal I fragment (Sal 1 sites flank the insert in fire multiple cloning site
of XEMBLB) and ligated into fire Sal I site of pUC19 (Y anisch-Perron et al., 1985)
to make plasmid p67gk. A 1.8 kb 331 II fragment, which contains fire coding
region of COR15a and approximately 1 kb of 3’ sequence, was subclorred into fire

Bax HI site of pBluescript (SK-) to make pBlue67.

Analysis of kanamycin resistance

Seeds were surface sterilized wifir 30% bleach/ 5% Triton X-100, washed wifir
sterile water five times and plated on GM media wifir Kanamycin. GM media
was made up of 0.05% 2—[N-morpholinolethane sulfonic acid (MES), 0.8%

phytagar and 1 bag/l of GibcoBRL Gamborg’ s B—S Medium wifir sucrose. The

63
pH was adiusbd toS.7wifir1 M KOH and 50mg/1Kanamycinwss added after

autoclaving. SeadswereplatedinssterilemsnnerbydrswingIOOrfioffire
seed/water mixture into a 1 ml pipetmsn tip, removingfiretip from fire
pipehanandgenfiytouchingfirefiptofiresurfsceoffireagartodepositseeda
Seedswerefirenseparsbdusingfiresametip. nreplateswcrewrappedwith
twopiscesofParafilm’andplantsweregrownatslightinbmityofaboutIOO
uErhflaK—hourphobperiodandtempersturesof23°Ct027°Cforaweekor
until resistant plants (green) were differentiable from sensitive plants (white or
yellow). '

Histoclremical analysis of GUS expression A

Platebgrown seedlings were tested for cold-inducible expression of GUS using
lristochemical staining (Jefferson, 1987). Leaves of plants first had remained at
growth conditions or first had been given an overnight treatment at 2°C were
excised and incubated at 37°C overnight in a solution containing 100 mM
sodium phosphate pH 7.0, 10 mM EDTA, 0.1% (v/v) Triton X-100, 2 mM 5-
bromo-4—chloro-3-indolyl-B-D-glucopyranoside (X-gluc), 1 mM KsFe(CN)s and
1 mM KsFe(CN)s. Chlorophyll was removed by soaking fire plant material

overnight in 50% (v/v) EtOH. The tissue was firen stored in 70% (v/v) EtOH.

Mutageneslsaudscreeuing

Mutagenesis wasdoneessentisllyasdescribed (HsughnandSomer-ville, 1986).
For each round of mutagenesis, approximately 25,000 (500 mg) seeds of
ArabidapsisWaLaleyn)fisnsgenichne1,6wereplacedin100mlsofa
0.3% solutionofEMS (efiryl mefiranesulfonate),leftinsfumehoodat room
temperature for 8 hours and swirled occasionally. Seeds were firen washed wifir
mesmosiswster 10 to 15 times, suspended in0.1% phytagarand planted
by pipeting fire seed/agar suspension onto 21 flats ofsoil(about1 seed per
squarecm). TheMIplanboffirefirstroundofmutagenesisweregrownunder
fluorescent lights on a shelving unit at ambient conditions, while firose of fire
sscondroundweregrowninagreenhouse. lnbofirplaoes,firripsarrdaphids
inkstedtheplantsJimitingseedproduction. Saferuinsecticidalsospwasused
onfiuphnhonfireslrelvingunitandDycsrbandOrfirenekppliedbyhcensed
persorurel)wereusedonfireplantsinfiregreenhouse. TheM2seedfiratwas
produced was pooled by flat and planted out in pots first were divided into
quadrants. Seedlings (approximately two weeks old) were screened for aberrant
GUSexpression. Plantsscreenedinfirewarmwereleftinfirechamberswhere
theyweregrowmwhilefirosescreenedinfirecoldweremovedtosrCcold
room overnight initially, but for four days when it was recognized first longer
coldtreatmmtreducedfirevsriabilityoffireMUGscreen. Ascalpelwasusedto
cutaleafpiecefromeachplantandforcepswereusedtoplsoeeachpieoein

250 ill of a solution containing 2 mM MUG (mefirylumbelliferyl glucuronide),

65
50 mM sodium phosphors pH 7.0 and 0.05% Triton X-100 in a microtiter plate
well. These were firen incubated at 37°C overnight fire first few times, but
subasquerrfithoursformoresensitivedemtion, andviewed onaUV
lightbox. The microtihr plates were divided into quadrants like the pots, and
when a positive well was detected, fire plants in fire appropriate quadrant were
flaggadwifirnumba'edtoofirpickssndanofirerleafpiecewasuted. Onces
putative mutant or mutants had been had been identified, all fire ofirer plants in
first pot were removed and fire putative mutant(s) was allowed to set seed. The
identification system used lists fire number of fire pool, followed by a dash and
the number of times seed had been planted from the pool, the quadrant in which
fire putative mutant had been found, fire number of fire putative mutant and
fhrsfly“w"forthewarmscreenedplsnhor“c"forfimsescreenedinfirecold.
As more work was done wifir a given mutant line, fire designation was

sometimesshortenedforeaseofuse.

Plant powtlr and stress treatment

Arabidopsis tlraliana a.) Heyn. ecotype RLD or Columbia was grown in
controlled environment chambers at 22°C wifir a 24 hour photoperiod (about 120
umol rrr'2 s4) as previously described (Gilmour et al., 1988). Humidity was not
controlled. After about firree weeks of growth, plants were eifirer harvested or
subjechd to one of several stress treatments. Plants were cold-acclimated by

placing pots in a 2°C cold room under constant cool-white fluorescent light (45-

66
55umolm4s4)forfourdays. Droughtsmsswasadmirusteredbyexcisingfire
shookfromfirepohplacingfiremonlScmWhatmanNoinlterpaperin
150mmx15mmplasticpefiiplateswifirthelidsopenfor45minums. Tlrelids
wesefirenplacedonfireplatesandplsntswereincubatedforfour tosixhours
under fluorescent lights (about 50 umol marl). Water loss was calculated by
dividingfireweightoffiredrought—sfiesudphnbbyfireirhufialfreshweight
andmultiplyingbyloo. Plantslostfrom18%to23%offireirinitialweight. The
ABAtreaMerrtconsistedofplacingexcisedplantsintoa150mmx15mmpetri
plate wifir 75 ml reverse osmosis water supplemented wifir 100 uM ABA (mixed
isomers, Sigma) and one drop (10-20 ul) Tween-20 (polyoxyefirylene-sorbitan
monolaurste 20, Sigma), covering firem wifir a piece of 15 cm thtman No. 1
filtarpapertokeepfiremsubmerged,placingthelidonfireplateandincubafing
them for 4 to 6 hours under fluorescent lights (about 50 umol m'zs'l). Water-

sosked plants underwent firis same treatment, except first fire ABA was omitted.

RNA extraction and fractionation

Total RNA was extracted essentially as described (Gilmour et al., 1%8) wifir a
few modifications. Frozen pulverized tissue was extracted wifir equal volumes
of phenol/chloroform/isopropyl alcohol (25/ 24/ 1, v/ v/ v) and extraction buffer
(1% (w/v) triisopropylnaphfirelene sulfonic acid, 6% (w/v) p—aminosslicylic

acid, 100 uM Tris-HCl pH 7.6, 50 mM‘ EGTA, 125 mM NaCl, 1%SDS, 10 mM

67
D'l'l')onice. 'I'lriswasfurfirerhomogenizedinfiretubeusinga'l‘ekmar’
Tissumizer, centrifuged (10,000 rpm in an SA600 rotor), and fire supernatant
med again with phenol/chloroform/isopropyl alcohol. This second set of
hrbesandallsubsequenthrbesandsolufiomexceptefiunolweremsdefreeof
RNase by fieatment wifir DEPC (Sambrook et al., 1989). Nucleic acids were
procipitabd with cold 95% efiranol, resuspended in 1 ml double-distilled water,
fiansferredtoamicrofugehrbeandprecipitstedoniceforanhourwifirl/l
volume 10 M Lic1. Thepelletswere precipitated againwifir95% EtOHand
resusperrdedin200ulwater. TheOmeasmeasuredtoestimstefire

concentration of RNA.

For fractionation, RNA (5 to 40 ug) Was dried down, resuspended in
formaldehyde loading buffer containing EtBr (about 100 ng/ ml), incubated at
68°C for 10-15 minutes and fractionated on denaturing formaldehyde agarose
gels (Sambrook et al., 1989). The RNA was firen transferred (Sambrook et al.,
1W9) to Magrrs NT membranes (MSl) using 10x SSPE made according to
instructions provided by Schleiclrer and Schuell. This recipe for SSPE is slighfiy
differentfrom firatdescribed bySambrooketal. (1%9), and iswhstwasused as
SSPE in all experiments. RNA was. UV cross-linked (StratalinkenStrstagene) to

firefilters.

Northern hybridization

For standard northern hybridization, blots were prehybridized, hybridized and
washsdinaRobbinsScientificModethybridizationovenaccordingto
standard-smmusubeletsl,1ss7)withsoaieasoditicstione Blotswere
incubatedwifirprehybridizationsolution, whichismadeupof50% formsmide,
5x $PE, 50 mM potassium phosphate pH 8, 5x Denlrardt’s solution (1x is

0.02% w/v ficoll, 0.02% w/v polyvinylpyrrolidone, 0.02% w/v bovine serum
albumin), 0.5% SDS, 100 ug/ ml sheared, denatured fish sperm DNA. at 42°C for
firree hours to overnight and in hybridization solution (50% formamide, 5x sees,
50 mM potassium phosphate pH 6.5, 1x Denhsrdt’ s solution, 0.5% SDS, 5%
dextran sulfate, 100 ug/ ml sheared denatured fish sperm DNA) at 42°C '
overnight. Blot were rinsed and firen given two 30-minute washes at room
temperature in 2x SSPE/0.5% SDS, firen washed two to fiuee times, for 15

minuhs each time, wifir 0.1x SSPE/0.5% SDS at 50°C.

Probes were made from gel-isolated fragments labeled wifir 32P by random
priming (Feinberg and Vogelstein, 1983). An Eco RI fragment from pHH67 was
used to visualize COR15 message and a Sac l/Hind III fragment from p81101.3

(Jefferson, 1987) served as a GUS probe.

69
PlantgenomicDNAextnctionandSouthernanalysis

Genomic DNA extraction for the analysis of the transgenes in 1,6 and for
analysis of putative mutants was done as described by Rogers and Bendich
(1m)exeept&atfisemixhsnofplantflssueandexh‘acflonbufferwuimubated
atGO'Cfor30minuhsbeforechloroformexh'actionandtheCI'AB '
(hexadscyltrimethylammonimn bromide) precipitation was left in the

refrigerator until flocculant precipitate was seen.

Total MW DNA or plasmid DNA was digested with various restriction
enzymes and fractionated by agarose gel electrophoresis (Sambrook et al., 1%9)
using TAB (40 mM Tris-acetate, 1 mM EDTA). The DNA was then transferred
(Sambrook et al., 1W9) to Magna NT membranes (M81) using 103: $PE. The

DNA was UV cross-linked (Stratalinker, Stratagene) to thefilm.

Southern blots were prehybridized and hybridized in a Robbins Scientific Model
400 hybridization oven using the nonfat dry milk method of Johnson et al.,
(1984) and washed using standard conditions (Sambrook et al., 1989), except that
all washes were done with 0.1% SDS/0.1x SSC. Gel-isolated fragments labeled
witthbyrandom priming (FeinbergandVogelstein, 1%3) wereusedas
probes. A Sac I/Hind III fragment of p31101.3 was used to detect GUS inserts, a

Bass HI/PstlfragmentoprIB710 (RothsteinetaL, 1987) wasused asa probe

70
fortheCaMVSSSpromobrandemRIfragmentopr-fllflwasusedto

visualize CORISa and COR15b sequences.

'I'hebinomialdish'ibution Nl/nilnz! (p)"i (W): wasused todeterminehow
many samplesofputative mutantlined-llflw‘shouldbetested to givea95%
probability that there is no segregation from the locus thought to contain the
linksdinserts. Nisthetotalnumberofsamples,niandnz arethenumberof
times co-segregation and the lack of co-segregation are seen, respectively,
pistheproporflonoffltefimethelackofabandisexpectedfromasegregafing
parant(1/4)andqistheproportionoffiietimeabandshouldbeseenfroma
segregatingparentCl/l). 'l'hepercentchanceofnotseeingsegregationwhen
these really was some is x. The exclamation mark denotes the factorial of the
numberpreceedingit. Theequationwassolvedfornzafternihadbeensetto
zeroandxhadbeensetto5%orless. Thisyieldedannzofeleven. Sinceniis

zero, thatmeansNis alsoelevenandelevensamples mustbehsbd.
Literature Cited

AusubelFM, BrentR, KingstonRE, MooreDD,Seidman]G,SmithIA,Struth
(1W7) Current Protocols in Plant Molecular Biology. Greene Publishing
Associates and Wiley Interscience, New York, Chichester, Brisbane, Toronto,

Singapore

71

Baker SS, Wilhelm KS, Thomashow MF (1994) The 5’-region of
dualism oor15a has sis-acting elements that confer cold-, drought- and ABA-
regulated gene expression. Plant Mol. Biol. 24: 701-713

BouchezD,CamilleriC,CaboucheM(1993)AbinaryvectorbasedonBasta
resistameforiuplaufs transformationofArsbidopsis thaliana. C. R. Acad. Sci.
316:1188-1193

Bowling SA, Guo A, Cao H, GordOn AS, Klessing DF, Dong X (1994) A mutation
in Arsbidopsis that leads to constitutive expression of sysbmic acquired
resistance. The Plant Cell 6:1845—1857

Brusslan JA, Karlin-Neumann GA, Huang 1., Tobin E (1993) An Arabidopsis
mutant with a reduced level of cab140 RNA is a result of cosuppression. The
Plant Cell 5: 667-677

Cao H, Bowling SA, Gordon AS, Dong X (1994) Characbrization of an
Arabidapsis mutant that is nonresponsive to inducers of systemic acquired
resistance. The Plant Cell 6:1583-1592

Feinberg AP, Vogelstein B (1%3) A technique for radiolabeling restriction
endonuclease fragments to high specific activity. Anal. Biochem. 132:6-13

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

Haiela RK, Horvath DP, Gilmour SJ, Thomashow MF (1990) Molecular cloning
and expression of cor (Cold-regulated) genes in Arabidopsis Minna. Plant Pysiol.
93:1246—1252

Haughn GW, Somerville CR (1986) Sulfonurea-resistant mutants of Arabidopsis
Midas. Mol. Gen. Genet. 204:430-434

Jefferson RA (1987) Assaying chimeric genes in plants: the GUS gene fusion
system. Plant Mol. Biol Rep. 5:387-405

Jefferson RA, Kavanagh TA,‘Bevan MW (1%?) GUS fusions: B—glucuronidase as
a sensitive and versatile gene fusion marker in higher plants. The EMBO Journal
63901-3907

KleeHJ, Horsch RB, HincheeMA, HeinMB, HoffmanNL(1987)Theeffectsof
overproduction of two Agrohacterium tumflcieus T-DNA auxin biosynthetic gene
products in transgenic petunia plants. Genes Dev. 1:86—96

72

levittJ(1%0)Responsesofplantstoenvironmentalstress. Chillingfreezing,
and high uperature stresses, second edition. Academic Press, New- York

Moffat B, Somerville C (1988) Positive selection for male-sterile mutants of
Arsbidopsis lacking adenine phosphoribosyl transferase activity. Plant Physiol.
86: 1150-1154

Negrutiul,BeeftinkF,JacobsM(1975)ArsbidopsisWasamodelsystemin
somaticcellgenetics. PlantSci.Lett.5:293-304

Nussaume L, Vincentz M, Caboche M (1991) Constitutive nitrate reductase: a
dominant conditional marker for plant genetics. The Plant Iourn. 1:267-274

Perera RJ, Linard CG, Signer ER (1993) Cytosine deaminase as a negative
selectable marker for Arabidopsis. Plant Mol. Biol. 23:793-799

Rogers 80, Bendich A] (1988) Extraction of DNA from plant tissues. In so
Gelvin, RA Schilperoort, eds., Plant Molecular Biology Manual. Kluwer
Academic Publishers, Dordrecht, Boston, London, pp A6/1-11 '

Sambrook J, Fritsch EF, Maniatis T (1%9) Molecular Cloning: A laboratory
manual, second edition. Cold Spring Harbor Laboratory, Cold Spring Harbor,
New York

Susek RE, Ausubel FM, Chery J (1998) Signal transduction mutants of Arabidopsis
uncouple nuclear CAB and RBCS gene expression from chloroplast
development. Cell 74: 787-799

Takahashi T, Naito S, Komeda Y (1992) The Arabidopsis HSP18. 2 promoter/ GUS
gene fusion in transgenic Arsbidopsis plants: a powerful tool for the isolation of
regulatory mutants of the heat-shock response. The Plant Journal 2:751-761

Thomashow MF (1993) Genes induced during cold acclimation in higher plants.
In, PL Steponkus, ed., Advances in low temperature biology, vol. 2. JAI Press
Ltd., London, pp 183-210

Thomashow MF (1994) Arabidopsis as a model for studying mechanisms of plant
cold tolerance. In, EM Meyerowitz, CR Somerville, eds., Arabidopsis. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, New York, pp 807-834

Valvekens D, Van Montagu M, Van Lijsebettens M (1988) Wm
tunefnciens-mediated transformation of Arabidopsis thalians root explants by
using kanamycin selection. PNAS 855536-5540

73
Yanisch-Perron c, Vieira I, Messing I (1985) Improved M13 phage cloning

vectors and host strains: nucleotide sequences of the M13mp18 and pUC19
vectm's. Gene 33:108-119 ‘

Chapter 3:

Analysis of a genomic clone containing COR15a and. COR15b
Summary

hordectoiacnitateaieintecpcetauonototndieoowmegnhdonandpmtein
analysis of the cold-regulated gene COR15, experiments were carried out to
determinewhetheritwas a memberofa small gene family. Initial genomic
SouthernsofArsbidopsisludshownflutCORlSwaseitherasinglegeneora
member of a small gene family (Hajela RK, Horvath DP, Gilmour SJ,
Thomashow MF, 1990, Plant Physiol. 93:1246-1252) and further genomic
Southern analysis indicated that there was a second gene which hybridized to a
cDNA corresponding to COR15. A genomic clone containing COR15 was found
to also contain this second member of the COR15 gene family, which was then
named COR15b. COR15 was renamed COR15a. The two genes are in the same
relative 5’ to 3’ orientation and their predicted coding regions are 82% identical
atthenucleicacidsequencelevel. COR15aandtheproteinpredictedtobe

encoded by COR15b were compared to each other and a model for their

74

75
secondary and tertiary structure is proposed. COR15b is inducible by low

“pasture and possibly by ABA, but not by drought. Gene fusion experiments

indicah that this lack of drought inducibility is most likely promoter-based.

Introduction

Arsbidopsisflrdiushablehmcreaumfmedngtolerameatlow,mnfreezing

temperatures in a process known as cold acclimation (Gilmour et al., 1988).
During this process, numerous genes are induced, including a gene known at the
beginning of this work as COR15 (Hajela et al., 1990). COR15 mRNA is
inducible by drought stress and by the exogenous application of abscisic acid as
well as by low temperature, but not by heat shock. In vitro translation
experiments showed that COR15 encodes a 15 kDa polypeptide, COR15, which
is cleaved to 9.4 kDa upon import into chloroplasts. The mature form of COR15,
CORISm, is acidic, with a pI of 4.6, and is very hydrophilic. COR15m remains
soluble upon boiling, a property thought to be a consequence of this high
hydrophilicity (Lin and Thomashow, 1992).

A genomic clone containing COR15 was isolated in preparation for gene
regulation studies (see Chapter 2). Approximately 3 kb of the COR15 promoter

was identified in this clone and subsequent promoter deletion analysis showed

76
that the cold- drought- and ABA- inducibility of COR15 is promoter-based

(BakeretaL, 1994).

In the analysis of COR15, the question of whether it was a single gene or a
member of a small gene family was raised. COR6.6 and COR78 had been found
tebemembersofsmallgenefamilies (KurkelaandBorg-Frank. 1992and
Yamaguchi-Shinozaki and Shinozaki, 1993) and an extra band found on
westerns when examining the tramlated product of COR15 might be explained
by tlm presence of two related genes. Initial analysis using two restriction
enzymes gave a simple pattern, suggesting that COR15 was either a single gene
or was a member of a small gene family (Hajela et al., 1990). The restriction map
of the genomic clone isolated allowed for a more efficient detailed analysis, so
moregenomicSouthermwereper-formed, andevidencewas found forthe

presenceofasecondgene.

The genomic clone already isolated containing COR15 was found to also contain
this second gene, so COR15 was renamed COR15a and the second gene was
designated COR15b. This chapter describes the sequencing of 4.2 kb of this
clone, analysis of COR15b sequence, a knowledge-based model of the protein it

encodes, and analysis of COR15b lnducibility.

Results

Identification of COR15b and sequence comparison with the coding region of
CORl5a

Southern analysis of Arobidopsis (ecotypes Columbia and RLD) genomic DNA
wasperformedunderconditionsofhighandlowstflngermytolookfortlm
presence of a second gene. A cDNA clone of COR15, pHH67 (Hajela, 1990), was
used as a probe. Four restriction enzymes (Sail, HindIlI, N001 aid EcoRI) were
chosentogiveapredktablebandingpattermbasedonaresh'icfionmapofflm
gmmicclone. Indeed, in bothecotypes faintbands were seenin addition to the
expectedbandsontlmlowstfingencyblotforfllreeoftlmfourenzymesused
(data not shown). The absence of extra bands with the fourth enzyme, EcoRI,
which should give the largest fragment containing COR15, suggested that both
genssmightbecontained withinthisfragmentand thus bothbepresentonthe
genomic clone already isolated (see Chapter 2). Closer examination of
preliminary sequence about one kilobase downstream of COR15 did reveal
identity with the promoter of COR15. This second germ was designated

COR15b, and the original gene renamed COR15a.

Further sequence analysis was carried out. The sequence of approximately 4.2
kilobases, beginning about one kilobase upstream of the start of transcription for
COR15a and extending 448. bases past the stop codon of COR15b is shown

(Figure 3.1). The region of the genomic clone that corresponded to cDNAs of

agatcttttc
tacaaaatta
agattctcta
tctttaagga
attttaacct
gttagacaat
tgaattctat
atggtttqtg
cggtttattt
qccgacatac
ggaaatgaqa
ctaacatgta
gacctgcttt
aaagaacgaa
ttgqagtgta
9+1 com!»
atctcacttt

tctcacatct

g E g

tttagtatct
aqtatagatt
ttatcaaaga
qtataqctct

ctttgataaa

ttaaactcgt
sagatgagtc
ccacagtcac
qaqtaattat
tgqgagath
tcttcgcatq
taccaaaaac
gcagtgaatt
catctcccqt
cacattactc
acatgctcta
qctgttgagc
aattaqqcat
ttaqaaatat

cqttqaattt
aattattqca
qacttttttt
aatacatcaq
tttttttttt
ggcataaatt
accactqtaa
qtttcaactc
tqtgtaggct
atttgtttct
acqaccctct
agtttttgtt
ttttttcttc
cagaaacth
ataqqtctat

gttcatctaa

taaaqatctc

tgcctcataa
ttagtqqaaa
taaagattca
aaqtttattt

tactaaaata

agatatattq
tgttgtaatt
cgtcacattc
aaqtaacatc
stagccttgt
tccactattt
ttcaaaattt
acaatccqtg
cacaaacata
acatacccaa
tttcatattt
aaaatagtta
tqaaqtqtga
atctatatac

attttaqact
ttttctattc
tttaaagact
aqaqaaaaat
acaattqaca
tatttaaatt
aattcttaaa
atttgggcta
tttggtagat
attacaaatt
taagtagtac
accaaaaatg
ttctttctqa
qtaatqtqca
cqatctataa

aaactcctcc

tctcllflflc.

qtaagqatct
tatgtttqtt
tataccqtac
qatgattctq

aattttaatt

atccatatcc
tctattaatt
tttatgtttt
qccqtctaga
ctcttttqtq
actgtttgqt
99‘99Ctt9t
qaaggaaaaa
tcatctagaa
ccaatgtggg
atcaaaaaaa
tatttgaaag
gtttttgttt
tattagtagt

78

ttttttttaa
atattqaatt
gatctatgat
tattacgaaa
atcttttcac
aatccqtata
cqaqataaga
ttagttttac
ttgqgcttqc
taacaacaaa
tgqaaattga
castttgtat
aaaccacaaa
aaaaatatct
aacqatacta

tttcatttcc
AIIIC!IICI
CAGIOCIDCI
COCAAfllABI
ChflOCCAcqt
taqcaqgcaa
ttqaacttat
attatatatc
ttqactactt

tgaaaatgat

qqtcttacct

tctctcttca
tcacatcaca
qcaaaatatt
gtctttcaac
acgaaccaat
atgacatcga
aatcaacatt
ttatggtccq
gtcggtatgc
acatatctaa
aaaatattag
taacattggt
ttqtttattt
ttaactacag

tqqacttcat
agqagatgtt
caqaattcca
cgattctatt
aacaaaaatc
tatthcctt
ttattttcaq
atttaggctt
aaacccaaat
cgtcaataaa
aaaaaaqaaa
gtagccacaa
tatqattaca
tactcttaat
ttqqagatta

.ICC‘IIIIC
GIBCIHCECI
lCOOCICEOI
COI’OIICII

aaqtctacat
tqtttatggt
tttatqatca
tctatttttc

ttgqatatqt

atathhflll

aatcaqttaa

tqtttaatag
actqaaataa
caetaqacaa
99t994499‘
aaaagaggaa
attggtatga
gatccactta
gcaccaacaa
cacaccacat
taagctcatt
gcattgtaaa
atttataatt
aacattggag
tttgtactta

tttaaatttt
actgtccgtc
attttttttt
acaagtaatg
cacaagaaac
ctatqagaat
catqtaaaaa
qcaaccttgt
taacttgttg
tacacgtgaa
tctaqaaatg
tttcatqgcc
cqtggcctga
acqtqtaatt
gattcttctc

ttctttttlt
QCECICECII
ClflIlIOOOC
COOOCCEIhI

tctttcttct
atactatatt
tatttgacta
taqtttacat

qttttqaaac
ABCTQCAGI!
ABICEICGII
AIC!UCOOCO
III-flflllhh

tttCIIQCIC

tacttacaat
gatatgqtat
attaaataat
tgsgagcata
ctgctttttt
tacattgttg
qccattccaa
attatcacct
aattaaggag
cttggctqqt
tagcgttttt
atagtatasc
tattaggttc
attgaaaaaa

60

120
180
240
300
360
420
480
540
600
660
720
780
840
900

960

1020
1080
1140

1200
1260
1320
1380
1440

1500
1560
1620
1680

1740

1800
1860
1920
1980
2040
2100
2160
2220
2280
2340
2400
2460
2520
2580

tqttaaaagt
acttgtacat
aqtctagtta
tagccaaaaa
ttttcttttg
tcttattcga
gttttqqatt

tgttttaacc
tttcaqctta
ttaatgttga
tgctaacatq
gtaaccatag
taqttaaaga
ctcattggtc

9+1 comsb

ctcatctcac

tctttaaqag

attttctcgt
tacthccat
ataatcaact

aaaathttt

ttcaatagtt
agctgttaat
cattttaqtg
caaaatcttc
ttatgacaac
tatatatcca
tgaatgtaga
aaqctt

tttctccatc
tcctchlfloc
he‘llflfllflc
IOOIOOEIOC
lclhlILOC!
gactaaggct
gatagttatt
ataatqgcct

qtcaataatg

tacaataaga
tactcactaq
actacaaqta
tctcattttg
ccctaataat
tctcctgagt
aqaaatattt

tagctaattg
acgactaata
aagttgcaaa
tgtataacgq
aaatggttsc
taataqcaat
gagagatcta
ttaaaactct
GIIOECTEII
AAABCAGIOI
ICABCOCAAB

CAACIAABOC
atcttcatgt
ttattttggt
tgttgaggat
aatcttaaca
ICGAIBGIGB
IIABCCAAIB

AASCCCICII

tcagtttgtt
cacattttaa
acatagtagt
atcctcaatt
agcctaaaaa
gaaatctcat
tttttgtcaa

7'9

ctaaaaatga
catttttcct
taaaacagaa
ttataaccac
acgtaactag
gcqcaaaaat
taaaacgata

ttcttgtatt
resonances
OOIOISOOII
ascrcorsaa
qutgagtct
gtactcgcaa
taatcccatc

tgagtttcca

tatatatata
CCEEOOOOII
AAACSOCGIC
CCOGIAACAI

IBICCTOItt

gtaatttctt
atacgtatca
atttatgaqt
ataaagaaac
gatcaaaaca
‘99t9t9999
aagttqtaac

ctaaatagac
ttatatatat
atgctaacat
aacttgatgq
tacgaaccaa
atctagcact
ctattggagg
tattttcctc
!!CSCABIOO
OOOECABIOI
ZIIICOCCO!

atattctttt
aaactctcat
ataacagttt
tggtgtagat

tatatatagh

atacacaaag
atctctatcg
gtaaatatcg
ccgacctctt
cgaaaactct
cacacgtgta
ttagattttt
ccaaaaaaca
CAIOOOEEC!
iOOCCOGIAfl
lllhlfllflhc

cttttaacaa
atacattacg
tatacatact
atatgttttt

Ci‘CCCCtCI

tttcaccatt
aatqatttaa
t99‘ttC999
tgtgaaactg
cctaaaacat
ggcaatacat
caaaacaaaa

aaeaaaoczr
srcoeaoaoo
tgtagtcatg

tgaaatgaag
atataaacta
aaaaatctgt
atgactgttt
cascaattgt
actctctttg
agaacattgg

2640
2700
2760
2820
2880
2940
3000
3060
3120
3180
3240

3300
3360
3420
3480

3540

3600

3660

3720

3780

3840
3900
3960
4020
4080
4140
4200
4206

Figure 3.1 Sequence of a 4.2 kb region of genomic DNA containing COR15a and
COR15b. Noncoding sequences, including promoter regions, introns and 3’
nontranslated regions are shown in small letters, while all translated sequences
areinuppercase. Thearrows show thestartoftranscriptionforeachgerm.

80
COR15a was only sequenced on one strand, since it exactly matched the

sequenceoftlmcDNAclones. Thisconfirmed thattheintronfoundinoneoftlm
cDNAclones(pl-IH71.1)istlmonlyormpresentinCOR15aandthattlmreareno
otter differences between the cDNA and genomic versions ofCOR15a. The rest
ofthesequencingwasdoneonbothstrands. Thetwogenesareseparatedby
lessthanonekilobaseofDNAandareintlmsamerelativeS’ to3’ direction
(Figure 3.1). The transcriptional start site of COR15a was determined by primer
extension (Baker et al., 1994), while that of COR15b is predicted based on
comparison to that ofCORlSa and on tlm consensus sequence found at
transcriptional start sites (Joshi, 1%7). The location of the intron of CORISa is
known based on comparison to the sequence of the cDNA clone pLCT10A (Lin
and Thomashow, 1992), while the intron of COR15b is predicted based on
sequence comparison to COR15a and on consensus sequences found at
intron/exon boundaries (Brown, 1%6). The predicted coding regions and

introns are, respectively, 82% and 57% identical at the nucleic acid level.

Analysis of the protein predicted to be encoded by COR15b and comparison
with COR15a

The polypeptide predicted to be encoded by COR15b, designated COR15b, is
77% identical and 82% similar in amino acid sequence to COR15a and both have
four degenerate 13 amino acid repeat (Figure 3.2). Like COR15a, which is
known to be targeted to chloroplast (Lin and Thomashow, 1992), the N-terminal

sequence of COR15b has features typical of chloroplast transit peptides (Gavel

81

COR15b 1 MAMSLSGKVLSGHGSSFHNVGAKQSGVGTVRVGRKSELVVVAQRKKSLI!A*V 52
COR15a 1 ....f ..... T..a....—s ..... sf.a....q.qu...s ...... , .. *5.51

53 KSDGNILDDLNIATKKASDFVTDKTKEALADGBKTKDYIVERTIEANEtAIEI 105
52 .g ................................ a...V...ns.taD.Lgk. 104

106 AKKALDYVTEKGKEAGNKAAEFVEGKABEAKNATKS 141
105 OQOOMOOG ...... a ...... AOOOOQOOOd... 139

Figure 3.2 Alignment of COR15a and COR15b. Identical, biochemically similar
and different amino acids are indicated by dot, capital and lower-case letters,
respectively. The consensus cleavage site sequence for chloroplast transit
peptides (Joshi, 1%7) is double-underlined, with the predicted site of cleavage
being indicated by an asterisk. The four copies of the 13 amino acid repeat are
underlined. One gap, indicated by a dash, was introduced to optimize the
alignment.

82
and von Heijne, 1990). It has a high serine/flueonine content (18%), only one

addic residue (Glu), and a sequence beginning at amino acid 49, IYAV, that
resembles the loosely defined chloroplast signal sequence cleavage site V/ I-X-
A/C-A (Gavel and von Heijne, 1990). While the final valine of COR15b’s
predichdcleavagesimisabaentfromflmconsensueitisfoundinanumberof
fiesoquencesfromwhichtheconsensuswasderived. Finally, tlmreisan
arginine residue eight amino acids before this putative cleavage site, as would be
expected. It is, however, also possible that COR15b may be targeted to
mitochondria, since mitochondrial transit peptides have much in common with
chloroplasttransitpeptides (SollandAlefsen, 1993). Thepredicmd molecular
weighteftlmmahueCORIprroteinifitistargeted and processed, is9.6 kDa,
compared to9.4 kDa forthemature form ofCOR15a (LinandThomashow,

1992). Both are quite acidic, each with a predicted pl of 4.6.

Proteins similar to (20315

The BLAST algorithm tblastn (Altchul et al., 1990; blaatsncbinlntnihgov) was
used to search nonredundant databases, including GenBank (Release 92,
December 15, 1995) and the EMBL data library (Release 45.0), for sequences
similar to the predicted mature (208155 polypeptide (comm). The putative
mature forms of BN115, le9 and BN26 from Bracelet nopus (accession numbers
U14665, 868879 and 868727, respectively; Weretilnyk et al. 1993) and of a Brassics

dam cold responsive mRNA (accession number U16751; Jo H, Lee C, and

83
Sohn U, unpublished data) show amino acid sequence identity of 72.2, 68.9, 66.7

and 70%, respectively, with COR15b. Tlm other germs with tlm highest sequence
identity identified by this search are Group 3 late embryogermsis abundant
(LEA) promins isolamd from birch (accession number 218891; Puupponen-
Pimia et al., 1993), Arsbidopsis (accession number D64140; Yang H, Saitou T,
Harada H and Kamada H, unpublished data), soybean (accession number
U02966; Chow T, Hsing YC, Clmn Z, unpublisl'md data, and accession number
222872; Hsing et al., 1996, In press), carrot (accession number X16131; Franz et
al., 1%9) and wheat (accession number M72395; Curry and Walker-Simmons,
1993). These have 20 to 40% amino acid sequence identity to COR15b, primarily
in repeat regions characteristic of Group 3 LEA’s (Dure, 1993), which align with
the degeneram repeat of COR15a and COR15b. Neitl'mr the structure nor the

functionisknownforanyoftheseproteins

Modelling of COR15 structure

CORISbm also shows similarity to COR15am when structural modelling is done.
The secondary structures of both COR15am and COR15bm are predicted by
PHDsec (Figure 3.3A) to be primarily alpha-helical with short loops at each
terminusand two potentialturnsorunstructured regions separatingtmlices
(RostandSander,1993andRostandSander,1994). Thismethod alignstl‘m
sequence of interest with similar sequences from tl'm SW SPROT database and

uses this alignment to develop a prediction of secondary structure using the

 

 

 

 

 

 

 

 

 

 

It
AA.vxsosurfifibifiihrxflxgbrvroxrxeALADczxrxDYEVEEFTEINETKEEEKREKEB
sec LLL.unansnnanhsunshnnnnahnhunaa ....... Hansnnnaaannhnhaanaann
prob 999279999999988989999999989999744412238999999988899999999999
AAHYVTBKGKEAGRRIKEPVEERKEBRENNER3
sec ssh...unnnnaunaununnnnunn..LLL
prob 997434768999999999986687533889
B —
AA‘vxsosnrtfifitfllirxxhsorvrourxennaocexrKDYIVERFTEInsrXFIEXREXIB
sec LLLLLLL..aunnahnnh...ahhannnunn .............. L....nnnannnnnn
prob 999998643799999877221579999999912241212212232541318999999999
.Aa 7vrsxcxencfikhhsrvscxhsshxuhiks
sec nuanna....nnhuhnan..auauu..LLL
prob 999998134268999997215589822699
(I
——L 1 Alpha helix
O—C Beta sheet.
4: Turn
I)
—[:l——D—L l—l 1 Alpha helix
HI, Beta sheet
-c:r—-—-—-——c:3—4:::::e {:3 Turn

 

 

Ifigumefl3 IheihcfiknulofthetuxxnndarytnructunecflWCCNRISbnm.It.Iflflfllun:
secondary structure prediction based on alignment of COR15bm with proteins
found in a search of 8m SW ISSPROT database. B. PHDsec secondary structure
prediction based on alignment of COR15bm with COR15am and the predicted
mature forms of BN115, BN19 and BN26. For the PHDsec predictions, amino
acklsequence(AukxnuxxuuhnqrshmutunepmedkflhnrGuxaauulpmobabflflynonmfl»
aneshovna.ffiitlpredknsloops(I),hehces(fflhmmdlxflarfluxfls(E;tmrexhnuhedx
with 1 being the lowest and 9 tlm highest probability of accuracy. Dot show
Vvhenethelmnalkfihxnistuuuntahi IinescnnnfiPPflDseclmnxhktknuudenmmeluflkxnl
predicted by knowledge-based modelling using beta-lactamase from Bacillus
lidreuiforslis. C. Secondary structure prediction based on tlm algorithm of
Garnier and Robson. D. Secondary structure prediction based on the algorithm
ofCJunrammlFanmuuh FortheChunuenJhflmuulpredkxknlammlCJunrJhuunan,
psedidflonl,crusthomesinmflcatethelnmnunuuhofflhetnructunalteatuneruuilines
hadkufleitlabsence.

85
profileneuralrmtworkmefimdmostandSander, 1993). Armuralnetworkis

essmtiaflyasystemofarfificialintelligencewhichcanbetairmdinpattem
recognition (Rest and Snider, 1993). The variables which are optimized during
tfiflngammtfimpmgrmsofimcriteriaonwhichfimpredicfionare
based are not available for examination. However, it is known first the amount
ofsequenceconservationinfimaligned sequencesaidsPHDsecinpredicting
secondarystructure. Thispredictionmefilodisrated atanexpecmd72.1%
averagesccuracyforfimfiueestatesoflmlimstrandandloop (RostandSander,
1993). When the same program used proteins known to be highly similar to
COR15bm in its alignment, namely COR15am and fim predicted mature forms of
BN115, BN26 and BN19, the prediction was still highly helical wifil loops at bofil

endabuthadmoreregionsofuncertainstructurea'igureSSB).

PriormodellingusingfimlessaccuratealgorifilmsofGarnierand Robson
(Figure 3.3C) (Garnier et al., 1978) and Chou and Fasman (1978) (Figure 3.3D)
hadpredicted thehighlmlicity, butonlyfimrnethod ofChouand Fasman had
predicted turns separating helices. It was not expected first COR15m would
formasinglelonghelix,sincefilisisunusualforasolubleproteinandhasonly
been observed in calcium-binding and fiberoforming proteins (observation from
protein structural analysis by LA Kuhn and IA Tairmr). When COR15bm n?
modelledasasinglelmlixinlmlicalwheel format(SchifferandEdmundson,
1967), it is notably amphipathic (Figure 3.4). One side of fim lmlix is notably

enriched in hydrophobic amino acids, while fim other is highly hydrophilic.

 

Figure 3.4 COR15bm modeled as an amphipathic helix. Single letter codes of
the amino acids are shown. Hydrophobic residues (A, L, I, V and F) are shown
as filled circles, polar residues (S, T, G, N and Y) are shaded and charged
residues(D, EandK) arewhite.

87
Helicdnetmodelflng(DunnflLl968)furfiterreveahpotenfialionpaus(Figum

3.5)thatmaybeinvolvedinstabilizingfltehelixorhelicesMarquseeutd
“mlm

In the process of predicting the secondary structure of COR15am and COR15bm,
PHDssccomparedthemtotheproteinsintheSWISSPROTdatabaseand
W21 and12proteinsthatshowed moderatetolowlevelsofaminoacid
identity with COR15am and COR15bm, respectively (see Appendix). The
protein with the highest sequence identity to COR15am was the class A beta-
lactamasefromBacilluswreusMadgwickandWaley, 1%7), whichhadZBX
idenb’tyandaflx similarity over89'residues(thiscalculationisindependentof
thatprevidedbyPHDsec). Twoshortgaps, consistingofresidues36-39and
residues 64-70, were introduced into beta-lactamase to optimize the alignment.

COR15bm is 23% identical and 34% similar to B. areas beta-lactamase.

PHDthreader was also used to analyze COR15am and COR15bm. This program
combines the secondary structure and solvent accessibility predicted by PHDsec
and PHDsec (a solvent accessibility prediction program which also uses a neural
network) with amino acid identity of a given protein into a string of information
that utilizes 120 characters (20 amino acids multiplied by three secondary
struchue states and two states of solvent accessibility). This string of

information is then compared with the corresponding string of amino acid

 

 

 

 

 

 

 

Figure 3.5 Helical net model of COR15bm. Single letter codes of the amino
acids are used. Positively charged residues are shown as black circles and
negatively charges residues are shaded. Hydrophobic and polar residues are
white. The proximity to each other of positively and negatively charged
residues on successive turns of the helix may allow intrahelical electrostatic
interactions. This could stabilize the helical conformation.

89
identity, secondary structure and solvent accessibility calculated from each of

450mmdundantproteinsofknownterfiarysfiuctureinordertolookfor
potentialremotehomologsmrobimwifitsimilartertiarysfiucture, butno
significant sequence similarity to the protein of interest) (Root 1995a and Rost
1995b). PHDthreader returned an alignment of 20 potential remote homologs of
known structure for COR15am and for CORISbm (see Appendix). For bofit
COR15am andCORlell, fiseslignedproteinwifitfiiehighestsequenoe
similaritywas apolipophorinfllfrom fiteAfricanlocustlmotsmigrataria
(Breiter et al., 1991). However, the sequence identity of apolipophorin III with
COR15am is only 22% and with COR15bm is 16%, and there is one five-residue
gapinthealignment. Twoofiuerfactorsdiscourageduseofapolipophorinlllin
modelling. Onewasthelowreliabilityoffiiefiueaderbchnique. Themefirodis
only rated at 30% accuracy and fire information on fitreading given at fite world
widewebsite fixatcarries PHDthreaderwarns fitatpredictionswill likely be
wrong. Thesecondwas fixatfitelocationoffiteregionofapolipophorinIIIfitat
issimilartotheentiretyofCORlSamand ofCORISbm mapstofitreeoutoffive
helicesfitatformahelicalbundle. 'Apolipophorinis67and66residueslonger
than COR15am and COR15bm, respectively. The absence of fite ofiter two

helices is likely to affect the way he profiin folds.

90
Of the reported proteins with similarity to COR15am and COR15bm for which
firefiruedmmatomkresdufionsfiucmnkknowmbeta-lactamasewas

fire most similar, making it fire best candidab for knowledge-based modelling.

lthasbeenshownfisatproteinshaving greaterthanorequalto25% identity over
an80residuestrekharelikelytohavegreaterfiran70% tertiarystructure
idutity (Sander and Schneider, 1991), alfirough the gap opening penalty of 3.0
and gap elongation penalty of 0.1 reduce fire effective identifies of COR15am
and COR15bm wifir B. was beta-lactamase to 20.9% and 16.9% respectively.
Furthermore, COR15am and COR15bm are expected to have essentially fire same
sfiucmrebasedonfireir77% aminoacidsequenceidentitywifirnogapsor
insertions. Inastudyoffirerelationshipbetween sequenoeidentityand
structural similarity for 51 proteins (Hilbert et al., 1993), it was found first fire
alphacarbonsofproteinswifir77% sequenceidentitywillsuperimposetowifirin
0.5 A RMSD (root mean squared deviation). In ofirer words, fire structures of
consult and COR15bm should overlap. This means first it should be

reasonable to model COR15bm, as well as COR15am, on beta-lactamase.

Because fire sequence identity between COR15am and beta-lactamase was on fire
border of Sander and Schneider’ s firreshold (1991) of 25% identity over 80 amino
acids, modelling was continued wifir caution wifir fire understanding that fire

resulting structure could be evaluated for reasonableness based fire following

91
four criteria. First, does fie modelled structure correspond wifir fie PHDsec

secondary structure prediction? Second, are tllc majority of hydrophobic
residues buried and hydrophilic residues exposed? Third, are imertions in
COR15bm relative to beta-lactamase'located in loop or turn regions of fie
surface rather firan in fire middle of helices? Finally, does fie model comprise a
compact globular subdomain of beta-lactamase ratler firan extended, poorly

packed segments ofsecondary structure?

Becausefirefirree-dimensionalstructureoffiebeta-lactamasefrom B. musis
not available, and fiere are several structures available for ofier class A beta-
lactanrases, fire structurd conservation among firese class A beta-lactamases was
evaluated. The thee-dimensional structures of Class A beta-lactamases from
Staphylococcus mus (1blc, Glen and I-Ierzberg, 1992) and Bacillus Mum/arms
(4blm, Knox and Moews, 1991), were acquired from Protein Data Bank (Abola et
al., 1987). When firese structures and fie structure of fie class A beta-lactamase
from Escherichia mli (Strynadka et al., 1992) were superimposed using fie protein
modelling program Insightll (version 95.0), fiey were seen to have very similar
structures (Figure 3.6). Some of fie helices are at slighfiy different angles and

some loop regions vary slighfiy, but fire overall structure is fie same.

In quantification of this similarity, fire class A beta-lactamases from S. mus and

E. coli are 24.2% identical, but have an RMSD of fie alpha carbons of only 1.6 A.

Fig]
Qt!

inp‘

 

Figure 3.6 Superimposition of firms: Class A beta lactamases. The structures of
Class A beta lactamases isolated from Staphylococcus aureus (yellow ribbon),
Escherichia coli (blue ribbon) and Bacillus lidlenifbmlis (pink ribbon) were
superimposed using Insight II (version 95.0).

93
Thosciroms.musands.lidmiifomisarsso.7x identicalwifiranRMSDof

1.4A,andfirosefrom£.coliandB.lidlemWare28.9%identicalwifiran
RMSDofOJA. SanderandSchneider(1991)useanRMSDof2.5Aasfireir
upperlimitforacceptablesfiuchudsimflarity,sofiesenumbersarewenwifiun
this limit. Thus, while ofirer class A beh-lactamases may have slight differences
infieanglesoflelkuorloopnfieyarehighlylikelytohavethesame
secondarysfiuchlresandterfiaryfoldasfiebeta-lactamasefromB.m This
high level of structural similarity among beta-lactamases, given the moderate
sequence identity,alsosupportsfieidea thatfiestructureoffieCORISm
proteins may be quite similar to first of beta-lactamase, despite fie moderately

low sequence identity.

QassAbeta-lactamasesofknownstructurewerefirenevaluated forfireir
sequence similarity with COR15am and COR15bm. The beta-lactamase from
Bacillus lidlenifilrm's (Knox and Moews, 1991) had fie highest identity wifir bofir
COR15m proteins, having 15% identity and 29% similarity to COR15am and
16% identity and 28% similarity to COR15bm, so it was chosen for use in

modelling COR15bm.

Using InsightII, fiesidechainsofB. lidlenifwmiswhichvariedinfirealignment
wifir COR15bm were swapped for fire corresponding COR15bm side chains. The

rest of the beta-lactamase polypeptide not analogous wifir COR15bm was fien

94
deleted from fire COR15bm model. Tle resulting protein is comprised of a

firreehelixbrmdlewhichhasashetchofrelafivelyunsfimhiredamhroadds
(Figure3.7). Theaminoterminalsevenaminoacidsoffiemodelhavebeen
omitted,sinoefieyformalehxhrbeh—hctamase,butamsfionglypredictedto
bealoopinsecondarystructureanalysesofCOR15amandCOR15bm. These
residuesueinaregionofmidenfitybetweenfietwoproteimand,iffieywere
modelled as an amphipathic lelix, its hydrophobic residues would be exposed
to the solution, which would be fiermodynamically unlikely. It is more likely to
bealoopregionfiratpacksits twohydrophobicresiduesagainstfire

hydrophobic regions of fie protein.

The model was fien evaluated according to fire criteria previously proposed.
First, fieplaoementoflelicesinfirismodelissimilar, alfiroughnotidentical, to
fire PHDsec prediction which utilized a provided alignment of COR15am and
the predicted mature forms of BN115, BN19 and BN26 (see bars over PHDsec
predictions in Figure 3.3). The high sequence identity of fie proteins used in
fiuspndicfionshouldhrcreaseibaccuracyoverfiratoffiepredicfionbasedon
proteins of lower identity found in the SW ISSPROT database. Second,
hydrophobic residues are generally buried in 'fie structure, while hydrophilic
residues are surface-exposed. Third, residues first are inserhd in COR15bm
relative to beta-lactamase occur in surface-exposed loops (see Figure 3.7) and are

musnotlikelytohaveamajoreffectonfireoverallstructure. Basedon

 

Figure 3.7 Model of the proposed structure of COR15bm. Shown is the ribbon
of residues B78-3151 of the beta lactamase from B. licheni/brmis. The sidechains
of file B. lidleniformis protein have been replaced by file sidechains from
COR15bm, starting wifii COR15bm residue 8 (Ile) and ending wifii residue 90
(Ser). Two yellow arrows at the center of the picture identify sites where
residues of COR15bm were looped out of fiie alignment, and thus out of file
model. The arrow on fiie left marks file point where residues 36-39 of COR15bm
were omitted, and file arrow on the right shows where residues 64-70 were left
out. Sidechains are colored according to fiieir properties. Acidic residues ('D
and E) are red, basic residues (K) are blue, polar residues (S, T, G, N and Y) are
magenta and hydrophobic residues (A, L, I, V and F) are green.

96
comparisonswifiifiseeffecbofinserfionsinofiierproteim,fiwseimerfions

wouldhkelyextendfiwloophngfiuorpmsiblyextendfiuhefixlengfiisaieinz
et al., 1993). Finally, it can be seen first five modelled region corresponds to a

single, compactsubdomain offiie beta-lactamase protein (Figure 3.8).

It is likely, however, fiiat file sidechains are somewhat more tightly packed in
fiie actual protein, since fire modelled COR15bm has a surface loop formed by
firemainchain in which fiie sidechains are less tightly paCked in the modelled
COR15bmthaninbeta-lactamase. Thisresultsintherebeinga tunnelthrough
fiaepefipheryoffiiemodelledCORISbm,whilefi\ereisnohmnelfi\roughfite

corresponding loop of beta-lactamase.

Regulation of COR15a and COR15b mRNA

The finding fiiat fitere are two very similar members of fiie COR15 family led to
fire question of why file duplication might be beneficial. Are fite two genes
redundant or might one be present at a different time or place than fire ofiter? In
order to examine fiie inducibility of COR15a and COR15b individually,
oligonucleotides complementary to a divergent region of file 3’ end of COR15a
and COR15b were synfiiesized. Once fite hybridization conditions at which
these probes were specific had been empirically determined, fiiey were used on
norfiiern blots prepared from plants exposed to low temperature, drought or

ABA (Figure 3.9A). Bofii genes were induced by low temperature and treatment

 

Figure 3.8 Region of beta lactamase wifii similarity to COR15am and COR15bm.
Shown is a ribbon depiction of beta lactamase from B. lichenijbrmis with fiie .
region flat has amino acid sequence similarity With COR15am and COR15bm
colored yellow.

98

 

BDDDDDCW

COR15a

   

COR15b .

Figure 3.9 Comparison of the regulation of COR15a and COR15b. A. RNA from
warm- (\N), cold- (C), drought- (D) and ABA- (A) treated plants was fractionated
on duplicate gels and blotted. Each blot was probed with a gene specific probe
for COR15a or COR15b. B. Five drought-treated (D) RNA samples with cold (C)
and warm (W) controls. Again, duplicate blots were prepared and each was
hybridized with a probe specific for COR15a or COR15b.

- 99
with ABA. However, only COR15a was notably responsive to drought. This

result was confirmed with five more drought samples (Figure 3.93). Induction
was observed when they were probed with COR15a, but not wifit COR15b.
There were low levels of COR15b transcript detectable in one of file drought
samples, butfiiesewerecomparabletotramcriptlevelsseeninsomewarm
control samphs probed withCORle. No transcriptwas detected infitesesame
warm control samples probed With COR15a, so COR15b appears to have a low

background level of expression at 22°C

lnlightofthisdifferenceinregulafionacomparisonoffiiepromotenof
COR15a and comso raises some questions (Figure 3.10) Both contain at least
oneg-box, anelementfiiathasbeenfoundinplantgenes respondingtoavariety
of environmental stresses (Daugherty et al., 1994), and at least one DRE-like
element. The DRE has been shown to be involved in both cold- and drought-
regulated induction (Yamaguchi-Shinozaki and Shinozaki, 1994). Finally, a 14-
base sequence, designated fiie ”ll’ repeat” is found twice in fire promoter of
COR15b and once in file promoter of COR15a. It is not known if fitis' repeat
affects inducibility. It has already been demonstrated that the drought-
inducibility of COR15a is promoter-based (Baker et al., 1994). Is fite same true of
COR15b? If so, why does fire DRE-like element not cause drought inducibility?

This second question, while important, is beyond fire scope of fitis study.

100

corle -334 catttttcct ttatatatat atctc-t-a- -----tc-ga --q----tct
60:15. -322 l.C9-"'.I" --ees-s-ss

-299
-280

-254
-234

-206
-195

-167
-145

-121
-106

-72
-75

-25
-25

+26
+26

+61
+71

t‘ss 0.090 s

tgtaaatatc
OOOOOgOtOt

cacaacttga
OOOOOtOOCO

atagaa-atg
see-s at...

ttattcgata
--.-a.—-.c
11°;
--a-enoqtg
tt.at.....

tactattgqa

ctttcttgta
OCOCOOOtCO

ttaagag-tc
0.0.0 Otc. O

Coo--0000- soo‘ses... gse.tstese

-gt-aqccaa
tests-

g’bex

0-0.020'0. gg"‘sq.ee .c..ccc.ss

!L_£2222£___
--agttatta atgttqaaag ttgcaaataa -aa--c-aga aatgutanoa

l' Egzgnt

anatqotane
-..t

00....

nan-113. ole-nut

tqgoognoct

qttacacgta

‘OOOOOOOOI
g’bo:

gttaaaqata

tagttttgga

gqttagattt
O‘OOOOOOOC

tttattttcc
e 0,... e.“s

-ctCATG
tees...

‘Ctttttt"
geessssstt

a-ctagtacg
9C. s-s‘s--

atagcaatgc

-----e e a ct

ttctcattgg
90g... 0‘. O

ttctcatctc

tcccaaaaaa

----s s e 0 ac

atgtqtataa

-
O .0...

-ctt-ttggt
t...c..ct.

IICCIICQII

cog-see...

gcaaaaa-ta

togagagatc
..t.tc....
t? +1

actttctcca
.....q.t..

ttctttttlt

cgqttataac

-e-s-sg--e

----aa--cc
tctg..aa..

-a+actcttc
Cogs ------

tctagcactc
esot--eees

tataaaacga

tcttaaaact

----- catct
tctca.....

Figure 3.10 Comparison of fire promoters of COR15a and COR15b. Dots
represent identical bases and dashes mark gaps introduced for optimal
alignment. Promoter elements are bold and indicated by an overline or
underline for COR15a and COR15b, respectively. The start of transcription is
indicated by an arrow, and fire capitalized ATG is fire translation start codon.

101
AsaninitialassessmentofwhefirerfireCORlepromoterislow temperatureor

drought-responsive, two construct were prepared and introduced into
Arm. A 1,744 base pair fragment, including fire COR15b promoter and
part of fire COR15b coding region (—1212 to +535 relative to fire start of
transcription), and a300basepairfragmentoffireCOR15bpromoter (-226to
+75) were each fused to fire coding region of GUS to crash 101.35 and 101Nla3,
respecfively. Comb-act 101.135 was intended to be a translational fusion, but
sequencingrevealed firatfirerehad beenatwobasepairdeletionduringfire

doningprocesqsofireCORleandGUSmessagesareoutofframe.

Northern analysis of firree lines of plants transgenic for 101.138 revealed that fire
introduced gene is strongly responsive to low temperature, but not to drought
(Figure 3.11). This demonstrates firat induction by low temperature and fire lack
of drought inducibility are controlled by fire promoter, fire first exon, fire intron
orsomecombirrationoffirese. Twooffirefirreelinesof101N1a3 tested showed
the same responsiveness? to low temperature and lack of drought-inducibility,
but the firird showed slight drought induction (Figure 3.11). The reason for firis

isunclear.

102

101.138 101.1BS 10115.1BS

1391 141
WDA C WDA c WDSA c

M

101N1a3 lOlNla3 101Nla3
1323 1591 1791
WDA C WDA C WDA C

GUS . , ..

GUS

COR15

 

COR15

 

Figure 3.11 Regulation of COR15b promoter constructs. 101.135 consists of a
large fragment of the COR15b promoter (-1212 to +535 relative to the start of
transcription) fused to the GUS coding region, and 101Nla3 is the smaller
COR15b promoter fragment (-226 to +75 relative to transcription start) fused to
GUS. Blots of three lines of each are shown here, probed for both GUS and
endogenous COR15. Plants were grown at warm (normal) conditions (22°C, W),
drought-stressed (D), treated with ABA (A), or cold-treated (25°C, C).

103
Discussion

Analysis of genomic Soufinrns ofArebidopsis has revealed firat COR15 is a gene
family. Soufirer-n and sequence {analysis of a'genornic clone containing COR15a
hasfurfirershownfiratfinsecond memberoffirisfamilyissituatedlessfiran

1 kbdownstreamoffinfirst. Iffinreareofinrmembersoffirisfamily,finir
sequences are too highly divergent for finm to be detected wifir fire pHH67
(Hajela, 1990) probe at low stringency. Sequence analysis and comparison firen
indicated first firese two members of fire family share many characteristics. Tiny
sharehghsequerneidenfitymreinfinsameorientafionmrepredicbdtohave
anintroninfinsameplace, and finproteinsencodedbybofirappeartohavea
chloroplast signal peptide, to be about fin same size, and to have fin same pI

andstructure.

TlnrolesofalloffireCORproteinsinfinresponsetolowtemperatureare
currently unclear. It has been shown firat warm-grown plants expressing
transgenic COR15a are less susceptible to freezing damage, as measured by
chlorophyll fluorescence and by protoplast survival, firan are wildtype plants,
but fire mechanism for firis protection is unknown. COR15am and COR15bm
show high amino acid sequence identity wifir fin predicted mature forms of
3N115, 3N19 and BN26 from Brassica napus (Weretilnyk et al., 1993), with the

predicted protein product of a cold-responsive Brassica oleracea mRNA

C001

him

foldi

Ni

“10)“

Wt

“C01

104
(accession number U16751; )0 H, Lee C and Sohn U, unpublished data) and low

identitywithrepeatregionscharacteriflicofGroup3lEApronins (Dure,1993).
Withingroup3LEAproteim,finrearetwosubclasses. Tlndegernraterepeats
ofCORle show greater similarity wifirfinrepeats ofgroup3 LEA (I) (Dure,
1993),butCOR15bisacidic,asareproteinsclassifiedasgroup3LliAaI)(Curry
and Walker-Simmons, 1993). Unfortunately, neifirer structure nor function is

knownforanyofbaegroupsofproteins.

In analyzing fin function of a protein, it would be helpful to have a model of its
structure, so as to know which residues could be positioned to interact wifir a '
substrah, a solvent or anofinr member of a complex. Specific residues could
then be mutagenized to determine which are required for activity, and
comparisons to proteins of similar structure could be made. It is also lnlpful to
know fin structure so firat mutants can be designed to not disrupt protein
folding. Mutations are typically more permissible at fin solvent-exposed surface
urdmorelikelytorehtetopmteinfurnfionfinnresiduesburiedinfinpmtein
core. Also, mutations, including insertions and deletions, are typically only
allowed in turn- or loop- forming regions, rafirer firan wifirin secondary

structures.

Three secondary structure prediction programs have all indicated that CDRlSam

and CORlem are highly helical. In fire process of making firese predictions, it

~105
wasdiscoveredfiratpaflofaproteinofkrnwnsh'ucturemsubdomainofbeta-

lactamase, alignedwifirCORlSamwifirhighenoughaminoacidsequence
identity to suggest fin use of a knowledge-based modelling approach in

devdoping a model for COR15bm nrtiary structure.

Themodelwasconstructedwifirparticularcautionbecausefinsequerne
identity between CORISam or COR15bm and beta-lactamaae is at fin lower limit
for knowledge-based modelling (Sander and Schnieder, 1991) and because it
would be unexpected tor COR15am or COR15bm to function in antibiotic
cleavage, as does beta-lactamase. However, fire model does agree wifir fin
secondarysfiuchrrepredicfionofPHDsecurddoesconformtofinfeahiresof
known protein structures. Namely, hydrophobic residues are buried and
hydrophilic residuesareexpoaed to finsolvent. Tlntwo insertions occurringin
COR15bm relative to beta-lactamase are in loops between helices, and firus
would not disrupt fin overall struchrre of this domain. Finally, fire model shows
aglobularfeldofamphipafiricalphaheliceswhichcompriseonecompact
domainoffinbeta-lactamasestructure. Thisdomainhasveryfewinteractions
wifirfinofinrdomainsoitisexpected firatfindomaincorrespondingto
COR15bm could fold similarly in solution in fin absence of fin second domain.
However, fin sidechairrs in one of the surface loops are probably more tighfiy

packedinCOR15bmfiraninfinmodeLsincefinpredictedstructurehasa

106
tunnelfiuoughthepaiphery,whichisunexpectedforaglobuhrpmteinandis

notfoundinfinbeta-lactamasefromwhichfinmodelwasderived.

MoresupportforfinmodelcomesfmmfinobservationfirstfinCORlSm
prounsarepredicndbehighlyhelicalandamphipafiricandareknowntobe
mysolubleinaqusoussolution. ltisreasonabletofirinkfiratfinseproteins
eithermulfimaizeorfoldtoslueldfinhydrophobicfaceoffinhelixorlnlices
fromfinsurroundingsolvent. Itisnotyetknownwlnfirerfireymultimerize. If
finydonotmultimerfie,finnitisquitélikelyfirattheyfold,pmsiblyas
predicted by firis model which does gernrally have fin hydrophobic residues

buried.

ltisunlikelythatCORlSm isa beta-lactamase. Itcontaimfirreeoffinresidues
involved in catalysis (Ser, Lys and Gly) in fin same general region of fin
structureas beta-lactamase, butcorresponds toless firanhalfoffirebeta-
lactanrasemoleculeandlacksfindomaincontributingfinofinrhalfoffin
binding sin. It does, however, show some structural similarity to another class
ofmolscules, thelipoproteins, whichbind lipids fortransportfirroughfin
cytoplasm. When analyzed by firreading, fin mature forms of bofir COR15a and
COR15b were favorably compared to apolipophorin III from fin African locust

Leads nigrstsria (Breiter et al., 1991). While firis comparison was not useful for

107 .
modelling, it provoked some firought regarding fin possible function of

COR15am and COR15bm.

Theapolipophorinsininsects (BreiteretaL, 1991) andplantnompecific lipid
transfer proteins (nsLTPs) (I-ieirnmann et al., 1996) are composed of small
bundles of amphipafiric helices first are firought to spread finir hydrophobic
coreoverlipids, firussolubilizingfinmfortransportfirroughfinaqueous
cytoaol (Bseiter et al., 1991). Mature COR15m is similar in size to plant nsLTPs,
buthasalonlrafinrfiranfinhighpIcharacteristicoffireseproteins, andlacks
finir eight conserved cysteirns. Thus, fire COR15m proteins are not plant
nsLTPs, but may belong to a class of lipid-binding proteins first has not yet been
described. lrrdeed,acidicphospholipid exchangeproteinshavebeenreportedin
castorbeanfl'anakaandYamada, 1982)utdapinach(xadaetar, 1934),
althoughfireirsequencesarenotlmown. Also, ithasbeenfoundfiratfin
common structural feature of multiple amplripafiric lnlices is more important for
fin lipid transfer function of insect apolipoproteirrs firan amino acid identity
(Smifiretal, 1994). This lends credarce to fin idea thatCORlSm could have
lipid tramfer activity, even firough its sequence differs substantially from firat of
char lipid transfer proteins.

Because membranes are a principal site of damage during a freeze-finw cycle

and because membrane composition changes during cold acclimation, a lipid

108
transfer function for COR15m would not be unexpected. Ttn presence of

COR15am, and possibly COR15bm, in fin chloroplast would also be consistent
wifirapotenfialroleinlipidh-ansportsirnefinsynfinsisofglyoerolipids
originates in the chloroplast (Roughanet al., 1%0). In vitro assays such as firose
describedbyMiqueletal.(1%7)couldbeutilizedtodetermirnwlnfinrfin
COR15m proteins are able to facilitate fin transport of lipids. Ifsuch activity
couldbedebcndfinnfinmodelcouldbeusedfurfirertodirectamutafional
analysisofamhnacidresiduesinordertochfifywhkhresiduesarenecessary

for firis function.

Tlnsimilarityoffintwogenesandfinproteinsfinyarefiroughttoencode
suggeststhateifinrmoreofthegernproductisneededthancanbeproduced
fromonecopyoffingernorfintfintwofurnfionindifferentfimuplacesor
ways. Noneoffinsepossibilifieshavebeenruledoutalfiroughithasbeenseen
first fin principal difference between fire two genes is fin way firey are
regulated. COR15b, as studied by promoter/ GUS fusions and by a probe
specific for its transcript, has a low level of expression in warm-grown plants,
while COR15a does not (Baker et al., 1994). More notably, however, COR15a is
clearly inducible by drought stress, while COR15b is induced, if at all, to a

substantially lower level.

109 .
A construct containing a litfieiover a kilobase of fin COR15b promoter, fire first

exon and theintronofCORle, 101.138, was induciblein transgenic plants by
bwhmperahrre,butnotbydrouglrtindicafingfiratlow@rperahrrehrducfion
andfinlackofdroughtinducfionareregulatedbyfinpromoterorbyfinfirst
half of fin transcript. A second construct containing about 300 bases of promoter
sequence and no coding sequence, 101Nla3, gave more variable results. All of
finlinesnstedwereinduciblebylownmperaturefindicafingfiratfirisis
conkelledatfinpromoterlevel, butoneoffiremalsoshowed inductionin

drought-stressedplanh.

Wifir only three reliable samples to examine, it is difficult to know which is
aberrant. It is possible first an element which suppresses drought-inducible
expressionwasnotincluded infirissmallconstructand firatallfirreelines
shouldhaveshowndroughtinduction. Itisalsopossible firatallfirreelines
should not have shown drought-induction, similar to fin results from 101.135.
Alternatively, two of fin lines could be showing fin effects of trans-inactivation.
Thissaunsurfiikelynunefinconsfiucbaresfiflhrduciblebylowmperature.
Whateverthecase,finresultsseem toreflectfineffectofmepositionoffin
transgenesinfingenome. Thus, moreworkneeds tobedonetodetermirnwhy
fire endogenous COR15b gene is not responsive to drought stress, butfin current
evidence suggests that the lack of inducibility is promoter-based. More

transgenic lines have been generated, which could he hated, alfirough fire

110
possibility fiat fin DRE-like element present in fin COR15b promoterpis

pr'usarily responsible for induction due to low nmpersture but not drought is
enticing. Answeringfinquestionofwlnfinrornotfitiselementrespondsto

bofitstimuhcouldulfimatelybemoremformativefinnfiyingtoumavelfinse

transgenics.

Why does fin DRE-like element in fin promoter of COR15b not cause COR15b
RNAlevelstobeinduCiblebydroughtstress? ltispossible thatnnssageis
induced but also degraded more quickly under drought conditions than low
mperature. A construct containing fin coding region of COR15b driven by fin
CsMVSESpromoterlnsbeenmadetobstfitispossibility, butfinanswerisnot

yet known.

If, however, fin promoter of COR15b is not responsive to drought, there are two
masonable possibilities regarding fin role of fin DRE-like enment. Bases
surrounding fin CCGAC core could affect its responsiveness or it could require
interaction wifit a second promoter element. It has already been shown fitat fin
core g-box (ACCT) is not specific to ABA inducibility, but is required for
induction under numerous stresses (Foster et al., 1994). G-boxe's wifit slightly
differing surrounding sequences (Foster et al., 1994) and in combination wifii
differing promoter elements (Kawagoe et al., 1994, Slnn and Ho, 1995 and

Rattan et al., 1995) are instrumental in responses to different stresses. Thus it is

111 .
possiblethstthenucleofidessurmundingtheDRE-likeelementinfinCORle

promonr cause fitis promoterelement to be non-responsive to drought stress.
This would be congruent with fin observation fitat a COR15a promoter deletion
constructwhich includes only the DRE-like element identical to fin COR15b
DRE-likeehnentinthecontextoffinnative promoterisnotverydrought-
responsive, while COR15a promoter deletion constructs including two more
DRE-like elements, which have slighfiy differing surrounding sequences, are
more strongly responsive to drought (Baker et al., 1994). In ofinr experiments,
two DRE-like elements found in fin rd29A (COR78) promoter, which contain fin
DRE CCGAC core sequence, did not show drought-inducibility in tobacco

(Yamaguchi-Shinozaki and Shinozaki, 1994).

Alternatively, cooperation wifii a second Cir—acting element not present in this
Mmyberequned. 'I'lnresultsonamaguchi-ShinozakisndShinozaki
(1991) suggest that a single DRE derived from rd29A (COR78) is sufficient for
drought-inducible expression, since an 83 bp fragment containing fin upstream
DRE ora 53 bp fragmentcontsining findownstream DRE fused m a minimal
promoter derived from fin same gene and to GUS gives drought-inducible GUS
activity. Further, five copies of a 25 bp fragment containing fin proximal DRE,
when fused to fine minimal promoter of rd29A (COR78) driving GUS, conferred
drought-inducibility to GUS in transgenic tobacco (Yamaguchi-Shinozaki and

Shinozaki, 1994). It is possible that finre is a second promoter element contairnd

112
wifirineschofthesefragments, althoughitisnotlikely thatfinZS-base

fragmart would contain anofinr element. However, fin presence of multiple
copies of fin same elment might make up for fire lack of a coupling element. It
isalsopossiblefiratfinminimalpromoterusedcontainsfinrequired second
elemsntorthatfinfurntionoffinsepmmoterdemenbisdifferentintobacco
thaninArsbidapst’s. Arabidopsr’s ischilling'tolerantand is ableto acclimateto
freezing temperatures, while tobacco is chilling sensitive and unable to acclimate

to freezing.

In order to test wlntlnr fin DRE-like element from COR15a or COR15b is
inducible by both cold and drought stress or only by cold stress, experiments
comparing fin inducibility of constructs containing an oligonucleotide of fin
DRE from COR78, or fin DRE-like elements of COR15a or COR15b fused to a
minimalprornou'andareportershouldbecarried outinArabidopsis.
Constructs containing ofinr promoter elements along wifir fin DRE-like
elements or containing multiple copies of fin DRE-like element could give

insight into wlnfinr fin elements are functioning syrnrgistically.

Both COR15a and COR15b appear to be inducible by ABA in gene-specific
norfinrn analysis. This result was called into question when fin COR15b
promoter constructs were not seen to be induced by ABA (Figure 3.11), but fin

ABA induction of fin endogenous gene was weak enough first it is not clear that

113
inductionwould havebeendetectable. Tlngenespecificnorfinrnblotsneed to

be repeatd, since only orn ABA-induced RNA sample was examined initially,
andfinfiansgenicsmustbetestedwifirbetterABAinducfionbeforeitcanbe

unequivocally determirnd wlnfinr COR15b is responsive to ABA.

ThepherromenonofCORgernsexistingassmsllgernfanriMmembersof
whicharewifirinfirreekilobasesofeachofinronachromosomeandare
differentially regulated, has now been seen for all four COR gene families in
Arsbidopsr’s (Figure 3.12). COR6.6 and kinl are related to each ofirer essentially
fin same way COR15a and COR15b are related to each ofinr (Kurkela and Borg-
Franck, 1992). Both show induction by low temperature and ABA, but only
CORG.‘ is strongly drought-responsive and only COR6.6 is expressed at low
levels in control plants (Kurkela and Borg-Franck, 1992). However, Wang et al.
(1995)usedpromoter fusionsoffinsetwo genestoshow firatfinkinl and
COR6.6 promoters both responded to drought stress. This reason for firis
difference has not yet been reported. COR78 and 11293 differ in firat fin former
is more strongly cold-responsive, while fin latter is more strongly drought-
regulated (Yamaguchi-Shinozaki and Shinozaki, 1993), and COR47 is more
strongly responsive to drought and ABA firan is lti45 (also called lti29) (Welin et

al., 1994 and Welin et al., 1995).

114

—L'I:l I'l:l—

 

 

 

 

CORlSa COR15b
—li:r] "En—
kinl kinZ/COR6.6
Ill-— 1' [ll—-
«ms/mes rd29A/lh78/COR78
—E- 1'“—
mrS/um COR47

 

1kb

Figure 3.12 Genomic organization of COR gern families. Coding regions are
shown as filled boxes, while introns are depicted as open boxes. Transcription
start sites are indicated by arrows.

115 .
Tln members of each of fin four gene families share 49% (COR78 and 111298),

67% (COR47 and M45) and 91% (kind and COR6.6) amino acid identity
(Yamaguchi-Shinozaki and Shinozaki, 1993, Welin et al., 1995, Kurkela and
Borg-Franck, 1992 and Wang et al., 1995). These high amounts of identity
suggest that the members of a given gern family may have identical or reland
functions. Perhapsfinsmalldifferencesamongmembersofagernfamilymake
onemsmberbettersuindforprotectionoffinplantunderdroughtsfiess
conditions without fin addition of low temperature.

Thesharedspatialorganizationoffinmembersoffiresefamiliesisalso
intriguing. It is possible firat finir proximity to each ofinr is advantageous in
firatfinopmringoffinchromatininornregionallowsfintranscriptionoftwo
cold-inducible gerns for fin price of one. Wifir firis in mind, fin question of
where fin COR gene families are in relation to each ofinr on fin chromosome
becomes more interesting. Perhaps finre is orn loop of chromatin that contains
all four families and becomes available for active transcription wlnn fin cell is
exposed to low temperature. Alternatively, perhaps COR gern families are
scattered in fin genome, but are near other, less highly-expressed cold-inducible

genes.

116
Materials and Methods

Sequencing and sequence analysis

Sequencing of fin region of fin genomic subclorn corresponding to fin COR15a
cDNA pHH67 (Hajela et al., 1990), pBlue67, and of fin junction between fin
GUS gern and COR15b promoter of 101Nla3 and 101.1115 was performed using
finmefirodofSangeretal. (1977), usingdoublebstranded DNA nmplatesas
described previously (Lin and Thomashow, 1992). Deletions of fin genomic
DNA insert were gernrated by digestion with exonuclease III and mung bean
nuclease (Henikoff, 1987). Automated fluorescent sequencing of COR15b was
performed by fin MSU-DOE—PRL Plant Biochemistry Facility using fin ABI

Catalyst 800 for Taq cycle sequencing and fin 373A Sequencer for fin analysis of

products.

Nucleotide sequence analysis was carried out using MacVector'“ 3.5
(Isnternational Biotechnologies, Inc.), and protein analysis was done using
Pz'obean (DNAST AR), PHDsec and PHDfirreader (www address:
11%: / /www.embl-lnidelberg.de/ predictprotein/ phd_pred.html). Protein

modelling was performed using InsightII (version 95.0).

117
HantgsusmicDNAextncfionandSouthernanalysis

DNAextrscfionforfindenrminationoffinnumberofmembersdfinCORIS
genefamflywasperformedasdescribedbyRogersandBendich(1988)except
that fin mixture of plant tissue and extraction buffer was incubated at 68°C for
as minuhs before chloroform extraction and the CI'AB precipitation was left
unfilflooculantprecipitatewasseen. Thissometimestookfiu'eeorfourdays,
andfinsannprocedureusedwifirfinsameCl‘ABsyearorsolanrgave
varyingyields, so DNA preps used inanalyzing finCORle promoter/GUS
transgerninserbwere performed according toTai and Tanksley (1991) as
modifiedbyDougDahlbeck. Uptoo.15g1eaitirruewar'trozenma
microcenuifugembsinliquidnifiugmaushedwifiraminiafirrepesfieand
incubated at65°Cwifir 0.7m] pretreated extraction buffer (100 mM Tris-HG pH
8,50mM EDTA pH 8,500mM NaCl, 1.25% SDS,8.3 mN NaOI-I, 0.38% sodium
bisdfite) for 10 minutes. Then 0.22 ml 5 M potassium acetan was added, fin
“mph were incubated on ice for 40 minutes and microcentrifuged at 13,0“)
x-prn for3minutes. TlnsupernatantwasfilteredfirroughspieceofKimwipe’in
r- 1 ml pipetmau tip into another microcentrifuge tube and precipitated with 0.7
volumeofisopropanol. TlnpelletwasresuspendedinSMulTsE(50mMTris-
I-ICI pH 8, 10 mM EDTA pH 8), allowed to resuspend at 65°C for 5 minutes,
nimdwifirlSOulthammonium acetsnandmicrocentrifuged for3minuns.
I'll-e supematentwasfinntransferrodtoarnwtubeandDNAwasprecipitsnd

Witls 330 pl isoprOpanol. The pellet was waslnd twice with 70% EtOH, allowed

118
tsresuspendat65°Cfor5minutesin100ulTsEandprecipitatedwifir10ul

sodiumacetateand75ulisopropanol. Finally,finpelletwaswashedtwicewifir

70% EDI-LdriedandresuspendedinZSulsteriledoubledisfilledwater.

Total Arsbidapsr's DNA or plasmid DNA was digested wifir various restriction

arzymesandfracfiomtedbyagarosegelelecfiophoresisusingTAE,whkhis40
mM Tris-acetate/ 1 mM ED'I'A (Sambrook et al., 1989). Soufirern blots were finn
preparedusingstandardmefirods($ambrooketal.,1%9) andfinDNAwasUV

cross-linked (Stratalinker, Stratagern) to fin filters.

Standard genomic Soufinm blots were prehybridized and hybridized in a
Robbins Scientific Model 400 hybridizafion oven using fire nonfat dry milk
mefirod orohnson et al. (1%4) and waslnd using standard conditions
(Sambrook et al., 1%9), except first all washes were dorn wifir 0.1% SDS/0.1x
SSC. High stringency blots were prehybridized in 2x SSC/ 0.25% nonfat dry
milk/0.5% SDS at 55°C for firree hours, hybridized overnight in the same
solution at 68°C, and washed in 0.1x SSC/0.1% SDS. Tln first four 20-minute
washes were at room temperature and fin last two were at 68°C. Low stringency
blots were prehybridized in 6x $C/ 0.25% nonfat dry milk/0.5% SDS for firree
hours at 55°C, hybridized in fin same solution overnight at 64°C and washed
four times, for 20 minutes each time, at room temperature and twice, for 20

minutes each time at so—oonc in 6x mums sns. Gel-isolated fragments

119
labeledwifirfiPbyrandompr-imingmeinbergandVogelstein, 1%3) wereused

as probes. A SscI/Hisdm fragmentoprIl01.3 was used to denctGUS inserts,
and anEmRIfragmentoprH67wasused tovisualizeCORlSa andCORle

W

rum power aid stress treatment
WWQ.)l-leynecotypeRLDorColumbiswasgrownin
controlled environment chambers at 22°C wifir a 24 hour photoperiod (about 120
unrol rrr-2 s4) as previously described (Gilmour et al., 1%8). Humidity was not
controlled. After about firree weeks of growfir, plants were eifinr harvested or
subiscnd tooneofseveralstresstreatments. Plantswerecold-acclimated by
placing pots in a 15°C cold room under constant cool-wlu’te fluorescent light (45-
55umolm'3s") forfourdays. Droughtstresswas administeredbyexcisingfin
shootsfromfinpots,placingfirsmon15cmthtmanNo. 1 filterpaperin150
nun x15mmplasticpetriplateswifirfinlids openfor45minutes. Tlnlidswere
finnplscedonfinplatesandplantswereincubated forfour tosixhoursunder
fluorescent light (about 50 unrol m'3s4). Water loss was quantified by dividing
fireweightoffindrought—stressed plantsbyfinirinifialfreshweightand
multiplying by 100. Plants lost from 18% to 23% of finir initial fresh weight.

The ABA treatment of plants used in fin gene-specific norfinrn experiments
consisted of spraying plants to runoff wifir 50 MA ABA (mixed isomer, Sigma) in

0.02% polyoxyefirylern-sorbitan monolaurate 20 (Tween 20), covering fin pots

120
withSaranwraptoslowevaporafionandincubatingflieplanhformatm

AlloflurABAhestmenbwaeperfoi-medbypladngexcisedplantsintoalso
mmx15mmpetriplatewith75mlof100uMABA(mixedisomers,Sigma)and
onedrop(10-20 ul)'l'woen-20, mvm.mmmapued15mwmm
No.1filhrpapertokeepthmsubmerged,placingthelidontheplateand
incubatingthemfor4t06hoursunderfluorescentlighh(aboutSOumolm'zs'l).
Wat—hsoakedplanhunderwenttlussameh‘eatmentemeptduttheABAwu

omitted.

RNA «bastion and fractionation

Total RNA was extracbd essentially as described (Gilmour et al., 1%8) with a
few modifications. Frozen pulverized tissue was extracted with equal volumes
of phenol/chloroform/isoamyl alcohol (25/24/1, v/v/v) and extraction buffer
(1% w/v hihopropylnaphthelene sulfonic acid, 6% w/v p-aminosalicylic acid,
100 all Tris-HG pH 7.6, 50 mM EGTA, 125 mM NaCl, 1% w/v SDS, 10 mM
D'FDonice. 'l'hiswasfurtherhomogenizedinthetubeusinga'l’ekmar’
Tissumizer, centrifuged (10,000 rpm in the SA600 rotor), and the supernatant
ext-acted again with phenol/chloroform/isoamyl alcohol. This second set of
hibesandaflsubsequentsolutionsexceptethanolweremadefreeofRNaseby
treatment with DEPC (Sambrook et al., 1989). Nucleic acids were precipitated
withcold 95% ethanol, resuspended inl ml doubledistilled water, transferred

to a microcentrifuge tube and precipitated on ice for an hour with 1/4 volume 10

121
M LiCl. The pellets were precipitated again with 95% EtOH and resuspended

inZOOquater. TheOonwasmeasuredtoestimatetheconcentrationofRNA.

ForfracfionaflonRNAfitolOuflwasdrieddownandresuspendedin
formaldehyde loading buffer containing m: (about 100 ng/ml), incubated at
68C for 10-15 minutes and fractionated on denaturing formaldehyde agarose
yels($ambrooketal., 1%9). TheRNAwasthentransferied (SambrooketaL,
1%) to Magns NT membranes (MSI) using 10): $PE made according to
instructions provided by Schleicher and Schuell. This recipe for SSPB is slightly
4mm thatdescribed-bySambrooketal. (1989). RNAwasUVcross-

linked (Strahlinker, Stratagene) to the filters.

Primes extension
Primerextension, performedasdescribedbyHorvathetal. (19%) andusingthe
oligonucleotide S’CAGCTCCTGAGAAAGACATCGC-S’, was used to determine

the transcription initiation site of COR in cold-treated plants.

Oligonncleotide northern analysis

Oligonucleotide northerns were done using the gene specific oligonucleotides
S’TGCI'I'GAAATTAACTGATTAGGTAAGACCCB’ (COR15a) and
yGAACATGACTACATGAGTGGTTGAATCAGGS’ (COR15b), which were

end-labeledasdescribedbyleffetal,(1%7)withminormodifications (Baker,

122
19%). FirstmulofNOuCiy-“PATPPSNOCi/mmobweredriedand

resuspended in 1 pl oligonucleotide (100 ng/ pl), 1 ul T4 polynucleotide kinase,
1 ull MTris-PKleHB, 1 ulOJM MgCl, 2.5 “10.034 011‘, and3.5uldouble
distilled autoclaved water. This reaction mix was incubated for 30 minutes at
37°C, then inach'vated at 65°C for five minutes and brought to 100 ul with TB (10

mM TrisvHCl pH 8, 1 mM EDTA). Unincorporated nucleotides were removed

usingaSephadexG-25spuncolumn($ambrook,1989).

Hybridization was performed as described (Baker, 19%), except that the
temperatureusedwasGBGCinsteadofu‘C. Thistemperaturehadbeen
smpiricallyshown (Sambrooketal, 1989) togivegene-specific bindingofthe
two probes. Filters were incubated in prehybridization solution (5x $C, 10x
Denhardt’ s solution, 7% 506, 100 ug sheared, denatured fish sperm DNA) for
four hours at 63C, then in hybridization solution (5x SSC, 1% SDS, 10% dextran
sulfate, 100 pg sheared, denatured fish sperm DNA) at 63C overnight. The
filters were washed once in 5x $C/5% SOS/5x Denhardt’s solution at 63°C for
30minutes,onoein3xSSC/2% SDSat63°Cfor45minutesandoncein2x$C

for 10 minutes at room temperature.

Cloning of COR15b promoter constructs
In the construction of 101Nla3, an NlallI fragment containing approximately 300

bp of the COR15b promoter (-226 to it 75 relative to the transcription start site)

123 .
was ligahd into the Sphl site of pUC19 (Yanisch-Perron et al., 1985), such that

meS’sndofthefragment, relativetotheCORlecodingregion,isnearestthe
HindlflsiteoprC19. TheimertwasthencutoutwithHiudflIandealand

claudintopillOlJOeffersonJMcutwithHiudmandesL

M 101.135 was made byusing thelargefragmentofDNA'polymerasel
(Klenow) tofillintheS’ overhang ofa Scsl/ Hindlllfragmentwhichcontains
approximately 1.5 kbofthepromoter(-1212to+535) and aaboutathird ofdie
codingregionofCOR15b. Thisfragmentwascloned intotheSmalsiteof
pBlusscript (SK-9 such that the 5’ end relative to me coding region of COR15b is
nearestthePstlsiteofthepBluescriptpolylinker. Thisnewconstructwasthen
cut 5011/3me and the insert was gel-isolated and ligated into pB1101.1

(Jefferson, 1987), which had also been cut with Sell and Bani-II. This should

haveputthechloroplastsignalpeptideofCORleinframewithGUS.

Plant transformation

The p31 101.1 recombinant were mobilized into Agrobscterims tuucfaciens strain
C58C1 (pMP90) (Koncz and Schell, 1%6) or EHA105 (Hood et al., 1993) by
triparentalmah‘ng, usingstrain pRK2013asahelper. Theroottramformation
technique of Valvekens et al., (1,988) with the modifications described by Baker

et al., (1994) was used to transform Arabidapsis Minna (I...) Heyn, ecotype RLD.

. 124
Northern hybridization

For standard northern hybridization, blots were prehybridized, hybridized and
washedinaRobbinsScientific Model400hybridintionovenaocordingto
standard methods (Ausubel et al., 1987) with some modifications. Blots were
incubahd with prehybridization solution (50% v/v formamide, 5x 9592, so mM
potassium phosphate pH 8, 5x Denhardt’s solution, 0.5% $06, 100 ug/ml
sheared, denatured fishsperm DNA) at42'C forthreehours to overnightand in
hybridization solution (50% formamide, 5x mp2, so mM potassium phosphate
pH 6.5, 1x Denhardt’ s solution, 0.5% SDS, 5% dextran sulfate, 100 ug/ ml sheared
densttned fishsperm DNA) attZ‘Covernight. Blohwererinsed and thengiven
two Sill-minute washes at room temperature in 2x SSPE/0.5% SDS, then washed

two to three times, for 15 minutes each time, with 0.1x SSPE/0.5% 505 at 50°C.

Probesweremadefrom gel-isolated fragmentslabeledwithfll’byrandom
priming (Feinberg and Vogelstein, 1983). An [5le fragment from pI-IH67 was
used to visualize COR15 message, and a Sad/Hind III fragment from pB1101.3

(Jefferson, 1%?) served as a GUS probe.

125
Literature Cited

Abola EB, BernsteinFC, BryantSH, Koetzle'l'F, Weng](1987) Protein Data
Bank. In, FH Allen, G Bergerhoff, R Sievers, eds, Crystallographic Databases-
information content, software systems, scientific applications. Data Commission
of the International Union of Crystallography, Bonn, Cambridge, Chester,
pp107-132

Altschul AF, Gish W, Miller W, Myers EW, Lipman D] (1990) Basic local
alignment search tool. J. Mol. Biol. 215:403-410

Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith IA, Struhl K
(1987) Current Protocols in Plant Molecular Biology. Greene Publishing
Associates and Wiley lnterscience, New York, Chichester, Brisbane, Toronto,
Singapore

Baker $ (1990) Cloning, identification and characterization of fertilization
associated genes in Pinus strobus (I...) ovules. Ph. D. thesis, Ohio State University

Baker SS, Wilhelm KS, Thomashow MF (1994) The 5'-region of Arabidopsis
W 00115! has cis-acting elements that confer cold-, drought- and ABA-
regulated gene expression. Plant Mol. Biol. 24: 701-713

Breiter DR, Kanost MR, Benning MM, Wesenberg G, Law 1H, Wells MA,
Rayment l, Holden HM (1991) Molecular structure of an apolipoprotein
determined at 2.5-A resolution. Biochemistry 30:603-6m

Brown JWS (1%6) A catalogue of splice junction and putative branch point
sequences from plant introns. Nucl. Acids Res. 14:9549-9559

Chen CCH, Herzberg O (1992) Inhibition of beta-lactamase by clavulanate:
trapped intermediates in cry'ocrystallographic studies. J. Mol. Biol. 224:1103-
1113

Chou PY, Fasman GD (1978) Prediction of the secondary structure of proteins
from their amino acid sequence. Adv. Enzymol. 47: 45-148

Curry ],.Walker-Simmons MK (1993) Unusual sequence of group 3 (ll) LEA
mRNA inducible by dehydration stress in wheat. Plant Mol. Biol. 21:907-912

126

Daughtry CI, Rooney MF, Paul A-L, de Vetten N, Vega-Palas MA, Lu G,
Gurley WB, Fed R] (1994) Environmental stress and gene regulation. In,EM
Meyerowitz, CR Somerville, eds, Arabldopsis. Cold Spring Harbor Laboratory
Press, Cold Spring Harbor, New York, pp 769-806

DunruflP(1968)Theuseofhdicdnet-diagramstorepramrtpmteinstucm.
Biophys. J. 8:865-875

DureIflL(1993)Arepeafing11-meraminoaddmotifandphntdesiccationlhe
PlantJournal3z363-369

FeinbergAP,VogelsbinB(1%3)Abchniqueforradiolabelingrestr-iction
endonuclease fragment to high specific activity. Anal. Biochun. 132:6—13

Fests R, lnwa T, Chua N-H (1994) Plant bZIP proteins gather at ACGT
element. FASEB J. 8:192-200

Franz G, Hatzopoulos P, IonesT], Krauss M, Sung ZR (1%9) Molecular and
gsneticanalysisofanembryonicgene,DC8,fromDauarscaetL MoLGen.
Genet. 218:143-151

Garnier I, Osguthhorpe DI, Robson B (1978) Analysis of the accuracy and
implications of simple methods for predicting the secondary structure of gobular
proteins. J. Mol. Biol. 120:97-120

Gavel Y, von Heijne G (1990) A conserved cleavage-sit motif in chloroplast
transit peptides. FEBS Lett. 261:455-458

Gilmour SI, Hajela RK, Thomashow MF (1988) Cold acclimation in Arafldapais
Minna. Plant Physiol. 87:745-750

Hajela RK. Horvath DP, Gilmour SI, Thomashow MF (1990) Molecular cloning
and expression of car (cold-regulated) genes in Arsbidopsis thaliana. Plant Pysiol.
93:1246-1252

Hatton D, Sablowski R, Yung M-H, Smith C, Schuch W, Bevan M (1995) Two
classes of cis sequences contribute to tissue-specific expression of a PAL2
promoter in transgenic tobacco. The Plant Iourn. 7:859-876

HeinemannB,AndenenKV,NiehenPR,BechLM,PouhenFM(l996)Stucture
insolutionofafour-helixlipidbindingprobin. ProteinSci.5:13-23

Iieinz DW, Baase WA, Dahlquist FW, Matthews BW (1993) How amino-acid
insertions are allowed in an a-helix of T4 lysozyme. Nature 361:561-564

127

HenikoffSOQ?) UnidirectionaldigestionwithexonucleaseminDNA
sequence analysis. Meth. EnzymoL 155:156-165

HoodEEdeinSB,MekheuIS,HoekemaA(l993)NewAgmbactsriuwhelper
plasmidsforgenetransfertoplant. Transgenic Research 2:208-218

HorvathDP, McInmeyBK,ThomashowMF(1993) RegulationofArsbidopsis
W L. (Heyn) M8 inresponse to low temperature. Plant Physiol. 103:1047-
1053

ItingYC,Chan,ShihM,lbiehLClwa(1996)Unusualsequenouofgroup
3I£AmRNAinduciblebymaturationordryinginsoybeanseedn PlantMol.
BioLlnpress

Jefferson RA (1%?) Assaying chimeric genes in plant: the GUS gene fusion
system. Plant MoL Biol Rep. 5:387-405

Johnson DA, Gautch IW, Sportsman IR, Elder IH (1W4) Improved technique
ufilizingmnfatdrymflkforandysisofproteinsandnucldcaddstarnferredto
nitrocellulose. GeneAnaLTech. 1:3-8

IoshiCP(1%7)AninspectionofthedomainbetweenputativeTATAboxand
tanslstionalstartsitin79plantgeus. NucLAcidsRes.15:6643-6653

Kader 1c, Julienne M, Vergnolle c (1984) Purification and characterization of a
spinach leaf protein capable of transferring phospholipids from liposomes to
mitochondria or chloroplast. Eur. J. Biochem. 139:411-416

Kawagoe Y, Campbell BR, Murai N (1994) Synergism between CACGT G (g-box)
and CACCI‘G ciselement is required for activation of the bean seed storage

prouinfi-Min gene. The Plantlourn. 5:885-890

KonczQSchellI(l986)lhepromoterolerDNAgene5contolsthetissue-

specificexpressionofchimeric genescarriedbyanoveltypeongrabacteriunl
binary vector. Mol. Gen. Genet. 204:383-396

Knox IR, Moews PC (1991) Beta-lactamase of Bacillus lidlcmfurnis 749/ C.
Refinement at 2 A resolution and analysis of hydration. J. Mol. Biol. 220:435-455

Kurkela s, Borg-Franck M (1992) Structure and expression of m2, one of two
cold-andABA-indmedgenesofArsbidopsistllalm PlantMoLBioL19:689-692

128

Lin C, Thomashow MF (1992) DNA sequence analysis of a complementary
DNA for cold-regulated Arsla’dopsi’s gene csr15 and charachrization of the COR15
polypeptide. Plant Physiol. 99:519-525

Madg‘wickPJ,WaleySG(1%7) Beta-lactamaselfromBscillusareus. Structure
and sit-diractd mutagenesis. Biechm. J. 248:657-662

Marqusee S, Baldwin RL (137) Helix stabilization by Glu'-Lys* salt bridges in
shortpeptidssofdcnwodesign. PNAS84:88%-8902

MiqsrelhtllockMAJoyardLDerneA-LDubach-EKaderJ-QDouceR
(1987) Protein-mediated transfer of phosphatidylcheline from liposomes to
spinach chloroplast envelope membranes. Biechimica et Biophysica Acta
937:219-228 '

Puupponen-PimiaR,SaleheimeM,VasaraT, Ra R,Gaugech, KurtenU,
Knowles J, Ker-anen S, Kauppinen V (1993) Characterization of a birch (Bends
penduls Roth.) embryonic gene, BP8. Plant Mol. Biol. 23:423-428 ‘

Rogers S0, Bendich AJ (1988) Extraction of DNA from plant tissues. In SB
Gelvin, RA Schilpereort, eds., Plant, Molecular Biology Manual Kluwer
Academic Publishers, Dordrecht, Boston, London, pp A6/1-A6/ 11

RestB(1995s)Fitting1Dpredictionsinto3Dstr-uctiues. In,HBehr,SBrynak,
engrotinFolds. ADistance-‘basedAppreach. CRCPresnBocaRatorhpp132-
151

Rest B (1995b) TOPITS: Threading one-dimensional predictions into three-
dimensional structures. In, C Rawlings, D Clark, R Altman, L Hunter, T

ungauer, S Wodak, eds, The third international conference on inhlligent
systems for molecular biology (ISMB). AAA] Press, Menlo Park CA, pp314-321

Rest B,SanderC(1993) Predictionofpreteinsecendarystuctureatbettrthan
70% accuracy. J. Mol. Biol. 232:584-599

Rest B, Sander C (1994) Combining evolutionary information and neural
networks to predict secondary structure. Proteins 19:55-77

Rethstein SJ, Inlmers KN, Lettein RJ, Carozzi NB, Jayne SM, Rice DA (1987)
Promoter cassettes, antibiotic-resistance genes, and vectors for plant
transformation. Gene 53:153-161

129 ,
Reughan PG, Holland R, Slack'CR (1980) The role of chloroplast and
micresemal fractions in polar lipid synthesis from [144C acetate by cell-free
preparations from spinach (Spinach oleracea) leaves. Biochem. J. 188: 17-24

SanderC,SchneiderR(1991) Databaseefhemolegy—derived protein structures
andflrestr'ucturalmeaningefsequencealignment. Preteins9:56-68

SambreekJ, FritchEF, ManiatisT (1W9) MolecularCloning: Alaberatory
manual, second edition. Cold Spring Harbor Laboratory, Cold Spring Harbor,
New York

Sanger F, Nicklsn S, Coulsen AR (1977) DNA sequencing with chain-terminating
inhibitors. PNAS 74: W7

SchifiathEdmundsonAB(1967)Useofhelicalwheelsterepressntthe
stuctirreefpreteinsandidentifysegmentwithhelicalpotentials. Biophys.J.
7:121-135

Shen Q, He T-H D (1995) Functional dissection of an abscisic acid (ABA)-
inducible gene reveals two independent ABA-responsive complexes each
centainingag-bexandanevelcis-actingelement. ThePlantCell7:295-307

SmithAF,OwenLM,StebelLM,ChentLKanestMR,HannemanF.WdlsMA
(1994) Exchangeable apolipoprehins of insect share a common structural motif.
I.LipidRes.35:1976-1%4

Sell J, Alefsmi H (1993) The protein import apparatus of chloroplast. Physiol.
Plant. 87:433-440

ShyndkaNCJAdachilLJensenSE,JeluuK,SieleckiA,BetzelC,Sutth
JamesMNG (1992) Molecularstr'uctureoftheacyl-enzymeintermediateinfi-
lactamhydrolysisatlJAresolutien. Nature359z700-705

TaiTlLTanksleySD(1991)Arapid and inexpensivemethedforiselationef
total DNA from dehydrated plant tissue. Plant Mol. Biol. Rep. 8:297-303

Tanaka T, Yamada M (1982) Propertiesm of phospholipid exchange proteins from
germinated castor bean endosperms. In,JFGM Wintermans, PJC Kuiper, eds. .,
Biochemistry and metabolism of plant lipids. Elsevier Biomedical Press B. V.,
North-Holland, Amsterdam, pp 99-106

Valvekens D, Van Mentagu M, Van Lijsebettens M (1988) Was
trim-mediated transformation of Arabidopsis thaliana root explant by
using kanamycin selection. PNAS 85: 5536-5540

130

Wang H, Datla R, Georges F, Leewen M, Cutler A (1995) Promoters from kinl
and cor-6. 6, two homologous Arsbidopsis thalisns genes: transcriptional regulation
and gene expression induced by low tmperatiue, ABA, esmeticum and
dehydration. Plant Mol. Biol. 28:605-617 A

Welin BV, Gson A, Nylander M, Palva ET (1994) (Juracterizatien and
differential expression of tfltrr/Iea/rsb-like genes during cold acclimation and
drought stress in Arsbidopsis Minus. Plant Mol. Biol. 26:131-144

Welin BV, Olson A, Palva ET (1995) Structure and organization of two closely
related lew-temperature—induced tflrnfleq/rsb-like genes in Arsbidepsis thaliana L.
Heynh. Plant Mel. Biol. 29:391-395

Weretilnyk E, Orr W, White TC, In B, 8th (1993) Characterization of three
related low-temperature regulated cDNAs from winter W nspus. Plant
PhysioL 101:107-177

Yamaguchi-Shinezaki K, Shinozaki K (1993) Characterization of file expression
of a desiccation-responsive r1129 gene of Arsbidopsis thalitma and fire analysis of
it promoter in transgenic plant. Mol. Gen. Genet. 236:331-340

Yamaguchi-SlfimkiKSlu‘nenkiK(1994)Anevelds—actingelementman
Arsbidopsr's gene is involved in responsiveness to drought, low-temperature, or
high-salt stress. The Plant Cell 6:251-264

Yanisch—Perron c, Vieira J, Messing J (1985) Improved M13 phage cloning
vectors and host strains: nucleotide sequences of fire M13mp18 and pUC19
vectors. Gene 33:103-119

Zeff RA, Geier SS, Nafiiersen SC (1986) Molecular less variant of the murine
major histecempatibility complex: nonexpressien 'ef H-2K antigen associated
with marked reduction in H-2K mRNA as determined by eligenucleetide
hybridization analysis. Immunology 137:1366-1370

Chapter 4:
Examination of the low temperature induction profiles

of COR gene families

Summary

The low temperature induction profiles of fire COR6.6, COR15, COR47 and
COR78 gene families and of a GUS gene fused to file COR15a promoter were
examined in transgenic Arabidopsis. In each case, expression was first seen to
increase at about 16C and message levels gradually rose as fire ten'lperature was
lowered, reaching a maximum at about~8°C. This pattern was seen whefiier fire
planthadbeenshifteddewntwedegreesatatimeorhadbeendrepped from
growth temperature to file temperature of interest, suggesting first induction was
due to file low temperature per se, rafiier fitan to a given change in temperature.
TheinductionprofileseffileCORgenefamflieswerealseexaminedinnon-
transgenic pin-3 and fiidé mutants to see if decreased levels of polyunsaturated
fatty acids in fire plasmalemma er chloroplast membranes, respectively, would

affect low temperature regulation of file COR genes. The induction profiles were

131

132
indist'nguidublefromfiiouoffiaenen-mutntplant,filusfaihngtosuppert
theideafiratthelipidcempesitionefmembranesinfluencesthesignal

transduction pathway at which filese COR genes are induced.

Introduction

The examination of fire regulation of file COR78, COR47, COR15 and COR6.6
genefamflieswifiarespecttoavarietyofshesseshasshewnfilatatleastene
member of each family is inducible under cendih'ens of low temperature,
drought and he application of ABA (Hajela et al., 1990). For several of filese
genes, ithasbeenshewnfiratfitisinductionisatfilepremoterlevel(Bakeretal.,
1994, Wang et al., 1995, Horvafil et al., 1993, Yamaguchi-Shinezaki and
Slu'nozaki, 1993). However, at file start of firis work, experiment looking more
specifically at the temperature of induction or at the amount of induction at any
given wiperature wae only preliminary (Thomashow et al., 1990). A more
complete characterization of low temperature induction was undertaken to
extend file understanding of COR gene induction and to provide a foundation

for speculation regarding possible mechanisms of signal transduction.

FirstefalLitwasnetknewnifalleffileseCORgeneshavefilesame

temperature induction profile. If filere is only one low temperature signal

. 133 '
operative,thenitwouldbeexpectedthattheCOR geneswouldallrespond

similarly, but a difference in profile might suggest the exishnce of multiple
signals. Also unknown was the temperature or change in temperature required
forinduction. lnduciionisclearat4°C,butthebehavioroffl\egenefamilies
betweenZZ'Cands‘Chadnotbeenexamined. lndeed,thefinalhmperahire
mightnotbefiledetermining factor. Losandcolleagues (1993) haveshoWn that
the low temperatue'mducible gene desA of Syledsocptis PCC6803. is responsive
toachangeintemperatureofgreatertlunS‘Qratherthantoaspeciflc

hmperature.

Subsequent work from this group demonstrated that desA induction is a factor of
membrane fluidity (V igh et al., 1993), which raised the question of whether
Arabidapsis mutants deficient in the sanitation levels ofmembrane fatty acids
would show the same temperature induction profile as wildtype Arsbidapsis.
Plants containing mutations fid 2-3, in which levels of polyunsaturated fatty
acids are increased in extra-chloroplastic membrane lipids (Miquel and Browse,
1992), and pa 6, which have an M levels of polyunsaturated fatty acids in
chloroplast membrane lipids (Browse et al., 1989), were employed to examine

this question.

Finally, although it was known that induction at 4°C is regulahd at the promoter

level, it was not clear whether this was true of regulation at other temperatures.

134
Results

In order to answer these questions, the temperature induction profiles of
endogenous COR genes and of GUS message driven by the promomr of COR15a
were debs-mined. Plants containing the-900/+7B promohr fragmentofCORlSa
fusedtotheGUScoduigregionmakeretalqlflowaegmwnunderstandard
growfichamberconditions. PlantscontainingtheM-BorfidGmutationwere
includedinthesameexperiment. Aftertlueeweeksofgrowtlythetemperature
wasshiftedtwodegreesadayfrom16°Cto6°C Theplantsreceived24hoursof
exposuretoeachtemperatureandapotofeachgenotypewasharvested
immediately before the temperature was lowered again. Thus, the plants
lnrvestedaté‘Clndbeenexpoaedtoeachofthetemperahuesinfitesefles,

culminating with 24 hours at 6°C (Figure 4.1).

Northern analysis of the RNA extracted from these samples, using probes
specific for COR gene families (but not specific for individual members of those
families) or GUS, did not show any clear difference among the temperature
induction profiles of the COR gene families (data not shown) the hansgene
(Figure 4.2) or the find mutants. Message levels were detectable over background
at 10C, gradually W as the temperature was lowered, then appeared to

leveloffatandbelowB‘C.

135

Temperature shift (°C) Temperature drop (°C)

222222222222222222
1919191919191919
16161616161616
141414141414
1212121212
10101010

8 8 8

6 6

4

nunnnnnnn
19 l
16

14

 

 

Hr
10'
8

 

 

V
6

\DQVQUIIFUNHS
‘<

Y
4

Figure 4.1 Two methods of lowering the temperature. For both meflrods, plants
weregrownforabouttlueeweeksat22'C. Inthetemperatureshiftmethod,the
temperaturewasloweredbytwoortlu'eedegreesatatime. Apotofplantswas
harveshdafterexposureto24hoursofagivenhmperature,buthadalso
experienced 24-hours ateachofthehigher temperaturesintheseries. Inthe
temperaturedmpmethod,theplantswerekeptat22°Candfl|entakendhecfly
tooneofthelowertemperaturesintheseries. Thusmostoftheseplants
experiencedordyoneofflielowtemperahrresinthesefiesandexpefierweda
dropinhmperatureofuptolB‘C.

136

1,6
16 14 12 10 8 6°C. fad6 fad2-3
GUS as .

  
 
 

16 14 12 1O 8 6°C16 14 12 10 8 6°C

  

COR15 .
”seer > “an...“ saga» last.“

rRNA feleQFCt‘lvfiuuuulaurzms‘;

Figure 4.2 First temperature shift experiment. RNA isolated from 1,6 plants at
the temperatures listed above the lanes is shown probed for GUS (GUS is fused
to the COR15a promoter in these plants) and COR15. RNA isolated from fatty
acid deficient mutants fad6 and fad2-3 at the given temperatures is only shown
hybridized with a COR15 probe. Ethidium bromide stained rRNA is included to
show differences in loading.

137 g .
Uponrepetitionoftheexperimenttheplantsweresimilarlygrown, butthe

ran'eofthebmperahueswasbroadenedtoirdudembtandardgrowth
hmperature),19'Cand4‘C. 'l'hiswasdoneinordertodeterminewhetherthere
wasagradualincreaaeintheamountofmessageatanybmperaturebelow22°c

or whether 164C was the highest temperature at which induction occurred.

Inaconcurrauexpa'imentdesignedhdetermhuwhethauductionduebflris
gradual lowering of temperature would differ from that following a single drop
in temperature, two sets of the plants containing the COR15a promoter fused to
fireGUScodingregionweregrowntogetherforthreeweeks. Atthispoint,one
setofpohwushifteddownintemperahireuiustdescribed,whilefliesecond
setwasmaintainedat22°C Asthefirstsetreachedeachlowtemperature,apot
fromthissecondsetwasincubatedwithitatthattemperaturefor24hoursand
thenharvesbd. Asanexample,beyondtheinitialthreeweeksofgrowth,the
plants of a given pot from the first set of plants would have experienced 22°C,
19'C, 16°C, 14‘Cand 12°C, each for 24 hours, while corresponding plants from

mesecondaetwouldhaveexperiencedfourdaysof22°Cand24hoursof12°C

Norfllernanalysisoffltissecondsetofexpefimenhconfimedflreresultsfrom
thefirstand showed thatwhetherflietemperaturewasloweredbytwoortluee
degreesatatimeorwas droppedmoredramatically, thelevelofinduction at

each temperature was the comparable (Figure 4.3). Message levels were first

138

1,6 shifted

221916141210 8 6 4°C

 
  

 

COR78
COR47 5;

COR6.6
1,6 dropped

2222L191614 1210 8 640C

COR15

 

Figure 4.3 Second temperature shift experiment. Shown are northems of 1,6
plants that were either shifted gradually down in temperature by two or three
degrees at a time (1,6 shifted) or were dropped from 22°C to the temperature
indicated above each lane (1,6 dropped). Lane 22L in the temperature drop
series refers to a sample that was kept at 22°C for the duration of the experiment
to control for the extended period of incubation at 22°C. The blot of 1,6 which
was shifted down in temperature is shown here probed with COR78, COR47,
COR15 and COR6.6. The rRNA is shown to indicate differences in loading. The
COR6.6 probe used here gave very strong signal relative to the other probes,
which is thought to be due to the probe itself, rather than to the message to
which it hybridizes.

139
dehctable over background at 16°C, gradually increased as fire temperature was

lowered,andagainappearedtoleveloffatandbelow8°C.
Discussion

In fire examination of possible mechanisms ofsignal transduction for the COR
genes, severslquestions aboutthenatureoffireinductionarise. Whatis the
temperature induction profile of the COR genes? Do all fire COR genes follow
the same pattern wifir respect to low temperature? Are firey induced by a
specific temperature, or by fire change in temperature? Is fire temperature
induction profile promoter-based? Since membranes are a primary site of low
hmperature injury and acclimation (Steponkus, 1984), do mutations first alter
firesaturationofmembrane lipids influence the temperature atwhich fireCOR

genesareinduced?

Inductionofmessageforallfouroffiregenefamilieswasfirstseenatlm
although it may increase ata slighfiy higher temperature, since 17°C and 18°C
werenotexamirred. Thelevel ofinductionthengraduallyincreasedasfire
mperature was lowered, and appeared to reach a maximum at 8°C. This was
fiuewhefirerfirephnhreachedagiventemperahrreinimrementsorinasingle
drop,wfingenuappearbberespondingdifferenfiytodifferenttemperamres,

irrespective of the change in temperature. Plants that received fire three degree

140
shiftfrom22tol9°Cslwwednourdmfiomwhilefiwsegivarafirreedegreeslufi

from]9to16°Cdidshowslightinduction,stronglysuggestingfiratfire
Wandnotmeshiftisresponsibleforinduction.

Rwuahomlearwhefirerfiresfighthrducfionatfiehighertemperamresinfire
serias(16,14and12°C)wasduetothetemperatureitselfortofirerateatwhich
massa'eaocumulahs. PerhapsittakaslongatoaccumulateCORgsnemRNA
at14°CfiranatI°C,buttheywouldreachfiresamelevelifgivenenoughtime.
Comparisonoffireexpresdonpafiemoffiretemperatureshiflandtemperahrre
drop experiments suggested again firat the temperature itself is responsible for
thelevelofurducfionandfintifmducfionisdoweratluglerbmperahrres,“
stillappearstoreachsaturatienwithin24hours. Otlrerwise,theplantsfirathad
experiermedeachhemperanneinfireseriesshouldhavehadalugherlevelof

induction firan those which had only received one low temperature.

Thelevehofh'anscriptalsosuggestfiutfireremaybeonemajorincreasein
massagelevelsbetwaenl9and16°Candasecondbetween10and8°C This
may point toward two mechanisms of low temperature induction of COR genes:
onewhich operahsat16-10°Candasecond which involvedininduction stand
below0°C. 'I'lrerecouldbetwosensorsoflowtemperature,twosignal
transduction pathways, or perhaps one transcriptional pafirway and a second

pathway which controls rates of mRNA degradation. Repetition of the

141
experiment and quantification of message levels at each hmperature should help

clarifyfiu’sintriguingpossibility. Furfirerwerkcouldthenbedirectedto

examining signal transducfion in the two unperature ranges separately.

Also, sinoetheinductionofGUS message drivenbytheCORlSa promoter
followsfiremmepathrnasfiratoffireendogenousCORngenefamfly,itwould
appearfiratfirehmperatureinducfionpmfileofCORlSaispromohmbased. It
lnspuviouslybeenshownfintofireraspectsoflowtemperahsreinducfionare
principallypromoter-based infireCOR6.6 gene family (Wangetal, 1995) and
forCOR78 (Hervathetal,1993andYamaguchi-Shinozakiand$hinozald,1994),
soitwould beconsisbnt for their promoters tobehavelikefiratofCORISa. Use
ofthepromotersferfiresecondmemberoffireCORflgenefamily,lti65(Nordin
etal., 1993), alsocalledrd29b (Yamaguchi-Shinozaki and Shinozaki, 1994), and
formembers offireCOR47 gene family'(Welinetal., 1995) inreporter gene

fusionshasnotyetbeendescribed.

Finally, fire lack of alteration in fire induction pattern of the fad mutants suggests
firstbsdkmembranemhrrafionisnotasigrufkantfacbrinfiremducfionoffirese
CORgenes. Thisfailstosupportamodelinwhichachangeinmembrane
fluidity is fire signal prompting low temperature gene induction. However, it is
sfiflpessiblefiratfireconfomafionofaspecificmembnnecomponentorfiuta

microenvironmmt formed in membranes exposed to low temperature might be

142
involved in low “pasture signal transduction. It has been shown first

cakiumaelecfiveionchannelsinfirephsmalemmaofonionceflsaresemifiveto

lowhmparaturasin(DingandPickard,1993).

Cakiumhasbeenshenglyimplicatedmfirereguhfionofcold-respmuivegenes
indfalfaMenroyandDhindsal995). Ithasbsenshownthatcoldshockto5°C
orO'Ccausaariseinfireinh'aoellularconoenfi'afionofcakiumintobacco
(KnightetaL,1992). ThiscouldnotexplaintheinductioneffireCORgenesat
16°CunlesscakiumlevekmArebidopdsincreaaeatahigherhmperamnfiunin
tobaeco,sinoecoldshockt910°Corhigherintobaccodidnotcausearisein
intracellular calcium (Knightet al., 1992). It is more encouraging to note that
calcium influx wasseeninalfalfa protoplasts at temperaturesashighas15°C
(MonroyandDhindsa,1995). Coldshockwifiricewahrelicitedanincreasein
urbaeefluhrcakiumlevehmArabidopsisMghtetaL,1996),butcakiumkvels
inArstidopsisatofires-temperatureshavenotbeenreported. Workwouldneed
to be done to bat whether temperature—responsive channels, particularly
channelsreguhtingcakiumfluxesmouldbefoundinhabidopsisandtobst

whathmperaturescausedfireeffect.

Thepossibilityremains that COR gems inArebidopsismightbeinducible upon
sehrrntonormalgrowfirtemperatures followingheatshock. Ithasbeenseen

first fire COR genes BN1 15 and BN28 in Brassics aapus L. cv. Westar (a spring

143
cultivar), are induced when two-week-old plants are moved from a two-hour

heat shock of 42°C to normal growfir conditions for four hours (Krishna et al.,
1995). However, when Weretilnyketal. (1993) studied BN115 in Brassica napus
chletneuflawintercultivar)fireyallowedfirreeweekoldplantstorecoverat
20°Cfortwotordnelwunaftera42°Cheatslwckbefomhawesfingleaffissue,
anddidnotseeinduction. Perhapsfirediffereneewasduetofirecultivarorto
fireageoftheplantsused. Arabidopsiswillnotsurvivembutisnotharmed
by37°C (DauglrertyetaL,1994). Testing COR gene expressionafterthe15°C
dropfrom37°Ct022°CshouldshowwhefirerornotCORgeneinductionis
responsive to a change in temperature, especially since fire 14°C drop from 22°C
to 8°C was enough to induce a high level of expression of all of fire COR gene

families tested.

Th f'uuling first message first showed an increase over background at 16°C was
unexpected. Initialworkhad orrlyslrowninductionat10°Cor12°Candcolder,
and had shown first fire COR15 gene family was induced at a higher temperature
firan fire other families (Thomashow et al., 1990). Howeva', fire plants used in
firoae experiments had been grown in petri plates and not in pots (Horvafir,
personal communication), which may account for fire difference. There are
differences in humidity, temperature transfer and nutrient availability, to name

a few, between fire two growfir methods.

144
ThehductienefCORgenesatmper-aturesashighaslfiCalsoracallsfire

quesfionefwhefirerfireyareinvolvedinacclimafiontofreeflngorinfire
adiuaherrttometabolismatlownronfreezingbmperatures. Perhapsfireyallow
physiologicalchsngesmatarebeneficialforbofir. Coldacclimationisusually
effectsdat2or4°C,alfiroughwheathasbeenshowntoacclimatesomewhat
undera10°Cday/8°Cniglrthmperatrrreregime(Gustaetal.,1%2),andpotato
hashewnincreasedfresdngtolerameafterexposumtotemperaturesulughu
12°C(Chenandl..i,1980). Ineachcase,firelevelofacclimationachievedatfirese
higher bmperatures was inferior to first reached when lower hmperatures were
used,butwasstillsubstantial. Itisnotknown‘atwhatbmperaturesabove4'C
Wwiflaccflmate.11uscouldbebsted,dfiwughitseemsfikelyfintfiw
temperatures of COR gene induction and of acclimation would overlap enough
to leave fire question of COR gene function open. More questions regarding fire
function of use COR genes could be probably be answered using plants first
failtoexpressallmembersofanyoffireCORgenefamilhs. Antisense
techniques have not yet been successful in producing such plants.

The bmperature induction profiles described here should be useful in designing
futureexperinrentsinwhichfirestandardtemperaturesusedforlow
hmperature induction (24°C) prohibitively slow plant growfir and development
or in which a weaker induction is preferred. I

145
Thuepeofilesmayahogiveduecfiontoexperimenbintendedtoduddatefire

signaloperativeinCORgeneinduction. Whateverfiresignslis,itmustbe
responsivetomperatureassuch,beresponsivetoatemperahrreashighas
IFQandbediffaenfiallyresponsivetodifterenttemperatures. Perhapsfire
signahatdifferenttsmmrahuesaresimfiar,butcomprisedofdifhrent
components. Also,itwouldappearlikelyfirstfiresamesignalisresponsiblefor
inducingallfourCORgenefamilies,sincefireytsllshowfiresamepatternof
induction. This is consistent wifir fire observation first fire core (CCGAC) of a
psemotsrelementshowntobesufficimtforlowbmperatureanddrought
inducibility, fire DRE (Yamaguclri-Shinozski and Slrinozaki, 1994), is found in

n1 COR 3.... promoters examined to date (Thomashow, 1994, Wang et al., 1995).
Materials and Methods

Plant growth and stress treatment

Areh'dopsr’s W (I...) Heyn. was grown in controlled environment chambers at

22°C wifir a 24 hour photoperiod (about 120 umol rrr'2 s4) as preViously described

(Gilmour et al., 1%8). Humidity was not controlled. After about firree weeks of

growfir, plants were exposed to lowered temperatures.

Infireflrstexperimerrtfiretemperamresetfingoffirechamberwuchangedto
16°C,firentol4°C,12°C,10°C,8°Cand6°C,for24hourseach. Atfireerrdofeach

24 hour period, one pot of each genotype was removed and fire plants harvested

146
andfroaeninliquidnitrogen. ThisharvestingwascomplebwifirinlSminutes

orlessofremovalfromfirechsmber. Thechamberstabilized ateachnew
temperature within about 10 minums of being reset (empirical observation). A
Tempscribe'clrartsscorderplaoed infirechamberconfirmed firatfire
tsmpsrahrrechmgedquicklyanddisphyedlifileornofluchrafionduflngeach
amputees. '

lnthesecondexperimentorreset(9pots)of1,6plantswasmovedtoasecond
clumberandmaintainedat22°C,whflefirerestoffirephnhwereleftfirfireflmt
chamber. Aftsr24hours,apotofplantswasharvestedfromeaclrgenotypein
eachchsmber,firehmperaturesetting inthefirstchamberwaschangedto19°C,
andonepotoflfifromfiresecondchamberwasmovedbackintofirefirst
chamber. Afhranofirer24houmfireplantsinfirepotfirathadbeenmovedinto
the19°Cchanrhrandonepotofeachoffiregenotypesofplantsfirathadbeenin
firechamberwereharvesbd. Iffirereweretoofewmtsformenumberof
samplesneededmhalfpotofplantswasharvesbd. Anotherpotfromfire
secondchamberwasmovedintofirefirstandfiretemperaturewaschangedto
16°C. Thissequence was repeated for 14°C,12°C, 10°C, 8°C,6°Cand4°C. Before
thelastpotofplantswasmovedintofire4°Cchamber,halfoffireplantswere
harvestedsofiutfirelmrg-termeffectofhavingmovedfirephnhmtofiresecond

chambercouldbemonitored. Theliglrtlevelsinfiresecondchsmberwere

.147
slighfiyhigherfiranfioaeinfirefirstandmeplantsfirsthadbeenmoved

gradually became anthocysnic.

RNAextractienandfr-actionation

Total RNA was extracted essentially as described (Gilmour et al., 1988) with a
fewmodificationa. Frozenpulverizedtissuewasextrachdwifirequalvolumes
of phenol/chloroform/isopropyl alcohol (25/ 24/ 1, v/v/v) and extraction buffer
(1% w/v triisepropylnaphfirelene sulfonic acid, 6% w/v p-anrinosalicylic acid,
100 uM Tris-HCl pH 7.6, 50 mM BGTA pH 8, 125 mM NaCl, 1% SDS, 10 mM
D'l'l')osrioe. 'I'lrisrnixturewasfurfirerhomogenizedinfiretubeusinga
rm’ “rm, centrifuged (10,000 rpm in fire SA600 rotor), and fire
supernatsrrtextracted again with phenol/chloroform/isopropyl alcohol. This
secondsetoftubesandaflsubsequenttubmandsolufionsexceptefinnolwere
as. free of RNase by treatment wifir DEPC (Sambrook et al., 1989). Nucleic
acidsweseprecipitatedwifircold95$ Mmuspendedinlmldouble
disfilledwater,fiansferredtoamiaefugehrbeandprecipitatedonioeforan
hour with 1/4 volume 10 M BC]. The pellets were precipitated again wifir 95%
EOHandresuspendedin200uldoubledistilledwater. momma

measured to estimate fire concentration of RNA.

Forfractionation,RNA(5to40ug)wasdrieddownandresuspendedin

formaldehyde loading buffer containing EtBr (about 100 ng/ ml), incubated at

148
68°C for 10—15 minuhs and fractionated on denaturing formaldehyde agarose

gals(Sambrooketal.,1%9). TheRNAwasfirenfiansferred (Sambrooketal.,
1”)»“agnaN’l'membranes(MSI)using10x$PBmadeaocordingto
instructions provided by Schleiclrer and Schuell. This recipe for $PE is slighfiy
differentfromfiratdescribedbySambrooketaLO989). RNAwasUVcross-
linked (Stratalinker, Stratagene) to the filters.

For norfirern hybridizations, blots were prehybridized, hybridized and washed
in a Robbins Scientific Model 400 hybridization oven according to standard
mefirods (Ausubel et al., 1987) wifir some modifications. Blots were incubated
with psehybridization solution (50% formamide, 5x SSPE, 50 mM Potassium
phoaphate pH I, 5x Denhardt’s solution, 0.5% 505, 100 ng/ml sheared,
denaturedfishsperm DNA) at42°Cforfiueehourstoovernightandin
hybridization solution (50% formamide, 5x SSPE, 50 mM potassium phosphate
pH 6.5, 1x Denhardt' s solution, 0.5% SDS, 5% dextran sulfab, 100 ug/ ml sheared
derutured fislrsperm DNA) at42°Covernight Blotswere rinsed and firen given
two 30-minute washes at room temperature in 2x SSPE/0.5% SDS, firen washed

two to firree times, for 15 minutes each time, wifir 0.1x SSPB/0.5% SDS at 50°C.

Probesweremadefromgel—isolatedfrsgmentslabeledwifir”Pbyrandom
priming (Feinberg and Vogelstein, 1%3). An Eco RI fragmentfrom pHH67 was

used to visualize COR15 message, a Sac I/ Hind 1]] fragment from pBI101.3

149
Oefleraon, 1N7)servedasaGUSprobe,andEmRIfragmentsfromfirecDNA

clones pHH7.2 (Gilmour et al.,1992): pI-IH28 (Hajela et al., 1990) and pHHZ9
(Gilmour et al., 1992) were used to detect message from COR47, COR78, and

COR6.6, respectively.
Literature Cited

AusubelFM,BrentR,KingstonRE, MooreDD,SeidmanIG,SmifirIA,Str-uhll(
(1%?) Current Protocols in Plant Molecular Biology. Greene Publishing
Associates and Wiley Interscience, New York, Clrichester, Brisbane, Toronto,

Singapore

Baker 55, Wilhelm KS, Thomashow MF (1994) The 5’-region of Arabidopsis
W cor15a has cis-acting elements first confer cold-, drought- and ABA-

regulated gene expression. Plant Mol. Biol. 24: 701-713

Browse], Kunst L, Anderson S, Hugly S, Somerville C (1%9) A mutant of
Arsbidopsis deficient in fire chloroplast 16:1 / 18:1 dessturase. PlantPhysiol.
90:522-529 '

Chen H-H, Li PH (1980) Characteristics of cold acclimation and deacclimation in
tuber-bearing Solarium species. Plant Physiol. 65:1146-1148

Daughtery CI, Rooney MF, Paul A-L, de Vetten N, Vega-Palas MA, Lu G, Gurley
WB, Ferl RI (1994) Environmental stress and gene regulation. In, EM
Meyerowitz, CR Somerville, eds, Arabidapsis. Cold Spring Harbor Laboratory
Press,ColdSpringHarbor,NewYork, pp769-806

Ding JP, PiCkard BC (1993) Modulation of mechanosensitive calcium-selective
cation channels by temperature. The Plant Iourn. 3:713-720

Feinberg AP, Vogelstein B (1%3) A technique for radiolabeling restriction
endorurclease fragments m high specific activity. Anal. Biochem. 132:6-13

' 150
Gusts LV, Fowler DB, Tyler NI (1982) Factors influencing hardening and
survivalirrwinterwheat. In, PH Li, ASaksi, eds, Plantcold hardiness and
freezing stress, mechanisms and crop implications, v. 2. Academic Press, New
York, London, Paris, San Diego, San Francisco, San Paulo, Sydney, Tokyo,
Toronto, pp 23-40

Haiela RK, Horvath DP, Gflmour SI, Thomashow MF.(1990) Molecular cloning
and expression of car (cud-regulated) genes in Arabidopsis thaliana. Plant Physiol.
93:1246-1252

Horvath DP, McLsrney BK, Thomashow MF (1993) Regulation of Arabidopsis
We 1.. (Heyn) c0078 in response to low hmperature. Plant Physiol. 103:1047-
1053

GilmourSI, Hajela RK, Thomashow MF (1%8) Coldacclinration inArabidopsis
Wises. Plant Physiol. 87:745-750

Gilmour SI, Artus NN, Thomashow MF (1992) cDNA sequence analysis and
expression of two cold-regulated genes of Arabidopsis thaliaua. Plant Mol. Biol.
18. 13-21

IefteraenRAflMAssayingchimericgenesinplants: fireGUSgenefusion
syshm. PlantMoLBiolRep.5:387-405

Knight MR, Campbell AK, Snrifir SM, Trewavas A] (1991) Transgeru‘c plant
aequofinreporbfireeffechoftouchandcold-shockaruiehdtononcytophsmic
calcium. Nature 352:524-526

Knight H, Trewavas A], Knight MR (1996) Cold calcium signaling ln
Arabidopsis involves two cellular pools and a change m calcium signature after
acclimation. The Plant Cell 8: 489-503

Krishna P, Sacco M, Clrerutti IF, Hill S (1995) Cold-induced accumulation of
hap90 transcript in W sepals. Plant Physiol. 107:915-923

Miquel M, Browae I (1992) W mutants deficient in polyunsaturated
fatty acid synthesis. Thelour. Biol. Chem. 267: 1502-1509

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

Ssnrbrook 1, mod. BF, Maniatis r (1939) Molecular Cloning: A lalmm manual,
second edition. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York

151 '

SteponkusPLflfloRoleoftheplasmamembraneinfieezinginjuryandcold
acclimation. ArmRev.PlantPhysiol.35:543—584

ThonashowMF(l994)Ambidcpsisuamodelforshrdyingmeclnnismsofplant
cold tolerance. In, EM Meyerowitz, CR Somerville, eds., Arsbidopsis. Cold
SprmgHarborIaboratoryPress,ColdSpringHarbor,NewYork,pp807-834

Thomashow MP, Gilmour S], Haiela R, Horvath D, Lin C, Guo W (1990) Studies
on cold acclimation in Arabidopsis finalised. In, AB Bennett, SD O'Neill, eds.,
Who'd Biotedualogy. Wiley-Lisa, New York, pp 305-314

WangHDaflaRGeorgesF, LoewenM,CutlerA(1995) Promotersfromh’nl
and (116.6, two homologous Arsbidapsis W genes: transcriptional regulation
and gene expression induced by low temperature, ABA, osmoticum and
dehydration. Plant Mol. Biol. 28:605-617

Welin BV, Olson A, Palva ET (1995) Structure and organization of two closely
related low-temperahrre-induced Mali-like genes in Arabidopsis Minna L.
Heynh. Plant Mol. Biol. 29:391-395

Yunaguchi-Shinonki K. Shinozaki K (1993) Characterization of the expression
of a desiccation-responsive rd29 gene of Arsbidopsis Minna and the analysis of
ibpromoterintransgenicplants. Mol.Gen.Genet. 236531-340

Yan'raguchi-Shinozald K, Shinozald K (1994) A novel cis-acting element in an
Arsbr’dopsis gene is involved in responsiveness to drought, low-hmperature, or
high-salt stress. The Plant Cell 6:251-264

APPEND IX

APPENDIX

152

SEQUENCE ALIMNT OUTPUT FRO! PHDIOC AND PHDthroador'

PHDsec: Abbreviations

NAXHONIALIGNNENT HEADER:

ID
STRID
PIDE
NBIN
LALI
NGAP
LGAP
LSEQZ
ACCNUN
NAME

ABBREVIATIONS POR SUNHARY

identifier of aligned (honologous) protein

903 identifier (only for known structures)

percentage of pairwise sequence identity
percentage of weighted similarity

number of residues aligned

nunber of insertions and deletions (indels)
number of residues in all indels

length of aligned sequence
SwissProt accession nunber

one-line description of aligned protein

COR15a-

NAXHOIIALIGNNINT HEADER: SUMMARY
STRID IDI NBIN LALI NGAP LGAP LENZ

ID
b1a1_bacce
h1g_strpu
tola_ecoli
edc8_dauca
tpna_xenla
dnak_bacne
vg2 4_bp-1 S
tpna_rante
cyli_bovin
le22_goshi
cyli_hunan
vinc_hunan
tpna_brare
subv;bacsu
lnb1_hunan
nodu_drone
rsp3_chlre
tola_haein
drpf_crapl
dyhc_yeast
sp2b_bacsu

38
34
33
32

24
38
35
33
24
27

68

HNHHHNUNNUNHNNHNNHOON

11
0
0
3

18

14
8

18
9
7

16

15

18

20

10

QM.H~IO

306
217
421
555
284
605
132
284
667
302
598
1065
284
806
“1786
544
516
372
201
4000
332

ACCNUN
910424
907796
919934
920075
001173
905646
005231
913105
935662
913940
935663
918206
913104
929141
P07942
913469
912759
P44678
923283
936022
937575

NAN!

EETA-LACTANASE PRECURSOR,
HISTONE Hl-GANNA, LATE.
TOLA PROTEIN.

ENBRYONIC PROTEIN 0C8 (CL
TROPOHYSIN ALPHA CHAIN,
DNAK PROTEIN (HEAT SHOCK
GENE 24 PROTEIN (6924).
TROPONYSIN ALPHA CHAIN,
CYCLIN I.

LATE ENDRYOGINESIS ABUNDA
CYCLIN (PRAGNENT).
VINCULIN.

TROPONYSIN ALPHA CHAIN,
MINOR EXTRACELLULAR PROTE
LANININ BETArl CHAIN PREC
DNA-BINDNG PROTEIN NODUL
RADIAL SPOKE PROTEIN 3.
TOLA PROTEIN.
DESSICATION-RELATED PROTE
DYNEIN HEAVY CHAIN, CYTOS
STAGE II SPORULATION PROT

[llnbesur CKNRISMHI cxuntisuusxi]

NAXNONIALIGNNINT:

l
predict_h191 AIEDGIILDD
bla1_hacce ..........
h1g_strpu ..........
tola_ecoli ADAKAKAIAD
edc8_dauca AIVBRINTDY
tpna_xenla .KBTIDILDK
dnak_bacle .......IDK
vg24_bpn15 ......TIDA
tpna_brare HIBTIDILDK
cyli_bovin AKKDTISTdd
le29_goshi ........DV
cyli_hunan .KYTKYTKKD
vinc_hunan PIGIBQIRGA
tpna_brare .KAIEDBLDK
subv;bacsu ....GNSLNN
lnb1_hunan ......LLBB
nodu_drone BIAAGLIDDB
rsp3_chlre AKHIAILOGK
tola_haein ..........
drpt_crapl ..........
dyhc_yeast .......Vll
sp2b_bacsu ..GrGLGLlA

51
predict_h191 Lcusarxaaa
bla14bacce LGIIAIAKVR
h1g_strpu VAKPAKKAAA
tola_ecoli LKKKAIAAIA
edc8_dauca IAQKABIAK!
tpnn_xenla KLBIABKAAD
dnak_bacne lnNBADQLV!
vg24_bpn15 PDLEEDDDDE
tpna_rante KLBIABKAAD
cyli_bovin KGKKDSK...
1e29_goshi rsrnruraar
cyli_hunan ISTDASSGDS
vinc_hunan LAKUVATALQ
tpna_brare KLBIAEKAAD
subv;bacsu ersAKVHGY
lnb1_hunan 1LTSIBSBIA
nodu_drone VDBSDDDBBA
rsp3_chlre IAQAEEAANA
tola_haein KAEAEAKAKA
drpt_crapl ASHKANGAAR
dyhc_yeast AIBBIKKILK
sp2b_bacsu OTSGAIKOAA

LNEATKKASD
LOQNSTKKLD
.TKKTKKVKK
AKAAEEAAKK
AXDKGREGGD
YSEALKDAOE
NGIVNVRAKD
EREEVKKRSD
YSEALKDAQE
SKDAKKGKKE
KNAAKGKSSE
TKKNAKKSSD
LAEARkkERD
YSEALKDAOE
PDNATSTALD
AKRASKSATD
AEEDEEYNSD
ELEAVRRRPT
.EEAKAKAAE
..ASQSOGRQ
LNKTLSKKST
LNISGNKEAS

YVIIXBKIAA
YSD.......

KPAKKAAKPA
AAAEARKKAA
KAAOKABETK
ESERGHKVIE
TTEKTLKDLE
LVDEYSLQVC
ESBRGHKVIE
..KDNKKKDA
RGARKAEEAK
KDERdeKDT
NLQTKTNRap
ESERGNKVIE
NKEDDVKALN
ASEETLFNAS
PVEKPVSKKS
KHEADKABAA
AAEAKAdeE
BTNDKAKETY
VOEekRKEVV
OTEGTYKTIA

153

PVTDKTKEAL
EVITYTKEDL
PAAKKAKKPA
AAADAKKKAE
VAAOKAEEAK
KLELSDKKAT
LGTNkiKSST
DVTVELKPLL
KLELAEKKAT
SKKDKKKDAK
NRQATTEKAR
AESEDSKDAK
DILRSLGEIS
KLELAEKKAT
WANSEGVVAV
vtADNVKEAL
DEEDDDDDEL
FVLRELKPAV
IAAQKAXOEA
QVSENAEDAK
ELTEKEKEAR
ARASLEDSLG

uxaarrarcx
uracurnrnx
Kxaaxraxxa
rraarxaxar
rxacrytaox
unasrLorIo
cxvrrarvrx
nrraxarnrr
NRaiELQEIQ
xxnarsrnar
rvvarxarca
KKYPESTDTE
axaavurrcx
NRanELOEIQ
uxrvrnvrac
QRISBLERNV
exasrxszru
rxanararaa
axaarsaxnx
NAAS....GK
Hxsronrrrr
voacxrsurx

ADGEKAKDYV
VDYSP....V
AKKPAAKKRA
AEAAKAAAEA
EKAKHAKDTT
DAEGDVASlr
GLSDDEIDRN
KLGQKAREAV
DAEADVASlr
KDAASDAESG
ELADSAKENI
KDSKKVKKNV
ALTSKLADLR
DAEGDVASlr
TSNGNSGPNG
EEAEKAQ.VA
EPGEVSK...

ASADAVE...

EAKAKLEAEA
KKPSETTDSL
8..TLDKHLH
SOTAKAGD..

89
AGIAKDATK
IGGPKBYEK
AKPAK....
AEKKAAAEK
AGEAKDTT.
LKEAKH...
ANEAKDALK
ATKPK....
LKEAKH...
SGDSKDAKK
AEETKKKNE
SGDAKDARN
IEOAO....
LKEAKH...
IGEAKD...
EELKRKAAO
RGIPK....
AEEQK....
ADQA.....
AGELKDKTQ
ILEAQRGVK
GAE......

IN IS! IOIHAI [snall letters nark an insertion]

50
VEKNSETADT
TEKHVDTGNT

AEKAIKOADO
....SEGADE
....AAAAEL
KAKAVAEAKA
KHKTSEATDS
EONESERKDE
.....T8ADK

Fillssnrz «DOIEIStss

MAXHOM.ALIGNMINT HEADER:

ID
r1a0_netva
crtc_caeel
c1pl_lacla
ifz_bacst
noes_pig
nysc_chick
hs71_leina
tebb_oxyno
noes_nouse
noes_hunen
htr2_natph
ynp9_caeel

MAXHOM.ALLIGNMENT:

38 33 64 1
35 33 68 2
33 21 79 3
33 35 76 2
32 36 78 2
31 26 81 2
31 22 81 3
31 24 85 2
31 36 78 2
31 35 78 2
30 27 79 3
30 22 90 3

SUMMARY
STRID IDI'NSIM LALI NGAP LGAP LEN2

154

14 336
8 395
43 763
11 741
2 576
33 1102
46 634
13 385
2 576
2 576
19 534
22 690

ACCNUM NAME

915826
927798
006716
904766
926042
929616
912076
916458
926041
926038
942259
934562

ACIDIC RIBOSOMAL PROTEIN
CALRETICULIN PRECURSOR.
ATP-DEPENDENT PROTEASE AT
INITIATION [ACTOR IP-Z-
NOESIN (MEMBRANE-ORGANIEI
MYOSIN HEAVY CHAIN,CARDI
MITOCHONDRIAL HEAT SHOCK
SUBUNIT).

MOESIN (MEMBRANE-ORGANIEI
MOESIN (MEMBRANE-ORGANIEI
PROTEIN II) (MPP-II).
HYPOTHETICAL 79.2 KB PROT

1

predict_h208 vxsocarrnc

rla0_netva
crtc_caeel
clpl_lacla
i£2_bacst

noes_pig

nysc_chick
hs71 lei-a
tebb oxyno
:noes nouse
noes hunan
htrZ natph
ynp9_caeel

......ITDS
......VIDR
..SDEEIPDD
.......LKO
........TK
.....NVIRV
..RHTALQAA
..... ..LKO
.......LKD
.......VKA
LKSENEKLLA

51

predict_h208 ASIIAKKALD

r1a0_netva
crtc_caeel
clpl_lacla
it2_bacst
noes_pig
nysc_chick
hs71_leinn
tebb_oxyno
noes_nouse
noes_hunan
htr2_natph
ynp9_caee1

ATEEAPKAET
AIEEARKKAE
YRGSFEENIK
AKKKGKEPA.

LNEATKKASD
DAKAVSVESA
VEEAEAHAAE
ENEIQKPAOK
VKEAAKPAN.

IEEQTKKAOQ
LDEMTRLMND
VNEPTAAALA
INKTVKGDN.

IEEQTKKAQQ
IEEQTKKAQQ
LAEETKAATO
KNEOIKKKSH

YVTEKFKEAG
KKEEKKEEAA
EEKEAKKDDD
QLVEEVKAAG
....KGKKOA
KASdQKKTQE
LLREQYEEEQ
RVQEKAKCEL
FVQEKGKDAL
QASdQKKTQE
QAsdQKKTQE
QSTDAOADAA
LASSRDKAE:

PVTDKTKEAL
PITEKTADII
.....TPDKL
PCRRRKKNPL
....KKKGAA
ELEEQTRRAL
LTTQKTKLOS
YGMDKTKDSL
.LVDISKVAD
ELEEQTRRAL
ELEEQTRRAL
tVQDRTQTTV
PVODETRKAI

IKAAIIVIGK
9AA.......

EEEKEEEEGH
nIAGEEVKBL
APAAKOVPQP
QLALIHAILT
EAKAElsKEN
SSAHefITAN
NKAADHTDGA
QLASEHABLT
QLALEHABLT
QKATTMVEDM
eKTLKVLKSI

I! I3! IORHAI [snall letters nark an

ancrxrxnrr
AGDEALDDDL
xrvrxrxx..

nvcrsqxrav
xcxrrxnrra
arroraxnao
rncrrvnora
IadLALSDYI
ascxxcxvna
ELEQERKRAO
rLrornanQ
Dbrnrrsoov
rxrrxsxvrr

90
AEEAKNATKS
DE........
ADIIK.....
AKKEKELPK.
ARISQ.....
AEVAOWRTK.
ADGAQH....
KVKGGAKBK.
ARISQ.....
ARISQ.....
AATSEQ....
LTESEKAHTT

insertion]
50
VERTIEANET
REQISSSAVV
.EKADEETRK
VEGllEAGTQ
QQQEKKAPOA
SEAEKLAKeq
eTKSKNALAH
LEEPRKTSGI
GIVKASAseG
SEABKLAKeq
SEAEKLAKeq
vEDTVDALee
TELEQOADQT

155

mm: [TOPITS]

TOPITS ALIGNMENTS HEADER: PARAMETERS

smin - -1.00 : minimal value of alignment metric

smex - 2.00 : maximal value of alignment metric

go - 2 : gap open penalty

ge - 0.2 : gap elongation penalty

lenl - 89 : length of search sequence, i.e., your protein

TOPITS ALIGNMENTS HEADER: ABBREVIATIONS

RANK : rank in alignment list, sorted according to z-score
EAL! : alignment score

LALI : length of alignment

IDIL : number of residues inserted

NDIL : number of insertions

IALI : alignment score: note: hits with z>3 more reliable
PIDI ' : percentage of pairwise sequence identity

LINZ : length of aligned protein structure

102 : PD! identifier of aligned structure

IAHIZ : name of aligned protein structure

TOP!!! ALIGIHIITS HIADIR: ACCURACY

rested on 80 proteins, 109113 found the
correct remote homologue in about 30% of
the cases, detection accuracy was higher
for higher z-scores (ZALI):

ZALI>0 : 1st hit correct in 33% of cases
ZALI>3 : 1st hit correct in 50% of cases
ZALI>3.5 :

1st hit correct in 60% of cases

llflDtluneeckmr: ‘CCICLSaus
rorrrs streams-rs aransa: summary
sasx EAL! LALI IDIL sort rat: 0100
1 01.00 09 5 1 2.37 22
2 77.17 00 19 5 2.10 23
3 75.50 00 19. 5 2.15 23
4 75.50 00 19 5 2.15 23
5 75.35 07 3 2 2.09 15
5 75.00 00 4 2 2.00 15
7 75.07 00 2 1 2.07 15
0 74.50 00 4 2 2.05 15
9 74.55 09 9 3 2.05 12
10 73.90 05 4 1 2.02 13
11 71.50 00 17 3 1.90 17
12 71.55 09 14 3 1.90 17
13 71.40 09 4 1 1.09 17
14 71.40 00 35 3 1.09 15
15 71.03 05 14 3 1.00 20
15 70.03 00 20 3 1.07 15
17 70.53 03 12 3 1.05 15
10 70.47 00 22 5 1.05 19
19 70.42 09 24 5 1.05 10
20 70.10 09 20 4 1.03 10
lHlDtflrreeuisru cxn115110
routs streams-r uranra: smear
m1: m1 mm 1021. 14021. 211.1 0100 1.0112
1 70.07 90 30 4 2.15 20 153
2 77.55 09 11 2 2.09 10 421
3 77.22 09 11 3 2.07 21 154
4 77.20 09 11 4 2.07 24 154
5 75.42 09 11 4 2.04 24 154
5 75.42 90 10 2 1.99 22 421
7 75.20 90 17 4 1.90 12 504
0 74.57 90 15 2 1.95 17 190
9 74.52 09 17 3 1.95 11 144
10 74.23 90 20 4 1.93 14 170
11 73.07 09 3o 7 1.92 30 153
12 73.03 90 27 5 1.92 19 500
13 73.03 90 27 5 1.92 19 500
14 73.77 90 35 7 1.91 22 197
15 73.50 09 9 2 1.90_ 11 139
15 72.95 90 27 5 1.07 19 500
17 72.95 90 27 5 1.07 19 500
10 72.42 90 5 3 1.05 19 725
19 72.30 00 0 1. 1.04 11 304
20 72.10 90 37 5 1.04 24 377

156

L882 102
153 laep
153 ngm
153 21h2
153 llh3
144 llpe
421 lses
139 list
421 lsry
146 luas
157 lbcf
197 lcol
289 lbab
384 lhtm
170 lfha

NAHBZ

LIPOPHORIN III

_ID: 1:

HEMOGLOBIN (AOUO,MET)
HEMOGLOBIN (CYANO,MET)
LIPOPROTEIN-*E3 (ILDLS RE
YL-TRNA SYNTHETASE (E.C.6
LIPOPROTEIN-*E4 (ILDLS RE
YL-TRNA SYNTHETASE (E.C.6
TERIAL ASPARTATE RECEPTOR
CTERIOPERRITIN (CYTOCHROM
LICIN *A.(C-TERMINAL DOMA
MOGLOBIN THIONVILLE ALPHA
MAGGLUTININ ECTODOMAIN (8
RRITIN (H-CHAIN) MUTANT (

 

127
118
147
524
148
154

102

laep
lses
2spo
2mge

lsry
ldlc

llpe
ltha

1cpc

lcol
11e4

lcpc

lvsg
lhtm
ldsb

2ccy
thr
2hbg
lddt
losa

2spo

TOCHROME $C(PRIME)
OHEMERYTHRIN
MOGLOBIN (DEOXY)

LMODULIN

NAMIZ

LIPOPHORIN III

YL-TRNA SYNTHETASE (E.C.6
GLOBIN (MET) MUTANT‘NITH
GLOBIN (MET) MUTANT WITH
MYOGLOBIN MUTANT WITH INI
YL-TRNA SYNTHETASE (E.C.6
TAPENDOTOXIN CYRIIIA.(BT1
GANESE SUPEROXIDE DISMUTA
LIPOPROTEIN-*E3 (/LDL$ RE
RRITIN (H-CHAIN) MUTANT (
OGLOBIN (PERRIC IRON-M
PHYCOCYANIN

PHYCOCYANIN

LICIN 1”AND-TERMINAL DOMA
OLIPOPROTEIN-*E4 (ILDLS R
PHYCOCYANIN

PHYCOCYANIN

RIANT SUREACE GLYCOPROTEI
MAGGLUTININ ECTODOMAIN (3
EA (DISULPIDE BOND FORMAT

PHTHERIA TOXIN (DIMERIC)

OGLOEIN (MET) MUTANT WITH