Molecular analysis of an abscisic acid deficient mutant Viviparous14 in maize

MISSING IMAGE

Material Information

Title:
Molecular analysis of an abscisic acid deficient mutant Viviparous14 in maize
Physical Description:
Book
Language:
English
Creator:
Tan, Bao-Cai
Publication Date:

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
aleph - 28638526
oclc - 38827141
System ID:
AA00026806:00001


This item is only available as the following downloads:


Full Text










MOLECULAR ANALYSIS OF AN ABSCISIC ACID DEFICIENT MUTANT,

VMPAROUS14 OF MAIZE











BY

BAO-CAI TAN

















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1997

















ACKNOWLEDGEMENTS





I am deeply grateful to Dr. Donald R. McCarty, chairman of my supervisory

committee for his valuable advice, guidance and encouragement throughout the course of

this study. I want to thank members of my graduate committee Drs. L. Curtis Hannah,

Karen E. Koch, William B. Gurley, and Nigel Richards for their valuable suggestions to

the completion of this study. Specially, I want to thank Dr. Kenneth Cline for his help in

localizing VIVIPAROUS14 in choloroplasts. I wish to thank Dr. Phil Becraft and Dr.

Ralph Henry for valuable discussions. I also wish to thank Mr. Xian-Yue Ma, Dr. Chien-

Yuan Kao, Dr. Masaharu Suzuki, Dr. Shailesh Lal and Dr. Joseph Cicero for their help. I

want to thank Ms. Janine Shaw, Mr. Mike McCaffery, Mr. Wayne Avigne, and Mr. Dale

Haskell for allowing me to frequently use their lab equipment. Finally, I owe thanks to my

wife, Wen-Tao Deng, for her constant support on every aspect during this study.













ii














TABLE OF CONTENTS




ACKNOWLEDGMENTS ........................................................................................... ii

A B STR A C T ................................................................................................................... iv

CHAPTERS

1. REVIEW OF LITERATURE............................................................................ 1

ABA Biosynthesis in Fungi and Plants ............................................................. 1
Major Functions of ABA in Higher Plants ...................................... ..... 11
The ABA Related Mutants in Higher Plants............................... ......... 23
Summary and Future Perspectives ..................................... ... ............ 27

2. MOLECULAR ANALYSIS OF VIVIPAROUS14, A DEVELOPMENTALLY
SPECIFIC ABSCISIC ACID BIOSYNTHETIC MUTANTS OF MAIZE ............... 31

Introduction ................................................. ................................................ 31
Materials and Methods ........................................................ 34
R esults. .............................................................................................................. 37
D iscussion .......................................................................................................... 58

3. VIVIPAROUS14 ENCODES A 9-CIS NEOXANTHIN/9-CIS
VIOLAXANTHIN DIOXYGENASE OF ABA BIOSYNTHESIS IN MAIZE......... 66

Introduction ............................................................................. ....................... 66
Materials and Methods. ........................................................ 68
R esults ............................................................................................................... 70
D iscussion .......................................................................................................... 77

4. LOCALIZATION OF VP 14, A SPECIFIC 9-CIS EPOXY CAROTENOID
DIOXYGENASE INVOLVED IN ABA BIOSYNTHESIS .................................. 82

Introduction ................................................. ................................................ 82
Materials and Methods. ................................................................................ 84
R esults. .............................................................................................................. 86
D iscussion................................................................................................. 95



iii









5. A FUNCTIONAL DUPLICATE OF VIVIPAROUS14 CONFERS A
DISTINCTIVE TISSUE-SPECIFIC EXPRESSION PATTERN IN MAIZE........... 98

Introduction ................................................. ................................................ 98
Materials and Methods ........................................................ 99
R esults. ............................................................................................................ 100
D iscussion..................................................................................................... 109

6. SUMMARY AND CONCLUSIONS ..................................... 115

LIST OF REFERENCES............................... 116

BIOGRAPHICAL SKETCH .................................................................................... 127






































iv















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements of the Degree of Doctor of Philosophy

MOLECULAR ANALYSIS OF AN ABSCISIC ACID DEFICIENT
MUTANT VIVIPAROUS14 IN MAIZE

By

BAO-CAI TAN

December 1997

Chairman: Donald R. McCarty
Major Department: Plant Molecular and Cellular Biology

Abscisic acid (ABA), a key regulator of seed maturation, dormancy and stress

responses throughout plant development, is proposed to be synthesized by oxidative

cleavage of 9-cis epoxy-carotenoids. Carotenoid cleavage is the first committed and

presumed regulatory step in ABA biosynthesis. In this study, a new ABA deficient mutant

viviparous14 (vp]4) was identified in maize. Although vpl4 plants were fully viable and

non-wilting in field, detached leaves of non-stressed mutant seedlings grown in the

greenhouse showed markedly higher rates of water loss than the wild type. Mutant

embryos exhibited normal sensitivity to exogenous ABA, and the ABA content of

developing mutant embryos was 70% lower than the wild type, indicating a defect in ABA

biosynthesis. The HPLC analysis of the vpl4 embryos detected no significant changes in

the epoxy-carotenoid precursors of ABA, and the cell extracts of vpl4 embryos efficiently









converted xanthoxin to ABA, suggesting that vp14 is blocked in the cleavage of 9-cis

epoxy carotenoids.

The VpJ4 gene was molecularly cloned by transposon tagging. The Vpl4 mRNA

was detected in embryos and seedling roots, but not in nonstressed leaves. Southern blot

analysis and cloning of related sequences indicated that VpJ4 belongs to a gene family.

Several related gene sequences were cloned and the DNA sequenced. A Vpl4-like

duplicate was mapped to chromosome 5S and found to potentially encode a protein with

93% amino acid identity to VP14, suggesting that it may have a function equivalent to

VP14. The VpJ4 mRNA and related transcripts were induced by water stress in leaves.

The VP14 amino acid sequence showed significant similarity to lignostilbene dioxygenase

of bacteria and RPE65, a protein found in the mammalian retinal pigment epithelium.

Purified recombinant VP 14 as expressed in E. col specifically cleaves 9'-cis-neoxanthin/9-

cis-violaxanthin to produce xanthoxin and a C2s allenic/epoxy apo-aldehyde. The

implications of this demonstration of specific enzymatic cleavage of carotenoids for

understanding the mechanism of vitamin A biosynthesis in mammalians are discussed.

Evidence that the VP14 protein is localized in the chloroplast of plant cells was obtained

by an in vitro chloroplast protein import assay. Precursor VP14 was imported into a

soluble fraction of chloroplast lysate and processed. A significant amount of imported

VP14 was also found associated with thylakoid membrane with a topology of facing

stroma. Furthermore, VP14 was subject to a fast turnover with a half life of about 30

minutes, suggesting that degradation may play a role in controlling the enzyme abundance.






vi














CHAPTER 1
REVIEW OF LITERATURE


The discovery of the plant hormone, abscisic acid (ABA), was associated with

two independent lines of research, one involved isolation of substances which accelerate

leaf abscission in cotton by Addicott's group, the other included a search for substances

which cause bud dormancy in woody plants by Wareing's group. The search for an

abscission accelerator from young cotton fruits identified a compound "abscisin I'I"

(Ohkuma et al. 1963). The search for a dormancy inducer from Betula pubescens led to

the discovery of a substance, named "dormin" (Eagles and Wareing, 1964). Soon

thereafter, the two substances were found to be chemically identical in structure

(Cornforth et al. 1965). Since then, a uniform name was given as abscisic acid. The

naturally occurring ABA is the S-enantiomer. Whereas the R-enantiomer has been

reported to be biologically active in most cases, it is inactive in regulating stomata

aperture (Addicott 1983). The biologically active configuration of ABA is 9-cis ABA,

and all trans ABA is biologically inactive.



ABA Synthetic Pathways in Fungi and Plants

Although its structure was elucidated more than 30 years ago (Cornforth et al.

1965), the ABA biosynthetic pathway had remained elusive until the late 1980's. Studies

of the ABA biosynthetic pathway in higher plants have encountered great difficulties

compared with the similar studies on other plant hormones such as ethylene and


1






2


cytokinins. ABA exists at very low concentrations (107 10-8 M), even in water stressed

tissues and maturing embryos where it is around 10'6 M. Thus, analysis of ABA was

technically challenging. Furthermore, the major obstacle is the poor incorporation of

isotope labeled precursors such as '4C-MVA (mevalonic acid) and 14CO2 into the

intermediates and ABA such that one of the most important approaches to studying

metabolic pathways has been of little use in studying ABA biosynthesis (Zeevaart and

Creelman 1988, Walton and Li 1995). The poor incorporation of '4C-MVA into ABA led

to a number of assumptions about precursor pools and localization of ABA synthesis.

Milborrow (1983) suggested that ABA may be synthesized in the chloroplast; thus, the

fed precursors have to cross two layers of chloroplast membrane to reach ABA synthetic

enzymes. In fact, currently it is widely believed that ABA is synthesized in plastids.

Another explanation is the existence of a relatively large pool of precursors such that the

incorporated labeled precursor was largely diluted. Alternatively, the pool used for ABA

biosynthesis may be separated from the one reached by conventional feeding

experiments. Some of those assumptions turned out to be correct as now we know that

ABA is derived from an oxidative cleavage of carotenoids which exists abundantly and

exclusively in plastids of higher plants.


The Direct C j Pathway of ABA Synthesis in Fungi

The discovery in 1977 that a rose pathogen, the fungus Cercospora rosicola,

produces and excretes relatively large quantities of the naturally occurring enantiomer of

ABA into its growth medium initiated work on the biosynthetic pathway in that organism

(Asante et al 1977). The hypothesis was then very simple that higher plants might use

the same pathway as in fungi. Radioactive isotope labeled MVA was fed to C. rosicola






3











-lop- ABA
S cow c lI






s -0 -lop- AC CO OH
e-iomylidaic pahway


MVA-D FPP j -p oo -t -- ABA






I -imlylidmc pohway





Fig. 1-1. The direct ABA biosynthetic pathway found in the fungi Cercospora rosicola
(a-ionylidene pathway), C. cruenta (y-ionylidene pathway), C. pini-densiflorae, and
Botrytis cinerea. A single arrow between two compounds does not necessarily indicate
that only one enzyme step is involved. (Adapted from Zeevaart and Creelman 1988)






4


and shown to be converted to I'-deoxy-ABA, a major accumulated intermediate. Purified

I'-deoxy-ABA fed to the fungus was converted to ABA with a good yield. Analysis of

the labeled-intermediates in C. rosicola fed [1,2-'3C]-acetate revealed that ABA is

synthesized via an isoprenoid pathway as indicated in Fig. 1-1 (Bennett et al. 1981,

1984). The radioisotope labeled intermediates were followed a chasing pattern, MVA,

farnesyl pyrophosphate (FPP), a-ionylidene, a -ionylidene ethanol, a-ionylidene acetic

acid, 4'hydroxy-ionylidene acetic acid, I'-deoxy-ABA in C. rosicola.

There seemed to be variations in other fungal species regarding ABA synthesis,

because similar assays revealed different radio-labeled intermediates. The major

difference may be the order of the hydroxylation and oxidation of ionylidene. In B.

cinerea, 1', 4'-t-diol of ABA was shown to be the immediate precursor of ABA (Hirai

1986). In C. pini-densifloorae, ABA seems to be derived from 4'-hydroxy-ionylidene

acetic acid via 1'4'-t-diol of ABA because the latter can be converted at a higher rate

(Okamoto et al 1987). In C. cruenta, the radio-labeled MVA was converted to ABA in

the following order, y-ionylidene ethanol, 4'-hydroxy-r-ionylidene acetic acid, l',4'-

dihydroxy-y-ionylidene acetic acid and ABA (Oritani et aL 1985, Oritani and Yamamoto

1985). Although different fungal species may use one or more of the suggested pathways,

all the intermediates have 15 carbons and are structurally analogous to ABA. Fluridone, a

carotenoid synthesis inhibitor, did not inhibit ABA synthesis in C. rosicola indicating

carotenoids are not involved in the ABA biosynthesis (Oritani and Yamamoto 1985).

In a search for a similar pathway in higher plants, tissues from a variety of plant

species were fed with '4C-MVA (Zeevaart and Creelman 1988). In contrast to fungi, only

a very low incorporation into ABA was detected. Feeding of bean and avocado fruits






5


with ionylidene or I'-deoxy- ionylidene did not produce ABA as was observed in fungi

(Zeevaart and Creelman 1988). The only report that showed ABA could be directly

converted from Cis substrates was provided by Robertson (Milborrow 1983), in which

'4C-phytoene and 3H-MVA were fed to avocado fruits. Both "4C and 3H were found in P-

carotene, but only the 3H label was found in ABA. This experiment was later regarded as

inconclusive because the possible existence of a separate O-carotene pool that is

inaccessible to ABA synthetic enzymes was not ruled out and the C14 incorporation of

into xanthophylls was not examined (Zeevaart et al 1989).


The Indirect C40 Pathway of ABA Synthesis in Higher Plants

For about thirty years since the elucidation of ABA structure, there has been a

debate as to whether ABA is synthesized in higher plants via a direct C1, pathway

(sesquiterpenoid) or via an indirect C40 (apo-carotenoid pathway). Because little evidence

supported the direct pathway in higher plants and due to the emerging evidence in

support of an indirect pathway, a universal pathway of ABA biosynthesis was proposed

for higher plants (Zeevaart et aL 1989, Parry et al 1992, Walton and Li 1992 1995).

Three lines of evidence confirmed that ABA is synthesized predominantly from an

oxidative cleavage of 9-cis xanthophylls (oxygenated carotenoids). 1). Carotenoid

mutants in maize which block the synthesis of certain carotenoids also blocked the

synthesis of ABA strongly indicating that ABA is derived the from carotenoid precursors.

The albino, viviparous mutants of maize, vp2, vp5, vp7,vp9, y3, y9, w3 have reduced

levels of ABA in embryos and cause precocious germination of developing seeds

(Robertson 1955, Neill et al. 1986). The albino phenotype presumably results from













PHYTOENE
+ vp2, vpS

PHYTOFLUENE
S w3

C-CAROTENE
4- vp9,y9
NEUOPORENE


LYCOPENE


y-CAROTENE 6- CAROTENE
4- 4-vp7

P- CAROTENE a- CAROTENE
4- 4 y3
ft-CRYPTOXANTHIN a- CRYPTOXANTHIN


ZEAXANTHIN LUTEIN





Fig. 1-2. The biosynthesis of zeaxanthin and the blocked steps of several
viviparous mutants in maize. Blocked steps are marked with lines crossed the
arrows for the mutants labeled on the right.












ZEAXANTHIN
I epoxidase
ANTHERAXANTHIN
epoxidase

ALL-RANSVIOLAXANTHSIN ALL-TRANSNEOXANTHIN
isomerase 4 4 isomerase

9-CIS VIOLAXANTHIN 9. '-CIS NEOXANTHIN
dioxygenase \ / dioxygenase
XANTHOXIN
I oxidase
ABA ALDEHYDE
I oxidase
ABA




Fig. 1-3. The proposed ABA biosynthetic pathway in higher plants [Slightly
modified from versions of Zeevaart and Creelman (1988) and Walton and Li
(1995)].








photobleaching of chlorophylls in the absence of photoprotection provided by

carotenoids. Such mutants are typically seedling lethal. Further biochemical studies have

located those mutations to the carotenoid biosynthesis pathway as indicated in Fig. 1-2.

2). Incorporation of "0 from 02 into the aldehyde oxygen but not into the ring of ABA in

water stressed leaves in a variety of species clearly supports an oxidative cleavage of a

double bond of 9-cis xanthophylls (Creelman and Zeevaart 1984, Creelman et aL 1987,

Li and Walton 1987, Parry et aL 1988, Gage et aL 1989; Zeevaart et aL 1989). In

addition, a decrease of four xanthophylls (all trans violaxanthin, all trans neoxanthin, 9-

cis neoxanthin and 9-cis violaxanthin) equals stoichiometrically to the increase of ABA

and its metabolites in water stressed etiolated bean leaves (Parry et al 1989; Li and

Walton 1990a). 3). An ABA deficient mutant, abal of Arabidopsis, was isolated by

Koornneef et al. (1982) and was shown to be genetically impaired in the epoxidation of

zeaxanthin to form violaxanthin (Duckham et al 1991, Rock and Zeevaart 1991). This is

an essential step for ABA biosynthesis via the proposed C40 carotenoid pathway

(Zeevaart and Creelman 1988). Recently, the homologous gene of abal in tobacco (aba2)

has been cloned using heterologous transposon tagging(Ac) and the epoxidase activity

was confirmed for the protein expressed in E. coli (Marin et aL 1996). In the present

study, an ABA deficient mutant of maize, viviparousl4, is shown to be blocked in the

oxidative cleavage of 9-cis xanthophylls. The purified recombinant protein expressed as a

GST-fusion in E. coli can cleave specifically the presumed substrates, 9-cis neoxanthin

and 9-cis violaxanthin. These molecular data confirm the existence of the C4o pathway in

higher plants. It has to be pointed out, however, that there is, so far, no conclusive

evidence ruling out the existence of an alternative Cis pathway in higher plants.






9


The proposed C40 ABA biosynthetic pathway in higher plants is shown in Fig. 1-

3. Xanthophyll cycle intermediate zeaxanthin is epoxidated to form all trans violaxanthin

which is evidently isomerized to produce a 9-cis violaxanthin and 9-cis neoxanthin. The

isomerization step is strongly supported by the fact that all trans-xanthoxin can not be

converted to biologically active ABA by cell free extracts (Zeevaart and Creelman 1988,

Walton and Li 1989). As to whether isomerization is an enzymatic or photodynamic

reaction is still not clear. The existence of specifically 9-cis isomers, in fact, favors the

enzymatic reaction view. Because 9-cis neoxanthin is about 10 times more abundant than

9-cis violaxanthin in leaves, the former is suggested to be the primary substrate for

oxidative cleavage to produce xanthoxin, a compound that inhibits cell growth and has a

skeleton similar to ABA. Xanthoxin is further isomerized and oxidized to produce ABA.

The oxidative cleavage reaction is the first committed step in the proposed ABA

biosynthetic pathway. Substantial evidence indicates that it is also the highly regulated

step of this pathway. In contrast to the endogenous level of ABA in most tissues,

carotenoids and xanthophylls are far more abundant than ABA. In particular, 9-cis

xanthophylls are found at much higher concentrations than ABA (Parry et aL 1990, Li

and Walton 1990a), indicating that substrate production is not the limiting factor unless

there is a separate pool for ABA biosynthesis. The endogenous level of the cleavage

product, xanthoxin, is extremely low in all tissues. When xanthoxin is incubated with

enzyme extracts from most tissues of bean, xanthoxin was quickly converted to ABA,

indicating the enzymes downstream of xanthoxin are constitutively expressed (Sindhu

and Walton 1987). Application of the protein synthesis inhibitor cycloheximide to

stressed leaves reduced ABA accumulation, but did not affect the enzymatic conversion






10


ofxanthoxin to ABA compared to the stressed leaves without cycloheximide, suggesting

that those enzymes are constitutively active and not stress inducible under conditions that

caused a greater than 40 fold increase in endogenous ABA level (Li and Walton 1990b,

Sindhu and Walton 1990). These results indicate that the oxidative cleavage of 9-cis

xanthophylls is likely to be the key regulated step in this pathway.


The Questions

In terms of ABA biosynthesis in higher plants, the possibility that a direct

pathway may exist is not completely ruled out. The aba2 mutant of tobacco which seems

to be a null mutation due to an Ac element insertion into the coding region of the Aba2

gene still retained 30% of its endogenous ABA (Marin et al 1996). Molecular analysis

indicated that Aba2 is a single copy gene based on Southern blot analysis. Thus, the

endogenous ABA in that mutant may be from one of three sources, a separate pathway

like the indirect pathway, leaky mutation, or existence of redundant pathways leading to

the production of violaxanthin including photodynamic conversions, etc. Given the

important functions of ABA throughout a plant life cycle, a total knockout of ABA

biosynthesis might be lethal to a plant during early embryogenesis. In that case, this class

of mutants may never be isolated. The direct pathway, if it exists in higher plants, may

only play a very minor role. If so, it is not surprising that mutants in such a pathway have

not yet been found.

One key question with respect to the proposed indirect pathway is the

demonstration of the oxidative cleavage of 9-cis xanthophylls to xanthoxin as it bridges

the carotenoid and ABA biosynthetic pathways. Mutants impaired in other steps

including zeaxanthin epoxidation and conversion of ABA aldehyde to ABA have been









found and are fairly well characterized (Taylor 1992, Rock and Zeevaart 1991, Duckham

et al 1992). Recently, a new ABA deficient mutant of Arabidopsis, aba2, was shown to

be blocked in the step converting xanthoxin to ABA-aldehyde (Schwartz et al. 1997).

However, no mutants that are blocked in the presumably key regulated step, the

oxidative cleavage of epoxy-carotenoids, have been isolated. Notabilis of tomato (Parry

and Horgan 1992) and willy of pea (Duckham et al. 1989) which are not blocked in the

conversion of xanthoxin to ABA and have normal carotenoids and xanthophylls and are

thus possible candidates for mutants in the oxidative cleavage step. However, the highly

leaky nature possibly due to overlapping expression of redundant genes, as discussed

above, has hindered the characterization of these mutants.



Major Functions of ABA in Higher Plants

Regulation of Seed Maturation. Dormancy and Germination by ABA

ABA has been implicated in the control of many events during seed formation

including maturation (Rock and Quatrano 1995), storage protein synthesis (Finkelstein et

aL 1985), desiccation tolerance acquisition (Chandler and Robertson 1994, McCarty

1995), and the onset and maintenance of dormancy (Koornneef and Karssen 1994). In

addition, ABA inhibits certain germination promoting processes such as expression of

hydrolytic enzymes like a-amylase (Jacobsen et al. 1995), which re-mobilize the storage

reserves in endosperm during germination.

Physiological studies indicated that ABA concentration peaks in developing seeds

around the time of maximum fresh weight in many species just prior to the acquisition of

desiccation tolerance and to onset of dormancy (Koornneef and Karssen 1994, Chandler






12


and Robertson 1994). Many proteins have been identified that are positively regulated by

ABA in seeds, and these may play important physiological roles in developmental arrest

or seed maturation. A group of seed storage proteins, the late embryogenesis abundant

proteins (LEA), were found to be activated by ABA (reviewed by Rock and Quatrano

1995). Application of ABA to germinating seeds not only arrests germination, but also

initiates synthesis of LEA proteins, indicating that ABA is a key signal for expression of

Lea genes. Several Lea genes have been cloned and through studies of promoter

structure, a conserved ABA responsive cis-element (ACTG core element) has been

identified (Marcotte et al 1992, Rock and Quatrano 1995). Rich in hydroxylated and

hydrophilic amino acids, LEAs are believed to be osmoprotectants which help protect

cells of seed tissues from desiccation (Rock and Quatrano 1995).

Genetic evidence for ABA function in seed development and dormancy was

revealed by the studies of ABA deficient and ABA insensitive seed mutants. Deficiency

of ABA synthesis in developing maize kernels, occurs in the vp2, vp5, vp7 and vp9

mutants. Such mutants cause the embryo to bypass dormancy and germinate precociously

before completion of seed maturation, i.e. vivipary (reviewed by McCarty 1995). These

mutant kernels neither develop desiccation tolerance nor accumulate certain LEAs,

suggesting that ABA is absolutely required for the acquisition of desiccation tolerance,

and the onset and maintenance of seed dormancy. In addition, sensitivity of embryos to

ABA seems also involved in the regulation of seed development and maturation.

Embryos of sprouting-resistant cereal cultivars have been shown to be more sensitive to

ABA than sprouting-sensitive ones (Walker-Simmons 1987). The ABA insensitive (abi)

mutants of Arabidopsis abi3 (Koornneef et al. 1984), abi4 and abi5 (Finkelstein 1994),






13


and the mutant of maize vpl (Robertson 1965, McCarty et al 1991) exhibit non-dormant

and desiccation intolerant phenotypes, clearly indicating the function of ABA in

controlling late seed development and dormancy.

ABA is a major determinant of the onset and maintenance of seed dormancy,

while the plant hormone gibberellin (GA) controls the release from dormancy and

initiation of germination processes. ABA and GA play their roles sequentially but also

overlap each other. Thus, transition from dormancy to germination is closely related to

the environmental and developmental regulation of ABA and GA biosynthesis. As an

antagonist of ABA, GA induces the expression of genes that promote utilization of

stored seed reserves such as hydrolytic enzymes, whereas ABA inhibits their expression.

One example is amylase expression (Jacobsen et al. 1995). The early hypothesis that a

balance of GA and ABA controls dormancy and germination by a similar fashion as

auxin and cytokinin controls of root and shoot differentiation, may not be entirely correct

as dormancy is not affected in many GA deficient mutants (Rock and Quatrano 1994).

The interaction between sources of ABA found in an embryo may be complex.

Furthermore, dependence of embryo maturation and dormancy on ABA signaling appears

to vary among plant species. The fact that embryos at early stages of development can be

cultured in vitro and develop into plants without undergoing dormancy indicates that

dormancy is affected by maternal factors. However, in maize, homozygous ABA

deficient, viviparous kernels of the vp5 mutant develop in the presence of genetically

heterozygous maternal tissues, indicating that ABA synthesized within the kernel itself

determines the maturation processes and the onset of dormancy. Robertson (1952), using

chromosome T-B translocation techniques, generated kernels that were homozygous vp5






14

in the embryo, but heterozygous in the endosperm. Such embryos still displayed a

viviparous phenotype. This proved that endosperm is not an essential source of embryo

ABA in maize. Thus, acquisition of desiccation tolerance and dormancy of embryo are

determined by the ABA synthesized in embryo itself (McCarty 1995, Rock and Quatrano

1995). It may be possible that maternal tissues surrounding the embryos generate a signal

other than ABA to activate ABA synthesis in developing embryos.

In Arabidopsis, ABA deficient mutants are typically not viviparous unless the

activity of another key regulator of seed maturation, Abi3, is attenuated (Koornneef and

Karssen 1994). A maternal ABA effect was also suggested by the fact that abal, abi3

double mutant F2 seeds that develop with heterozygous maternal tissues were not

viviparous, whereas, double mutant seeds that develop on homozygous abal plants in the

next generation are viviparous (Koornneef and Karssen 1994). The aba2 mutant of

tobacco which is homologous to abal of Arabidopsis is also able to complete seed

maturation and to develop desiccation tolerant seeds (Marin et al. 1996). Although ABA

was not directly measured in aba2 seeds, the ABA deficiency was evident in the

phenotype of reduced dormancy. Aba2 encodes the zeaxanthin epoxidase in ABA

synthesis and is a single copy gene (Marin et al. 1996). Thus, the dependence of seed

maturation and dormancy on ABA signaling appears to be markedly different among

Arabidopsis, tobacco and maize.


Regulation of Stomatal Closure

Resistance to carbon dioxide intake and transpirational water loss by leaves are

controlled by stomatal pores typically located on both surfaces of plant leaves. The

aperture of stomatal pores is controlled by changes in the turgor of the two surrounding






15


guard cells. Loss of turgor, as in case of water deficit, leads to closing of stomatal pores

and hence help protect leaves from further water loss. Guard cell turgor is directly related

to K+ influx and efflux which is controlled at least in part by ABA.

The action of ABA on stomatal closure is manifested in a number of ABA

deficient mutants. Deficiency in ABA typically causes a wilty phenotype because of a

failure of mutant plants to efficiently close their stomata. These mutants include abal,

aba2, aba3 of Arabidopsis; aba2 of tobacco; sitiens and flacca of tomato; wilty of pea;

and droopy of potato (Taylor 1991). Two ABA insensitive mutants (abil and abi2) of

Arabidopsis also exhibit a wilty phenotype suggesting that an ABA signal transduction

cascade is possibly involved in the stomatal closing (Giraudat 1995). Physiological

studies indicate that stomatal opening is strongly inhibited by exogenous application of

ABA onto leaves or peeled epidermis of many species (Mansfield and McAinsh 1995). In

many plant species, bulk ABA levels increase dramatically in leaves exposed to water

stress (reviewed by Ingram and Bartels 1996). However, when the time course is taken

into account, the closing of stomata generally occurs before any significant changes of

ABA content can be detected in the leaves. One possible explanation for this time

discrepancy may involve a fast localized accumulation of ABA around or in guard cells,

through either re-localization of ABA among leaf cells or highly localized de novo

synthesis of ABA, e.g. only in guard cells or the neighboring cells. In support of this

suggestion, a rapid increase in the ABA concentration of a single guard cell was shown to

be correlated with the closure of stomata in stressed leaves (Harris and Outlaw 1991).

However, this study did not identify the source of the ABA. These results also suggest

that the dramatic accumulation of ABA in stressed leaves may have functions other than






16


to close the stomata. It may mediate other processes involved in plant acclimation to the

environment.

The sources of ABA in leaves appear to be complex. It is now well established

that ABA is synthesized in roots (Cornish and Zeevaart 1988). When growing roots come

in contact with drying soil, they produce ABA in increased quantities, which enters the

xylem and is transported to the leaves where it may inhibit stomatal opening (Zhang and

Davies 1989). This occurs before the shortage of soil moisture causes any measurable

change in the water status of the leaves. Thus it is suggested that the early stages of soil

drying in field grown plants lead to the production of ABA which is transported as a

chemical signal to the leaves, where it causes a reduction in transpiration and prevents a

decline in water potential or a loss of turgor. When leaves are detached, the increased

ABA level apparently results from the de novo synthesis because transcription and

translation inhibitors can prevent this increase (Li and Walton 1990b, Ingram and Bartels

1996). Thus, accumulated ABA in stressed leaves apparently has at least two sources,

roots and local synthesis. It is possible that in field grown plants, the sources of ABA that

cause stomatal closure are determined by the severity of water deficit; mild water deficit

will cause increased synthesis of ABA in roots which translocates to leaf guard cells;

severe and progressive water deficit will induce ABA synthesis in the leaf itself. The

study of field grown maize by Tardieu and Davies (1992) has shown that there is a good

correlation between stomatal conductance and ABA concentration in the xylem sap and

this may be the best above-ground indicator of the water status of the root system.

ABA modulates the activities of three major classes of ion channels, inward-,

outward-rectifying K+ channels and anion channels located on the guard-cell plasma






17


membrane (Armstrong et al. 1995). How the ABA signal cascade functions is still

unknown. When guard cells are exposed to ABA, the first detected event is an influx of

positive charges leading to an initial depolarization of the membrane (Thiel et al 1992).

The influxes of Ca+ and H+ are proposed to act as secondary messengers to activate

Ca2+ sensitive and voltage sensitive anion channels. This causes long term depolarization

and a massive anion efflux across the guard cell membrane (Schroeder and Hagiwara

1990). It is suggested that slow anion channels could be the rate limiting step in

controlling stomata closing. Using patch-clamp techniques, it was found that the slow

anion channel is strongly activated by ABA and this activation can be suppressed by the

protein phosphatase inhibitor, okadaic acid. this suggests a kinase/phosphatase signal

cascade (Pei et al 1997). In the ABA-insensitive (abi) mutants ofArabidopsis, abil and

abi2, failure to close stomata is associated with loss of the slow anion channel activation

caused by ABA. ABII encodes a protein phosphatase 2C, a component in ABA signaling

pathway (Meyer et al. 1994, Leung et at 1994). The membrane depolarization is

proposed to generate the driving force for K+ efflux through the outward-rectifying K+

channels. This in turn causes the loss of turgor in guard cells, and closing of stomata

(Armstrong et al 1995, Pei et al 1997). Conversely, ABA also inhibits the inward-

rectifying K+ channel activity (Blatt 1992, Schroeder and Keller 1992).

Guard cells of tobacco transformed with the abil dominant allele have reduced

ABA activation of the outward-rectifying K+ channel and also show less sensitivity of

the inward-rectifying K+ channels to ABA (Armstrong et al 1995). The presence of

broad-range protein kinase antagonists, H7 and staurosporine restores the sensitivity,

suggesting a kinase may be a negative regulator. A similar study of the ion channels also






18


indicated loss of slow anion channel activation by ABA in abil and abi2 mutants (Pei et

aL 1997). ABA-induced anion channel activation and stomatal closing were suppressed

by protein phosphatase inhibitors in both mutants. However, kinase inhibitors did not

rescue the ABA activation of slow anion channels in abi2 mutants, indicating that abi2

may be an upstream component in ABA signal cascade. Since most of the ABA signaling

cascade as well as ABA receptors in plasma membrane of guard cells have not been

identified, the positions of ABIl and ABI2 in that cascade remain to be resolved. There

are reports that Ca2+ dependent protein kinases (CDPK1 & CDPKla) may also be

involved in the ABA signal transduction pathway (Sheen et al 1996).

Another question that remains elusive is the location of ABA receptors which

hold the key to the ABA signal transduction cascade. It is believed that they are located

on the outer surface of guard cell plasma membrane sensing ABA from outside the guard

cells when water deficit occurs (Hartung and Davies 1991). However, micro-injection of

a physiological concentration of ABA into the cytoplasm can trigger the K+ efflux and

close stomata, and the extent of stomata closing is correlated with the extent of ABA

uptake by the guard cells, thus suggested the existence of internal ABA receptors

(Schwartz et al 1994). An increase of ABA content inside the guard cells is correlated

with the closing of stomata (Harris and Outlaw 1991), suggesting the existence of

intercellular ABA receptors.


Adaptation to Stress Environments

The survival of a plant in nature frequently requires a capacity to withstand

extremes of a stressful environment including drought, extreme temperature, salinity,

wounding and pathogen infection. Plants have developed two overall mechanisms to






19


survive, stress avoidance and stress tolerance (Chandler and Robertson 1994). Avoidance

is achieved by developing specialized adaptations in architecture, such as a highly

developed root system to avoid drought stress. Whereas stress tolerance depends on a

combination of rapid and longer-term physiological responses to stresses, such as closing

stomata to prevent from further water loss and accumulation of a class of proteins to

protect cells from damage (Skriver and Mundy 1990, Chandler and Robertson 1994,

Ingram and Bartels 1996).

The involvement of ABA in stress acclimation is supported by several lines of

evidence: (1) ABA accumulates during stress including drought, flooding, low

temperature, wounding and pathogen infection (Ingram and Bartels 1996). Wright and

Hiron (1969) first detected a 40 fold increase of ABA in detached, wheat leaves.

Subsequent studies have reported similar phenomena in many plant species (Ingram and

Bartels 1996). The mechanism of stress-dependent ABA biosynthesis is still not fully

understood (Hartung and Davies 1991). (2) Plants develop freezing tolerance when

treated with ABA under nonacclimating conditions (Lang et al 1989). The ABA

deficient mutant abal of Arabidopsis does not develop cold acclimation and this can be

complemented by exogenous application of ABA (Heino et al 1990). (3) Under drought

stress conditions, ABA is shown to be required for primary root elongation, and

elongation of primary roots is inhibited in ABA deficient vp5 of maize (Sharp et al

1994). (4) Proteins that are induced by environmental stresses and associated stimuli

such as RABs (responsive to ABA) can be induced by applying ABA (Skriver and

Mundy 1990, Mantyla et al 1995).






20


One class of ABA proteins that are strongly induced by ABA in the late

embryogenesis is the LEAs. Although functions of those proteins are still unknown,

synthesis of LEAs is correlated with the acquisition of desiccation tolerance in embryos.

The direct relationship between LEAs and ABA was established by ABA treatment of

cultured embryos, which induced LEA expression (Ingram and Bartels 1996). Recent

molecular studies have also characterized the rapid induction of RAB genes in leaves or

roots. These novel genes have been isolated from several species by differential screening

of cDNA libraries constructed from ABA treated leaves or roots. Some of these proteins

are also expressed during the maturation phase of embryo development, thus may be

functionally related to LEAs (Rock and Quatrano 1995). Both LEA and RAB proteins

share one predominant feature; high hydrophilicity and a high content of uncharged and

hydroxylated amino acids. Conserved domains are postulated to be functionally

important in desiccation protection (Dure et aL 1989), possibly through interactions with

other embryo proteins and membranes. However, the RAB17 of maize was found

localized in the nucleus and is suspected to play a role in the nuclear protein transport

(Goday et al 1994).

The regulation of ABA responsive gene expression is mediated by the promoters

of these genes. ABA response elements (ABRE) which contain the palindromic motif

CACGTG with a G-box core element (Giuliano et al. 1988) are found in RABs and

LEAs, and have been extensively studied in Em gene of wheat and the rice RAB 16 gene

(Skriver et al. 1991, Guiltinan et al 1990). Recently, the modular nature of the abscisic

acid response complex (ABRC) was further refined in a barley EM gene, HVA1. The

promoter unit necessary and sufficient for abscisic acid (ABA) induction of gene






21


expression consists of a G-box and one of several "coupling elements" (CEs) (Shen et aL

1995, 1996). Different combinations of the G-box with different CEs may modulate

different gene responses in different tissues.

ABA is also involved in processes that lead to the establishment of freezing

tolerance. However, little is known about how ABA triggers those processes. Thus far,

few mutants have been isolated that have an altered freezing tolerance phenotype. A

putative freezing sensitive mutant (frsl) of Arabidopsis has a reduced cold acclimation

response and this response can be compensated partially by application of ABA (refer to

Quatrano et al. 1997). The molecular nature underlying this mutation should be of

importance to the understanding of freezing tolerance mechanisms. (Quatrano et al. 1997)


Inhibition of Growth

Physiological studies have established that exogenous application of ABA leads

to inhibition of vegetative growth. At the molecular level, ABA regulates the expression

of a large number of genes especially under stressful environments. Many of these ABA

up-regulated genes have been cloned by differential screening; however, their functions

remains unknown. The ABA down-regulated genes have drawn little attention, althrough,

most appear to promote cell growth. An example is the inhibition of the a-amylase

promoter by synergistic interaction of ABA and VP1 (Hoecker et al 1995).


Defense Responses

The proteinase inhibitor II (Pin2) is a model system for wound induced gene

activation (Pena-Cortes and Willmitzer 1995). The Pin2 gene family is the best studied

examples of genes which are systematically activated upon mechanical damage of plants.






22


The involvement of ABA in gene activation processes following mechanical damage of

the plant tissues is supported by the fact that the endogenous ABA concentration rises

three- to five-fold upon wounding (Pena-Cortes and Willmitzer 1995). Exogenous

application of ABA on non-wounded plants can activate the expression of Pin2 in a

pattern analogous to wounding ( Pena-Cortes et al 1991). In the ABA deficient droopy

mutant of potato (Quarrie 1982) and sitiens mutant of tomato (Taylor et al 1988,

Duckham et al 1989), wounding failed to stimulate the expression of Pin2, and did not

increase ABA in leaves (Pena-Cortes and Willmitzer 1995). Spraying ABA on wounded

droopy and sitiens leaves resulted in an activation of Pin2, thus strongly indicating that

ABA is at least one of the signals to trigger Pin2 gene activation upon wounding.

Drought stress induced ABA accumulation does not activate Pin2, whereas wounding of

the stressed leaves of potato and tomato activates Pin2 transcription (Pena-Cortes et aL

1989). This implys that multiple factors including ABA may account for the wounding

activation of gene expression.

Several other wound induced genes have been cloned which show a pattern of

wound induction similar to Pin2 (Hildmann et aL 1992). ABA is involved in the

activation of those genes. Molecular analysis of promoters from several genes and Pin2

revealed an ABA responsive element (ABRE) that is similar to that found in ABA

regulated genes including Leas, Em and Rab16. However, mutagenesis of this element

did not affect ABA regulated expression (Lorberth et aL 1992). This suggests that

different factors in addition to the presence of ABA are necessary for the induction of this

set of genes in contrast to Em. Consistent with the multiple functions of ABA throughout

plant development, the signaling pathways for ABA that regulate transcription of






23

different genes are expected to be complex. In addition to ABA, jasmonic acid is also

involved in the activation of related genes in response to wounding (Pena-Cortes and

Willmitzer 1995).


Abscission and ABA

Although abscisic acid was initially isolated as a plant hormone that enhanced

organ abscission, this nomenclature is misleading. Application of ABA to plant leaves or

flowers does not result in organ abscission (Addicott 1983). Ethylene is a plant hormone

that more directly controls organ abscission (Abeles et al. 1992). Application of ethylene

or its precursor 1-aminocylopropane-l-carboxylic acid (ACC) strongly promotes organ

abscission. The link between ABA and organ abscission may be linked to an interaction

of the two hormones. It is possible that ABA is involved in the senescence processes in

the separation layer cells in organs undergoing abscission (Abeles et al. 1992). ABA

increases ethylene production and abscission in aged citrus organs (Riov et al 1990).

Water stress may cause leaf abscission in cotton and citrus, and inhibition of ABA

biosynthesis was reported to reduce leaf abscission (Gomez-Cadenas et al 1996).


The ABA Related Mutants in Higher Plants

ABA Insensitive Mutants

The abil, abi2, and abi3 were isolated by Koornneef et al. (1984), and the abi4

and abi5 were isolated by Finkelstein (1994). These mutants were selected by

germination of EMS mutagenized Arabidopsis seeds on media containing ABA at

concentrations that completely inhibit germination of wild type seeds. The abi3, abi4 and

abi5 mutations affect seed development specifically. The severe abi3 alleles isolated






24

later, abi3-3 (Nambara et al 1994), abi3-4 abi3-5 (Ooms et al. 1993), exhibit a

phenotype of nondormancy and desiccation intolerance. The viviparousl of maize

exhibits a similar seed specific phenotype, and in addition controls anthocyanin synthesis

in the seed (Robertson 1955, McCarty et al. 1991). Cloning of Vpl indicated that it

encodes a novel transcription factor (McCarty et al 1991). It can activate CI, and EM

promoters in a synergistic manner in combination with ABA (Kao et al 1996, Hoecker et

at 1995, Hottori et at 1992, Vasil et aL 1995), but acts as a repressor of ca-amylase

(Hoecker et al 1995). A B3 domain of VP1 has a cooperative of DNA binding activity to

the sph element of Cl promoter (Suzuki et aL 1997). The sequence of ABI3 revealed

strong homology to VP1 (Giraudat et aL 1992), and thus ABI3 is considered an ortholog

of VP1( McCarty 1995). Unlike abi3, plants of abil and abi2 display a wilty leaf

phenotype in normal growth conditions (Koornneef et aL 1984, Giraudat 1994). ABI1

encodes a 2C type of protein phosphatase (Meyer et al. 1994, Leung et al 1994). ABI1 is

suggested to be a signal transduction component involved in many processes including

embryogenesis, ABA responses in guard cells (Amstrong et aL 1995, Pei et al 1997) and

mitotic cell division in roots (Leung et aL 1994). A puzzle remains as to how a missense

mutation (Gly-180 to Asp) in the mutant abil allele gives rise to a dominant phenotype.

A search for abil homologous sequences, surprisingly, led to the cloning of Abi2, and the

mutation in abi2 was exactly as in abil, a Gly-180 to Asp (Leung et aL 1997). Thus,

ABI1 and ABI2 encode very similar protein phosphatases.


Carotenoid Biosvnthetic Mutants

In maize, several viviparous mutants are associated with an albino phenotype as

indicated by vp2, vp5, vp7, vp9, w3, y9, etc. One function of the carotenoids is to protect






25

the photosynthetic apparatus from photobleaching in a high light situation. Blocking the

synthesis of certain carotenoids causes these mutants to be albino and generally lethal.

The steps blocked in the various mutants are shown in Fig. 1-1. As ABA is synthesized

from cleavage of carotenoids, certain carotenoid mutants are also ABA biosynthetic

mutants. Those mutants provided key evidence that ABA is derived from carotenoids but

have generated little information about the committed (specific) steps of the ABA

biosynthetic pathway (Zeevaart and Creelman 1988).


ABA Biosynthetic Mutants

A number of ABA deficient mutants have been isolated from a variety of plant

species as indicated in Table 1. Most of those mutants (ftc, sit, droopy, clrl, ibal, etc.)

affect the last step of ABA biosynthesis, the oxidation of ABA-aldehyde to produce

ABA. The tomato mutant, Notabilis, is less well characterized largely due to its leaky

nature, but is suspected to be impaired in the oxidative cleavage of 9-cis neoxanthin to

xanthoxin (Parry et aL 1988, Parry and Hogan 1992a, b, Taylor et al. 1988). Similarly, a

pea wilty (wt) ABA deficient mutant is suspected to be blocked in the same step, since it

can convert ABA aldehyde to ABA (Duckham 1989). No differences in xanthophyll and

carotenoid profiles were found in wt mutant and wild type plants (Duckham et aL 1991).

More detailed reviews including these mutants can be found elsewhere (McCarty 1995,

Rock and Quatrano 1995, Taylor 1991). The following summarizes the current advances

in this area.

The Arabidopsis abal was isolated in a screen for revertants of the GA deficient

gal mutant population (Koornneef et aL 1982). Abal is impaired in the conversion of

zeaxanthin to violaxanthin (Duckham et aL 1991, Rock & Zeevaart 1991). The Aba2






26


mutant of tobacco was isolated by heterologous Ac tagging in tobacco and shown to

encode zeaxanthin epoxidase by expression in E coli. (Marin et al 1996). The tobacco

aba2 gene is able to complement the abal phenotype of Arabidopsis through T- DNA

transformation, indicating that it is an otholog of abal (Marin et al 1996). Additional


Table 1-1. Abscisic acid deficient mutants

Species Mutant Phenotype Blocked step References*

Arabidopsis thaliana abal nd, wilty, red. height Zexan -> Vioxan Koonneef 1982; Rock 1991

aba2 red. dormancy, wilty Xan -> ABA-ald Leon 1996; Schwutz 1997

aba3 red. dormancy, wilty ABA-ald -> ABA Leon 1996; Schwatz 1996

Nicotiana plumbaginifolia aba2 nd, wilty Zeaxan -> Vioxan Mamn 1996

(tobacco) cdrl cytokinin R, wilty ABA-ald-> ABA Pa"y 1991

ibal auxin R, wilty ABA-ad -> ABA Bioun 1990

Lycopersicon esculentum flacca wilty ABA-ald -> ABA Parry 1988

(tomato) silten wilty ABA-ald -> ABA Prry 1988

notabilis slightly wilty ? Tayor 1988; Parry 1991

Solanum phureja (Potato) droopy wilty ABA-ald -> ABA Dudacdm 1989

Pisum sativum (pea) willy slightly wilty ? Pny 1991

Hordeum vulgare (barley) nar2a wilty ABA-ald -> ABA Wa&er-Simmona 1989

Zea may (maize) vp*2274 viviparous this study MoCrty (1995)

(vp 4)



Note: nd, nondomant; red., reduced; R, resistance.

* Only the first author was listed to save space.



Arabidopsis ABA deficient mutants were isolated in a screen for seed able to germinate

in the presence of the gibberellin biosynthesis inhibitor paclobutrazol. Two new ABA






27


deficient mutants were reported that have reduced dormancy and excessive rate of leaf

water loss (Leon-Kloosterziel et al. 1996). Biochemical characterization indicated that

aba2 is blocked in the conversion ofxanthoxin to ABA-aldehyde and aba3 blocked in the

last step of ABA biosynthesis, i.e. the conversion of ABA aldehyde to ABA (Schwartz et

al. 1997).


Candidates for Mutations in the Cleavage Step

Two ABA deficient mutants, wilty and notabilis of pea and tomato, are blocked at

undetermined steps in ABA biosynthetic pathway (Taylor 1991). Biochemical analysis of

the two mutants did not detect any defects in either the xanthophyll profiles or the

conversion of xanthoxin to ABA (Parry and Horgan 1992a, b, Taylor 1991). However,

the highly leaky nature of both mutations greatly compromised any conclusions based on

these data. Several newly isolated viviparous mutants of maize may potentially provide

new information about the pathway (McCarty 1995). Viviparous mutants of maize,

characterized so far, are all associated with ABA, either through perception or synthesis.

Thus, mutants with normal carotenoids but decreased ABA levels are potentially ABA

biosynthetic mutants. One mutant, named vp14-2274 (McCarty 1995), was analyzed in

this study and is shown to be impaired in the oxidative cleavage step, the conversion of 9-

cis xanthophylls to xanthoxin. Further study of these mutants is expected to fiurther

improve our understanding of ABA function and biosynthesis in plants.



Summary and Future Perspectives


ABA, a sesquiterpene plant hormone, has multiple functions throughout the plant

life cycle (Addicott 1983, Zeevaart and Creelman 1988, Taylor et aL 1992, Walton and Li






28


1995). ABA plays important roles in the inhibition of growth, control of stomatal

aperture, organ abscission, embryo development, dormancy, tolerance of stresses and

plant defense responses (Zeevaart and Creelman 1988, Hartung and Davies 1991, Black

1991, Prescott and John 1996). In addition, ABA is particularly interesting with regard to

its molecular regulation because hormone levels increase 10- to 50-fold within hours of

environmental perturbations (Walton and Li 1995). Considerable effort has been directed

toward dissection of the ABA biosynthetic pathway since elucidation of the chemical

structure of ABA in 1965 (Addicott). We now know that in certain fungi (Cercospora

rosicola, etc) ABA is synthesized from a Cis isopropenoid via a so called direct pathway

(Bennett et al 1984, Zeevaart and Creelman 1988). In contrast, higher plants synthesize

ABA largely from a C40 carotenoid (zeaxanthin) via a so called indirect pathway using an

oxidative cleavage of 9 '-cis xanthophylls to produce xanthoxin. Xanthoxin is converted

through two additional steps to ABA (Zeevaart and Creelman 1988, Walton and Li

1995). The cleavage step is the first committed step leading to ABA biosynthesis and

substantial evidence has indicated that it is likely to be the key regulated step in ABA

biosynthesis (Zeevaart and Creelman 1988, Walton and Li 1995). The epoxidase which

converts zeaxanthin to all-trans violaxanthin has recently been cloned in tobacco and the

enzyme activity was confirmed (Marin et aL 1996). However, none of the enzymes that

are specific to ABA synthesis have been isolated, particularly the hypothesized

dioxygenase that cleaves carotenoids.

Mutants have proven to be very powerful in the elucidation of biosynthetic

pathways, particularly in the case of ABA biosynthesis, where conventional isotope

labeling has not been feasible. A single recessive mutation such as vp5 has provided






29


important evidence that ABA is synthesized from carotenoids. The abal mutant of

Arabidopsis and aba2 of tobacco have led to the conclusion that zeaxanthin is a precursor

of ABA synthesis and that an epoxidation of zeaxanthin to form violaxanthin is required

for ABA synthesis (Rock and Zeevaart 1991, Buckham et al 1992, Marin et al 1996).

Other mutants like sittens, flacca, etc. have confirmed the last step of ABA synthesis

(Taylor 1992). Recently, an ABA deficient mutant of Arabidopsis, aba2, was shown to

be blocked in the step of converting xanthoxin to ABA-aldehyde (Schwarts et al. 1997).

However, mutants that block other intervening steps in the ABA synthetic pathway are

still lacking, particularly mutants blocking the proposed oxidative cleavage of epoxy-

carotenoids. Mutants that potentially regulate ABA biosynthesis are also required. Thus,

successful isolation of new, developmentally specific ABA deficient mutants provide an

important way to the study of ABA biosynthesis and regulation (McCarty 1995).

Eventually, the combination of molecular cloning, genetics, biochemistry and

physiological studies will lead to a better understanding mechanisms for ABA action in

plants. This will include developmental and environmental regulation of ABA synthesis,

as well as how ABA regulates the expression of other genes to exert its function as well.

Agricultural applications resulting from manipulation of ABA synthesis in crops will also

be possible.

Mutants are also required for explanation of the ABA signal transduction cascade

since it is the key to understanding how ABA functions in plants. ABA signal

transduction pathways are likely complex, as indicated by the functions of ABA

throughout plant life cycle. Potential components of the ABA transduction pathway such

as Abil and Abi2 have begun to emerge (Meyer et al. 1994, Leung et al. 1994 1997). A






30


recent mutant in Arabidospsis which confers an enhanced response to ABA (era) encodes

the beta subunit of farnesyl transferase, a putatively negative regulator of ABA signal

transduction pathway (Cutlers et aL 1996). However, the key details of how the ABA

signal is perceived and transduced in plant cells are far from clear.














CHAPTER 2
MOLECULAR ANALYSIS OF VIVIPAROUSI4, A DEVELOPMENTALLY
SPECIFIC ABSCISIC ACID BIOSYNTHETIC MUTANT OF MAIZE


Introduction

The plant hormone, abscisic acid (ABA) plays a key role in the regulation of seed

maturation and dormancy (McCarty 1995, Rock and Quatrano 1995) and mediates plant

responses to a variety of stress conditions including wounding (Pena-Cortes and

Willmitzer 1995), drought (Vartanian et al. 1994, Giraudat 1995) and cold acclimation

(Gilmour et al. 1991). ABA is synthesized from carotenoids as manifested in the ABA

deficient mutants of maize (e.g. vp2, vp5, vp7, and vp9) that are in fact blocked in the

early steps of carotenoid synthesis (Neill et al. 1987). Biochemical studies (reviewed by

Zeevaart and Creelman 1988) and analyses of mutants that block ABA synthesis

(reviewed by Taylor 1991) have led to a proposed indirect ABA synthetic pathway in

higher plants (refer to Chapter 1, Zeevaart and Creelman 1988, Walton and Li 1995). The

first committed step in the pathway is the cleavage of two potential substrates, 9-cis

violaxanthin and 9-cis neoxanthin to form xanthoxin which is subsequently converted to

ABA via an ABA-aldehyde intermediate.

Mutants blocked in ABA synthesis have been isolated in a variety of plant species

including maize, Arabidopsis, tomato, tobacco, potato, etc. (Robertson 1955, Koornneef

at al. 1984, Leon-kloosterziel et al. 1996, Marin et al. 1996, Neill et al. 1987). ABA

deficient mutants of maize are viviparous, i.e. developing kernels fail to enter


31






32


developmental arrest in the mid to late stage of seed maturation and germinate while still

attached to the plant (Neill et al. 1986 1987, Robichaud et al. 1987). The viviparous

mutants of maize, w3, y3, vp2, vp5, vp7 and vp9 are blocked in early steps of carotenoid

biosynthesis, and thus also display albino (carotenoid deficient) and seedling lethal

phenotypes due to photobleaching of chlorophyll (Neill et al. 1986, refer to Fig. 1-1 for

blocked steps). These mutants are not considered specific to ABA synthesis and have

contributed little to the studies of the ABA specific portion of the biosynthetic pathway

(Zeevaart and Creelman 1988). ABA deficient mutants in other species (tomato,

Arabidopsis, tobacco, etc) are generally not viviparous indicating that dependence of seed

maturation on ABA signaling varies markedly among these species (McCarty 1995). The

abal mutant of Arabidopsis exhibits enhanced sensitivity to water stress and a reduced

requirement for gibberellin hormone synthesis during seed germination (Koornneef et al.

1986, 1989). abal is deficient in the epoxy carotenoids, such as neoxanthin and

violaxanthin, due to a genetic lesion in the epoxidation of zeaxanthin (Rock and Zeevaart

1991, Duckham et al. 1991). The aba2 mutant of tobacco, recently isolated and cloned by

transposon (T-DNA/Ac) tagging, is homologous to abal of Arabidopsis and encodes a

zeaxanthin epoxidase (Marin et al. 1996). The flacca and sitiens mutants of tomato

(Sindhu and Walton 1988), and the aba2 and aba3 mutants of Arabidopsis (Schwartz et

al. 1997) genetically define two steps downstream of the cleavage reaction that allow

conversion of xanthoxin to ABA. Thus far, mutants blocked in several steps of the

pathway have been isolated. None of these mutants are known to be blocked in a key step

that connects Cao epoxy-carotenoids and C15 xanthoxin, i.e. the cleavage of epoxy-

carotenoids. The notabilis mutant of tomato is the only reported candidate for the






33

carotenoid oxidative cleavage (Parry et al. 1992 a, b). However, the blocked step has been

biochemically difficult to locate in notabilis largely as a result of its highly leaky nature.

Circumstantial evidence indicates that the oxidative cleavage of epoxy-carotenoids

is the key regulated step in the ABA biosynthetic pathway. The carotenoid precursors of

ABA are widely distributed in plant tissues. Epoxidation of zeaxanthin was ruled out to

play a regulating role in ABA synthesis (Marin et al. 1996). The potential immediate

precursors in the proposed reaction (Creelman et al. 1992, Walton and Li 1995), 9-cis

violaxanthin and 9-cis neoxanthin, are far more abundant than the endogenous ABA

levels in leaves of maize, pea, tomato and other species (Parry et al. 1990, Li and Walton

1987). This indicated that steps prior to formation of the 9-cis xanthophylls are unlikely to

limit ABA synthesis. The enzyme activities downstream of the oxidative cleavage step

required for conversion of xanthoxin to ABA also appear to be constitutively active in

most plant tissues (Sindhu and Walton 1987). The enzymatic activities required for

conversion of xanthoxin to ABA, as assayed in vitro, are not regulated by drought stress,

and are also not inhibited by protein synthesis and transcription inhibitors in stressed

leaves (Li and Walton 1990). Accumulation of ABA by stresses can be inhibited by

inhibitors of both transcription and translation in a variety of species (Ingram and Bartels

1996). These studies have led to the hypothesis that the oxidative cleavage of 9-cis

xanthophylls to form xanthoxin is the key regulated step in the ABA synthetic pathway

(Sindhu and Walton 1987, Creelman et al. 1992).

As suggested by McCarty (1995), a potentially important approach for identifying

mutants in regulated steps of ABA synthesis is to look for mutants that affect ABA

synthesis in specific organs or phases of development. In maize, evidence that all the






34

known ABA deficient mutants are viviparous indicated that vivipary is possibly a sensitive

seed specific phenotype for detection of ABA deficient mutants. To generate clonable

ABA deficient mutants in maize, we screened active Robertson's Mutator transposon

populations for viviparous seed mutants. To avoid mutants that were blocked in

carotenoid biosynthesis, we specifically looked for mutants that were viviparous but fidly

viable as homozygotes and had normal carotenoid pigmentation (McCarty 1995). Here we

report evidence that one of the ABA deficient mutants, viviparousl4 (vpl4), is blocked in

the cleavage step of epoxy-carotenoids. Molecular analysis of the VpJ4 gene indicates that

it encodes a protein related to lignostilbene dioxygenase, a bacterial enzyme which

catalyzes a double bond cleavage reaction with striking similarity to the carotenoid

cleavage step of ABA biosynthesis. We show that VpJ4 and related genes are

developmentally regulated and induced in leaves by drought stress.


Materials and Methods

The p]14-2274 and vp14-3250 mutants were identified in active Robertson's

Mutator lines. Wild type W22, M14, Q66, Q67, Q77, Q79 strains that founded the

Mutator population were a gift of Donald S. Robertson, Iowa State University. The

nonsegregating wild type NS-2274 and homozygous vp14-2274 mutant strains used for

biochemical and molecular analysis were extracted from a segregating vp14-2274 stock

that had been maintained by self pollination and selection of heterozygous plants.

ABA Determination and HPLC Analysis of Carotenoids and Xanthophvls

Embryos were harvested in Florida at 16, 18 and 20 days post-pollination and frozen in

liquid nitrogen. Extraction, purification and analysis of carotenoids and xanthophylls by






35


HPLC was carried out in collaboration with Dr. Jan Zeevaart's group at Michigan State

University using methods described earlier (Rock and Zeevaart 1991). ABA extractions,

purification, and quantification by GC-MS with electron capture detector (ECD) were

performed as described (Leon-Kloosterziel et al. 1996b).

Determination of ABA Sensitivity in Culture Ears were harvested from green

house grown plants at 16 and 18 days after pollination (DAP) and surface sterilized by

submersion in a 70% ethanol solution containing 1% dish detergent for 5 minutes.

Embryos were harvested aseptically and placed in magenta boxes containing MS medium

(pH 5.7), 0.2% phytagel and the indicated concentration of abscisic acid (Robichaud and

Sussex 1986). Embryos were cultured at 25 C in a growth chamber and shoot, root

length and fresh weight were measured after four days. The data shown are the means of

15 to 20 16 DAP embryos for each treatment. 18 DAP embryos showed a similar

response.

Southern and Northern Blot Analysis Genomic DNA was isolated from maize

seedling leaves as described by Dellaporta (1983). Approximately 10 Atg of DNA was

digested with the indicated restriction enzymes and resolved by agarose gel

electrophoresis. The gel was denatured then blotted onto nylon membrane (Sambrook et

al 1989). The membrane was UV-linked and hybridized in a Hybaid Chamber (Bio-Rad)

according to the method of Church and Gilbert (1984). The membranes were washed in a

40 mM sodium phosphate buffer (pH 7.2), 1% SDS solution at 65 C for 2-3 hours for

moderate stringency and at 70 C for high stringency.

Total RNA was extracted from maize tissues (1-2 g) in TriZol solution according

to manufacturers instructions (BRL) and purified further by precipitation with






36


isopropanol. Poly(A)-enriched RNA was prepared using PolyATtract according to

manufacturers instructions (Promega) and quantified spectrophotometrically. 1 pag of

Poly(A)*-RNA was resolved in a 1.2% agarose gels containing formaldehyde (Sambrook

et al. 1989). After transfer, nylon membranes were UV-linked and probed as described for

southern hybridization. The probes were radiolabeled with the Random Primer DNA

Labeling System (BRL) in the presence of 32Pa-dCTP (3000 Ci/mmol, DuPont).

Construction and Screening of Genomic and cDNA Libraries Approximately

100 uIg genomic DNA prepared from homozygous mutant or wild type plants (vp14-2274,

vp14-3250, wild type) was digested with appropriate restriction enzymes, then size

fractionated by centrifugation at 85,000 xg at 4 C for 24 hours through a 10-40% linear

sucrose gradient prepared in sterile TE buffer (pH 8.0). Fractions containing the fragment

of interest were confirmed by Southern blotting and the concentrated DNA was ligated

into an appropriate lambda phage cloning vectors (X-gtlO was used to clone the 2.5 kb

xhol fragment, X-ZAP was used to clone the EcoRI fiagments). The ligated DNA was

packaged into lambda phage according to the manufacture's instructions (Stratagene).

A wild type embryo cDNA library was constructed in X-Zap from 5 utg of

poly(A)-mRNA prepared from 18 day post pollination inbred W22 embryos. cDNA was

prepared and cloned using the X-ZAP Express cDNA Synthesis Kit according to the

manufacture's instruction (Stratagene). Lambda phage libraries were plated, lifted on

nylon membrane and probed by DNA hybridization as described above for Southern and

northern blots.

A 2.5 kb XhoI Mul containing genomic fragment isolated from vp14-2274 was

subcloned in pBluescript (Promega) and PCR was used to amplify and clone a 1 kb






37


sequence that flanked the Mul insertion element. A oligonucleotide primer specific for the

inverted terminal repeat of Mu was used as the 5' primer (5'-CCATAATGGCAATTAT-

CTC-3') and the T7 sequencing primer for the vector was used as the 3' primer

(TAATACGACTCACTATAGGG-3'). The resulting -1 kb fragment was subcloned to

pBluescript-SK.

Conversion of Xanthoxin to ABA in Cell Free Enzyme Extract This assay was

performed by Dr. Zeevaart's group at Michigan State University on embryos prepared in

Gainesville. Embryos were dissected from self-pollinated ears of homozygous mutant and

wild type segregants of an F3 family. Embryos were homogenized in 0.2 M potassium

phosphate (pH7.5) and 10 mM DTT and the extract was centrifuged to remove insoluble

material. The supernatant soluble protein fraction was desalted using a G-25 Sephadex

spin column. Enzyme assays contained 1 mM PMSF, 0.25 mM EDTA, 1 gg 9-cis

xanthoxin and the crude enzyme extract. The amount of ABA produced was determined

by GC-MS.


Results

Isolation of Viviparous14 Mutant

All of the characterized viviparous mutants of maize (vpl, vp2, vpS, etc.) are

blocked either in the perception or in the synthesis of ABA (McCarty et al. 1991, Neill et

al. 1987). A search for viviparous mutants in maize was considered to be an efficient way

to isolate ABA deficient mutants. In a screen of the maize strains containing active

Robertson's Mutators transposons, two alleles of a new recessive mutant,

viviparousl4(vpl4), were identified and designated as vp14-2274 and vp14-3250. The






38

vpJ4 mutant kernels have a weak viviparous seed phenotype. In field growth conditions,

the embryo shoot axis of the mutant seeds typically elongates but frequently does not

rupture the pericarp to initiate precocious germination. Thus, most mutant seeds are

desiccation tolerant and are capable of germinating. The penetrance of the viviparous

phenotype is variable from season to season. In the green house growth conditions where

plants were well watered, the penetrance was relatively higher, as shown in Fig. 2-1, and

most of the vp14-2274 kernels that grew under these conditions also germinated fully to

form shoots and roots. Some shoots were strong enough to break the pericarp while

others were restrained by the pericarp. Unlike vpl (McCarty et al 1989), anthocyanin

pigmentation is not affected by the vpJ4 mutation, and if rescued, these germinated

kernels can develop into healthy, normal plants (Fig. 2-1B).

The B-A translocations in maize allow recessive genes to be located on the correct

chromosome arm in the Fl population. At the second pollen division, the supernumerary

B chromosome frequently nondisjoins so that one sperm cell of a pollen grain has two B

chromosome and the other has none. Various maize T-B lines were created such that the

B chromosome carries a specific segment translocated from an A chromosome. If a

mutant is located on a translocated segment, a cross between a mutant and the specific T-

B line will produce some embryos displaying the mutant phenotype in Fl because

nondisjunction will produce hemizygous embryos deficient in the tested wild type gene. In

crosses using homozygous vpl4-2274 and vpl4-3250 as female to a pollen parent that

carried the T-B1La B-chromosome translocation (Birchler 1995), a similar



















0


















Fig. 2-1. The phenotype of vp14. A. The vivipary of a homozygous vpl4-2274 ear grown in green house. B. These
viviparous kernels were rescued before desiccation and developed into healthy plants. The wild type plants were
non-segregating Vp14.






40


viviparous phenotype was uncovered, indicating that vpl4 is located on the long arm of

chromosome 1. To resolve the relationship of vpl4 relative to other viable viviparous

mutants of maize, homozygous vp14-2274 plants were crossed to vp8 (located on 1L,

Robertson 1955), vpIO (Smith and Neuffer 1992) and other unmapped, though

phenotypically similar viviparous mutants including vp*3286 and vp*3239 (McCarty

1995). The vp14-2274 mutant complemented each of these mutants, indicating that vp14

is a new locus. It was thus named viviparous]4 (vpl4).


vp14 Is a Developmentally Specific ABA Deficient Mutant

In the relatively harsh Florida Summer field growth conditions, the vpl4 mutant

plants showed no discernible phenotype compared to wild type plants (Fig. 2-1B). In

contrast to other known ABA deficient mutants such as abal of Arabidopsis (Rock and

Zeevaart 1991), sitiens andflacca of tomato (Taylor 1991), and aba2 of tobacco (Marin

et al. 1996), vp14 leaves were not prone to wilting. When grown in the green-house

where plants were watered daily, the detached leaves of vpl4 mutant seedlings showed a

distinctly greater rate of water loss in comparison to wild type siblings and an inbred W22

(Fig. 2-2D). This difference was detected within 5 minutes following leaf detachment,

indicating that the stomata in vpl4 leaves did not close as quickly as those of the wild type

leaves. However, determination of the bulk levels of ABA in leaves detected no significant

difference between vp14 mutant and wild type siblings (refer to Fig. 2-8B).This apparent

contradiction raised a question as to what role changes in the bulk ABA levels have in

regulating stomata. It has been previously documented that in stressed leaves stomata

close before any increase in bulk ABA can be detected (Walton et






41






A
80
60





S20





C &5 100
I 80
60







440
S20O
F 60 |

20










0 10 20 30 40 50 60

Minutes After Detachment

Fig. 2-2. Response of exogenous ABA on germination of embryos cultured in vitro
(A.B.C), and water loss rates of detached leaves (D) of wild type and vpl4 mutants.
Embryos harvested aseptically were placed on MS medium supplied with different
concentrations of ABA. Relative fresh weight (A), shoot(B) and root elongation(C)
as a percentage of a control (without ABA) were measured after four days. diamond:
W22, circle: wild type, square: vp14-2272, triangle: vp14-3250.






42


al. 1977). Harris and Outlaw (1991) reported that the increase of ABA in guard cells was

faster than changes in other leaf cells when stress was imposed, and more importantly, the

guard cell ABA increase correlated with the stomata aperture. Thus, one possible

explanation of the water loss phenotype in vpl4 was that it affected a small pool of ABA

that regulated stomata aperture. If so, this fast water loss phenomenon may be suppressed

in the field grown vpl4 plants by stress induced ABA since ABA can translocate easily.

To test whether vivipary of vpl4 can be inhibited by exogenous ABA, developing

embryos were cultured in vitro in media containing different concentrations of ABA.

Embryos 16 and 18 days after pollination were used because that stage allows fully

developed embryos to be isolated before any sign of axis elongation occurs. After 4 days,

growth parameters were measured. The vpl4 mutant embryos exhibited the same

sensitivity to ABA as did wild type (Fig. 2-2). Precocious germination of vpl4 and wild

type embryos was also inhibited by exogenous ABA at the same level, i.e. germination

was completely inhibited at 10"5 M ABA. Embryo growth was inhibited at ABA

concentrations greater than 10" M, consistent with previous reports on other ABA

deficient mutants (Robichaud et al. 1986). It was noted that root growth of vpl4 mutant

embryos showed an enhanced sensitivity to ABA inhibition compared to the wild type

(Fig. 2-2C). The severe allele vp14-2274 was somewhat more sensitive than the less

severe allele vp14-3250, in this respect. Together these data suggested that the vivipary of

vpl4 is likely attributed to a deficiency in ABA. However, the enhanced ABA sensitivity

of the mutant roots is not explained by this conclusion.

In collaboration with Dr. Jan A. D. Zeevaart's lab at Michigan State University,

the endogenous levels of ABA in embryos were determined using GC-MS. Embryos of






43


Table 2-1. ABA levels are reduced in vp14-2274 embryos.

DAP* WT sib vpl4-2274

ABA ng/g fr.wt. (%)

16 83.3 (100) 23.3 (27.8)

18 127.5 (100) 36.1 (28.3)

20 71.0(100) 43.7(61.5)

*DAP: days after pollination


16, 18, and 20 days after pollination (DAP) were sampled in order to span the time of

ABA regulated gene expression in embryos (McCarty et al. 1991). The vpl4-2274 embryo

contained only 28% of the wild type level of ABA at 16 and 18 DAP, and 61% as much as

wild type at 20 DAP (Table 2-1). Visible elongation of the embryo shoot axis in the

mutant is generally observed at 20 DAP. The significant level of residual ABA may

account for the relatively weak penetrance of the vp14-2274 phenotype. Lower but still

significant levels of residual ABA (-10% of wild type) are also evident in embryos of the

strongly viviparous carotenoid deficient mutants (e.g. vp5 ) of maize (Neill et al. 1986).

The extremely wilty mutant of tobacco aba2 also retains 23-48% of wild type ABA in leaf

tissues (Marin et al. 1996). The origin of this residual ABA is still unknown. ABA levels in

other vegetative organs such as leaves and roots appeared to be less affected since no

differences were detected in nonstressed root and leaf tissues. The total stress induced

ABA synthesis measured after 5 h in detached vp14-2274 leaves was about 25% lower

than the wild type siblings (data not shown). Together these data indicated that vpJ4 is an






44

ABA deficient mutant that predominantly affects the ABA synthesis in developing

embryos.


Analysis of ABA Biosvnthetic Pathway Intermediates in vp14 Embryos

ABA is proposed to be synthesized through a sequence of reactions that

includesepoxidation of zeaxanthin to violaxanthin, subsequent isomerization of all-trans

violaxanthin to 9-cis-violaxanthin, followed by an oxidative cleavage of 9-cis xanthophylls

to produce xanthoxin which is subsequently converted to ABA (Zeevaart and Creelman

1988, Walton and Li 1995). To locate the blocked step, the C4o carotenoids and

xanthophyll intermediates in vpl4 mutant embryos were analyzed by HPLC in

collaboration with Dr. Zeevaart's lab at MSU. No significant differences in the levels of

C4o carotenoid precursors were detected between vpl4 and the wild type embryos at 16

or 18 DAP (Fig. 2-3). The identities of those absorption peaks were calibrated either by

standard compounds or by published spectral data (Molnar and Szabolcs 1979, Rock and

Zeevaart 1991, Parry and Horgan 1992). The mutant embryos have comparable amounts

of zeaxanthin, violaxanthin, and slightly increased 9-cis violaxanthin and 9-cis neoxanthin

contents. The blocked step in vpl4 is clearly different than in other viviparous maize

mutants such as vp2, vpS, vp7 and vp9 which are depleted in those ABA synthetic

precursors (Neill et al. 1986). The vp14 mutation also differs from the ABA deficient

mutant of Arabidopsis, abal, or the equivalent mutant of tobacco aba2, which are

blocked in the epoxidation of zeaxanthin and have markedly lower violaxanthin, 9-cis

xanthophyll contents (Rock and Zeevaart 1991, Duckham et al. 1991, Marin et al. 1996).

The normal or slightly increased 9-cis violaxanthin and 9-cis neoxanthin levels also







45






46


indicated that vpl4 is unlikely to be blocked in the double bond isomerization step. These

results suggested that the blocked step in vpJ4 is downstream of the 9-cis xanthophylls in

the pathway. The remaining steps include oxidative cleavage of 9-cis xanthophylls to form

xanthoxin, xanthoxin conversion to ABA-aldehyde, and oxidation of ABA-aldehyde to

ABA.


Table 2-2: Xanthoxin conversion to ABA in vp14-2274 embryos.

Embryos (20 DAP) mg protein ABA(ng)

WT sib 100 18.9+/-2.1

vpl4-2274 100 35.0+/-1.4



Many of the known ABA deficient mutants are blocked in the conversion of ABA

aldehyde to ABA possibly because this step requires a molybdenum-Fe cofactor, the

synthesis of which requires several steps. These mutants include aba3 of Arabidopsis

(Leon-Kloosterziel et al. 1996, Schwartz et al. 1997),flacca, sitiens of tomato (Parry and

Horgan 1988), ckrl, ibal of tobacco (Parry et al. 1991, Bitoun et al. 1990), nar2a of

barley (Walker-Simmons et al. 1989), and droopy of potato (Duckham et al. 1989). A

convenient approach to assess the steps downstream of xanthoxin is to incubate xanthoxin

with cell free extracts and assay for production of ABA. The ability of vpl4 mutant

embryos to convert xanthoxin to ABA was determined in collaboration with Dr. Zeevaart.

In this assay, homozygous vp14-2274 embryos were found to be fully capable of

converting xanthoxin to ABA, and did so, surprisingly, at a somewhat rate higher than

wild type (Table 2-2). The reason why these enzymes are up-regulated in the mutant






47

remains to be determined, but these data indicated that vpl4 was not blocked in the steps

involved in xanthoxin conversion to ABA. Based on our knowledge of the proposed ABA

biosynthetic pathway, the results of enzymatic assay and HPLC analysis of precursors

suggested that vp14 is blocked in the cleavage of 9-cis xanthophylls to form xanthoxin.

Theoretically, the suggested vp14 block would be directly confirmed by the

detection of a depletion of the product xanthoxin. However, determination of xanthoxin is

experimentally difficult and may not be conclusive in any case because the normal level of

xanthoxin in all the plant tissues is reportedly extremely low (Nonhebel and Milborrow

1987, Parry et al. 1988 1990).These data provided a compelling though circumstantial

case, that vpl4 is the long-sought mutant blocked in the oxidative cleavage step of the

ABA biosynthetic pathway.


Molecular Analysis of viviparous 4

Because both alleles of vpl4 were isolated from maize strains containing active

Robertson's Mutator (Mu) transposons, it is possible that they were tagged by Mu

insertions. To clone the vpl4 gene, vpl4 was outcrossed to W22 in order to reduce copy

numbers of the Mu family elements. Statistically, one outcross to a low Mu copy number

strain such as inbred W22 is expected to reduce the copy number by 50%. Then a

segregating population was created from the outcrosses. Genomic DNA from lines

segregating the mutant phenotype upon selfing (Segregating, S) were analyzed by

Southern blot hybridization to compare lines that did not segregate (Nonsegregating, NS,

i.e. homozygous for wild type Vpl4). An internal Ttdllll fragment of the Mul was used

as a probe because it will detect most members of the Mu transposon family. Southern






48

blot hybridization identified a 2.5kb XhoI fragment that co-segregates with the mutant

(Fig. 2-4A). This fragment was cloned by screening a subgenomic X-gtl0 library

constructed using size selected DNA enriched in 2.5 kb fragments. Sequencing of this

clone confirmed the presence of a Mul insertion in that fragment. The Mul flanking

sequence (-~1 kb) was amplified by PCR and used subsequently to isolate overlapping wild

type genomic clones and cDNA clones from an embryo cDNA library.

To test whether or not the cloned sequence was from Vpl4, we analyzed vp14-

3250, an independent allele that was also isolated from an active Mu line. Southern blot

analysis of vp14-2274 and vp14-3250 DNA showed evidence that each contained a Mul

sized polymorphism that was not present in any of six possible progenitor strains that

founded the Robertson's mutator population (Fig. 2-4B). The corresponding 6.0 and

7.5kb EcoRI fragments were cloned from vp14-3250 and subsequent DNA sequencing

detected a Mul insertion approximately 1 kb upstream of the vp14-2274 Mul insertion in

the 7.5 kb fragment (Fig. 2-4C).

Restriction fragment length polymophorism (RFLP) mapping of Vp14 probes using

recombinant inbred populations (Ben Burr and Frances Burr, pers. comm.: Maize Genome

Center, University of Missouri) indicated that vpl4 is located approximately 50 map units

proximal to vp8 on 1L in the vicinity of the bz2 locus. This mapping location was in full

agreement with the TB translocation experiment which placed vpl4 on 1L. In addition, a

possible duplicate locus was detected and mapped to chromosome 5S. This finding is

consistent with a large body of RFLP evidence that 1L and 5S are duplicated






49


A B

4

s NS 2 00 Z

-s o

lkb(5S)
(, 7.5kb

S6.0kb (1L)





2.5 kb MO


Probe: Mul P1 P2

Nul
vp14-2274


E SNX X NX XE

ATG TAG

P2 P1
vp14-3250
Mut


Fig. 2-4. A. Souther analysis of a family segregating vp]4-2274 using a fragment
of Mu1 as a probe. DNA was pooled from several seedlings of each line. The arrow
points to a 2.5 kb fragment identified as co-segregating with the vp]4 mutant. B. DNAs
of six possible progenitors, vp]4 mutant and wild type plants were hybridized with
a Vpl4 specific probe (PI as indicated in C). On the right, W22 DNA was hybridized
by probe2 (P2) at low stringency. The closed arrows point to a 6 kb vp]4 (mapped on
IL) and a putative duplicate (11 kb mapped on 5S). Open arrows point to the closely
related fragments. C. The restriction enzyme map of a 6Kb EcoRI fragment containing
the Vp14 gene and the locations of the Mul insertion sites in each allele. The seguences
corresponding to two probes used are marked as PI and P2. E, EcoRI; N, Nodt; X, Xhol,
Nc, NcoI.









50











CAACAACAGACTACGGAGGACCCCG CCC AAAACCGATCCCCGCAAAAACCCCGCAGCGATCCACCAATTAGT 100

TGCAGGGTCTCGCCCCGCACKCTCTGTTTCCATACACCGGCACCTGCC GGTCAGGGCCGGGCCCTCCM ^ 200
q g I a p p t s v a i h r h I p a r s r a r a a n s v r f a p r a

CGTCAGCTCCGTGCCGCCCGOCGAGTGCGTCTCCACAAGCCCGTC;CGACTGCC TGCGCCGTCCAGGAAG CGCCA 300
v a a v p p a e c 1 q a p f h k p v a d 1 p a p s r k p a a i a v

CCAGGGCACGCCGCGXC(XGAGGAAAGCGGAGGCGGCAAGAGCAGC TCAACTTGTTCCACGC(;GGG CCGCGCTC CGG 400
p g h a a a p r k a e g g k k q I n I f q r a a a a a I d a f e g

GGTTCGG AGTCCTCGC ACGGGCTCCCAGCACGGCCGACCCGGCCGTGCAGATCGCCGGCAACGTT C 500
f v a n v 1 e r p h g 1 p s t a d p a v q i a g n f a p v g e r p

GCCCGTGCACGGCTCCCCGTCTCCGGCCATCCCGCCCTTCATCGACGGGGTCTACGCGCGCAACG CCC GACCGTCGCGGG 600
p v h e I p v s g r i p p f i d g v y a r n g a n p c f d p v a g

CACCACXTCTTCGACGGCCGGTGGTGCACGCGCTG C AAACGGCGCCGCCGAGTCCTACCGCCGC AG GGG 700
h h l f d g d g m v h a 1 r i r n g a a e s y a c r f t e t a r 1 r

GCCAGGAGMCGCGCGTCGGCCGCCCGTCT=T AAGGCCATTGGCGGCTGCACGGGCACTCCGSATCGCGTCAC 800
q a r a i g r p v f p k a i g e 1 h g h a g i a r l a l f y a r a

CGCGTGCGCTCGGGACCCTCGCCGGACCGGGTGGCAACGCCGGCCTCGTCTACTTCAACG GCC CATG GAGGA 900
a c g l v d p s a g t g v a n a g l v y f n g r 1 1 a m s e d d 1

CACTAC GTCGACCTCGAGACCGTCGGCCGCTACGACTTCCG AGCCGG CCATGATCG A 1000
p y h v r v a d d g d 1 e t v g r y d f d g ql g c a m i a h p k I

TGGACCCGGCCACCGGGAGCTCACGCGCTCAGCTACACGTCATCAAGAGGCCGTACCTCAAGTATTTTTCAGCCCC CC CCGA 1100
d p a t g l 1 h a ls y d v i k r p y 1 k y f y f r p d g t k s d

CGACGTGGAGATCCC(CTGGAGCAGCCCACGATGATCCACGACTTCGCCATCACCGAGAACTTCGTGGTTGTGCCCGACCACCAGTGGTGTTAAGT 1200
d v e i p l e q p t m i h d f a i t a n f v v v p d h q v v f k 1

CAGGAGATGCTGCGCGGCGGGTCGCCCTGTGCTGGAAGGAGGCGGCGGC AAACCGCGGACGCGTCGGAGATGG 1300
q e a I r g g a p v v 1 d k e k t s r f g v 1 p k h a a d a s e m a

CGTGGGTGGACGT4CCGGACTGCTTCTGCTTCCACCTGTGGAAC GTGGG0GGACGA0CGACGGGCGAGGTG TGATCGGCTCC A C 1400
w v d v p d c f c f h 1 w n a w e d e a t g e v v v i qg c m t p

CGCCGACTCCATCTTCAACGAGTCCGACGAGCGCCTGGAGAGCGTGCTAGAGATCC G CGGCCGGGCCGGCCCGCGCCCGG 1500
a d s i f n e a d e r 1 e s v I t e r 1 d a r t g r as t r r a v

CTGCGCCGTCGCAGCAGGAGAACCTGGAGGGGGCATGGTACCGCAACCTGCGGGCCGCGAGACGGTAC GGGAGGT 1600
1 p p s q q e n 1 e v g m v n r n 1 l g r e a r y a y l a v a e p w

GGCAAGGTCGGTTGCCAAGGAGGACCTGTC GGTGCGAGCTCACCAAGTGAGTA GAG GTCGGCGGACCGCCT 1700
p k e s g f a k e d I s t g e 1 t k f e y g e g r f g g e p c f v
TCCCATGGAXGGCCGCGGCCCACCCGCGCGGCGTGCTCACCTTCGTCCACGACGCGC GCCGGCACGTCGGAGCTACTTG 1800
p m d p a a a h p r g e d d g y v 1 t f v h d e r a g t s e 1 1 v

GTCAATCGCGAC-ATCCGGCTGGAGGCCACGGTTCACTGCGTCCCGCGTGCCCTTCGGCTTCCACGGCCCTTC 1900
v n a d i r 1 e a t v q I p r v p f g f h g t f i t g q e 1 e a

CCAGGCGGCTGACCGGCTCCACGTTTCTCGGAGGAGAACAGAGGAGCCAGCCTTGGATCAGGGGAGAAGCACCAGAGGGAGCCCAGACA CC 2000
q a a .

CCGGGGT CTTCTCTTTGCCTTTTTTTACATTATTTATTTCACTAGTGT T 2100
AATTGTATATGGCAGCTTAGAGAGAGAGAGAGATTAGTAGAAAGGCGCCCAGCTCGTAGCTTAACAGCTGGTGCTAGTATATACA 2200
TATTTTTrTTT CTTCTTGTCTTTACCCTTTCCCTTTGGATGACATGGATGTGCATCCAGCTC 2300
ACTGGGGTGTCTGGGATCTTGGTTGCTGTTACTTGCTCGCCATTGCCACCCATTGCTGCCGCT TTGTATAT 2400
GTGTTCCCGTGGCTATCGGTGCGTrGTACATTTGTTCACAGTATATGACTGATGGTATTAAAATAAGAACCGGTGACGGCTTCTGTTTCAAAAAAAAA 2500






Fig. 2-5. Sequence of a near full length Vpl4 cDNA and the deduced amino acid sequence
from an open reading frame. Vpl4 encodes a 604 amino acid protein with a computed

molecular mass of 65594 daltons. The putative chloroplast targeting transit peptide at the

N-terminus is underlined and a putative amphipathic helix is marked with a bold line.






51


segments in the maize genome (Helentjaris et al. 1988). Southern blot analysis using the

cDNA sequence as a probe suggests that vpl4 belongs to a small gene family. At

moderately high stringency, two major EcoRI fragments (6.0 and 11/3.5 kb) in inbred

W22 genomic DNA were detected corresponding to the 1L and 5S copies respectively.

Four to six fragments hybridized at lower stringency (Fig. 2-4B).


Vpl4 Is an Intronless Gene

To reveal the structure of the VpJ4 gene, a cDNA library was constructed in 7-

ZAP using PolyA enriched RNA isolated from the 18 DAP W22 inbred embryos. Ten

nearly fill length cDNA clones were recovered from 109 primary recombinants when the

cDNA library was screened with the 1kb flanking probe (P2). Sequencing of the cDNAs

and the 6.0 kb EcoRI genomic clone indicated that vpl4 contains no introns in the

transcribed region. A putative TATA box was located in the genomic sequence 80 bp

upstream of the longest cDNA. Analysis of the cDNA identified a 1812 bp open reading

frame predicted to encode a 604 amino acid protein with a calculated molecular mass of

65.5 kilo-daltons (Fig. 2-5).


VP14 Protein Is Related to a Cleavage Dioxygenase in Bacteria

The BLAST algorithm was used to search for related protein sequences in the non-

redundant NCBI protein database. VP14 had a strong similarity to lignostilbene

dioxygenase (LSD) of Pseudomonas paucimobilis (Kamoda et al. 1993a) and a weaker,

but significant similarity to human RPE65, a protein highly specific to the retinal pigment

epithelium of the eye (Hamel et al. 1993), and its mammalian homologs. In addition, two

related proteins are potentially encoded in the Synechocytis cyanobacterial genome. A






52


multiple sequence alignment of the proteins shown in Fig. 2-6 suggested strongly that the

LSD, RPE65 and VP14 are related proteins. The proteins align closely at their C-termini.

Several blocks of similar sequence in proteins are clustered around four conserved

histidines and one potentially conserved tyrosine residue. In dioxygenases of known

structure, conserved histidine and tyrosine residues are typical ligands for the non-heme

iron cofactor. It has been confirmed that the VP14 protein expressed in E. coli as a GST-

fusion protein contains significant amounts of bound non-heme iron (data not shown).

LSD also contains non-heme iron (Kamoda et al. 1993a, b). These results strongly

reinforce the conclusion that LSD and VP14 define a new class of non heme iron proteins.

The VP14 sequence includes an additional 100 amino acid N-terminal extension relative to

LSD and RPE65. This N-terminal peptide is consistent with the properties of a chloroplast

transit peptide (Cline and Henry 1996) and may serve to target VP14 to plastids where the

initial step of ABA synthesis is believed to occur (Zeevaart and Creelman 1988).

LSD catalyses an oxidative cleavage of the central double bond of lignostilbene to

form two molecules of the corresponding aldehyde (vanillin) as illustrated in Fig. 2-6.

This activity has been confirmed for the recombinant protein expressed in E. coli (Kamoda

et al. 1993a, b). This oxidative cleavage reaction is chemically analogous to the oxidative

cleavage of the double bond in xanthophylls that initiates ABA biosynthesis in plants (Fig.

2-7). These results together with the biochemical characterization, supported the

hypothesis that VpJ4 encodes the 9-cis xanthophyll cleavage dioxygenase of the ABA

biosynthetic pathway. Of course, still more convincing evidence would be the direct

demonstration of the VP14 cleavage activity in vitro.










UK11 --14*. PL:T e --------------- -- 1Le ,I-,lEp- 111 254 1u+u.+ M--... -.'-l -DL Y* FlllU v---




.-l 1 -:o --. .11 .. --1 321 . Ba I.-- --- L------ ----- Bm. ----- .C 1 K53
a_ 1*0 &; *#L W-llIE--SVL xt' ISO 306 113-- losII




-" 1 1 -T L S ....*L L ,,18 V.o-- ,,j ,,, 25 S 5 I 76 To. .,, u....' D ,---o -7


"L a -- '-- -- #A E'in'IB0 0 Y' l3---* hl- a m 4., 31. ... <0,. -------- <.It 1 Ex
Ito 1r r rf31 a L Vti W.C - PiT------- La 35 3. i iI------



37.2 ln L - ,sr 0 LL S2 2?a +m I
61i a* NS .+ -. - S L F "-------
















Fig. 2-6. Aligment ofVP1 with homologous proteins. LSD: lignostilbene dioyigenase 6 bacteria. RPE65: a 65 kd protein
specific to retinal pig epithelium cells of bovine. Syn(D0914, o) and Syn2 (90914, or. Four highly conserved

histidine residues were marked witgh asterisks.3 3 6 ILL S IS R 17 n----------
1*1r 240 60043 3 3- 1 133eV- 1114 404 13t 131-.I
nL 1O2 ----Ta U1Pr#. Sri L8II S45 Ve" M IEP ;?Gx I-







histine It -,or r d wr marL ked wiP ast .--------







histidine, residues wvere m~arked with asterisks.






54



VP14



02



HO 0
Sxanihoin








!\ I

OH

ABA

















Fig. 2-7. Proposed reactions catalyzed by VP14 and the reaction catalyzed
by lignostilbene dioxygenase (LSD) reaction (Kamoda : Saburi, 1993b).
-H






55


VpJ4 Expression Is Developmentally and Environmentally Regulated

The expression of Vpl4 was examined in embryos by northern blot analysis (Fig.

2-7A). A 2.6 kb transcript was detected in the developing embryos of wild type, i.e. W22

and the corresponding non-segregating Vp14-2274 (NS-2274) strain, whereas a series of

size altered transcripts were detected in the homozygous vp14-2274 mutant embryos. The

Vpl4 mRNA levels were greatly reduced in the homozygous vp14-3250 embryos, and a

similar pattern of altered sized transcripts was detected. The increased size of the mutant

vp14-2274 and vpl4-3250 transcripts (about 4.0 kb) is consistent with a transcriptional

readthrough of the Mul insertion (1.4 kb) in wild type Vpl4 gene (2.6 kb). The smaller

sized transcripts in the vpl4 mutant embryos may the aberrantly spliced products due to

the Mul insertion. Considering the evidence that Vpl4 belongs to a gene family and the

probe may hybridize to homologous sequences, the 4.0 kb transcript in vp14-2274 was

used to identify vpl4 specific RNA in the northern hybridization analysis. In Northern

analysis of other tissues, this mutant vpl4-2274 transcript (4.0 kb) was also detected in

roots of the mutant seedlings indicating that the VpJ4 gene is normally expressed in roots

as well as embryos (Fig. 2-8A). No Vpl4 message was detected in leaf tissues that were

grown in normal conditions.

Stress induction of ABA synthesis has been shown in a variety of plant species

(Wright and -iron 1969, Ingram and Bartels 1996), and this induction can be inhibited

either by transcriptional or by translational inhibitors (Li and Walton 1990), indicating

gene expression is involved in this process. Previous studies have suggested that the

cleavage reaction may be the key regulated step in the ABA biosynthetic pathway






56




A Embryo Let Root



ii 11








4 k4.0: kb::...::::.. .:,:
2.62 kkb A





B 2

N S N S N S
Probe:
4.0 kb

1 2.6 kb


P2 3.5 kb (





7 514 7 332 6 232
ABA (ng/g tissue)



Fig. 2-8. A. Northern analysis using P2 as a probe. The aberrant sized transcripts in vp]4
mutants were marked with open arrows. B. Water stress induction of Vpl4 expression in
detached leaves as probed with Vpl4 specific probe (P1) or induction of homologous genes
as probed with P2. A hybridization with a SUSI (sucrose synthase) probe was provided as
control. The ABA levels in those leaves were provided under the lanes.






57


(Zeevaart and Creelman 1988, Walton and Li 1995). Thus, it was predicted that

expression of the key 9-cis xanthophyll cleavage enzyme might be stress inducible.

To determine whether VpJ4 expression is environmentally regulated, detached

leaves were subjected to a water stress treatment at room temperature for 6 hours. This

resulted in a 40 70 fold increase of endogenous ABA levels in the vp14-2274 and wild

type leaves (Fig. 2-8B). Two probes were used in the northern blot hybridization of

stressed leaves, probe 1 (PI), a 3' untranslated sequence of Vpl4 that is specific to Vpl4

as evidenced by its specific hybridization to Vpl4 transcript (Fig. 2-4B), and broad

specificity probe 2 (P2) that can cross hybridize to homologous sequences in W22

genomic Southern (Fig. 2-4B). The Vpl4 message was dehydration inducible as indicated

by the strong signal of vp14-2274 transcript (4.0 kb) in the stressed vp14-2274 leaves and

the strong hybridization of Pl to the 2.6 kb transcript in stressed wild type leaves. Using

P2 at low stringency, Vpl4-related gene expression was detected. A 3.5 kb transcript

increases in stressed leaves, and a smaller VpJ4-related transcript was down regulated

upon stress. A probe from the Sus] (sucrose synthase) gene was a control for variation in

poly(A)-RNA loading. The loading is about equal for nonstressed and stressed samples

within each genotype, although it is variable among different genotypes. In any case, these

data indicated, as predicted, that expression of Vpl4 and related genes are induced by

water stress in plant leaves. This correlates with the stress induced ABA accumulation.

However, the vpl4 mutant only partially blocks stress induced ABA synthesis in leaves.






58




Discussion

Biochemical analysis shows that the vp14 mutant selectively blocked ABA

synthesis in developing embryos and suggests that the Vpl4 gene has a role in

developmental control of hormone synthesis in maize. Furthermore, the sequence

similarity of VP14 to LSD provided evidence that vpl4 encodes a dioxygenase enzyme

responsible for oxidative cleavage of 9-cis xanthophylls to xanthoxin.


Variable Dependence of Seed Maturation on ABA in Different Plant Species

ABA is a key regulator of seed maturation and dormancy (Koomnneef and Karssen

1994, McCarty 1995, Rock and Quatrano 1995). As documented by ABA deficient

mutants (e.g. vp5, vp14, etc) and the ABA insensitive mutant, vpl mutant of maize, a

failure to synthesize or to perceive the ABA signal results in a viviparous phenotype.

Viviparous seeds bypass the process of maturation, fail to acquire desiccation tolerance,

and initiate germination processes while kernels are still attached to the mother plant.

Although some key regulators of the seed maturation pathway such as the VP1

and ABI3 factors are highly conserved between maize and Arabidopsis, the dependence of

the seed maturation pathway on ABA signaling differs markedly between the two species.

In Arabidopsis, an ABA dependence of seed maturation is manifest only in mutant

backgrounds in which the ABI3 function is attenuated (Koornneef et al. 1994). The

mature seed phenotype of the abal mutant shows that the activity of ABI3 is sufficient to

result in a normal seed maturation in the absence of ABA. This view is further supported

by the fact that even the severe ABA deficient mutants of Arabidopsis are able to






59


complete seed maturation and form desiccation tolerant embryos (Giraudat 1995). In

tobacco, the ABA deficient mutant of aba2 which exhibits severe wilting, also completes

the normal seed maturation processes although a reduced dormancy was observed (Marin

et al. 1996).

In maize, the dependence of seed maturation on ABA appears to be relatively

tight. For the vpl4 mutant, a 70% reduction in bulk ABA levels in embryos during mid

embryogeny is evidently sufficient to disrupt seed maturation and thus cause an

incompletely penetrant viviparous phenotype. This finding is consistent with the evidence

that several strongly penetrant viviparous mutants (e.g. vp5) of maize have lower, but still

significant levels of residual ABA (-10 % of wild type) in the seed (Neill et al. 1986).

Overall it suggests that a threshold level probably greater than 30% of the wild type level

ABA is required to complete embryo development in vivo. Hence the viviparous, or more

precisely the elongated embryo shoot axis phenotype, provides a very sensitive screen for

ABA synthetic mutants in maize. This may have contributed to the isolation of vpl4 and

other new potentially ABA related mutants in maize (McCarty 1995). Together these

results suggest that dependence of the seed maturation pathway on ABA signaling is

variable among plant species, and in Arabidopsis the default activity of the ABI3 pathway

is normally sufficient to complete seed maturation.


Overlapping Sources of ABA Synthesis in Plant Development

The significant levels of residual ABA in embryos and normal levels in vegetative

tissues of homozygous vpJ4 mutants suggested a possibility of overlapping sources of

ABA. Three possible sources may account for the residual ABA in vpl4 embryos: 1) the






60

existing alleles may be leaky, 2) related genes may be expressed at low levels in the seed,

3) ABA may be transferred from the surrounding maternal tissues that express related

genes. Both alleles of vpl4 are caused by a Mul insertion, vpl4-2274 in the central

portion of coding region and vpl4-3250 in the 3' untranslated region (Fig. 2-4). It seems

unlikely that both alleles of vpl4 are leaky because in order to produce a functional VP14

protein, a perfect slicing of the Mul insertion is likely required from the mutated form of

the transcript. Northern blot analysis detected a 4.1 kb transcript that is consistent in size

with a Mul insertion in the normal 2.6 kb wild type VpJ4 transcript, and smaller than 2.6

kb sized transcripts that is presumably due to alternative splicing or premature termination

resulted from the Mul insertion (Fig. 2-8). A low level of Vpl4 related gene expression in

embryos is supported by the recovery of a Vpl4-like sequence from an embryo cDNA

library. Partial sequencing of this clone revealed that the potentially encoded protein is

highly homologous to VP14 (data not shown). However, a screen of another cDNA

library constructed from mRNA from W22 embryos at 18 DAP did not recover this

sequence among the ten VpJ4 clones that were isolated. One possible explanation is that

timing of expression of VpJ4 and this Vp]4-like gene differs during embryo development.

With respect to a third possibility, the normal ABA levels in vegetative tissues of

vpl4 mutants and the observation that mutant plants develop normally under field

conditions implies strongly that the VpJ4 gene accounts for only a subset of the ABA

biosynthetic activity in the plant. Partial suppression of the vp14 phenotype by ABA

transferred from maternal tissues might explain the variable penetrance of the viviparous

phenotype we observe under field conditions. Maternal ABA synthesis induced by stress

can significantly affect seed development in maize (Ober and Setter 1992). Thus, these






61

results suggest that embryo ABA may come from at least two sources, synthesis by the

embryo itself which accounts for a majority of the ABA, and ABA translocation from

other tissues such as leaves or roots. The endosperm evidently does not contribute

significant ABA to the embryos, consistent with the earlier conclusions of Robertson

(1952) and Neill et al. (1983). Using T-B translocations in maize, they created kernels

with embryos homozygous for vp5, but wild type endosperm. The phenotype of these

embryos were still viviparous and ABA deficient.

Several lines of data provided indirect evidence that functionally VP14 equivalent

genes are expressed in vegetative tissues of the plant. On moderate stringency Southern

blots, the 1 kb VpJ4 probe (P2) hybridized to nine EcoRI fragments of inbred W22 DNA

(Fig. 2-4A). Sequencing of a Vpl4 duplicated sequence (the 11 kb EcoRI fragment in Fig.

2-4A) indicated that it may potentially encode a protein with similar function to VP14 (see

chapter 5). The northern blot analysis of leaves also detected expression of related genes

(Fig. 2-8B). In addition, we have isolated several distinct, but related cDNAs from root

and leaf libraries (B. C. Tan and D. R. McCarty, unpublished data). The normal levels of

endogenous ABA in leaves and roots clearly indicate that ABA is synthesized in

vegetative tissues of the mutant. And the activity of ABA synthesis is only slightly affected

in stressed vpl4 -2274 leaves since ABA can still rise about 40 fold, although not as much

as in wild type (Fig. 2-8B). Data base searches identified several VpJ4 related sequences

in the rice, maize and Arabidopsis EST collections. For these reasons, we suggest that

Vpl4 belongs to a small family of differentially regulated genes that contribute to

developmental control of ABA biosynthesis in plants.






62


Role of Vpl4 in Regulation of Stomata in Leaves and in Root Development

Many of the characterized ABA deficient mutants display a wilty plant phenotype,

indicating that regulation of stomata closure is affected in these mutants (refer to Chapter

1 for details). Field grown vp14 mutants did not show any discernible plant phenotype in

addition to the seed vivipary. However, detached leaves of the vp14 mutant seedlings

grown in the greenhouse with abundant water supply show enhanced rates of water loss.

This feature suggested that stomatal closure is affected by the vpl4 mutation.

Measurement of the ABA levels in leaves of these plants detected no significant difference

between vp14 mutants and the wild type siblings, suggesting that the bulk pool of ABA is

unchanged. Northern blot analysis has confirmed that in addition to embryo expression,

Vpl4 is also expressed in roots and is stress inducible in leaves (Fig. 2-8), indicating the

potential involvement of VpJ4 in the ABA synthetic activity in these tissues. However,

VpJ4 may only account for a small set of the ABA detected in those tissues based on these

results. In many plant species, bulk ABA level increases dramatically in stressed leaves

(reviewed by Ingram and Bartels 1996). However, when the time course is taken into

account, the dosing of stomata generally occurs before any significant changes in ABA

content can be detected in the leaves. One possible explanation for this time discrepancy

may involve a rapidly localized accumulation of ABA in or around guard cells, either

through re-distribution of ABA among leaf cells or through highly localized de novo

synthesis of ABA, e.g. only in guard cells or the neighboring cells. In support to this

suggestion, a rapid increase in the ABA concentration of a single guard cell is correlated

with the closure of stomata in stressed leaves (Harris and Outlaw 1991). However, this

study did not identify the sources of the ABA. Thus, the role of Vpl4 in regulating






63

stomata closure may be explained by the existence of a highly localized pool of ABA in

leaves which specifically affects stomatal closure. ABA in this pool may result from re-

mobilization of ABA from other sites such as roots, or from a highly localized expression

of Vp14, in a limited number of cells, which may not be detected by northern blot analysis.

Vpl4 is expressed in normal roots which may be a significant source of ABA in leaves that

regulates stomatal aperture (Fig. 2-8). In several species, variations in stomatal

conductance are well correlated with the ABA present in xylem, and the initial stomatal

close occurs before the shortage of soil moisture causes any measurable change in the

water status of the leaves (Zhang and Davies 1989).

Another feature of germinating vp14 embryos is that their roots show a slightly

enhanced sensitivity to exogenous ABA compared to wild type. ABA is required for root

elongation in stressed conditions based on evidence that the ABA deficient vp5 mutant

showed poor root elongation compared to wild type (Sharp et al. 1994). The role of ABA

in root development and adaptation to environments is less understood. However, the root

phenotype of vpl4 may be explained by an increasing hormone sensitivity to compensate

for the reduced ABA synthesis.


Evidence that VP14 Is the 9-cis Xanthophyll Cleavage Dioxygenase

Deficiency of ABA in the vp14 embryos indicates that ABA biosynthesis is blocked

(Table 2-1). The existence of normal or slightly higher than normal levels of 9-cis

violaxanthin and 9-cis neoxanthin in vpl4 embryos suggests that steps leading to synthesis

of these two ABA precursors are not affected. In this respect, vpl4 is clearly different

from two related ABA deficient mutants, abal ofArabidopsis and aba2 of tobacco, which






64

accumulate zeaxanthin and deplete violaxanthin due to a block in the epoxidation step

(Duckham et al. 1991, Marin et al. 1996, Rock and Zeevaart 1991). A significant increase

of 9-cis xanthophylls does not occur in vpl4 embryos, possibly because ABA

concentration is normally much lower and only accounts for a small fraction of

xanthophyll synthesis (Parry and Horgan 1992). The fact that cell free extract from the

vpl4 embryos can convert xanthoxin to ABA ruled out the possibility that vpl4 is blocked

downstream of xanthoxin. Thus, by elimination, these data suggest that the blocked step in

vpJ4 is the oxidative cleavage of 9-cis xanthophylls to yield xanthoxin.

Two other possible approaches may provide further evidence for this conclusion:

1) confirmation that xanthoxin pools are depleted in the mutant, 2) direct measurement of

the enzymatic activity converting 9-cis xanthophylls to xanthoxin, in vitro. However, this

has proven technically difficult. Firstly, it was reported that xanthoxin levels are extremely

low in plant tissues, which in turn provided the evidence that downstream enzymatic

activities in converting xanthoxin to ABA may be constitutively active (Sindhu et al. 1987,

Li and Walton 1990). Experimentally, direct measurement of xanthoxin in developing

embryos is not feasible. With respect to the second approach, an in vitro assay of the

xanthophyll cleavage activity has not been developed. In vitro assays of the carotenoid

cleavage activity may be hindered by the extremely low abundance of the enzyme,

insolubility of the substrates, and complications arising from nonspecific cleavage

reactions by enzymes such as lipoxygenases (Creelman et al. 1992, Schwartz et al. 1997).

In fact, it is these difficulties and complications that have greatly hindered previous

progress in the search and demonstration of such a dioxygenase in the ABA biosynthetic

pathway. The incomplete characterization of the notabilis mutant of tomato (Parry and






65


Horgan 1992, Taylor 1991), has been due to a combination of its leaky nature and the lack

of more precise, direct approaches.

Molecular cloning of Vpl4 provides an independent line of evidence at the

molecular level. VP14 showed high similarity to a bacterial enzyme, lignostilbene

dioxygenase (LSD) (Kamoda et al. 1993). Based on the proposed ABA biosynthetic

pathway, cleavage of 9-cis xanthophylls was predicted to involve a dioxygenase type

enzyme (Walton and Li 1995, Creelman et al. 1992), which breaks the conjugated double

11,12 bond and produces xanthoxin and a C22 apo-aldehyde, each with an aldehyde group.

No such dioxygenases have been reported in the plant or animal kingdoms (Prescott and

John 1996). The reaction catalyzed by LSD is highly analogous chemically to the

proposed oxidative cleavage of 9-cis xanthophylls. Indeed, the bacterial LSD is the only

known enzyme capable of catalyzing the specific oxidative cleavage of a conjugated

double bond to form two aldehyde products and would, in the absence of other evidence,

provide a compelling model for the carotenoid cleavage enzyme (Fig. 2-7). Lipoxygenases

can also generate a series of aldehyde cleavage products of carotenoids in vitro, but do so

nonspecifically (Creelman et al. 1992, Schwartz et al. 1997). These results, together with

biochemical analyses of the vpl4 mutant, build a strong though circumstantial case that

VP14 encodes the cleavage dioxygenase in ABA synthesis. Recent analyses of the purified

recombinant VP14 have confirmed that it catalyzes oxidative cleavage of 9-cis

violaxanthin to xanthoxin in viro (next chapter).














Chapter 3
Viviparous] 4 Encodes a 9'-cis Neoxanthin/9-is Violaxanthin Dioxygenase of Abscisic
Acid Biosynthesis in Maize


Introduction

Abscisic acid (ABA), a sesquiterpene plant hormone, is involved in regulation of

many physiological processes throughout plant development, including germination,

transpiration (stomata aperture and root conductivity), seed maturation, dormancy and

tolerance to environmental stresses (Zeevaart and Creelman 1988, Walton and Li 1995,

Giraudat 1995, Ingram and Bartels 1996). ABA is a particularly interesting hormone with

regard to its regulation. The levels of ABA can rise or fall about two magnitudes within 4

to 8 hours in response to environmental and developmental changes (Walton and Li 1995).

In fungi, ABA is synthesized from Cs15 intermediates (farnesyl pyrophosphate) via a so

called "direct pathway"(Zeevaart and Creelman 1988). While in higher plants, ABA is

proposed to be synthesized from carotenoids. Blockage of carotenoid biosynthesis in

several viviparous mutants of maize (vp2, vp5, vp7 and vp9) also prevents ABA

biosynthesis (Neill et al. 1986, Taylor 1991). Incorporation of "0 from 02 in the -CHO

group instead of the ring of ABA strongly suggested oxidative cleavage of epoxy-

carotenoids as the origin of ABA as opposed to direct synthesis from a Cis intermediate

(Zeevaart et al. 1989). Biochemical analysis of stressed leaves indicated 9-cis violaxanthin

or/and 9-cis neoxanthin are probable substrates that are cleaved to produce xanthoxin (Li

and Walton 1990, Parry et al. 1990). Xanthoxin is subsequently converted to ABA


66






67

(Zeevaart and Creelman 1988, Walton and Li 1995). In support of this pathway, a number

of mutants that are blocked at several of the individual steps of the pathway have been

isolated (reviewed by Taylor 1991, Leon-Kloosterziel et al. 1996, Schwartz et al. 1997,

Marin et al. 1996). Many of the characterized ABA specific mutants are blocked in the

conversion of ABA aldehyde to ABA (Taylor 1991). An ABA deficient mutant of

Arabidopsis, abal, is shown to be blocked in the epoxidation of zeaxanthin (Duckham et

al. 1991, Rock and Zeevaart 1991). A zeaxanthin epoxidase was recently cloned in

tobacco and was shown to complement the abal mutant in transgenic Arabidopsis.

However, direct evidence for an enzymatic activity with the proposed specificity for

epoxy-carotenoid cleavage in ABA biosynthesis has not yet been found due to difficulties

arising from low abundance and liability of the enzyme (Walton and Li 1995, Schwartz et

al. 1997). Development of an in vitro assay has been hindered by the presence of

lipoxygenases and peroxidases, which oxidatively cleave carotenoids in a non-specific

manner (Creelman et al. 1992).

Biochemical characterization of a seed specific ABA deficient mutant of maize,

viviparousl4 (vpl4), indicates that it is blocked in the cleavage of epoxy-carotenoids in

the ABA biosynthetic pathway (Tan et al. 1997, Chapter 2). The existence of normal or

perhaps slightly higher levels of 9-cis neoxanthin and 9-cis violaxanthin in mutant embryos

indicates that vp14 is not blocked in steps prior to the formation of 9-cis carotenoids. The

extracts of the mutant embryos are fully active for enzymatic activities required for

conversion of xanthoxin to ABA. Together, these results suggested that vpl4 is impaired

in the oxidative cleavage step of epoxy-carotenoids in ABA biosynthesis. Experimental

results described in the previous chapter show that VP14 is related to a bacterial enzyme






68


lignostilbene dioxygenase (LSD) (Kamoda and Saburi 1993a, b) which catalyzes a

reaction that is strikingly analogous to the oxidative cleavage of 9-cis epoxy carotenoids in

the ABA biosynthetic pathway in plants. In this chapter, purified recombinant VP14 as

expressed in K coli is shown to specifically cleave 9'-cis neoxanthin and 9-cis

violaxanthin to produce the ABA precursor xanthoxin, and a corresponding C2s apo-

aldehyde. Consistent with other dioxygenases (Prescott and John 1996), recombinant

VP14 contains a non-heme iron which is critical to its activity.


Material and Methods


Expression of VP14 Recombinant Protein A fragment containing the coding

region and 3' untranslated sequences of the Vpl4 cDNA was amplified by polymerase

chain reaction (PCR) using a forward primer (5'-ATGCGGATCCATGCAGGGTCTCG-

CCCCG) and a T7 reverse primer. A BamHi site was engineered in the 5 primer very

close to the first ATG to facilitate cloning. The amplified fiagment was ligated into the

BamHI and EcoRI sites of pGex-2T expression vector to express a GST-VP14 fusion

protein in E. coli, JM109. Synthesis of the fusion protein was induced by adding 0.1 mM

IPTG to mid-log phase culture grown at 32 C in 2xYT-G medium supplemented with 0.4

mM Ferrous iron. GST-VP14 was purified by binding to Glutathione Sepharose 4B and

VP14 was isolated by digestion with thrombin according to the manufacture's instruction

(Pharmacia). Purified VP14 was quick frozen in aliquots in liquid nitrogen and shipped on

dry ice for enzymatic assay at Michigan State University.

Determination of Iron The iron content of the purified VP14 protein was

determined by a colorimetric method described by Perciva et a1.(1991). 30%






69

trichloroacetic acid was added to the protein solution to a final concentration of 5% to

release the non-heme iron. After centrifuigation to remove the precipitate, the 105 pil

supernatant was transferred to a new tube and 20 0p saturated ammonium acetate, 12.5 id

ascorbate (0.12M) and 12.5 Wl FerroZine (0.25M) reagent were added. After incubation at

room temperature for 30 minutes, absorbance at 562 nmn was determined. The protein

concentration was measured by Bradford method using BSA (bovine serum albumin) as of

standard (Bradford 1976) using bovine serum albumin (BSA) as a standard..

Preparation of the Substrates The substrates used for VP14 enzymatic assay test

were isolated by Steve Schwartz and Jan Zeevaart at Michigan State University as

described (Schwartz et al. 1997). Briefly, crude thylakoid membranes were isolated from

spinach leaves. The carotenoids were saponified and partitioned to ether and then resolved

on a semiprep lPorasil HPLC column (Waters). Fractions were collected and the

identities of prepared substrates were confirmed by spectrum analysis in comparison with

standards or published data (Molnar and Szabolcs 1979, Parry and Horgan 1991). 9-cis

zeaxanthin was prepared by iodine isomerization of the all-trans zeaxanthin (Zechmeister

1962).

Enzymatic Activity of Recombinant VP14 Protein The enzyme assays were

performed by Schwartz and Zeevaart. About 6 pg VP14 was incubated with each

substrate in a reaction buffer containing 100 mM BisTris, pH 6.7, 0.05% Triton X-100, 10

mM ascorbate, 5 PM FeSO4 and 1 mg/ml catalase. Assays were performed at room

temperature under red light to minimize photo-oxidation of the precursors and product.

Purified 9-cis neoxanthin was initially used to test for cleavage. The products were

analyzed by NP-HPLC and they were C25 apo-aldehyde fragment and xanthoxin. For






70

determination of substrate specificity, VP14 was incubated with purified all-trans/9-cis

neoxanthin, all-trans/9-cis violaxanthin and all-trans zeaxanthin / 9-cis zeaxanthin

suspended in Triton X-100 and ascorbate. After the reaction, the products were resolved

on a TLC plate and stained for aldehyde with 2,4-dinitrophenylhydrazine.

Mass-Spectrometrv Analysis of Reaction Products As described in detail by

Schwartz et al. (1997), the identity of xanthoxin was confirmed by GC-MS of the TMSi

derivative according to Gaskin and MacMillan (1991). The Cu2 apo-aldehydes were

analyzed by electron impact mass spectrometry with direct inlet sampling by Dr. Douglas

Gage at the mass spectrometry facility at Michigan State University. The expected

fragmentation pattern of both Cs product was derived from published spectra (Molnar

and Szabolcs 1979, Parry and Horgan 1991).


Results


Recombinant VP14 Protein Contains Non-heme Iron

Biochemical characterization and molecular analysis have provided compelling

evidence indicating that Vpl4 encodes the 9-cis epoxy-carotenoid cleavage dioxygenase

(Chapter 2, Tan, et al. 1997). To directly test the proposed activity of VP14, the coding

region of the Vpl4 cDNA was fused to a glutathion-S-transferase (GST) gene and

expressed in E. coli (Fig. 3-1A). SDS-PAGE analysis of the affinity purified proteins

detected a 90 kd fusion protein (GST-VP14) that was induced by IPTG. Cleavage of that

fusion protein by the site specific proteinase thrombin (LVPR4GS) produced two major

proteins, the GST domain and the VP14 recombinant protein which presumably carried

two additional amino acids (Gly-Ser-) at its amino terminus. Binding of GST to






71


A E
U) :(IE 0L




GST-VP14
VP14






GST

B

Fe2+ Standard (gg/ml)
0 8 16 24 32 40 48













GST VP14 VP148T Fd





Fig. 3-1. Expression and purification of recombinant VP14 from E. coli (A)
and determination of iron in the protein (B). The recombinant VP14 has extra
Gly-Ser- at its N-terminus; VP148T bears a 30 amino acid truncation in the
putative transit peptide at N-terminus. Fd stands for ferrodoxin.






72


glutathione sepharose 4B facilitated an easy purification of recombinant VP14. Because

proper folding of VP 14 may require the presence of ferrous iron, the E. coli cultures were

supplied with 0.4 mM iron. The purified protein was analyzed for iron using the cleavage

produced GST protein as a control. As shown in Fig. 3-1B, purified recombinant VP14

contains significant amounts of iron. Accurate determination of iron ratio in prepared

VP14 by FAB-MS in collaboration with Dr. B. W. Smith (Department of Chemistry,

University of Florida) revealed a stoichiometry about 0.3, which hints that either some

protein lost their iron or the active form of the protein is an oligomer. Later assays which

showed that addition of ferrous iron proportionally increased the activity of VP14 favored

the first possibility.


VP 14 Cleaves 9-cis Neoxanthin into Xanthoxin and a CZ Product

In collaboration with Steve Schwartz and Jan Zeevaart (Michigan State

University), the recombinant VP14 protein was tested for cleavage activity using 9-cis

neoxanthin as a substrate. The cleaved products were analyzed by HPLC and thin layer

chromatograph (TLC). The expected cleavage products, xanthoxin and the C25 allenic

apo-aldehyde, were identified by their UV/VIS absorption spectra and by mass spectra

obtained by Dr. Douglas Gage at the mass spectrometry facility at Michigan State

University (Schwartz et al. 1997). The fragmentation patterns for the epoxy- and allenic-

C25 cleavage products were nearly identical to published spectra for these compounds

(Molnar and Szabolcs 1979, Parry and Horgan 1991). Exact mass measurements

described in Schwartz et al. 1997 match the theoretical mass of the isomeric C25 cleavage

products from neoxanthin and violaxanthin (C25.HiO3) is 382.2508. The measured mass of






73


the C25 product from neoxanthin was 382.2498 with an error of -2.6 ppm. The measured

mass of the C2s product from violaxanthin was 382.2501 with an error of-1.8 ppm. Thus,

the cleavage is at the 11,12 double bond of 9-cis violaxanthin and 9-cis neoxanthin to

produce one molecule of xanthoxin (Cis) and one molecule of C25 apoaldehyde (allenic

and epoxy apoaldehyde). This cleavage of 9-cis epoxy-carotenoids by VP14 is well

matched with the proposed reaction in the ABA biosynthetic pathway. Xanthoxin was

detected with the absorbance at 280 nm, the 9-cis neoxanthin and the C25 cleavage

product were detected at 412 nm. As shown in Fig. 3-2, VP14 cleaves 9-cis neoxanthin to

produce two products, xanthoxin (C15) and a C25 apoaldehyde which was identified by

mass-spectrometry to be C2s allenic apo-aldehyde with molecular mass of 382.2498 with

an error of -2.6 ppm. Quantification of the products revealed an equimolar ratio of

xanthoxin and the C25 product, consistent with a specific cleavage at the 11-12 double

bond of the polyene chain. Non-enzymatic cleavage resulting from photo-oxidation or

Fenton reaction would result in random cleavage at different double bond positions

(Schwartz et al. 1997).

A number of factors were tested to optimize the VP14 cleavage activity and the

results indicated that molecular oxygen (co-substrate), ferrous iron, and a detergent were

all necessary for the in vitro cleavage activity (Schwartz et al. 1997). According to the

1'02 incorporation studies in stressed leaves (Creelman and Zeevaart 1984, Creelman et al

1987, Li and Walton 1987, Parry et al. 1988, Zeevaart et al. 1989), the molecular oxygen

is the co-substrate and each atom of 02 is incorporated into the -CHO group of the

reaction products, xanthoxin and C2s apo-aldehyde (Parry and Horgan 1991). Depletion of

02 in the reaction completely abolished the cleavage activity. A supplement






74


A



i~i'i.iii. i B." ... ..al eh
Xanthoxin

C2 allenic apo-
aldehyde



control +VP14 Xt.
Std.



B



9-c"s neoxanthin 9-cis neoxanthin
-VP14 |
Control




C2. apo-aldehyde xanthoxin







A412 profile A2o profile



Fig. 3-2. TLC analysis (A) and NP-HPLCanalysis (B) of the reaction products after
9-cis neoxanthin was incubated with recombinant VP14 purified from E. coli. The
TLC plate was stained by 2,4-dinitrophenylhydrazine to detect aldehydes. The test
was performed in collaboration with Steve Schwartz and Jan Zeevaart.






75


of ferrous iron, but not ferric iron, in the reaction mixture resulted in an about 3 fold

increase in the cleavage activity. Addition of divalent cation chelator EDTA led to a

complete inhibition of the cleavage activity, while removal of EDTA and addition of Fe2+

sufficiently restored the activity. These results support the idea that ferrous iron is a

critical co-factor of this epoxy-carotenoid dioxygenase. This is consistent with the

substoichiometric amount of iron detected in the purified recombinant VP14 protein. The

requirement for Triton X-100 in the assay is probably due to its role in solubilizing the

substrate. It may not reflect the in vivo reaction environment since carotenoids are shown

to exist in association with plastid membranes (Bartley and Scolnik 1995, Markwell et al.

1992). A slight increase in xanthoxin production by adding catalase is possibly caused by

reducing the nonenzymatic degradation of the substrate, e.g. by a Fenton reaction

(Schwartz et al. 1997). Ascorbate was added to the reaction to maintain Fe2+ in the

reduced state.


9-cis Configuration Defines the Substrate Specificity of VP14

Further studies of recombinant VP14 performed at Michigan State University

determined the substrate specificity of the cleavage reaction. The reaction products from

various carotenoid substrates were separated on TLC plates and stained with

2,4-dinitrophenyl hydrazine to detect aldehyde as shown in Fig. 3-3. The all trans isomers

of violaxanthin and neoxanthin were not cleaved by VP14. Xanthoxin and the predicted

Czs products are present only in reactions containing the 9-cis violaxanthin and 9-cis

neoxanthin as indicated by the xanthoxin standard.





76


Ptt

tS







VP14 + + + + + + Std
atZ 9cZ aN 9cV atN 9cN Xan




a "A 9cV



atN
9cN R1



atZ 9cZ R2


Fig. 3-3. Staining of a TLC plate that resolved the reaction products of specific
substrate incubated with recombinant VP14 (Top, adapted from Schwartz et al,
1997). The structure of the tested substrates were shown below. atV= all trans
violaxanthin, atN= all trans neoxanthin atZ= all trans neoxanthin, and 9c= 9-cis
isomers. R1 and R2 represented the corresponding residue in all trans configuration.






77

9-cis zeaxanthin was also cleaved, presumably yielding a Cis and a C25

apoaldehyde which suggested that the 9-cis configuration is a primary determinant of

VP14 substrate specificity. Cleavage of 9-cis zeaxanthin does not produce xanthoxin, and

this reaction is not known in plants possibly because that zeaxanthin exists only in the all

trans configuration. The 9-cis zeaxanthin used in above reaction was formed by iodine

isomerization of the all-trans zeaxanthin (Schwartz et al. 1997). Other mono-epoxy

carotenoids, antheraxanthin and lutein epoxide also exist in the all-trans configuration

(Parry et al. 1990) and, as expected, their all-trans isomers were not cleaved by VP14

(Schwartz, Tan, Zeevaart and McCarty, unpublished data). Thus, Vpl4 encodes the 9-cis

neoxanthin/9-cis violaxanthin dioxygenase of the ABA biosynthetic pathway in maize, and

this specific carotenoid cleavage activity has been directly demonstrated in vitro.


Discussion


Cloning and expression of VP14 reinforced the conclusion made by physiological

and biochemical characterization of the mutant that vpl4 is blocked in the oxidative

cleavage of oxygenated carotenoids. Furthermore, the activity of VP14 provided

convincing evidence to the identity of the VpJ4 clone. VP14 as expressed in E. coli can

specifically cleave the 11, 12 double bond of 9-cis oxygenated carotenoids into a C25 and a

C1s product, each with an aldehyde group (Fig. 3-3). Xanthoxin, the precursor of

subsequent ABA biosynthetic reactions, is formed only by cleavage of 9-cis violaxanthin

and 9'-cis neoxanthin. And one molecule of 9'-cis neoxanthin or 9-cis violaxanthin

produces one molecule of xanthoxin and one molecule of C25 apoaldehyde. Thus, it is in

good agreement with the proposed reactions of an oxidative cleavage of 9-cis






78



P-carotene



Zeaxanthin



Antheraxanthin






HO CM,






Ho0xyt -, at, at,
nil-iran& vioiaxthin









0, o + C
Dioxygcae 3
-HO atH aCC,










C. H .I l h





VPH4 as a dioxygenasae. he cleavd double bond by VP4 was marked by a
arrow. 9-cs neoxanthi vioand 9cs volath a the substrates to produce




oxin and only diff sidues were shown in e boxs.
cL-abscisic acid


Fig. 3-4. The ABA biosynthetic pathway in higher plants and the function of
VP14 as a dioxygenasae. The cleaved double bond by VP14 was marked by a
arrow. 9'-Cs neoxanthin and 9-cis violaxanthin are the substrates to produce
xanthoxin and only different residues were shown in the boxes.






79


xanthophylls in the ABA biosynthetic pathway in plants as shown in Fig. 3-4 (Zeevaart

and Creelman 1988, Walton and Li 1995). The fate of the Cs fragments in vivo is not

known. The affinity of VP14 for 9'-cis neoxanthin and 9-cis violaxanthin was not

determined. Because the former is more abundant than the latter in a number of species

including maize, tomato, bean and spinach (Li and Walton 1990, Parry and Horgan 1992),

9'-cis neoxanthin might be the major substrate of VP14. The decrease of 9'-cis neoxanthin

in stressed leaves of a number of species accounted for most of the increase of ABA (Li

and Walton 1987, 1990, Parry et al. 1990, Sindhu et al. 1988).

VP14 is also able to cleave 9-cis zeaxanthin, which lacks the epoxide on the ring

compared to the normal substrates, suggesting that the 9-cis configuration is a primary

factor that determines the substrate specificity. It is reasonable to speculate that both the

ionone ring and the distance between the 9-cis double bond and ring structure also

determine the substrate specificity. Inferred from that, all of the 9-cis isomers such as 9-cis

lutein, 9-cis antheraxanthin and 9-cis zeaxanthin may potentially serve as the substrates of

VP14. However, the 9-cis isomers of these carotenoids are not found in plants (Parry et

al. 1991). The 9-cis zeaxanthin used for cleavage reaction was made by iodine

isomerization of the all-frans-zeaxanthin in vitro (Zechmeister 1962). The predominant

cis xanthophyll in either light or dark grown bean leaves was found to be 9-cis neoxanthin,

and 9-cis violaxanthin only accounted for a small percentage (Li and Walton 1990). What

role, if any, light plays in the isomerization of all-trans carotenoids to generate 9-cis

isomers needs to be answered, however, roots grown in darkness normally contain 9-cis

carotenoids. The specific abundance of 9-cis violaxanthin and 9-cis neoxanthin in dark

grown leaves suggests involvement of a specific isomerase in the trans to cis conversion.






80


The rapid and specific increase of 9-cis violaxanthin and 9-cis neoxanthin in stressed dark

grown bean leaves is consistent with that hypothesis (Li and Walton 1990).

The lignostilbene dioxygenases from Pseudomonas and VP14 comprise a novel

class of dioxygenases that catalyze similar double bond cleavage reactions and generate

products with an aldehyde group. The conserved sequences have also been identified two

ORFs in the complete genomic sequence of the cyanobacterium Synechocystis. Functions

of the gene products have not yet been determined. However, related cyanobacteria are

known to synthesize at least two apo-carotenoids from 3-carotene, cyclocitral and (3-

ionone (Fresnedo et al. 1991). The demonstration of VP14' role in cleaving carotenoids

may clarify vitamin A biosynthesis in animals as it was proposed that a similar cleavage

reaction of 3-carotene to generates vitamin A This 15, 15-dioxygenase has been

controversial (Yeum 1995, Wolf and Phil 1995). Thus, conserved sequences may be useful

in identifying such carotenoid cleavage dioxygenases in the animal intestine.

The significant similarity of VP14 to RPE65 of the human retinal pigment

epithelium is intriguing in the light of recent evidence that REP65 may catalyze an

isomerization of all trans retinol to 11 -cis retinol (Micheal Redmond, NIH, unpublished

data). RPE65 was cloned and found to be expressed specifically in the retinal pigment

epithelium cells (Hamel et al. 1993). VP14 does not have any detectable isomerase

activity as indicated by testing all-trans substrates. However, some of the VP14

homologous genes may encode the isomerase as a large number of vpl4 related genes in

plants are detected by DNA hybridization and database searches. (Fig. 2-4, chapter 2). It

has been suggested that 9-cis isomerase, if exists, should also be stress inducible as

synthesis of 9-cis violaxanthin and neoxanthin is promoted upon stress (Li and Walton






81


1990). Further study of those homologous sequences may possibly resolve these

questions.














Chapter 4
Localization ofVP14, a 9-cis Violaxanthin/9-cis Neoxanthin
Dioxygenase Involved in ABA Biosynthesis


Introduction

Abscisic acid (ABA) is involved in the regulation of various plant processes

including seed maturation, dormancy and plant responses to a variety of stress conditions

(Ingram and Bartels 1996, McCarty 1995, Pena-Cortes and Willmitzer 1995, Rock and

Quatrano 1995). ABA is an apo-carotenoid derived from oxidative cleavage of 9-cis

epoxy carotenoids, a step that is widely believed to be the key regulated step in ABA

synthesis (Creelman et al. 1992, Zeevaart and Creelman 1988), whereas the enzymes

required for conversion of xanthoxin to ABA appear to be constitutively active in most

plant tissues (Sindhu and Walton 1987). Thus, xanthoxin does not accumulate to

significant levels in normal or water stressed plant tissues.

In maize, the cleavage dioxygenase was cloned by transposon tagging from an

ABA deficient viviparous 4 mutant that is blocked in that step (Tan et al. 1997, Refer to

Chapter 2). VP14 recombinant protein expressed in E. coli was shown to cleave

specifically 9-cis violaxanthin and 9-cis neoxanthin to produce xanthoxin, an immediate

precursor for ABA synthesis (Schwartz et al. 1997). ABA biosynthetic enzymes that

utilize carotenoids as substrates are believed to be localized in plastids since carotenoids

exist exclusively in plastid membranes. In tobacco, the epoxidase which converts

zeaxanthin to all frans-violaxanthin was found targeted to chloroplasts (Marin et al.


82






g3

1996). It appears that the cleavage product, xanthoxin, can be converted to ABA in the

cytosolic extraction, indicating the subsequent enzymes may be cytosolically localized

(Sindhu and Walton 1988, Walton and Li 1995). The substrates for VP14, 9-cis

violaxanthin and 9-cis neoxanthin are abundant in both thylakoid membranes and envelope

membranes of plastids and are not water soluble (Rock and Zeevaart 1991, Li and Walton

1990b). The localization of VP14 protein is of great importance in terms of understanding

regulation of the cleavage activity and substrate accessibility following stress induction

(Tan et al. 1997). Using an plastid protein import system, we show that VP14 is targeted

to chloroplasts and predominantly exists in the soluble stroma fraction. A small fiaction of

VP14 was also found associated with thylakoid.

Proteolysis is an essential process for many aspects of plant physiology and

development. These include housing keeping functions such as removing misfolded

proteins, generating peptide hormones, processing of organellar and secreted proteins

through specific cleavage, and programmed cell death (Vierstra 1996). Protein turnover

may also play an important role in both development and environmental regulation of

metabolism by reducing the abundance of key enzymes and regulatory proteins. Because

VP14 catalyzes a highly regulated step in ABA biosynthetic pathway, and is responsible

for the rapid and transit increase of ABA synthesis within hours in water stressed tissues,

post transcriptional regulation involving proteolysis is likely a means of regulatory

mechanism. Application of cytosolic protein synthesis inhibitor cycloheximide to

unstressed bean leaves caused a 50% decrease of ABA within 4 hours after treatment,

while treatment of stressed leaves with cycloheximide caused a faster decrease in ABA

levels (Li and Walton 1990a). Based on these data, it was predicted that the key steps






84


controlling ABA synthesis are likely to have a half life of several hours. The enzymatic

cleavage of 9-cis xanthophylls is a good candidate for the regulated step since the

downstream enzyme activities are not affected by stress treatment (Zeevaart and Creelman

1988). To address this possibility, the post-translational regulation of VP14 protein was

studied in chloroplast. It was found that VP14 is quickly turned over with a half life of less

than 30 minutes after import into chloroplast.


Materials and Methods

Plasmid Constructs Higher rates of in vitro transcription and translation efficiency

have been achieved by shortening the space between the start code and SP6 promoter

(Tranel et al. 1995). To remove the 5' untranslated sequence of Vpl4, the cDNA was

amplified from the first ATG with a engineered BamHI site to facilitate the cloning

(forward primer 5'-ATGCGGATCCATGCAGGGTCTCGCCCCG, reverse primer T7).

The resulting product ligated into pGem-3Z (Promega) under control of the SP6

promoter. The plasmid was named pSP6-Vp]4.

In Vitro Transcription and Translation pSP6-Vpl4 DNA was linearized by

digestion in the 3' linker region with EcoRI. The linear DNA was purified from an agarose

electrophoresis gel by electroelution. The in vitro transcription reaction was carried out as

described by Cline et al. (1993).The transcription reaction contained 5 mM DTT, 50

units/ml RNasin (Promega), 0.5 mM NTP's, 50 p.g/ml BSA, 1.5mM diguanosine

triphosphate [m7G(5')ppp(5)G, Pharmacia]. 50 jig linearized plasmid DNA/ml, 500

units/ml SP6 DNA polymerase in 1 x SP6 DNA polymerase buffer (Promega). The

reaction was incubated at 40 C for 60 minutes. The messenger RNA was precipitated






85


with ethanol after removing the proteins by phenol/chloroform extraction. The messenger

RNA was quantitated and tested for in vitro translation efficiency in a wheat germ cell free

system (Promega).

A typical 50 tl translation reaction included 30 tl of premade wheat germ cell-free

extract, 5 tJ premix of all amino acids except leucine, 5 tl 5xBuffer, 5 Wl in vitro

synthesized RNA (-~ 40 ng), 5 ld RNase-free water containing 3H-leucine (3000 Ci/mol,

DuPont). The mixture was incubated at 25 C for 1 hour and placed on ice to stop

translation. Immediately before the import assay, translation product was diluted with

equal volume of 60 mM leucine in 2x import buffer (IB, 50 mM hepes/KOH pH 8 and

0.33 M sorbitol).

Isolation of Pea Chloroplasts Leaves harvested from dwarf peas grown in

vermiculite for 10 days at 20 C, 150 pE fluorescent light were homogenized in a buffer

containing 50 mM Hepes/KOH (pH 7.5), 0.33M sorbitol, 1 mM MgC12, 1 mM MnCI2. 2

mM EDTA, 5 mM sodium ascorbate and 0.2% BSA. Homogenate was filtered through

one layer of miracloth and centrifuged at 2000 g (3200 rpm) for 3 minutes in a Beckman

swinging bucket rotor. The pellet was resuspended in grinding buffer (GR) and loaded

onto a precentrifuged (50,000 g, 40 minutes) 35% Percoll gradient in GR buffer

supplemented with 10 mM glutathione. The gradient was centrifuged at 2000 g for 15

minutes in a swinging bucket rotor and a lower band which contained the intact

chloroplasts was removed and diluted 3-fold with IB. The chloroplasts were pelleted at

1500 g and finally resuspended in IB at 1 mg chlorophyll/ml as determined

spectrophotometrically.






86


Protein Import Assay The protein import assay was carried out according to

Cline et al. ( 1993) in the 1 x LB. All the import reactions were carried at 25 C under light

with 5 mM Mg-ATP except where ATP and light dependent import was tested. The

unimported proteins were removed by washing the chloroplasts with IB buffer and the

surface adhering protein removed by treatment with thermolysin. The chloroplasts were

repurified through a 35% percol cushion, centrifugation and rinsed once with IxIB. The

chloroplasts were lysed by suspension in a hypertonic solution containing 20 mM EDTA.

For subfractionation of membrane and soluble compartments, chloroplasts were lysed in

HKM on ice for 5 to 10 min. Lysis was monitored by using percoll cushion centrifugation

to separate the unlysed chloroplasts. Lysed chloroplasts were collected by centrifugation

at 4,200 rpm for 8 min. The pellet fraction contained thylakoids and was rinsed twice with

IxIB before lysis for proteins. The supernatant fraction was further centrifuged at 18,000

g for 2 hours at 4 C to pellet the envelope membrane fraction, the supernatant fraction

collected as the stroma fraction. All the subfractions were lysed in 20 mM EDTA and

denatured in 15% SDS at 80 C for 10 minutes, then analyzed in 12% SDS-PAGE. The

gel was treated with DMSO and PPO/POPOP then dried. The radioactive protein was

detected fluorography (Cline et al. 1993).


Results

The N-Terminal Sequence ofVP14 Has Features of a Chloroplast Transit Peptide

VpJ4 encodes an 11, 12 9-cis epoxy carotenoid dioxygenase that cleaves 9-cis

violaxanthin and 9-cis neoxanthin to produce xanthoxin, a precursor of abscisic acid

(ABA) as indicated in Fig. 4-1B (Schwartz et al. 1997, Tan et al. 1997). This cleavage






87








A
MQGLAPPTSVSIHRHLPARSRARASNSVRFSPRAVSSVPP
AECLQAPFHKPVADLPAPSRKPAAIAVPGHAAAPRKAEGG
KKQLNLFQRAAAAALDAFEEGFVA ......











9'-civiolaxanthin
C25 ,poaddhyde



SC25 allenic
xandwun qoadshyde


9'-dsaneixntdn











Fig. 4-1. The sequence of 104 amino acid N-terminus extension of VP14
that does not align to LSD and RPE65 (A) and the reactions catalyzed by
VP14 in ABA biosynthesis (B). The putative transit peptide region was
underlined with single line and a putative amphipathic a-helix region was
underlined with a double line.






88


reaction is the first committed step in the ABA biosynthetic pathway in plants, and

substantial evidence has indicated that it is also the regulated step controlling synthesis of

ABA. VP14 is homologous to a bacterial lignostilbene dioxygenase (LSD) which catalyzes

a very similar cleavage of a conjugated double bond to produce products with -CHO

group (Kamoda et al. 1993a, b), and to a human eye retinal epithelium specific protein

RPE65 (Hamel et al. 1993). A three way alignment of VP14 with LSD and RPE65

indicates that VP14 has an -100 amino acid extension at the amino terminus as shown in

Fig.4-1A. Blast searches using the extension sequence did not reveal any significant

similarity to known proteins. Removal of 30 amino acids from N-terminus does not affect

the activity (data not shown), suggesting that this extension may have other functions.

Although no conserved sequence has been found among the chloroplast targeting

transit peptides, there are some features shared by the stroma-targeting domains(STD),

such as being rich in hydroxylated residues and lacking in acidic residues (Cline and Henry

1996). The sequence of the first 40-50 amino acids at N-terminus were consistent with a

STD. It contains 20% percent serine, compared to 5 % for the whole protein, and is

positively charged (pH 12.10) in contrast to the full protein which has an isoelectric point

of pi 5.72. Similar features are also observed in the N-terminus of ABA2 in tobacco, an

epoxidase that is upstream of VP14 in the ABA biosynthetic pathway (Marin et al. 1996).

Also noticed was a putative amphipathic a-helix that was highly conserved among a maize

homolog and Arabidopsis homologs (Deng and McCarty, unpublished data). Amphipathic

a-helices are potentially involved in protein-protein interactions or anchoring protein to

membranes (Carr et al. 1993).






89

Chloroplast proteins synthesized in cytoplasm can be tested in an in vitro system in

which proteins are radiolabeled during in vitro translation and incubated with intact

chloroplasts. Chloroplast targeted proteins will use the protein trafficking pathway to

reach their organellar destination which is inaccessible to proteinase treatment. The

specific protein can be detected by analyzing the radioactivity of proteins isolated from

chloroplasts after proteinase treatment. VP14 was tested for trafficking to chloroplasts

using an import assay. To do so, full length Vpl4 cDNA was placed under the control of

SP6 promoter in pGEM-3Z (Promega) plasmid, then transcribed and translated in vitro

using SP6 DNA polymerase and wheat germ cell free systems, respectively. Because the

full length Vpl4 cDNA has a 100 nucleotide 5' untranslated sequence in front of the first

ATG, a poor translation efficiency was detected. In order to enhance translation

efficiency, the 5' UTR was removed by PCR amplification and by subcloning of Vpl4

coding region into the pGEM-3Z plasmid. Upon incubation of in vitro translated VP14

with the intact pea chloroplasts, the VP14 protein was imported into a chloroplast

compartment that is not accessible to the proteinase thermolysin (Fig 4-2A). In addition,

upon import into chloroplasts, VP14 was processed to a smaller molecular weight with the

loss of about 4.5 kd. This is consistent with proteolytic cleavage and removal of a N-

terminal peptide. The control protein, light-harvesting-complex-protein (LHCP), a well

characterized chloroplast thylakoid protein that is involved in photosynthesis, is processed

by losing its transit peptide to produce a mature protein of about 28 kd. As indicated in

Fig.2B, similar to pLHCP, import of VP14 requires energy provided by either ATP or

light which further confirmed the chloroplast localization of VP14. Thus, these data

indicate that VP14 is localized in chloroplasts.







90













VP14 LHCP VP14 LHCP

Chioropaat + + + + ATP + + + + -
"henmolysin + + hv + + + +


pVPI4, .
VP14






pL-CP














Fig. 4-2. Import of VPl4 into chloroplasts and its energy dependence. 3H-leucine
labeled precursor VP14 or LHCP were incubated with c hloroplasts and then
treated with thennolysin in the presence or absence of ATP or light.






91

VP14. a Soluble Protein Is Localized in Stroma and Thylakoid Membrane

The next question is where inside the chloroplast VP14 is localized. Chloroplasts

have six different compartments including outer and inner envelope membranes, envelope

interspace, stroma, thylakoid membrane and thylakoid lumen. Substrates of VP14, 9-cis

neoxanthin and 9-cis violaxanthin, have been reported in both thylakoid membranes and

envelope membranes (Li and Walton 1987, Zeevaart and Creelman 1988, Parry and

Hogan 1991). They presumably exist in the membranes through interaction with

membrane associated proteins, as found in a carotenoid-protein complex (Markwell et al.

1992). The sub-organelle localization of VP14 reflects to some extent the likely site of

interaction between substrates and enzyme. To further locate VP14, chloroplasts were

subfractionated into three fractions, the soluble fraction which presumably contains the

stromal proteins and interspace, the thylakoid fraction that includes all thylakoid

membrane and lumen proteins, and envelope fraction which includes outer and inner

envelope proteins. VP14 was found predominantly in the soluble fraction, suggesting that

it may be a stroma protein (Fig. 4-3). Moreover, there was a significant amount of VP14

associated with thylakoid. Further treatment of the thylakoid with thermolysin digested

VP14, indicating that the protein was bound on the outer face of thylakoid membrane.

LHCP, a thylakoid membrane protein was partially protected from thermolysin digestion

consistent with the evidence that the protein is embedded in the membrane so that it is not

accessible to an external proteinase (Fig. 4-3).


Evidence of Fast Turnover of Imported VP14

In the import assays, it was noticed that small sized radiolabeled proteins were also






92









VP14 LHCP
TP E S T Tt TP E S T Tt




















Fig. 4-3. Localization of mature VP14 inside the ch.oroplast. After
incubated with I H-pVPI, purified chiiiiiiiiiiiiiiiiiiiiiiiioroplasts were lysed andiiiiiiii
inbated with 3H-pV4, puiied ................... .erelysed..














separated into subfractions of stroma (S), thylakoids (T) and
envelopes (E). An equal amount of thylakoids were treated with
thermolysin (Tt).






93


detected in the soluble fraction, which we hypothesized might be degradation fragments of

VP14. Because stress induction of ABA synthesis is transient, the presumed key regulated

step, the dioxygenase may be subject to rapid regulation at both the transcription level and

protein level. We have detected using both monoclonal and polyclonal antibodies that

VP14 protein exists at very low levels in even stressed leaves of maize in which Vpl4

transcripts were found increased significantly and ABA levels were increased about 40-

fold (Data not shown). One possibility is that VP14 may undergo fiast turnover such that

the steady state level is low. Using the import assay, we measured the rate of turnover of

mature VP14 in isolated chloroplasts compared to a control thylakoid protein, LHCP.

After the import, precursors were removed by thermolysin which was subsequently

removed by repurifying the chloroplasts. The chloroplasts were incubated at 28 C under

light for different periods (Materials and Methods), the level of the imported mature VP14

was determined. As shown in Fig. 4-4A, level of imported VP14 was quickly declined

within 60 minutes, while no significant decrease in LHCP was observed over the same

period time (data not shown).

It has been reported that mis-targeted and misfolded proteins are likely to be

degraded in chloroplasts (Halperin and Adam 1996, Vierstra 1996). As VP14 was found

both in the soluble and thylakoid fractions, we examined the degradation pattern in both

fractions. The degradation of VP14 was about the same in either the soluble fraction and

the thylakoid fraction (Fig. 4-4B). Because VP14 requires a non-heme iron in its active

form (chapter 3), we considered the possibility that iron may be needed to stabilize the

protein. To test this, ferrous iron and ferric iron were supplied to the chloroplasts before






94






A
Time after import (min)
TP 0 20 40 60








B Envelope Stroma Thylakoid
0 20 40 60 0 20 40 60 0 20 40 60mim


-iron




Fe2+












Fig. 4-4. VP14 is rapidly degraded after import into the pea chloroplasts.
VP14 was translated in vitro in presence of 3H-leucine and imported into
chloroplasts. Then the precursor was washed off and the chloroplasts were
tested for time-course of transport. A. whole chloroplasts; B. fractionated
chloroplasts; TP=translation precursor.




Full Text

PAGE 1

MOLECULAR ANALYSIS OF AN ABSCISIC ACID DEHCIENT MUTANT, VIVIPAR0USJ4 0 MAIZE BY BAO-CAI TAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1997

PAGE 2

ACKNOWLEDGEMENTS I am deeply grateful to Dr. Donald R. McCarty, chairman of my supervisory committee for his valuable advice, guidance and encouragement throughout the course of this study. I want to thank members of my graduate committee Drs. L. Curtis Hannah, Karen E. Koch, William B. Gurley, and Nigel Richards for their valuable suggestions to the completion of this study. Specially, I want to thank Dr. Kenneth Cline for his help in localizing VIVIPAROUS 14 in choloroplasts. I vAsh to thank Dr. Phil Becraft and Dr. Ralph Henry for valuable discussions. I also wish to thank Mr. Xian-Yue Ma, Dr. ChienYuan Kao, Dr. Masaharu Suzuki, Dr. Shailesh Lai and Dr. Joseph Cicero for their help. I want to thank Ms. Janine Shaw, Mr. Mike McCafFery, Mr. Wayne Avigne, and Mr. Dale Haskell for allowing me to frequently use their lab equipment. Fmally, I owe thanks to my wife, Wen-Tao Deng, for her constant support on every aspect during this study. ii

PAGE 3

TABLE OF CONTENTS page ACKNOWLEDGMENTS u ABSTRACT iv CHAPTERS L REVffiW OF LITERATURE 1 ABA Biosynthesis in Fungi and Plants 1 Major Functions of ABA in Higher Plants 11 The ABA Related Mutants in Higher Plants 23 Summary and Future Perspectives 27 2. MOLECULAR ANALYSIS OF VIVIPAROUS J4, A DEVELOPMENTALLY SPECIFIC ABSCISIC ACID BIOSYNTHETIC MUTANTS OF MAIZE 31 Introduction 31 Materials and Methods 34 Results 37 Discussion 58 3 VIVIPAROUS J 4 ENCODES A 9-CIS NE0XANTHIN/9-C/^ VIOLAXANTHIN DIOXYGENASE OF ABA BIOSYNTHESIS IN MAIZE 66 Introduction 66 Materials and Methods 68 Results 70 Discussion 77 4. LOCALIZATION OF VP14, A SPECIFIC 9-CIS EPOXY CAROTENOID DIOXYGENASE INVOLVED IN ABA BIOSYNTHESIS 82 Introduction 82 Materials and Methods 84 Results 86 Discussion 95 iii

PAGE 4

5. A FUNCTIONAL DUPLICATE OF VIVIPAR0US14 CONFERS A DISTINCTIVE TISSUE-SPECIFIC EXPRESSION PATTERN IN MAIZE 98 Introduction 98 Materials and Methods 99 Results 100 Discussion 109 6. SUMMARY AND CONCLUSIONS 115 LIST OF REFERENCES 1 16 BIOGRAPfflCAL SKETCH 127 iv

PAGE 5

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial FulfiUment of the Requirements of the Degree of Doctor of Philosophy MOLECULAR ANALYSIS OF AN ABSCISIC ACID DEFICIENT MUTANT V1VIPAR0US14 IN MAIZE By BAO-CAI TAN December 1997 Chairman: Donald R. McCarty Major Department: Plant Molecular and Cellular Biology Abscisic acid (ABA), a key regulator of seed maturation, dormancy and stress responses throughout plant development, is proposed to be synthesized by oxidative cleavage of 9-cis epoxy-carotenoids. Carotenoid cleavage is the first committed and presumed regulatory step in ABA biosynthesis. In this study, a new ABA deficient mutant viviparousN (ypl4) was identified in maize. Although vpl4 plants were fially viable and non-wilting in field, detached leaves of non-stressed mutant seedlings grown in the greenhouse showed markedly higher rates of water loss than the wild type. Mutant embryos exhibited normal sensitivity to exogenous ABA, and the ABA content of developing mutant embryos was 70% lower than the wild type, indicating a defect in ABA biosynthesis. The HPLC analysis of the vpl4 embryos detected no significant changes in the epoxy-carotenoid precursors of ABA, and the cell extracts of vpl4 embryos efficiently

PAGE 6

converted xanthoxin to ABA, suggesting that vp]4 is blocked in the cleavage of 9-cis epoxy carotenoids. The VpI4 gene was molecularly cloned by transposon tagging. The Vp]4 mRNA was detected in embryos and seedling roots, but not in nonstressed leaves. Southern blot analysis and cloning of related sequences indicated that VpJ4 belongs to a gene family. Several related gene sequences were cloned and the DNA sequenced. A VpI4-\ike duplicate was mapped to chromosome 5S and found to potentially encode a protein with 93% amino acid identity to VP 14, suggesting that it may have a function equivalent to VP 14. The Vp]4 mRNA and related transcripts were induced by water stress in leaves. The VP 14 amino acid sequence showed significant similarity to lignostilbene dioxygenase of bacteria and RPE65, a protein found in the mammalian retinal pigment epithelium. Purified recombinant VP 14 as expressed mE. coli specifically cleaves 9'-cw-neoxanthin/9c/j-violaxanthin to produce xanthoxin and a C25 allenic/epoxy apo-aldehyde. The implications of this demonstration of specific enzymatic cleavage of carotenoids for understanding the mechanism of vitamin A biosynthesis in mammalians are discussed. Evidence that the VP 14 protein is localized in the chloroplast of plant cells was obtained by an in vitro chloroplast protein import assay. Precursor VP 14 was imported into a soluble fi'action of chloroplast lysate and processed. A significant amount of imported VP 14 was also found associated with thylakoid membrane with a topology of facing stroma. Furthermore, VP 14 was subject to a fast turnover with a half life of about 30 minutes, suggesting that degradation may play a role in controlling the enzyme abundance. vi

PAGE 7

CHAPTER 1 REVIEW OF LITERATURE The discovery of the plant hormone, abscisic acid (ABA), was associated with two independent Hnes of research, one involved isolation of substances which accelerate leaf abscission in cotton by Addicott's group, the other included a search for substances which cause bud dormancy in woody plants by Wareing's group. The search for an abscission accelerator from young cotton fruits identified a compound "abscisin H" (Ohkuma et al. 1963). The search for a dormancy inducer from Betula pubescem led to the discovery of a substance, named "dormin" (Eagles and Wareing, 1964). Soon thereafter, the two substances were found to be chemically identical in structure (Comforth et al. 1965). Since then, a uniform name was given as abscisic acid. The naturally occurring ABA is the S-enantiomer. Whereas the R-enantiomer has been reported to be biologically active in most cases, it is inactive in regulating stomata aperture (Addicott 1983). The biologically active configuration of ABA is 9-cis ABA, and all trans ABA is biologically inactive. ABA Synthetic Pathways in Fungi and Plants Although its structure was elucidated more than 30 years ago (Comforth et al. 1965), the ABA biosynthetic pathway had remained elusive until the late 1980's. Studies of the ABA biosynthetic pathway in higher plants have encountered great difficulties compared with the similar studies on other plant hormones such as ethylene and 1 1

PAGE 8

cytokinins. ABA exists at very low concentrations (10" 10 M), even in water stressed tissues and maturing embryos where it is around 10"^ M. Thus, analysis of ABA was technically challenging. Furthermore, the major obstacle is the poor incorporation of isotope labeled precursors such as ^^C-MVA (mevalonic acid) and into the intermediates and ABA such that one of the most important approaches to studying metabolic pathways has been of little use in studying ABA biosynthesis (Zeevaart and Creelman 1988, Wahon and Li 1995). The poor incorporation of ''*C-MVA into ABA led to a number of assumptions about precursor pools and localization of ABA synthesis. Milborrow (1983) suggested that ABA may be synthesized in the chloroplast; thus, the fed precursors have to cross two layers of chloroplast membrane to reach ABA synthetic enzymes. In fact, currently it is widely believed that ABA is synthesized in plastids. Another explanation is the existence of a relatively large pool of precursors such that the incorporated labeled precursor was largely diluted. Alternatively, the pool used for ABA biosynthesis may be separated from the one reached by conventional feeding experiments. Some of those assumptions turned out to be correct as now we know that ABA is derived from an oxidative cleavage of carotenoids which exists abundantly and exclusively in plastids of higher plants. The Direct C ij Pathway of ABA Synthesis in Fungi The discovery in 1977 that a rose pathogen, the fungus Cercospora rosicola, produces and excretes relatively large quantities of the naturally occurring enantiomer of ABA into its growth medium initiated work on the biosynthetic pathway in that organism (Asante et aL 1977). The hypothesis was then very simple that higher plants might use the same pathway as in fungi. Radioactive isotope labeled MVA was fed to C. rosicola

PAGE 9

3 fXf^ ABA K^*^ COOH MVA Y •ionylidene pathway Fig. 1-1. The direct ABA biosynthetic pathway found in the fungi Cercospora rosicola (a-ionyHdene pathway), C. cruenta (y-ionylidene pathway), C. pini-demiflorae, and Botrytis cinerea. A single arrow between two compounds does not necessarily indicate that only one enzyme step is involved. (Adapted from Zeevaart and Creelman 1988)

PAGE 10

4 and shown to be converted to I'-deoxy-ABA, a major accumulated intermediate. Purified I'-deoxy-ABA fed to the fungus was converted to ABA with a good yield. Analysis of the labeled-intermediates in C. rosicola fed [l,2-''C]-acetate revealed that ABA is synthesized via an isoprenoid pathway as indicated in Fig. 1-1 (Beimett et al. 1981, 1984) The radioisotope labeled intermediates were followed a chasing pattern, MVA, famesyl pyrophosphate (FPP), a-ionylidene, a -ionylidene ethanol, a-ionylidene acetic acid, 4'hydroxy-ionylidene acetic acid, I'-deoxy-ABA in C. rosicola. There seemed to be variations in other fungal species regarding ABA synthesis, because similar assays revealed different radio-labeled intermediates. The major difference may be the order of the hydroxylation and oxidation of ionylidene. In B. cinerea, V, 4'-t-diol of ABA was shown to be the immediate precursor of ABA (Hirai 1986). In C. pini-densifloorae, ABA seems to be derived from 4'-hydroxy-ionylidene acetic acid via r4'-t-diol of ABA because the latter can be converted at a higher rate (Okamoto et aL 1987). In C. cruenta, the radio-labeled MVA was converted to ABA in the following order, y-ionylidene ethanol, 4'-hydroxy-r-ionylidene acetic acid, r,4'dihydroxy-y-ionylidene acetic acid and ABA (Oritani et al 1985, Oritani and Yamamoto 1985) Although different &ngal species may use one or more of the suggested pathways, all the intermediates have 15 carbons and are structurally analogous to ABA. Fluridone, a carotenoid synthesis inhibitor, did not inhibit ABA synthesis in C. rosicola indicating carotenoids are not involved in the ABA biosynthesis (Oritani and Yamamoto 1985). In a search for a similar pathway in higher plants, tissues from a variety of plant species were fed with '^C-MVA (Zeevaart and Creelman 1988). In contrast to fungi, only a very low incorporation into ABA was detected. Feeding of bean and avocado fruits

PAGE 11

with ionylidene or I'-deoxyionylidene did not produce ABA as was observed in fiingi (Zeevaart and Creelman 1988). The only report that showed ABA could be directly converted from Cu substrates was provided by Robertson (Milborrow 1983), in which **C-phytoene and ^H-MVA were fed to avocado fruits. Both ^'*C and 'H were found in Pcarotene, but only the label was found in ABA. This experiment was later regarded as inconclusive because the possible existence of a separate P-carotene pool that is inaccessible to ABA synthetic enzymes was not ruled out and the C^'* incorporation of into xanthophylls was not examined (Zeevaart et al. 1989). The Indirect C 4 0 Pathway of ABA Synthesis in Higher Plants For about thirty years since the elucidation of ABA structure, there has been a debate as to whether ABA is synthesized in higher plants via a direct C15 pathway (sesquiterpenoid) or via an indirect C40 (apo-carotenoid pathway). Because little evidence supported the direct pathway in higher plants and due to the emerging evidence in support of an indirect pathway, a universal pathway of ABA biosynthesis was proposed for higher plants (Zeevaart et al. 1989, Parry et al. 1992, Walton and Li 1992 1995). Three lines of evidence confirmed that ABA is synthesized predominantly from an oxidative cleavage of 9-cis xanthophylls (oxygenated carotenoids). 1). Carotenoid mutants in maize which block the synthesis of certain carotenoids also blocked the synthesis of ABA strongly indicating that ABA is derived the from carotenoid precursors. The albino, viviparous mutants of maize, ^J)2, vp5, vp7,vp9, y3, y9, w3 have reduced levels of ABA in embryos and cause precocious germination of developing seeds (Robertson 1955, NeilletaL 1986). The albino phenotype presumably results from

PAGE 12

6 PHYTOENE _^ vp2, vp5 PHYTOFLUENE w3 C-CAROTENE 4^ vp9,y9 NEUOPORENE i LYCOPENE i Y-CAROTENE PCAROTENE 4P-CRYPTOXANTHIN 1 ZEAXANTHIN 5CAROTENE aCAROTENE 4aCRYPTOXANTHIN i LUTEIN Fig. 1-2. The biosynthesis of zeaxanthin and the blocked steps of several viviparous mutants in maize. Blocked steps are marked with lines crossed the arrows for the mutants labeled on the right.

PAGE 13

ZEAXANTHIN i epoxidase ANTHERAXANTHIN epoxidase ALL-TRANS VIOLAXANTHIN isomerase ALL-TRANS NEOXANTHIN isomerase 9-C/5 VIOLAXANTHIN 9 '-CZS NEOXANTHIN dioxygenase dioxygenase XANTHOXIN I oxidase ABA ALDEHYDE I oxidase ABA Fig. 1-3. The proposed ABA biosynthetic pathway in higher plants [SUghtly modified from versions of Zeevaart and Creelman (1988) and Wahon and Li (1995)].

PAGE 14

t photobleaching of chlorophylls in the absence of photoprotection provided by carotenoids. Such mutants are typically seedling lethal. Further biochemical studies have located those mutations to the carotenoid biosynthesis pathway as indicated in Fig. 1-2. 2). Incorporation of '^0 from O2 into the aldehyde oxygen but not into the ring of ABA in water stressed leaves in a variety of species clearly supports an oxidative cleavage of a double bond of 9-cis xanthophylls (Creelman and Zeevaart 1984, Creelman et al. 1987, Li and Walton 1987, Parry et aL 1988, Gage et al. 1989; Zeevaart et al. 1989). In addition, a decrease of four xanthophylls (all trans violaxanthin, all trans neoxanthin, 9cis neoxanthin and 9-cis violaxanthin) equals stoichiometrically to the increase of ABA and its metabolites in water stressed etiolated bean leaves (Parry et al. 1989; Li and Wahon 1990a). 3). An ABA deficient mutant, aba J of Arabidopsis, was isolated by Koomneef et al. (1982) and was shown to be genetically impaired in the epoxidation of zeaxanthin to form violaxanthin (Duckham et al. 1991, Rock and Zeevaart 1991). This is an essential step for ABA biosynthesis via the proposed C40 carotenoid pathway (Zeevaart and Creelman 1988). Recently, the homologous gene of abal in tobacco (aba2) has been cloned using heterologous transposon tagging(/4c) and the epoxidase activity was confirmed for the protein expressed in E. coli (Marin et al. 1996). In the present study, an ABA deficient mutant of maize, viviparousJ4, is shown to be blocked in the oxidative cleavage of 9-c/j xanthophylls. The purified recombinant protein expressed as a GST-fusion in E. coli can cleave specifically the presumed substrates, 9-cis neoxanthin and 9-cis violaxanthin. These molecular data confirm the existence of the C40 pathway in higher plants. It has to be pointed out, however, that there is, so far, no conclusive evidence ruling out the existence of an alternative Cu pathway in higher plants.

PAGE 15

The proposed C40 ABA biosynthetic pathway in higher plants is shown in Fig. 13. Xanthophyll cycle intermediate zeaxanthin is epoxidated to form all trans violaxanthin which is evidently isomerized to produce a 9-cis violaxanthin and 9-cis neoxanthin. The isomerization step is strongly supported by the fact that all fram-xanthoxin can not be converted to biologically active ABA by cell free extrarts (Zeevaart and Creelman 1988, Walton and Li 1989). As to whether isomerization is an enzymatic or photodynamic reaction is still not clear. The existence of specifically 9-cis isomers, in fact, favors the enzymatic reaction view. Because 9-cis neoxanthin is about 10 times more abundant than 9-cis violaxanthin in leaves, the former is suggested to be the primary substrate for oxidative cleavage to produce xanthoxin, a compound that inhibits cell growth and has a skeleton similar to ABA. Xanthoxin is further isomerized and oxidized to produce ABA. The oxidative cleavage reaction is the first committed step in the proposed ABA biosynthetic pathway. Substantial evidence indicates that it is also the highly regulated step of this pathway. In contrast to the endogenous level of ABA in most tissues, carotenoids and xanthophylls are far more abundant than ABA. In particular, 9-cis xanthophylls are found at much higher concentrations than ABA (Parry et al. 1990, Li and Walton 1990a), indicating that substrate production is not the limiting factor unless there is a separate pool for ABA biosynthesis. The endogenous level of the cleavage product, xanthoxin, is extremely low in all tissues. When xanthoxin is incubated with enzyme extracts from most tissues of bean, xanthoxin was quickly converted to ABA, indicating the enzymes downstream of xanthoxin are constitutively expressed (Sindhu and Walton 1987). Application of the protein synthesis inhibitor cycloheximide to stressed leaves reduced ABA accumulation, but did not affect the enzymatic conversion

PAGE 16

10 of xanthoxin to ABA compared to the stressed leaves without cycloheximide, suggesting that those enzymes are constitutively active and not stress inducible under conditions that caused a greater than 40 fold increase in endogenous ABA level (Li and Walton 1990b, Sindhu and Walton 1990). These results indicate that the oxidative cleavage of 9-cis xanthophylls is likely to be the key regulated step in this pathway. The Questions In terms of ABA biosynthesis in higher plants, the possibility that a direct pathway may exist is not completely ruled out. The aba2 mutant of tobacco which seems to be a null mutation due to an Ac element insertion into the coding region of the Aba2 gene still retained 30% of its endogenous ABA (Marin et al. 1996). Molecular analysis indicated that Aba2 is a single copy gene based on Southern blot analysis. Thus, the endogenous ABA in that mutant may be from one of three sources, a separate pathway like the indirect pathway, leaky mutation, or existence of redundant pathways leading to the production of violaxanthin including photodynamic conversions, etc. Given the important functions of ABA throughout a plant life cycle, a total knockout of ABA biosynthesis might be lethal to a plant during early embryogenesis. In that case, this class of mutants may never be isolated. The direct pathway, if it exists in higher plants, may only play a very minor role. If so, it is not surprising that mutants in such a pathway have not yet been found. One key question with respect to the proposed indirect pathway is the demonstration of the oxidative cleavage of 9-cis xanthophylls to xanthoxin as it bridges the carotenoid and ABA biosynthetic pathways. Mutants impaired in other steps including zeaxanthin epoxidation and conversion of ABA aldehyde to ABA have been

PAGE 17

11 found and are fairly well characterized (Taylor 1992, Rock and Zeevaart 1991, Duckham et aL 1992). Recently, a new ABA deficient mutant of Arabidopsis, aba2, was shown to be blocked in the step converting xanthoxin to ABA-aldehyde (Schwartz et al. 1997). However, no mutants that are blocked in the presumably key regulated step, the oxidative cleavage of epoxy-carotenoids, have been isolated. Notabilis of tomato (Parry and Horgan 1992) and wilty of pea (Duckham et al. 1989) which are not blocked in the conversion of xanthoxin to ABA and have normal carotenoids and xanthophylls and are thus possible candidates for mutants in the oxidative cleavage step. However, the highly leaky nature possibly due to overlapping expression of redundant genes, as discussed above, has hindered the characterization of these mutants. Major Functions of ABA in Higher Plants Regulation of Seed Maturation. Dormancy and Germination by ABA ABA has been implicated in the control of many events during seed formation including maturation (Rock and Quatrano 1995), storage protein synthesis (Finkelstein et aL 1985), desiccation tolerance acquisition (Chandler and Robertson 1994, McCarty 1995), and the onset and maintenance of dormancy (Koomneef and Karssen 1994). In addition, ABA inhibits certain germination promoting processes such as expression of hydrolytic enzymes like a-amylase (Jacobsen et al. 1995), which re-mobilize the storage reserves in endosperm during germination. Physiological studies indicated that ABA concentration peaks in developing seeds around the time of maximum fresh weight in many species just prior to the acquisition of desiccation tolerance and to onset of dormancy (Koomneef and Karssen 1994, Chandler

PAGE 18

12 and Robertson 1994). Many proteins have been identified that are positively regulated by ABA in seeds, and these may play important physiological roles in developmental arrest or seed maturation. A group of seed storage proteins, the late embryogenesis abundant proteins (LEA), were found to be activated by ABA (reviewed by Rock and Quatrano 1995). Application of ABA to germinating seeds not only arrests germination, but also initiates synthesis of LEA proteins, indicating that ABA is a key signal for expression of Lea genes. Several Lea genes have been cloned and through studies of promoter structure, a conserved ABA responsive c/s-element (ACTG core element) has been identified (Marcotte et al. 1992, Rock and Quatrano 1995). Rich in hydroxy lated and hydrophilic amino acids, LEAs are believed to be osmoprotectants which help protect cells of seed tissues from desiccation (Rock and Quatrano 1995). Genetic evidence for ABA function in seed development and dormancy was revealed by the studies of ABA deficient and ABA insensitive seed mutants. Deficiency of ABA synthesis in developing maize kernels, occurs in the vp2, vp5, vp7 and vp9 mutants. Such mutants cause the embryo to bypass dormancy and germinate precociously before completion of seed maturation, i.e. vivipary (reviewed by McCarty 1995). These mutant kernels neither develop desiccation tolerance nor accumulate certain LEAs, suggesting that ABA is absolutely required for the acquisition of desiccation tolerance, and the onset and maintenance of seed dormancy. In addition, sensitivity of embryos to ABA seems also involved in the regulation of seed development and maturation. Embryos of sprouting-resistant cereal cultivars have been shown to be more sensitive to ABA than sprouting-sensitive ones (Walker-Simmons 1987). The ABA insensitive (abi) mutants of Arabidopsis abi3 (Koomneef et al. 1984), abi4 and abi5 (Finkelstein 1994),

PAGE 19

13 and the mutant of maize vpl (Robertson 1965, McCarty et al. 1991) exhibit non-dormant and desiccation intolerant phenotypes, clearly indicating the fiinction of ABA in controlling late seed development and dormancy. ABA is a major determinant of the onset and maintenance of seed dormancy, while the plant hormone gibberellin (GA) controls the release from dormancy and initiation of germination processes. ABA and GA play their roles sequentially but also overlap each other. Thus, transition from dormancy to germination is closely related to the environmental and developmental regulation of ABA and GA biosynthesis. As an antagonist of ABA, GA induces the expression of genes that promote utilization of stored seed reserves such as hydrolytic enzymes, whereas ABA inhibits their expression. One example is amylase expression (Jacobsen et al. 1995). The early hypothesis that a balance of GA and ABA controls dormancy and germination by a similar fashion as auxin and cytokinin controls of root and shoot differentiation, may not be entirely correct as dormancy is not affected in many GA deficient mutants (Rock and Quatrano 1994). The interaction between sources of ABA found in an embryo may be complex. Furthermore, dependence of embryo maturation and dormancy on ABA signaling appears to vary among plant species. The fact that embryos at early stages of development can be cultured in vitro and develop into plants without undergoing dormancy indicates that dormancy is affected by maternal factors. However, in maize, homozygous ABA deficient, viviparous kernels of the vp5 mutant develop in the presence of genetically heterozygous maternal tissues, indicating that ABA synthesized within the kernel itself determines the maturation processes and the onset of dormancy. Robertson (1952), using chromosome T-B translocation techniques, generated kernels that were homozygous vp5

PAGE 20

M in the embryo, but heterozygous in the endosperm. Such embryos still displayed a viviparous phenotype. This proved that endosperm is not an essential source of embryo ABA in maize. Thus, acquisition of desiccation tolerance and dormancy of embryo are determined by the ABA synthesized in embryo itself (McCarty 1995, Rock and Quatrano 1995). It may be possible that maternal tissues surrounding the embryos generate a signal other than ABA to activate ABA synthesis in developing embryos. In Arabidopsis, ABA deficient mutants are typically not viviparous unless the activity of another key regulator of seed maturation, Ahi3, is attenuated (Koomneef and Karssen 1994). A maternal ABA effect was also suggested by the fact that abal, abi3 double mutant F2 seeds that develop with heterozygous maternal tissues were not viviparous, whereas, double mutant seeds that develop on homozygous abal plants in the next generation are viviparous (Koomneef and Karssen 1994). The aba2 mutant of tobacco which is homologous to abal of Arabidopsis is also able to complete seed maturation and to develop desiccation tolerant seeds (Marin et al. 1996). Although ABA was not directly measured in abal seeds, the ABA deficiency was evident in the phenotype of reduced dormancy. Aba2 encodes the zeaxanthin epoxidase in ABA synthesis and is a single copy gene (Marin et al. 1996). Thus, the dependence of seed maturation and dormancy on ABA signaling appears to be markedly different among Arabidopsis, tobacco and maize. Re gulation of Stomatal Closure Resistance to carbon dioxide intake and transpirational water loss by leaves are controlled by stomatal pores typically located on both surfaces of plant leaves. The aperture of stomatal pores is controlled by changes in the turgor of the two surrounding

PAGE 21

u guard cells. Loss of turgor, as in case of water deficit, leads to closing of stomatal pores and hence help protect leaves from further water loss. Guard cell turgor is directly related to influx and efflux which is controlled at least in part by ABA. The action of ABA on stomatal closure is manifested in a number of ABA deficient mutants. Deficiency in ABA typically causes a wilty phenotype because of a failure of mutant plants to efficiently close their stomata. These mutants include abal, aba2, aba3 of Arabidopsis, aba2 of tobacco; sitiens and flacca of tomato; wilty of pea; and droopy of potato (Taylor 1991). Two ABA insensitive mutants {abil and abi2) of Arabidopsis also exhibit a wilty phenotype suggesting that an ABA signal transduction cascade is possibly involved in the stomatal closing (Giraudat 1995). Physiological studies indicate that stomatal opening is strongly inhibited by exogenous application of ABA onto leaves or peeled epidermis of many species (Mansfield and McAinsh 1995). In many plant species, bulk ABA levels increase dramatically in leaves exposed to water stress (reviewed by Ingram and Bartels 1996). However, when the time course is taken into account, the closing of stomata generally occurs before any significant changes of ABA content can be detected in the leaves. One possible explanation for this time discrepancy may involve a fast localized accumulation of ABA around or in guard cells, through either re-localization of ABA among leaf cells or highly localized de novo synthesis of ABA, e.g. only in guard cells or the neighboring cells. In support of this suggestion, a rapid increase in the ABA concentration of a single guard cell was shown to be correlated with the closure of stomata in stressed leaves (Harris and Outlaw 1991). However, this study did not identify the source of the ABA. These results also suggest that the dramatic accumulation of ABA in stressed leaves may have functions other than

PAGE 22

to close the stomata. It may mediate other processes involved in plant acclimation to the environment. The sources of ABA in leaves appear to be complex. It is now well established that ABA is synthesized in roots (Cornish and Zeevaart 1988). When growing roots come in contact with drying soil, they produce ABA in increased quantities, which enters the xylem and is transported to the leaves where it may inhibit stomatal opening (Zhang and Davies 1989). This occurs before the shortage of soil moisture causes any measurable change in the water status of the leaves. Thus it is suggested that the early stages of soil drying in field grown plants lead to the production of ABA which is transported as a chemical signal to the leaves, where it causes a reduction in transpiration and prevents a decline in water potential or a loss of turgor. When leaves are detached, the increased ABA level apparently results fi"om the de novo synthesis because transcription and translation inhibitors can prevent this increase (Li and Walton 1990b, Ingram and Bartels 1996). Thus, accumulated ABA in stressed leaves apparently has at least two sources, roots and local synthesis. It is possible that in field grown plants, the sources of ABA that cause stomatal closure are determined by the severity of water deficit; mild water deficit will cause increased synthesis of ABA in roots which translocates to leaf guard cells; severe and progressive water deficit will induce ABA synthesis in the leaf itself The study of field grown maize by Tardieu and Davies (1992) has shown that there is a good correlation between stomatal conductance and ABA concentration in the xylem sap and this may be the best above-ground indicator of the water status of the root system. ABA modulates the activities of three major classes of ion channels, inward-, outward-rectifying K+ channels and anion channels located on the guard-cell plasma

PAGE 23

It membrane (Armstrong et al. 1995). How the ABA signal cascade functions is still unknown. When guard cells are exposed to ABA, the first detected event is an influx of positive charges leading to an initial depolarization of the membrane (Thiel et aL 1992). The influxes of Ca^* and are proposed to act as secondary messengers to activate Ca2+ sensitive and voltage sensitive anion channels. This causes long term depolarization and a massive anion efflux across the guard cell membrane (Schroeder and Hagiwara 1990). It is suggested that slow anion channels could be the rate limiting step in controlling stomata closing. Using patch-clamp techniques, it was found that the slow anion channel is strongly activated by ABA and this activation can be suppressed by the protein phosphatase inhibitor, okadaic acid, this suggests a kinase/phosphatase signal cascade (Pei et al. 1997). In the ABA-insensitive (abi) mutants of Arabidopsis, abil and abi2, failure to close stomata is associated with loss of the slow anion channel activation caused by ABA. ABIl encodes a protein phosphatase 2C, a component in ABA signaling pathway (Meyer et al. 1994, Leung et aL 1994). The membrane depolarization is proposed to generate the driving force for K+ efflux through the outward-rectifying K+ channels. This in turn causes the loss of turgor in guard cells, and closing of stomata (Armstrong et aL 1995, Pei et al. 1997). Conversely, ABA also inhibits the inwardrectifying K+ channel activity (Blatt 1 992, Schroeder and Keller 1 992). Guard cells of tobacco transformed with the abil dominant allele have reduced ABA activation of the outward-rectifying K+ channel and also show less sensitivity of the inward-rectifying K+ channels to ABA (Armstrong et aL 1995). The presence of broad-range protein kinase antagonists, H7 and staurosporine restores the sensitivity, suggesting a kinase may be a negative regulator. A similar study of the ion channels also

PAGE 24

It indicated loss of slow anion channel activation by ABA in abil and abi2 mutants (Pei et aL 1997). ABA-induced anion channel activation and stomatal closing were suppressed by protein phosphatase inhibitors in both mutants. However, kinase inhibitors did not rescue the ABA activation of slow anion channels in abi2 mutants, indicating that abi2 may be an upstream component in ABA signal cascade. Since most of the ABA signaling cascade as well as ABA receptors in plasma membrane of guard cells have not been identified, the positions of ABIl and ABI2 in that cascade remain to be resolved. There are reports that Ca2+ dependent protein kinases (CDPKl & CDPKla) may also be involved in the ABA signal transduction pathway (Sheen et al. 1996). Another question that remains elusive is the location of ABA receptors which hold the key to the ABA signal transduction cascade. It is believed that they are located on the outer surface of guard cell plasma membrane sensing ABA from outside the guard cells when water deficit occurs (Hartung and Davies 1991). However, micro-injection of a physiological concentration of ABA into the cytoplasm can trigger the K+ efflux and close stomata, and the extent of stomata closing is correlated with the extent of ABA uptake by the guard cells, thus suggested the existence of internal ABA receptors (Schwartz et aL 1994). An increase of ABA content inside the guard cells is correlated with the closing of stomata (Harris and Outlaw 1991), suggesting the existence of intercellular ABA receptors. Adaptation to Stress Environments The survival of a plant in nature frequently requires a capacity to withstand extremes of a stressful environment including drought, extreme temperature, salinity, wounding and pathogen infection. Plants have developed two overall mechanisms to

PAGE 25

19 survive, stress avoidance and stress tolerance (Chandler and Robertson 1994). Avoidance is achieved by developing specialized adaptations in architecture, such as a highly developed root system to avoid drought stress. Whereas stress tolerance depends on a combination of rapid and longer-term physiological responses to stresses, such as closing stomata to prevent from further water loss and accumulation of a class of proteins to protect cells from damage (Skriver and Mundy 1990, Chandler and Robertson 1994, Ingram and Bartels 1996). The involvement of ABA in stress acclimation is supported by several lines of evidence: (1) ABA accumulates during stress including drought, flooding, low temperature, wounding and pathogen infection (Ingram and Bartels 1996). Wright and Hiron (1969) first detected a 40 fold increase of ABA in detached, wheat leaves. Subsequent studies have reported similar phenomena in many plant species (Ingram and Bartels 1996). The mechanism of stress-dependent ABA biosynthesis is still not fully understood (Hartung and Davies 1991). (2) Plants develop freezing tolerance when treated with ABA under nonacclimating conditions (Lang et al. 1989). The ABA deficient mutant cAal of Arabidopsis does not develop cold acclimation and this can be complemented by exogenous application of ABA (Heino et aL 1990). (3) Under drought stress conditions, ABA is shown to be required for primary root elongation, and elongation of primary roots is inhibited in ABA deficient vp5 of maize (Sharp et al. 1994). (4) Proteins that are induced by environmental stresses and associated stimuli such as RABs (responsive to ABA) can be induced by applying ABA (Skriver and Mundy 1990, MantylaetaL 1995).

PAGE 26

20 One class of ABA proteins that are strongly induced by ABA in the late embryogenesis is the LEAs. Although Unctions of those proteins are still unknown, synthesis of LEAs is correlated with the acquisition of desiccation tolerance in embryos. The direct relationship between LEAs and ABA was established by ABA treatment of cultured embryos, which induced LEA expression (Ingram and Bartels 1996). Recent molecular studies have also characterized the rapid induction of RAB genes in leaves or roots. These novel genes have been isolated from several species by differential screening of cDNA libraries constructed from ABA treated leaves or roots. Some of these proteins are also expressed during the maturation phase of embryo development, thus may be fiinctionally related to LEAs (Rock and Quatrano 1995). Both LEA and RAB proteins share one predominant feature; high hydrophilicity and a high content of uncharged and hydroxylated amino acids. Conserved domains are postulated to be functionally important in desiccation protection (Dure et al. 1989), possibly through interactions with other embryo proteins and membranes. However, the RAB 17 of maize was found localized in the nucleus and is suspected to play a role in the nuclear protein transport (Goday et aL 1994). The regulation of ABA responsive gene expression is mediated by the promoters of these genes. ABA response elements (ABRE) which contain the palindromic motif CACGTG with a G-box core element (Giuliano et al. 1988) are found in RABs and LEAs, and have been extensively studied in Em gene of wheat and the rice RAB 16 gene (Skriver et al. 1991, Guiltinan et al. 1990). Recently, the modular nature of the abscisic acid response complex (ABRC) was further refined in a barley EM gene, HVAl. The promoter unit necessary and sufficient for abscisic acid (ABA) induction of gene

PAGE 27

21 expression consists of a G-box and one of several "coupling elements" (CEs) (Shen et al. 1995, 1996). Different combinations of the G-box with different CEs may modulate different gene responses in different tissues. ABA is also involved in processes that lead to the establishment of freezing tolerance. However, little is known about how ABA triggers those processes. Thus far, few mutants have been isolated that have an ahered freezing tolerance phenotype. A putative freezing sensitive mutant (frsl) of Arabidopsis has a reduced cold acclimation response and this response can be compensated partially by application of ABA (refer to Quatrano et al. 1997). The molecular nature underlying this mutation should be of importance to the understanding of freezing tolerance mechanisms. (Quatrano et al. 1997) Inhibition of Growth Physiological studies have established that exogenous application of ABA leads to inhibition of vegetative growth. At the molecular level, ABA regulates the expression of a large number of genes especially under stressful environments. Many of these ABA up-regulated genes have been cloned by differential screening; however, their functions remains unknown. The ABA down-regulated genes have drawn little attention, althrough, most appear to promote cell growth. An example is the inhibition of the a-amylase promoter by synergistic interaction of ABA and VP 1 (Hoecker et aL 1995). Defense Responses The proteinase inhibitor II (Pinl) is a model system for wound induced gene activation (Pena-Cortes and Willmitzer 1995). The Pin2 gene family is the best studied examples of genes which are systematically activated upon mechanical damage of plants.

PAGE 28

22 The involvement of ABA in gene activation processes following mechanical damage of the plant tissues is supported by the fact that the endogenous ABA concentration rises threeto five-fold upon wounding (Pena-Cortes and Willmitzer 1995). Exogenous application of ABA on non-wounded plants can activate the expression of Pin2 in a pattern analogous to wounding ( Pena-Cortes et aL 1991). In the ABA deficient droopy mutant of potato (Quarrie 1982) and sitiens mutant of tomato (Taylor et al. 1988, Duckham et aL 1989), wounding failed to stimulate the expression of P/2, and did not increase ABA in leaves (Pena-Cortes and Willmitzer 1995). Spraying ABA on wounded droopy and sitiens leaves resulted in an activation of Pin2, thus strongly indicating that ABA is at least one of the signals to trigger Pin2 gene activation upon wounding. Drought stress induced ABA accumulation does not activate Pin2, whereas wounding of the stressed leaves of potato and tomato activates Pm2 transcription (Pena-Cortes et al. 1989). This implys that muhiple factors including ABA may account for the wounding activation of gene expression. Several other wound induced genes have been cloned which show a pattern of wound induction similar to Pm2 (Hildmann et aL 1992). ABA is involved in the activation of those genes. Molecular analysis of promoters fi-om several genes and Pm2 revealed an ABA responsive element (ABRE) that is similar to that found in ABA regulated genes including Leas, Em and RabI6. However, mutagenesis of this element did not affect ABA regulated expression (Lorberth et aL 1992). This suggests that different factors in addition to the presence of ABA are necessary for the induction of this set of genes in contrast to Em. Consistent with the multiple functions of ABA throughout plant development, the signaling pathways for ABA that regulate transcription of

PAGE 29

23 different genes are expected to be complex. In addition to ABA, jasmonic acid is also involved in the activation of related genes in response to wounding (Pena-Cortes and Willmitzer 1995). Abscission and ABA Although abscisic acid was initially isolated as a plant hormone that enhanced organ abscission, this nomenclature is misleading. Application of ABA to plant leaves or flowers does not result in organ abscission (Addicott 1983). Ethylene is a plant hormone that more directly controls organ abscission (Abeles et al. 1992). implication of ethylene or its precursor 1-aminocylopropane-l-carboxylic acid (ACC) strongly promotes organ abscission. The link between ABA and organ abscission may be linked to an interaction of the two hormones. It is possible that ABA is involved in the senescence processes in the separation layer cells in organs undergoing abscission (Abeles et al. 1992). ABA increases ethylene production and abscission in aged citrus organs (Riov et aL 1990). Water stress may cause leaf abscission in cotton and citrus, and inhibition of ABA biosynthesis was reported to reduce leaf abscission (Gomez-Cadenas et aL 1996). The ABA Related Mutants in Higher Plants ABA Insensitive Mutants ; *. The abil, abil, and abi3 were isolated by Koomneef et al. (1984), and the abi4 and abiS were isolated by Finkelstein (1994). These mutants were selected by germination of EMS mutagenized Arabidopsis seeds on media containing ABA at concentrations that completely inhibit germination of wild type seeds. The aZ>/5, abi4 and abiS mutations affect seed development specifically. The severe abi3 alleles isolated

PAGE 30

24 later, abi3-3 (Nambara et aL 1994), abi3-4 abi3-5 (Ooms et aL 1993), exhibit a phenotype of nondormancy and desiccation intolerance. The viviparousl of maize exhibits a similar seed specific phenotype, and in addition controls anthocyanin synthesis in the seed (Robertson 1955, McCarty et aL 1991). Cloning of Vpl indicated that it encodes a novel transcription factor (McCarty et aL 1991). It can activate CI, and EM promoters in a synergistic manner in combination with ABA (Kao et al. 1996, Hoecker et aL 1995, Hottori et aL 1992, Vasil et al. 1995), but acts as a repressor of a-amylase (Hoecker et al. 1995). A B3 domain of VPl has a cooperative of DNA binding activity to the sph element of CI promoter (Suzuki et aL 1997). The sequence of ABI3 revealed strong homology to VPl (Giraudat et aL 1992), and thus ABI3 is considered an ortholog of VP1( McCarty 1995). Unlike abi3, plants of abil and abi2 display a wilty leaf phenotype in normal growth conditions (Koomneef et aL 1984, Giraudat 1994). ABIl encodes a 2C type of protein phosphatase (Meyer et al. 1994, Leung et al. 1994). ABIl is suggested to be a signal transduction component involved in many processes including embryogenesis, ABA responses in guard cells (Amstrong et aL 1995, Pei et al. 1997) and mitotic cell division in roots (Leung et al. 1994). A puzzle remains as to how a missense mutation (Gly-180 to Asp) in the mutant abil allele gives rise to a dominant phenotype. A search for abil homologous sequences, surprisingly, led to the cloning of Abil, and the mutation in abil was exactly as in abil, a Gly-180 to Asp (Leung et aL 1997). Thus, ABIl and ABI2 encode very similar protein phosphatases. Carotenoid Biosvnthetic Mutants In maize, several viviparous mutants are associated with an albino phenotype as indicated by vpl, vp5, vp7, vp9, w3,y9, etc. One function of the carotenoids is to protect

PAGE 31

25 the photosynthetic apparatus from photobleaching in a high light situation. Blocking the synthesis of certain carotenoids causes these mutants to be albino and generally lethal. The steps blocked in the various mutants are shown in Fig. 1-1. As ABA is synthesized from cleavage of carotenoids, certain carotenoid mutants are also ABA biosynthetic mutants. Those mutants provided key evidence that ABA is derived from carotenoids but have generated little information about the committed (specific) steps of the ABA biosynthetic pathway (Zeevaart and Creelman 1988). ABA Biosynthetic Mutants A number of ABA deficient mutants have been isolated from a variety of plant species as indicated in Table 1. Most of those mutants (/7c, sit, droopy, ckrl, ibal, etc) affect the last step of ABA biosynthesis, the oxidation of ABA-aldehyde to produce ABA. The tomato mutant, Notabilis, is less well characterized largely due to its leaky nature, but is suspected to be impaired in the oxidative cleavage of 9-cis neoxanthin to xanthoxin (Parry et al. 1988, Parry and Hogan 1992a, b, Taylor et al. 1988). Similarly, a pea wilty (wt) ABA deficient mutant is suspected to be blocked in the same step, since it can convert ABA aldehyde to ABA (Duckham 1989). No differences in xanthophyll and carotenoid profiles were found in wt mutant and wild type plants (Duckham et aL 1991). More detailed reviews including these mutants can be found elsewhere (McCarty 1995, Rock and Quatrano 1995, Taylor 1991). The following summarizes the current advances in this area. The Arabidopsis abal was isolated in a screen for revertants of the GA deficient gal mutant population (Koomneef et al. 1982). Abal is impaired in the conversion of zeaxanthin to violaxanthin (Duckham et al. 1991, Rock & Zeevaart 1991). The Aba2

PAGE 32

26 mutant of tobacco was isolated by heterologous Ac tagging in tobacco and shown to encode zeaxanthin epoxidase by expression in E. coli. (Marin et aL 1996). The tobacco aba2 gene is able to complement the abal phenotype of Arabidopsis through TDNA transformation, indicating that it is an otholog of abal (Marin et al. 1996). Additional Table 1-1. Abscisic acid deficient mutants Species Mutant Phenotype Blocked step References* Arabidopsis thaliana abal nd, wilty, red. height Zeaxan -> \^oxan Koonmeef 1982; Rock 1991 aba2 red. dormancy, wilty Xan > ABA-ald Leoo 1996; Schwartz 1997 aba3 red. dormancy, wilty ABA-ald -> ABA Leon 1996; Schwartz 1996 Nicotiana piumbagtmfolia aba2 nd, wilty Zeaxan -> \^oxan Marin 1996 (tobacco) dcrl cytokinin R, wilty ABA-ald -> ABA Pany 1991 ibal auxin R, wilty ABA-ald -> ABA Bitoiuil990 Lycopersicon esculentum flacca wilty ABA-ald -> ABA Parry 1988 (tomato) sitten wilty ABA-ald -> ABA Parry 1988 notabilis slightly wilty ? Taylor 1988; Pairy 1991 Solanum phureja (Potato) droopy wilty ABA-ald -> ABA Duckham 1989 Pisum sativum (pea) wilty slightly wilty ? Parry 1991 Hordeum vulgare (bariey) nar2a wilty ABA-ald -> ABA Walker-Simmoiu 1989 Zea may (maize) vp*2274 viviparous this study McCarty (1995) (vpl4) Note: nd, nondormant; red, reduced; R, resistance. Only the first author was listed to save space. Arabidopsis ABA deficient mutants were isolated in a screen for seed able to germinate in the presence of the gibberellin biosynthesis inhibitor paclobutrazol. Two new ABA

PAGE 33

27 deficient mutants were reported that have reduced dormancy and excessive rate of leaf water loss (Leon-Kloosterziel et al. 1996). Biochemical characterization indicated that aba2 is blocked in the conversion of xanthoxin to AB A-aldehyde and aba3 blocked in the last step of ABA biosynthesis, i.e. the conversion of ABA aldehyde to ABA (Schwartz et al. 1997). Candidates for Mutations in the Cleavage Step Two ABA deficient mutants, wilty and notabilis of pea and tomato, are blocked at undetermined steps in ABA biosynthetic pathway (Taylor 1991). Biochemical analysis of the two mutants did not detect any defects in either the xanthophyll profiles or the conversion of xanthoxin to ABA (Parry and Horgan 1992a, b, Taylor 1991). However, the highly leaky nature of both mutations greatly compromised any conclusions based on these data. Several newly isolated viviparous mutants of maize may potentially provide new information about the pathway (McCarty 1995). Viviparous mutants of maize, characterized so far, are all associated with ABA, either through perception or synthesis. Thus, mutants with normal carotenoids but decreased ABA levels are potentially ABA biosynthetic mutants. One mutant, named vp 14-227 4 (McCarty 1995), was analyzed in this study and is shown to be impaired in the oxidative cleavage step, the conversion of 9cis xanthophylls to xanthoxin. Further study of these mutants is expected to further improve our understanding of ABA function and biosynthesis in plants. Summary and Future Perspectives ABA, a sesquiterpene plant hormone, has multiple functions throughout the plant life cycle (Addicott 1983, Zeevaart and Creelman 1988, Taylor et aL 1992, Walton and Li

PAGE 34

28 1995). ABA plays important roles in the inhibition of growth, control of stomatal aperture, organ abscission, embryo development, dormancy, tolerance of stresses and plant defense responses (Zeevaart and Creelman 1988, Hartung and Davies 1991, Black 1991, Prescott and John 1996). In addition, ABA is particularly interesting with regard to its molecular regulation because hormone levels increase 10to 50-fold within hours of environmental pertuttations (Walton and Li 1995). Considerable effort has been directed toward dissection of the ABA biosynthetic pathway since elucidation of the chemical structure of ABA in 1965 (Addicott). We now know that in certain fiingi (Cercospora rosicola, etc) ABA is synthesized from a Cis isopropenoid via a so called direct pathway (Bennett et al. 1984, Zeevaart and Creelman 1988). In contrast, higher plants synthesize ABA largely from a C40 carotenoid (zeaxanthin) via a so called indirect pathway using an oxidative cleavage of 9 '-cis xanthophylls to produce xanthoxin. Xanthoxin is converted through two additional steps to ABA (Zeevaart and Creelman 1988, Walton and Li 1995). The cleavage step is the first committed step leading to ABA biosynthesis and substantial evidence has indicated that it is likely to be the key regulated step in ABA biosynthesis (Zeevaart and Creelman 1988, Walton and Li 1995). The epoxidase which converts zeaxanthin to all-trans violaxanthin has recently been cloned in tobacco and the enzyme activity was confirmed (Marin et al. 1996). However, none of the enzymes that are specific to ABA synthesis have been isolated, particularly the hypothesized dioxygenase that cleaves carotenoids. ? Mutants have proven to be very powerful in the elucidation of biosynthetic pathways, particularly in the case of ABA biosynthesis, where conventional isotope labeling has not been feasible. A single recessive mutation such as vp5 has provided

PAGE 35

29 important evidence that ABA is synthesized from carotenoids. The abal mutant of Arabidopsis and abal of tobacco have led to the conclusion that zeaxanthin is a precursor of ABA synthesis and that an epoxidation of zeaxanthin to form violaxanthin is required for ABA synthesis (Rock and Zeevaart 1991, Buckham et al. 1992, Marin et al. 1996). Other mutants like sittens, flacca, etc. have confirmed the last step of ABA synthesis (Taylor 1992). Recently, an ABA deficient mutant of Arabidopsis, aba2, was shown to be blocked in the step of converting xanthoxin to ABA-aldehyde (Schwarts et al. 1997). However, mutants that block other intervening steps in the ABA synthetic pathway are still lacking, particularly mutants blocking the proposed oxidative cleavage of epoxycarotenoids. Mutants that potentially regulate ABA biosynthesis are also required. Thus, successfril isolation of new, developmentally specific ABA deficient mutants provide an important way to the study of ABA biosynthesis and regulation (McCarty 1995). Eventually, the combination of molecular cloning, genetics, biochemistry and physiological studies will lead to a better understanding mechanisms for ABA action in plants. This will include developmental and environmental regulation of ABA synthesis, as well as how ABA regulates the expression of other genes to exert its function as well. Agricultural applications resulting from manipulation of ABA synthesis in crops will also be possible. Mutants are also required for explanation of the ABA signal transduction cascade since it is the key to understanding how ABA ftinctions in plants. ABA signal transduction pathways are likely complex, as indicated by the functions of ABA throughout plant life cycle. Potential components of the ABA transduction pathway such as Abil and Abi2 have begun to emerge (Meyer et al. 1994, Leung et al. 1994 1997). A

PAGE 36

30 recent mutant mArahidospsis which confers an enhanced response to ABA {era) encodes the beta subunit of famesyl transferase, a putatively negative regulator of ABA signal transduction pathway (Cutlers et aL 1996). However, the key details of how the ABA signal is perceived and transduced in plant cells are far from clear.

PAGE 37

CHAPTER 2 MOLECULAR ANALYSIS OF VIVIPAROUS 14, A DEVELOPMENTALL Y SPECIFIC ABSCISIC ACID BIOSYNTHETIC MUTANT OF MAIZE Introduction The plant hormone, abscisic acid (ABA) plays a key role in the regulation of seed maturation and dormancy (McCarty 1995, Rock and Quatrano 1995) and mediates plant responses to a variety of stress conditions including wounding (Pena-Cortes and Willmitzer 1995), drought (Vartanian et al. 1994, Giraudat 1995) and cold acclimation (Gihnour et al. 1991). ABA is synthesized from carotenoids as manifested in the ABA deficient mutants of maize (e.g. vp2, vp5, vp7, and vp9) that are in fact blocked in the early steps of carotenoid synthesis (Neill et al. 1987). Biochemical studies (reviewed by Zeevaart and Creelman 1988) and analyses of mutants that block ABA synthesis (reviewed by Taylor 1991) have led to a proposed indirect ABA synthetic pathway in higher plants (refer to Chapter 1, Zeevaart and Creehnan 1988, Walton and Li 1995). The first committed step in the pathway is the cleavage of two potential substrates, 9-cis violaxanthin and 9-cis neoxanthin to form xanthoxin which is subsequently converted to ABA v/a an ABA-aldehyde intermediate. Mutants blocked in ABA synthesis have been isolated in a variety of plant species including maize, Arabidopsis, tomato, tobacco, potato, etc. (Robertson 1955, Koomneef at al. 1984, Leon-kloosterziel et al. 1996, Marin et al. 1996, Neill et al. 1987). ABA deficient mutants of maize are viviparous, i.e. developing kernels fail to enter 31

PAGE 38

32 developmental arrest in the mid to late stage of seed maturation and germinate while still attached to the plant (Neill et al. 1986 1987, Robichaud et al. 1987). The viviparous mutants of maize, w3, y3, vp2, vp5, vp7 and vp9 are blocked in early steps of carotenoid biosynthesis, and thus also display albino (carotenoid deficient) and seedling lethal phenotypes due to photobleaching of chlorophyll (Neill et al. 1986, refer to Fig. 1-1 for blocked steps). These mutants are not considered specific to ABA synthesis and have contributed little to the studies of the ABA specific portion of the biosynthetic pathway (Zeevaart and Creelman 1988). ABA deficient mutants in other species (tomato, Arahidopsis, tobacco, etc) are generally not viviparous indicating that dependence of seed maturation on ABA signaling varies markedly among these species (McCarty 1995). The abal mutant Arahidopsis exhibits enhanced sensitivity to water stress and a reduced requirement for gibberellin hormone synthesis during seed germination (Koomneef et al. 1986, 1989). abal is deficient in the epoxy carotenoids, such as neoxanthin and violaxanthin, due to a genetic lesion in the epoxidation of zeaxanthin (Rock and Zeevaart 1991, Duckham et al. 1991). The aba2 mutant of tobacco, recently isolated and cloned by transposon (T-DNA/Ac) tagging, is homologous to abal oi Arabidopsis and encodes a zeaxanthin epoxidase (Marin et al. 1996). The flacca and sitiem mutants of tomato (Sindhu and Walton 1988), and the aba2 and abaS mutants oi Arabidopsis (Schwartz et al. 1997) genetically define two steps downstream of the cleavage reaction that allow conversion of xanthoxin to ABA. Thus far, mutants blocked in several steps of the pathway have been isolated. None of these mutants are known to be blocked in a key step that connects C40 epoxy-carotenoids and C15 xanthoxin, i.e. the cleavage of epoxycarotenoids. The notabilis mutant of tomato is the only reported candidate for the

PAGE 39

33 carotenoid oxidative cleavage (Parry et al. 1992 a, b). However, the blocked step has been biochemically difficult to locate in notabilis largely as a result of its highly leaky nature. Circumstantial evidence indicates that the oxidative cleavage of epoxy-carotenoids is the key regulated step in the ABA biosynthetic pathway. The carotenoid precursors of ABA are widely distributed in plant tissues. Epoxidation of zeaxanthin was ruled out to play a regulating role in ABA synthesis (Marin et al. 1996). The potential immediate precursors in the proposed reaction (Creelman et al. 1992, Walton and Li 1995), 9-cis violaxanthin and 9-cis neoxanthin, are far more abundant than the endogenous ABA levels in leaves of maize, pea, tomato and other species (Parry et al. 1990, Li and Walton 1987). This indicated that steps prior to formation of the 9-cis xanthophylls are unlikely to limit ABA synthesis. The enzyme activities downstream of the oxidative cleavage step required for conversion of xanthoxin to ABA also appear to be constitutively active in most plant tissues (Sindhu and Walton 1987). The enzymatic activities required for conversion of xanthoxin to ABA, as assayed in vitro, are not regulated by drought stress, and are also not inhibited by protein synthesis and transcription inhibitors in stressed leaves (Li and Walton 1990). Accumulation of ABA by stresses can be inhibited by inhibitors of both transcription and translation in a variety of species (Ingram and Bartels 1996). These studies have led to the hypothesis that the oxidative cleavage of 9-cis xanthophylls to form xanthoxin is the key regulated step in the ABA synthetic pathway (Sindhu and Walton 1987, Creelman et al. 1992). As suggested by McCarty (1995), a potentially important approach for identifying mutants in regulated steps of ABA synthesis is to look for mutants that affect ABA synthesis in specific organs or phases of development. In maize, evidence that all the

PAGE 40

34 known ABA deficient mutants are viviparous indicated that vivipary is possibly a sensitive seed specific phenotype for detection of ABA deficient mutants. To generate clonable ABA deficient mutants in maize, we screened active Robertson 's Mutator transposon populations for viviparous seed mutants. To avoid mutants that were blocked in carotenoid biosynthesis, we specifically looked for mutants that were viviparous but fiiUy viable as homozygotes and had normal carotenoid pigmentation (McCarty 1995). Here we report evidence that one of the ABA deficient mutants, viviparousN {ypl4), is blocked in the cleavage step of epoxy-carotenoids. Molecular analysis of the Vpl4 gene indicates that it encodes a protein related to lignostilbene dioxygenase, a bacterial enzyme which catalyzes a double bond cleavage reaction with striking similarity to the carotenoid cleavage step of ABA biosynthesis. We show that Vpl4 and related genes are developmentally regulated and induced in leaves by drought stress. Materials and Methods The vpl4-221A and vp]4-3250 mutants were identified in active Robertson's Mutator lines. Wild type W22, MM, Q66, Q67, Q77, Q79 strains that founded the Mutator population were a gift of Donald S. Robertson, Iowa State University. The nonsegregating wild type NS-2274 and homozygous vp 14-227 4 mutant strains used for biochemical and molecular analysis were extracted fi-om a segregating vpl4-221A stock that had been maintained by self pollination and selection of heterozygous plants. ABA Determinat ion and HPLC Analvsis of Carotenoids and XanthophvUs Embryos were harvested in Florida at 16, 18 and 20 days post-pollination and fi-ozen in liquid nitrogen. Extraction, purification and analysis of carotenoids and xanthophylls by

PAGE 41

35 HPLC was carried out in collaboration with Dr. Jan Zeevaart's group at Michigan State University using methods described earlier (Rock and Zeevaart 1991). ABA extractions, purification, and quantification by GC-MS with electron capture detector (ECD) were performed as described (Leon-Kloosterziel et al. 1996b). Determination of ABA Sensitivity in Culture Ears were harvested fi-om green house grown plants at 16 and 18 days after pollination (DAP) and surface sterilized by submersion in a 70% ethanol solution containing 1% dish detergent for 5 minutes. Embryos were harvested aseptically and placed in magenta boxes containing MS medium (pH 5.7), 0.2% phytagel and the indicated concentration of abscisic acid (Robichaud and Sussex 1986). Embryos were cultured at 25 C in a growth chamber and shoot, root length and fi-esh weight were measured after four days. The data shown are the means of 15 to 20 16 DAP embryos for each treatment. 18 DAP embryos showed a similar response. ; Southern and Northern Blot Analysis Genomic DNA was isolated from maize seedling leaves as described by Dellaporta (1983). Approximately 10 ^g of DNA was digested with the indicated restriction enzymes and resolved by agarose gel electrophoresis. The gel was denatured then blotted onto nylon membrane (Sambrook et al 1989). The membrane was UV-linked and hybridized in a Hybaid Chamber (Bio-Rad) according to the method of Church and Gilbert (1984). The membranes were washed in a 40 mM sodium phosphate buffer (pH 7.2), 1% SDS solution at 65 C for 2-3 hours for moderate stringency and at 70 C for high stringency. Total RNA was extracted from maize tissues (1-2 g) in TriZol solution according to manufacturers instructions (BRL) and purified fiirther by precipitation vwth

PAGE 42

36 isopropanol. Poly(A)*-enriched RNA was prepared using PolyATtract according to manufacturers instructions (Promega) and quantified spectrophotometrically. 1 ^g of Poly(A)^-RNA was resolved in a 1.2% agarose gels containing formaldehyde (Sambrook et al. 1989). After transfer, nylon membranes were UV-linked and probed as described for southern hybridization. The probes were radiolabeled with the Random Primer DNA Labeling System (BRL) in the presence of "P-a-dCTP (3000 Ci/mmol, DuPont). Construction and Screening of Genomic and cDNA Libraries Approximately 100 Jig genomic DNA prepared fi"om homozygous mutant or wild type plants (yp 14-227 4, vpl4-3250, wild type) was digested with appropriate restriction enzymes, then size fi-actionated by centriftigation at 85,000 xg at 4 C for 24 hours through a 10-40% linear sucrose gradient prepared in sterile TE buffer (pH 8.0). Fractions containing the fi-agment of interest were confirmed by Southern blotting and the concentrated DNA was ligated into an appropriate lambda phage cloning vectors (X,-gtlO was used to clone the 2.5 kb xhol fi-agment, X-ZAP was used to clone the EcoRI fi-agments). The ligated DNA was packaged into lambda phage according to the manufacture's instructions (Stratagene). A wild type embryo cDNA library was constructed in X-Zap fi-om 5 pig of poly(A)'^-mRNA prepared fi-om 18 day post pollination inbred W22 embryos. cDNA was prepared and cloned using the X-ZAP Express cDNA Synthesis Kit according to the manufacture's instruction (Stratagene). Lambda phage libraries were plated, lifted on nylon membrane and probed by DNA hybridization as described above for Southern and northern blots. A 2.5 kb Xhol Mul containing genomic fragment isolated from vpl4-2274 was subcloned in pBluescript (Promega) and PCR was used to amplify and clone a 1 kb

PAGE 43

37 sequence that flanked the MuJ insertion element. A oligonucleotide primer specific for the inverted terminal repeat of MuJ was used as the 5' primer (5'-CCATAATGGCAATTATCTC-3') and the T7 sequencing primer for the vector was used as the 3' primer (TAATACGACTCACTATAGGG-3'). The resulting ~1 kb fragment was subcloned to pBluescript-SK. Conversion of Xanthoxin to ABA in Cell Free Enzyme Extract This assay was performed by Dr. Zeevaart's group at Michigan State University on embryos prepared in (jainesville. Embryos were dissected from self-pollinated ears of homozygous mutant and wild type segregants of an F3 family. Embryos were homogenized in 0.2 M potassium phosphate (pH7.5) and 10 mM DTT and the extract was centrifiiged to remove insoluble material. The supernatant soluble protein fraction was desalted using a G-25 Sephadex spin column. Enzyme assays contained 1 mM PMSF, 0.25 mM EDTA, 1 \ig 9-cis xanthoxin and the crude enzyme extract. The amount of ABA produced was determined byGC-MS. Results Isolation of Yiviparousl 4 Mutant g All of the characterized viviparous mutants of maize (vpJ, vp2, vp5, etc) are blocked either in the perception or in the synthesis of ABA (McCarty et al. 1991, Neill et al. 1987). A search for viviparous mutants in maize was considered to be an efficient way to isolate ABA deficient mutants. In a screen of the maize strains containing active Robertson's Mutators transposons, two alleles of a new recessive mutant, viviparous] 4(\p J 4), were identified and designated as vp 14-227 4 and vp 14-3250. The

PAGE 44

38 vpl4 mutant kernels have a weak viviparous seed phenotype. In field growth conditions, the embryo shoot axis of the mutant seeds typically elongates but fi-equently does not rupture the pericarp to initiate precocious germination. Thus, most mutant seeds are desiccation tolerant and are capable of germinating. The penetrance of the viviparous phenotype is variable from season to season. In the green house growth conditions where plants were well watered, the penetrance was relatively higher, as shown in Fig. 2-1, and most of the vp 14-2274 kernels that grew under these conditions also germinated fully to form shoots and roots. Some shoots were strong enough to break the pericarp while others were restrained by the pericarp. Unlike vpl (McCarty et al. 1989), anthocyanin pigmentation is not afifected by the \>pl4 mutation, and if rescued, these germinated kernels can develop into healthy, normal plants (Fig. 2IB). The B-A translocations in maize allow recessive genes to be located on the correct chromosome arm in the Fl population. At the second pollen division, the supernumerary B chromosome fi-equently nondisjoins so that one sperm cell of a pollen grain has two B chromosome and the other has none. Various maize T-B lines were created such that the B chromosome carries a specific segment translocated fi-om an A chromosome. If a mutant is located on a translocated segment, a cross between a mutant and the specific TB line will produce some embryos displaying the mutant phenotype in Fl because nondisjunction will produce hemizygous embryos deficient in the tested wild type gene. In crosses using homozygous vpl 4-2274 and vpl 4-3250 as female to a pollen parent that carried the T-BlLa B-chromosome translocation (Birchler 1995), a similar

PAGE 46

40 viviparous phenotype was uncovered, indicating that vpl4 is located on the long arm of chromosome 1. To resolve the relationship of vpl4 relative to other viable viviparous mutants of maize, homozygous vp 14-2274 plants were crossed to vp8 (located on IL, Robertson 1955), vplO (Smith and NeufFer 1992) and other unmapped, though phenotypically similar viviparous mutants including vp*3286 and vp*3239 (McCarty 1995) The vp 14-2274 mutant complemented each of these mutants, indicating that vpl4 is a new locus. It was thus named viviparmsN (ypl4). vpl4 Is a Developmentally Specific ABA Deficient Mutant In the relatively harsh Florida Summer field growth conditions, the vpl4 mutant plants showed no discernible phenotype compared to wild type plants (Fig. 2IB). In contrast to other known ABA deficient mutants such as abal of Arabidopsis (Rock and Zeevaart 1991), sitiens and flacca of tomato (Taylor 1991), and aba2 of tobacco (Marin et al. 1996), vpl4 leaves were not prone to wilting. When grown in the green-house where plants were watered daily, the detached leaves of vpl4 mutant seedlings showed a distinctly greater rate of water loss in comparison to wild type siblings and an inbred W22 (Fig. 2-2D). This difference was detected within 5 minutes following leaf detachment, indicating that the stomata in vpl4 leaves did not close as quickly as those of the wild type leaves. However, determination of the bulk levels of ABA in leaves detected no significant difference between vpl4 mutant and wild type siblings (refer to Fig. 2-8B).This apparent contradiction raised a question as to what role changes in the bulk ABA levels have in regulating stomata. It has been previously documented that in stressed leaves stomata close before any increase in bulk ABA can be detected (Walton et

PAGE 47

41 1 Minutes After Detachment Fig. 2-2. Response of exogenous ABA on germination of embryos cultured in vitro (A.B.C), and water loss rates of detached leaves (D) of wild type and vpJ4 mutants. 3 Embryos harvested aseptically were placed on MS medium supplied with different concentrations of ABA. Relative fresh weight (A), shoot(B) and root elongation(C) as a percentage of a control (without ABA) were measured after four days, diamond: W22, circle: wild type, square: vpl 4-2272, triangle: vp 14-3250.

PAGE 48

42 al. 1977). Harris and Outlaw (1991) reported that the increase of ABA in guard cells was faster than changes in other leaf cells when stress was imposed, and more importantly, the guard cell ABA increase correlated with the stomata aperture. Thus, one possible explanation of the water loss phenotype in vpl4 was that it affected a small pool of ABA that regulated stomata aperture. If so, this fast water loss phenomenon may be suppressed in the field grown vpl4 plants by stress induced ABA since ABA can translocate easily. To test whether vivipary of vpl4 can be inhibited by exogenous ABA, developing embryos were cultured in vitro in media containing different concentrations of ABA. Embryos 16 and 18 days after pollination were used because that stage allows fiilly developed embryos to be isolated before any sign of axis elongation occurs. After 4 days, growth parameters were measured. The vpl4 mutant embryos exhibited the same sensitivity to ABA as did wild type (Fig. 2-2). Precocious germination of vpl4 and wild type embryos was also inhibited by exogenous ABA at the same level, i.e. germination was completely inhibited at 10"' M ABA. Embryo growth was inhibited at ABA concentrations greater than 10"* M, consistent with previous reports on other ABA deficient mutants (Robichaud et al. 1986). It was noted that root growth of vpl4 mutant embryos showed an enhanced sensitivity to ABA inhibition compared to the wild type (Fig. 2-2C). The severe allele vp 14-2274 was somewhat more sensitive than the less severe allele vpJ4-3250, in this respect. Together these data suggested that the vivipary of vpI4 is likely attributed to a deficiency in ABA. However, the enhanced ABA sensitivity of the mutant roots is not explained by this conclusion. In collaboration with Dr. Jan A. D. Zeevaart's lab at Michigan State University, the endogenous levels of ABA in embryos were determined using GC-MS. Embryos of

PAGE 49

-: ^ .-43 Table 2-1. ABA levels are reduced in vp 14-227 4 embryos. DAP* WT "iih vnl 4-7774 ABA ng/g fr.wt. (%) 16 83.3 (100) 23.3 (27.8) It 127.5(100) 36.1 (28.3) 20 71.0(100) 43.7(61.5) *DAP: days after pollination 16, 18, and 20 days after pollination (DAP) were sampled in order to span the time of ABA regulated gene expression in embryos (McCarty et al. 1991). The vpl4-2274 embryo contained only 28% of the wild type level of ABA at 16 and 18 DAP, and 61% as much as wild type at 20 DAP (Table 2-1). Visible elongation of the embryo shoot axis in the mutant is generally observed at 20 DAP. The significant level of residual ABA may account for the relatively weak penetrance of the vp 14-227 4 phenotype. Lower but still significant levels of residual ABA (-10% of wild type) are also evident in embryos of the strongly viviparous carotenoid deficient mutants (e.g. vp5 ) of maize (Neill et al. 1986). The extremely wilty mutant of tobacco aba2 also retains 23-48% of wild type ABA in leaf tissues (Marin et al. 1996). The origin of this residual ABA is still unknown. ABA levels in other vegetative organs such as leaves and roots appeared to be less affected since no differences were detected in nonstressed root and leaf tissues. The total stress induced ABA synthesis measured after 5 h in detached vp 14-227 4 leaves was about 25% lower than the wild type siblings (data not shown). Together these data indicated that vpl4 is an

PAGE 50

44 ABA deficient mutant that predominantly affects the ABA synthesis in developing embryos. Analysis of ABA Biosynthetic Pathway Intermediates in vpI4 Embryos ABA is proposed to be synthesized through a sequence of reactions that includesepoxidation of zeaxanthin to violaxanthin, subsequent isomerization of all-trans violaxanthin to P-c/j-violaxanthin, followed by an oxidative cleavage of 9-c/s xanthophylls to produce xanthoxin which is subsequently converted to ABA (Zeevaart and Creelman 1988, Walton and Li 1995). To locate the blocked step, the C40 carotenoids and xanthophyll intermediates in vp]4 mutant embryos were analyzed by HPLC in collaboration with Dr. Zeevaart' s lab at MSU. No significant differences in the levels of C40 carotenoid precursors were detected between \pJ4 and the wild type embryos at 16 or 18 DAP (Fig. 2-3). The identities of those absorption peaks were calibrated either by standard compounds or by published spectral data (Mobiar and Szabolcs 1979, Rock and Zeevaart 1991, Parry and Horgan 1992). The mutant embryos have comparable amounts of zeaxanthin, violaxanthin, and slightly increased 9-cis violaxanthin and 9-cis neoxanthin contents. The blocked step in ^pJ4 is clearly different than in other viviparous maize mutants such as \p2, vp5, vp7 and vp9 which are depleted in those ABA synthetic precursors (Neill et al. 1986). The vpl4 mutation also differs fi-om the ABA deficient mutant of Arabidopsis, abal, or the equivalent mutant of tobacco aba2, which are blocked in the epoxidation of zeaxanthin and have markedly lower violaxanthin, 9-cis xanthophyll contents (Rock and Zeevaart 1991, Duckham et al. 1991, Marin et al. 1996). The normal or slightly increased 9-cis violaxanthin and 9-cis neoxanthin levels also

PAGE 51

45

PAGE 52

indicated that vpl4 is unlikely to be blocked in the double bond isomerization step. These results suggested that the blocked step in vpl4 is downstream of the 9-cis xanthophylls in the pathway. The remaining steps include oxidative cleavage of 9-c xanthophylls to form xanthoxin, xanthoxin conversion to ABA-aldehyde, and oxidation of ABA-aldehyde to ABA. Table 2-2: Xanthoxin conversion to ABA in ip7'/-227^ embryos. Embryos (20 DAP) mg protein ABA(ng) WTsib 100 18.9+/-2.1 vpl4-2274 100 35.0+/-1.4 Many of the known ABA deficient mutants are blocked in the conversion of ABA aldehyde to ABA possibly because this step requires a molybdenum-Fe cofactor, the synthesis of which requires several steps. These mutants include aba3 of Arabidopsis (Leon-Kloosterziel et al. 1996, Schwartz et al. 1997), flacca, sitiens of tomato (Parry and Horgan 1988), ckrl, ibal of tobacco (Parry et al. 1991, Bitoun et al. 1990), nar2a of barley (Walker-Simmons et al. 1989), and droopy of potato (Duckham et al. 1989). A convenient approach to assess the steps downstream of xanthoxin is to incubate xanthoxin with cell free extracts and assay for production of ABA. The ability of vpl4 mutant embryos to convert xanthoxin to ABA was determined in collaboration v^dth Dr. Zeevaart. In this assay, homozygous vp 14-2274 embryos were found to be fiilly capable of converting xanthoxin to ABA, and did so, surprisingly, at a somewhat rate higher than wild type (Table 2-2). The reason why these enzymes are up-regulated in the mutant

PAGE 53

47 remains to be determined, but these data indicated that vpl4 was not blocked in the steps involved in xanthoxin conversion to ABA. Based on our knowledge of the proposed ABA biosynthetic pathway, the results of enzymatic assay and HPLC analysis of precursors suggested that vpN is blocked in the cleavage of 9-cw xanthophylls to form xanthoxin. Theoretically, the suggested vpl4 block would be directly confirmed by the detection of a depletion of the product xanthoxin. However, determination of xanthoxin is experimentally diflScult and may not be conclusive in any case because the normal level of xanthoxin in all the plant tissues is reportedly extremely low (Nonhebel and Milborrow 1987, Parry et al. 1988 1990).These data provided a compelling though circumstantial case, that vpl4 is the long-sought mutant blocked in the oxidative cleavage step of the ABA biosynthetic pathway. Molecular Analysis of viviparousN Because both alleles of vpl4 were isolated fi-om maize strains containing active Robertson's Mutator (Mu) transposons, it is possible that they were tagged by Mu insertions. To clone the vpJ4 gene, vp]4 was outcrossed to W22 in order to reduce copy numbers of the Mu family elements. Statistically, one outcross to a low Mu copy number strain such as inbred W22 is expected to reduce the copy number by 50%. Then a segregating population was created fi-om the outcrosses. Genomic DNA fi-om lines segregating the mutant phenotype upon selfing (Segregating, S) were analyzed by Southern blot hybridization to compare lines that did not segregate (Nonsegregating, NS, i.e. homozygous for wild type VpJ4). An internal TthllJI fi-agment of the MuJ was used as a probe because it will detect most members of the Mu transposon family. Southern

PAGE 54

48 blot hybridization identified a 2.5kb Xhol fragment that co-segregates with the mutant (Fig. 2-4 A). This fi-agment was cloned by screening a subgenomic X-gtlO library constructed using size selected DNA enriched in 2.5 kb fi-agments. Sequencing of this clone confirmed the presence of a Mul insertion in that fi-agment. The Mul flanking sequence (~1 kb) was amplified by PGR and used subsequently to isolate overiapping wild type genomic clones and cDNA clones from an embryo cDNA library. To test whether or not the cloned sequence was from VpJ4, we analyzed vpI43250, an independent allele that was also isolated from an active Mu line. Southern blot analysis of vp 14-2274 and vpl 4-3250 DNA showed evidence that each contained a Mul sized polymorphism that was not present in any of six possible progenitor strains that founded the Robertson's mutator population (Fig. 2-4B). The corresponding 6.0 and 7.5kb EcoRl fragments were cloned from vpl 4-32 50 and subsequent DNA sequencing detected slMuI insertion approximately 1 kb upstream of the vpl4-2274 Mul insertion in the 7.5 kb fragment (Fig. 2-4C). Restriction fragment length polymophorism (RFLP) mapping of Vpl 4 probes using recombinant inbred populations (Ben Burr and Frances Burr, pers. comm.: Maize Genome Center, University of Missouri) indicated that vpl 4 is located approximately 50 map units proximal to vp8 on \L in the vicinity of the bz2 locus. This mapping location was in full agreement with the TB translocation experiment which placed vpl4 on IL. In addition, a possible duplicate locus was detected and mapped to chromosome 5S. This finding is consistent with a large body of RFLP evidence that IL and 5S are duplicated

PAGE 55

49 if it II 2.5 kb^;*| S NS B CM CM 4 ^ s o a a a z § ^^Imw^' '^^mh^ 'WMI^ '^mI^ '^iMiN^CN 11kb(5S) 7.5kb e.Okb (1L) Probe: Mu1 PI P2 Mul vp14-2274 vp14^2S0 Mul Fig. 2-4. A. Southern analysis of a family segregating vpl4-2274 using a fragment oiMul as a probe. DNA was pooled from several seedlings of each line. The arrow points to a 2.5 kb fragment identified as co-segregating with the vpl4 mutant. B. DNAs of six possible progenitors, \>pl4 mutant and wild type plants were hybridized with !LVpl4 specific probe (PI as indicated in C). On the right, W22 DNA was hybridized by probe2 (P2) at low stringency. The closed arrows point to a 6 kb vpl4 (mapped on IL) and a putative duplicate (1 1 kb mapped on 5S). Open arrows point to the closely related fragmaits. C. The restriction enzyme map of a 6Kb EcoRI fragment containing the VpJ 4 gene and the locations of the Mm7 insertion sites in each allele. The seguences corresponding to two probes used are marked as PI and P2. E, EcoRI; N, NotI; X, Xhol, Nc, Ncol.

PAGE 56

50 C:AACAACAGACTACGGAGGA(XG
PAGE 57

51 segments in the maize genome (Helentjaris et al. 1988). Southern blot analysis using the cDNA sequence as a probe suggests that vp]4 belongs to a small gene family. At moderately high stringency, two major EcoRI fragments (6.0 and 11/3.5 kb) in inbred W22 genomic DNA were detected corresponding to the IL and 5S copies respectively. Four to six fragments hybridized at lower stringency (Fig. 2-4B). Vpl^ Is an Intronless Gene To reveal the structure of the VpJ4 gene, a cDNA library was constructed in XZAP using PolyA enriched RNA isolated from the 18 DAP W22 inbred embryos. Ten nearly foil length cDNA clones were recovered from lO' primary recombinants when the cDNA library was screened with the Ikb flanking probe (P2). Sequencing of the cDNAs and the 6.0 kb EcoRI genomic clone indicated that vpJ4 contains no introns in the transcribed region. A putative TATA box was located in the genomic sequence 80 bp upstream of the longest cDNA. Analysis of the cDNA identified a 1812 bp open reading frame predicted to encode a 604 amino acid protein with a calculated molecular mass of 65.5 kilo-daltons (Fig. 2-5). VP 14 Protein Is Related to a Cleavage Dioxveenase in Bacteria The BLAST algorithm was used to search for related protein sequences in the nonredundant NCBI protein database. VP14 had a strong similarity to lignostilbene dioxygenase (LSD) of Pseudomonas paucimobilis (Kamoda et al. 1993a) and a weaker, but significant similarity to human RPE65, a protein highly specific to the retinal pigment epithelium of the eye (Hamel et al. 1993), and its mammalian homologs. In addition, two related proteins are potentially encoded in the Synechocytis cyanobacterial genome. A

PAGE 58

multiple sequence alignment of the proteins shown in Fig. 2-6 suggested strongly that the LSD, RPE65 and VP 14 are related proteins. The proteins align closely at their C-termini. Several blocks of similar sequence in proteins are clustered around four conserved histidines and one potentially conserved tyrosine residue. In dioxygenases of known structure, conserved histidine and tyrosine residues are typical ligands for the non-heme iron cofactor. It has been confirmed that the VP 14 protein expressed in E. coli as a GSTfusion protein contains significant amounts of bound non-heme iron (data not shown). LSD also contains non-heme iron (Kamoda et al. 1993a, b). These results strongly reinforce the conclusion that LSD and VP 14 define a new class of non heme iron proteins. The VP14 sequence includes an additional 100 amino acid N-terminal extension relative to LSD and RPE65. This N-terminal peptide is consistent with the properties of a chloroplast transit peptide (Cline and Henry 1996) and may serve to target VP 14 to plastids where the initial step of ABA synthesis is believed to occur (Zeevaart and Creelman 1988). LSD catalyses an oxidative cleavage of the central double bond of lignostilbene to form two molecules of the corresponding aldehyde (vanillin) as illustrated in Fig. 2-6. This activity has been confirmed for the recombinant protein expressed in E. coli (Kamoda et al. 1993a, b). This oxidative cleavage reaction is chemically analogous to the oxidative cleavage of the double bond in xanthophylls that initiates ABA biosynthesis in plants (Fig. 2-7). These results together with the biochemical characterization, supported the hypothesis that Vpl4 encodes the 9-cis xanthophyll cleavage dioxygenase of the ABA biosynthetic pathway. Of course, still more convincing evidence would be the direct demonstration of the VP 14 cleavage activity in vitro.

PAGE 59

53 H H n M m ^ flf t> a M W It f >* ft M H M o H i< I *t 4 4 *4 H •> R H h Ai X Ik t4 W M a II h k M ^ al 14 M <* ^ It 4 N o> M >• a o a >4 i • O P>i Pi l4 M M M H > M a M ft M M a ft :#( H > I ft M M M a ft M m V 9 t* >) >l ft l< < k a H w 'V n M M H a tt ^ Ih N K M M > 4 M H M •! Q M M M i H M I H M I h I h i I M M H ( • U t ft I tt O M I 9 0 H n o C ij 01 > 1 u I I t v4 l V M m m o m h k *t N M & 4 M B li Ik H h > W rl I* 'V M M M 4 Q li N N ^ M M > J w M H •4 n H 31 tSMHK h > << 4 a F M <4 M H • >4 N o ft j ft a • a a M • a M iiiiCfiiit: P a M a a a =iiiiii:;ijiiii:Tiliii: a a ei > a H N • M H H n M h lb Ik n • M it > 3 M H )• ^ x K M 4 a K Ik M • M ( > ftJ M K M •< Q R lb H ^ W M t iJ M t4 H H W M H lO V M H M 4 n h h Ik Ik H h M E> a M M • O 4 4 e M M M ft n t4 M 4 e ^ Ik M 01 & iJ o -a at 't-i o O ft 1 s ^ 2 2 -2 o J2 .a I J3 4> 2 i

PAGE 60

54 VP14 ABA Lignostilbene dioxygenase: Fig. 2-7. Proposed reactions catalyzed by VP14 and the reaction catalyzed by lignostilbene dioxygenase (LSD) reaction (Kamoda & Saburi, 1993b).

PAGE 61

55 VpI4 Expression Is Developmentally and Environmentally Regulated The expression of VpJ4 was examined in embryos by northern blot analysis (Fig. 2-7 A). A 2.6 kb transcript was detected in the developing embryos of wild type, i.e. W22 and the corresponding non-segregating Vp 14-2274 (NS-2274) strain, whereas a series of size altered transcripts were detected in the homozygous vp 14-227 4 mutant embryos. The Vpl4 mRNA levels were greatly reduced in the homozygous vpl4-3250 embryos, and a similar pattern of altered sized transcripts was detected. The increased size of the mutant vpl4-2274 and vpl4-3250 transcripts (about 4.0 kb) is consistent with a transcriptional readthrough of the A/m7 insertion (1.4 kb) in v^dld type Vpl4 gene (2.6 kb). The smaller sized transcripts in the vpl4 mutant embryos may the aberrantly spliced products due to the Mul insertion. Considering the evidence that Vpl4 belongs to a gene family and the probe may hybridize to homologous sequences, the 4.0 kb transcript in vpl4-2274 was used to identify vpl4 specific RNA in the northern hybridization analysis. In Northern analysis of other tissues, this mutant vpl4-2274 transcript (4.0 kb) was also detected in roots of the mutant seedlings indicating that the Vpl4 gene is normally expressed in roots as well as embryos (Fig. 2-8A). No Vpl4 message was detected in leaf tissues that were grown in normal conditions. Stress induction of ABA synthesis has been shown in a variety of plant species (Wright and Hiron 1969, Ingram and Bartels 1996), and this induction can be inhibited either by transcriptional or by translational inhibitors (Li and Walton 1990), indicating gene expression is involved in this process. Previous studies have suggested that the cleavage reaction may be the key regulated step in the ABA biosynthetic pathway

PAGE 62

56 Embryo Leaf Root 7 514 7 332 6 232 ABA (ng/g tissue) Fig. 2-8. A. Northern analysis using P2 as a probe. The aberrant sized transcripts in vpl4 mutants were marked with open arrows. B. Water stress induction of Vpl4 expression in detached leaves as probed with Vpl4 specific probe (PI) or induction of homologous genes as probed with P2. A hybridization with aS'f/5'i (sucrose synthase) probe was provided as control. The ABA levels in Aose leaves were provided under the lanes.

PAGE 63

57 (Zeevaart and Creelman 1988, Walton and Li 1995). Thus, it was predicted that expression of the key 9-cis xanthophyll cleavage enzyme might be stress inducible. To determine whether VpJ4 expression is environmentally regulated, detached leaves were subjected to a water stress treatment at room temperature for 6 hours. This resulted in a 40 70 fold increase of endogenous ABA levels in the vp 14-227 4 and wild type leaves (Fig. 2-8B). Two probes were used in the northern blot hybridization of stressed leaves, probe 1 (PI), a 3' untranslated sequence of Vpl4 that is specific to Vpl4 as evidenced by its specific hybridization to Vpl4 transcript (Fig. 2-4B), and broad specificity probe 2 (P2) that can cross hybridize to homologous sequences in W22 genomic Southern (Fig. 2-4B). The Vpl4 message was dehydration inducible as indicated by the strong signal of \pl4-2274 transcript (4.0 kb) in the stressed vpl4-2274 leaves and the strong hybridization of PI to the 2.6 kb transcript in stressed wild type leaves. Using P2 at low stringency, f^i-Z-related gene expression was detected. A 3.5 kb transcript increases in stressed leaves, and a smaller VpN-rtXiAoA transcript was down regulated upon stress. A probe fi-om the Susl (sucrose synthase) gene was a control for variation in poly(A)-RNA loading. The loading is about equal for nonstressed and stressed samples vwthin each genotype, although it is variable among different genotypes. In any case, these data indicated, as predicted, that expression of Vpl4 and related genes are induced by water stress in plant leaves. This correlates with the stress induced ABA accumulation. However, the \>pl4 mutant only partially blocks stress induced ABA synthesis in leaves.

PAGE 64

58 Discussion Biochemical analysis shows that the ^pJ4 mutant selectively blocked ABA synthesis in developing embryos and suggests that the Vpl4 gene has a role in developmental control of hormone synthesis in maize. Furthermore, the sequence similarity of VP 14 to LSD provided evidence that ^pJ4 encodes a dioxygenase enzyme responsible for oxidative cleavage of 9-cw xanthophylls to xanthoxin. Variable Dependence of Seed Maturation on ABA in Different Plant Species ABA is a key regulator of seed maturation and dormancy (Koomneef and Karssen 1994, McCarty 1995, Rock and Quatrano 1995). As documented by ABA deficient mutants (e.g. vp5, vpl4, etc) and the ABA insensitive mutant, vpl mutant of maize, a failure to synthesize or to perceive the ABA signal results in a viviparous phenotype. Viviparous seeds bypass the process of maturation, fail to acquire desiccation tolerance, and initiate germination processes while kernels are still attached to the mother plant. Although some key regulators of the seed maturation pathway such as the VP 1 and ABI3 factors are highly conserved between maize and Arabidopsis, the dependence of the seed maturation pathway on ABA signaling differs markedly between the two species. In Arabidopsis, an ABA dependence of seed maturation is manifest only in mutant backgrounds in which the ABI3 flinction is attenuated (Koomneef et al. 1994). The mature seed phenotype of the aba J mutant shows that the activity of ABI3 is suflBcient to resuh in a normal seed maturation in the absence of ABA. This view is further supported by the fact that even the severe ABA deficient mutants of Arabidopsis are able to

PAGE 65

complete seed maturation and form desiccation tolerant embryos (Giraudat 1995). In tobacco, the ABA deficient mutant of aba2 which exhibits severe wilting, also completes the normal seed maturation processes although a reduced dormancy was observed (Marin etal. 1996). In maize, the dependence of seed maturation on ABA appears to be relatively tight. For the vpJ4 mutant, a 70% reduction in bulk ABA levels in embryos during mid embryogeny is evidently sufficient to disrupt seed maturation and thus cause an incompletely penetrant viviparous phenotype. This finding is consistent with the evidence that several strongly penetrant viviparous mutants (e.g. vpS) of maize have lower, but still significant levels of residual ABA (-10 % of wild type) in the seed (Neill et al. 1986). Overall it suggests that a threshold level probably greater than 30% of the wild type level ABA is required to complete embryo development in vivo. Hence the viviparous, or more precisely the elongated embryo shoot axis phenotype, provides a very sensitive screen for ABA synthetic mutants in maize. This may have contributed to the isolation of vpI4 and other new potentially ABA related mutants in maize (McCarty 1995). Together these resuks suggest that dependence of the seed maturation pathway on ABA signaling is variable among plant species, and in Arabidopsis the default activity of the ABI3 pathway is normally sufficient to complete seed maturation. Overlapping Sources of ABA Svnthesis in Plant Development The significant levels of residual ABA in embryos and normal levels in vegetative tissues of homozygous vpl4 mutants suggested a possibility of overlapping sources of ABA. Three possible sources may account for the residual ABA in vpl4 embryos. 1) the

PAGE 66

60 existing alleles may be leaky, 2) related genes may be expressed at low levels in the seed, 3) ABA may be transferred from the surrounding maternal tissues that express related genes. Both alleles of vpl4 are caused by a Mul insertion, vpl4-2274 in the central portion of coding region and vp 14-3250 in the 3' untranslated region (Fig. 2-4). It seems unlikely that both alleles of vpl4 are leaky because in order to produce a functional VP 14 protein, a perfect slicing of the Mul insertion is likely required from the mutated form of the transcript. Northern blot analysis detected a 4. 1 kb transcript that is consistent in size with a A/i/7 insertion in the normal 2.6 kb wild type Vpl4 transcript, and smaller than 2.6 kb sized transcripts that is presumably due to alternative splicing or premature termination resuhed from the A/wy insertion (Fig. 2-8). A low level of Vpl4 related gene expression in embryos is supported by the recovery of a Vpl4-X\\iQ sequence from an embryo cDNA library. Partial sequencing of this clone revealed that the potentially encoded protein is highly homologous to VP 14 (data not shown). However, a screen of another cDNA library constructed from mRNA from W22 embryos at 18 DAP did not recover this sequence among the ten Vpl4 clones that were isolated. One possible explanation is that timing of expression of Vpl4 and this Vpl4-VikQ gene differs during embryo development. With respect to a third possibility, the normal ABA levels in vegetative tissues of vpl4 mutants and the observation that mutant plants develop normally under field conditions implies strongly that the Vpl4 gene accounts for only a subset of the ABA biosynthetic activity in the plant. Partial suppression of the vpl4 phenotype by ABA transferred from maternal tissues might explain the variable penetrance of the viviparous phenotype we observe under field conditions. Maternal ABA synthesis induced by stress can significantly affect seed development in maize (Ober and Setter 1992). Thus, these

PAGE 67

• ^ ... gj results suggest that embryo ABA may come from at least two sources, synthesis by the embryo itself, which accounts for a majority of the ABA, and ABA translocation from other tissues such as leaves or roots. The endosperm evidently does not contribute significant ABA to the embryos, consistent with the earlier conclusions of Robertson (1952) and Neill et al. (1983). Using T-B translocations in maize, they created kernels with embryos homozygous for ^J>5, but wild type endosperm. The phenotype of these embryos were still viviparous and ABA deficient. Several lines of data provided indirect evidence that fimctionally VP 14 equivalent genes are expressed in vegetative tissues of the plant. On moderate stringency Southern blots, the 1 kb VpI4 probe (P2) hybridized to nine EcoRI fragments of inbred W22 DNA (Fig. 2-4A). Sequencing of a VpJ4 duplicated sequence (the 1 1 kb EcoRI fragment in Fig. 2-4 A) indicated that it may potentially encode a protein with similar fiinction to VP 14 (see chapter 5). The northern blot analysis of leaves also detected expression of related genes (Fig. 2-8B). In addition, we have isolated several distinct, but related cDNAs from root and leaf libraries (B. C. Tan and D. R. McCarty, unpublished data). The normal levels of endogenous ABA in leaves and roots clearly indicate that ABA is synthesized in vegetative tissues of the mutant. And the activity of ABA synthesis is only slightly affected in stressed ^J}J4 -2274 leaves since ABA can still rise about 40 fold, although not as much as in wild type (Fig. 2-8B). Data base searches identified several Vpl4 related sequences in the rice, maize and Arabidopsis EST collections. For these reasons, we suggest that Vpl4 belongs to a small family of differentially regulated genes that contribute to developmental control of ABA biosynthesis in plants.

PAGE 68

62 Role of Vpl4 in Regulation of Stomata in Leaves and in Root Development Many of the characterized ABA deficient mutants display a wilty plant phenotype, indicating that regulation of stomata closure is affected in these mutants (refer to Chapter 1 for details). Field grown vpI4 mutants did not show any discernible plant phenotype in addition to the seed vivipary. However, detached leaves of the vpl4 mutant seedlings grown in the greenhouse with abundant water supply show enhanced rates of water loss. This feature suggested that stomatal closure is affected by the vpl4 mutation. Measurement of the ABA levels in leaves of these plants detected no significant difference between vpl4 mutants and the wild type siblings, suggesting that the bulk pool of ABA is unchanged. Northern blot analysis has confirmed that in addition to embryo expression, Vpl4 is also expressed in roots and is stress inducible in leaves (Fig. 2-8), indicating the potential involvement of Vpl4 in the ABA synthetic activity in these tissues. However, Vpl4 may only account for a small set of the ABA detected in those tissues based on these results. In many plant species, bulk ABA level increases dramatically in stressed leaves (reviewed by Ingram and Bartels 1996). However, when the time course is taken into account, the closing of stomata generally occurs before any significant changes in ABA content can be detected in the leaves. One possible explanation for this time discrepancy may involve a rapidly localized accumulation of ABA in or around guard cells, either through re-distribution of ABA among leaf cells or through highly localized de novo synthesis of ABA, e.g. only in guard cells or the neighboring cells. In support to this suggestion, a rapid increase in the ABA concentration of a single guard cell is correlated with the closure of stomata in stressed leaves (Harris and Outlaw 1991). However, this study did not identify the sources of the ABA. Thus, the role of Vpl4 in regulating

PAGE 69

63 stomata closure may be explained by the existence of a highly localized pool of ABA in leaves which specifically affects stomatal closure. ABA in this pool may resuh fi-om remobilization of ABA fi-om other sites such as roots, or fi-om a highly localized expression of Vpl 4, in a limited number of cells, which may not be detected by northern blot analysis. Vpl 4 is expressed in normal roots which may be a significant source of ABA in leaves that regulates stomatal aperture (Fig. 2-8). In several species, variations in stomatal conductance are well correlated with the ABA present in xylem, and the initial stomatal close occurs before the shortage of soil moisture causes any measurable change in the water status of the leaves (Zhang and Davies 1989). Another feature of germinating ypl4 embryos is that their roots show a slightly enhanced sensitivity to exogenous ABA compared to wild type. ABA is required for root elongation in stressed conditions based on evidence that the ABA deficient vp5 mutant showed poor root elongation compared to wild type (Sharp et al. 1994). The role of ABA in root development and adaptation to environments is less understood. However, the root phenotype of vpl4 may be explained by an increasing hormone sensitivity to compensate for the reduced ABA synthesis. Evidence that VP 14 Is the 9-cis Xanthoohvll Cleavage Dioxvpenase Deficiency of ABA in the vpl 4 embryos indicates that ABA biosynthesis is blocked (Table 2-1). The existence of normal or slightly higher than normal levels of 9-cis violaxanthin and 9-cis neoxanthin in vpl4 embryos suggests that steps leading to synthesis of these two ABA precursors are not affected. In this respect, vpl4 is clearly different fi-om two related ABA deficient mutants, abal of Arabidopsis and aba2 of tobacco, which

PAGE 70

64 accumulate zeaxanthin and deplete violaxanthin due to a block in the epoxidation step (Duckham et al. 1991, Marin et al. 1996, Rock and Zeevaart 1991). A significant increase of 9-cis xanthophylls does not occur in ^pJ4 embryos, possibly because ABA concentration is normally much lower and only accounts for a small fi-action of xanthophyll synthesis (Parry and Horgan 1992). The fact that cell free extract fi-om the vpl4 embryos can convert xanthoxin to ABA ruled out the possibility that ^pJ4 is blocked downstream of xanthoxin. Thus, by elimination, these data suggest that the blocked step in vpJ4 is the oxidative cleavage of 9-c xanthophylls to yield xanthoxin. Two other possible approaches may provide further evidence for this conclusion: 1) confirmation that xanthoxin pools are depleted in the mutant, 2) direct measurement of the enzymatic activity converting 9-cis xanthophylls to xanthoxin, in vitro. However, this has proven technically difiScuh. Firstly, it was reported that xanthoxin levels are extremely low in plant tissues, which in turn provided the evidence that downstream enzymatic activities in converting xanthoxin to ABA may be constitutively active (Sindhu et al. 1987, Li and Walton 1990). Experimentally, direct measurement of xanthoxin in developing embryos is not feasible. With respect to the second approach, an in vitro assay of the xanthophyll cleavage activity has not been developed. In vitro assays of the carotenoid cleavage activity may be hindered by the extremely low abundance of the enzyme, insolubility of the substrates, and complications arising fi-om nonspecific cleavage reactions by enzymes such as lipoxygenases (Creelman et al. 1992, Schwartz et al. 1997). In fact, it is these difficulties and complications that have greatly hindered previous progress in the search and demonstration of such a dioxygenase in the ABA biosynthetic pathway. The incomplete characterization of the notabilis mutant of tomato (Parry and

PAGE 71

65 Horgan 1992, Taylor 1991), has been due to a combination of its leaky nature and the lack of more precise, direct approaches. Molecular cloning of VpJ4 provides an independent line of evidence at the molecular level. VP 14 showed high similarity to a bacterial enzyme, lignostilbene dioxygenase (LSD) (Kamoda et al. 1993). Based on the proposed ABA biosynthetic pathway, cleavage of 9-cis xanthophylls was predicted to involve a dioxygenase type enzyme (Walton and Li 1995, Creelman et al. 1992), which breaks the conjugated double 1 1,12 bond and produces xanthoxin and a C25 apo-aldehyde, each with an aldehyde group. No such dioxygenases have been reported in the plant or animal kingdoms (Prescott and John 1996). The reaction catalyzed by LSD is highly analogous chemically to the proposed oxidative cleavage of 9-cis xanthophylls. Indeed, the bacterial LSD is the only known enzyme capable of catalyzing the specific oxidative cleavage of a conjugated double bond to form two aldehyde products and would, in the absence of other evidence, provide a compelling model for the carotenoid cleavage enzyme (Fig. 2-7). Lipoxygenases can also generate a series of aldehyde cleavage products of carotenoids in vitro, but do so nonspecifically (Creebnan et al. 1992, Schwartz et al. 1997). These resuhs, together with biochemical analyses of the vpl4 mutant, build a strong though circumstantial case that VP 14 encodes the cleavage dioxygenase in ABA synthesis. Recent analyses of the purified recombinant VP 14 have confirmed that it catalyzes oxidative cleavage of 9-cis violaxanthin to xanthoxin in vitro (next chapter).

PAGE 72

Chapter 3 ViviparousN Encodes a 9'-c/5 Neoxanthin/9-c/j Violaxanthin Dioxygenase of Abscisic Acid Biosynthesis in Maize Introduction Abscisic acid (ABA), a sesquiterpene plant hormone, is involved in regulation of many physiological processes throughout plant development, including germination, transpiration (stomata aperture and root conductivity), seed maturation, dormancy and tolerance to environmental stresses (Zeevaart and Creelman 1988, Walton and Li 1995, Giraudat 1995, Ingram and Bartels 1996). ABA is a particularly interesting hormone with regard to its regulation. The levels of ABA can rise or fall about two magnitudes within 4 to 8 hours in response to environmental and developmental changes (Walton and Li 1995). In fungi, ABA is synthesized from Cis intermediates (famesyl pyrophosphate) via a so called "direct pathway"(Zeevaart and Creelman 1988). While in higher plants, ABA is proposed to be synthesized from carotenoids. Blockage of carotenoid biosynthesis in several viviparous mutants of maize (yp2, vp5, vp7 and vp9) also prevents ABA biosynthesis (Neill et al. 1986, Taylor 1991). Incorporation of '*0 from O2 in the -CHO group instead of the ring of ABA strongly suggested oxidative cleavage of epoxycarotenoids as the origin of ABA as opposed to direct synthesis from a C15 intermediate (Zeevaart et al. 1989). Biochemical analysis of stressed leaves indicated 9-cis violaxanthin or/and 9-cis neoxanthin are probable substrates that are cleaved to produce xanthoxin (Li and Walton 1990, Parry et al. 1990). Xanthoxin is subsequently converted to ABA 66

PAGE 73


PAGE 74

68 lignostilbene dioxygenase (LSD) (Kamoda and Saburi 1993a, b) which catalyzes a reaction that is strikingly analogous to the oxidative cleavage of 9-cis epoxy carotenoids in the ABA biosynthetic pathway in plants. In this chapter, purified recombinant VP 14 as expressed in K coli is shown to specifically cleave 9'-c/j neoxanthin and 9-cis violaxanthin to produce the ABA precursor xanthoxin, and a corresponding C25 apoaldehyde. Consistent with other dioxygenases (Prescott and John 1996), recombinant VP 14 contains a non-heme iron which is critical to its activity. Material and Methods Expression of VP 14 Recombinant Protein A fragment containing the coding region and 3' untranslated sequences of the Vpl4 cDNA was amplified by polymerase chain reaction (PGR) using a forward primer (5'-ATGCGGATCCAlGCAGGGTCTCGCCCCG) and a T7 reverse primer. A BamHI site was engineered in the 5 primer very close to the first ATG to facilitate cloning. The amplified fragment was ligated into the BamHI and EcoRI sites of pGex-2T expression vector to express a GST-VP14 fijsion protein in £. coli, JM109. Synthesis of the fiision protein was induced by adding 0.1 mM IPTG to mid-log phase culture grown at 32 C in 2xYT-G medium supplemented with 0.4 mM Ferrous iron. GST-VP14 was purified by binding to Glutathione Sepharose 4B and VP 14 was isolated by digestion with thrombin according to the manufacture's instruction (Pharmacia). Purified VP 14 was quick frozen in aliquots in liquid nitrogen and shipped on dry ice for enzymatic assay at Michigan State University. Determination of Iron The iron content of the purified VP 14 protein was determined by a colorimetric method described by Perciva et al.(1991). 30%

PAGE 75

69 trichloroacetic acid was added to the protein solution to a final concentration of 5% to release the non-heme iron. After centrifligation to remove the precipitate, the 105 \il supernatant was transferred to a new tube and 20 ^1 saturated ammonium acetate, 12.5 ^1 ascorbate (0.12M) and 12.5 ^il FerrbZine (0.25M) reagent were added. After incubation at room temperature for 30 minutes, absorbance at 562 ran was determined. The protein concentration was measured by Bradford method using BSA (bovine serum albumin) as of standard (Bradford 1976) using bovine serum albumin (BSA) as a standard.. Preparation of the Substrates The substrates used for VP 14 enzymatic assay test were isolated by Steve Schwartz and Jan Zeevaart at Michigan State University as described (Schwartz et al. 1997). Briefly, crude thylakoid membranes were isolated fi-om spinach leaves. The carotenoids were saponified and partitioned to ether and then resolved on a semiprep uPorasil HPLC column (Waters). Fractions were collected and the identities of prepared substrates were confirmed by spectrum analysis in comparison with standards or published data (Mohiar and Szabolcs 1979, Parry and Horgan 1991). 9-cis zeaxanthin was prepared by iodine isomerization of the all-trans zeaxanthin (Zechmeister 1962). Enzymatic Activity of Recombinant VP 14 Protein The enzyme assays were performed by Schwartz and Zeevaart. About 6 \ig VP 14 was incubated with each substrate in a reaction buffer containing 100 mM BisTris, pH 6.7, 0.05% Triton X-100, 10 mM ascorbate, 5 ^iM FeS04 and 1 mg/ml catalase. Assays were performed at room temperature under red light to minimize photo-oxidation of the precursors and product. Purified 9-cis neoxanthin was initially used to test for cleavage. The products were analyzed by NP-HPLC and they were C25 apo-aldehyde fi-agment and xanthoxin. For

PAGE 76

70 determination of substrate specificity, VP 14 was incubated with purified all-trcms/9-cis neoxanthin, aU-trans/9-cis violaxanthin and all-trans zeaxanthin / 9-cis zeaxanthin suspended in Triton X-100 and ascorbate. After the reaction, the products were resolved on a TLC plate and stained for aldehyde with 2,4-dinitrophenylhydrazine. Mass-Spectrometry Analysis of Reaction Products As described in detail by Schwartz et al. (1997), the identity of xanthoxin was confirmed by GC-MS of the TMSi derivative according to Gaskin and MacMillan (1991). The C23 apo-aldehydes were analyzed by electron impact mass spectrometry with direct inlet sampling by Dr. Douglas Gage at the mass spectrometry facility at Michigan State University. The expected fi-agmentation pattern of both C25 product was derived fi-om published spectra (Molnar and Szabolcs 1979, Parry and Horgan 1991). Results Recombinant VP 14 Protein Contains Non-heme Iron Biochemical characterization and molecular analysis have provided compelling evidence indicating that Vpl4 encodes the 9-cis epoxy-carotenoid cleavage dioxygenase (Chapter 2, Tan, et al. 1997). To directly test the proposed activity of VP 14, the coding region of the VpJ4 cDNA was fused to a glutathion-S-transferase (GST) gene and expressed in E. coli (Fig. 3-1 A). SDS-PAGE analysis of the affinity purified proteins detected a 90 kd fusion protein (GSTVP 14) that was induced by IPTG. Cleavage of that fusion protein by the site specific proteinase thrombin (LVPRIGS) produced two major proteins, the GST domain and the VP 14 recombinant protein which presumably carried two additional amino acids (Gly-Ser-) at its amino terminus. Binding of GST to

PAGE 77

71 (0 ^ I CO £ CD + > GST-VP 14 VP14 GST B 8 Fe^"^ Standard (|ag/ml) 16 24 32 40 Iff GST VP14 VP148T Fd Fig. 3-1. Expression and purification of recombinant VP 14 from E. coli (A) and determination of iron in the protein (B). The recombinant VP 14 has extra Gly-Serat its N-terminus; VPH^^ bears a 30 amino acid truncation in the putative transit peptide at N-terminus. Fd stands for ferrodoxin.

PAGE 78

72 glutathione sepharose 4B facilitated an easy purification of recombinant VP 14. Because proper folding of VP 14 may require the presence of ferrous iron, the E. coli cultures were supplied with 0.4 mM iron. The purified protein was analyzed for iron using the cleavage produced GST protein as a control. As shown in Fig. 3-lB, purified recombinant VP14 contains significant amounts of iron. Accurate determination of iron ratio in prepared VP14 by FAB-MS in collaboration with Dr. B. W. Smith (Department of Chemistry, University of Florida) revealed a stoichiometry about 0.3, which hints that either some protein lost their iron or the active form of the protein is an oligomer. Later assays which showed that addition of ferrous iron proportionally increased the activity of VP 14 favored the first possibility. VP 14 Clea ves 9-cis Neoxanthin into Xanthoxin and a Cj ^ Product In collaboration with Steve Schwartz and Jan Zeevaart (Michigan State University), the recombinant VP 14 protein was tested for cleavage activity using 9-cis neoxanthin as a substrate. The cleaved products were analyzed by HPLC and thin layer chromatograph (TLC). The expected cleavage products, xanthoxin and the C25 allenic apo-aldehyde, were identified by their UVmS absorption spectra and by mass spectra obtained by Dr. Douglas Gage at the mass spectrometry facility at Michigan State University (Schwartz et aL 1991). The fi-agmentation patterns for the epoxyand allenicC25 cleavage products were nearly identical to published spectra for these compounds (Mobiar and Szabolcs 1979, Parry and Morgan 1991). Exact mass measurements described in Schwartz et al. 1997 match the theoretical mass of the isomeric C25 cleavage products fi-om neoxanthin and violaxanthin (C2JH34O3) is 382.2508. The measured mass of

PAGE 79

73 the C25 product from neoxanthin was 382.2498 with an error of -2.6 ppm. The measured mass of the C25 product from violaxanthin was 382.2501 with an error of-1.8 ppm. Thus, the cleavage is at the 11,12 double bond of 9-cis violaxanthin and 9-cis neoxanthin to produce one molecule of xanthoxin (C15) and one molecule of C25 apoaldehyde (allenic and epoxy apoaldehyde). This cleavage of 9-cis epoxy-carotenoids by VP 14 is well matched with the proposed reaction in the ABA biosynthetic pathway. Xanthoxin was detected with the absorbance at 280 nm, the 9-cis neoxanthin and the C23 cleavage product were detected at 412 nm. As shown in Fig. 3-2, VP14 cleaves 9-cis neoxanthin to produce two products, xanthoxin (C15) and a C25 apoaldehyde which was identified by mass-spectrometry to be C23 allenic apo-aldehyde with molecular mass of 382.2498 with an error of -2.6 ppm. Quantification of the products revealed an equimolar ratio of xanthoxin and the C25 product, consistent with a specific cleavage at the 11-12 double bond of the polyene chain. Non-enzymatic cleavage resulting from photo-oxidation or Fenton reaction would result in random cleavage at different double bond positions (Schwartz et al. 1997). A number of factors were tested to optimize the VP 14 cleavage activity and the resuhs mdicated that molecular oxygen (co-substrate), ferrous iron, and a detergent were all necessary for the in vitro cleavage activity (Schwartz et al. 1997). According to the ^*02 incorporation studies in stressed leaves (Creelman and Zeevaart 1984, Creelman et al 1987, Li and Walton 1987, Parry et al. 1988, Zeevaart et al. 1989), the molecular oxygen is the co-substrate and each atom of O2 is incorporated into the -CHO group of the reaction products, xanthoxin and C25 apo-aldehyde (Parry and Morgan 1991). Depletion of O2 in the reaction completely abolished the cleavage activity. A supplement

PAGE 80

74 B -VP14 Control +VP14 9-cis neoxanthin L 9-cis neoxanthin .^^^ J.. C25 apo-aldehyde \ xanthoxin 1 A4,2 profile A280 profile Fig. 3-2. TLC analysis (A) and NP-HPLCanalysis (B) of the reaction products after 9-cis neoxanthin was incubated with recombinant VP14 purified fi-om E. coli. The TLC plate was stained by 2,4-dinitrophenylhydrazine to detect aldehydes. The test was performed in collaboration with Steve Schwartz and Jan Zeevaart.

PAGE 81

75 of ferrous iron, but not ferric iron, in the reaction mixture resulted in an about 3 fold increase in the cleavage activity. Addition of divalent cation chelator EDTA led to a complete inhibition of the cleavage activity, while removal of EDTA and addition of Fe2+ sufficiently restored the activity. These results support the idea that ferrous iron is a critical co-factor of this epoxy-carotenoid dioxygenase. This is consistent with the substoichiometric amount of iron detected in the purified recombinant VP 14 protein. The requirement for Triton X100 in the assay is probably due to its role in solubilizing the substrate. It may not reflect the in vivo reaction environment since carotenoids are shown to exist in association with plastid membranes (Hartley and Scolnik 1995, Markwell et al. 1992). A slight increase in xanthoxin production by adding catalase is possibly caused by reducing the nonenzymatic degradation of the substrate, e.g. by a Fenton reaction (Schwartz et al. 1997). Ascorbate was added to the reaction to maintain Fe^"^ in the reduced state. 9-cis Configuration Defines the Substrate Specificity of VP 14 Further studies of recombinant VP 14 performed at Michigan State University determined the substrate specificity of the cleavage reaction. The reaction products from various carotenoid substrates were separated on TLC plates and stained with 2,4-dinitrophenyl hydrazine to detect aldehyde as shown in Fig. 3-3. The all trans isomers of violaxanthin and neoxanthin were not cleaved by VP14. Xanthoxin and the predicted C25 products are present only in reactions containing the 9-cis violaxanthin and 9-cis neoxanthin as indicated by the xanthoxin standard.

PAGE 82

76 .jjik jiik VP14 ++ + + + .+std atZ 9cZ aA/ 9cV a/N 9cN Xan Fig. 3-3. Staining of a TLC plate that resolved the reaction products of specific substrate incubated with recombinant VP 14 (Top, adapted from Schwartz et al, 1997). The structure of the tested substrates were shown below. a/V= all tram violaxanthin, a/N= all trans neoxanthin a/Z=== all trans neoxanthin, and 9c= 9-cis isomers. Rl and R2 represented the corresponding residue in all trans configuration.

PAGE 83

9-C/5 zeaxanthin was also cleaved, presumably yielding a Cu and a C25 apoaldehyde which suggested that the 9-cis configuration is a primary determinant of VP 14 substrate specificity. Cleavage of 9-c/j zeaxanthin does not produce xanthoxin, and this reaction is not known in plants possibly because that zeaxanthin exists only in the all trans configuration. The 9-cis zeaxanthin used in above reaction was formed by iodine isomerization of the all-trans zeaxanthin (Schwartz et al. 1997). Other mono-epoxy carotenoids, antheraxanthin and lutein epoxide also exist in the all-trans configuration (Parry et al. 1990) and, as expected, their all-trans isomers were not cleaved by VP 14 (Schwartz, Tan, Zeevaart and McCarty, unpublished data). Thus, Vpl4 encodes the 9-cis neoxanthin/9-CM violaxanthin dioxygenase of the ABA biosynthetic pathway in maize, and this specific carotenoid cleavage activity has been directly demonstrated in vitro. Discussion Cloning and expression of VP 14 reinforced the conclusion made by physiological and biochemical characterization of the mutant that vpl4 is blocked in the oxidative cleavage of oxygenated carotenoids. Furthermore, the activity of VP 14 provided convincing evidence to the identity of the Vpl4 clone. VP 14 as expressed in E. coli can specifically cleave the 1 1, 12 double bond of 9-cw oxygenated carotenoids into a Czj and a Cis product, each with an aldehyde group (Fig. 3-3). Xanthoxin, the precursor of subsequent ABA biosynthetic reactions, is formed only by cleavage of 9-cis violaxanthin and 9'-CM neoxanthin. And one molecule of 9'-ck neoxanthin or 9-cis violaxanthin produces one molecule of xanthoxin and one molecule of C2S apoaldehyde. Thus, it is in good agreement with the proposed reactions of an oxidative cleavage of 9-c/j

PAGE 84

78 cu-abscisic acid Fig. 3-4. The ABA biosynthetic pathway in higher plants and the function of VP14 as a dioxygaiasae. Hie cleaved double bond by VP14 was marked by a arrow. 9"-cis neoxanthin and 9-cis violaxanthin are the substrates to produce xanthoxin and only different residues were shown in the boxes.

PAGE 85

xanthophylls in the ABA biosynthetic pathway in plants as shown in Fig. 3-4 (Zeevaart and Creehnan 1988, Walton and Li 1995). The fate of the C25 fragments in vivo is not known. The affinity of VP14 for 9'-c/5 neoxanthin and 9-cis violaxanthin was not determined. Because the former is more abundant than the latter in a number of species including maize, tomato, bean and spinach (Li and Walton 1990, Parry and Horgan 1992), 9'-cw neoxanthin might be the major substrate of VP14. The decrease of 9'-cw neoxanthin in stressed leaves of a number of species accounted for most of the increase of ABA (Li and Walton 1987, 1990, Parry et al. 1990, Sindhu et al. 1988). VP 14 is also able to cleave 9-cis zeaxanthin, which lacks the epoxide on the ring compared to the normal substrates, suggesting that the 9-cis configuration is a primary factor that determines the substrate specificity. It is reasonable to speculate that both the ionone ring and the distance between the 9-cis double bond and ring structure also determine the substrate specificity. Inferred from that, all of the 9-cis isomers such as 9-cis lutein, 9-cis antheraxanthin and 9-cis zeaxanthin may potentially serve as the substrates of VP 14. However, the 9-cis isomers of these carotenoids are not found in plants (Parry et al. 1991). The 9-cis zeaxanthin used for cleavage reaction was made by iodine isomerization of the all-trans-zeaKmtim in vitro (Zechmeister 1962). The predominant cis xanthophyll in either light or dark grown bean leaves was found to be 9-cis neoxanthin, and 9-cis violaxanthin only accounted for a small percentage (Li and Walton 1990). What role, if any, light plays in the isomerization of all-trans carotenoids to generate 9-cis isomers needs to be answered, however, roots grown in darkness normally contain 9-cis carotenoids. The specific abundance of 9-cis violaxanthin and 9-cis neoxanthin in dark grown leaves suggests involvement of a specific isomerase in the trans to cis conversion.

PAGE 86

80 The rapid and specific increase of 9-cis violaxanthin and 9-cis neoxanthin in stressed dark grown bean leaves is consistent with that hypothesis (Li and Walton 1990). The lignostilbene dioxygenases from Pseudomortas and VP 14 comprise a novel class of dioxygenases that catalyze similar double bond cleavage reactions and generate products with an aldehyde group. The conserved sequences have also been identified two ORFs in the complete genomic sequence of the cyanobacterium Synechocystis. Functions of the gene products have not yet been determined. However, related cyanobacteria are known to synthesize at least two apo-carotenoids fi-om P-carotene, cyclocitral and Pionone (Fresnedo et al. 1991). The demonstration of VP14' role in cleaving carotenoids may clarify vitamin A biosynthesis in animals as it was proposed that a similar cleavage reaction of P-carotene to generates vitamin A. This 15, 15-dioxygenase has been controversial (Yeum 1995, Wolf and Phil 1995). Thus, conserved sequences may be usefiil in identifying such carotenoid cleavage dioxygenases in the animal intestine. The significant similarity of VP 14 to RPE65 of the human retinal pigment epithelium is intriguing in the light of recent evidence that REP65 may catalyze an isomerization of all trans retinol to U-cis retinol (Micheal Redmond, NIH, unpublished data). RPE65 was cloned and found to be expressed specifically in the retinal pigment epithelium cells (Hamel et al. 1993). VP 14 does not have any detectable isomerase acti\aty as indicated by testing all-trans substrates. However, some of the VP 14 homologous genes may encode the isomerase as a large number of vpl4 related genes in plants are detected by DNA hybridization and database searches. (Fig. 2-4, chapter 2). It has been suggested that 9-cis isomerase, if exists, should also be stress inducible as sjmthesis of 9-cis violaxanthin and neoxanthin is promoted upon stress (Li and Walton

PAGE 87

81 1990). Further study of those homologous sequences may possibly resolve these questions.

PAGE 88

Chapter 4 Localization of VP 14, a 9-cis Violaxanthin/9-c/5 Neoxanthin Dioxygenase Involved in ABA Biosynthesis Introduction Abscisic acid (ABA) is involved in the regulation of various plant processes including seed maturation, dormancy and plant responses to a variety of stress conditions (Ingram and Bartels 1996, McCarty 1995, Pena-Cortes and Wilhnitzer 1995, Rock and Quatrano 1995). ABA is an apo-carotenoid derived from oxidative cleavage of 9-cis epoxy carotenoids, a step that is widely believed to be the key regulated step in ABA synthesis (Creelman et al. 1992, Zeevaart and Creelman 1988), whereas the enzymes required for conversion of xanthoxin to ABA appear to be constitutively active in most plant tissues (Sindhu and WaUon 1987). Thus, xanthoxin does not accumulate to significant levels in normal or water stressed plant tissues. In maize, the cleavage dioxygenase was cloned by transposon tagging from an ABA deficient viviparous 14 mutant that is blocked in that step (Tan et al. 1997, Refer to Chapter 2). VP14 recombinant protein expressed in E. coli was shown to cleave specifically 9-cis violaxanthin and 9-cis neoxanthin to produce xanthoxin, an immediate precursor for ABA synthesis (Schwartz et al. 1997). ABA biosynthetic enzymes that utilize carotenoids as substrates are believed to be localized in plastids since carotenoids exist exclusively in plastid membranes. In tobacco, the epoxidase which converts zeaxanthin to all fraw^-violaxanthin was found targeted to chloroplasts (Marin et al. 82

PAGE 89

83 1996). It appears that the cleavage product, xanthoxin, can be converted to ABA in the cytosolic extraction, indicating the subsequent enzymes may be cytosolically localized (Sindhu and Walton 1988, Walton and Li 1995). The substrates for VP14, 9-cis violaxanthin and 9-cis neoxanthin are abundant in both thylakoid membranes and envelope membranes of plastids and are not water soluble (Rock and Zeevaart 1991, Li and Walton 1990b). The localization of VP 14 protein is of great importance in terms of understanding regulation of the cleavage activity and substrate accessibility following stress induction (Tan et al. 1997). Using an plastid protein import system, we show that VP 14 is targeted to chloroplasts and predominantly exists in the soluble stroma fraction. A small fraction of VP 14 was also found associated with thylakoid. Proteolysis is an essential process for many aspects of plant physiology and development. These include housing keeping functions such as removing misfolded proteins, generating peptide hormones, processing of organellar and secreted proteins through specific cleavage, and programmed cell death (Vierstra 1996). Protein turnover may also play an important role in both development and environmental regulation of metabolism by reducing the abundance of key enzymes and regulatory proteins. Because VP 14 catalyzes a highly regulated step in ABA biosynthetic pathway, and is responsible for the rapid and transit increase of ABA synthesis within hours in water stressed tissues, post transcriptional regulation involving proteolysis is likely a means of regulatory mechanism. Application of cytosolic protein synthesis inhibitor cycloheximide to unstressed bean leaves caused a 50% decrease of ABA within 4 hours after treatment, while treatment of stressed leaves with cycloheximide caused a faster decrease in ABA levels (Li and Walton 1990a). Based on these data, it was predicted that the key steps

PAGE 90

84 controlling ABA synthesis are likely to have a half life of several hours. The enzymatic cleavage of 9-cjs xanthophylls is a good candidate for the regulated step since the downstream enzyme activities are not a£fected by stress treatment (Zeevaart and Creelman 1988). To address this possibility, the post-translational regulation of VP 14 protein was studied in chloroplast. It was found that VP 14 is quickly turned over with a half life of less than 30 minutes after import into chloroplast. Materials and Methods Plasmid Constructs Higher rates of in vitro transcription and translation eflBciency have been achieved by shortening the space between the start code and SP6 promoter (Tranel et al. 1995). To remove the 5' untranslated sequence of Vp]4, the cDNA was amplified fi^om the first ATG with a engineered BamHI site to facilitate the cloning (forward primer S'-ATGCGGATCCATCCAGGGTCTCGCCCCG, reverse primer T7). The resuhing product ligated into pGem-3Z (Promega) under control of the SP6 promoter. The plasmid was named pSP6-VpJ4. In Vitro Transcription and Translation pSP6-Vpl4 DNA was linearized by digestion in the 3' linker region with EcoRI. The linear DNA was purified fi"om an agarose electrophoresis gel by electroelution. The in vitro transcription reaction was carried out as described by Cline et al. (1993).The transcription reaction contained 5 mM DTT, 50 units/ml RNasin (Promega), 0.5 mM NTP's, 50 ng/ml BSA, 1.5mM diguanosine triphosphate [m7G(5')ppp(5')G, Pharmacia]. 50 ^ig linearized plasmid DNA/ml, 500 units/ml SP6 DNA polymerase in 1 x SP6 DNA polymerase buflFer (Promega). The reaction was incubated at 40 C for 60 minutes. The messenger RNA was precipitated

PAGE 91

85 with ethanol after removing the proteins by phenol/chloroform extraction. The messenger RNA was quantitated and tested for in vitro translation efficiency in a wheat germ cell free system (Promega). A typical 50 ^il translation reaction included 30 ^1 of premade wheat germ cell-free extract, 5 nl premix of all amino acids except leucine, 5 ^il SxBufFer, 5 jil in vitro synthesized RNA (~ 40 ng), 5 |il RNase-free water containing ^H-leucine (3000 Ci/mol, DuPont). The mixture was incubated at 25 C for 1 hour and placed on ice to stop translation. Immediately before the import assay, translation product was diluted Avith equal volume of 60 mM leucine in 2x import buffer (IB, 50 mM hepes/KOH pH 8 and 0.33 M sorbitol). Isolation of Pea Chloroplasts Leaves harvested from dwarf peas grown in vermiculite for 10 days at 20 C, 150 fluorescent light were homogenized in a buffer containing 50 mM Hepes/KOH (pH 7.5), 0.33M sorbitol, 1 mM MgCb, 1 mM MnCl2, 2 mM EDTA, 5 mM sodium ascorbate and 0.2% BSA. Homogenate was filtered through one layer of miracloth and centrifuged at 2000 g (3200 rpm) for 3 minutes in a Beckman swinging bucket rotor. The pellet was resuspended in grinding buffer (GR) and loaded onto a precentrifiiged (50,000 g, 40 minutes) 35% PercoU gradient in GR buffer supplemented with 10 mM glutathione. The gradient was centrifiiged at 2000 g for 15 minutes in a swinging bucket rotor and a lower band which contained the intact chloroplasts was removed and diluted 3-fold with IB. The chloroplasts were pelleted at 1500 g and finally resuspended in IB at 1 mg chlorophyll/ml as determined spectrophotometrically.

PAGE 92

86 Protein Import Assay The protein import assay was carried out according to Cline et al. ( 1993) in the 1 x IB. All the import reactions were carried at 25 C under light with 5 mM Mg-ATP except where ATP and light dependent import was tested. The unimported proteins were removed by washing the chloroplasts with IB bufifer and the surface adhering protein removed by treatment with thermolysin. The chloroplasts were repurified through a 35% percol cushion, centri&gation and rinsed once with IxIB. The chloroplasts were lysed by suspension in a hypertonic solution containing 20 mM EDTA. For subfractionation of membrane and soluble compartments, chloroplasts were lysed in HKM on ice for 5 to 10 min. Lysis was monitored by using percoll cushion centrifugation to separate the unlysed chloroplasts. Lysed chloroplasts were collected by centrifugation at 4,200 rpm for 8 min. The pellet fraction contained thylakoids and was rinsed twice with IxIB before lysis for proteins. The supernatant fraction was ftirther centrifiiged at 18,000 g for 2 hours at 4 C to pellet the envelope membrane fraction, the supernatant fraction collected as the stroma fraction. All the subfractions were lysed in 20 mM EDTA and denatured in 15% SDS at 80 C for 10 minutes, then analyzed in 12% SDS-PAGE. The gel was treated with DMSO and PPO/POPOP then dried. The radioactive protein was detected fluorography (Cline et al. 1993). Results The N-Terminal Sequence of VP14 Has Features of a Chloroplast Transit Peptide Vpl4 encodes an 11, 12 9-cis epoxy carotenoid dioxygenase that cleaves 9-cis violaxanthin and 9-cis neoxanthin to produce xanthoxin, a precursor of abscisic acid (ABA) as indicated in Fig. 4-lB (Schwartz et al. 1997, Tan et al. 1997). This cleavage

PAGE 93

87 Fig. 4-1. The sequence of 104 amino acid N-terminus extension of VP14 that does not align to LSD and RPE65 (A) and the reactions catalyzed by VP14 in ABA biosynthesis (B). The putative transit peptide region was underlined with single line and a putative amphipathic a-helix region was underiined with a double line.

PAGE 94

88 reaction is the first committed step in the ABA biosynthetic pathway in plants, and substantial evidence has indicated that it is also the regulated step controlling synthesis of ABA. VP 14 is homologous to a bacterial lignostilbene dioxygenase (LSD) which catalyzes a very similar cleavage of a conjugated double bond to produce products with -CHO group (Kamoda et al. 1993a, b), and to a human eye retinal epithelium specific protein RPE65 (Hamel et al. 1993). A three way alignment of VP14 with LSD and RPE65 indicates that VP 14 has an -100 amino acid extension at the amino terminus as shown in Fig.4-1A. Blast searches using the extension sequence did not reveal any significant similarity to known proteins. Removal of 30 amino acids fi-om N-terminus does not affect the activity (data not shown), suggesting that this extension may have other fiinctions. Although no conserved sequence has been found among the chloroplast targeting transit peptides, there are some features shared by the stroma-targeting domains(STD), such as being rich in hydroxylated residues and lacking in acidic residues (Cline and Henry 1996). The sequence of the first 40-50 amino acids at N-terminus were consistent with a STD. It contains 20% percent serine, compared to 5 % for the whole protein, and is positively charged (pH 12.10) in contrast to the fiall protein which has an isoelectric point of pi 5.72. Similar features are also observed in the N-terminus of ABA2 in tobacco, an epoxidase that is upstream of VP 14 in the ABA biosynthetic pathway (Marin et al. 1996). Also noticed was a putative amphipathic a-helix that was highly conserved among a maize homolog and Arabidopsis homologs (Deng and McCarty, unpublished data). Amphipathic a-helices are potentially involved in protein-protein interactions or anchoring protein to membranes (Carr et al. 1993).

PAGE 95

Chloroplast proteins synthesized in cytoplasm can be tested in an in vitro system in which proteins are radiolabeled during in vitro translation and incubated with intact chloroplasts. Chloroplast targeted proteins will use the protein trafiGcking pathway to reach their organellar destination which is inaccessible to proteinase treatment. The specific protein can be detected by analyzing the radioactivity of proteins isolated fi-om chloroplasts after proteinase treatment. VP 14 was tested for traflBcking to chloroplasts using an import assay. To do so, full length Vpl4 cDNA was placed under the control of SP6 promoter in pGEM-3Z (Promega) plasmid, then transcribed and translated in vitro using SP6 DNA polymerase and wheat germ cell fi^ee systems, respectively. Because the fiill length Vpl4 cDNA has a 100 nucleotide 5' untranslated sequence in fi-ont of the first ATG, a poor translation efficiency was detected. In order to enhance translation eflSciency, the 5' UTR was removed by PGR amplification and by subcloning of Vpl4 coding region into the pGEM-3Z plasmid. Upon incubation of in vitro translated VP 14 with the intact pea chloroplasts, the VP14 protein was imported into a chloroplast compartment that is not accessible to the proteinase thermolysin (Fig 4-2A). In addition, upon import into chloroplasts, VP 14 was processed to a smaller molecular weight with the loss of about 4.5 kd. This is consistent with proteolytic cleavage and removal of a Nterminal peptide. The control protein, light-harvesting-complex-protein (LHCP), a well characterized chloroplast thylakoid protein that is involved in photosynthesis, is processed by losing its transit peptide to produce a mature protein of about 28 kd. As indicated in Fig.2B, similar to pLHCP, import of VP14 requires energy provided by either ATP or light which fiirther confirmed the chloroplast localization of VP 14. Thus, these data indicate that VP 14 is localized in chloroplasts.

PAGE 96

90 VP14 LHCP VP14 LHCP Chloroplasl -++-++ ATP + + + + TP TP Thermolysin .. + -.4+ + Fig. 4-2. Import of VP14 into chloroplasts and its energy dependence. ^H-leucine labeled precursor VP14 or LHCP were incubated with c hloroplasts and then treated with thermolysin in the presence or absence of ATP or light.

PAGE 97

91 VP 14, a Soluble Protein Is Localized in Stroma and Thvlakoid Membrane The next question is where inside the chloroplast VP 14 is localized. Chloroplasts have six different compartments including outer and inner envelope membranes, envelope interspace, stroma, thylakoid membrane and thylakoid lumen. Substrates of VP 14, 9-cis neoxanthin and 9-cis violaxanthin, have been reported in both thylakoid membranes and envelope membranes (Li and Walton 1987, Zeevaart and Creelman 1988, Parry and Hogan 1991). They presumably exist in the membranes through interaction with membrane associated proteins, as found in a carotenoid-protein complex (Markwell et al. 1992). The sub-organelle localization of VP 14 reflects to some extent the likely site of interaction between substrates and enzyme. To further locate VP 14, chloroplasts were subfractionated into three fractions, the soluble fraction which presumably contains the stromal proteins and interspace, the thylakoid fraction that includes all thylakoid membrane and lumen proteins, and envelope fraction which includes outer and inner envelope proteins. VP 14 was found predominantly in the soluble fraction, suggesting that it may be a stroma protein (Fig. 4-3). Moreover, there was a significant amount of VP 14 associated with thylakoid. Further treatment of the thylakoid with thermolysin digested VP 14, indicating that the protein was bound on the outer face of thylakoid membrane. LHCP, a thylakoid membrane protein was partially protected from thermolysin digestion consistent with the evidence that the protein is embedded in the membrane so that it is not accessible to an external proteinase (Fig. 4-3). Evidence of Fast Turnover of Imported VP 14 In the import assays, it was noticed that small sized radiolabeled proteins were also

PAGE 98

92 pVP14_ VP14~ VP14 LHCP TPESTTt TPESTTt pLHCP— LHCP— r. I. 1 Fig. 4-3. Localization of mature VP 14 inside the chloroplast. After incubated with ^H-pVP14, purified chloroplasts were lysed and separated into subfractions of stroma (S), thylakoids (T) and envelopes (E). An equal amount of thylakoids were treated with thermolysin (Tt).

PAGE 99

detected in the soluble fraction, which we hypothesized might be degradation fragments of VP 14. Because stress induction of ABA synthesis is transient, the presumed key regulated step, the diojQTgenase may be subject to rapid regulation at both the transcription level and protein level. We have detected using both monoclonal and polyclonal antibodies that VP 14 protein exists at very low levels in even stressed leaves of maize in which VpI4 transcripts were found increased significantly and ABA levels were increased about 40fold (Data not shown). One possibility is that VP 14 may undergo fast turnover such that the steady state level is low. Using the import assay, we measured the rate of turnover of mature VP 14 in isolated chloroplasts compared to a control thylakoid protein, LHCP. After the import, precursors were removed by thermolysin which was subsequently removed by repurifying the chloroplasts. The chloroplasts were incubated at 28 C under light for different periods (Materials and Methods), the level of the imported mature VP 14 was determined. As shown in Fig. 4-4 A, level of imported VP 14 was quickly declined within 60 minutes, while no significant decrease in LHCP was observed over the same period time (data not shown). It has been reported that mis-targeted and misfolded proteins are likely to be degraded in chloroplasts (Halperin and Adam 1996, Vierstra 1996). As VP 14 was found both in the soluble and thylakoid fractions, we examined the degradation pattern in both fractions. The degradation of VP 14 was about the same in either the soluble fraction and the thylakoid fraction (Fig. 4-4B). Because VP14 requires a non-heme iron in its active form (chapter 3), we considered the possibility that iron may be needed to stabilize the protein. To test this, ferrous iron and ferric iron were supplied to the chloroplasts before

PAGE 100

94 Time after import (min) TP 0 20 40 60 B Envelope Stroma Thylakoid 0 20 40 60 0 20 40 60 0 20 40 60 min Fig. 4-4. VP 14 is rapidly degraded after import imo the pea chloroplasts. VP14 was translated //I v/Yro in presence of ^H-leucine and imported into chloroplasts. Then the precursor was washed off and the chloroplasts were tested for time-course of transport. A. whole chloroplasts; B. fractionated chloroplasts; TP=translation precursor.

PAGE 101

95 import of VP14. VP14 was still subjected to degradation after addition of iron (Fe"^, Fe*'), although it seemed that Fe*^ may stabilize VP 14 to some extent. Discussion VP 14 Is Localized in Both Stroma and Thylakoid Membrane i> The N-terminal sequence of VP 14 showed certain features of a stromal targeting domain (STD) such as a high percentage of hydroxylated amino acids and positively charged amino acids as reviewed by Cline and Henry (1996). In vitro synthesized VP 14 was imported into isolated pea chloroplasts and localized in the soluble stromal fraction and thylakoid membrane fraction (Fig. 4-3). The soluble fraction was considered to contain proteins mainly from the stroma, but proteins that exist between two envelope membranes could potentially be released in this fraction during chloroplast lysis. In fact, there was a faint band detected in envelope fraction (Fig. 4-3), which was likely due to the contamination by thylakoids, because a faint band was also observed in envelope fraction in the control of stroma localized small subunit of LHPC. Experimentally it is difficult to isolate envelopes absolutely free of thylakoid contamination. A small amount of thylakoids lysed during experiment produces membrane vesicles that are inseparable from envelopes. However, the abundant presence of VP 14 in the soluble fraction but not with envelopes favors a stromal localization of VP14. The sequence of VP 14 has a putative amphipathic a-helix which is also conserved in its homologs in Arabidopsis (Deng, McCarty, Unpublished data). Reports have indicated that such kinds of protein structure have been implicated in protein-protein interaction or anchoring proteins onto membranes (Carr 1991). The substrates of VP 14,

PAGE 102

96 9-cis neoxanthin and 9-cis violaxanthin are found chiefly in envelopes and thylakoids membranes. Thus anchoring of the soluble VP 14 protein could be essentially to make it accessible to either one or both substrates. Biochemical labeling studies using '"'C-MVA indicate that separate pools of carotenoids are used for ABA biosynthesis (Parry and Horgan 1991). VP14 was found in association with thylakoid membrane with a topology of facing stroma, as indicated by its accessibility by thermolysin (Fig 4-3B). Based on this indirect evidence, we propose that the localization of VP 14 in thylakoid membranes is possibly regulated by environmental and developmental factors. The amphipathic a-helical domain could anchor the protein to the membrane and provide to access the membrane localized substrates. With immunolocalization of VP 14 by monoclonal antibody, this idea will be investigated in later studies. Post-translational Regulation Is One of the Mechanisms for its Regulation The import assay of VP 14 revealed rapid turnover of mature VP 14 inside the chloroplasts. We identified the fast turnover as a possible regulatory mechanism in plants. As discussed previously, the dioxygenase encoded by Vpl4 is considered to the regulated step in ABA biosynthetic pathway based several lines of evidence (refer to chapter 3). The relative abundance of its substrates and the constitutively active state of the downstream enzymes that convert xanthoxin to ABA make the oxidative cleavage reaction the regulatory step in the pathway. The ways in which environmental and developmental factors regulate VP14 was not known. Our preliminary data indicated that the VP14 level was extremely low in even stressed maize leaves, even though Vpl4 steady state transcript levels were increased dramatically by water stress. Because ABA levels increase by up to

PAGE 103

97 40-fold in stressed leaves (Wright and Hiron 1969, Chandler and Robertson 1994), and rewatering can quickly restore the synthesis to normal (Ingram and Bartels 1996), the de novo synthesized VP 14 is likely to be degraded or inactivated quickly such that it does not lead to ABA production. Studies which show inhibition of ABA biosynthesis by cycloheximide indicate that a key regulated enzyme in ABA biosynthetic pathway could have a half life as short as 1 hour (Li and Walton 1990a). Consistent with that, we found that VP14 was quickly degraded inside the chloroplasts with a half life about 30 minute in the isolated chloroplasts, and this fast turn-over rate was not likely a resuh of either mistargeted or abnormally folded protein. Addition of ferrous iron can result in a correct folding of VP 14, thus activate the cleavage activity (Chapter 3). However, Fe3+ did not prevent degradation of VP 14. The fast turnover is also unlikely due to a heterologous environment for VP 14, a maize protein in pea chloroplast, as a number of such cases have been reported that do not turnover rapidly (Cline and Henry 1996). Thus, we suggest that fast turn over of VP 14 is a possible regulatory mechanism controlling ABA biosynthesis is plants.

PAGE 104

CHAPTER 5 A FUNCTIONAL DUPLICATE OF VIVIPAR0US14 CONFERS A DISTINCTIVE TISSUE-SPECIFIC EXPRESSION PATTERN IN MAIZE Introduction Abscisic acid (ABA) is synthesized by oxidative cleavage of 9-cis xanthophylls to form xanthoxin, which is subsequently converted to ABA. The vpl4 mutant selectively blocks ABA synthesis in the developing seed and suggests that the Vpl4 gene has a role in developmental control of hormone synthesis in maize. Vpl4 encodes the dioxygenase enzyme responsible for oxidative cleavage of 9-cis xanthophylls to xanthoxin (Schwartz et al. 1997) which is believed to be the key regulated step in the ABA biosynthetic pathway (Zeevaart and Creebnan 1988, Walton and Li 1995). Biochemical analysis of the mutants indicated that there is about a 70% reduction of ABA in developing embryos, whereas no reduction of ABA levels was found in other tissues including leaves and roots (Tan, et al. 1997). Two possible explanations may account for the residual ABA in vpl4 mutants: 1) The existing vpl4 mutant alleles are leaky, 2) Functionally redundant VpJ4-like genes exist. A leaky expression of a functional VP14 protein from the vpJ4-2274 allele would likely require splicing of the Mu J insertion from the transcript. We detect little evidence of wild type size transcripts in mutant embryos on northern blots (Tan et al. 1997, refer to chapter 2). Furthermore, it is hard to explain that the vp 14-2274 mutation can specifically affect ABA biosynthesis in embryos and retain normal function in other tissues if vpl 42274 is leaky. Several lines of indirect evidence favor the second possibility. First, the 98

PAGE 105

99 observation that ypl4 does not markedly affect ABA levels in vegetative tissues and that mutant plants develop normally under field conditions implies strongly that the Vpl4 gene accounts for only a subset of the ABA biosynthetic activity in the plant. Thus, partial suppression of the \>pl4 phenotype by ABA transferred from maternal tissues might explain the variable penetrance of the viviparous phenotype we observe under field conditions. Ober and Setter (1992) reported that maternal ABA synthesis induced by stress can significantly affect seed development in maize. Secondly, detached vpl4 leaves accumulate ABA nearly as much as the wild type, suggesting the expression of a fimctionally active dioxygenase in stressed leaves. Thirdly, we detected evidence of related genes on moderate stringency Southern and northern blots. In addition, several related sequences are found in the rice, maize and Arabidopsis expressed sequence tag (EST) collections. Thus, the evidence points to overiapping expression of fimctionally equivalent ^7^-like genes with different patterns of tissue specific expression. Here we reported the cloning and sequencing of a highly homologous Vpl4 duplicate gene (Ecdl) that was mapped to short arm of chromosome 5. The distinctive expression pattern of Ecdl relative to Vpl4 provided evidence for a potentially important mechanism in regulating hormone biosynthesis in higher plants. Materials and Methods The plant genotypes used and growth conditions were described as in Chapter 2. Southern Blot and North ern Blot Analv.sis As described in Chapter 2. Construction a nd S cree n i ng of Clone T.ihrarips For cloning of the 3.2 kb EcoRI fi-agment identified by hybridization to the Vpl4 conserved probe (P2), about 100 \xg

PAGE 106

100 genomic DNA was digested with EcoRI to completion, then size fractionated on a 1040% sucrose in TE (pH 8.0) gradient. The gradient was centrifuged at 85,000 g in a Beckman SW41 rotor at 4 C for 24 hours. DNA in each fraction was precipitated in 70% ethanol and the fractions tested by Southern blot to locate the fraction containing the 3.2 kb fragment. DNA in this fraction was ligated to X,-ZAP (Stratagene) arms and packaged into lambda phage. The library was screened with the Vpl4 P2 probe (refer to Fig. 2-4 in Chapter 2). The resulted lambda phage was in vivo excised by a helper phage to yield a Bluescript plasmid containing 3.2 kb EcoRI fragment, named p£cJ7. Water Stress Treatment Leaves of well watered maize leaves of about 2 week of age were detached and put on the lab bench at room temperature until they lost 15% of their fresh weight to transpiration. Then they were sealed in plastic bags and kept at room temperature for 6 hours. Leaves then were frozen in liquid nitrogen and stored at -80 C until use. Results Molecular and biochemical analysis of vpl4 mutants indicate that vpl4 is specifically blocked in the epoxy-carotenoid cleavage in the ABA biosynthesis in seeds (Tan et al. 1997). The expressed recombinant VP14 protein was shown to specifically cleave the 11,12 double bond of 9'-c/j neoxanthin and 9-cis violaxanthin to generate xanthoxin as indicated in Fig. 5-1 A (Schwartz et al. 1997). Southern blot analysis using the conserved sequence of Vpl4 (P2, refer to Chapter 2) revealed multiple EcoRI fragments in the inbred W22 and Q69 genomic DNAs, suggesting that Vpl4 belongs to a gene family (Fig. 5IB). In support for this, several Vp]4-\ikc sequences have been

PAGE 107

101 B W22 Q68 9-cis violaxandiin \ vpl4 \ • 9-cis neoxanthin vpl4 Xanthoxin I I ABA 4i 3.2 kb(SS) Vpl4 duplicate Ikb s 2 Ecdl I I I I I II I I I I L — (OX 111 Coding region Homology with Vpl4 80% cDNA 92% 74% Fig. 5-1. Cloning of a Vpl4 duplicate, Ecdl in maize. A. The blocked step of ABA biosyntiiesis in vp7-^ mutant. B. A low stringency Southern analysis of W22 and Q68 DNA witfi a probe (P2) of a highly conserved region of Vpl4. C. Restriction map of the cloned 3.2 kb EcoRI fragment and the open reading frame. The homologous region with Vpl4 cDNA was marked.

PAGE 108

recovered from genomic and cDNA libraries (Data not shown). A 3.2 kb EcoRI fragment in Q69, which was polymorphic in inbred W22, showed an almost equal hybridization intensity as VpJ4, suggested that it may contain a closely P^7'^-related sequence. To search genes that potentially encode proteins functionally equivalent to VP 14, this fragment was cloned from a subgenomic library constructed from enriched 3.2 kb EcoRI fragments (See Materials and Methods). Sequencing of this fragment revealed a strong homology to VpJ4 cDNA (Fig. 5-lC). This clone, designated Ecdl for potentially encoding an epoxy-carotenoid dioxygenase like VpJ4, suggested an identical gene structure as compared to VpI4. The DNA alignment with VpJ4 cDNA indicated that the coding region is not interrupted by any introns, a feature of VpJ4 but quite rare in other eukaryotic genes. The position of the open reading frame and a putative TATA box were highly conserved between Vpl4 and EcdJ (Fig. 5-2). The coding sequence had a 92% identity to VpJ4 at nucleotide level, indicating that EcdJ is a recent duplication of Vpl4. The lower homology to Vpl4 at the 5' and 3' flanking sequences hinted at a diversification of expression during evolution. This clone was mapped by the Missouri mapping center to a location between RFLP markers umc83b and csu222b on the short arm of chromosome 5. There is accumulating RFLP evidence that a segment of the maize chromosome IL which include the Vpl4 gene, was duplicated during the evolution of maize (Helentjaris et al. 1988 ). Molecular clock analysis based on synomynous nucleotide substitutions in the Vpl4 and Ecdl coding regions indicates that the duplication occurred 12.60.5 million years ago (J. Doebley, unpublished). This is in line with evidence of other duplications in maize such as

PAGE 109

103 GAATTCATTCTTGGACTC GCGGCCCGTTCAAACCGCCCACCACTCTCCTCTCCTCCACCCGCQCCGCCCGTCGCCCCGTCATTTCTTTTTCTCCCACCCCCCAGTC^ T(Xa:TGCTCTGTCXGCGACTCCGCCrTCACTCBJUttlLCAACCGCCGGCTCCTATCCC^ caacagtggagagtccatcctccgccgocgcccaggcaataacccactcgatcacgcagaaacccccgcacgcgatcccacgccaattcagccaccgccg atgcagggtctcgccccgcccacctctgtcrccatacaccggcaoctgccggccgggtccagggcccgggcctccaactccgtcaggttctcgccgcgcg mqglapptsvsihrhlpagsrarasnsvrfspra ccgtccgctccgtgccgcacgagtgccgccaggcgccgttccacgccgacctgccggcgccgtccaagaagcccaccgccattgccgtcccgaggcacgc vrsvphecrqapfhadlpapskkptaiavprha cgcggcgccgcgcaagtctggcggcggcggcggcaagaagcagctcaacctattccagcgcgcggcggcggccgcgctcgacgcgttcgaggaggggttc aaprksgggggkkqlnlfqraaaaaldafeegf gtggacaac(jiw:tcgagcggccccacgggctgcccagcacggccgacccagccgtgcagatcgccggcaacttcgcgcccgtcggggagaggcc(x:c^ vdnvlerphglpstadpavqiagnfapvgerppv tgcgcggtctccccgtctccgggcgtatcccgcccttcatcagcggcgtctacgcgcgcaacggcgtcaacccctgcttcgaccccgtcggcgggcacca rglpvsgrippfisgvyarngvkpcfdpvgghh cxttcttcgacggcgacggcatggtgcacgcgctgcggatacgcgacggtgtcgacgagtcctacgcctgccgcttcaccgagaccgcgcgcctgacccag lfdgdgmvhalrirdgvdesyacrftetarltq gagcgcgcggtcggccgccccgtcttccccaaggccatcggcgagctgcacggccactccgggatcgcgcgcctcgcx;ctgttctacgcgcgcgccgcgt eravgrpvfpkaigelhghsgiarlalfyaraac gcggcctcgtcgacccctcggccggcaccggcgtcgccaacgccggactcgtctacttcaacggccgcctcctcgccatgtccgaggacgacctcccgta glvdpsagtgvanaglvyfngrllamseddlpy ccacgtccgcgtcgcggacgacggcgaoctcgagaccgtcggccgctacgacttcgacggccagctcggctgccccatgatcgcgcaccccaagctggac hvrvaddgdletvgrydfdgqlgcpmiahpkld ccggccaccggggagctgcacgcgctcagctacgaggtcgtcaggaggccctacctcaagtacttctacttcaggcccgacggcaccaagtccgacgacg patgelhalsyevvrrpylkyfyfrpdgtksddv tggagatcccgctggcccagcccaccatgatccacgacttcgccatcaccgagaacctggtcgtggtgcccgaccaccaggtggtgttcaagctgcagga eiplaqptmihdfaitenlvvvpdhqvvfklqe gatgctgcgcggcgggtcgcccgtggtgctggacagggagaagacgtcgcgcttcggcgtcctcccgaagcgcgccgcggacgcgtcggagatggcgtgg mlrggspvvldrektsrfgvlpkraadasemaw gtggacgtgccggactgcttctgcttccacctgtggaacgcgtgggaggacgaggcgacgggcgaggtggtggtgatcggctcctgcatgacccccgccg vdvpdcfcfhlwnawedeatgevvvigscmtpad actccatcttcaacgagtcggacgtgcggctcgagagcgtgctgacggagatccggctggacgcgcgcacgggccggtcgacgcgccgcgccgtcctgcc sifnesdvrlesvlteirldartgrstrravlp gccgtcgcggcaggtgaacctggaggtgggcatggtgaaccgcaacctcctggggcgcaggacgcggtacgcgtacctcgcggtggccgagccgtggccc psrqvnlevgmvnrnllgrrtryaylavaepwp aaggtctcgggcttcgccaaggtggacctggccacgggcgagctcgccaggttcgagtacggcgagggccggttcggcggcgagccctgcttcgtgccca kvsgfakvdlatgelarfeygegrfggepcfvpm tggactccgccgcggcccacccgcgcggcgaggacgacgggcacgtgctcgccttcgtccacgacgagcgcgccggcacgtccgagctcctggtggtcaa dpaaahprgeddghvlafvhderagtsellvvn tgccgccgacatgcggctggaggcx:accgtccggctcccgtcccgcgtgcccttcggcttccacggcaccttcatcacgggcgcggagctcgaggcccag aadmrleatvrlpsrvpfgfhgtfitgaeleaq gcctgagcggagctcgaggcccaggcctgagtgactcagctccacctttcttggaggaggaggaacagaggagccatggatcagggggaggagtcgccag A agggagcctagatcagttccccggggtcttcccgtcxccatctcaccacagtctttacagttgcttttgtttttcttttcatttcagttcacactagtgt aagtaaaattgggatagtggcagcttagagagagattattagtagcagtagggagagagagagagagaggagaaaggcgcccagctcgtagcttttcagc tgctgcttgcttctagagatcgagcagccagagctcagctagtggtggctcttgctagtattccccctccttccttttcctttgatgacatggatatgca tccatctccagctggggtgtctaggatcggttgctgctacttgatcgccattgccacccagttgctgctgctgctgctgctgaggtgttgtgtgtgtaca tttgttcattaatataataataatatgatgattgtataagaattaagaacggtgacggtttctgttgggagatcgaatttcgtttgtctgtgtgtcgtga tgccggccgggacaagcagagccttcgctgtcgacgggcacacagcagcattgcgttgctggtgtggtgcgtgccagacgcagacgggaatctgcttcct ccacagcaaccgggcgatcctctgctttttgtccacttgaccacttgtgcgcggggggcgcgctttggttctttgctgccaaccaatgtactccgatgtc gtaaggaccgcgtgtgtttggacgctgctgctgctggtggtggtgtcgatttatgcccaacaacaagtggacgacggggctatgtgctaacgcacggggg tgaaccggggatttgtcgtgtctgcgatgcggttcgtcgctgtagatgatgaggtcatgagggcttccaacttccaagatgctgtccagcgtcgacgggc attgcgatcgtgtggagcaagcttcagagagaacttgattccgagtcgggcggctcgctgctgggcaactcttcttctgttcgcccaactcaactggatc ttgaattc

PAGE 110

104 CI and PI (Cone et al. 1993). The open reading frame of Ecdl encodes a 601 amino acid protein with a calculated molecular mass of 65.5 kd, very close to VP 14 (Fig. 5-2). The alignment of the deduced protein sequence of ECDl with VP 14 provided further evidence that ECDl may function as an VP14-like epoxy-carotenoid dioxygenase. The two proteins share an identity of 93% of amino acid sequences, and a similarity of 99% (Fig. 5-3). Gaps in Ecdl only include a triple amino acid (KPV) deletion close to the possible cleavage site of the transit peptide (refer to Chapter 4), and one missing alanine residue at the C-terminus of the protein. In a multiple sequence alignment with closely related proteins found in the database, the conserved histidine and tyrosine residues including the surrounding clusters of sequences present in the VP 14, LSD, REP65, SYNl and SYN2 proteins (Chapter 2) are also conserved in ECDl. Those residues are suggested to chelate the co-factor iron which is proven to be critical to the activity of VP 14 (Chapter 3). Features of a putative transit peptide, a high percentage of hydroxylated amino acids (serine is 20% in the first 50 amino acids) and positively charged residues are also conserved. In addition, a putative amphipathic a-helix appears to be highly conserved between ECDl and VP14 (boxed region in Fig. 5-3). These features strongly support the idea that Ecdl is a Vpl4 duplicate. The high degree of identity between ECDl and VP 14 proteins strongly indicated the ECDl may have the same fiinction as VP14, i.e. cleavage of 9-cis neoxanthin and 9cis violaxanthin, provided that Ecdl is expressed in plants. Furthermore, if at least two functionally identical enzymes (isoenzymes) exist in maize, it would be of great interest to know how they are regulated or coordinated throughout plant developmental processes. The expression pattern of Ecdl was studied by northern blot analysis and compared to

PAGE 111

105 VP14 MQGLAPPTSVSIHRHLPARSRARASNSVRFSPRAVSSVPPAECLQAPFHKPVADLPAPSRK 61 53 **************** *(;* ************* *i^* *H* *R* **** ****** j^* 57 VP14 PAAIAVPGHAAAPRK--AEGGKKQIJILFQR*AAAALDMEEGFV7\NVLERPHGLPSTADPA 120 gg *j*****j^*******3Q(^** ****************** *****p************ **** ^]^g VP14 VQIAGNFAPVGERPPVHELPVSGRIPPFIDGVYARNGANPCFDPVAGHHLFDGDGMVHALR 181 53 ****************J^Q***********3* ******Y*******Q****** *** ** ** ^ig VP14 IRNGAAESYACRFTETARLRQERAIGRPVFPKAIGELHGHSGIARLALFYARAACGLVDPS 242 53 **Q*YQ*************>ji****y** ************************* ********* 240 VP14 AGTGVANAGLVYFNGRLLAMSEDDLPYHVRVADDGDLETVGRYDFDGQLGCAMIAHPKIDP 303 JjJ ************************************************* **p********* VP14 ATGELHAISYDVIKRPYLKYFYFRPDGTKSDDVEIPLEQPTMIHDFAITENLVWPDHQW 364 53 **********£ *Yj^**** ************* ******^******** ******* ******** A VP14 FKLQEMLRGGSPWLDKEKTSRFGVLPKHAADASEMAWVDVPDCFCFHLWNAWEDEATGEV 425 53 ****************|^***********{^** ******************* *********** VP14 WIGSCMTPADSIFNESDERLESVITEIRLDARTGRSTRRAVLPPSQQVNLEVGMVNRNLL 486 3g ******************Y***************************I^************** VP14 GRETRYAYLAVAEPWPKVSGFAKVDLSTGELTKFEYGEGRFGGEPCFVPMDPAAAHPRGED 547 5g *R* ********************* *^* AR* *************************** 54 5 VP14 DGYVLTFVHDERAGTSELLWNTADMRLEATVQLPSRVPFGFHGTFITGQELEAQAA 604 53 *[{* A* ************** *A* ******* *R* ************** *****— gQi Fig. 5-3. Alignment of the VP 14 duplicate with VP 14. indicates identical; gap, ^ conserved histidine residues in LSD, REP65, SYNl and SYN2. Underlined region refers to a putative transit peptide. Boxed region refers to a putative amphipathic a-helix.

PAGE 112

106 VpJ4 in both water stressed and nonstressed leaves. Because VpJ4 and EcdJ were highly homologous in the coding region, a less conserved 3' untranslated sequence was used as gene specific probe for EcdJ. To avoid ambiguity in the event that VpI4 and Ecdl may have roughly equal sized mRNA's and cross hybridization, vpJ4 homozygous mutants were used as a positive control for the VpJ4 transcript since &MuJ insertion in the coding region resulted in a 4.0 kb transcript (Fig. 2-8, Chapter 2). As indicated in northern blot analysis in Fig. 5-4, there were four bands detected in the water stressed leaves of vpJ4 mutant by a VpJ4 coding sequence probe, P2 (the probe was illustrated in Fig. 2-4C, chapter II). The 4.0kb marks the vpl4 message containing a 1.4 kb MuJ insertion. Vpl4 messages were induced in water stressed leaves, whereas little was detected in the unstressed leaves. In addition, two bands smaller than the VpI4 transcript (2.6 kb) were evidently detected by this probe, indicating the expression of related genes. It is important to keep in mind that VpJ4 and Ecdl have a 95% identity in nucleotide sequence in the P2 region. When the same blot was hybridized with the 3' untranslated sequence of the EcdJ genomic DNA which has a 74% identity to VpJ4 as indicated in Fig. 5-lC, a little cross hybridization to the 4.0 kb band in vpJ4-2274 and vpJ4-3250 was expected. However, a strong hybridization to one of the smaller bands (-1.5 kb) was detected. The intensity strongly suggested that it is a transcript fi-om the EcdJ gene which may be spliced or transcribed fi-om an alternative start site. The sequence of this smaller size transcript is currently under investigation. The weak hybridization to P2 probe suggested that this transcript may not have an intact P2 corresponding region as indicated in the genomic ORF. Furthermore, this smaller transcript appeared to be expressed constitutively in leaves

PAGE 113

107 Probe CM CM Si 3 NS ST NS ST NS ST NS ST Vp14 (P2) ::;xi^:: 3^ <> 4.0 kb ^ 2.6 kb 1.5 kb Fig. 5-4. Expression pattern of Vpl4 and EcdJ'm water stressed (ST) and non-stressed (NS) leaves of inbred W22, Vpl4 and two vpJ4 alleles. The same blot was hybridized with three different probes. Each lane was loaded with roughly 500 ng poly(A)-RNA.

PAGE 114

108 VpJ4 (P2) Ecdl specific 1^ wt wt vp J 4-2274 •g Root leaf leaf (4 ns St ns St ns st 4.0 kb 2.6 kb 1.5 ld> 4.0 kb 2.6 kb l.Skb Fig. 5-5. Northern blot analysis of EcdJ expression in embryos and stressed roots, st, water stressed; ns, nonstressed.

PAGE 115

109 and not affected by water stress. The reason for the lack of this smaller message in stressed homozygous vp 14-227 4 mutants was not known. Similarly, the expression of Ecd J was explored in developing embryos and roots. Because Vpl4 was found to be expressed in both tissues (Tan et al. 1997, Chapter 2), the stressed mutant vp 14-227 4 was used as a control for Vpl4 cross hybridization when Ecdl specific probe was used. As shown in Fig. 5-5, under a high stringency, the P2 probe hybridized to Vpl4 transcripts shown in embryos and nonstressed roots as 2.6 kb of wild type size, and in water stressed vpl4-2274 leaves as 4.0 kb of mutated form as shown previously (Tan et al. 1997). Surprisingly, P2 did not detect any signal in the water stressed roots, suggesting that at least Vpl4 expression is down-regulated in stressed roots. This is markedly different fi^om Vpl4 expression in leaves where it is strongly water stress inducible as also indicated in this blot and previously (Tan et al. 1997). As mentioned above, the P2 probe may also recognize Ecdl based on 95% identity in the Ecdl genomic sequence provided that the Ecdl transcript is the same size as Vpl4. When the same blot was probed by the Ecdl specific probe, three bands of different sizes were detected (Fig. 5-5). A 2.6 kb transcript was detected present in the nonstressed roots, and a transcript between the 2.6 and 1.5 kb in stressed roots. The barely visible of the 2.6 kb signal in the embryo and faint signal of the 4.0 kb transcript which seems to be overloaded, argue that the 2.6 kb transcript present in nonstressed roots is Ecdl. Discussion The biochemical analysis of mutant vpl4 and molecular cloning of Vpl4 gene led to a suggestion that multiple functionally identical but differentially expressed Vpl4 related

PAGE 116

110 Fig. 5-6. Maps of chromosomes IL and 5S showing locations of duplicated markers Only markers that are duplicated on both chromosomes are indicated. The order of the markers is according to MMN (1996), while the distance between markers was not precisely indicated. Centromeres were marked with circles. Arrows indicate orientations of the chromosomes.

PAGE 117

Ill genes may control the biosynthesis of ABA throughout plant development. Different patterns of tissue specific expression have been found in a number of gene families. Differentially regulated gene families are emerging as an important mechanism underlying developmental control of biochemical pathways in plants. The ACC synthase genes that encode the key regulated step in ethylene hormone biosynthesis also form differentially regulated gene family in plants (Liang et al. 1989). It remains to be seen whether key steps in other hormone synthetic pathways will have this feature. The duplication feature of VpJ4 in chromosome 5S may be specific to maize. There is a large body of RFLP evidence of a duplicated chromosome segment on IL and 5S of maize that is presumably a residue of a tetraploid ancestry (Helentjaris et al. 1988). In an analysis of existing RFLP markers neighboring VpJ4 and its duplicate Ecdl on chromosome IL and 5S respectively (Maize Genetics Cooperation Newsletter 70 1996), we found clusters of conserved RFLP sequences between the two (Fig. 5-6). The order of the markers suggested that duplication is inverted between IL and 5S. The highly conserved gene structure and sequence between VpJ4 at IL and EcdJ at 5S provided additional molecular evidence for such a duplication. The 93% identity in amino acid sequence indicates that the two proteins are likely to have the same enzymatic activity, i.e. cleaving 9'-cis neoxanthin and 9-cis violaxanthin to produce xanthoxin in the ABA biosynthetic pathway. The distinctly different pattern of EcdJ expression compared to VpJ4 provides an explanation for the normal levels of ABA in tissues like roots, and perhaps the residual ABA detected in embryos because low level expression can not be ruled out by northern bolt analysis. The relatively divergent presumptive promoter region of EcdJ may substantiate the tissue specific expression. By analogy to other duplicate

PAGE 118

112 Fig. 5-7. Diagram of Ecdl and Vpl4 gene structure and possible spicing in Ecdl.

PAGE 119

113 gene pairs that have been analyzed in maize (Chandler et. al. 1989, Cone et. al. 1993), we believe that VpJ4 and Ecdl have evolved distinctive patterns of tissue specific gene expression. Instead of having a complex elements in the promoter of a key gene in a pathway, plants may choose simple divergent promoters for a family of functionally related genes to achieve the regulation and integration of developmental and environmental signals. As to how important EcdJ is to the plant remains to be determined. The absence or reduced level of EcdJ transcripts in vp 14-227 4 is unexplained. One possibility is that the tested line of vp 14-2274 is a double mutant of vpl4 and ecdl such that it tends to be more penetrant than vpl4-3250. An outcross of this specific line to inbred like W22 should resolve this question. If it is a double mutant, and then the Ecdl may also only contribute another small subset of ABA in plants like Vpl4 does. It may be possible that other Vpl4 related genes besides Ecdl might encode proteins functionally equivalent to VP 14. Vpl4 homologous cDNA's recovered from roots and another clone from embryos may be the candidates for such genes. An interesting finding is that in leaves the transcript detected by a probe of 3' untranslated region of Ecdl is smaller than Vpl4. That probe has a 60% homology to Vpl4 which is evidently reflected by the weak hybridization to Vpl4 mRNAs (4.0 and 2.6 kb. Fig. 5-4), suggesting that the strong signal is very likely the Ecdl transcript. One possibility is that an internal sequence was spliced out as an intron as shown in the diagram (Fig. 5-7). As hybridization of the conserved 1 kb P2 probe to the smaller sized Ecdl transcript was much weaker than to Vpl4, and the homology of that region is as high as 95% at nucleotide level, it appears that the smaller transcript might have a spliced

PAGE 120

114 sequence in that region, such as an intron. Based on the sequence alignment of VP 14 with Ecdl, such a spliced massage would not be translated into a functional enzyme. It is possible that the splicing of an internal sequence in the coding region of Ecdl might play a role in the regulation of the expression of that gene, and that splicing may be developmentally and environmentally regulated. However, the northern blot analysis with 3' sequence probe of Ecd J did not detect any significant increase in the 2.6 kb message, suggesting that there is not a significant increase in the unspliced version at least in dehydrated leaves. Further cloning of the cDNA and northern analysis using the splicing sequence as a probe should provide more insights to that question.

PAGE 121

SUMMARY AND CONCLUSIONS By searching for viable and green viviparous mutants from Robertson's mutator lines, we have isolated a new ABA deficient mutant, viviparousl4 (vpJ4) that specifically blocks ABA biosynthesis in seeds. It was proven to be an important mutant as biochemical analysis indicated that ^J}14 is blocked in the oxidative cleavage of 9-cis xanthophylls, the first committed step in the ABA biosynthetic pathway. Vpl4 was cloned by transposon tagging and the sequence shares strong homologies to lignostilbene dioxygenase of pseudomonas and RPE65 of mammals. VP 14 as expressed in E. coli can specifically cleave 9'-c/j neoxanthin and 9-cis violaxanthin to produce xanthoxin, an immediate precursor of ABA. VP 14 was synthesized as a precursor and processed upon import into chloroplasts. Vpl4 belongs to a gene family that potentially plays a role in the coordinated regulation of ABA biosynthesis in different tissues at different developmental stages and in response to environmental signals. A duplicate of Vpl4 was cloned and it may encode a protein functionally equivalent to VP14, which is expressed with a distinaly different in development. VP14 is subject to fast turnover (half life -30 minutes) which taken together suggested that expression of Vpl4 may be regulated transcriptionally and post-transcriptionally. 115

PAGE 122

LIST OF REFERENCES Abeles, F.B., Morgan, P.W., and Saltveit, M.E. Jr. (1992). Ethylene in Plant Biology, 2nd ed. Academic Press, New York. Adam, Z. (1996). Protein stability and degradation in chloroplasts. Plant Mol. Biol 32 773-783. Addicott, F.T. (ed.), (1983). Abscisic Acid. Praeger, New York. Armstrong, J., Leung, A, Grabov, J., Brearley, J., Giraudat, J., and Blatt, M.R. (1995). Sensitivity to abscisic acid of guard-cell K+ channels is suppressed by abil-1, a tmidixAArabidopsis gene encoding a putative protein phosphatase. Proc. Natl. Acad Sci.USA92, 9520-9524. Assante, G., Merlini, L., and Nasini, G. (1977). (+)Abscisic acid, A metabolite of the fungus Cercospora rosicola. Experientis 33, 1556-57. Bartley, G.E., and Scolnik, P A. (1995). Plant carotenoids. Pigments for photoprotection, visual attraction, and human health. Plant Cell 7, 1027-1038. Bennett, R.D., Norman, S.M., and Maier, V.P. (1981). Biosynthesis of abscisic acid from [1,2-"C2] acetate in Cercospora rosicola. Phytochemistry 20, 2343-2344. Bennett, R.D., Norman, S.M., and Maier, V.P. (1984). Biosynthesis of abscisic acid from famesol derivatives in Cercospora rosicola. Phytochem. 23, 1913-1915. Birchler, J.A. (1994). Marker systems for B-A translocations. In, Maize Handbook. Freeling, M., and Walbot, V. eds, SpringerVerlag, New York, Inc. pp 330-331. Bitoun, R., Rousselin, P. and Caboche, M. (1990). A pleitropic mutation results in crossresistance to auxin, abscisic acid and paclobutrazol. Mol. Gen. Genet. 220, 234-239. Black, M. (1991). Involvement of ABA in the physiology of developing and mature seeds. In: Abscisic Acid, Physiology and Biochemistry. Davies, D.W. and Jones, H.G. eds. Bios Scientific Publishers Limited, Oxford, UK. pp. 81-98. Blatt, M.R. (1992). K+ channels of stomatal guard cells. Characteristics of the inward rectifier and its control by pH. J. Gen. Physiol. 99, 615-624. Bradford, M. (1976). A rapid and sensitive method for the quantitation of microgram quantities ofprotein utilizing the principle of protein-dye binding. Anal Biochem 72 248-254. Carr, D.W., Stofko-Hahn, R.E., Fraser, I.D., Bishop, S.M., Acott, T.S., Brennan, R.G., and Scott, J.D. (1991). Interaction of the regulatory subunit (RH) of cAMPdependent protein kinase with RH-anchoring proteins occurs through an amphipathic helix binding motif J. Biol. Chem. 266, 14188-14192. Chandler, P.M., and Robertson, M. (1994), Gene expression regulated by abscisic acid and its relation to stress tolerance. Ann. Rev. Plant Physiol. Plant Mol Biol 45 113141. 116

PAGE 123

117 Chandler, V.L., Radicella, J.P., Robbins, T P., Chen, J., and Turks, D. (1989). Two regulatory genes of the maize anthocyanin pathway are homologous: isolation of B utilizing i? genomic sequence. Plant Cell 1, 1175-1183. Church, G.M., and Gilbert, W. (1984). Genomic sequencing. Proc. Natl. Acad. Sci. USA 81, 1991-1995. Cline, K., and Henry, R. (1996). Import and routing of nucleus-encoded chloroplast proteins. Ann. Rev. Cell. Dev. Biol. 12, 1-26. Cline, K., Henry, R., Li, C. and Yuan, J. (1993). Multiple pathways for protein transport into or across the thylakoid membrane. EMBO J. 12, 4105-41 14. Cone, K.C., Cocciolone, S.M., Burr, F.A., and Burr, B. (1993). Maize anthocyanin regulatory gene PI is a duplicate of CI that functions in the plant. Plant Cell 5, 17951805. Comforth, J.W., Milborrow, B.V., Ryback, G. and Wareing, P.P. (1965). Identity of sycamore 'dormin' with abscisin n. Nature 205, 1269-1270. Cornish, K., and Zeevaart, J.A.D. (1988). Phenotypic expression of wild-type tomato and three wilty mutants in relation to abscisic acid accumulation in roots and leaflets of reciprocal grafts. Plant Physiol. 87, 190-194. Creelman, R.A., and Zeevaart, J.A.D. (1984). Incorporation of oxygen into abscisic acid and phaseic acid from molecular oxygen. Plant Physiol. 75, 166-169. Creelman, R.A., Bell, E., and Mullet, J.E. (1992). Involvement of a lipoxygenase-like enzyme in abscisic acid biosynthesis. Plant Physiol 99, 1258-1260. Creelman, R.A., Gage, D.A., and Zeevaart, J.AD. (1987). Abscisic acid biosynthesis in leaves and roots of Xanthium strumarium. Plant Physiol. 85, 726-732. Cutler, S., Ghassemian, M., Bonetta. D., Cooney. S., and McCourt, P. (1996). A protein famesyl tansferase involved in abscisic acid signal transduction in Arabidopsis. Science 273, 1239-1241. Dellaporta, S. (1983). A plant DNA minipreparation, Version II. Plant Mol. Biol. Rep. 1,19-21. Duckham, S.C., Taylor, I.B., Lindforth, R.S.T., Al-Naieb, R.J., Marples, B.A and Bov^an, W.R. (1989).The metabolism of cw-ABA aldehyde by the wilty mutants of potato, pea Arabisopsis thaliana. J. Exp. Bot. 217, 901-905. Duckham, S C., Taylor, I.B., Linforth, R.S.T., Al-Naieb, R.J,. Marples, B.A, Bowan, W.R. (1991). Abscisic-acid-deficient mutants at the aba gene locus of Arabidopsis thaliana are impaired in the epoxidation of zeaxanthin. Plant Cell Environ. 14, 601606. Dure, L m.. Crouch, M., Harada, J., Ho, TD., and Mundy J. (1989). Common amino acid sequence domains among the LEA proteins of higher plants. Plant Mol Biol 12 475-486. Eagles, C.F., and Wareing P.P. (1964). The role of growth substances in the regulation of bud dormancy. Physiol. Plant. 17, 697-709.

PAGE 124

118 Finkelstein, R.R. (1994). Mutations at two new Arabidopsis ABA responsive loci are similar to abi3 mutations. Plant J. 5, 765-771. Finkelstein, R.R., Somerville, C.R. (1990). Three classes of abscisic acid (ABA)insensitive mutations of Arabidopsis define genes that control overlapping subsets of ABA response. Plant Physiol. 94,1172-1179. Fresnedo, O., Gomez, R., and Serra, J.L. (1991). Carotenoid composition in the cyanobacterium Phormidium laminosum. Effect of nitrogen starvation. FEBS Lett. 282,300-304. Gage, D.A., Fong, F., and Zeevaart J.A.D. (1989). Abscisic acid biosynthesis in isolated embryos of Zea mays L. Plant Physiol. 89, 1039-1041. Gaskin P., and MacMillan P. (1992). GC-MS of gibberellins and related compounds /w; Methodology and a Library of Reference Spectra. Cantock's Bristal, UK, S885. Giraudat, J. (1995). Abscisic acid signaling. Current Opin. Cell Biol. 7, 232-240. Giraudat, J., Hugee, B.M., Valon, C, Smalle, J., Parcy, F., and Goodman, H.M. (1992). Isolation of the Arabidopsis ABI3 gene by positional cloning. Plant Cell 4, 12511261. Giuliano, G., Pichersky, E., Malik, V.S., Timko, M.P., Scolnik, P.A., and Cashmore, A. R. (1988). An evolutionarily conserved protein binding sequence upstream of a plant light-regulated gene. Proc. Natl. Acad. Sci. USA 85, 7089-7093. Goday, A., Jensen, A.B., Culianez-Macia, F.A., Mar, AM., Figueras, M.J., Torrent, M. and Pages S.M. (1994). The maize abscisic acid-responsive protein Rabl7 is located in the nucleus and interacts with nuclear localization signals. Plant Cell 6, 351-360. Gomez-Cadenas, A., Tadeo, R.T., Talon, M., and Primo-millo, E. (1996). Leaf abscission induced by ethylene in water-stressed intact seedlings of Leopatra mandarin requires abscisic acid accumulation in roots. Plant Physiol. 1 12, 401-408. Guiltinan, M.J., Marcotte, W.R., and Quatrano, R.S. (1990). A plant leucine zipper protein that recognizes an abscisic acid response element. Science 250, 267-271 Halpersin, T., and Adam, Z. (1996). Degradation of mistargeted OEE33 in the chloroplast stroma. Plant Mol. Biol. 30, 925-933. Hamel, CP., Tsilou, E., Pfeffer, B.A., Hooks, J.J., Detrick B. and Redmond, T.M. (1993). Molecular cloning and expression of RPE65, a novel retinal pigment epithelium-specific microsomal protein that is post-transcriptionally regulated in vitro. J. Biol. Chem. 268, 15751-15757. Harris, M.J. and Outlaw, W.H. Jr. (1991). Rapid adjustment of guard-cell abscisic acid levels to current leaf-water status. Plant Physiol. 95, 171-173. Hartung, W. and Davies, W.J. (1991). Drought-induced changes in physiology and ABA. In: Abscisic Acid. Physiology and Biochemistry. Bios Scientific Publishers. Oxford UK. p61-80. Hattori, T., Vasil, V, Rosenkrans, L., Hannah, L.C., McCarty, D.R. and Vasil, I K. (1992). The Viviparous] gene and abscisic acid activate the CJ regulatory gene for

PAGE 125

119 anthocyanin biosynthesis during seed maturation in maize. Gene Development 6, 609-618. Helentjaris, T., Weber, T.D.F. and Wright, S. (1988). Duplicate sequences in maize and identification of their genomic locations through restriction fragment length polymorphisms. Genetics 6, 609-618. Hildmann, T., Ebneth, M., Pena-Cortes, H., Sanchez-Serrano, J.J., Willmitzer, L. and Prat S. (1992). General roles of abscisic and jasmonic acids in gene activation as a result of mechanical wounding. Plant Cell 4, 1 157-1 170. Hirai, N. (1986). Abscisic acid. In: Chemistry of Plant Hormones, N. Takahashi, ed. Boca Raton, CRC Press, pp20 1-248. Hoecker, U., Vasil, I.K., McCarty, D R. (1995). Integrated control of seed maturation and germination programs by activator and repressor functions of Viviparous] of Maize. Gene Development 9, 2459-2469. Ingram, J. and Bartels, D. (1996). The molecular basis of dehydration tolerance in plants. Ann. Rev. Plant Physiol. Plant Mol. Biol. 47, 377-403. Jacobsen, J.V., Gubler, F., and Chandler, P.M. (1995). Gibberellin action in germinated ceral grains. In, Plant Hormones, Physiology, Biochemistry and Molecular Biology. Davies, P.J. (ed.), Kluwer Academic Publishers, Dordrecht, Netherlands, pp.246-271. Kamoda, S. and Saburi, Y, (1993b). Structure and enzymatical comparison of lignostilbene-a,P-dioxygenase isozymes, I, H, and III, from Pseudomonas paucimobilis TMY1009. Biosci. Biotech. Biochem.57, 93 1-934. Kamoda, S. and Saburi, Y. (1993a). Cloning, expression, and sequence analysis of a lignostilbene-a, P-dioxygenase gene from Pseudomonas paucimobilis TMY 1 009. Biosci. Biotech. Biochem. 57, 926-930. Kao, C.Y., Cocciolone, S.M., Vasil, I.K., and McCarty, D R. (1996). Localization and interaction of the cis-acting elements for abscisic acids, VTVIPAROUSl, and light activation of the Cl gene of maize. Plant Cell 8, 1 1 7 1 1 1 79. Koomneef, M., and Karssen, CM. (1994). Seed dormancy and germination. In: Arabidopsis. Meyerowitz, E.M., and Somerville, C.R. eds. Cold Spring Harbor Laboratory Press, Cold Spring harbor, NY. pp.3 13-334. Koomneef, M., Hanhart, C.J., Hilhorst, H.W.M., and Karssen, CM. (1989). In vitro inhibition of seed development and reserve protein accumulation in recombinants of abscisic acid biosynthesis and responsiveness mutants m Arabidopsis thaliana. Plant Physiol. 90, 463-469. Koomneef, M., Joma, M L., Brinkhorst-van der Swan, D.L.C, and Karssen, CM (1982). The isolation of abscisic acid deficient mutants by selection of induced revertants in non-germinating gibberellin-sensitive lines of Arabidopsis thaliana. Theor. Appl. Genet. 61, 385-393. Koomneef, M., Reuling, G, and Karssen, CM. (1984). The isolation and characterization of abscisic acid-insensitive mutants of Arabidopsis thaliana. Physiol Plant 61, 377-383.

PAGE 126

120 Lang, v., Mantyla, E., and Welin, B. (1994). Alterations in water status, endogenous abscisic acid content, and expression of rabl8 gene during the development of fiQQzm^XoXermct'mArabidopsisthaliana. Plant Physiol. 104, 1341-1349. Leon-Kloosteriel, K.M., Gil, M.A., Ruijs, G.J., Jacobsen S.E., Olszewski, N.E., Schwartz S.H., Zeevarrt, J.A.D., and Koomneef, M. (1996). Isolation and characterization of abscisic acid-deficient ^rai/dlcp^w mutants at two new loci. Plant J. 10, 655-661. Leon-Kloosterziel, K.M. (1996) Arabidopsis mutants with a reduced seed dormancy. Plant Physiol. 110, 233-240. Leung, J., Bouvier-Durand, M., Morris, P., Guerrier, D., Chefdor, F., and Giraudat, J. (1994). Arabidopsis ABA response gene abil, features of a calcium-modulated protein phosphatase. Science 264,1448-1452. Leung, J., Merlot, S., and Giraudat, J. (1997). Ihe Arabidopsis abscisic acid insensitive2 (ABI2) and ABU genes encode homologous protein phosphatases 2C involved in abscisic acid signal transduction. Plant Cell 9, 759-771. Li, Y., and Walton, D.C. (1987). Xanthophyll and abscisic acid biosynthesis in waterstressed bean leaves. Plant Physiol. 85, 910-915. Li, Y., and Walton, D.C. (1990a). Effects of cycloheximide on abscisic acid biosynthesis and stomatal aperture in bean leaves. Plant Physiol. 93, 128-130. Li, Y., and Wahon, D.C, (1990b). Violaxanthin is an abscisic acid precursor in waterstressed dark-grown bean leaves. Plant Physiol. 92, 551-559. Liang, X., Abel, S., Keller, J.A., Shen, N.F., and Theologis, A. (1992). The 1aminocyclopropane-l-carboxylate synthase gene family of Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 89, 11046-11050. Lorberth, R., Danman, DC, Ebnath, M., and Sanchez-serrano, J. (1992). Promoter elements involved in environmental and developmental control of potato proteinase inhibitor 11 expression. Plant J. 2, 477-486. Mansfield, T.A., and McAinsh, MR. (1995). Hormones as regulators of water balance. In: Plant Hormones: Physiology, Biochemistry and Molecular Biology. Davies, P.J. (ed ), Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 598-616. Mantyla, E., Lang, V., and Palva, E.T. (1995). Role of abscisic acid in drought-induced fi-eezing tolerance, cold acclimation and accumulation of LTI78 and RablS protein xti Arabidopsis thaliaria. VlaxAVhysiol 107, 141-148. Marcotte, W.R., Guiltinan, M.J. and Quatrano, R.S. (1992). ABA-regulated gene expression, cis-acting sequences and trans-acting factors. Biochem. Soc. Trans. 20, 93-97. Marin, E., Nussaume, L., Quesada, A., Gonneau, M., Sotta, B., Hugueney, P., Frey, A., and Marion-Poll., A. (1996). Molecular identification of zeaxanthin epoxidase of Nicotiana plumbaginifolia, a gene involved in abscisic acid biosynthesis and corresponding to the ABA locus of Arabidopsis thaliana. EMBO 15, 233 1-2342.

PAGE 127

121 Markwell, J., Bruce, B.D., and Keegstra, K. (1992). Isolation of a carotenoid-containing sub-membrane particle from the chloroplastic envelope outer membrane of pea {Pisum sativum). J. Biol. Chem. 267, 13933-13937. McCarty, D.R, and Carson, C.B. (1990). Molecular genetics of seed maturation in maize Plant Physiol .81, 267-272. McCarty, D.R. (1995). Genetic control and integration of maturation and germination pathways in seed development. Ann. Rev. Plant Physiol. Plant Mol. Biol. 46, 71-93. McCarty, D R., Carson, C.B., Stinard, P.S., and Robertson, D.S.(1989). Molecular analysis of viviparous-1, an abscisic acid insensitive mutant of maize. Plant Cell 1, 523-532. McCarty, D.R., Hattori, T., Carson, C.B, Vasil, V., and Vasil, I.K. (1991). The viviparous] developmental gene of maize encodes a novel transcriptional activator. Cell 66, 895-905. Meyer, K., Leube, M P, and Grill, E. (1994). A protein phosphatase 2C involved in ABA signal transduction in y4raZ)/
PAGE 128

'>fl>v122 Okamoto, M., Hirai, N., and Koshimizu, K. (1987). Occurrence and metabolism of 1', 4'trans-diol of abscisic acid. Phytochem. 26, 1269-1271. Ooms, J.J.J., Leon-Kloosterziel, K.M., and Bartels, D. (1993). Acquisition of desiccation tolerance and longevity in seeds ofArabidopsis thaliana. A comparative study using abscisic acid-insensitive aZ>/5 mutants. Plant Physiol. 102, 1185-1191. Oritani, T., and Yamashita, K. (1985). Biosynthesis of (+)-abscisic acid in Cercospora cruenta. Agric. Biol. Chem. 49, 245-249. Oritani, T., Niitsu, M., Kato, T., and Yamashita, K. (1985). Isolation of (2Z, 4E>yionylideneethanol from Cercospora cruenta, a fungus producing (+)-abscisic acid. Agric. Biol. Chem. 49, 2819-2822. Parry, A.D., and Morgan, R. (1991). Carotenoids and abscisic acid (ABA) biosynthesis in higher plants. Physiol. Plant. 82, 320-326. Parry, A.D., and Morgan, R. (1992a). Abscisic acid biosynthesis in roots. I. The identification of potential abscisic acid precursors, and other carotenoids. Planta 187, 185-191. Parry, A.D., and Morgan, R. (1992b). Abscisic acid biosynthesis in roots, n. The effects of water-stress in wild-type and abscisic-acid-deficient mutant (notabilis) plants of Lycopersicon esculentum Mill. Planta 187, 192-197. Parry, A.D., Babiano, M.J., and Morgan, R. (1990). The role of cw-carotenoids in abscisic acid biosynthesis. Planta 182, 118-128. Parry, A.D., Blonstein, A.D., Baviano, M.J., King, P.J., and Morgan, R. (1991). Abscisicacid metabolism in a wihy mutant of Nicotiana plumbaginifolia. Planta 183, 237-243. Parry, A.D., Neill, S.J., and Morgan, R. (1988). Xanthoxin levels and metabolism in wildtype and wilty mutants of tomato. Planta 173, 397-404. Parry, A.D., Neill, S.J., and Morgan, R. (1990). Measurement of xanthoxin in higher plant tissues using '^C labeled internal standards. Phytochem. 29, 1033-1039. Pei, Z.M., Kuchitsu, K., and Ward, J.M. (1997). Differential abscisic acid regulation of guard cell slow anion channels in Arabidopsis wild-type and abil and abi2 mutants. Plant Cell 9, 409-23. Pena-Cortes, M., and Willmitzer, L. (1995). The role of hormone in gene activation in response to wounding. In: Plant Hormones: Physiology, Biochemistry and Molecular Biology. Davies, P.J. (ed.), Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 395-414. Pena-Cortes, H., Sanchez-Serrano, J. J., and Mertens, R. (1989). Abscisic acid is involved in the wound-induced expression of the proteinase inhibitor II gene in potato and tomato. Proc. Natl. Acad. Sci. USA 86, 9851-9855. Percival, M.D. (1991). Muman 5-lippoxygenae contains an essential iron. J. Biol. Chem. 266, 10058-10061. Prescott, AG., and John, P. (1996). Dioxygenases, molecular structure and role in plant metabolism. Ann. Rev. Plant Physiol. Plant Mol. Biol. 47, 245-271.

PAGE 129

123 Quarrie, S.A. (1982). Droopy, a wilty mutant of potato deficient in abscisic acid. Plant Cell Environ. 5, 23-26. Quatrano, R., Battels, D., Ho, T.D., and Pages, M. (1997). New insights into ABAmediated processes (Meeting Report). Plant Cell 9, 470-475. Roberson, D.S. (1952). The genotype of the endosperm and embryo as it influences vivipary in maize. Proc. Natl. Acad. Sci. USA 38, 580-583. Roberson, D.S. (1955). The genetics of vivipary in maize. Proc. Natl. Acad. Sci. USA 40, 745-760. Robichaud, C, and Sussex, I.M. (1986). The response of vivioarous-l and wild type embryos of Zea mays to culture in the presence of abscisic acid. J. Plant Physiol. 126, 235-242. Rock, CD., and Quatrano, R.S. (1995). The role of hormones during seed development. In: Plant Hormones: Physiology, Biochemistry and Molecular Biology. Davies, P.J. (ed.), Kluwer Academic Publishers, Dordrecht, Netherlands, pp.671-697. Rock, CD., and Zeevaart, J. AD. (1991). The aba mutant of Arabidopsis thaliana is impaired in epoxy-carotenoid biosynthesis. Proc. Natl. Acad. Sci. USA 88, 74967499. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular Cloning, A Laboratory Manual (2nd ed ). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Schroeder, J.I., and Hagiwara, S. (1990). Repetitive increases in cytosolic Ca^^ of guard cells by abscisic acid activation of nonselective Ca^* permeable channels. Proc. Natl. Acad .Sci. USA 87, 9305-9309. Schroeder, J.I., and Keller, B .U. (1992). Two types of anion channel currents in guard cells with distinct vohage regulation. Proc. Natl. Acad. Sci. USA 89, 5025-5029. Schwartz, A, Wu, W.H., Tucker, E.B., and Assmann, S.M. (1994). Inhibition of inward channels and stomatal response by abscisic acid, an intracellular locus of phytohormone action. Proc. Natl. Acad. Sci. USA 91, 4019-4023. Schwartz, S.H., Leon-Kloosterziel, K.M., Koomneef, M., and Zeevaart, J.A.D. (1997). Biochemical characterization of the aba2 and aba3 mutants in Arabidopsis thaliana. Plant Physiol. 114, 161-166. Schwartz, S.H., Tan, B.C. Gage, D A, Zeevaart, J.A.D., and McCarty, D R. (1997). Specific oxidative cleavage of carotenoids by VP 14 of maize. Science 276, 18721875. Sharp, R.E., Wu, Y., Voetberg, G.S., Saab, I.N., and LeNoble, M.E. (1994). Confirmation that abscisic acid accumulation is required for maize primary root elongation at low water potentials. J. Exp. Bot. 45, 1743-1751. Sheen, J. (1996). Ca^'^-dependent protein kinases and stress signal transduction in plants. Science 274, 1900-1902.

PAGE 130

124 Shen, Q., and Ho, T.H. (1995). Functional dissection of an abscisic acid (ABA)-inducibIe gene reveals two independent ABA-responsive complexes each containing a G-box and a novel cis-acting element. Plant Cell 7, 295-307. Shen, Q., Zhang, P., and Ho, T.H. (1996). Modular nature of abscisic acid (ABA) response complexes, composite promoter units that are necessary and sufficient for ABA induction of gene expression in barley. Plant Cell 8, 1 107-1 1 19. Sindhu, R.K., and Wahon, D C. (1987). Conversion of xanthoxin to abscisic acid by cellfree preparations from bean leaves. Plant Physiol. 85, 916-21 Sindhu, R.K., and Walton, D C. (1988). Xanthoxin metabolism in cell free preparations from wild type and wilty mutants of tomato. Plant Physiol. 88, 178-182. Sindhu, R.K., and Wahon, D.C. (1990). Abscisic aldehyde is an intermediate in the enzymatic conversion of xanthoxin to abscisic acid in Phaseolus vulgaris L. leaves. Plant Physiol. 93, 689-694. Skriver, K., and Mundy, J. (1990). Gene expression in response to abscisic acid and osmotic stress. Plant Cell 2, 503-512. Skriver, K., Olsen, F.L., Rogers, J.C., and Mundy, J. (1991). cis-acting DNA elements responsive to gibberellin and its antagonist abscisic acid. Proc. Natl. Acad. Sc.i USA 88, 7266-7270. Smith, J.D., and NeufFer, M.G. (1992). ViviparouslO, a new viviparous mutant in maize. Maize Genet. Coop. Newslet. 66, 34. Suzuki, M., Kao, C.Y., and McCarty, D R. (1997). The conserved B3 domain of VIVIPAROUS 1 has a cooperative DNA binding activity. Plant Cell 9, 799-807. Tan, B.C., Schwartz, S.H, Zeevaart, J.A.D., and McCarty, D R. (1997). Genetic control of abscisic acid biosynthesis in maize. Proc. Natl. Acad. Sci. USA. 94, 12235-12240. Tardieu, F., Davies, W.J. (1992). Stomatal response to abscisic acid is a fiinction of current plant water status. Plant Physiol. 98, 540-5. Taylor, I B. (1991). Genetics of ABA synthesis. In: Abscisic Acid, Physiology and Biochemistry. Bios Scientific Publishers Limited. Oxford, UK, pp. 23-38. Taylor, I.B., Linforht, R.S.T., Al-naieb, R.J., Bowan, W.R., and Marples, B.A. (1988). The wilty tomato mutants flacca and sitiens are impaired in the oxidation of ABAaldehyde to ABA. Plant Cell Environ. 1 1, 739-745. Thiel, G., Blatt, M R., and Fricker, M.D. (1993). Modulation of channels in Vicia stomatal guard cells by peptide homologs to the auxin-binding protein C-terminus. Proc. Natl. Acad. Sci. 90, 1 1493-1 1497. Tietz, D., Dorffling, K., Whorle, D., Erxleben, I., Liemann, F. (1979). Identification by combined gas chromatography-mass spectrometry of phaseic acid and dihydrophaseic acid and characterization of further abscisic acid metabolites in pea seedlings Planta 147, 168-73.

PAGE 131

125 Tranel, P. J., Froehlich, J., Goyal, A. and Keegstra, K. (1995). A component of the chloroplastic protein import apparatus is targeted to the outer envelope membrane via a novel pathway. EMBO J. 14, 2436-2446. Vartanian, N., Marcotte, L., and Giraudat, J., (1994). Drought rhizogenesis in Arabidopsis thalina. Differential responses of hormonal mutants. Plant Physiol. 104, 761-767. Vasil, v., Marcotte, W.R., Cocciolone, S.M., Vasil, I.K., Quatrano, R.S., and McCarty, D.R. (1995). Overlap of Viviparous 1 (VPl) and abscisic acid response elements in the Em promoter, G-box elements are sufficient but not necessary for VPl transactivation. Plant Cell 7, 1511-1518. Vierstra, R.D. (1996). Proteolysis in plants, mechanisms and functions. Plant Mol. Biol. 32, 275-302. Walker-Simmons, M. (1987). ABA levels and sensitivity in developing wheat embryos of sprouting resistant and susceptible cultivars. Plant Physiol. 84, 61-66. Walker-Simmons, M., Kudma, D.A., and Warner, R.L. (1989). Reduced accumulation of ABA during water stress in a molybdenum cofactor mutant of barley. Plant Physiol. 90, 728-733. Walton, D.C., and Li, Y. (1995). Abscisic acid biosynthesis and metabolism. In: Plant Hormones, Physiology, Biochemistry and Molecular Biology. Davies, P.J. (ed.), Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 140-157. Walton, D.C., Galson, E., Harrison, M. A. (1977). The relationship between stomatal resistance and abcisic-acid levels in leaves of water-stessed bean plants. Planta 133, 145-148. Wolf, G., and Phil, D. (1995). The enzymatic cleavage of P-carotene is still controversial. Nutrition Reviews 53, 134-137. Wright, S.T.C., and Hiron, R.W.P., (1969). (+) abscisic acid, the growth inhibitor induced in detached wheat leaves by a period of wilting. Nature 224, 719-720. Yamaguchi-Shinozaki, K. and Shinozaki, K. (1994). A novel cis-acting element in an Arabidopsis gene is involved in responsiveness to drought, low-temperature, or highsalt stress. Plant Cell 6, 251-264 Yeum, K.J., Leekim, Y.C., Yoon, S., Lee, K.Y., Park, I S., Lee, K.S., Tang, G.W., Russell, R.M., and Krinsky, N.L (1995). Similar metabolites formed from P-carotene by human gastric-mucosa homogenates, lipoxygenase, or linoleic-acid hydroperoxide. Arch. Biochem. Biophy. 321, 167-174. Zechmeister L. (1962). Cis-trans Isomeric Carotenoids, Vitamin A, and Arylpolyenes. Academic Press, New York Zeevaart, J.A.D., and Creelman, R.A. (1988). Metabolism and physiology of abscisic acid. Ann. Rev. Plant Physiol. Plant Mol. Biol. 39, 439-73.

PAGE 132

126 Zeevaart, J.A.D., Heath, T.G., and Gage, D.A., (1989). Evidence for a universal pathway of abscisic acid biosynthesis in higher plants from incorporation pattern. Plant Physiol. 91, 1594-1601. Zhang, J. Schurr, U., and Davies, W.J. (1987). Control of stomatal behavior by abscisic acid which apparently originates in the roots. J. Exp. Bot. 38, 1 1 174-1 181. Zhang, J., and Davies, W.J. (1987). Increased synthesis of ABA in partially dehydrated root tips and ABA transport from roots to leaves. J. Exp. Bot. 38, 2015-2023. Zhang, J., and Davies, W.J. (1990). Changes in the concentration of ABA in xylem sap as a function of changing soil water status can account for changes in leaf conductance and growth. Plant Cell Environ. 13, 277-285.

PAGE 133

Biographical Sketch Bao-Cai Tan was bom on October 26, 1963, in Henan Province, People's Republic of China. He completed high school in 1980 and entered Lanzhou University. He received his B. S. in plant physiology in 1984, and master's degree in plant physiology in 1987 at Lanzhou University. He was hired as an assistant professor in the Department of Biology at Lanzhou University in 1990 and left that position in 1992 as he moved to the University of Florida to begin his Ph.D. study in plant molecular and cellular biology. 127

PAGE 134

I certify that I have read this study and that in my opinion it conform to acceptable standards of scholarly presentation and is fully adequate, in ^ope and quality, as a dissertation for the degree of Doctor of Philosophy. Donald IvNfcGarty,' Professor of Plant Molecular and Cellular Biology I certify that I have read this study and that in my opinion it conform to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philos ophy L. Curtis Hannah Professor of Plant Molecular and Cellular Biology I certify that I have read this study and that in my opinion it conform to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Karen E. Koch Professor of Plant Molecular and Cellular Biology I certify that I have read this study and that in my opinion it conform to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. William B. Gurley Associate Plant Molecular \ Biology id Cellular I certify that I have read this study and that in my opinion it conform to acceptable standards of scholarly presentation and is fully ad^qjiate, in ^ope and quality, as_a dissertation for the degree of Doctor of Philosophy. Nigel G. Richards Associate Profess of Chemistry

PAGE 135

This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1997 Dean, College of^ Agriculture Dean, Graduate School