, m ' A - V 7- T '3 fr; 0%“; "3'..- ,fifmfi
‘ ' :2 ! VII ‘5‘ ' é J 7%.}??? 5W3?" 5*‘Efi’fi‘t‘f’ig‘é 'fi’b‘g‘é‘:

‘7 ' ’r ““5%:u ‘9‘ ~'_
: vi A 1‘ ‘ 1:55.: .;.?£;\-;§€.v

., .Is‘
m: yaw“: :1'.
2‘ “1X" _ :f‘rzfirpfitchfl V1
~ .1 "fitgiv.m:1.,-_
3 gm: ' a
w

w
7 y a
.f :3. fl)..-
;- “.5. “9;:
£71 .95.". ~ 1-“,
J J

L._., . ‘ 1. Lit ;

 

.
mam v‘§'€€‘m

if"?
‘(u .4»,
~~ «“ ~¢

u . :3"?
a“ “I .
.._..."mv?(hg~.¢.
«a. u.-.
“-1“:

1-,
..‘ .. . .~.
. «:3 » v “m ~
.. a - $4,: "1.3:"; .- . ~ .. , ’ 4 ' .. ‘
. L :9: , 4 , . 0" 4 . ' :1 fwfi. if - ' ~' ~ Kr."
Fwd-r.- . ‘ ' . at W. . ‘ ‘* ’ ‘1’ was? arm-.2351: wag-«w 1-1
w r.5-..’~IJ:].7 ‘ , . {I - — , - t- . .- \hnhn‘»\l .~..r ‘
4 «a. .-. . J. ,- (. .. mm.- o ‘ Jar-«~5- u- ..
L unfufi» . — - ’ J‘v .
.. - . . 45 “a ‘1 M4- .. t N. , ‘1
f. ,» . I 1., NM »4 I." J r v 3‘." .53 - .J‘N‘ xx 0
sum 3. nun} «a. In afixlg. «
'v-uk . - 4:». 3'». ".1 .Jy;
(I “‘22." viva!!!“ hifid‘i‘ :fiu‘ u
. ‘{.:."v...17 .rw's‘uu»
.fh’L‘I‘L! a u

’1”
‘K

n

f

2‘.‘
r‘ld m:

‘11

xi w”
. ..;J{ 3341*,"
"4-: J 1‘ —‘ l

«a» -'
x.‘ .;I' A
5"3.‘ §2~

 

'VIJ‘VL'
u.
“%
v.
w. A’ZL !
a» Aw
I '3."
11“-

_,

 

_‘...,_",

,1:
l
1.1
_¢
‘5‘

' '1
1- t: ,
1'

.ul .
‘ nin-

a? .44;
l‘n .‘E

t".

II'

 

--. L“ :...'
“'an -

MICHIGAN STATE U NERSITY IES

If! I/flil/IHWIII

 

 

 

 

 

 

1m {NIH/II WWI/w

3 1293 00901 1564

 

 

This is to certify that the

dissertation entitled

BIOCHEMICAL GENETICS OF ABSCISIC ACID BIOSYNTHESIS

presented by

Christopher DaIe Rock

has been accepted towards fulfillment
of the requirements for

Ph.D. Genetics/Botany

degree in

 

 

We»

Major professor

 

 

‘Dam ApriI 23, 1991

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

r’ \
LIBMRY
Michigan State
University

 

 

L

fl

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVG. FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution
chS-at

 

fl

BIOCHEMICAL GENETICS OF ABSCISIC ACID BIOSYNTHESIS

By

Christopher Dale Rock

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Program in Genetics

Department of Botany and Plant Pathology

1991

65’7”“ 4 57 33

ABSTRACT
BIOCHEMICAL GENETICS OF ABSCISIC ACID BIOSYNTHESIS
by
Christopher Dale Rock

The iron chelators a,a-dipyridyl and o-phenanthroline inhibited
abscisic acid (ABA) biosynthesis in water-stressed leaves of
Xanthium snumarium. This supports the hypothesis that a non-heme
oxygenase is involved in ABA biosynthesis. The onium-type plant
growth retardant PC-QO'Z inhibited ABA biosynthesis, but the anti-
transpirant LAB 173711 did not.

ABA-aldehyde and ABA-trans-diol were quantified in apple
(Mains domestics) fruits and leaves. 18O-Labeling experiments
established that [1°O]ABA-a1dehyde was largely unlabeled in the side
chain carbonyl group due to exchange of the carbonyl oxygen with
water. This can explain the unusual labeling patterns of [1°O]ABA in
labeling experiments with apple fruits. Based on specific activity of
[1°O]ABA and [190]ABA-trans-diol, it is concluded that ABA-trans-diol
is a catabolite of ABA. Results of feeding studies with [21-16]ABA-
aldehyde and FHSJABA support the biosynthetic relationship ABA-

aldehyde --> ABA --> ABA-trans-diol. A parallel pathway of trans-

ABA biosynthesis is prOposed.

ABA and vans-ABA biosynthesis and metabolism were
investigated in the ABA-deficient mutants of tomato (Lyc0persicon
esculentum), potato (Solanum tuberosum), barley (Hordeum vulgare)
and Arabidopsis thaliana. In the tomato, potato, and barley mutants,
a high percentage of [”OJABA and [190] trans-ABA was doubly-
labeled in the side chain carboxyl group. Feeding studies in tomato
with FHGJABA-alcohol and 18O2 established that this doubly-carboxyl-
labeled [‘BOJABA was synthesized from [1801ABA-alcohol with
incorporation of molecular oxygen. ABA-alcohol oxidation was
inhibited by carbon monoxide, which indicates the involvement of a
cytochrome P-450 monooxygenase. This minor shunt pathway from
ABA-aldehyde --> ABA-alcohol --> ABA operates in all species
examined and is an important component of ABA biosynthesis in
mutants impaired in ABA-aldehyde oxidation.

Analysis of carotenoids established that the aba mutant of
Arabidopsis is impaired in epoxy-carotenoid biosynthesis. This
result supports the hypothesis that ABA is synthesized from
xanthOphylls via oxidative cleavage of the epoxy-carotenoids

violaxanthin and neoxanthin. Trans-ABA biosynthesis was less

afiected than ABA biosynthesis in the aba genotypes. Feeding
experiments suggest that Uans—xanthoxin may be a precursor to ABA
in Arabidopsis.

The xanthophyll cycle was utilized to specifically 18O label the
epoxy group of violaxanthin in Spinach (Spinacia oIeracea) leaves.
Subsequent ABA biosynthesis resulted in 18O incorporation
specifically into the l’-hydroxyl group of [1301ABA. This supports the

precursor role of violaxanthin in ABA biosynthesis.

Dedicated to my mother

ACKNOWLEDGEMENTS
I would like to thank those who helped me in my work; Ian
Zeevaart, Doug Gage, Tim Heath, and Kate Noon. I would also like
to thank my committee members, Andrew Hanson, Hans Kende, Iohn
Ohlrogge, and Chris Somerville. I wish to thank Austen D.
Warburton for his support and guidance since my father’s death.
And thanks to my friends, and eSpecially Amy, for helping me

through a challenging and rewarding time of my life.

Having pried through the strata, analyzed to a hair, counsel’d
with doctors and calculated close,

I find no sweeter fat than sticks to my own bones.

Walt Whitman
"Song of Myself"

Leaves of Grass

TABLE OF CONTENTS

Page

List of Tables .................................................................................. xv

List of Figures ................................................................................ xx

List of Abbreviations ..................................................................... xxiii

Chapter 1. INTRODUCTION ................................................................. 1

1.1. Ovemew 2

1.2. ABA Biosynthesis .................................................................... 4

1.3. Statement of Purpose ............................................................ 7

1.4. Literature Cited ...................................................................... 8
Chapter 2. THE IRON CHELATORS a, a-DIPYRIDYL AND 0-
PHENANTHROLINE, AND THE GROWTH RETARDANT

FC-90'I, INHIBIT ABSCISIC ACID BIOSYNTHESIS ................ 13

2.1. Introduction ............................................................................ 14

2.2. Materials and Methods ........................................................ 16

2.3. Results and Discussion ........................................................ 17

2.3.1. Feeding Studies with Iron Chelators .................. 17

Page

2.3.2. Feeding Studies with LAB 173711 ....................... 22
2.3.3. Feeding Studies with FC-907 ............................... 24
2.4. Literature Cited ..................................................................... 26

Chapter 3. ABSCISIC (ABA) -ALDEHYDE IS A PRECURSOR

TO, AND 1’-4’- TRANS-ABA—DIOL A CATABOLITE OF,

ABA IN APPLE ................................................................................ 30
3.1. Abstract ................................................................................... 31
3.2. Introduction ............................................................................ 31
3.3. Materials and Methods ........................................................ 31
3.3.1. Plant Material ........................................................... 31
3.3.2. Extraction and Purification of Metabolites ........ 32
3.3.3. Quantifications ......................................................... 32
3.3.4. MS ............................................................................... 32
3.3.5. Chemicals ................................................................. 32
3.4. Results ..................................................................................... 33
3.4.1. ABA Biosynthesis in Fruits and Leaves .............. 33

3.4.2. ABA-Aldehyde and ABA- Trans-Diol in Apple

Fruits and Leaves .................................................... 34

Page
3.4.3. Quantification of ABA-Aldehyde and ABA-
Trans-Diol .................................................................. 36
3.4.5. Feeding Studies with Deuterated Substrates... 36
3.5. Discussion ............................................................................... 38
3.5.1. ABA-Aldehyde as a Precursor to, and ABA-
Trans-Diol as a Catabolite of ABA. ....................... 38

3.5.2. ABA Biosynthesis in Apple Fruits and

Leaves ......................................................................... 38
3.6. Acknowledgements ............................................................... 38
3.7. Literature Cited ...................................................................... 38

Chapter 4. ABSCISIC (ABA) ~ALCOHOL IS AN INTERMEDIATE

IN ABA BIOSYNTHESIS IN A SHUNT PATHWAY FROM

ABA-ALDEHYDE ............................................................................. 40
4.1. Abstract .................................................................................... 41
4.2. Introduction ............................................................................. 42
4.3. Materials and Methods ........................................................ 45
4.3.1. Plant Material ........................................................... 45
4.3.2. Feeding Experiments .............................................. 46
4.3.3. ABA, ABA-CE, and PA Analysis ............................ 46

Page
4.3.4. Chemicals ................................................................. 47
4.4. Results ..................................................................................... 47
4.4.1. ABA and Trans-ABA Biosynthesis in the ABA-
Deficient Tomato Mutants ...................................... 47
4.4.2. ABA and Trans-ABA Biosynthesis in the ABA-
Deficient Potato and Barley Mutants ................... 59

4.4.3. ABA-Alcohol Oxidation in the Hacca and

Sitiens Mutants .......................................................... 62
4.5. Discussion ............................................................................... 68
4.6. Literature Cited .................................................................... 76

Chapter 5. THE aba MUTANT OF ARABIDOPSIS TIflILIANA

IS IMPAIRED IN EPOXY-CAROTENOID BIOSYNTHESIS ..... 80
5.1. Abstract ................................................................................... 81
5.2. Introduction ........................................................................... 82
5.3. Materials and Methods ........................................................ 86

5.3.1. Plant Material ........................................................... 86

5.3.2. ABA. ABA-Glucose Ester (ABA-GE), and
Phaseic Acid (PA) Analysis .................................... 87

5.3.3. Carotenoid Determinations ................................... 89

Page
5.4. Results ..................................................................................... 91

5.4.1. ABA Biosynthetic Capacity is Negatively

Correlated with the Phenotypic Severity

Associated with the aba Alleles ............................ 91
5.4.2. ABA Precursor Pool Size is Correlated with

ABA Biosynthesis in aba Genotypes .................... 95
5.4.3. Epoxy-Carotenoid Deficiency and Zeaxanthin

Accumulation are Correlated with the Small

ABA Precursor Pool in the aba Genotypes ........ 97
5.5. Discussion .............................................................................. 102
5.6. Literature Cited ..................................................................... 103

Chapter 6. 2- TRANS-ABA BIOSYNTHESIS AND THE
METABOLISM OF ABA-ALDEHYDE AND XANTHOXIN

IN WILD TYPE AND THE aba MUTANT OF

ARABIDOPSIS THALWVA ........................................................... 107
6.1. Abstract ................................................................................... 108
6.2. Introduction ............................................................................ 108
6.3. Materials and Methods ...................................................... 111

6.3.1. Plant Material ........................................................... 111

Page
6.3.2. Feeding EXperirnents ............................................. 112
6.3.3. Extraction, Purification, and Quantitation
of Metabolites ............................................... 1 13
6.4. Results ..................................................................................... 114
6.4.1. ABA and Trans-ABA Biosynthesis and
Metabolism ................................................................ 1 14

6.4.2. ABA-Aldehyde and Xanthoxin Metabolism in

aba Genotypes .......................................................... 123
6.5. Discussion .............................................................................. 128
6.6. Literature Cited ..................................................................... 133

Chapter 7. 18O INCORPORATION INTO VIOLAXANTHIN AND
ABSCISIC ACID (ABA) VIA THE XANTHOPHYLL CYCLE

SUPPORTS VIOLAXANTHIN AS A PRECURSOR OF ABA... 136

7.1. Introduction ............................................................................ 137
7.2. Materials and Methods ....................................................... 139
7.3. Results and Discussion ........................................................ 141
7.4. Literature Cited ..................................................................... 149
Chapter 8. SUMMARY ............................................................................ 152
8.1. Literature Cited ..................................................................... 155

.lppe

Appe

Appe

AIlpe

Page
Appendix A. PARTITION COEFFICIENTS OF ABA-ALDEHYDE,
ABA-TRANS—DIOL AND ABA IN THREE DIFFERENT
SOLVENTS AS A FUNCTION OF pH ....................................... 156
Appendix B. STABILITY OF DEUTERIUM LABEL IN ABA-
ALDEHYDE AND ABA-TRANS-DIOL AT VARIOUS pHs ....... 158
Appendix C. CORRECTION FACTORS FOR CAROTENOID
INTEGRATION DATA .................................................................. 159

Appendix D. CALCULATIONS OF NATURAL ISOTOPE

CONTRIBUTIONS TO MASS SPECTRAL DATA. ................... 160
D.1. Mass Spectral and SIM Data ............................................ 160
D2. MS/MS Data .......................................................................... 162
D3. FAB-MS of Carotenoids ...................................................... 167
D4. Literature Cited .................................................................... 172

xiv

LIST OF TABLES

Table Page
3.1. Biosynthesis and Catabolism of ABA in Mutsu Fruit and

Leaves under 18O2 .......................................................................... 33
3.1L Extent of ABA Functional Group 18O-Labeling in Mutsu

Leaf, Immature and Mature Fruit Tissues ................................ 34
3.111. 18O Incorporation in the Side Chain of ABA-Aldehyde,

ABA-trans-Diol and ABA Extracted from the Same Fruit

Sample after 48 h under 18O2 ...................................................... 35
3.1V. Comparision of cis and trans-ABA-Aldehyde and

Xanthoxin Isomer Ratios by Various GLC Detection

Methods ........................................................................................... 36
3.V. Quantification of ABA, ABA-trans-Diol and cis- plus trans-

ABA-Aldehydes in Mutsu Fniit and Leaves ............................. 37
3.VI. Metabolism of Deuterated Compounds Fed to Mutsu Fruit

Tissue ............................................................................................... 37

T9;

4.2.

4.4.

4.5.

Table Page
4.1. Ouantitation of ABA, trans-ABA and Catabolites from
Unstressed Leaves of Wild Type and ABA-Deficient
Tomato Mutants ............................................................................. 48
4.2. Ouantitation of 18O-Labeled ABA, trans-ABA and
Catabolites from WaterStressed Leaves of Wild Type and
ABA-Deficient Tomato Mutants in 1302 for Various Lengths
of Time ............................................................................................. 50
4.3. Extent and Position of ABA and trans-ABA 18O-Labeling in
Water-Stressed Leaves of Wild Type and ABA-Deficient
Mutants in 1802 for 4 h .................................................................. 56
4.4. Quantitation of ABA, trans-ABA and Catabolites from
Heterozygous (Normal) and Homozygous (Mutant) drOOpy
Potato and from Wild Type and the Molybdenum Cofactor
Mutant (A234) of Barley Before and After 8 h Water
Stress ................................................................................................ 60
4.5. Extent and Position of ABA and trans-ABA l"O-Labeling in
Water-Stressed Leaves of Heterozygous (Normal) and
Homozygous (Mutant) droopy Potato and in Wild Type

and the Molybdenum Cofactor Mutant (A234) of Barley

 

Table

4.8.

5.2.

8.3. ,

Table Page

4.6.

4.7.

4.8.

5.1.

5.2.

5.3.

in ”02 for 8 h .................................................................................. 63
Ouantitation of Labeled ABA from Wild Type and FIacca
Tomato Leaves Fed Deuterated ABA-Aldehyde or ABA-
Alcohol 8 h in 18O2 ......................................................................... 65
Inhibition by Carbon Monoxide of ABA-Alcohol Oxidation

in Sitiens Leaves ............................................................................ 67
Extent and Position of ABA and trans-ABA Labeling in
Water-Stressed Leaves of the ABA-Deficient Tomato

Mutant Sitiens Incubated in the Presence or Absence of

50% Carbon Monoxide for 8 h .................................................... 69
Quantitation of mO-Labeled ABA and Catabolites in Water-
Stressed Leaves of Wild Type and Three aba Genotypes

of ArabidOpsis after Incubation for 4 or 8 hr in 1802 .............. 93
18O Incorporation into the Ring-Attached Oxygens of ABA

fiom Water-Stressed Leaves of Wild Type and Three aba
Genotypes of ArabidOpsis after Incubation in l"O2 for 4

or 8 hr ............................................................................................... 96
Ouantitation of Carotenoids and ChlorOphylls from Leaves

of Wild Type and Three aba Genotypes of ArabidOpsis ...... 101

6.2.

6.3.

8.4.

6.5.

7.2.

Table Page

6.1.

6.2.

6.3.

6.4.

6.5.

7.1.

7.2.

Quantitation of ABA, trans-ABA and Catabolites from

Leaves of ArabidOpsis Wild Type and aba Genotypes

Before and After 24 h Water Stress ........................................... 116
Quantitation of 1E3O-Labeled ABA, trans-ABA and

Catabolites from Water-Stressed Leaves of ArabidOpsis

Wild Type and aba Genotypes after 24 h in 18O2 ................... 118
Extent and Position of ABA and trans-ABA 18O-Labeling

from Water-Stressed Leaves of Wild Type and the aba
Genotype Incubated 24 h in 1802 ............................................... 121
Conversion of FHBJABA-Aldehyde to ABA and ABA

Catabolites in Leaves of ArabidOpsis Wild Type and aba
Genotypes ....................................................................................... 125
Conversion of Xanthoxin to ABA and ABA Catabolites in
Leaves of Arabi'dOpsis Wild Type and aba Genotypes

in 1°02 ................................................................................................ 127
130 Enrichment of XanthOphylls from Spinach Leaves after
Running the XanthOphyll Cycle in "’02 for 4 h ........................ 144
Position and Extent of ABA Labeling in Spinach Leaves

after Running the Xanthophyll Cycle in 18O2 for 4 h .............. 146

Table Page
A.l. Partition Coefficients of ABA-Aldehyde, ABA-trans-Diol

and ABA in Three Different Solvents as a Function of pH... 156
B.l. Stability of Deuterium Label in ABA-Aldehyde and ABA-

trans-Diol at Various pHs ............................................................. 158
CI. Correction Factors for Carotenoid Integration Data .............. 159
D.1. Example of SIM Data Correction for 13C and 180 Natural

Abundance Contributions ............................................................ 163
D2. Example of MS/MS Data Correction for 13C Natural

Abundance Contributions ............................................................ 168
D.3. Calculation of ABA Labeling Patterns from Corrected SIM

and MS/MS Data ............................................................................ 169
DA. Example of [‘°O]Violaxanthin FAB-MS Data Correction for

13C Natural Abundance Contributions ...................................... 173

LIST OF FIGURES

Figure Page
1.1. The structure of S-(+)-abscisic acid (ABA) .............................. 3
1.2. Proposed pathways to ABA, with characterized ABA

2.1.

2.2.

2.3.

2.4.

2.5.

3.1.

biosynthetic mutants ..................................................................... 5
Structures of a,a-dipyridyl (A), o-phenanthroline (B), LAB
173711 (C), and FC-907 (D) ........................................................ 15
The effect of various a,a-dipyridyl concentrations on ABA
accumulation in Xanthiurn leaves ............................................... 18
The effect of various o-phenanthroline concentrations on

ABA accumulation in Xanthi'um leaves ...................................... 19
The efi'ect of various LAB 173711 concentrations on ABA
accumulation in Xanthi'um leaves ............................................... 23
The efiect of various FC-907 concentrations on ABA
accumulation in Xanthium leaves ............................................... 25

Relationship between the rate of ethylene evolution and

Figure Page

3.2.

3.3.

3.4.

4.1.

4.2.

5.1.

5.2.

6.1.

ABA biosynthesis in Mutsu apple fruit ...................................... 33
GLC-NCI-MS of apple fruit ABA-aldehyde synthesized

under 1"02 ........................................................................................ 34
MS-MS of ABA-aldehyde containing zero, one, or two 18O
atoms ................................................................................................ 35
GLC-NCI-SIM chromatogram of ABA-aldehyde from Mutsu
fruit with [2H6]ABA-aldehyde internal standard ....................... 36
GC-NCI-MS/MS of [1802]Me-ABA (m/z = 282) from water-
stressed leaves of wild type Rheinlands Ruhm (A),

flacca (B), and sitiens (C) tomato in 1802 for 8 h ................... 55
PrOposed pathways of ABA and vans-ABA biosynthesis

from neoxanthin isomers (violaxanthin is also a

cleavage substrate) ........................................................................ 70
The structures of: A) zeaxanthin; B) trans-violaxanthin;

C) 9’-cis-neoxanthin; D) (S)-(+)-abscisic acid ....................... 84
HPLC chromatograms of carotenoids extracted from

leaves of wild type Landsberg erecta and the aha-4
genotype of Arabidopsis thaliana ............................................... 99

Pr0posed pathways of abscisic acid and trans-abscisic acid

Figure Page

biosynthesis from violaxanthin (neoxanthin is also a

substrate for cleavage) in ArabidOpsis thaliana ..................... 129
7 .1. The xanthophyll cycle (after Yamarnoto, 1979) ........................ 138
7.2. Positive ion FAB-MS spectra of the molecular ion cluster

of all-trans-violaxanthin from unlabeled (A) and 18O-

labeled (B) spinach leaves run through the xanthophyll

cycle for 4 h .................................................................................... 142

ABA

ABA-GE

ABA-trans-diol

C

Ci

ECD

EI

FAB

FID

flc

fr

CC or GLC

h

LIST OF ABBREVIATIONS

(S)-(+)-abscisic acid

B-D -gluc0pyranosyl abscisate
l’,4’-trans-ABA-diol

Celsius

Curie(s)

cultivar

einstein

electron capture detection
electron impact mass spectrometry
fast atom bombardment
flame ionization detection
flacca

fresh

gas-liquid chromatography

hour(s)

2H
3H

HPLC

15"

NCI

not

PA

deuterium

tritium

high performance liquid
chromatography

partition coefficient

liter

molarity (moles per liter)

odd electron negative molecular ion
positive molecular ion

methyl ester derivative

minute(s)

mass spectrometry
GLC-NCI-collisionally activated
dissociation-tandem mass spectrometry
mass to charge ratio

normal (concentration)

negative chemical ionization
notabilis

ortho

phaseic acid

xxiv

PCI

re:

PCI

%

SE

SIM

sit

tert

WOI’VVI

positive chemical ionization
percent

retention time

standard error of the mean
selected ion monitoring
sz'ti'ens

trans

tertiary

weight

ultraviolet radiation

volume

CHAPTERI

INTRODUCTION

 

hghe
CYEJIC

mm

rdud

1.1. OVERVEW

Abscisic acid (ABA; Figure 1.1) is a sesquiterpene plant growth
substance which was discovered in 1963 (for comprehensive review
of chemical aspects, see Hirai, 1986). ABA has been found in all
higher plants examined as well as in eukaryotic algae and
cyanobacteria (Hirsch et a1., 1989). It has even been identified in
mammalian brain (Chen et a1., 1988; Le-Page-Degivry et a1., 1986).

The physiological roles of ABA in plants are diverse and
include: closure of stomata, regulation of growth and development,
promotion of seed dormancy, and adaptation to environmental
stresses (for review of ABA physiology and metabolism see Zeevaart
and Creelman, 1988). The mechanisms of ABA action are unlcnown,
although regulation of gene expression by trans-acting factors is one
component of ABA signal transduction (Guiltinan et a1., 1990; Mundy
et a1., 1990; Slcriver and Mundy, 1990). Identification of a putative
ABA receptor has been reported (Hocking et a1., 1978; Homberg and
Weiler, 1984) but these reports have not been reproduced. ABA
regulation of intracellular calcium concentrations has recently been

demonstrated (McAinsh et a1., 1990; Schroeder and Hagiwara, 1990;

 

 

theA

McC.

Hints

3
Gehring et 31., 1990), suggesting that calcium is a second messenger

in ABA action. The three ABA-insensitive mutants of Arabidopsis
thaliana (abi-I, -2, -3) are likely to correspond to blocks in the ABA
signal transduction pathway(s) (Koornneef et a1., 1984, 1989;
Finkelstein and Somerville, 1990). The maize vivz‘parous-I (vp-I)
mutant is ABA-insensitive and identifies a gene which is involved in
the ABA response during seed deve10pment (Robichaud eta1., 1980;

McCarty et a1., 1989).

 

Figure 1.1. The structure of S-(+)-abscisic acid (ABA).

 

till

pc

lit:

he

'i

4
1.2. ABA BIOSYNTHESIS

Although the structure of ABA has been known for 25 years,
the biosynthetic pathway has still not been fully elucidated. Two
possible pathways are: the "direct" pathway involving a C15 precursor
derived from cyclization of famesyl perphosphate, or the "indirect"
pathway involving epoxy-carotenoids and xanthoxin as intermediates.
Some phytopathogenic fungi synthesize ABA via a direct pathway
(Neill et a1., 1984; Okamoto et a1., 1988) which involves ionylidene
derivatives. In higher plants the evidence for the direct pathway is
low level incorporation of mevalonic acid into ABA (Milborrow, 1974;
Hartung et 31., 1981; Cowan and Railton, 1987). These data are not
inconsistent with the indirect pathway hypothesis. The evidence for
the indirect pathway is more substantial: a) The viviparous mutants
of maize are blocked in the early stages of carotenoid biosynthesis
and are ABA—deficient (Figure 1.2; Moore and Smith, 1985; Neill et
a1., 1986). b) The carotenoid biosynthesis inhibitors fluridone and
norflurazon also inhibit ABA biosynthesis (Moore and Smith, 1984;
Henson, 1984; Ouarrie and Lister, 1984; Gamble and Mullet, 1986).

c) 18Oz-Labeling experiments with

S

Mevalonic acid

 

 

   
 

~17
Farniyl pyrophosphate
.L
PhytoJe‘ne
Phytcfene
g Carotene
Neurdlsvporene
Lycopene
\/
5 Carotene Y Carotene
1 1
a Carotene B Carotene
a Crprtoxanthin B Cryptoxanthin
Lu‘tt‘n Zeaxahtin

Lutein epoxide 7 Antheraxanthin
\A / 1
4”

Violaxanthin
Xanthoxin l,
i l o‘\ .
Z Neoxanthin
. . . Abscisic
AbsCiStc ACld ‘— aldehyde
flc \
3’: Trans -
Abscis'c alcohol
CKR1 '
A234 (narZa)

 

 

flundone
vp—2
vp-5

W3

y-9
vp-9

vp-7 = p5

y3=al

Figure 1.2. Pr0posed pathways to ABA, showing characterized

biosynthetic mutants. -

pcs
113‘
wt:

ire

6

water-stressed leaves show 180 incorporation into the side chain
carboxyl group of ABA, but little incorporation in the ring oxygen
positions (Creelinan et a1., 1987; Zeevaart et a1., 1989), indicating
that there is a large ABA precursor pool (presumably xanthophylls)
which contains oxygens on the ring. d) Xanthoxin, a Cu,-
breakdown product of epoxy-carotenoids, is found in plants (Parry et
a1., 1990) and is readily converted to ABA in vivo (Taylor and
Burden, 1973; Parry et a1., 1988) and by cell-free plant extracts
(Sindhu and Walton, 1988; Sindhu et a1., 1990). e) Most recently, a
1:1 correlation on a molar basis between decreases in trans-
violaxanthin and 9’-cis-neoxanthin levels and concomitant increases
in ABA and its catabolites has been shown for dark-grown, water-
stressed bean leaves (Li and Walton, 1990; Parry et a1., 1990).

There are ABA-deficient chlorophyll-containing mutants in
potato (Quarrie, 1982), tomato (Tal and Nevo, 1973; Parry er a1.,
1988; Taylor et a1., 1988; Sindhu and Walton, 1988), barley (Walker-
Sirnrnons et a1. 1989), Arabidopsfs thali'ana (Koornneef et a1., 1982;
Koomneef, 1985), pea (Wang et a1., 1984), maize (Neill et a1., 1986)
and Nicotiana plumbagi'nifoli’a (Parry et a1., 1991). The flacca and

sitiens of tomato, drOOpy potato, the molybenum cofactor mutant

 

hidlv
1

PE

ARV

‘ Ii.‘

7
A234 (narZ) of barley and the recently discovered wilty mutant of N.

plumbagi'm'foli‘a, GIG], are blocked in ABA-aldehyde oxidation
(Taylor et a1., 1988; Duckham et a1., 1989; Walker-Simmons et a1.,
1989; Sindhu et a1., 1990; Parry et a1., 1991); the flacca and sitiens
mutants and the wilty Nicotiana mutant accumulate trans-ABA-
alcohol as a result of the biosynthetic block (Figure 1.2; Linforth et
a1., 1987; Taylor et a1., 1988; Parry et a1., 1991). The wilty mutant of
pea can accumulate ABA in response to water stress (Wang et a1.,
1984) and it is therefore unclear whether it is an ABA biosynthetic

mutant.

1.3. STATEMENT OF PURPOSE

The goal of the research presented in this dissertation was to
elucidate the biosynthetic pathway of ABA in higher plants. Toward
this end, three approaches were used: 1) Experiments aimed at
identifying inhibitors of ABA biosynthesis; 2) Feeding of stable
isotOpes ("302, Fifi-labeled ABA precursors) and analysis of labeled
ABA by mass spectrometry to study ABA biosynthesis in vivo; and 3)

characterization of ABA biosynthesis in the ABA-deficient mutants of

tom 1C

Cher.

8

tomato, potato, barley, and Arabidopsis thaliana.

1.4. LITERATURE CITED

Chen FSC, MacTaggart IM, Wang LCH, Westly IC (1988) Analysis of
abscisic acid in the brains of rodents and ruminants.
Agric Biol Chem 52: 1273-1274

Cowan AK, Railton ID (1987) The biosynthesis of abscisic acid in a
cell-free system from embryos of Hordeum vulgare. I Plant
Physiol 131: 423-431

Creelrnan RA, Gage DA, Stults IT, Zeevaart IAD (1987) Abscisic acid
biosynthesis in leaves and roots of Xanthi'um struman’um. Plant
Physiol 85: 726-732

Duckham SC, Taylor 1B, Linforth RST, Al-Naieb RI, Marples BA,
Bowman WR (1989) The metabolism of cis ABA-aldehyde by
the wilty mutants of potato, pea and Arabidopsis thaliana. I
Exp Bot 40: 901-905

Finkelstein RR, Somerville CR (1990) Three classes of abscisic acid
(ABA) -insensitive mutations of ArabidOpsis define genes that

control overlapping subsets of ABA responses. Plant Physiol
94: 1 172-1 179

Gamble PE, Mullet IE (1986) Inhibition of carotenoid accumulation
and abscisic acid biosynthesis in fluridone-treated dark-grown
barley. EurI Biochem 160: 117-121

Gehring CA, Irving HR, Parish RW (1990) Effects of auxin and
abscisic acid on cytosolic calcium and pH in plant cells. Proc
Natl Acad Sci USA 87: 9645-9649

Guiltinan MI, Marcotte WR, Quatrano RS (1990) A plant leucine
zipper protein that recognizes an abscisic acid response
element. Science 250: 267-271

Ha

He

9

Hartung W, Heilrnann B, Gimmler H (1981) Do chloroplasts play a
role in abscisic acid synthesis? Plant Sci Lett 22: 235-242

Henson IE (1984) Inhibition of abscisic acid accumulation in seedling
shoots of pearl millet (Penm'setum amen'canum L. Leeke)

following induction of chlorosis by norflurazon. Z
Pflanzenphysiol 114: 35-43

Hirai N (1986) Abscisic acid. In N Takahashi, ed, Chemistry of Plant
Hormones. Boca Raton, FL, CRC Press, pp 201-248

Hirsch R, Hartung W, Gimmler H (1989) Abscisic acid content of
algae under stress. Bot Acta 102: 326-334

Hocking TI, Clapham I, Cattell KI (1978) Abscisic acid binding to
subcellular fractions from leaves of Vicia faba. Planta 138: 303-
304

Hornberg C, Weller EW (1984) High-affinity binding sites for abscisic
acid on the plasmalemrna of Vicia faba guard cells. Nature 310:
321-324

Koornneef M (1985) Genetic aspects of abscisic acid. In] King, ed,
Plant Gene Research. Berlin, Springer, pp 35-54

Koomneef M, Hanhart CI, Hilhorst HMW, Karssen CM (1989) In vivo
inhibition of seed development and reserve protein
accumulation in recombinants of abscisic acid biosynthesis and
reSponsiveness mutants of Arabidopsfs thaliana Plant Physiol

90: 463-469

Koornneef M, Ioma ML, Brinkhorst-van der Swan DLC, Karssen CM
(1982) The isolation of abscisic acid (ABA) deficient mutants by
selection of induced revertants in non-gerrninating gibberellin
sensitive lines of ArabidOpsis thab'ana (L.) Hey-uh. Theor Appl
Genet 61: 385-393

Koomneef M, Reuling G, Karssen CM (1984) The isolation and
characterization of abscisic acid-insensitive mutants of
Arabidopsis thaliana. Physiol Plant 61: 377-383

Le Pa;
Lil, 1

Life:

Mibc

MOO}
Moe:

Mu:

as» - «—-'

10

Le Page-Degivry M-T, Bidard I-N, Rouvier E, Bulard C, Lazdunski M
(1986) Presence of abscisic acid, a phytohorrnone, in the
mammalian brain. Proc Natl Acad Sci USA 83: 1155-1158

Li Y, Walton DC (1990) Violaxanthin is an abscisic acid precursor in
water-stressed dark-grown bean leaves. Plant Physiol 92: 551-
559

Linforth RST, Bowman WR, Griffin DA, Marples BA, Taylor IB (1987)
2-trans-ABA-alcohol accumulation in the wilty tomato mutants
flacca and sitiens. Plant Cell Environ 10: 599-606

McAinsh MR, Brownlee C, Hetherington AM (1990) Abscisic acid-
induced elevation of guard cell cytosolic Ca2+ precedes
stomatal closure. Nature 343: 186-188

McCarty DR, Carson CB, Stinard PS, Robertson DS (1989) Molecular
analysis of viviparous-I: an abscisic acid-insensitive mutant of
maize. Plant Cell 1: 523-532

Milborrow BV (1974) Biosynthesis of abscisic acid by a cell-free
system. Phytochemistry 13: 131-136

Moore R, Smith ID (1984) Growth, graviresponsiveness and abscisic
acid content of Zea mays seedlings treated with fluridone.
Planta 162: 342-344

Moore R, Smith ID (1985) Graviresponsiveness and abscisic acid

content of roots of carotenoid-deficient mutants of Zea mays L.
Planta 164: 126-128

Mundy I, Yamaguchi-Shinozaki K, Chua N-H (1990) Nuclear proteins
bind conserved elements in the abscisic acid-responsive
promoter of a rice rab gene. Proc Natl Acad Sci USA 87: 1406-
1410

Neill SI, Horgan R, Parry AD (1986) The carotenoid and abscisic acid
content of viviparous kernels and seedlings of Zea mays L.
Planta 169: 87-96

Neill SI, Horgan R, Walton DC (1984) Biosynthesis of abscisic acid.

 

Par.

Vii/‘5

11

In A Crozier, IR Hillman, eds, The Biosynthesis and Metabolism
of Plant Hormones. Soc Exper Biol Seminar Ser 23, Cambridge,
Cambridge University Press, pp 43-70

Okamoto M, Hirai N, Koshimizu K (1988) Biosynthesis of abscisic
acid. Mem Coll Agric Kyoto Univ 132: 79-115

Parry AD, Babiano MI, Horgan R (1990) The role of cis-carotenoids
in abscisic acid biosynthesis. Planta 182: 118-128

Parry AD, Blonstein AD, Babiano MI, King PI, Horgan R (1991)
Abscisic acid metabolism in a wilty mutant of Nicotiana
plumbagi'nifolia. Planta 183: 237-243

Parry AD, Neill SI, Horgan R (1988) Xanthoxin levels and metabolism
in the wild-type and wilty mutants of tomato. Planta 173: 397-
404

Parry AD, Neill SI, Horgan R (1990) Measurement of xanthoxin in
higher plant tissues using 13C labelled internal standards.
Phytochemistry 29: 1033-1039

Quarrie SA (1982) Droopy: a wilty mutant of potato deficient in
abscisic acid. Plant Cell Environ 5: 23-26

Quarrie SA, Lister PG (1984) Evidence of plastid control of abscisic
acid accumulation in barley (Hordeum vulgare L.). Z
Pflanzenphysiol 114: 295-308

Robichaud CS, Wong I, Sussex IM (1980) Control of in vitro growth
on viviparous embryo mutants of maize by abscisic acid. Dev
Genet 1: 325-330

Schroeder II, Hagiwara S (1990) Repetitive increases in cytosolic
Ca2+ of guard cells by abscisic acid activation of nonselective
Ca"+ permeable channels. Proc Natl Acad Sci USA 87: 9305-
9309

Skriver K, MundyI (1990) Gene expression in response to abscisic
acid and osmotic stress. Plant Cell 2: 503-512

 

Sin:

12

Sindhu RK, Griffin DH, Walton DC (1990) Abscisic aldehyde is an
intermediate in the enzymatic conversion of xanthoxin to

abscisic acid in Phaseolus vulgan's L. leaves. Plant Physiol 93:
689-694

Sindhu RK, Walton DC (1988) Xanthoxin metabolism in cell-free
preparations from wild type and wilty mutants of tomato. Plant
Physi0188: 178-182

Tal M, Nevo Y (1973) Abnormal stomatal behavior and root
resistance, and hormonal imbalance in three wilty mutants of
tomato. Biochem Genet 8: 291-300

Taylor HF, Burden RS (1973) Preparation and metabolism of 2—[”C]-
cis,trans-xanthoxin. I Exp Bot 24: 873-880

Taylor IB, Linforth RST, Al-Naieb RI, Bowman WR, Marples BA (1988)
The wilty mutants flacca and sitiens are impaired in the
oxidation of ABA-aldehyde to ABA. Plant Cell Environ 11: 739-
745

Wang TL, Donkin ME, Martin ES (1984) The physiology of a wilty
pea: abscisic acid production under water stress. I Exp Bot
35: 1222-1232

Walker-Simmons M, Kudra DA, Warner RL (1989) Reduced
accumulation of ABA during water stress in a molybdenum
cofactor mutant of barley. Plant Physiol 90: 728-733

Zeevaart IAD, Creelrnan RA (1988) Metabolism and physiology of
abscisic acid. Annu Rev Plant Physiol Plant Mol Biol 39: 439-
473

Zeevaart IAD, Heath TG, Gage DA (1989) Evidence for a universal
pathway of abscisic acid biosynthesis in higher plants from 18O
incorporation patterns. Plant Physiol 91: 1594-1601

CHAPTER 2
THE IRON CHELATORS a,a-DIPYRIDYL AND o-PI-IENANTHROLINE,
AND THE GROWTH RETARDANT FC-QO'Z,

INHIBIT ABSCISIC ACID BIOSYNTHESIS

13

1 4
2.1. INTRODUCTION

Vifith the exception of fluridone and norflurazon, which inhibit
the early desaturation steps of carotenoid biosynthesis (Boger and
Sandmann, 1989) and thereby inhibit the indirect pathway of ABA
biosynthesis by decreasing the precursor pool of xanthophylls
(Henson, 1984; Quarrie and Lister, 1984; Moore and Smith, 1985;
Gamble and Mullet, 1986; Li and Walton, 1987), no inhibitors of ABA
biosynthesis in plants have been discovered. Paclobutrazol, an
inhibitor of cytochrome P-450 enzymes involved in gibberellin and
sterol biosynthesis (Hedden, 1983; Benveniste, 1986), also inhibits
ABA biosynthesis in the fungus Cercospora rosicoIa (Norman et a1.,
1986) but only marginally in apple leaves (Wang et a1., 1987).

The present study was undertaken to determine if the iron
chelators, a,a-dipyridyl and o-phenanthroline (Figure 2.1), which
inhibit dioxygenases involved in proline hydroxylation (Holleman,
1967; Chrispeels, 1970) and gibberellin biosynthesis (I-Iedden and
Graebe, 1982; Hedden, 1983), could also inhibit ABA biosynthesis.
This would support the hypothesis that a dioxygenase is involved in

cleavage of epoxy-carotenoids during ABA biosynthesis (Creelrnan

15

O
D

Figure 2.1. Structures of a,a-dipyridyl (A), o-phenanthroline (B), LAB

173711 (C), FC-907 (D).

16
et a1., 1987). A second hypothesis was tested, w’z, that ABA

structural analogs can inhibit ABA biosynthesis by competitive
binding to biosynthetic enzymes, or by a feedback mechanism

(Milborrow, 1978).

2.2. MATERIALS AND METHODS

Xanthium struman’um, L., Chicago strain, was grown in a
greenhouse under a photoperiod of 20 h to keep the plants in the
vegetative state. Growing conditions were as described (Raschke
and Zeevaart, 1976). Leaves (average weight 5 g) with petiole were
detached using a razor blade and immediately immersed in water for
10 min. Stock solutions of compounds were diluted to various
concentrations in 3 mL of water and taken up by leaves through the
petioles under dim fluorescent light for 1.5 h. After the solutions
were taken up by the leaves, an additional 2 mL of water were taken
up for 1 h in darkness, then the blades were detached from the
petioles and water-stressed with a hair dryer until 12% of the fresh
weight was lost. Leaves were placed in plastic bags and incubated

in the dark at room temperature for 4.5 h and frozen.

17
If PA was to be analyzed, the leaf was bisected through the

midrib and one-half of the leaf frozen, while the other half was
rehydrated in distilled water for 5 min and incubated in the dark for
an additional 4 h and then frozen. ABA and PA were extracted,
purified and quantified as described (Zeevaart, 1980). The
compound LAB 173711 was the gift of Dr. I. Iung, BASF, D-6703
Lirnburgerhof, Germany. The compound FC-907 (N,N,N,-tr1methyl-l-
methyl-2’,6’,6’-trimethylcyclohex-2’-en-1’-yl) prop-2-enylarnmonium
iodide) was the gift of Dr. Y. Kamuro, Tsukuba Research
Laboratories, Ibaraki 300-26, Iapan. The iron chelators, a,a-dipyridyl
and o-phenanthroline, were purchased from Aldrich Chemical,

Milwaukee, WI.

2.3. RESULTS AND DISCUSSION

2.3.1. Feeding Studies with Iron Chelators

The effects of various concentrations of a,a-dipyridyl on ABA
accumulation in water-stressed Xanthi'um leaves are presented in
Figure 2.2. There was a clear inhibition of ABA accumulation at high

(3.2 mM) concentrations of a,a-dipyridyl. Figure 2.3 presents the

 

:—

‘p— I- C

A.“

ADA l

18

 

J

,j +HZOControl

ABA (ug/gtrwt)

 

 

 

 

 

IIII
a I I I I IIII' I I I I IIIII I I I I IIII' I I 1 I

.001 .01 .l l 10
Concentratlon of applied (,ac-dlpyrldyl (M, x 10 ' )

Figure 2.2. The effect of various a,a-dipyridyl concentrations on ABA
accumulation in Xanthi'um leaves. A leaf (approximately 5 g fr wt)
was fed 3 mL of water or a solution of a,a-dipyridyl, and was then

water-stressed for 4.5 h.

19

 

+ H 2 0 Control

ABA (pg/gtrwt)

 

 

I I IIII' I I I TIIII

 

1 I I I IIIII' I I I IIIII'

.001 .01 .1 1 10
-3
Applied o -Phenanthrollne Concentratlon ( M, x 10 )

Figure 2.3. The effect of various o-phenanthroline concentrations on
ABA accumulation in Xanthi'um leaves. Experimental conditions as

described in Figure 2.2.

20

results of a similar experiment with o-phenanthroline. Because there
are no reports in the literature of heme mono-oxygenase inhibition
by a,a-dipyridyl, we interpret these results as evidence for the
involvement of a non-heme dioxygenase in ABA biosynthesis. Iron
deficiency does not affect the synthesis of Violaxanthin (Morales et
a1., 1990), and carbon monoxide, an inhibitor of heme
monooxygenases, does not inhibit zeaxanthin epoxidation (data not
shown). Therefore, the most likely ABA biosynthetic step which
would involve a dioxygenase would be the cleavage step, where 9-
cis-violaxanthin or 9’-cis-neoxanthin is cleaved to xanthoxin and a
Czs-apocarotenal with insertion of 02. This hypothesis is analogous
to the cleavage enzyme which synthesizes vitamin A (retinal) from B-
carotene in animals (Ganguly and Sastry, 1985) and B-carotene
oxidase of Microcysti's (Iiittner and Hoflacher, 1985). Evidence in
support of the cleavage enzyme hypothesis, aside from the
circumstantial evidence for the indirect pathway of ABA biosynthesis
(Li and Walton, 1990; Parry et a1., 1990, Chapter 5), is the
incorporation of 1"0 from “’02 into the side chain carboxyl group of
ABA during stress-induced ABA biosynthesis (Creelrnan et a1., 1987),

and the observation that carbon monoxide, a specific inhibitor of

21

heme monooxygenases, does not inhibit stress-induced ABA
biosynthesis (data not shown; see Chapter 4).

It has been shown that a,a-dipyridyl causes increases in
seedling hypocotyl elongation rates (Barnett, 1970; Lang, 1976). It
was preposed that hydroxyproline synthesis in the cell wall might
play a causal role in cell wall extensibility and elongation. However,
Lang (1976) showed that light-induced inhibition of radish hypocotyl
elongation did not result in increased hydroxyproline deposition in
the cell wall and suggested that the effect of a,a-dipyridy1 on
elongation may be unrelated to its inhibition of hydroxyproline
biosynthesis. In light of the results of Figures 2.2 and 2.3, which
show inhibition of ABA biosynthesis by a,a-dipyridyl and o-
phenanthroline, we hypothesize that inhibition of ABA biosynthesis by
iron chelators may result in stimulation of hypocotyl elongation.
Bensen eta]. (1988) have shown that increased levels of ABA result
in inhibition of soybean hypocotyl elongation; it follows that reduced
levels of ABA associated with inhibition of ABA biosynthesis may
enhance elongation. This hypothesis is supported by experiments
With rice coleoptiles; treatment with fluridone caused an increase in

growth rate (Hoffmann and Kende, 1990). However, this

22

interpretation is complicated by the involvement of ethylene in
hypocotyl elongation. The ethylene-fonning enzyme requires iron
and ascorbate for activity (Ververdis and Iohn, 1991) and is inhibited
in vivo by o-phenanthroline (H. Kende, personal communication).
2.3.2. Feeding Studies with LAB 173711

The effect of various LAB 173711 concentrations on stress-
induced ABA accumulation in Xanthi'um leaves is shown in Figure
2.4. There was little inhibition of ABA accumulation, even at high
(3.2 mM) concentrations of LAB 173711. When leaves were re-
hydrated for 4 h after the water stress treatment, ABA levels
decreased and PA levels increased about 2 fold at all LAB173711
concentrations tested (data not shown). This suggests that LAB
173711 did not inhibit ABA hydroxylation or other metabolic
processes in general. LAB 173711 has been shown to have ABA-like
activity in physiological processes such as stomatal closing and
promotion of senescence (Iung and Grossmann, 1985; Schubert et
a1., 1991). If LAB 173711 has an inhibitory effect on ABA
biosynthesis, it is a subtle one, based on the inconclusive data of
Figure 2.4. A feedback mechanism which down-regulates ABA

biosynthesis at high ABA concentrations has been preposed

23

 

, q + H20 Control

ABA (uglgfw)

 

 

u I I I ITIII' I I I IIIII‘

.01 .l 1 10
Applied LAB 173711 Concentration (M, x 10'3)

 

Figure 2.4. The effect of various LAB 173711 concentrations of ABA
accumulation in Xanthi'um leaves. Experimental conditions as

described in Figure 2.2.

24
(Milborrow, 1978). If LAB 173711 functions as an ABA analog which

can interact with an ABA-receptor, then its effect at high
concentrations may be due to feedback inhibition of ABA
biosynthesis.
2.3.3. Feeding Studies with FC-907

The compound FC 907 (Figure 2.1D) is an onium type plant
growth retardant (Haruta et a1., 1974) similar to other quaternary
ammonium compounds such as AMO-1618 and chlorocholine
chloride (CCC), which inhibit ent-kaurene synthetase activity A
(Hedden, 1983). FC-907 also inhibits gibberellin biosynthesis in fungi
(Hedden et a1., 1977). Figure 2.5 shows the effects of various
concentrations of PC 907 on ABA accumulation in water-stressed
Xanthi‘um leaves. There is an inhibitory effect on ABA biosynthesis at
the highest (3.2 mM) concentration of PC 907. It is unknown whether
the inhibition of ABA accumulation is due to a specific effect on ABA
biosynthesis, or to a more general inhibition of metabolism.

In the inhibition studies described here, the extent of uptake
and the final intracellular concentrations of compounds were
unknown. This may account for the variability in ABA accumulation

at low concentrations of inhibiting compounds compared with

25

 

4 q + HZOControl

ABA(pg/gfrwt)

 

 

II'Ir

10

 

o I I I IIIII' I I I If‘II" I I

.01 .1 1
-3
Applled Fc-907 Concentration ( M. X 10 )

Figure 2.5. The effect of various concentrations of FC-907 on ABA
accumulation in Xanthium leaves. Experimental conditions as

described in Figure 2.2.

26
control leaves which took up water (Figures 2.2-2.5). Ultimately the

mechanism of ABA biosynthesis inhibition can only be resolved in

vitra with purified ABA biosynthetic enzymes.

2.4. LITERATURE CITED

Barnett NM (1970) Dipyridyl-induced cell elongation and inhibition of
cell wall hydroxyproline biosynthesis. Plant Physiol 45: 188-191

Bensen RI, Boyer IS, Mullet IE (1988) Water deficit-induced changes
in abscisic acid, growth, polysomes, and translatable RNA in
soybean hypocotyls. Plant Physiol 88: 289-294.

Benviste P (1986) Sterol biosynthesis. Annu Rev Plant Physiol 37:
275-308

Chrispeels MI (1970) Synthesis and secretion of hydroxyproline-
containing macromolecules in carrots. II. In viva conversion of
peptidyl proline to peptidyl hydroxyproline. Plant Physiol 45:
223-227

Creelrnan RA, Gage DA, Stults IT, Zeevaart IAD (1987) Abscisic acid
biosynthesis in leaves and roots of Xanthium struman'um. Plant
Physiol 85: 726-732

Ganguly I, Sastry PS (1985) Mechanism of conversion of fi-carotene
into vitamin A: central cleavage versus random cleavage. Wld
Rev Nutr Diet 45: 198-220

Haruta H, Yagi H, Iwata T, Tamura S (1974) Syntheses and plant
growth retardant activities of trimethyl-amrnanium compounds
containing a terpenoid moiety. Agr Biol Chem 38: 141-148

27

Hedden P (1983) In vitra metabolism of gibberellins. In A Crozier,
ed, The Biochemistry and Physiology of Gibberellins, Vol. I.
New York, Praeger, pp 99-149

Hedden P, Graebe IE (1982) Cofactor requirements for the soluble
oxidases in the metabolism of the Czo-gibberellins. I Plant
Growth Regul 1: 105-116

Hedden P, Phinney BO, MacMillan I, Sponsel VM (1977) Metabolism
of kaurenoids by Gibberella fujikurai in the presence of the
plant growth retardant, N,N,N,-trimethyl-l-methyl-2’,6’,6’-
nimethylcyclohex-2’-en-l’-y1) prop-2-enylammanium iodide.
Phytochemistry 16: 1913-1917

Henson IE (1984) Inhibition of abscisic acid accumulation in seedling
shoots of pearl millet (Pennisetum amen'canum [L.] Leeke)
following induction of chlorosis by norflurazon. Z
Pflanzenphysiol 114: 35-43

Hoffmann S, Kende H (1990) The role of ABA in the growth of rice
coleoptiles. Plant Physiol 93: S71.

Holleman I (1967) Direct incorporation of hydroxyproline into protein
of sycamore cells incubated at growth-inhibitory levels of
hydroxyproline. Proc Natl Acad Sci USA 57: 50-54

Iung I, Grossmann K (1985) Effectiveness of new terpenoid
derivatives, abscisic acid and its methyl ester on transpiration
and leaf senescence of barley. I Plant Physiol 121: 361-367

Iiittner F, Hofléichler B (1985) Evidence of B-carotene 7,8(7’,8’)
oxygenase (B-cyclocitral, crocetindial generating) in
Micracystis. Arch Microbiol 141: 337-343

Lang W (1976) Biosynthesis of extensin during normal and light-
inhibited elongation of radish hypocotyls. Z Pflanzenphysiol 78:
228-235

Li Y, Walton DC (1987) Xanthophylls and abscisic acid biosynthesis

28

in water-stressed bean leaves. Plant Physiol 85: 910-915

Li Y, Walton DC (1990) Violaxanthin is an abscisic acid precursor in
water-stressed dark-grown bean leaves. Plant Physiol 92: 551-
559

Milborrow BV (1978) Abscisic acid. In DS Letham, PB Goodwin, TIV
Higgins, eds, Phytohorrnones and Related Compounds: A
Comprehensive Treatise, Vol 1. Elsevier/North-Holland, New
York, pp 295-347

Morales F, Abadia A, Abadia I (1990) Characterization of the
xanthOphyll cycle and other photosynthetic pigment changes

induced by iron deficiency in sugarbeet (Beta vulgan’s L.) Plant
Physiol 94: 607-613

Moore R, Smith ID (1984) Growth, gravireSponsiveness and abscisic
acid content of Zea mays seedlings treated with fluridane.
Planta 162: 342-344

Norman SM, Bennett RD, Paling SM, Meier VP, Nelson MD (1986)
Paclobutrazole inhibits abscisic acid biosynthesis in
Cercaspara rasicaIa. Plant Physiol 80: 122-125

Parry AD, Babiano MI, Horgan R (1990) The role of cis-caratenoids
in abscisic acid biosynthesis. Planta 182: 118-128

Quarrie SA. Lister PG (1984) Evidence of plastid control of abscisic
acid accumulation in barley (Hordeum vulgare L.). Z
Pflanzenphysial 114: 295-308

Raschke K, Zeevaart IAD (1976) Abscisic acid content, transpiration,
and stomatal conductance as related to leaf age in plants of
Xanthium strumarium L. Plant Physiol 58: 169-174

Sandmann G, Boger P (1989) Inhibition of carotenoid biosynthesis by
herbicides. In P Bager, G Sandmann, eds, Target Sites of
Herbicide Action. CRC Press, Boca Raton, FL, pp 25-44

 

29

Schubert I, Réser K, Grossmann K, Sauter H, Iung I (1991)
TranSpiration-inhibiting abscisic acid analogs. I Plant Growth
Regul 10: 27-32

Ververdis P, John P (1991) Complete recovery in vitra of ethylene-
forming enzyme activity. Phytochemistry 30: 725-727.

Wang SY, Sun T, Ii ZL, Faust M (1987) Effect of paclobutrazale on
water stress-induced abscisic acid in apple seedling leaves.
Plant Physiol 84: 1051-1054

Zeevaart IAD (1980) Changes in the levels of abscisic acid and its
metabolites in excised leaf blades of Xanthium strumaiium
during and after water stress. Plant Physiol 66: 672-678

 

CHAPTER 3
ABSCISIC (ABA) -ALDEHYDE IS A PRECURSOR TO,
AND 1’, 4’-TRANS-DIOL A CATABOLITE OF,

ABA IN APPLE

30

Pg" 3*,

R... ‘\\
.Ilclx‘»«

 

Plant Physrol. (1990) 93. 915—923
0032-0889/90/93/091 5/09/501 ,00/0

31

Received for publication Decemw 7, 1989
Accepted March 19. 1990

Abscisic (ABM-Aldehyde Is a Precursor to, and
1’,4’-trans-ABA-Diol a Catabolite of, ABA in Apple1

Christopher D. Roclt and Jan A. D. Zeevaart‘

Michigan State University-Department of Energy Plant Research Laboratory, Michigan State University,
East Lansing, Michigan 48824-1312

ABSTRACT

Previous ”0 labeling studies or abscisic acid (ABA) have
shown that apple (Mains domestics Borkh. cv Granny Smith) truits
synthesize a majority of [“OIABA with the label incorporated in
the 1’-hydroxyl position and unlabeled in the carboxyl group (JAB
Zeevaart, TG Heath, DA Gage [1989] Plant Physiol 91: 1594-
1001). it was proposed that exchange of "O in the side chain
with the medium occurred at an aldehyde intermediate stage of
ABA biosynthesis. We have isolated ABA-aldehyde and 134’-
trans-ABA-diol (ABA-trans-diol) from “O-labeled apple lrult tissue
and measured the extent and position or "0 incorporation by
tandem mass spectrometry. "O-Labeiing patterns or ABA-alde-
hyde, ABA-trans-diol, and ABA indicate that ABA-aldehyde is a
precursor to, and ABA-trans-diol a catabolite ot, ABA. Exchange
at “O in the carbonyl ol ABA-aldehyde can be the cause or loss
of “0 from the side chain of [“OIABA. Results at leading exper-
iments with deuterated substrates provide further support for the
precursor-product relationship of ABA-aldehyde _. ABA -o ABA-
trans-diol. The ABA-aldehyde and ABA-trans-diol contents of
fruits and leaves were low. approximately 1 and 0.02 nanograms
per gram fresh weight for ABA-aldehyde and ABA-trans-dloi.
respectlvely. while ABA levels in fruits ranged from 10 to 200
nanograms per gram lresh weight. ABA biosynthesis was about
10-lold lower in fruits than in leaves. In truits, the majority of ABA
was conjugated to n-o-glucopyranosyl abscisate, whereas in
leaves ABA was mainly hydroxylated to pluseic acid. Parallel
pathways tor ABA and trans-ABA biosynthesis and conIugatlon
In lnrlts and leaves are proposed.

 

Evidence now strongly favors the indirect pathway of ABA
biosynthesis from carotenoids. First. carotenoid biosynthetic
mutants of maize are ABA-deficient (l8). and fluridone. a
specific inhibitor of carotenoid biosynthesis. inhibits ABA
biosynthesis (8, l3). Second. heavy oxygen is incorporated
from ”‘0: into the side chain of ABA during stress-induced
biosynthesis. suggesting oxidative cleavage of a large precursor
pool that contains oxygen functions on the ring (6. 7). Third.
Li and Walton (l4) have provided evidence that 9-cis-violax-
anthin and 9’-cis-neoxanthin are precursors to ABA in dark-
grown bean leaves. They showed a reduction in the levels of

’ Supported by the US. Department of Energy under contract DE-
AC02-7bERO-l338. by the National institutes of Health grant
DRR00480 to the MSU-NIH Mass Spectrometry Facility. by Na-
tional Science Foundation grant DMB-8703847 to J. A. D. 2.. and
by a Monsanto Graduate Fellowship to C. D. R.

915

these two xanthophylls during water stress, with a concomi-
tant increase in ABA and its catabolites PA2 and dihydro-
phaseic acid. on a molezmole basis. In contrast to higher
plants. fungi synthesize ABA through a- and 7-ionylidene
intermediates (29).

Xanthoxin is an apocarotenoid and a probable intermediate
in ABA biosynthesis (22. 24). ABA-aldehyde has been postu-
lated as the immediate precursor of ABA because the ABA-
deficient tomato mutants flacca and sitiens are unable to
convert xanthoxin or ABA-aldehyde to ABA in vivo (21, 25)
or in vitro (23). ABA-Irans-diol is an endogenous compound
of plants and has been proposed as both a precursor to (20,
2 l ). and a catabolite of ABA (26).

Zeevaart er al. (30) reported that ripe avocado and apple
fruits. in contrast to stressed leaves. incorporated ”‘0 predom-
inantly into the l'-hydroxyl position of ABA. The apparent
conflict of these results with the indirect pathway was recon-
ciled by hypothesizing an exchange of ”‘0 label in the side
chain carbonyl with water at an aldehyde intermediate stage
of ABA biosynthesis. The present work was undertaken to
determine if ABA-aldehyde is an endogenous compound in
apple and could be the cause of the unusual ABA "'O-labeling
patterns in fruits. We also investigated the role of ABA-trans-
diol in ABA biosynthesis.

MATERIALS AND METHODS
Plant Material

Young. fully expanded apple (Ma/us domestica Borkh. cv
Mutsu) leaves were frozen in liquid N;, or allowed to lose
l2% fresh weight and subsequently incubated in 20%
I"0::80% N: (v/v) for 12 or 24 h as previously described (7).
The cultivars Granny Smith and Mutsu were used in post-
harvcst apple fruit experiments. Immature apple fruits were
approximately 1 l0 (1 post-anthesis. Alternatively, Mutsu ap-
ple fruits stored approximately 150 d under hypobaric con-
ditions (0.05 aim at O'C to prevent ripening) were used.
Cortical tissue (skin and core excised) was frozen or sliced
into wedges and incubated with "‘03 or exposed to air for 48

i Abbreviations: PA. phaseic acid: ABA-GE. B-D-glucopyranosyl
abscisate: ABA-trans-diol. l’-4'-rrans-ABA-diol: EC D. electron cap-
ture detection: El. electron impact: FlD. flame ionization detection:
fr. fresh: M‘. odd electron negative molecular ion: Me-. methyl ester:
MS-MS. GLC-NCl-collisionally activated dissociation-tandem mass
spectroscopy: NCl. negative chemical ionization: PCI. positive chem-
ical ionization: SIM. selected ion monitoring: !-. trans.

GO

32

916 ROCK AND ZEEVAART

h before freezing. 0: consumption and ethylene evolution
were measured as described elsewhere (6. l2).

For feeding experiments. a 10 pH solution of deuterated
substrate (approximately 2 rig/g fresh weight) in 2% (w/v)
KCl plus 0.05% (v/v) Tween 20 (l) was vacuum-infiltrated
and incubated at room temperature in the dark for 24 or 48
h. Control tissue was autoclaved for 20 min before incubation
with substrate for 48 h.

Extraction and Purification of Metabolites

Frozen tissue was extracted overnight at 4 'C in acetone
plus 0.0l % (w/v) 2.6—di-Iert-butyl-4-methylphenol and 0.25%
(v/v) glacial acetic acid. The extract was drawn off and the
tissue homogenized in an additional volume of extraction
solvent with a Polytron (Bn'nkmann Instruments). The ho-
mogenate was filtered and 30 mL 1.0 M phosphate buffer (pH
8.3) was added to the combined fractions. The acetone was
evaporated using a rotary evaporator at 35 'C and the aqueous
extract adjusted to pH 7.2 with 6 N KOH. The extracts were
filtered through a cellulose membrane to remove precipitates.
With leaf extracts. C hl and lipids were removed by partition-
ing one to three times with an equal volume of hexanes.
Measurement of the partition coefficient of an ABA-aldehyde
standard indicated less than 8% loss from the aqueous fraction
by partitioning with one volume of hexanes. and greater than
90% extraction of ABA-aldehyde from the aqueous fraction
with one volume of ethyl acetate or dicthyl ether (data not
shown). The aqueous extract was partitioned three times with
an equal volume of ethyl acetate to give the neutral fraction
containing ABA-aldehyde. The aqueous phase was adjusted
to pH 3.0 with 6 N HCI and partitioned four times with one
volume of ethyl acetate to give the acidic fraction containing
ABA. ABA-trans-diol. PA. and a portion of the ABA-GE (4).
The organic fractions were taken to dryness under vacuum or
N;. and were further purified by HPLC.

ABA. ABA-GE, and PA in the acidic fraction were sepa-
rated by reverse phase HPLC as previously described (6. 30).
ABA-rrans-diol coeluted with ABA and r-ABA from 23.5 to
26.5 min. The ethanol from the eluates was evaporated and
the remaining water was removed by lyophilization. ABA-
aldehyde from leaf neutral fractions was also purified using
the same reverse phase HPLC conditions as described above:
it eluted from 21 to 25 min. The ethanol was evaporated and
an equal volume of 1.0 M phosphate buffer (pH 8.3) was
added; the ABA-aldehyde was then partitioned four times
into equal volumes of diethyl ether and dried under N; for
normal phase HPLC. Apple fruit extracts did not require
reverse phase HPLC for ABA-aldehyde purification.

The ABA plus ABA-trans-diol fraction was methylated with
ethereal diazomethane and purified by normal phase HPLC
with a uPorasil 30 x 0.4 cm column (Waters Associates).
Elution was with a gradient of ethyl acetate in hexanes from
IO to 50% in 20 min at 2.0 mL/min. UV absorbance was
monitored at 270 nm. Me—ABA was collected from l4.9 to
17.6 min. and Me-ABA-rrans-diol from l8.8 to 2i min.

The fraction containing ABA-GE and PA was hydrolyzed
with 2 N NI-LOH for 2 h at 60°C to yield free ABA. The
samples were dried. methylated and purified by the same

Plant Physiol. Vol. 93. 1990

HPLC system as for Me-ABA and Me-ABA-(rans-diol. except
a gradient of ID to 60% ethyl acetate in hexanes in IO min
was used. Me-A BA and Me-t-ABA were collected from 12 to
13.8 min. and Mc-PA from 14.5 to 16.5 min. ABA-aldehyde
was purified using the same conditions. except UV absorbance
was monitored at 282 nm. ABA-aldehyde was collected from
l4.5 to I65 min. ABA-alcohol fractions were also collected
from 18.5 to 21 min.

Quantitications

ABA-aldehyde and ABA-trans-diol were quantified by the
method of isotope dilution (17) using deuterium-labeled in-
ternal standards. [3Hx]ABA-aldehyde was synthesized by ex-
change of the ring hydrogens of ABA-alcohol in l N NaOD
for 2 d. followed by oxidation in chloroform with manganese
oxide N ( 10) to [3H.]ABA-aldehyde. [3H7]ABA-trans-diol was
synthesized by reduction of [thlABA with sodium borodeu-
teride (l I) and purified by reverse phase semipreparative
HPLC. Deuterated standards were quantified by GLC-FID
and GLC-EC D for ABA-aldehyde and Me-ABA-trans-diol,
respectively. using standard curves of the unlabeled com-
pounds. Quantification of ABA-aldehyde and Mc-ABA-Irans-
diol standards was confirmed by UV absorbance in ethanol
at 285 and 262 nm. using a molar extinction coefficient of
25.000 and 20.000. respectively (1 l)(C.D. Rock. unpublished
observations). As determined by GLC-NCl-MS. the [2H6]
ABA-aldehyde internal standard contained trace amounts of
unlabeled ABA-aldehyde and was 56% hexadeutero-labeled.
The [’HleBA-Irans-diol was 60% heptadeutero-labeled with
a trace of unlabeled compound. A calibration curve was made
using various amounts of unlabeled and [3H.]ABA-aldehydc.
A linear NCl-SIM response for m/z = 248 and 254 was
observed when molar ratios of unlabeled ABA-aldehyde to
[:HtlABA-aldehyde were greater than 0.2. indicating that
reliable measurements could be made when deuterated stand-
ard was added to extracts within a factor of five of the
endogenous concentration of compound.

ABA. ABA-GE. and PA were quantified by GLC-ECD and
addition of ’H-labeled standards to correct for losses during
extraction and purification (5). Percent recoveries for ABA
and PA averaged approximately 50%. and typically l5% for
ABA-GE because only a small fraction partitioned into the
organic phase (4). In experiments with apple leaves where the
entire extract was lyophilized instead of partitioned. ABA-GE
recoveries were typically 50%.

GLC-NCl-MS and GLC-NCI-SIM were performed on a
JEOL AX-505H double focusing mass spectrometer equipped
with a Hewlett-Packard 5890 gas chromatograph and a direct
source inlet. The column used was a DB-l capillary (30 m x
0.25 mm. film thickness 1.0 am; or 30 m x 0.326 mm. film
thickness 0.25 um: J&W Scientific. Inc., Rancho Cordova,
CA) injected in splitless mode with He as the carrier gas (flow
rate 2.5 mL/min). GLC conditions were: oven temperature
programmed from 50 to 180 'C at 35 'C/min. followed im-
mediately by a temperature gradient from ISO to 260 'C at

33

ABA. ABA-ALDEHYDE. AND ABA-trans-DIOL IN APPLE 917

 

r = 0.878

3
H

 

czu‘ (nL/girwtlh)
H

°i

 

 

‘T I v

10 2'0 30 40
l‘aOIABA ( ng/g trwt/48 h)

Figure 1. Relationship between the rate of ethylene evolution and
ABA biosyntheSis in MutSU apple trUit. Cortical tissue from post~
harvest apples of varying degrees oi ripeness was incubated 48 h
under "0;. The content of |‘°O)ABA was calculated from ABA levels
measured by GLC-ECD multiplied by the percentage of [“OIABA
ennchment in the same samples measured by GLC-NCI-SIM of M”
Ethylene measurements 2 se.

‘7 1' Y

ID ’C/min. Methane was used as the reagent gas for NCI and
PCI.

MS—MS was performed on a Finnegan TSQ 70 tn'ple quad-
rupole mass spectrometer as described (30) with modifica-
tions. The collision energy in the second quadrupole was 9
eV for ABA-aldehyde. 2 eV for Me-ABA-rrans-diol. and 3 eV
for Me-ABA with an 0; pressure of 0. 19 Pa.

For analysis of mass spectral data. ion intensity values were
normalized by subtracting the naturally abundant "C and '“O
isotope contributions. To express "‘0 label incorporation into
the side chain of ABA and metabolites as a percentage of total
“‘0 incorporation. the relative abundance of M'. (M + 2)".
(M + 4)“. and (M + 6)‘ was measured by GLC-NCI-SIM and
the percent of total ["O]ABA for each M’ calculated. The
percent ”‘0 enrichment of the side chain was measured by
MS-MS. The product of these two measurements for each
respective M' yields the incorporation into the side chain as
a percent of total ”‘0 incorporation.

The amount of 2H-labeled product synthesized in feeding
experiments was calculated from the ratio of deuterated to
endogenous metabolite measured by GLC-SIM and multi-
plied by the amount of the endogenous compound measured
by GLC-EC D of control tissue frozen at the beginning of the
CXperiment. The extent of 2H exchange by [lHlABA-GE
during hydrolysis was small (data not shown). and resulted in
a slight underestimation of [IHlABA-GE quantities in feeding
c"periments.

Chemicals

"01(97-98% enrichment) was purchased from Cambridge
Isotopes Laboratories (Wobum. MA). Synthetic (3:)-ABA-
aldehyde and (2t)-ABA-alcohol were gifts of Dr. M. Soukup.
“Offmann-LaRoche Inc., Basel. Switzerland.

RESULTS
ABA Biosynthesis in Fruits and Leaves

The relationship between ethylene evolution and ABA bio-
synthesis as measured by ”‘O incorporation in postharvest
Mutsu fruit tissue is shown in Figure l. Although the data do
not establish a causal relationship between ethylene produc-
tion and ABA biosynthesis. the two biosynthetic processes are
clearly linked. The postharvest increase in ABA biosynthesis
was related to the ripening stage of the fruits. Each fruit began
its climacteric at a different time. between 5 and 30 d after
harvest. presumably due to the slightly different stage of
ripening of each individual fruit. With apple fruits stored
under hypobaric conditions. ethylene evolution began im-
mediately after removal from storage and thus the fruits were
synchronized with respect to ABA biosynthesis (data not
shown). A correlation between the climacteric and ABA levels
has been reported in avocado fruits (2).

Results on the biosynthesis and catabolism of ABA in fruits
and water-stressed leaves are shown in Table I. Fruit tissue
had a IO-fold lower rate of ABA biosynthesis than stressed
leaves on a fresh weight basis. ABA levels averaged 200 ng/g
fresh weight in these fruits and decreased about 70% during
the labeling period with a concomitant increase in ABA-GE.
In stressed leaves the ABA content was initially 360 ng/g fresh
weight and increased approximately 50% during 24 h under
"‘03. No I-ABA was detected in these samples. although some
postharvest apple fruits did contain significant amounts of (-
ABA. as has been reported (3).

In fruit. [“‘O]ABA was metabolized to ["O]ABA-GE to a
much greater extent than it was to [“‘O]PA. whereas leaves
metabolized a larger percentage of ABA to PA than to ABA-
GE (Table I). As measured by MS—MS. [”‘O]ABA-GE (as Me-
ABA) did not lose ”‘0 from the carboxyl group during puri-
fication or hydrolysis. as evident from the equal "0 enrich-
ment in the side chain of ABA and ABA-GE from the same
samples (data not shown). The 4’-kcto group of ABA ex-
changes with water under basic conditions (9). thus ["0]
ABA-GE 4’-keto label was lost during hydrolysis and the
values of [”‘OlABA-GE in Table I are therefore underesti-
mated. In fruits. relatively more ["Olr-ABA-GE than ["0]

 

Table l. Biosynthesis and Catabolism of ABA in Mutsu Fruit and
Leaves under “0?

Amounts of (“0)ABA and [“Olrnetabolites synthesized were cal-
culated lrom the levels measured by GLC-ECD and mutliplied by the
percent "0 ennohment as determined by GLC-NCl-SIM oi the M-
(Me-ABA, m/z = 280. 282. 284; Me—PA, m/z = 296. 298. 300. 302).
No t-ABA was detected in these samples.

 

 

. Fruit Stressed Leaves
‘GLabeledCompound 48h"02 24".”:
ng/g fresh wt/time
ABA 29.8 t 6.9‘ 313.0 :i: 58.5
PA 0.2 t 0.1 69.4 t 14.6
ABA-GE 0.5 t 0.3 19.8 t 6.1
t—ABA-GE 3.0 :t 1.3 3.0 t 0.7

 

' Average of three experiments 1 se.

 

 

Ttbltll Er
.557 ”a:
Emery 93

39'9“: 3‘? “a

$3.!

.9 -.a
View“: ea
c: . M.

V

I ~ ‘r
v"

N". ,
Gioeti‘

(3.25:3
\‘_
.|-s

.. L071
0‘ . A
if“ "1015‘

34

918 ROCK AND ZEEVAART

 

Table II. Extent of ABA Functional Group “O'Labeiing in Mutsu
Leaf. Immature and Mature Fruit Tissues

Except for column 3. each class of ["‘OIABA molecules is inter-
preted as havrng the majority of label in the 1'-hydroxyl group (30).

Percent oi Total [“OIABA

 

 

 

 

U‘ilabeled in ,.
Tissue_{Treatment carOOxyl One 0 atom n carboxyl
(M+2) (M+4) (M‘2) (M+4) (M+6)'
1 2 3 4 5
Stressed leaf
12 h "02 9.9 Trace 89 4 0.4 0.1
24 h “02 16.4 Trace 81.4 1.5 0.3

Immature fruit. 48 h "0;
42.8 Trace 52 9 0.9 3.3
30.6 0.1 62.6 2.7 4.0
Mature fruit, 48 h "’02

5 d postharvest 61.2 0.3 6.8 31.1 0.3
28 d postharvest 64.8 0.6 5 4 29.2 NO'
150 d storage" 15.1 23.6 17.9 17.8 25.3

 

‘ Not detected ° Stored under hypobanc conditions

 

ABA-GE was synthesized: the converse was true for leaves
(Table I).

In Table II the various classes of labeled ABA from stressed
leaves. immature and mature fruits are expressed as percent
of the total [”‘OlABA. The class of [”‘OJABA containing two
“‘0 atoms in the carboxyl group (30) represented a negligible
fraction of total [”‘OlABA and is not included in Table II. In
all cases examined [”‘O]ABA not labeled in the carboxyl was
completely labeled at the l’-hydroxyl position. Two aspects
of ABA biosynthesis are demonstrated in Table II: the size
and turnover of the precursor pool containing oxygen atoms
on the ring. and the extent of the proposed carbonyl exchange
in an aldehyde which is a precursor to ABA (30). In stressed
leaves. the ABA synthesized over 24 h only slightly depleted
the precursor pool containing the ring oxygens of ABA (ap-
proximately 18% of [”‘O]ABA contained "‘0 on the ring after
24 h: sec columns I + 4 + 5). It can be inferred that there
was almost complete exchange of the side chain ”‘0 in a
doubly-labeled aldehyde intermediate. because the percent of
total [”‘OlABA in column 4 is only one-tenth of that in
column 1 (Table II). These two classes would be identical at
an aldehyde precursor stage before exchange of the side chain
carbonyl oxygen with water.

In immature fruit. at larger percentage of ”‘O-labeled ABA
molecules contained ”‘0 atoms on the ring as compared to
stressed leaves (approximately 35% compared with < l8% for
leaves: combine columns I. 4. and 5). This suggests that in
these fruits the precursor pool containing oxygens on the ring
turned over quickly. As in leaves. the proposed exchange of
the side chain ”‘0 of an aldehyde intermediate before conver-
sion to ABA was almost complete (compare columns 4 and
I). In mature fruit the majority of labeled ABA molecules

contained "0 on the ring. As in the other tissues examined.
"‘0 exchange of an aldehyde precursor was substantial
(Table II).

Plant Physiol. Vol. 93. 1990

ABA-Aldehyde and ABA-trans-Diol in Apple Fruits and
Leaves

ABA-aldehyde and ABA-trans-diol were present in both
fruit and leaf tissues. Figure 2 shows mass spectra of “‘0
labeled ABA-aldehyde and I-ABA-aldehyde isomers separated
by GLC. lsomerization. if it occurs in vivo. is slow because
the two isomers have significantly different “‘0 enrichment.
ABA-aldehyde was synthesized and/or metabolized more
quickly than r-ABA-aldehyde. as evidenced by the greater ”‘0
enrichment of the (is isomer than that of the trans isomer
(Fig. 2). On the other hand. in three ”‘0 labeling experiments.
ABA-trans-diol averaged l0% ”‘0 enrichment. while ABA
from the same extracts averaged 38% ”‘O incorporation.
Therefore. synthesis and/or turnover of the ABA pool was
faster than that of the ABA-Irans-diol pool in these fruits.

The fragmentation pathways by NC I of ABA-aldehyde and

T10 cis A

250/ 252 (magnlication 2.3)

248

Ion currrent

 

10
814 min)

 

 

 

 

 

 

 

 

 

e
0
r:
C
'u
e
a
.o
(
9 2m
,1 2a
a C
I on
I
col 0 .
111
‘0‘ p
230
20‘ r
152 215
o 'JAIJ‘ ' LA...
100 150 200 250

ml!

Figure 2. GLC-NCl-MS of apple fruit ABA-aldehyde synthesized
under "0,. Granny Smith cortical tissue was incubated for 48 h under
“0,. (A) ten chromatogram with profile of total ion current (110). ABA-
aldehyde isomer M“m/z = 248. and “O-ennched (M + 2)“ and (M +
4)”, m]: - 250. 252. (8) Mass spectrum of ABA-aldehyde. showing
"Oenrichment.Rr-9min18s(C)Massspectrumolt-ABA-
aldehyde. Rr 3 10 min.

Hulnllvo abundance

15%

’3

ABA. ABA-ALDEHYDE. AND ABA-frans-DIOL IN APPLE 919

 

 

 

 

 

 

100- A 215

2 d on eno' 24s
GO-i

160

y 230

20-4
121 165 205
O Y A 'l v A. J V v l v v -
100 140 160 220 260

 

 

 

 

 

 

 

 

 

 

e
we.
2 B 215 25°
'8 q .0 CHO
5
a ‘0‘ /
C . 160
ads Loss
3 i chain air-OH
'- O
o; 20‘ 111 23°
.5 “a 165 207 [rm
— I
c U ‘1 1.711 I" Jan-ll L'A L}L ' v v
100 140 160 220 260
100‘ 217
C 252
q
162
‘0‘ 232
20" “1 ‘23 159 202209 219 23‘
lei $2.121 'Ll' i-v .
°ioo no no 220 260

m/z

Figure 3. MS-MS of ABA-aldehyde containing zero. one. or two "0
atoms. MS-MS was performed on the sample of Figure 2 to determine
the position and extent of “O incorporation into ABA-aldehyde. ABA-
aldehyde and t-ABA-aldehyde were not separated by the GLC method
employed. so that spectra are of a mixture of ABA-aldehyde isomers.
(A) Unlabeled ABA-aldehyde wuth M‘ inset; (B) singly "Oolabeled
ABA-aldehydes with fragmentation products inset; (C) doubly “0-
labeled ABA aldehydes.

Me-ABA-trans-diol are analogous to those of Me-ABA (19):
(CD. Rock and T.G. Heath. unpublished observations). Fig-
ure 3 shows MS—MS spectra obtained from parent M' of
ABA—aldehyde containing zero. one or two "‘0 atoms. The
ion at m/z = 230 corresponds to loss of the 1 '-hydroxyl group
as water; m/z = 215 has lost the l’-hydroxyl and a gem-
methyl group from position 08' or 09’ as methanol. Two
other diagnostic ions are m/z - I60. which contains the 4'
keto group. and m/z = lll. which con'esponds to the side
chain. Measurement of isotope intensities of these ions allows
determination of “‘0 enrichment at each oxygen function of
ABA-aldehyde when the parent ions (M + 2)‘ (m/z =2 250)
and (M + 4)“ (m/z = 252) are analyzed by MS-MS.

It is apparent that the side chain carbonyl was only partially
"‘O-labeled in the (M + 2)” and (M + 4)’ species (see m/z -=
Ill and “3: 160 and I62: 2l5. 217 and 2l9). The 1'-
hydroxyl was predominantly "0 labeled in both the (M + 2)"

and (M + 4)’ species (loss of 20 atomic mass units as H:”‘O:
see m/z 230. 232. 234). and the 4’ keto group was partially
labeled in the (M + 2)“ and completely labeled in the (M +
4)" species (see m/z = I62). Thus. the [“‘O]ABA-aldehyde
pool consisted of molecules with “‘0 at each oxygen function.
the majority of the molecules containing ”‘0 in the I'-hy-
droxyl group. The "‘0 incorporation pattern of ABA-aldehyde
is correlated with that of [”‘O]ABA in fruits in which there is
lack of label in the side chain and predominant ”‘O-labeling
ofthe l’-hydroxyl oxygen (Table II).

In model experiments with synthetic ABA-aldehyde it was
established that the carbonyl oxygen of ABA-aldehyde ex-
changes with H30 with a half-life of approximately 9 min (t
2 SE. n = 9). but does not exchange “‘0 carbonyl label when
dissolved in ethyl acetate (data not shown). Rapid workup
and phase partitioning of aqueous extracts from "‘O-labeled
tissue may have allowed retention of some label in the car-
bonyl oxygen of'ABA-aldehyde from ”‘O-labeled samples (Fig.
3. Table III).

In Table III the percent '“Ooenrichment in the side chain is
compared between ABA-aldehyde. ABA. and ABA-trans-diol
extracted from the same sample and analyzed by MS—MS.
The law of mass action supports a precursor role for species
which have a higher specific activity of ”‘0 label than the
product. assuming biosynthetic precursor pools are homoge-
neous and the label does not exchange with the medium. The
second assumption is true for the carboxyl and l’-hydroxyl
groups of ABA (6). but not for the carbonyl of ABA-aldehyde.
which exchanges oxygen with water (see above). The 4’-keto
group of ABA can slowly exchange with water with a half-life
of days (6). ABA-rransdiol can slowly exchange the I’-hy-
droxyl with water. but not the 4'-hydroxyl (27). Because each
compound can exchange some oxygen function with water.
the "‘0 labeled (M + 4)‘ pools are converted to the (M + 2)’
pools for each compound. Thus. the weighted average of side
chain enrichment (factoring in the relative abundance of each
M‘) is presented in Table III. Taking into account the loss of

 

Table Ill. "’0 Incorporation in the Side Chain of ABA-Aldehyde.
ABA-trans-Diol and ABA Extracted from the Same Fruit Sample after
48 h under “‘0;

Percent side chain armament was calculated from intensities of
daughter ions measured by MS—MS and GLCoNCl-SIM of “Gen-
riched M' (Me-ABA m/z = 280. 282. 284; Me-ABA-trans-diol m/z -
282. 284; ABA-aldehyde m/z = 250. 252). ABA-aldehyde was meas-
ured as a mixture of cis- and trans-isomers. Experiments 1 to 4 were
performed with post-harvest Granny Smith cortex tissue. experiment
5 with Mutsu cortex lrom fruits stored under hypobanc conditions.

Percentof"0—LabeledMoleculesCmtmhg

 

 

 

 

ABAeldehydes ABA ABA-tremolo!
1 18.2 10.6 6.1
2 13.6 16.0 4.1
3 10.5 71.8 5.5
4 NA' 42.0 25.5
5 NA 59.4 35.2
' Not analyzed.

 

3A)
l;

K.)

In.

on....: 9 :5-

 

. .. - . E 1..
. ”Q \1
.FJ .i... .1 We
at .9... . a

ll

 

~ » I . 4 III. N4
Q h ?.l U . .1
.li . Id} \.\w n hi-RW N... \ AH
. u ,I. 3 q . Y . 1.). 5.x .\...
rt 1 i4: . a . .
. s 3.: Reid‘ AM

In. H1.»

36

 

 

 

 

 

 

920 ROCK AND ZEEVAART
tram 248
R- — areas - 4 53

'g' _

'5' mt: 254 (Weston 4.5) '

0

r:

2

. mrz 240
6 1 e e
R 1" mln)

Figure 4. GLC-NCl-SIM chromatogram of ABA-aldehyde from Mutsu
fruit wrth [’HGJABA-aldehyde internal standard. The sum of the ABA-
aldehyde plus t-ABA-aldehyde peak areas gave a ratio of unlabeled
to [’HelABA-aldehyde. from which the amount of unknown endoge-
nous ABAaldehyde was calculated.

l"O label from the ABA-aldehyde side chain carbonyl due to
exchange with the medium. a precursor role of ABA-aldehyde
for ABA can be inferred. In one experiment the side chain
label of ABA-aldehyde was actually higher than that of ABA.
In all cases the side chain specific activity of Me-ABA (percent
l“O in side chain of total “‘01 was greater than that of Me-
ABA-rrans-diol. Thus. these data suggest that ABA is a pre-
cursor to ABA-rrans-diol.

Quantification of ABA-Aldehyde and ABA-trans-Diol

The method of isotope dilution of deuterated internal stand-
ardsl 17) was used to quantify ABA-aldehyde and ABA-Irani-
diol. Retention and stability of deuterium label during extrac-
tion and purification procedures were confirmed by dissolving
deuterium-labeled compounds in 0.2 M phosphate buffers.
ranging from pH 3.5 to 10. for 6 d at room temperature.
followed by partitioning into ethyl acetate and analysis by
GLC-MS. There was less than 1% exchange of deuterium for
both standards over the range of pHs tested (data not shown).
Thus. the internal standards could be used for quantification.
provided the standards were added at concentrations less than
five times the endogenous sample concentrations to minimize
contamination of the nominal mass peak by unlabeled inter-
nal standard (see “Materials and Methods”).

Figure 4 shows a GLC-NCI-SIM chromatogram for quan-
tification of ABA-aldehyde. The deuterated material eluted
from the GLC column about 2 s earlier than the unlabeled
compound. The ratio of ABA-aldehyde to t-ABA-aldehyde as
measured by NCI-SIM was essentially constant. regardless of
the actual quantities of isomers present in a stock solution.
This suggested that the two isomers were not equally detected
by this method. Consequently. both isomers were then ana-
lyzed by various GLC detection methods. In Table IV the
phenomenon of ‘differential ionization’ of ABA-aldehyde iso-
mers is illustrated. FID generates a signal based on the mass
and organic composition of the compound being measured
“and is unbiased in the response toward different isomers of

Plant PhySiol. Vol. 93, 1990

similar or identical compounds. Positive ion detection meth-
ods (PCI and El) generally reflected the abundance of ABA-
aldehyde versus t-ABA-aldehyde isomers. but were relatively
less sensitive to ABA-aldehyde than FID. NCI and ECD
consistently gave ABA-aldehyde to I-ABA-aldehyde isomer
ratios which were less than unity. indicating that the two
isomers were differentially ionized. The degree of ABA-alde-
hyde ionization was affected by the pressure of reagent gas
(Table IV). The mechanism of this differential ionization is
unknown. Thermal isomerization of ABA-aldehyde in the
injection port of the gas chromatograph was ruled out as a
possible cause of observed isomer ratios (see Fig. 3). Xan-
thoxin and I-xanthoxin ratios were relatively constant regard-
less of the detection method employed (Table IV). even
though the chemical structure is quite similar to that of ABA-
aldehyde. NC I was the only mass spectroscopy method tested
which was sensitive enough to measure the small amounts of
ABA-aldehyde present in plants. It was apparent that ABA-
aldehyde isomers could not be quantified by NCI-SIM using
a deuterated internal standard because the cis isomer was
refractory to ionization. The relative amount of cis isomer
was reflected in the NC l-MS signal at high reagent gas pressure
(Table IV). Because the ratio of cis- to trans-ABA-aldehyde
isomers in the deuterated internal standard was known. and
the endogenous isomer ratio was similar (Fig. 4). the sum of
the two isomer concentrations was used as a measure of total
ABA-aldehydes.

In Table V the levels of ABA. ABA-truns-diol and ABA-
aldehyde plus I-A BA-aldehyde in leaves and apple fruit tissue
are presented. There was very little ABA-trans-diol in either
leaves or fruit. and no ABA-alcohol was detected. ABA-
aldehyde levels were low in all tissues examined. When ABA-
aldchyde concentrations are presented as percent of ABA
levels. there is a trend toward relatively higher ABA-aldehyde
levels from leaves to mature fruits (Table V).

Feeding Studies with Deuterated Substrates

Table VI summarizes feeding experiments addressing the
biosynthetic relationship of putative ABA precursors. [2H5]
ABA-aldehyde and and [szlt-ABA-aldehyde were converted
to [3H..]ABA and [’Hxli-ABA. which were further converted

 

Table IV. Comparison of cis- and trans-ABA-Aldehyde and
-Xanthoxin Isomer Ratios by Various GLC Detection Methods

An aliquot of a stock solution was chromatographed by GLC using
a DB-l capillary column and the isomers quantified by peak areas of
the detector responses. For mass spectrometric detection methods.
the total ion current was integrated.

Ratio of cis to trans Isomer Areas

 

 

 

Sample
FID PCI El NCILW' NClmu‘ ECD
ABA-aldehyde 55.4 10.9 2.3 0.5 0.9 1.1
cis- and trans-ABA-aldehydes 0.6 0.2 0.2 0.2 0.3 0.2
[’H.]ABA-aldehydes 5.1 1.1 1.0 0.2 0.4 0.6
cis- and trans-xanthoxins 0.3 0.3 0.4 0.3 0.3 0.4

 

'Methane chemical ionization gas pressure was 6.7 x10“ and
2.0 x 10" Pa for 'low' and 'high' measurements. respectively.

 

:N'H‘

..
11' "‘-

1“

.l
UE. ll.
Wigs

37

ABA. ABA-ALDEHYDE. AND ABA-trans—DIOL lN APPLE 921

 

Table V. Quantification of ABA. ABA -trans Dial and crs- plus trans ABA-Aldehydes in Mutsu Fruit and
Leaves

ABA was quantified by GC-ECD usmg [3H|ABA as an internal standard. ABA-aldehydes and ABA-
trans-diol were quantified by isotope dilution using 2H-labeled synthetic standards and measurement by
GLC-NCI-SIM of the M‘ as shown in Figure 4.

 

 

 

Sample
ABA-Aldehydes
Tissue __
ABA ABA-trans-diol ABA‘a‘dehyoes ABA x ‘00
(as + trans)

ng/g tr wt %
Unstressed leaf 362.80‘ 0.02 1.02 0.3
Immature fruut‘ 19.85 nx" 0.44 2.2
Mature fruitc 5.12 0.01 0.68 13.3

 

‘ Average of two experiments.

° Not analyzed.

C Average of four experiments.

 

to [3HtlABA-GE and [=H.]t-ABA-GE. There was no evidence
for isomerization of ABA to t-ABA (Table VI. [3H~]ABA-
trum-diol feed). or ABA-GE to t-ABA-GE ([3H.]ABA feed)
during the incubation. These data can be interpreted as show-
ing a direct conversion of ABA-aldehyde to ABA to ABA-
GE. and of t-ABA-aldehyde to r-ABA to r-ABA-GE. The rate
of metabolism of I-ABA to I-ABA-GE appears to be greater
than that of ABA to ABA-GE. as indicated by the lower levels
of I-ABA relative to ABA and higher levels of t-A BA-GE than
of ABA-GE. A higher rate of conjugation for t-ABA than for
ABA has been reported for other plants (16). The (-) enan-
tiomer of ABA is conjugated readily (16): thus. a significant
proportion of the [iHlABA-GE product would be (-)-ABA-
GE. if (—)-ABA-aldchyde was converted to (—)-ABA. The
conclusions reached for conversion of [BthABA-aldehyde to
ABA and ABA-GE isomers also hold for the [IHxlABA-

alcohol feeding experiment. Assuming the applied com-
pounds entered the cell to equal extents and were not con-
verted to products other than listed in Table VI. it can be
concluded that the rate of metabolism of [ZHolABA-alcohol
was slower than that of [3H6]ABA-aldehyde.

Conversion of [3H71ABA-trans-diol to ABA took place:
however. this oxidation occurred nonenzymatically as evi-
denced by the presence of [3H.]ABA in the autoclaved control
feed (see also ref. 27). [3H.]ABA was enzymatically reduced
to [3H.]ABA-trans-diol (Table VI). The extent of this conver-
sion was dependent on the concentration of ABA. When [3H6]
ABA was fed at a concentration approximately 400«fold
higher than the endogenous ABA level. the [:HblABA-trans-
diol produced was about 50 times the endogenous ABA-trans-
diol content: quantifications of ABA-trans-diol also support
a concentration dependence on ABA (Table V).

 

Table VI. Metabolism of Deuterated Compounds Fed to Mutsu Frurt Tissue

Plugs of cortex tissue (1.4 cm diameter) were vacuum-infiltrated wuth 2 mL of a 10 pH solution of substrate (except [’HylABA-trans-diol. 2.0
pH). The substrates [’HalABA-aldehyde and [2H61ABA-alcohol contained a significant amount of the respective trans isomer. The control tissue

was autoclaved before addition of substrate for the 48 h incubation period.

 

 

 

Product
Substrate
[’HIABA [’Hlt-ABA [’HIABA-trans-diol FHIAeA—GE [’Hy-ABA-GE
ng/g fresh art/time

[’H.]ABA-aldehyde

24 h 1.6 0.2 no 2.6

48 h 1.4 1.1 1.6 8.0

Control no' no no no
[’H.]ABA-alcohol

24 h 0.2 no no 0.3

48 h 0.8 no no 3.4

Control no no no no
[’HilABA-trans-diol

24 h 7.1 no 0.2 0.1

48 h 5.3 no 3.2 0.1

Control 12.5 no no no
MW

24 h 0.4 2.0 no

48 h 0.6 17.3 no

Control no no no

 

 

; Not detected.

 

.5.

311

BMW
iCataOoiit

Ehc‘r'k:
$1.17“ 1;
1331 ll“. 3::
75:3: :01.

is: .2: t:

:L‘:§ hit... ‘1
.‘ I n

ii‘51 IN
“‘5 i“ iv;

31' x "t
I, H

”W
“ifl,3nc
337?: 3

3’. .
i.

5‘

H

38

922 ROCK AND ZEEVAART

DISCUSSION

ABA-Aldehyde as a Precursor to. and ABA-trans-Diol as
a Catabolite of ABA

Evidence is presented that ABA-aldehyde is the immediate
precursor to ABA. and that ABA-trans-diol is a catabolite of
ABA in apple fruits and leaves. and possibly in all plants.
These conclusions are based on the following evidence. (a)
Assuming the labeled and unlabeled metabolite pools were in
equilibrium. there was greater flux through the A BA-aldehyde
pool than the l-ABA-aldehyde pool (Fig. 2). These data are
consistent with an enzymatic specificity toward the (is isomer
of ABA. the biologically active isomer in plants (28). There
was also greater flux through the ABA versus ABA-rrans-diol
pools (data not shown). which supports ABA as a precursor
to ABA-trans-diol. (b) Precursor-product relationships were
established by measurement of ”‘0 specific activity of metab-
olites extracted from I"O-labeled tissues (Table III). and by
feeding studies using deuterated substrates (Table VI). (c)
There were low endogenous levels of ABA-aldehyde and A BA-
trans-diol (Table V) and a concentration dependence of ABA-
Irans-diol on ABA.

ABA-aldehyde has been postulated as a precursor to ABA.
because the ABA-deficient ~flaa‘u and srtiens mutants of to-
mato are unable to convert xanthoxin or ABA-aldehyde to
ABA in vivo (21. 25) or in vitro (23). Taylor et al. (15. 25)
showed that these mutants reduce ABA-aldehyde to t-ABA-
alcohol. which accumulates to high levels. [:HxlABA-alcohol
is converted to ABA in apple fruits. but at lower rates than
[3Hx]ABA-aldehyde (Table VI).

ABA Biosynthesis in Apple Fruits and Leaves

The qualitative and quantitative data obtained by MS-MS
of metabolites from '“OJabeled tissues provide a means to
analyze the ABA biosynthetic pathway. especially rates of
turnover of precursor pools. Two conclusions can be drawn
from Table II regarding the precursor(s) of ABA. (a) Based on
ABA biosynthetic capacity of the tissues (Table I). and the
turnover rate of the pools as judged by ”‘0 incorporation on
the ring of ABA (Table II). the precursor pool containing ring
oxygens is large in leaves and much smaller in fruits. (b) In
apple. and perhaps in all plants. a significant amount of
exchange occurs between the medium and the carbonyl side
chain oxygen of an aldehyde precursor to ABA. This carbonyl
exchange would give rise to unlabeled ABA synthesized under
"‘0; in tissues which have a large precursor pool. ABA labeled
exclusively in the ring oxygens is observed only in tissues (like
aDole fruit) which have a small. rapidly depleted violaxanthin
Precursor pool. This tissue type synthesizes ABA labeled in
the ring oxygen functions. but without label in the carboxyl
group. The exchange can be at the ABA-aldehyde level (Fig
3). but could also occur in an aldehyde precursor prior to
ABA-aldehyde. such as xanthoxin. Parry er a]. (21) observed
a low amount of "‘0 incorporation from “‘0: into xanthoxin
"1 tomato leaves and suggested that lack of steady state ”‘0
enl'iChment might be due to carbonyl exchange.
ABA biosynthesis and catabolism in climacteric fruits is
under developmental control. in contrast to leaves where

Plant Physiol. Vol. 93. 1990

turgor pressure is a regulatory signal for ABA metabolism
( 29). The regulation of ABA biosynthesis and catabolism is
difTerent in leaves and fruits: hydroxylation of ABA to PA is
a minor pathway in fruits (Table I). but the major catabolic
pathway in leaves (Table I) (29). Fruits synthesize relatively
more of the trans than the cis isomer of ABA-GE: in contrast.
leaves synthesize more ABA-GE than t-ABA-GE (Table I).
There is no evidence to support enzymatic isomerization of
ABA to I-ABA. or of ABAGE to («ABAoGE (Table VI) (16).
We hypothesize that ABA and r-ABA biosynthesis and me-
tabolism proceed in parallel from xanthophylls to xanthoxin
isomers. through A BA-aldehyde isomers. to the corresponding
isomers of ABA and ABA~GE. Both cis- and trans-xanthoxin
isomers are present in plants. and t-xanthoxin is converted to
t-ABA ( 16. 21. 22). Sindhu and Walton (23) showed that bean
leaf extracts convert ABA-aldehyde and t-ABA-aldehyde pre-
dominantly to ABA and l-ABA. respectively. If the levels of
ABA-aldehyde plus t-ABA-aldehyde are similar in leaves and
fruits. but ABA levels are much lower in fruits (Table V),
then the latter should contain more t-ABA-aldehyde than
leaves. The relatively high I-ABA-GE levels in fruits are
evidence in support of this hypothesis. The large amounts of
I-ABA found in apple fruits (3) may not have been due to
isomerization as proposed. but by a loss of ABAoconjugating
activity in these apples. A precise measurement of ABA-
aldehyde and I-ABA-aldehyde levels would provide further
insight into synthesis of ABA and t-ABA in plants.

ACKNOWLEDGMENTS

We thank T.G. Heath and Dr. D.A. Gage for assistance with mass
spectrometry. Dr. D.R. Dilley and Dr. F.G. Dennis Jr. for providing
plant material. and Dr. M. Soukup. Hoffmann-LaRoche. Inc. for
providing synthetic ABA-alcohol and ABA-aldehyde.

LITERATURE CITED

1. Adams DO. Yang SF ( I979) Ethylene biosynthesis: identification
of I-aminocyclopropane-l-carboxylic acid as an intermediate
in the conversion of methionine to ethylene. Proc Natl Acad
Sci USA 76: I70-l74
. Adato I. Gazit S. BIumenfeld A (1976) Relationship between
changes in abscisic acid and ethylene production during rip-
ening of avocado fruits. Aust J Plant Physiol 3: 555—558
3. Bangertli F ( 1982) Changes in the ratio of (is-trans to trans-trans
abscisrc acid during ripening of apple fruits. Planta I55: 199-
203

4. Boyer GL. Zeevaart JAD(1982) Isolation and quantitation of)?-
D—glucopyranosyl abscrsate from leaves of A'anthiiim and spin-
ach. Plant Physiol '70: 227-231

5. Cornish K. Zeevaart JAD (I984) Abscisic acid metabolism in
relation to water stress and leaf age in Xanthium .rtrumarmm.
Plant Physiol 76: 1029—1035

6. Creelmari RA. Cage DA. Stults JT. Zeevaart JAD (1987) Ab-
scisic acid biosynthesis in leaves and roots of Xanlhium smi-
marmm. Plant Physiol 85: 726-732

7. Creelman RA. Zeevaart JAD (I984) Incorporation of oxygen
into abscisic acid and phaseic acid from molecular oxygen.
Plant Physiol 75: 166-169

8. Gamble PE. Mullet JE (1986) Inhibition of carotenoid accu-
mulation and abscisic acid biosynthesis in fluridoncotrcated
dark-grown barley. EurJ Biochem 160: 117-121

9. Gray RT. Mallaby R. Rybaclt G. Williams VP (1974) Mass
spectra of methyl abscisate and isotopically labeled analogs. J
Chem Soc Perkins Trans 2: 919-924

Id

limit! I

”in. \.
.‘ _\

: krlldr l

D-

\

ll“,

411).“

; lla“'nn

. ‘ilif.;p'

. \lll.\ll
" \t’ll.\l_

‘ \r'ljf};

‘ .

"Minn"

W.

In.

ABA. ABA-ALDEHYDE. AND ABA-rrans-DIOL IN APPLE

. Critter RJ. \\ allace TJ ( I959) The manganese dlt)\lLlL‘ oxidation

ol‘allylic alcohols. J Org Chem 24: lllfil-ltlfio

. Hirai N. Okamoto .\|. Koshimizu K ( I986) The l'.-I'-Iruny-diol

ol‘abscisic acid. a possible precursor ol abscisic acid in Built/M
tunvru.PhytochenNsuw ZSZIBhS—IXhh

. kende II. Hanson AD (I975) Relationship between ethylene

evolution and senescence in morning-glory Iloyyer tissue. Plant
Physiol 57: 523—527

. Li Y. \\ alton DC ( I987) \anthophylls and abscisic aCid biosyn-

thesis in water-stressed bean leaves. Plant Physiol 85: 910-915

. Li I. \\ alton DC (I990) Violaxanthin is an ABA precursor in

water-stressed dark-grown bean leaves. Plant Physiol 92: 551-
559

. Linforth RST. Bowman \I'R. Griffin DA. Marples BA. Taylor

I8 (1987) 2-Iruny-ABA-alcohol accumulation in the mill} to-
mato mutants ”um: and ynienv. Plant Cell Environ II) 599—
606

. Milborrow BV I I983) Pathvyays to and from abscisic and, In FT

Addicott. ed. Abscrsic Acid. Praeger. \(‘yy \ ork. pp 7e-) 1 |

. Neill SJ. Horgan R ( I987) -\l".SL‘lSlL‘ acid and related compounds.

In I. Rivier. A Crozier. eds. Principles and Practice of Plant
Hormone Analysrs. Vol I. Academic. London. pp 1 l l—lo7

. .\ei|l SJ. Horgan R. Parry .\D( I986) The carotenoid and abscisic

acid content of viviparous kernels and seedlings of [ca mars
L. Planta I69: 87-96
Netting AG. Milborrow BY. Vaughan GT. Lidgard R0 (I988)
The fragmentation of methyl abscisate and its 215 isomer in
methane positive and negative chemical ionization mass spec-
trometry. Biomed Environ Mass Spectrom I5: 375-389
Okamoto .\I. Hirai N. koshimizu K (I987) Occurrence and

39

30.

923

metabolism of I '.4’-trunv-diol ofabscisic acid. Phytochemistry
26:|369~l:7l

. Parry AD. Neill SJ. Horgan R (I988) Xanthoxin levels and

metabolism in the wild-type and wilty mutants of tomato.
Planta I73: 397-404

Sindhu RK. Walton DC (I987) Conversion of xanthoxin to
abscisic acid by cell-free preparations from bean leaves. Plant
Phy'Siol 85: 9l6-92l

. Sindhu RK. Walton DC (I988) Xanthoxin metabolism in cell-

free preparations from wild type and wilty mutants of tomato.
Plant Physrol 88: ”8482

. Taylor IIF. Burden RS (I972) Xanthoxin. a recently discovered

plant growth inhibitor. Proc R Soc Lond Ser B I80: 3l7-346

'. 'Iaylor IB. Linforth RST. Al-Naieb RJ. Bowman WR. Marples

B.-\ (1988) The wilty mutants (luau and sim'ns are impaired
in the oxidation of ABA-aldehyde to ABA. Plant Cell Environ
Ilt739-745

. Vaughan GT. Milborrow BV ( l987) The occurrence and metab-

olism of the I'.4’-diols of abscisic acid. Aust J Plant Physiol
I4: 593-604

Vaughan GT. Milborrow BY (I988) The stability of the l'.4’.

diols ol‘abscisic acid. Phytochemistry 27: 339443

. Walton DC( 1983) Structure-activity relationships ofabscisic acid

analogs and metabolites. In FT Addicott. ed. AbSClSlC Acid.
Praeger. New York. pp l l3—l46

. Zeevaart JAD. (‘reelman RA ( I988) Metabolism and physiology

ofabscisic acid. Annu Rev Plant Physiol Plant Mol Biol 39:
439—473

Zeevaart JAD. Heath TG. Gage DA (I989) Evidence for a
universal pathway ofabsCisic and biosynthesis in higher plants
from “O incorporation patterns. Plant Physiol 9|: 1594460)

CHAPTER 4
ABSCISIC (ABA) -ALCOHOL IS AN INTERMEDIATE IN ABA

BIOSYNTHESIS IN A SHUNT PATHWAY FROM ABA-ALDEHYDE

40

With l

41
4.1. ABSTRACT

It has previously been shown that the abscisic acid (ABA) -
deficient flacca and sitiens mutants of tomato are impaired in ABA-
aldehyde oxidation and accumulate trans-ABA-alcohol as a result of
the biosynthetic block (EB Taylor et a1. [1988] Plant Cell Environ 11:
739-745). Here we report that the flacca and sitiens mutants
accumulate Hans-ABA and trans-ABA-glucose ester, and that this
accumulation is due to trans-ABA biosynthesis. 18O Labeling of
water-stressed wild type and mutant tomato leaves and analysis of
[‘°O]ABA by tandem mass spectrometry shows that the tomato
mutants synthesize a significant percentage of their ABA and trans-
ABA as [1°OJABA with two 90 atoms in the carboxyl group. The
droopy mutant of potato and the A234 (narZa) molybdenum cofactor
mutant of barley, which are also impaired in ABA-aldehyde oxidation,
also synthesize more doubly-carboxyl-labeled ABA than wild type in
1t’0 labeling experiments. The droopy mutant accumulates t-ABA
and t-ABA-GE like the sitiens mutant. We further show, by feeding
experiments in tomato with FPIdABA-alcohol and "’02, that this
doubly-carboxyl-labeled ABA is synthesized from [1°OJABA-a1cohol

with incorporation of molecular oxygen. In vivo inhibition by carbon

42

monoxide of [21-161ABA-alcohol oxidation establishes the involvement
of a cytochrome P-450 monooxygenase. Likewise, CO inhibits the
synthesis of doubly-carboxyl-labeled ABA in 18O labeling
experiments. This minor shunt pathway from ABA-aldehyde to ABA-
alcohol to ABA Operates in all plants. For the ABA-deficient mutants
impaired in ABA-aldehyde oxidation, this shunt pathway is an

important source of ABA and is physiologically significant.

4.2. INTRODUCTION

Much progress has been made recently on the elucidation of
the biosynthetic pathway of ABA. The evidence is now conclusive
for the "indirect pathway," which proceeds by oxidative cleavage of
epoxycarotenoids to xanthoxin, which is sequentially metabolized to
ABA-aldehyde and ABA. Xanthoxin and ABA-aldehyde are
endogenous compounds of plants (Parry et a1., 1988; 1990b; Rock
and Zeevaart, 1990a) and the oxidase activities which convert these
compounds to ABA have been characterized in vitro (Sindhu and
Walton, 1988; Sindhu et a1., 1990). The viviparous mutants of maize
are blocked in the early stages of carotenoid biosynthesis and are

ABA-deficient (Moore and Smith, 1985; Neill et a1., 1986). Fluridone

d.

01

R:

43

and norfiurazon, specific inhibitors of carotenoid biosynthesis, also
inhibit ABA biosynthesis (Gamble and Mullet, 1986). 18O-Labeling
experiments show incorporation predominantly into the side chain
carboxyl group of ABA, suggesting oxidative cleavage of
xanthophylls which contain oxygens on the ring of ABA (Creelrnan
and Zeevaart, 1984; Creelman et a1., 1987; Zeevaart er a1., 1989). Li
and Walton (1990) showed a 1:1 relationship on a molar basis
between decreases in violaxanthin and 9’-c1's-neoxanthin and
increases in ABA and the catabolites PA and dihydrOphaseic acid in
dark-grown water-stressed bean leaves; similar results have been
obtained by others (Gamble and Mullet, 1986; Parry et a1., 1990a).
Rock and Zeevaart (1991) have characterized aba, the ABA-deficient
mutant of Arabidopsz’s thaliana (Koomneef et a1., 1982) as being
impaired in epoxy-carotenoid biosynthesis, which provides strong
evidence for the indirect pathway of ABA biosynthesis.

There are three non-allelic ABA-deficient mutants of tomato,
not, flc, and sit, which difi'er in their phenotypic severity of
tranSpiration rates and inability to close their stomata (Tal and Nevo,
1973; Neill and Horgan, 1985). Application of ABA can restore the
normal phenotype (Tal and Nevo, 1973; Neill and Horgan, 1985).

Feeding studies and enzymology have established that the flc and sit

44

mutants are blocked in ABA-aldehyde oxidation (Sindu and Walton,
1988, Sindhu et 31., 1990; Taylor et a1., 1988). The A234 (narZa)
molybdenum cofactor mutant of barley (Walker-Simmons et a1.,
1989), the droopy mutant of potato (Quarrie, 1982; Duckham et a1.,
1989), and the CKRI mutant of Nicotjana pIubagz'nifolia (Parry et 31.,
1991) have also been characterized as being blocked in ABA-
aldehyde oxidation.

Because the flc and sit mutants accumulate t-ABA-alcohol as a
result of the ABA-aldehyde oxidase biosynthetic block (Linforth et a1.,
1987, 1990; Taylor et a1., 1988), our observation that the flc and sit
mutants accumulate significant amounts of t-ABA—GE (Rock and
Zeevaart, 1990b) prompted us to investigate whether ABA-alcohol
could be a biosynthetic intermediate between ABA-aldehyde and
ABA. Our results confirm this hypothesis and establish the enzyme
involved in ABA-alcohol oxidation is a P-450 monooxygenase. This
shunt pathway is the source of doubly-carboxyl-labeled ABA

observed in “’02 labeling experiments with numerous species

(Zeevaart et a1., 1989).

4.3

ale

flit

WE

we

)
r,

I":

he

Pb

g)

45
4.3. MATERIALS AND METHODS

4.3.1. Plant Material
Seeds of Lyc0persicon esculentum Mill. cv Rheinlands Ruhm

and the near-isogenic mutants not, flc, and sit were germinated on
filter paper under continuous light and transplanted to pots which
were subirrigated with half-strength Hoagland’s solution. Plants
were grown for six to ten weeks in a high humidity growth chamber
and maintained on a diurnal cycle of 9 hr light (300 i113 - rn'2 ' s"), 23°
C and 15 hr dark, 20° C. Seeds of Hordeum vulgare L. cv Steptoe
and the molybdenum cofactor mutant A234 (narZa) were obtained
from Dr. Robert Warner, Washington State University, Pullman, WA
and grown for 8 d in vermiculite in a greenhouse, as described
(Walker-Simmons et a1., 1989). Seeds of heterozygous (Dr/dr) and
homozygous (dr/dr) dr00py potato (SoIanum tuberosum L., group
Phureja) were from Dr. Ian Taylor, University of Nottingham, UK.
and were grown as described for the tomato mutants.

For ABA biosynthesis studies the leaves were harvested and
water-stressed by dehydrating with a hair dryer until 14% of the fresh

weight was lost. The tissue was incubated under 20% 1‘30, : 80% N2

46

(v/v) (Creelrnan and Zeevaart, 1984) or air at room temperature in
the dark for various lengths of time and frozen in liquid N2.
4.3.2. Feeding Experiments

[2H3J5](:)ABA-alcohol and [2H3J5](:)ABA-aldehyde (97.2%
[w/w]) were synthesized as described (Rock and Zeevaart, 1990a).
Leaves of flc or sit were dipped in a 4 x 10‘5 M solution of substrate
containing 0.05% (v/v) Tween 20 and 0.2 % (v/v) ethanol, and were
immediately placed in an atmosphere containing either 20% 18Oz :
80% N2 (v/v) or air plus 10% or 50% (v/v) carbon monoxide, or N2,
and incubated in darkness for 8 h and frozen.
4.3.3. ABA. ABA-GE, and PA Analysis

ABA and catabolites were extracted with acetone plus 0.025%
(w/v) 2,5-di-tert-butyl-4-methylphenol and 0.25% (v/v) glacial acetic
acid, partitioned at pH 3.0 into ethyl acetate, and purified by reverse
phase HPLC as described (Rock and Zeevaart, 1991). The trans
isomers of ABA and ABA-GE had slightly shorter retention times in
the HPLC system than the respective cis isomers. Samples were
methylated with ethereal diazomethane and quantified by GC-NCI-
SIM as described (Rock and Zeevaart, 1991). For Me-ABA

synthesized from (”Had-labeled compounds, m/z = 278, 283, 284,

47

285, 286, and 288 were monitored. Corrections were made for the
natural abundance of stable isotopes. MS/MS was performed on a
Finnegan TSQ-70 triple-quadrupole mass spectrometer as previously
described (Rock and Zeevaart, 1990a).
4.3.4. Chemicals

1802 (97-98% enrichment) was purchased from Cambridge
IsotOpes Laboratories (Woburn, MA). Carbon monoxide (99.3%
pure) was from Scott Specialty Cases (Houston, TX) and was the gift
of Dr. Robert Creelrnan, Department of Biochemistry and Biophysics,

Texas A 8: M University.
4.4. RESULTS

4.4.1. ABA and Trans-ABA Biosynthesis in the ABA-Deficient Tomato
Mutants

Quantitation of ABA, t-ABA and their catabolites ABA-GE, t-ABA-GE
and PA in unstressed leaves of wild type and mutant tomato is
presented in Table 4.1. Consistent with the work of others (Tal and
New, 1973; Neill and Horgan, 1985; Linforth et a1., 1987), there was a

correlation between ABA levels and resistance to wilting in

48

8:95.830 95 mo magmas?

 

 

Nam 32 H mdm md + mfi m; + md 5 H mdm «no.3»
fiv ad H «.mw md H md md H m." 0d H fimm moon:
«.mm: Qm H New as H o.vm no H m4 mdm H mdmm Mango?
«.mmm Wm H «.mm Nun H mdm «6 H o._ 0.2 H «dam 832m mpg—52m
b: hm \m:
Rm www.mmma mQEE «a? ma mahogany

 

2mm no 55583 8.5 do :82 258200 3 32:89

925 8508308 pam 52 .SN .mC Eugen omgocosa 053805 Co “090 5 68m: 98 $5932

sass: cameos 2283.23:

as... x5. 2.? no 853 83825 23 85336 Ea 58$ .52 3 83285 .3 28¢.

49

unstressed leaves of the mutants; this correlation also holds for the
ABA catabolites ABA-CE and PA (Table 4.1). Furthermore, there is
an inverse correlation between t-ABA and t-ABA-CE accumulation
and phenotypic severity of the mutants (Table 4.1). The least severe
mutant, not, accumulated similar amounts of t—ABA and t-ABA-GE,
taken together, as wild type (Table 4.1). The most severe mutant,
sit, accumulated the most t—ABA and t-ABA-GE, while flc had
intermediate levels between not and sit (Table 4.1). These results
correlate with increased levels of t-ABA-alcohol in the flc and sit
mutants (Linforth et a1., 1987).

18O-Labeling of ABA in vivo and quantitation by MS allows
measurement of biosynthetic capacity of the ABA pathway (Zeevaart
et a1., 1989; Rock and Zeevaart, 1990a, 1991; Rock et a1., 1991).
Table 4.2 shows quantitation of 18O-labeled ABA, t-ABA and
catabolites from water-stressed leaves incubated under “’02 for
various lengths of time. Consistent with the results in Table 4.1,
there was a correlation between phenotypic severity of the mutants
and lower ABA biosynthetic capacity (Table 4.2). Furthermore, there
was a correlation between phenotypic severity of the mutants and

the lack of inhibition of t—ABA biosynthesis, as measured by the ratio

50

 

 

new ad: 2.. S no 22. a 8
mg m8 2 3 3 «d a a
«.5302
.42 3.2 mg 2m 38 a 8.
n5 3% mm 3 a.“ $3 a m
was «.2: 3 3 3 38 a v
553% mflsgsmopa
oEu \ ts h m \ m:
may—203302 «.2536 GEE.
58.; moemfi 8.; 5.550.; 5:: 5.; 528.; 33050
853902 no

 

.Bcofipoouo oz: Co 09222 .Smméoz

-00 .3 noun—82.6 203 mofionflofi 93 a .6m 53 Eugen 3020:0an @5329: ac 390 5 v05: 98 853:2

use... ac 985.2 Bong as No... a 2552 cameos £383

.32 Una OE 25> .«0 82mm..— Uommobméoum? Sod mflaonflmu U5” <m<.§ 4a Uganda—.02 «0 conflusmao .N.v 03w?

51

.voaoouop «OZ;

ooubmsm «oza

 

N;
Nam

Hm.
Wm

m6
md

42
3:
ed

mé
o4

m.m
ad
md

0095

G:

vd

no

3.9.2

Secs «.4 28a.

od
md

md
md
«6

O.“
m.—

0.:
fie
ad

gm
5v
98.2%
5mm
Sm
s¢

«comm

52
of cis-ABA and metabolites to trans-ABA and metabolites (Table

4.2). However, t-ABA and t-ABA-GE biosynthesis, taken together,
were still decreased in the mutants by more than 60 percent (Table
4.2).

The fragmentation of Me-ABA by NCI has been elucidated
(Netting et a1., 1987; Heath et a1., 1990). Analysis of [‘80]ABA by
MS/MS allows measurement of the position and extent of 18O
incorporation into the different oxygen functions of ABA (Zeevaart et
a1., 1989; Rock and Zeevaart, 1990a; Rock et a1., 1991). Three
phenomena involving the ABA biosynthetic pathway have been
characterized by this method. 1) The incorporation of a single 180
atom into the side chain carboxyl group of ABA, with gradual
incorporation, over time, of label into the ring oxygens. This was the
observation by Creelrnan eta]. (1987) which suggested there is a
large ABA precursor pool of xanthophylls that turns over slowly and
already contains the oxygens on the ring. 2) The loss of side chain
carboxyl 1"O label in [‘80]ABA due to carbonyl oxygen exchange
with the medium at an aldehyde (xanthoxin or ABA-aldehyde) stage
of biosynthesis. This exchange is proposed to be the mechanism

which gives rise to ring-labeled [‘QOJABA which is unlabeled in the

53

side chain carboxyl, as seen in apple fruit (Rock and Zeevaart,
1990a). 3) The phenomenon of doubly-carboxyl-labeled [‘90]ABA
which has been observed as a minor biosynthetic product in 18O-
labeling experiments with a number of species (Zeevaart et a1.,
1989).

Figure 4.1 shows MS/MS spectra of [‘BOZJABA (parent ion
m/z=282) fi'om water-stressed wild type, 110 and sit leaves under 1802
for 8 h. Rheinlands Ruhm wild type showed the typical pattern of
labeling; for doubly-labeled ABA the side chain ion (unlabeled m/z=
141) was shifted to m/z = 143, and dehydration of the l’-hydroxyl
group of [1°02]ABA yielded a loss of 20 atomic mass units to m/z =
262. This result shows that incorporation of 18O was predominantly
into the side chain carboxyl and l’-hydroxyl groups. Doubly-
carboxyl-labeled ABA, evident by the presence of ions at m/z = 145,
was a minor product in wild type leaves (Figure 4.1). However, in
[’°OZ]ABA from tie and sit the majority of molecules contained both
180 atoms in the side chain carboxyl group, evident by the
abundance of ions at m/z = 145 (Figure 4.1).

A Table 4.3 presents a quantitative analysis of the labeling

patterns of [1°O]ABA and [‘90] t-ABA from wild type and ABA-deficient

54

Figure 4.1. CC-NCI-MS/MS of [‘BOZJMe-ABA (m/z = 282) from water-
stressed leaves of wild type Rheinlands Ruhm (A), flacca (B), and
sitiens (C) tomato in “302 for 8 h. Inset: the structure of Me-ABA with

asterisks (') showing the predominant position of 18O atoms.

Relative Abundance

100

100

100

55

A. Rheinlands Ruhm wild type

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

' 282
side chain
‘ 143 CO*OCH 3
‘ ring dehydration
q 154 262
rALfi v ~_W f Llr Junk? A
B. Flacca
‘ 282
. 145
. J 152 264
. .r .L+fi f - r . 1%th .
C. Sitiens
‘ 282
" 145
, 152 264
Y Li J; f i v 741% fill:
150 200 250

300

6
5

00:00.00 37?

 

 

8d 8.0 8.2 .3 3. 3.8 so :2 <9:

8.0 we can an E 0.2. «.o 8.: <m<
88......

mod 3 mg: 3 ma :2 no 5 £22

3.0 3 «.2 S 3. m. a 3 8.: 52
8000:

:3 dz 3 dz 3 :8 «a 3. 5::

8.0 .02 3 .02 8.8 q: no 3.. £2
03282

88 no m2 3 S: 8.; «d 2.. 52.2

2: 88 so 3 3 3m 3 5 £2
8%: 882522

a a. \ E 8 8 \ a:
a o m. 4 m a _
52 8.8.8.3 stale. 5 8.8.8.3 83.3 3...: 8&3 85:3 8&3 83880
238289580 o... M8o o... 83 o.. 2:. o.. 8o o: can 9. as... o: 2o

 

33:50 E «:52 O: 03,—.

 

3.8280 2 82¢ o.. 2no

 

38.80 5 8232::

 

525.; .88. no 288..

 

.22 0:8 23.52.00 3 000293 0:03 00353 60200: 33:03.; 05 .8
0233 30:29:00 08 .0 0:8 m 3828 300:0 .525... go momma—0 :1 .SN a: 3:050 0:50:93 0:380:05 mo 30:0 5 330: 0:0 8:83)—

.218

8O... E 2050:. no 3:332 .:0_0a00.<m< 0:0 0&5. 25> .o 3.60.— pommohm .083 E 950384.02 $2.38.: 0:: <3 .0 :oEmom 0:: 32am .m.v 038...

5?

mutant leaves after a 4 h labeling experiment. The seven major
classes of [leOJABA molecules are represented as a percentage of
total [laOJABA (Table 4.3, columns 1-7). This allows a comparison of
the ABA biosynthetic reactions in wild type and mutant tissues.
Several points can be made from the labeling patterns. a) Because
ABA and t-ABA from the same samples can have significantly
different labeling patterns (Table 4.3), the law of mass action
supports the hypothesis that little isomerization of ABA to t-ABA has
occurred in vivo or in vitro. This interpretation is supported by
feeding studies with apple (Rock and Zeevaart, 1990a) and
measurement of ABA isomerization in vitro (Milborrow, 1970). b)
Side chain carbonyl exchange at an aldehyde intermediate stage of
ABA biosynthesis (Rock and Zeevaart, 1990a) was higher in the
mutants than wild type (Table 4.3, compare column 4 versus column
1 between wild type and mutants). In the wild type [’80]ABA pool
there was relatively more doubly-labeled ABA with one 1"0 atom on
the ring and one atom in the carboxyl group (Table 4.3, column 4)
than singly-labeled ABA with label on the ring (Table 4.3, column 1).
This is interpreted as a high ABA-aldehyde oxidase activity which

resulted in carbonyl label retention by rapid oxidation to ABA.

58

Conversely, the mutant ABA labeling patterns suggest there was
more carbonyl oxygen exchange because there was relatively more
singly-labeled ABA which was labeled on the ring (Table 4.3, column
1) than doubly-labeled ABA with one 180 in the carboxyl group and
one on the ring (Table 4.3, compare columns 4 versus 1). This
observation suggests that aldehyde oxidation was slower in the
mutants and is supported by previous work showing the 110 and sit
mutants are impaired in ABA-aldehyde oxidation (Sindhu and Walton,
1988; Sindhu et a1., 1990; Taylor et a1., 1988). c) From the MS/MS
analysis it can be concluded that the mutants synthesized a higher
percentage of [WOJABA which was doubly labeled in the carboxyl
group than wild type leaves (Table 4.3, columns 6 and 7). When this
doubly-carboxyl-labeled ABA was quantified on a gram fresh weight
basis, both wild type and mutant tissues synthesized approximately
equal but small amounts; significantly, 110 and sit synthesized more
doubly-carboxyl-labeled r-ABA than wild type (Table 4.3). This
result correlates with endogenous t-ABA—alcohol levels in wild type
and mutant leaves (Linforth et a1., 1987). Similar patterns of doubly-
carboxyl-labeled ABA were seen in 1802 experiments of longer

duration, and in [1°021ABA-GE and [‘BOSJPA (data not shown).

59
4.4.2. ABA and Trans-ABA Biosynthesis in the ABA-Deficient Potato

and Barley Mutants

In order to obtain supporting evidence for the hypothesis that
ABA-alcohol is an intermediate in ABA biosynthesis, we analyzed the
dr00py mutant of potato (Quarrie, 1982) and the molybdenum
cofactor mutant A234 (narZa) of barley (Walker-Simmons et a1.,
1989). These two mutants have been characterized as impaired in
ABA-aldehyde oxidation (Duckham et a1., 1989; Walker-Simmons et
a1., 1989), and therefore might exhibit similar patterns of ABA
biosynthesis as the flc and sit mutants. Table 4.4 presents ABA, t-
ABA and catabolite levels in unstressed and 8 h water-stressed
leaves of the mutants and wild types (for potato, a pseudo-wild type
Dr/dr heterozygote was used). The droopy potato mutant showed
the same phenotype as the 110 and sit tomato mutants on the basis of
t-ABA and t-ABA-GE accumulation and biosynthesis (Table 4.4). The
A234 molybdenum cofactor mutant of barley did not accumulate t-
ABA or t-ABA-GE (Table 4.4).

Analysis by MS/MS of [‘BOJABA and [1°O]t-ABA from an 8 h
labeling experiment with the potato and barley mutants showed

similarities and differences from the tomato mutant labeling patterns

60

«.mm H mfim «.8 H QC. Nd H a.“ «d H .m.mm a; H 56 0 0 0000005

92 H 9:. Qm H 95 ad H m._ Wm H va m4 H ad 00000000:
0300 0000>NoEom

«.80 H Numom: No + Na o.m H Num— mA H 0.0 gum" H fimmm 0 w 0000.26

mdm H Odom o; H md 0.0 H m.m Nd H ad Nam H m.mm 00000000:
0055 00093000003

 

990m

0000\00

 

0a 0300: 0302 000.. 000 000800 \ .2800

 

0000000098 020 “0 000020 05 00 3000000000000 >035 x00 HV 008000098

0005 Ho 0008 05 00 3000000000000 0.0.00 flaws—4.00 >0 00000000 0003 003003

300% 00003 0 m 0002 000
00ou0m >000m .00 3505 000002 00000000 000003302 05 000 05 0:3 5 000 980m .3020 900066
000960003“ 000 2000on 0000300083 5 005040000 000 5.0.00.0 a no 000050000 .06 0300.

61

0000000900 000..

00000000000 000m.

 

$55
0”:

amfimm
00.2

no.”
2

and
CA

nmd
Nd

amd
ad

0.0.280 3. 030.0.

06
Ho

Hm
Ho

3: g 0 03800
04 000000005
@0005 0m 00
0.0: .0 0 03800
o.m 00000000:
25 E; .8600:
00000

62

(Table 4.5). Both mutants showed increases in the percentage of
doubly-carboxyl-labeled ABA and t-ABA compared to their
respective wild types (Table 4.5, columns 6 and 7). However,
carbonyl exchange at an aldehyde intermediate stage of biosynthesis
did not appear affected in either the droopy potato or A234 MoCo
mutant of barley (Table 4.5, compare columns 4 versus 1 between
wild type and mutant ABA), in contrast with the tomato mutants
(Table 4.3). For t-ABA, carbonyl exchange was high in both wild
type and mutant tissue of potato and barley (Table 4.5, compare
columns 4 versus 1 for t-ABA).
4.4.3. ABA-Alcohol Oxidation in the Hacca and Sitiens Mutants

In order to directly test whether ABA-alcohol is a precursor to
doubly-carboxyl-labeled ABA, we fed [2H36](:)ABA-aldehyde and
[2H36](:)ABA-alcohol to wild type and flc leaves in 1802 for 8 h and
measured by MS/MS the extent of 18O incorporation into [2H351ABA
From the results presented in Table 4.6, it is clear that flc is impaired
in the oxidation of ABA-aldehyde, as previously shown (Sindhu et a1.,
1990; Taylor et a1., 1988). In the [2H3_6](:)ABA-alcohol feed, the high
oxidation rate in wild type was probably due to conversion of ABA-

alcohol to ABA-aldehyde, which occurs non-enzymatically in

63

.02 .02 .02 .02 «dd .d.z ms 552
0000 vd md dd 0.8 0000 ad a
EEG 0000300000:

 

 

 

 

 

 

 

0.08m
N. m w v m .0. _
mlll00 0m m00 00 000 00 000 00 m|||00 0m . 100 00 m.0n010lm
0.: 000 02 0000 02 020 02 000 02 Damn 02 025 02 000
$000000 5 00090 09 020 $000000 5 00000 02 000 3000000 5 000000000
0050.; Be 00 2880 003800 \ 8030

 

.3030036300 05 20202.00 .3 853a
0003 02ng 002000 300033.; 05 00 000000— 30029000 000 d 000 m 0000200 30000 550.; .00 000003 a

0 0 30 0.0... 5 >280 00 $005 000:: 88000 2:53.032 90 Ha SE. was a 05 2000 00800 05036 8028500
000 20000—6 00090000000: do 00>000 000000W0000>> 5 000000402 5.0000 000 a do 00200m 000 0000m— .m.v 00000.

64

00000000 002.

 

.Q.Z

Nd

.Q.Z

Nd

flown

md

ms .02 Hm“ odm .02 NE <93

md Nd 0.9 ado Nd m0. 03¢
«030.6 0m 2

0 d .02 md mdm 0000 md Marl

vd v.0 de Nag. _d v.” <m<

.0900“? 09: 25>
0.00000

md 0000 ad mdm 0d dd <93
Adm 0000 0H 0.00 .02 md 52
0000 000030.00:

0.0.280 0.0 2000.

65

Table 4.6. Ouantitation of Labeled ABA from Wlld Type and Hacca Tomato Leaves

Fed Deuterated ABA-Aldehyde or ABA-Alcohol 8 h in “’02

Unstressed leaves were dipped in a 4 x 10'5 M solution of deuterated

substrate (97.2% [31134) plus 0.05% Tween 20 and 0.2% ethanol (v/v) and

immediately placed in "303 (20 % [v/v] with N2} as remainder). Samples were

quantified by GC-NCI-SIM and 18O incorporation in the carboxyl group confirmed

by MS/MS.

 

Substrate / Genotype

Products

 

FthABA [‘BOZHHJABA

Percent 1“O

in side chain

 

[’HuJABA-aldehyde
Rheinlands Ruhm

Hacca

FHRJABA-alcohol
Rheinlands Ruhm

Ha cca

ng/gfirwt/811

1451.7

11.8

726.5

23.4

9.8

2.5

20.1

12.7

0.7%

17.6%

2.7%

35.3%

66

vitro (CD. Rock, unpublished observations). Leaves of flc converted
a higher percentage of [2H3J51(:+:)ABA-alcohol, than of [2H3J3](i)ABA-
aldehyde, to [‘802H361ABA (Table 4.6), which supports the hypothesis
that ABA-alcohol is converted to ABA by oxidation with molecular
oxygen.

In order to test the involvement of a P-4SO monooxygenase as
the molecular mechanism of ABA-alcohol oxidation in tomato, we
utilized carbon monoxide, a specific inhibitor of heme P-450
monooxygenases (Ortiz de Montellano and Reich, 1986). The
oxidation of ABA-alcohol was measured by GC-NCI-SIM of
[2H361ABA extracted from sit leaves incubated with [2H345](:)A.BA-
alcohol in the presence or absence of carbon monoxide or oxygen.
The results are presented in Table 4.7. In the absence of molecular
oxygen, there was no conversion to [ZHMJABA (Table 4.7). In the
presence of 10% or 50% (v/v) CO, conversion to PHMJABA was
inhibited 58% and 81% of control values, respectively. Similar results
Were obtained in vitro with a cell-free pumpkin endosperrn extract
CF. Pantauzzo and C. Rock, unpublished results) analogous to that of

Gillard and Walton (1976). We further tested the monooxygenase

I"lyraothesis by measuring [’°O]ABA labeling patterns in sit leaves

67

Table 4.7. Inhibition by Carbon Monoxide of ABA-Alcohol Oxidation in Sitiens

Leaves

Sitiens leaves were dipped in 1 x 10'5 M [2H3GJABA'31COhOI plus 0.05%
(v/v) Tween 20, 0.2% ethanol and incubated in the dark for 8 h in air, N2, or air
plus 10% or 50% (v/v) carbon monoxide. Samples were quantified by GC-NCI-

SIM. Mean of three experiments (t SE).

 

 

Treatment [ZHGJABA Percent Inhibition
ng/ g fr wt/ 8 11
Air 11.9 i 2.8
N; ND.‘ 100%
10% CO 5.0 i 0.7 58%
50% CO 2.2 1; 0.5 81%

 

‘Not detected, two experiments

68

incubated under 18O2 or 18O2 plus 50% carbon monoxide. If doubly-
carboxyl-labeled ABA is synthesized by a monooxygenase, the
presence of CO in this experiment should Specifically inhibit the
labeling of this class of [‘BOJABA molecules. The results in Table 4.8
confirm this prediction. The percentage of doubly-carboxyl-labeled
ABA and t-ABA (Table 4.8, columns 6 and 7) was significantly
decreased by CO treatment. When ABA and t-ABA from this
experiment were quantified, more than 70% inhibition of doubly-
carboxyl-labeled ABA and t-ABA biosynthesis was observed (Table
4.8). The results of these experiments support the conclusion that a
cytochrome P-450 monooxygenase is responsible for ABA-alcohol

oxidation in viva.

4.5. DISCUSSION

Based on the results presented here, we conclude that ABA-

alcohol is an intermediate in ABA biosynthesis in a shunt pathway
from ABA-aldehyde that involves reduction of ABA-aldehyde to ABA-
alcohol and oxidation of ABA-alcohol to ABA via a cytochrome P-450

rmonooxygenase (Figure 4.2). A number of species convert

00.00.00 .02.

 

 

 

«86 dz 2 dz no 38 3 «.3 £5
as S 3: 3 S. «a... I. 3 <5
00 u .z N .o._ 205 :8... can .. :8
mod .3. 2: dz 3 08 3.2 08 5.:
x: 2 08 3 0m «.3 82. as $2
”3.1330 03.33“-
a a \ E s s \ a:
9 a c m s a a _
6 E: 8.33 32.3 20.5 was was 30.3 33 gas .832...
453302280 0.. 25 o._ 28 0.. oz... 0.. So 9. as 0.. 2:. o.. 8o

 

3.8030 s 8.2: 0.. 2:.

 

3.80.00 5 :32 O: 0:0

 

3.8230 5 3.32::

 

528.; 33. a 2520.:

 

0:95.003 oz: .00 0:0 :30 0:003" .292 0:0

25.52.00 >0 000230 003 02903 dogma.“ 30.903... 05 .0 000000— 2303500 003 d 0:0 a «5:200 .0030 £50.; .00 00000.0 =<

z m 3.. 0010.32 :0900 *3 we 09.00% :0 09.0005 05 5 «O:
5 0000205 0:005 0:052 90:58 E2002“??? 05 _o 0260: 0ommo0mvu000>> 5 95030..— O: 490300 0:0 4% ac :ogmom 0:0 0:036— dd 030,—.

ES
0005 $2
:0 :0

o:

0 “oz

70

00300.0- -<m<

f8\
_

 

+

5.85:9.

3529...?

OI

0
1090/ no N :0
10:0 / /

<m<-m:0.=

.0c00_0.<m<-m:00

/ 00>:00_0.<m<-m|l.c000
:0
0:0 /

030503.050

 

I
0:0 / /0
: 0030:0000- 0 «0%

OI .5 = OI
./ / / / / / / /O

71

0000000000 30000 :000 00: m0:
0:002: 0.5008: 05 S 8002 0000000005 05 ”00000 000 0:05.: 000008 00030035 05. .0390
200 505 85000 0: 0005:8000 a :0 03500 8.00:: 00:. .A0000mnam 0005000 0 0000 0: 5558005

30:80“ 550080: :80 00055003 50:00 0:0 52 .8 0.003509 0000000“ dd 005E

72
exogenous ABA-aldehyde to ABA-alcohol (Duckham et a1., 1989).

The Ho and sit tomato mutants and dmOpy potato, which is probably
homologous to sit (Duckham er a1., 1989), also isomerize a significant
amount of ABA-aldehyde to t-ABA-alcohol (Duckham et a1., 1989;
Linforth et a1., 1990). This isomerization and accumulation in the flc
and sit mutants of tomato (Linforth et a1., 1987) gives rise to t-ABA
and t-ABA-GE (Table 4.1) as a result of t-ABA-alcohol oxidation
(Figure 4.2). This conclusion is supported by results of 18O labeling
studies which show a higher proportion of doubly-carboxyl-labeled
ABA and t-ABA in the mutants of tomato (Table 4.3), barley, and
potato (Table 4.5) blocked in ABA-aldehyde oxidation. Because
ABA-alcohol from flc and sit is labeled in the primary alcohol group
in 18O labeling experiments (Linforth et a1., 1987; Rock and Zeevaart,
unpublished results), we conclude that reduction of ABA-aldehyde
results in at least partial trapping the 1"0 present in the side chain,
which is then oxidized by molecular oxygen to give doubly-carboxyl-
labeled ABA in 1"O labeling experiments. The oxidation with 1802 of
[ZHMJ (:t)ABA-alcohol to [‘BOZHMMBA (Table 4.6), plus the in viva

- inhibition by C0 of FH3,61(:)ABA-alcohol oxidation (Table 4.7) and

ABA with two 180 atoms in the carboxyl (Table 4.8) indicate that a

73

cytochrome P-450 monooxygenase is the catalyst for ABA-alcohol
oxidation in tomato.

The labeling data (Table 4.3) and feeding studies (Table 4.6)
indicate that the shunt pathway of ABA—alcohol oxidation is not
enhanced in the flc and sit tomato mutants to maximize ABA
biosynthesis. We conclude that this pathway is a minor source of
ABA in plants, but is physiologically significant in mutants which are
impaired in ABA-aldehyde oxidation. Taken together with the
quantitation data of Table 4.2, which show that wild type tomato
synthesizes more t-ABA and t-ABA-GE than the flc and sit mutants, it
can be concluded that oxidation of t-ABA-alcohol only accounts for
about 10-15 percent of total t-ABA biosynthesis. We propose that the
remainder of t-ABA is synthesized by a parallel pathway from all-
trans-epoxy-carotenoids (Figure 4.2).

Duckham eta]. (1989) proposed that synthesis of [2H,]c1's-ABA-
alcohol from [2H1](:)ABA-aldehyde was an artifact produced by
conversion of the unnatural (—)ABA-aldehyde enantiomer in flacca
and sitiens feeding experiments. Our data indicate that (+)-cis—ABA-
alcohol is an endogenous metabolite of plants because (+) ABA is

the naturally occurring enantiomer (Milborrow, 1970). Cris-ABA-

74
alcohol is likely the biosynthetic precursor to t-ABA-alcohol.

18O-Labeling patterns in ABA and t-ABA from tomato suggest
that the loss of ABA-aldehyde oxidase activity in the Ho and sit
mutants (Sindhu et a1., 1990) also results in an increase in ABA—
aldehyde carbonyl oxygen enchange (Table 4.3, compare columns 4
versus 1 between wild type and mutants). The not mutant also
shows this increased carbonyl exchange phenotype, although it has
been shown not to be impaired in xanthoxin oxidation (Parry et a1.,
1988) and only slightly reduced in ABA-aldehyde oxidation (Taylor et
a1., 1988). Based on our data (Table 4.3) and those of Taylor eta].
(1988), ABA-aldehyde oxidase may be affected by the not mutation.
Taylor and Tarr (1984) crossed the tomato mutants and obtained all
the possible double mutants; all these recombinants showed
phenotypic interactions between the gene products of not, 110 and
sit. The function of these genes in ABA-aldehyde oxidation are
unknown. Because the not mutant is extremely leaky, and the
mechanisms involved in [IBOJABA labeling have not been completely
elucidated, and carbonyl exchange patterns are somewhat variable
(compare Tables 4.3 and 4.8, columns 4 versus 1), we cannot

conclude from our data that the site of action of the not mutation is

75
ABA-aldehyde oxidase.

The droopy mutant, which is probably homologous to sit based
on map location (Duckham et 31., 1989), synthesizes large amounts
of t-ABA; the MoCo mutant of barley does not (Table 4.4). The
droopy potato and barley A234 mutants do not exhibit decreased
side chain carboxyl group labeling in [‘BOJABA (Table 4.5, compare
columns 4 versus 1 between wild type and mutant ABA), as is seen
in the tomato mutants. In contrast to tomato wild type t-ABA (Table
4.3), potato and barley wild types exchange more carbonyl label
than they retain in t-ABA (Table 4.5, compare columns 4 versus 1 for
t-ABA). Because doubly-labeled ABA with one 180 atom on the ring
and one in the side chain carboxyl (Tables 4.3 and 4.5, column 4)
can be synthesized from either ring-labeled ABA-alcohol or doubly-
labeled ABA-aldehyde, it is difficult to interpret the ABA and t-ABA
labeling patterns in terms of individual enzyme activities, unless the
precursor pool sizes and labeling patterns are known. The recently
discovered CKRI mutant of N. plumbaginifoiia is impaired in ABA-
aldehyde oxidation, and may be useful to study the mechanism of
carbonyl label retention in [1801ABA. It is not known if the mutations

studied here affect the concentrations of ABA-aldehyde oxidase, or

76

the enzyme's catalytic activity, or ABA-aldehyde isomerization, or the
reduction of ABA-aldehyde to ABA-alcohol. Purification and
characterization of the wild type and mutant enzyme activities
involved in the late steps of ABA biosynthesis may resolve some of

these questions.

4.6. LITERATURE CITED

Creelman RA, Gage DA, Stults IT, Zeevaart IAD (1987) Abscisic acid
biosynthesis in leaves and roots of Xanthium struman’urn. Plant
Physiol 88: 726-732

Creelman RA. Zeevaart IAD (1984) Incorporation of oxygen into
abscisic acid and phaseic acid from molecular oxygen. Plant
Physi0175: 166-169

Duckham SC, Taylor EB, Linforth RST, Al-Naieb RI, Marples BA,
Bowman WR (1989) The metabolism of cis ABA-aldehyde by
the wilty mutants of potato, pea and Arabidopsis thaliana. I
Exp Bot 40: 901-905

Gamble PE, Mullet IE (1986) Inhibition of carotenoid accumulation
and abscisic acid biosynthesis in fluridone-treated dark-grown
barley. EurI Biochem 160: 117-121

Gillard DP, Walton DC (1976) Abscisic acid metabolism by a cell-
free preparation from Echinocystis lobata liquid endOSperm.
Plant Physiol 58: 790-796

Heath TC, Gage DA, Zeevaart IAD, Watson IT (1990) Role of
molecular oxygen in fragmentation processes of abscisic acid

methyl ester in electron capture negative ionization. Org
Mass Spectrom 26: 665-663

77

Koornneef M, Ioma ML, Brinkhorst-van der Swan DLC, Karssen CM
(1982) The isolation of abscisic acid (ABA) deficient mutants by
selection of induced revertants in non-germinating gibberellin
sensitive lines of Arabidopsis thaIiana (L.) Heynh. Theor Appl
Genet 61: 386-393

Li Y, Walton DC (1990) Violaxanthin is an abscisic acid precursor in
water-stressed dark-grown bean leaves. Plant Physiol 92: 551-
569

Linforth RST, Bowman WR, Griffin DA, Marples BA, Taylor IB (1987)
Z-u‘ans-ABA-alcohol accumulation in the wilty tomato mutants
flacca and sitiens. Plant Cell Environ 10: 699-606

Linforth RST, Taylor IB, Duckham SC, Al-Naib RI, Bowman WR,
Marples BA (1990) The metabolism and biological activity of
ABA analogues in normal and flacca mutant tomato plants.
New Phytol 115: 517-621

Milborrow BV (1970) The metabolism of abscisic acid. I Exp Bot 21:
17-29

Moore R, Smith ID (1986) GravireSponsiveness and abscisic acid
content of roots of carotenoid-deficient mutants of Zea mays L.
Planta 164: 126-128

Neill SI, Horgan R (1986) Abscisic acid production and water
relations in wilty tomato mutants subjected to water deficiency.
I Exp Bot 36: 1222-1231

Neill SI, Horgan R, Parry AD (1986) The carotenoid and abscisic acid
content of viviparous kernels and seedlings of Zea mays L.
Planta 169: 87-96

Netting AG, Milborrow BV, Vaughan GT, Lidgard R0 (1988) The
fragmentation of methyl abscisate and its 215' isomer in methane
positive and negative chemical ionization mass spectrometry.
Biomed Environ Mass Spectrom 15: 376-389

Ortiz de Montellano PR, Reich NO (1986) Inhibition of cytochrome P-

78

450 enzymes. In PR Ortiz del Montellano, ed, Cytochrome P-
460: Structure, Mechanism, and Biochemistry. Plenum, New
York, pp 273-314

Parry AD, Babiano MI, Horgan R (1990a) The role of cis-carotenoids
in abscisic acid biosynthesis. Planta 182: 118-128

Parry AD, Blonstein AD, Babiano MI, King PI, Horgan R (1991)
Abscisic-acid metabolism in a wilty mutant of Nicotiana
plumbaginifolia. Planta 183: 237-243

Parry AD, Neill SI, Horgan R (1988) Xanthoxin levels and metabolism
in the wild-type and wilty mutants of tomato. Planta 173: 397-
404

Parry AD, Neill SI, Horgan R (1990b) Measurement of xanthoxin in
higher plant tissues using 1“C labelled internal standards.
Phytochemistry 29: 1033-1039

Quarrie SA (1982) DroOpy: a wilty mutant of potato deficient in
abscisic acid. Plant Cell Environ 6: 23-26

Rock CD, Zeevaart IAD (1990a) Abscisic (ABA)-aldehyde is a
precursor to, and l’,4’-trans—diol a catabolite of, ABA in apple.
Plant Physiol 93: 915-923

Rock CD, Zeevaart IAD (1990b) ABA-alcohol is an intermediate in
abscisic acid biosynthesis. Plant Physiol 93: SS

Rock CD, Zeevaart IAD (1991) The aba mutant of Arabidopsis
thaliana is impaired in epoxy-carotenoid biosynthesis. Proc
Natl Acad Sci USA, submitted

Rock CD, Heath TG, Zeevaart IAD (1991) Trans-ABA biosynthesis
and xanthoxin metabolism in wild type and the aba mutant of
Arabidopsis thaliana. I Exp Bot, submitted

Sindhu RK, Griffin DH, Walton DC (1990) Abscisic aldehyde is an
intermediate in the enzymatic conversion of xanthoxin to
abscisic acid in Phaseoius vulgan’s L. leaves. Plant Physiol 93:

79
689-694

Sindhu RK, Walton DC (1988) Xanthoxin metabolism in cell-free
preparations from wild type and wilty mutants of tomato. Plant
Physiol 88: 178-182

Tal M, Nevo Y (1973) Abnormal stomatal behavior and root
resistance, and hormonal imbalance in three wilty mutants of
tomato. Biochem Genet 8: 291-300

Taylor IB, Linforth RST, Al-Naieb RI, Bowman WR, Marples BA (1988)
The wilty mutants flacca and sitiens are impaired in the
oxidation of ABA-aldehyde to ABA. Plant Cell Environ 11: 739-
745

Taylor IB, Tarr AR (1984) Phenotypic interactions between abscisic
acid deficient tomato mutants. Theor Appl Genet 68: 116-119

Walker-Simm0ns M, Kudra DA, Warner RL (1989) Reduced
accumulation of ABA during water stress in a molybdenum
cofactor mutant of barley. Plant Physiol 90: 728-733

Zeevaart IAD, Creelman RA (1988) Metabolism and physiology of
abscisic acid. Annu Rev Plant Physiol Plant Mol Biol 39: 439-
473

Zeevaart IAD, Heath TG, Gage DA (1989) Evidence for a universal
pathway of abscisic acid biosynthesis in higher plants from 1"O
incorporation patterns. Plant Physiol 91: 1594-1601

CHAPTER 5
THE aba MUTANT OF ARABIDOPSIS THALIANA IS IMPAIRED

IN EPOXY -CAROTENOID BIOSYNTHESIS

80

81
5.1. ABSTRACT

The three mutant alleles of the ABA locus of Arabidopsis
thaliana result in plants that are deficient in the plant grth
regulator abscisic acid (ABA). We have used 18O2 to label ABA in
water-stressed leaves of mutant and wild type Arabidopsis. Analysis
by selected ion monitoring and tandem mass spectrometry of
[‘BOIABA and its catabolites, phaseic acid and ABA-glucose ester,
indicates that the aba genotypes are impaired in ABA biosynthesis
and have a small ABA precursor pool of compounds which contain
oxygens on the ring, presumably oxygenated carotenoids
(xanthophylls). Quantitation of the carotenoids from aba and wild
type leaves establishes that the aba alleles cause a deficiency of the
epoxy-carotenoids violaxanthin and neoxanthin and an accumulation
of their biosynthetic precursor, zeaxanthin. These results provide
evidence that ABA is synthesized via oxidative cleavage of epoxy-
carotenoids (the "indirect pathway"). This is the first report of a
carotenoid mutant in higher plants which undergoes normal
greening. The aba alleles provide an opportunity to study the

physiological role of epoxy-carotenoids in photosynthesis.

82
5.2. INTRODUCTION

Abscisic acid (ABA) is a sesquiterpenoid plant growth
regulator involved in many physiological and develOpmental
processes such as transpiration, germination and dormancy, and
adaptation to environmental stresses, e.g. drought and chilling, and
pathogen attack (Pena-Cortes et a1., 1989; for review see Zeevaart
and Creelman, 1988). Although the structure of ABA (Figure 6.1D)
has been known for 26 years (Ohkuma et a1., 1966), the biosynthetic
pathway in higher plants has not been fully elucidated. The
evidence favoring the indirect pathway from xanthOphylls, as
opposed to the direct pathway from farnesyl pyrophoSphate
(Milborrow, 1974), can be summarized as follows. 1) The viviparous
mutants of maize, Vp-Z, Vp-S, vp-Y, and Vp-Q, are blocked in the early
stages of carotenoid biosynthesis and are ABA-deficient (Moore and
Smith, 1986; Neill et a1., 1986). 2) The carotenoid biosynthesis
inhibitors fluridone and norflurazon also inhibit ABA biosynthesis
(Gamble and Mullet, 1986; Moore and Smith, 1984). 3) "303-
Labeling experiments with water-stressed leaves show 18O

incorporation into the side chain carboxyl group of ABA, but little

83

Figure 5.1. The structures of: A) zeaxanthin; B) trans-violaxanthin; C)

9’-cis-neoxanthin; D) (S)-(+)-abscisic acid.

 

84

\\\\\\\\\

OH

85

incorporation in the oxygen functions on the ring (Creelman et a1.,
1987; Rock and Zeevaart, 1990; Zeevaart et a1., 1989), indicating that
there is a large ABA precursor pool (presumably xanthophylls)
which contains oxygens on the ring (see Figure 5.1). 4) Xanthoxin,
a 015- metabolite of epoxy-carotenoids, is found in plants (Parry et
a1., 1990b) and is readily converted to ABA in vivo (Parry et a1., 1988;
Taylor and Burden, 1973) and in vitro (Sindhu et a1., 1990). 5) A 1:1
correlation on a molar basis between decreases in trans-violaxanthin
and 9’-cis-neoxanthin (Figure 5.1) levels and concomitant increases
in ABA and its catabolites has been shown for dark-grown, water-
stressed bean leaves (Li and Walton, 1990; Parry et a1., 1990a). In
contrast to higher plants, phytOpathogenic fungi synthesize ABA via a
direct pathway from farnesyl pyrophoSphate (Neill et a1., 1984; *
Okamoto et a1., 1988).

Koornneef et a1. (1982) isolated three alleles, (aha-1, -3, -4), of
a single locus (ABA) from ArabidOpsis thaiiana and showed a
correlation between ABA deficiency in seeds of the aba genotypes
and the phenotypic severity of leaf transpiration rates, reduced A
growth, and reduced seed dormancy. Application of ABA to the

mutant plants restored the normal phenotype. Here we report laO

86

labeling studies and quantitation of ABA and carotenoids in leaves of
wild type and the three aba genotypes. The results show a
correlation between ABA deficiency associated with the aba alleles

and epoxy-carotenoid deficiency.
5.3. MATERIALS AND METHODS

5.3.1. Plant Material

Seeds of ArabidOpsis thaiiana ecotype Landsberg erecta
(collection number W20) and the mutant genotypes aha-3 [isolation
mutant G4 (Koomneef er a1., 1982); collection number W122], aba-I
(A26; W21), and aha-4 (A73; W123) were obtained from Dr. Maarten
Koornneef, Agricultural University, Wageningen, The Netherlands.
The genetic nomenclature recommended at the Third International
ArabidOpsis Meeting (East Lansing, MI, April 1987) is used (see
Bleecker et a1., 1988). The aba-I genotype also included the
recessive markers ttg (transparent testa glabra) and yi (yellow
inflorescence). Seeds were germinated on 1% agar in Petri dishes
for 2 weeks following storage at 4°C for two days to break

dormancy. Seedlings were transplanted to trays containing a

87

mixture of perlite/vermiculite/peat moss (1:1:1, v/v/v). Plants were
grown in a high humidity growth chamber maintained on a diurnal
cycle of 9 h light (300 umol - m‘2 - s“) at 23° C and 15 h dark at 20°
C. When stem elongation started, approximately 7 weeks after
germination, rosette leaves were harvested and frozen in liquid N2.
For water stress experiments, detached rosettes were dehydrated
with a hair dryer until 14% of the fresh weight was lost. The stressed
material was immediately incubated in 20% 18O2 : 80% N2 (v/v) or air
at room temperature in the dark for 4 or 8 h and frozen in liquid N2.
The ”Oz (97-98% enrichment) was from Cambridge Isotopes
Laboratories (Woburn, MA).
5.3.2. ABA, ABA-Glucose Ester (ABA-GE), and Phaseic Acid (PA)
Analysis

Frozen tissue was extracted overnight at 4° C with acetone
containing 0.01% (w/v) 2,6-di-tert-butyl-4-methy1phenol and 0.25%
(v/v) glacial acetic acid. To each extract was added 20,000 dpm
PmABA, PHIABA-GE, and [3HIPA (Cornish and Zeevaart, 1984) to
quantify losses during purification. The samples were homogenized
with a Polytron (Brinkmann, Westbury, NY) and filtered, and 10 ml

1.0 M potassium phosphate buffer (pH 8.2) was added. The acetone

88

was evaporated at 35° C using a rotovapor (Brinkmann), and the
aqueous solution was passed through a cellulose filter to remove
precipitated material. The pH was adjusted to 3.0 with 6 M HC1 and
ABA, ABA-GE and PA were partitioned five times into an equal
volume of ethyl acetate. The combined organic fraction was
evaporated and the residue chromatographed by reverse phase
HPLC with a uBondapak C18 semipreparative 0.78 x 30 cm column
(Waters, Milford, MA). A convex gradient (Waters no. 5) was run in
20 min from zero to 50% ethanol [plus 1% (v/v) acetic acid in water]
at a flow rate of 2.5 ml/min. The eluant was monitored by UV
absorbance at 262 nm. The fractions containing ABA-GE isomers
plus PA (17-19 min) and ABA isomers (20.5-24 min) were collected
and dried. The ABA-GE was hydrolyzed to the free acid by 2 M
NH,OH for 2 hr at 60° C and dried. Free ABA (from ABA-GE) was
then separated from PA with a Nova-Pak C18 0.39 x 15 cm analytical
column (Waters) using a 20 min linear gradient from 10% to 60%
methanol [in water containing 1% (v/v) glacial acetic acid] at a flow
rate of 1.5 ml/min. The retention times of PA and ABA were 17 min
and 22 min, reSpectively. Typical recoveries for ABA and PA were >

70% and > 50% for ABA-CE.

89
The ABA, ABA-GE (free acid) and PA samples were methylated

with ethereal diazomethane, and a portion of each sample was
quantified by GC—selected ion monitoring with a IEOL AH-505 double
focussing mass spectrometer equipped with a Hewlett-Packard
5890A gas chromatograph and a 30 m, 0.259 mm internal diameter
DB-23 capillary column (I. & W. Scientific, Rancho Cordova, CA) with
He as the carrier gas. Flow rate was 1 ml/min. The GC oven
temperature was programmed from 80° C to 200° C at 40°/min, then
from 200° C to 250° C at 10°/min. Standard curves of ABA-methyl
ester and PA-methyl ester with ABA-ethyl ester as an internal
standard were constructed for quantitation. ABA-Methyl ester ions
were monitored at m/z= 278, 280, 282, and 284 for l"O-labeled
samples and at m/z= 294, 296, 298, 300, and 302 for PA-methyl ester.
Corrections were made for the natural abundance of stable isotopes
by subtracting the theoretical contribution from the measured ion
abundance. Tandem mass spectrometry was performed on a
Finnegan TSQ-70 triple-quadrupole mass spectrometer as previously
described (Rock and Zeevaart, 1990).

5.3.3. Carotenoid Determinations

Carotenoids were extracted according to Britton (1985) with

9O

modifications. Frozen tissue (approximately 1 g) was extracted
overnight at 4° C in 40 ml methanol containing 1% (w/v) sodium
bicarbonate plus 0.1% (w/v) 2,6-di-tert-butyl-4-methy1phenol and
homogenized. Samples were manipulated in dim light to avoid
isomerization of carotenoids. The extract was filtered, and the
chlorophyll a and b concentrations were determined according to
Holden (1976). The samples were diluted tenfold with water, NaCl
was added to saturation, and the samples were repeatedly
partitioned with 40 ml diethyl ether until no color remained in the
aqueous fraction. The ether was evaporated, and the samples were
saponified with 10% (w/v) KOH (in methanol) for 3 h at room
temperature under a stream of N2. The samples were again diluted
tenfold and partitioned with diethyl ether. The ethereal extracts
were stored overnight at -70° C, and ice crystals were removed by
filtration. The extracts were evaporated, and the residue was
resuspended in 90% hexanes, 10% ethyl acetate (v/v) and
chromatographed by normal phase HPLC with a uPorasil semi-
preparative 0.78 x 30 cm column (Waters) using a linear gradient
from 10% to 100% ethyl acetate in 65 min at a flow rate of 2.5 ml/min.

The major carotenoids were collected and identified by their

91

absorbance maxirna, fine structure spectra, and acid shifts of
absorbance maxirna (Braumann and Grimme, 1981; Britton, 1985).
Cis isomers were characterized by rechromatography of iodine-
catalyzed trans isomers and by spectrophotometric analysis and
predicted equilibrium stoichiometries (Khachik et a1., 1986; Molnar
and Szabolcs, 1980). In addition, the identities of violaxanthin,
neoxanthin and antheraxanthin isomers were confirmed by their
retention times in reverse phase HPLC systems (Li and Walton, 1990;
Parry et a1., 1990a).

Carotenoids were quantified by integration of the area under
the absorbance curve at 450 nm. A standard curve of B-carotene
was constructed, and corrections were made for differences in
Specific extinction coefficients (Braumann and Grirnme, 1981; Britton,

1985; our unpublished data).

5.4. RESULTS

5.4.1 . ABA Biosynthetlc Capacity is Negatively Correlated with the

Phenotypic Severity Associated with the aba Alleles

The aba alleles differ in their phenotypic severity of increased

92

leaf tranSpiration, reduced growth, and reduced seed dormancy,
which is correlated with reduced endogenous levels of ABA
(Koomneef et 31., 1982). Plants homozygous for the aha-3 allele
have close to normal growth rates, while aba-I and aha-4 plants are
more pronounced in their negative effects on plant size and vigor
(Koomneef et a1., 1982; our unpublished data). Biosynthesis of ABA
in leaves of wild type Landsberg erecta and aba genotypes was
determined by 18O-labeling and quantitation of de novo [’BOIABA and
metabolites. The results (Table 5.1) show that the phenotypic
severity of the alleles was correlated with reduced ABA biosynthesis.
Although leaves of plants homozygous for the least severe allele,
aha-3, accumulated ABA and catabolites which were about 50% of
wild type turgid levels (our unpublished data), aha-3 plants
synthesized ABA and its catabolites, PA and ABA-GE, at about 3% of
the wild type capacity in 18O-labeling experiments (Table 5.1).
Leaves of plants homozygous for the more severe alleles, aha-1 and
aha-4, had correspondingly lower ABA biosynthetic capacities (Table

5.1).

93

m.v m4 fio ad 3 m

Nam : 0008 m4 .3 ¢
n-0Q0

flaw“ v.92 od fimv B m

9mm: «.2: ad fimm E v

0.600 9030.050:

05.0 \ 390.: £000» 0 \ 0.:

 

Eon. 5m m0.<m< <m<

 

853%: 333-0... 8288

 

.flz0sfi0dx0 05 mo 0m000>< 0509208 so“ 0050.00.00 .3 00050300 “0.00 3508 505 00 0053030

0003 00353 .363 40 0 “0055036 30030 Bozo—“0:0 «o 0005 9:30.85 E 003: 0.0 8:832

Nos 5 2 m s a 8. 803:2: 5% 0500300 do 8588 Ba «2:. as
new 25. BE do 85.3 0385633 :50 85898 9a 52 383.0: no 805050 .3 «Ba.

94

0003000 “oz "42

 

flu
“A

42

.42

md
vd

no
fio

mo

000.0

8.483 E 2%...

o;
md

a."
N6

2 m

2 a
4.2:.

2 m

2 4
73a.

95
8.4.2. ABA Precursor Pool Size is Correlated with ABA Biosynthesis

in aba Genotypes
The position and extent of 18O incorporation into the ABA
molecule can be determined by analysis of 18O-labeled ABA by
tandem mass Spectrometry (Rock and Zeevaart, 1990; Zeevaart et a1.,
1989). Comparison of 18O incorporation, by wild type and aba
Plants, into the ring-attached oxygens of ABA (Table 5.2) indicates a
correlation between ABA biosynthesis (Table 5.1) and turnover of the
ABA precursor pool containing oxygens on the ring (presumably
xanthophylls). Leaves of plants homozygous for the most severe
allele, aha-4, had the highest percentage of 18O incorporation into
the ring-attached oxygens of [IBOIABA After 4 hr of water stress in
an “Oz atmosphere, almost 60% of [’SOIABA from aha-4 plants was
labeled in the ring l’-hydroxyl group (Table 5.2), whereas in the wild
type Landsberg tissue only 4% of the [‘SOIABA contained 18O at this
position. Plants homozygous for the less severe aba alleles had
intermediate levels of 18O ring incorporation (Table 5.2), which
indicates that ABA precursor pool turnover was correlated with
phenotypic severity of the alleles (Koomneef et a1., 1982). In the

wild type Landsberg erecta plants, incorporation of 18O in the

96
Table 5.2. 18O Incorporation into the Ring-Attached Oxygens of ABA from Water-

Stressed Leaves of Wlld Type and the Three aba Genotypes of Arabidopsis after
Incubation in 1802 for 4 or 8 h

Mutants are listed in order of increasing phenotypic severity (Koomneef et
a1., 1982). Samples were analyzed as their methyl ester derivatives by GC-selected
ion monitoring and tandem mass spectrometry (Rock and Zeevaart, 1990). The
label was always present in the l’-hydroxyl group (Zeevaart et a1., 1989); in
addition, 5-10% of the l’-hydroxyl-1abeled ABA was also labeled in the 4’-keto

group.

 

Cenotype/ 1802 Incubation Percent of total [180]ABA

Time labeled in the ring-attached oxygens

 

Landsberg erecta

4 hr 4.3%

8 hr 8.7%
aha-3

4 hr 6.9%

8 hr 8.0%
aha-1

4 hr 29.8%

8 hr 27.4%
aha-4

4 hr 58.6%

8 hr 60.2%

 

97

ring-attached oxygens was six to fifteen times greater, on a fresh
weight basis, than the incorporation by the aba plants. [This
calculation is the product of [IBOIABA levels (Table 5.1) and the
respective percentage of [IQOIABA which is ring-labeled (Table 5.2)].
Thus, the ABA precursor pool containing oxygens on the ring is
much smaller in the aba plants than in wild type plants.

5.4.3. Epoxy-Carotenoid Deficiency and Zeaxanthin Accumulation
are Correlated with the Small ABA Precursor Pool in aba Genotypes
The results of the 18O2 incorporation studies with the aba
genotypes (Tables 5.1 and 5.2) suggested that the xanthophyll levels

of aba plants are reduced. Figure 5.2 shows chromatograms of
carotenoids from leaves of wild type and aha-4 plants. It is clear
that aha-4 plants had reduced levels of the major epoxy-carotenoids
trans-violaxanthin and 9’-cis-neoxanthin (Figure 5.2, peaks 5 and 7),
and that this mutant accumulated zeaxanthin (peak 3). Ouantitation
of carotenoids and chlorophst from leaves of Landsberg erecta and
the three aba genotypes is presented in Table 3. Zeaxanthin, the
biosynthetic precursor to the epoxy-carotenoids antheraxanthin,
violaxanthin, and neoxanthin (Jones and Porter, 1986), accumulated

in leaves of all the aba genotypes. B-Carotene, the precursor to

98

Figure 5.2. HPLC chromatograms of carotenoids extracted from
leaves of wild type Landsberg erecta and the aha-4 genotype of
ArabidOpsis thaiiana. Numbered peaks are: l, fi-carotene; 2, lutein;
3, zeaxanthin; 4, antheraxanthin; 5, trans-violaxanthin; 6, 9-cis-

violaxanthin; 7, 9’-cis-neoxanthin.

Absorbance, 450 nm

99

 

0.5 -

.O
O
l

 

 

 

 

.0
\l
1

0.0

 

 

 

 

 

 

 

 

 

Time, min

Landsberg erecta
5
l 7
3 4 6
__/k - l' ' '
' ' l I l r f
aha-4 3
2 .
i
l
“—j\ 4 5 6 7
_ L I llLJaLu '
, , I , . , r . r . .
15 30 45 60

100

zeaxanthin, also accumulated in aha-4 plants (Table 5.3). There was
a reduction of tans-violaxanthin and 9’-cis-neoxanthin in all aba
plants. Lutein, the most abundant xanthophyll in wild type leaves
and a product of the a-carotene branch pathway (Iones and Porter,
1986), was significantly reduced in the aba genotypes. Furthermore,
the quantitative differences in zeaxanthin and B-carotene
accumulation (Table 5.3) are correlated with the phenotypic severity
(Koomneef et a1., 1982) of the different aba alleles. From these
results we conclude that the ABA precursor pool containing oxygens
on the ring is composed of the epoxy-carotenoids violaxanthin and
neoxanthin.

Total carotenoids and chlorOphylls were not significantly
different in wild-type and aba plants on a gram fresh weight basis
(Table 5.3). However, the lower chlorophyll b level in aha-4 plants
is of interest because chlorophyll b and epoxy-carotenoids are
associated predominantly with the light-harvesting photosynthetic
complexes (Siefermann-Harms, 1985), and epoxy-carotenoids are
neccessary for assembly of photosystem II light-harvesting

complexes in vitro (Plumley and Schmidt, 1987).

d... v m 60>. 0...... 0.2. .0200... 2.000500... .0032. .000. .... 000.0 00o. mod v k 60>. 0:3 000. 3.080
3:85:00 23:2. 0988.. .o: u 0.2. .08. ... 0%.: oz... 8... v m .00... 3.3 ES. .380... 3:85:00 .532. .8058. 0.8 8.5.05.

 

101

 

mm... H mm H mod. H S... H 2.... H cod H E... H 30 H «a... H and H
cod omm. 0.8. E... no.2 N... .m... 0.0 .N.Nm .03 6.2.. 0.000
vmd H mm H mm... H mod H mod H mod H :d H and H mm. H No. H
mod cm: 0.8. . ..o no.2 md .06 ad .03 5.3 ..m« 7000
S... H mm H 2... H mod H vod H mod H «.6 H end H Nod H and H
ov.m o8. 0.3. .md 3.0.2 md 36 0. 5.3 5.0m Eva 0-000
omd H mm H mm... H 2d H mm... H mod H mad H 2.0 H 0...... H a... H 3...” H 0.0000
80 cm: flaw. odm m. ed ms. ad a. odm 9mm 9030000..
590.: 0000 u \ a:

£0 :50 :30 00600300 02.00502 02.00502 0.5000035 550000.05 0.0.000 .5500n00N .5005 000.200-.. 0.50000

6520 -2200 $0-0 00005 0.50 000$. 0.0002

W

3000000000 5.0 43009030 .0030 .00 H. 0000
0.000.000.0000. o3. .300000000 05 .o 0002 009002 000 3000.02 0. 00000000 00 00.00000. 00.9.2300 000 30030.00 000 3000.00. 5.3

00.00000 0.03 0.00.0 5.0 o. 03. 0.00. 000.00.. .Ammm. 40 .0 3000.8on 300.60 0.300000 .0 000.0 00.000000. 0. 00.0.. 0.0 00.0..0 .0052

mwmmopsma .o 000.co000 000 00...... 0... 000 00>... 25> .o 00:00.. .00.. 0.3000520 000 «0.9.0.300 .o 02.500000 .00 030,—.

1 02
5.5. DISCUSSION

The data presented here (Tables 8.2, 5.3) provide strong
correlative evidence for the indirect pathway of ABA biosynthesis
from violaxanthin and neoxanthin. If a direct pathway from farnesyl
pyrophosphate exists in ArabidOpsis, it is of negligible physiological
importance. Prom extrapolation of the data in Tables 5.1 and 5.3, we
predict that a complete loss of epoxy-carotenoids would result in
absence of ABA and would be lethal.

The correlations between phenotypic severity (Koomneef et a1.,
1982), reduced ABA biosynthesis (Tables 5.1, 5.2) and reduced
epoxy-carotenoid content (Table 5.8) suggest that the ABA locus
affects an enzyme which functions in the epoxidation of xanth0phylls. 9
Such an enzyme has been identified in chlor0plast envelopes
(Costes et a1., 1979) and as a component of the xanth0phyll cycle
(Siefermann and Yamamoto, 1975), which is involved in the
dissipation by zeaxanthin of excess energy in photosynthesis
(Demmig et a1., 1987; Demmig-Adams et a1., 1990). The data
presented here imply that one and the same epoxidase is involved in
the xanth0phyll cycle and in epoxy-carotenoid biosynthesis. The

mutations in the different alleles are presumably leaky, and the

103

residual epoxidase activity determines the rate-limiting step of
violaxanthin and neoxanthin biosynthesis, and consequently of ABA
biosynthesis, in aba plants.

Because aba plants have increased ABA precursor pool
turnover (Table 5.2), these genotypes may be useful to study the
regulation of epoxy-carotenoid and ABA biosynthesis. The
involvement of violaxanthin, neoxanthin, and zeaxanthin in
photosynthetic processes makes the aba alleles of ArabidOpsfs a
promising experimental tool to investigate the function of epoxy-

carotenoids and the xanth0phyll cycle in plants.

5.6. LITERATURE CITED

Braumann T, Grimme LH (1981) Reversed-phase high performance
liquid chromatography of chlor0phylls and carotenoids.
Biochim Bi0phys Acta 637: 8-1'Z

Britton G (1985) General carotenoid methods. Meth Enzymol 111:
113-149

Bleecker AB, Estelle MA, Somerville C, Kende H (1988) Insensitivity
to ethylene conferred by a dominant mutation in Arabidopsis
thaliana. Science 241: 1086-1089

Cornish K, Zeevaart IAD (1984) Abscisic acid metabolism in relation

to water stress and leaf age in Xanthium struman’um. Plant
Physiol'16: 1029-1035

104

Costes C, Burghoffer C, onard 1, Block M, Douce R (197 9)
Occurrence and biosynthesis of violaxanthin in isolated
Spinach chlor0plast enve10pes. FEBS Lett 103: 17-21

Creelman RA, Gage DA, Stults IT, Zeevaart IAD (1987) Abscisic acid
biosynthesis in leaves and roots of Xanthfum struman’um. Plant
Physiol 85: 726-732

Demrnig B, Wmter K, Kruger A, Czygan F-C (1987) Photoinhibition
and zeaxanthin formation in intact leaves: a possible role of the
xanthophyll cycle in the dissipation of excess light energy.
Plant Physiol 84: 218-224

Demmig-Adams B, Adams WW IH, Heber U, Neirnanis S, Wmter K,
Krliger A, Czygan F-G, Bilger W, Bjorkrnan O (1990) Inhibition
of zeaxanthin formation and of rapid changes in radiationless
energy dissipation by dithiothreitol in spinach leaves and
chlor0p1asts. Plant Physiol 92: 293-301

Gamble PE, Mullet IE (1986) Inhibition of carotenoid accumulation
and abscisic acid biosynthesis in fluridone-treated dark-grown
barley. Eur I Biochem 160: 117-121

Holden M (1976) Chlor0phylls. In TW Goodwin, ed, Chemistry and
Biochemistry of Plant Pigments, V01 2. Academic, New York. pp
1-3'1

Iones BL, Porter W (1986) Biosynthesis of carotenes in higher
plants. CRC Crit Rev Plant Sci 3: 295-324

Khachik P, Beecher GR, Whittaker NP (1986) Separation,
identification, and quantification of the major carotenoid and
chlor0phyll constituents in extracts of several green vegetables
by liquid chromatography. I Agric Food Chem 34: 603-616

Koomneef M, Ioma ML, Brinkhorst-van der Swan DLC, Karssen GM
(1982) The isolation of abscisic acid (ABA) deficient mutants by
selection of induced revertants in non-germinating gibberellin
sensitive lines of Arabidopsfs thaliana (L.) Heynh. Theor Appl
Genet 61: 385-393

105

Li Y, Walton DC (1990) Violaxanthin is an abscisic acid precursor in
water-stressed dark-grown bean leaves. Plant Physiol 92: 551-
559

Milborrow BV (1974) The chemistry and physiology of abscisic acid.
Annu Rev Plant Physiol 25: 259-307

Molnar P, Szabolcs I (1980) Occurrence of lS—cz’s-violaxanthin in
Viola tn'coIor. Phytochemistry 19: 623-627

Moore R, Smith ID (1984) Growth, gravireSponsiveness and abscisic
acid content of Zea ma ys seedlings treated with fluridone.
Planta 162: 342-344

Moore R, Smith JD (1985) Graviresponsiveness and abscisic acid
content of roots of carotenoid-deficient mutants of Zea mays L.
Planta 164: 126-128

Neill S], Horgan R, Parry AD (1986) The carotenoid and abscisic acid
content of viviparous kernels and seedlings of Zea mays L.
Planta 169: 87-96

Neill SI, Horgan R, Walton DC (1984) Biosynthesis of abscisic acid.
In A Crozier, IR Hillman, eds, The Biosynthesis and Metabolism
of Plant Hormones. Cambridge Univ, Cambridge, UK, pp 43-70

Ohkurna K, Addicott PT, Smith OE, Thiessen WE (1965) The structure
of abscisin II. Tetrahedron Lett 29: 2529-2535

Okamoto M, Hirai N, Koshimizu K (1988) Biosynthesis of abscisic
acid. Mem Coll Agric Kyoto Univ 132: 79-115

Parry AD, Babiano M], Horgan R (1990a) The role of cis-carotenoids
in abscisic acid biosynthesis. Planta 182: 118-128

Parry AD, Neill SI, Horgan R (1988) Xanthoxin levels and metabolism
in the wild-type and wilty mutants of tomato. Planta 173: 397-
404

Parry AD, Neill S], Horgan R (1990b) Measurement of xanthoxin in

106

higher plant tissues using 13C labelled internal standards.
Phytochemistry 29: 1033-1039

Pena-Cortes H, Sénchez-Serrano I], Mertens R, Willmitzer L, Prat S
(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

Plurnley PG, Schmidt GW (1987) Reconstitution of chlor0phyll a/b
light-harvesting complexes: xanthophyll-dependent assembly
and energy transfer. Proc Natl Acad Sci USA 84: 146-150

Rock CD, Zeevaart IAD (1990) Abscisic (ABA)-aldehyde is a
precursor to, and l’,4’-trans-ABA-diol a catabolite of, ABA in
apple. Plant Physiol 93: 915-923

Siefennann D, Yamamoto HY (1975) Properties of NADPH and
oxygen-dependent zeaxanthin epoxidation in isolated
chlor0plasts. Arch Biochem BiOphys 171 70-77

Siefermann-Harms D (1985) Carotenoids in photosynthesis. I.
Localization in photosynthetic membranes and light-harvesting
function. Biochirn Biophys Acta 811: 325-355

Sindhu RK, Griffin DH, Walton DC (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

Taylor HF, Burden RS (1973) Preparation and metabolism of 2-[14C]-
cis,trans-xanthoxin. J Exp Bot 24: 873-880

Zeevaart IAD, Creelman RA (1988) Metabolism and physiology of
abscisic acid. Annu Rev Plant Physiol Plant Mol Biol 39: 439-
473

Zeevaart IAD, Heath TC, Gage DA (1989) Evidence for a universal
pathway of abscisic acid biosynthesis in higher plants from 16O
incorporation patterns. Plant Physiol 91: 1594-1601

CHAPTER 5
2-TRANS-ABSCISIC ACID BIOSYNTHESIS AND THE
METABOLISM OF ABA-ALDEHYDE AND XANTHOXIN ‘
IN WILD TYPE AND THE

3178 MUTAN T OF ARABEOPSIS THALIANA

107

108
6.1. ABSTRACT

Abscisic acid (ABA) and 2-trans-ABA (t-ABA) biosynthesis were
studied in wild type Landsberg erecta and the three allelic aba
mutants of Arabidopsis thalz’ana (L.) Heynh., which are impaired in
epoxy-carotenoid biosynthesis. Labeling experiments with 18O2 and
mass spectrometric analysis of [‘BO]ABA and its catabolites ABA-
glucose ester (ABA-GE) and phaseic acid (PA), and t-ABA and t-
ABA-GE, showed that t-ABA biosynthesis was less affected than ABA
biosynthesis by mutations at the ABA locus. The aha-4 allele caused
the most severe impairment of ABA biosynthesis compared with the
other two mutant alleles aba-I and aha-3, yet aha-4 plants
synthesized as much t-ABA as wild type Landsberg erecta plants.
Feeding experiments with RS-PHSJABA-aldehyde isomers and
unlabeled xanthoxin suggest that t-xanthoxin and t-ABA-aldehyde are

precursors to ABA and t-ABA in Arabidopsis.

1 09
6.2. INTRODUCTION

The abscisic acid (ABA)-deficient mutants of Arabidopsis
thaliana (L.) Heynh. and tomato (LyCOpersicon esculentum Mill.)
have been paramount to the understanding of ABA biosynthesis.
Recently, it has been shown that the aba alleles of ArabidOpsis are
impaired in epoxy-carotenoid biosynthesis (Rock and Zeevaart,
1991), which provides strong evidence for the indirect pathway of
ABA biosynthesis from the epoxy-carotenoids violaxanthin and
neoxanthin, through 2-c1's-xanthoxin and 2-cis-ABA-aldehyde to ABA
(for review see Zeevaart and Creelman, 1988). The flacca and
sia'ens mutants of tomato are impaired in ABA-aldehyde oxidation
(Sindhu and Walton, 1988; Taylor et a1., 1988) and accumulate t-ABA-
alcohol (Linforth et a1., 1987). ABA-deficient mutants in potato
(Solanum tuberosum, group Phureja; Duckham et a1., 1989),
Nicotiana plumbagz'nifolia (Parry et a1., 1991), and barley (Hordeum
vulgare; Walker-Simmons et a1., 1989) have also been characterized
as impaired in ABA-aldehyde oxidation. We have demonstrated that
the flacca and sitfens mutants of tomato synthesize a significant

percentage of ABA from ABA-alcohol via a shunt pathway from ABA-

110
aldehyde (Rock et a1., 1991). In these mutants, most t-ABA and the

catabolite t-ABA-GE are synthesized from t-ABA-alcohol.

With the exception of the shunt pathway from t-ABA-alcohol,
little work has been done on the biosynthesis of t-ABA, since this
isomer is biologically inactive (Walton, 1983). Since trans-ABA is
rapidly esterified to t-ABA-GE (Milborrow, 1970), it may have been
overlooked in ABA biosynthesis studies. Apple fruits can accumulate
significant amounts of t—ABA during ripening (Bangerth, 1983), and
this t-ABA is not the result of isomerization of ABA (Rock and
Zeevaart, 1990). 2-Trans-xanthoxin, a metabolite of epoxy-
carotenoids and precursor to trans-ABA (Taylor and Burden, 1973), is
the predominant isomer found in plants (Parry et a1., 1988, 1990).

The identification of [1°O]t-ABA-aldehyde in 18O labeling studies with
apple fruits suggested the existence of a parallel pathway of t-ABA
biosynthesis from epoxy-carotenoids through t-xanthoxin and t-ABA-
aldehyde (Rock and Zeevaart, 1990).

Here we report results of feeding and 18O labeling studies with
leaves of wild type Landsberg erecta and the three aba genotypes of
Afabidopsis thaliana. The results suggest that t-ABA is synthesized

via a pathway distinct from t-ABA-alcohol oxidation. Quantitation of

111
ABA, t-ABA and catabolites indicates that plants homozygous for the

most ABA-deficient allele, aha-4, are capable of synthesizing

significant amounts of t-ABA.

6.3. MATERIALS AND METHODS

6.3.1. Plant Material

Seeds of ArabidOpsis thaliana (L.) Heynh., ecotype Landsberg
erecta (collection number W20) and the mutant genotypes aha-3
[isolation mutant G4 (Koomneef et a1., 1982); collection number
W122], aba-I (A26; W21), and aha-4 (A73; W123) were obtained from
Dr. M. Koornneef, Agricultural University, Wageningen, The
Netherlands. The aba-l genotype also carried the recessive
markers ttg (tranSparent testa/glabrous) and y1' (yellow
inflorescence). Seeds were germinated on 1% agar in Petri dishes
for 2 wks following storage at 4°C for two d to break dormancy.
Seedlings were transplanted to trays containing a mixture of
perlite/vermiculite/peat moss (1:1:1, v/v/v). Plants were grown in a
high humidity growth chamber and maintained on a diurnal cycle of

9 h light (300 uE ~ In2 ' s"), 23° C and 15 h dark, 20° C. When stern

112

elongation started, approximately 7 weeks after gennination, rosette
leaves were harvested and frozen in liquid N2 or used for feeding
and 18O2 labeling experiments. For ABA biosynthesis studies the
tissue was water-stressed with a hair dryer until 14% of the fresh
weight was lost. The tissue was incubated in 20% 18O2 : 80% N2
(v/v) (Creelman and Zeevaart, 1984) or in air, at room temperature in
the dark for 24 h, and frozen in liquid N2. 18O2 (97-98% enrichment)
was purchased from Cambridge Isotopes Laboratories (Wobum, MA,
USA).
6.3.2. Feeding Experiments

RS-[ZHMJABA-aldehyde [97.2% (w/w)] was synthesized as
described (Rock and Zeevaart, 1990). 2-Cis- and trans-xanthoxin
were obtained by potassium permanganate oxidation of violaxanthin
(Burden and Taylor, 1970), extracted from leaves of Xanthium
stmman’um (Britton, 1985) and purified via an open ODS column as
described by Sindhu and Walton (1987). Reaction products were
chromatographed by normal phase HPLC on a uPorasil
semipreparative 0.78 x 30 cm column (Waters, Milford, MA, USA)
with a linear gradient from 10 to 60% (v/v) ethyl acetate in hexanes

in 23 min at a flow rate of 2.5 cm3 min'1 and the eluant monitored by

113

UV absorbance at 280 nm. Xanthoxin isomers and butenone were
collected from 28 to 31 min and dried under a stream of N2. The
xanthoxin isomers were separated from butenone by reverse phase
HPLC on a uBondapak semipreparative 0.78 x 30 cm column
(Waters) with a linear gradient of 20-60% (v/v) ethanol (in water) in
25 min at a flow rate of 2.5 cm3 min". Butenone eluted at 12 min and
xanthoxin isomers at 18 min. Samples were taken to dryness under
a stream of N2. Xanthoxin and ABA-aldehyde isomers were
separated and purified by reverse phase HPLC using a 4 pm
Novapak C18 Radial-PAK 0.8 x 10 cm cartridge (Waters) and a linear
gradient from 20% to 60% (v/v) methanol (in water) in 45 min at a
flow rate of 2.5 cm3 min“. Trans-ABA-aldehyde eluted from 22.7 to
23.5 min, and cis-ABA-aldehyde from 25.5 to 26.8 min. Trans-
xanthoxin eluted from 29.8 to 31 min, and cis-xanthoxin from 32 min
to 33.3 min. Substrate purity was confirmed by GC-flame ionization
detection, and the isomers were quantified by UV absorbance
(Burden and Taylor, 1970; Rock and Zeevaart, 1990). A 2.5 ug cm“1
aqueous solution (1 x 10'2 mol ms) of substrate containing 0.05% (v/v)
Tween 20 was vacuum infiltrated into leaf tissue from seven to 16

plants of wild type or aba genotypes, reSpectively, to give similar

114

fresh weights. The tissue was placed in air, or 20% ”Oz : 80% N2
(v/v) for xanthoxin experiments, incubated in darkness for 8 h, and

then frozen.

6.3.3. Extraction, Purification, and Quantitation of Metabolites

ABA, t-ABA, ABA-GE, t-ABA-GE, and phaseic acid (PA) were
extracted and purified by reverse phase HPLC as described (Rock
and Zeevaart, 1991). The 2-trans isomers of ABA and ABA-GE had
slightly shorter retention times in this HPLC system than the
corresponding 2-c1's-isomers. Samples were methylated with
ethereal diazomethane and quantified by GC-negative chemical
ionization-selected ion monitoring as described (Chapters 4, 5).
ABA-aldehyde was extracted and quantified by the isotOpe dilution
method as described by Rock and Zeevaart (1990). Tandem mass
spectrometry (MS/MS) was performed on a Finnegan TSQ-7O triple-
quadrupole mass spectrometer as previously described (Rock and

Zeevaart, 1990).

l 15
6.4. RESULTS

6.4.1. ABA and Trans-ABA Biosynthesis and Metabolism

Initial characterization of the aba genotypes established a
negative correlation between tranSpiration rates and ABA
accumulation in seeds and immature green siliques (Koomneef et a1.,
1982); the three aba genotypes also showed a correlation between
ABA levels, reduced growth rates, and reduced seed dormancy.
Table 6.1 confirms this result for leaf tissue. There was a negative
correlation between phenotypic severity of the aba alleles and ABA,
t-ABA and catabolite levels in unstressed tissue, with the exception
of t-ABA and t-ABA-GE accumulation in plants homozygous for the
most severe allele aha-4. Trans-ABA levels in unstressed aha-4
plants were higher than in wild type, and t-ABA-GE was only slightly
less than in wild type (Table 6.1). After 24 h water stress, Landsberg
erecta accumulated ABA and its catabolites ABA-GE and PA, when
taken together, to levels about six times higher than unstressed
leaves (Table 6.1). The combined contents of t-ABA and t-ABA-GE
in wild type increased only about two-fold (Table 6.1). In leaves of

the aba genotypes, ABA and the major catabolite PA increased

. Mitt-N uC l0>601~ pruned moazoafluflo DEG (mt-:QCEQ i: ‘0 503323030

I FE ‘A‘ISHCCOO
.~.mv 0~n~flp-

6

11

.Bsofifioauo 03. mo 2.3803?

 

 

wd ad ad 04. ed .: vw pommobm

.m.m mm; H QN :3 H Hm :d H ¢._ 2.: H a; @3353:
YQO

m.“ md o." ad ad .: vm pommobm

Ed H m._ de H Hm «fiv H Qm 3.0 H Ho mmd H Nd panache:
7QO

ed v4 04 vd Wm h vm pommobm

54 H Nam mvd H m ~ and H o w. mod H o o vmd H ad pommmbmcb
9QO

mvdm H «62 mad H N.“ «Ya H We Ed H md 3.2 H mdm ; Iva pommoum

om.m H Qmm on a H ad OWN H Num 8 o H ad mm; H 3. vommobmab
gumbo Scamp—84

“a H9553 m0.<m< <52 a

 

Cum H: coho @8982 H :moEV .bsmfi? god 3 5:

20585096058

 

653538 :2 Ufloo_oV:oumqu_ 3250:“.

guanoPUO >3 609535 0.83 «2ng Sagan 0329859 053205 go 390 5 team: 08 8:83)—

mmobm .833 : vm ~82 93 280m 330sz
mam. 9.8 as ESP finQoEg .«0 3234 an «Smonflmu paw. 5.3g 5 no command—5 4.0 £nt

 

6.2
Cata

gent

117

slightly in response to 24 h water stress. ABA-GE levels were more
variable in both wild type and mutant tissue (Table 6.1). Consistent
with the relatively high levels of t-ABA and t-ABA-GE in unstressed
aha-4 tissue, leaves of this mutant allele also accumulated a
significant amount of t-ABA and t-ABA-GE after 24 h water stress
(Table 6.1). Thus, although ABA biosynthesis is impaired by the aba
mutation, in the case of the aha-4 genotype, t-ABA biosynthesis is not
reduced as compared to Landsberg wild type.

The capacity for ABA biosynthesis can be measured by 1"O-
labeling studies and quantification of [‘GO]ABA and catabolites by
GC-negative chemical ionization-mass spectrometry (Rock et a1.,
1991; Rock and Zeevaart, 1990, 1991; Zeevaart et a1., 1989). Table
6.2 presents the quantification of 18O-labeled ABA, t-ABA and
catabolites from water-stressed leaves of wild type and aba
genotypes in laO2 for 24 h. Consistent with the results of Table 6.1,
ABA biosynthesis was reduced in the aba genotypes and negatively
correlated with phenotypic severity (Koomneef er a1., 1982)
associated with the individual aba alleles (Table 6.2). Furthermore,

the accumulation of t-ABA and t-ABA-GE in aha-4 plants (Table 6.1)

was due to de novo synthesis of t-ABA and t—ABA-GE (Table 6.2).

 

13.71530 DEG “23‘-

118

.pfloflop “oz 2

528.; 58 33355 :3 .22... o... as: 3 22 52 5:9; .

 

md

.02

 

v4 v.0 v4 m4 Twas
mé d2 Adz womb v.0 m4 7QO
m.m 3 md Nd md 9m «:QO
0.8 3: am 3 3 «an mama 938:3
.3359; m0.<m<¢ HO...“ mnuimEOaL €52 ”Om; find—”Os;
8503908 $850500

8% a... 8mm

 

.2 «a _.Em_oz fig 3 a:

 

.Ecofinoauo 025 mo 00823 653.305 com p302$éoumfic£

Ragga ogmmocOO >3 pogcmsq 203 $353 sun—96m cabocoga mfimmouofi Co 390 5 noun: 0.8 $5352

No... a 2 a 65 "2588 8... ea 25

33 aaonfiew .0 8:3 3803-3sz Boa 358.8 98 «3.88 .mmm 3335.. Lo sousufiso .3 saws

I‘E

18C
Oxi.
and
(elk

Carb

1 19
When t-ABA biosynthesis ([180] t-ABA plus [IBOJt-ABA-GE) in wild

type and aba plants was expressed as a fraction of ABA biosynthesis
([‘°O]ABA + [‘80]ABA-GE + [leOJPA) , a positive correlation
between relative t-ABA biosynthesis and phenotypic severity was
observed (Table 6.2, ratio cis/trans metabolites). Thus, consistent
with the elevated t-ABA levels in aha-4 plants (Table 6.1), 18O
labeling studies also showed that the aba alleles had less effect on t-
ABA than ABA biosynthesis (Table 6.2).

The fragmentation by NCI of Me-ABA has been elucidated
(Heath et a1., 1990: Netting et a1., 1988). By MS/MS of [1°O]ABA from
in vivo labeling experiments it is possible to quantify the
incorporation of 18O into each oxygen atom of ABk the side chain
carboxyl, the ring 4’-keto, and the l’-hydroxyl group. Four possible
reactions in ABA biosynthesis have been characterized by this
method of analysis (Rock et a1., 1991; Rock and Zeevaart, 1990): l)
1"O incorporation into the side chain carboxyl group as a result of
oxidative cleavage of epoxy-carotenoids, such as 9-cis-violaxanthin
and 9’-cis-neoxanthin. 2) Turnover of the large ABA precursor pool
(epoxy-carotenoids), resulting not only in 18O incorporation into the

carboxyl group, but also into the oxygens on the ring of [IBOJABA

C0

ab.

120
3) Synthesis of ABA from ABA-alcohol in a minor shunt pathway from

ABA-aldehyde, which yields doubly-carboxyl-labeled [IBO]ABA via
incorporation of 18O from a second molecular oxygen. 4) Side chain
carbonyl oxygen exchange with water at an aldehyde intermediate
stage of ABA biosynthesis, resulting in loss of side chain carboxyl
label in [‘°O]ABA. Analysis by MS/MS of [leO]ABA and [won-ABA
from stressed wild type and aha-4 plants in 18O2 for 24 hr is
presented in Table 6.3. The [leO]ABA molecules are qualitatively
distinguished by the number of 18O atoms and their position of
incorporation, giving rise to five major classes of [1°O]ABA (columns
1-5). The percentage of total [leO]ABA in each class allows a
comparison of the ABA biosynthetic pathways in the wild type and
aha-4 genotypes. Landsberg erecta showed the typical pattern of
18O incorporation in ABA from stressed leaves. The bulk of [“1ij
was singly labeled, and the label was present in the carboxyl group.
This pattern of incorporation is due to the large epoxy-carotenoid
pool which turns over slowly; 18O is incorporated into the side chain
of ABA during oxidative cleavage of the xanthOphyll precursor (Rock
and Zeevaart, 1990; Zeevaart et a1., 1989). In plants homozygous for

the aha-4 allele, 65 percent of the [‘BOJABA was labeled in the

1

2

1

 

3 0.8 3: as «.8 «a:
S can 3a 3 3m 52
View
as n: .92 .2 _.: <9:
3 g as 3 Q £2

80030 90303303

 

 

 

 

 

m v m N 3
man 5 :3 E mac 3.. me. 5 ms. 5
02 030N 02 QGO O... 030N 0.: 03h. 02 0:0
3383300 5 9:80 O: 03>. 3333300 5 E06 02 0:0 333380 5 30303233
$28.; 32 Lo 2830 03200

 

H03: So: am can «550.; 392 mo «$6 :05 $2 003 5350.; “003033.335. .zoEmoQ _>No.—p>3..3 05 30 3030303 20303800 030
.m p30 m «55:00 E030 £55.; .30 0.00.020 =< .Amgmsc 3.050500% 0008 50333.00 >3 3032030 0303 m0EEmm

a a as No... 5 3385 2528
0.030 05 pad 093. 35> go «2803 p0mm0hmé0fi>> 50.3 5.055 p5 a E mam—0303.02 Mo 3030053 3:0 BBQ .md 0.30?

 

122
oxygens on the ring, while in wild type only 14 percent of the

[‘eO]ABA was labeled on the ring (Table 6.3, compare columns 1
plus 4 between wild type and mutant). This result indicates that the
epoxy-carotenoid precursor pool was turning over rapidly in the
mutant, consistent with the observed low levels of epoxy-carotenoids
(Rock and Zeevaart, 1991). In both wild type and aba genotypes, 1‘-
ABA from the same material contained a higher percentage of ”O
label in the ring positions than ABA (Table 6.3, compare columns 1
plus 4 between ABA and t-ABA). Based on the law of mass action,
the large differences in labeling patterns between [1°O]ABA and
[mOJt-ABA from the same samples indicate that isomerization of ABA
was not the source of t-ABA. This conclusion is supported by
feeding studies with apple (Rock and Zeevaart, 1990) and
measurements of ABA isomerization in vino (Milborrow, 1970).
IsotOpe exchange studies have shown that [1°O]ABA which is
unlabeled in the side chain carboxyl group can arise in mO-labeling
experiments by carbonyl oxygen exchange with the medium at an
aldehyde intermediate stage of ABA biosynthesis (Rock and
Zeevaart, 1990). Thus, xanthoxin or ABA-aldehyde with two 1"O

atoms could be oxidized to doubly-labeled ABA (9.9:, Table 6.3,

123

column 4), or could exchange the carbonyl label and then be
oxidized to ABA that is only labeled on the ring (Table 6.3, column
1). The data in Table 6.3 suggest that carbonyl exchange was
slightly higher in the aha-4 plants than in wild type, because ABA
from Landsberg erecta tissue retained almost six tirnes more label in
the side chain than it lost the label, but the aha-4 plants lost a
greater percentage of side chain label (Table 6.3, compare columns
4 versus 1 between wild type and mutant). [”0] Trans-ABA from
both wild type and aha-4 plants showed a higher degree of aldehyde
carbonyl exchange than [IBOJABA from the same sample (Table 6.3,
compare columns 4 and l betweenABA and t-ABA).

[mOJABA with two laO-atoms in the carboxyl group (Table 6.3,
column 5) is synthesized from [‘BOJABA-alcohol with incorporation of
18O by a minor shunt pathway from ABA-aldehyde (Rock et a1., 1991).
The minor shunt pathway of ABA biosynthesis from ABA-alcohol was
not affected in the aha-4 genotype (Table 6.3, compare column 5
between wild type and mutant). This is unlike the flacca and sia‘ens
mutants of tomato, which are blocked in ABA-aldehyde oxidation
(Sindhu and Walton, 1988; Taylor et a1., 1988) and synthesize a

significant percentage of ABA and t-ABA via ABA-alcohol and t-ABA-

124
alcohol, respectively (Rock et a1., 1991).

6.4.2. ABA-Aldehyde and Xanthoxin Metabolism in aba Mutant
Alleles

In order to determine if the ABA locus affects aldehyde
oxidation, C-2 isomers of RS-[ZHGJABA-aldehyde were fed to leaves of
wild type and aba genotypes. In agreement with the work of
Duckham et a1. (1989), mutations at the ABA locus do not affect ABA-
aldehyde oxidation (Table 6.4). There was evidence for
isomerization of ABA-aldehyde and t-ABA-aldehyde to their
respective C-2 isomers with subsequent oxidation to t-ABA and ABA,
followed by conjugation to their glucose esters (Table 6.4). This is
interpreted to mean that there was little or no interconversion
between ABA-GE and t-ABA—GE, which is supported by results of
feeding studies with apple fruits (Rock and Zeevaart, 1990) and
tomato shoots (Milborrow, 1970).

Xanthoxin is a postulated intermediate in ABA biosynthesis
between epoxy-carotenoid cleavage and ABA-aldehyde oxidation
(Sindhu et a1., 1990; Zeevaart and Creelman, 1988). In order to test
whether the ABA locus affects xanthoxin metabolism, C-2 isomers of

xanthoxin were fed to wild type and aba genotypes in the presence

 

o.vm mAm 0.3 9.82 adv 03>30Em.<m<.w:m5

3E. 3 ea «.2: 4.8: «B88358»;
4.28
So so 3 38 EN 8388.43.88
38m «.0. n2 «.8 «.80 8383.098
78m
$8 3. ms 38 m8 83835.88
3% S 0.: So. 32: 83834038

80030 90330303

 

125

$225 008$.er moamfizmm Emisfi «ma-”E

 

0Hmbm3=m \ 0&30500
.2 a .290; 88 3 a:

 

85053090 .8350 003 mo 98 :83 $3m0m 9339208 :2 “09003003030553 30033030 260503.00 >3 30093538 0303
83933 .3 m 30.3 33p 05 5 “00.03205 98 030mb H03 05 95 30030235 Song 003 29503 cm 3003? $86 9:58:00
cases 88 4.8 u 88 .8 :8 u as seen 8283.53-93 Lo 8838 as as «2 u a 18 9. 3 0

00330300 030 paw
09? 25> “@533 Co 00503 5 00503900 p5 5.058 a o. B0500— 0p>302$a§ uo 8330230 .vd £309

12.6

of 18Oz. The presence of 1802 during the xanthoxin feed allowed
subtraction of any ABA synthesized from endogenous substrates
during the experiment, because this ABA was 18O-labeled.
Exogenous xanthoxin would not incorporate 180 during oxidation to
ABA except via ABA-alcohol (Rock et 31., 1991), which is a minor
pathway in Arabidopsis and amounts to only about 2% of the total
ABA (Table 6.3, see below). The results presented in Table 6.8
show that the conversion of xanthoxin to ABA was not affected by
the ABA locus. There was considerable isomerisation of t-xanthoxin,
which was metabolized to ABA and then PA (Table 6.5). Cis-
xanthoxin was also isomerised, but to a lesser extent than t-
xanthoxin. It has been shown that t-xanthoxin is converted to t-ABA
in tomato plants (Taylor and Burden, 1973; Parry et 31., 1988). Only
small quantities of [‘BOJABA and [190] t-ABA were detected in the
xanthoxin isomer feeding experiments (data not shown); if ABA and
t-ABA were synthesized from ABA-alcohol and t-ABA-alcohol, a
significant amount of ABA from exogenous xanthoxin would have
been 18O-labeled (Rock er a1., 1991). The apparent higher xanthoxin
and ABA-aldehyde oxidizing activity of the mutant tissue may be a

result of the smaller size of the mutant leaves, which resulted in

127
Table 6.6. Conversion of Xanthoxin Isomers to ABA, trans-ABA and Catabolites in

Leaves of Arabidopsis Wild Type and aba Genotypes in 1302

A 2.5 pg cm‘1 (1 x 10'2 mol m“) solution of cis- or trans-xanthoxin (both
isomers 94% pure) containing 0.05% Tween 20 (vol/vol) was vacuum infiltrated into
leaf tissue, which was then put under 20% 1302. 80% N3 for 8 h. ABA, t-ABA, and
catabolites were quantified by CBC-negative chemical ionization-selected ion
monitoring and endogenous (time = 0 h) and de novo ABA and catabolites (= ‘30-
1abeled after 8 h) were subtracted from the total to give the amounts of ABA and
catabolites synthesized from applied xanthoxin. Results from one of three
experiments with similar results.

 

 

ng (g fresh weight)" (8 h)‘1

 

 

Genotype/Subs trate
ABA t—ABA ABA-GE t-ABA-GE PA

Landsberg erecta

cis-xanthoxin 14.1 60.0 0.0‘ 1.7 705.8

trans-xanthoxin 3.3 61.0 0.0 1.4 492.7
aha-I

cis-xanthoxin 38.5 29.2 0.0 0.9 2788.7

trans-xanthoxin 8.8 94.7 0.0 0.0 431.5
aha-4

cis-xanthoxin 128.2 104.4 7.6 1.0 3716.6

trans-xanthoxin 67.1 242.0 1.5 12.9 2302.0

 

‘Afier correction for endogenous material.

128

complete infiltration of substrate into the tissue (CD. Rock,

unpublished observations).

6.6. DISCUSSION

The results presented here suggest that t-ABA in Arabidopsis is
synthesized primarily through a pathway which is distinct from t-ABA-
alcohol oxidation (Rock et 31., 1991). In the presence of 1802, ABA-
alcohol and t-ABA-alcohol can be oxidized to ABA and t-ABA,
respectively (Figure 6.1), with incorporation of 180 into the carboxyl
group (Rock et a1., 1991). In Arabidopsis, this pathway accounts for
only slighly more than 1 percent of the total [‘°O]ABA and [1°O]t-
ABA, based on abundance of doubly-carboxyl-labeled ABA and t-
ABA (Table 6.3, column 5). Feeding experiments by Duckham eta].
(1989) with deuterated ABA-aldehyde did not result in significant
synthesis of t-ABA-alcohol in Arabidopsis. The results of our ABA-
aldehyde isomer feeding experiments (Table 6.4) support the
conclusion that isomerization of ABA-aldehyde to give t-ABA-alcohol
and t-ABA is only a few percent of the total ABA biosynthesis. The

most likely source of t-ABA, therefore, is via a parallel pathway from

129

<m<-mcm.=

1000/
1000

/I

0/0
_ocoo_a-<m<
m
:o :o/ Mo :0
Klwvmzo :0 /
ou>cou.a-<m<.mca=

82208.3( \
010 O 0
IO
/
5.85:8? mica:

:_x05cux
o: a
010/
YOIO\JN\O/b\H

—————> I
/0

_/g

 

/ /O.'"g

55:58.03

  

5.253.203. win-o. m

 

130

.0886 005000 53> 55050 0.8 8500000
5506?: "850.00 >50; 3 00.08% 3 8353 00.85 9a. 00ng0 no. 905QO 0 0m? mm 5558805

55583? Bob Emmfigmofi Bow 880308200 98 200 06630 mo mhmgfima 03305 Ad 0.53m

131

aH-a'ans-epoxy-carotenoids (Rock et a1., 1991; Rock and Zeevaart,
1990; Figure 6.1).

There was greater isomerization in the t-xanthoxin and RS-
[zHe]t-ABA-a1dehyde feeding experiments than in the cis isomer
experiments (Tables 6.4 and 6.5). These observations suggest that t-
xanthoxin and t-ABA-aldehyde may be precursors to ABA in
Arabidopsis (Figure 6.1). Similar feeding experiments with xanthoxin
isomers in tomato shoots by Taylor and Burden (1973) and Parry et
a1. (1988) showed some isomerisation and conversion to ABA
isomers; however, labeled ABA catabolites were not quantified in
these experirnents. High t-xanthoxin : xanthoxin ratios have been
found in all tissues and Species so far examined (Parry et a1., 1988,
1990). Xanthoxin and t-xanthoxin interconvert non-enzymatically in
vitro with an equilibrium that favors the trans isomer (Parry et a1.,
1990). Although precautions were taken to minimize isomerization,
we cannot rule out artifactual isomerization of xanthoxin isomers
during the feeding experiments. Our results suggest that t-xanthoxin
was isomerized in vivo to xanthoxin to a greater extent than
xanthoxin to t-xanthoxin, which is against the expected direction of

the isomerization reaction (xanthoxin --> t-xanthoxin; Parry et a1.,

 

132
1990). In the present study, endogenous ABA-aldehyde and t-ABA-

aldehyde were present in leaves of wild type and aba genotypes
(data not shown). Thus, a parallel pathway from all-trans-epoxy-
carotenoids may give rise to t-ABA and ABA in Arabidopsis (Figure
6.1).

The cause(s) of the relatively high t-ABA biosynthetic
capacities in the aba genotypes (Table 6.2) are not understood.
One possible explanation is that epoxy-carotenoid deficiency in the
aba genotypes results in an altered regulation of the parallel
pathways to ABA and t-ABA The cleavage enzyme may utilize trans-
violaxanthin as a substrate when 9-cis-violaxanthin is limiting;
isomerisation of the resultant t-xanthoxin would contribute to ABA
biosynthesis, but would also result in greater t-ABA biosynthetic
rates. The observation that tans—epoxycarotenoids are more
reduced in the aba genotypes than 9-cis-epoxy—carotenoids (Rock
and Zeevaart, 1991) is not inconsistent with this hypothesis. Isolation
and characterization of the enzymes involved in ABA biosynthesis
may resolve questions of substrate Specificity and the multiple

pathways of ABA and t-ABA biosynthesis.

 

l 33
6.6. LITERATURE CITED

Bangerth F (1982) Changes in the ratio of cis-trans to trans-trans
abscisic acid during ripening of apple fruits. Planta 155: 199-
203.

Britton G., 1985. General carotenoid methods. Meth Enzymol 111:
113-149

Burden RS, Taylor HF (1970) The structure and chemical
transformations of xanthoxin. Tetrahed Lett 47: 4071-4074

Creelman RA, Zeevaart IAD (1984) Incorporation of oxygen into
abscisic acid and phaseic acid from molecular oxygen. Plant
Physiol 78: 166-169

Duckham SC, Taylor IB, Linforth RST, Al-Naieb RI, Marples BA,
Bowman WR (1989) The metabolism of cis ABA-aldehyde by
the wilty mutants of potato, pea and ArabidOpsis thaliana. I
Exp Bot 40: 901-905

Heath TG, Gage DA, Zeevaart IAD, Watson IT (1990) Role of
molecular oxygen in fragmentation processes of abscisic acid

methyl ester in electron capture negative ionization. Organ
Mass Spectrom 25: 655-663

Koomneef M, Ioma ML, Brinkhorst-van der Swan DLC, Karssen CM
(1982) The isolation of abscisic acid (ABA) deficient mutants by
selection of induced revertants in non-germinating gibberellin
sensitive lines of ArabidOpsis thaliana (L.) Heynh. Theor Appl
Genet 61: 385-393

Linfonh RST, Bowman WR, Griffin DA, Marples BA, Taylor IB (1987)
Z-fi’ans-ABA-alcohol accumulation in the wilty tomato mutants
flacca and sia'ens. Plant Cell Environ 10: 599-606

Milborrow BV (1970) The metabolism of abscisic acid. I Exp Bot 21:
17-29

134

Netting AG, Milborrow BV, Vaughan GT, Lidgard R0 (1988) The
fragmentation of methyl abscisate and its 25' isomer in methane
positive and negative chemical ionization mass spectrometry.
Biomed Environ Mass Spectrom 15: 376-389

Parry AD, Blonstein AD, Babiano MI, King PI, Horgan R (1991)
Abscisic-acid metabolism in a wilty mutant of Mcon‘ana
plumbagz'nifolia. Planta 183: 237-243

Parry AD, Neill SI, Horgan R (1988) Xanthoxin levels and metabolism
in the wild-type and wilty mutants of tomato. Planta 1'13: 397-
404

Parry AD, Neill SI, Horgan R (1990) Measurement of xanthoxin in
higher plant tissues using 13C labelled internal standards.
Phytochemistry 29: 1033-1039

Rock CD, Heath TG, Gage DA, Zeevaart IAD (1991) Abscisic (ABA)-
alcohol is an intermediate in ABA biosynthesis in a shunt
pathway from ABA-aldehyde. Plant Physiol, submitted

Rock CD, Zeevaart IAD (1990) Abscisic (ABA)-a1dehyde is a
precursor to, and 1’,4’-fi'ans-diol a catabolite of, ABA in apple.
Plant Physiol 93: 915-923

Rock CD, Zeevaart IAD (1991) The aba mutant of Arabidopsfs
thalfana is impaired in epoxy-carotenoid biosynthesis. Proc
Natl Acad Sci USA, submitted

Sindhu RK, Griffin DH, Walton DC (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

Sindhu RK, Walton DC (1987) Conversion of xanthoxin to abscisic
acid by cell-free preparations from bean leaves. Plant Physiol
88: 916-921

Sindhu RK, Walton DC (1988) Xanthoxin metabolism in cell-free
preparations from wild type and wilty mutants of tomato. Plant

135
Physiol 88: 178-182

Taylor HF, Burden RS (1973) Preparation and metabolism of 2-[“C]-
cis,trans-xanthoxin. I Exp Bot 24: 873-880

Taylor IB, Linforth RST, Al-Naieb RI, Bowman WR, Marples BA (1988)
The wilty mutants flacca and sitfens are impaired in the
oxidation of ABA-aldehyde to ABA. Plant Cell Environ 11: 739-
745

Walker-Simmons M, Kudra DA, Warner L (1989) Reduced
accumulation of ABA during water stress in a molybdenum
cofactor mutant of barley. Plant Physiol 90: 728—333

Walton DC (1983) Structure-activity relationships of abscisic acid
analogs and metabolites. In FT Addicott. ed, Abscisic Acid.
Praeger, New York, pp 113-146

Zeevaart IAD, Creelman RA (1988) Metabolism and physiology of
abscisic acid. Annu Rev Plant Physiol Plant Mol Biol 39: 439-
473

Zeevaart IAD, Heath TG, Gage DA (1989) Evidence for a universal
pathway of abscisic acid biosynthesis in higher plants from 18O
incorporation patterns. Plant Physiol 91: 1594-1601

CHAPTER 7
18O INCORPORATION INTO VIOLAXANTHIN AND ABSCISIC ACID
VIA THE XANTHOPHYLL CYCLE SUPPORTS VIOLAXANTHIN AS

A PRECURSOR OF ABSCISIC ACID

136

1 37
7.1. INTRODUCTION

The xanthophylls have been shown to function in a number of
physiological processes of importance to plants, such as scavenging
of oxygen radicals produced by photooxidation (Cogdell, 1989),
photosynthetic light harvesting (Sieferrnann-Harms, 1985), assembly
of light harvesting complexes (Plumley and Schmidt, 1984; Humbeck
et a1., 1989), ABA biosynthesis (Rock and Zeevaart, 1991), and
dissipation of excess light energy through the “xanthophyll cycle"
(Demmig et a1., 1987; Demmig-Adams et 31., 1990). The xanthophyll
‘ cycle is the reversible, light-induced deepoxidation/epoxidation of
the epoxy-carotenoids violaxanthin and antheraxanthin in the
thylakoid membrane (Figure 7.1)(Siefermann—Harms, 1977;
Yamamoto, 1979). The product of deepoxidation is zeaxanthin,
which has recently been shown to function in the protection of the
photosynthetic apparatus by non-photochemical quenching of high-
chlorophyll fluorescence, thus preventing photoinhibition by high
photon flux densities (Demmig et a1., 1987; Demmig-Adams et a1.,
1990). In order to study the biosynthetic relationship of xanthophylls

and ABA, we have utilized the xanthophyll cycle to specifically

138

OH
O\ \ \ \ \ \ \ \ \
HO . .
Violaxanthin
dark NADPH 1L light
02
OH
O\ \ \ \ \ \ \ \ \
HO
antheraxanthin
dark "ADP” 1L light
02
OH
\ \ \ \ \ \ \ \ \
HO

zeaxanthin

Figure 7.1. The xanthophyll cycle (after Yamarnoto, 1979).

139
pulse-label the epoxide group of violaxanthin and antheraxanthin

with 18O and follow 18O incorporation into the 1’-hydroxy1 position of
[‘80]ABA synthesized after water-stress in air. Results of a similar

experiment with bean leaves have been reported by Li and Walton

(1987).

7.2. MATERIALS AND METHODS

Spinach (Spinacia oIeracea L., Savoy Hybrid 612, Harris Seed
00., Rochester, NY) was grown as described (Zeevaart, 1971). Fully
expanded leaves were excised and immediately weighed and frozen
in liquid N2 for determination of the initial ABA content. About 80 g
fresh weight of leaves were arranged in a monolayer in a custom-
designed 28 cm x 38 cm x 1 cm vacuum-tight plexiglas case fitted
with an O-ring and lined with wet paper towels. The case was
sealed with C-clamps and evacuated and flushed with N2 four times.
The case was illuminated through an icewater bath with a 400 W
halogen lamp ("multivapor," General Electric) at a photon flux
density of 900 (13 rn‘2 - S‘1 for 50 min; this was the length of time

found to give the maximum conversion of violaxanthin to zeaxanthin

 

140

(data not shown). The case was evacuated and 18O2 (97-98%
enrichment, Cambridge Isotope Laboratories, Wobum MA)
introduced to 20% (v/v) and filled with N2. The tissue was incubated
for 4 h in the dark in 18Oz, then removed to air and a portion frozen
for ABA and xanthophyll determinations. This length of time was
sufiicient for complete regeneration of the violaxanthin pool (data not
shown). The remainder of the sample was water-stressed with a hair
dryer until 14% of the fresh weight had been lost, then incubated in
air for 2 h and frozen. A control experiment was performed with
leaves which were treated as described above, except N2 was
substituted for 18O2 after irradiation, and 1802 was substituted for air
after water stress.

ABA and xanthophylls were extracted and purified as
described (Rock and Zeevaart, 1991). ABA was quantified as the
methyl ester derivative by GC-ECD using a standard curve of ABA-
Me with dieldrin as an intemal standard. Incorporation in [‘°O]ABA-
Me was measured by GC-NCI-SIM and MS/MS as described in Rock
and Zeevaart (1990) and corrected for 13C isotope contributions.
Incorporation of 18O in tans-violaxanthin, 9’-cfs-neoxanthin,

antheraxanthin, zeaxanthin, and lutein was measured by positive ion

141
FAB-MS on a IEOL-HXIlO double focussing mass spectrometer by

direct probe in a matrix of nitrobenzyl alcohol (Vetter and Meister,
1985). Nominal mass molecular ions of violaxanthin and neoxanthin
gave m/z = 600.4, antheraxanthin m/z = 584.4, lutein, and
zeaxanthin, m/z = 568.4. These compounds showed the typical
xanthophyll fragment ions of m/z = M+-80, m/z = 221, and m/z =
181 (Budzikiewicz et a1., 1970). The percent 18O incorporation was
calculated after correcting for 13C isotope abundance and proton

adduct formation, described in Appendix D).
7.3. RESULTS AND DISCUSSION

In order to establish a biosynthetic relationship between
xanthOphylls and ABA, the amounts of 18O-labeled precursor and
product and the position of 18O label must be determined. Figure 7.2
shows the molecular ion (M+) obtained by FAB-MS of all-trans-
violaxanthin before and after running the xanthophyll cycle for 4 h in
”Oz. About 80% of 18O-labeled violaxanthin was doubly-labeled (see
Appendix D, Table D4).

The 1"O enrichment of xanthophst pulse-labeled via the

142

 

 

 

 

 

 

 

3 ' A
M+°
600

2 .-

1 ..
0
° | |
g [11: I‘Ll‘l [3-1!Ijllill'll'll‘Ll'll‘lL'll T1311: 1211': 1'! 1 I'll:111 1'14'1 1‘: 1'1 1'11'1 1'1 1'. 1'1 1
'0 570 580 590 600 610 620 630 640
C
3
.Q
<
(D
.2
:6 B M4" M 4+-
09 4 600 +
I * 604

3 .

2 a

 

 

 

 

 

 

 

 

1 1
l'Lljll'lLlltLlYll!lrll'llrlj'lltlJ‘I ll1 . . ll'lltll'l-JIVIl‘ll'll'lI:L‘11:1'LL'.1'“'..ITL
570 580 590 600 610 620 630 640

 

m / 2
Figure 7.2. Positive ion FAB-mass Spectra of the molecular ion
cluster of all-trans-violaxanthin from unlabeled (A), and 18O-labeled
(B) spinach leaves run through the xanthophyll cycle, 4 h in 1"Oz,

The asterisk (*) indicates a matrix ion.

 

143

xanthophyll cycle is presented in Table 7.1. This analysis reflects
the synthesis and turnover rates of the compounds under conditions
where violaxanthin and antheraxanthin are de-epoxidated in N2 in the
light and re-epoxidated in 20% 16O2 in the dark Violaxanthin and
antheraxanthin were highly enriched with 18O (Table 7.1). The lack
of complete labeling in violaxanthin was presumably due to
inaccessability of a fraction of the violaxanthin pool to the
xanthophyll cycle enzymes (Siefermann and Yamamoto, 1974). The
greater enrichment of antheraxanthin compared to violaxanthin
(Table 7.1) was due to its smaller pool size and more rapid turnover
(data not shown) as an intermediate between zeaxanthin and
violaxanthin. Zeaxanthin showed some 18O incorporation into the
ring hydroxyl functions (Table 7 .1), presumably as a result of de
nova synthesis from B-carotene. Antheraxanthin and violaxanthin
showed a similarly small enrichment of two and three 18O atoms, or
three and four 18O atoms, respectively, as zeaxanthin (see Appendix
D, Table D4). Neoxanthin and lutein are not substrates in the
xanthophyll cycle (Yamamoto and Higashi, 1978) and were not ”O-
labeled (Table 7 .1).

Analysis of [IBOIABA by MS/MS allows determination of the

 

144
Table 7.1. I°O Enrichment of Xanthophylls from Spinach Leaves After Running

the Xanthophyll Cycle in ”O; for 4 h

Xanthophylls were purified by normal phase HPLC and analyzed by FAB-

MS.

 

 

Compound 180 Enrichment (percent of molecules)
Trans-violaxanthin 51%
Trans-antheraxanthin 75%
Trans-zeaxanthin 3%
9’-cis-neoxanthin < 1%
Trans-lutein ND‘

 

‘ Not detected

145
position and extent of 18O label (Rock and Zeevaart, 1990). In Table

7.2 the levels of [IGOIABA and the percentage of [laOIABA in each
oxygen function are presented. When [‘BOIABA was analyzed from
tissue frozen immediately after 18O labeling, only 20% of the [1°O]ABA
was labeled in the l’-hydroxyl function (Table 7.2). In this treatment
the bulk of [1°O]ABA was carboxyl-labeled (Table 7.2); it is the
typical labeling pattern of ABA (Creelman et a1., 1987). When themO
xanthophyll cycle material was water-stressed and allowed to
synthesize ABA in air after the 4 h 18O2 pulse-labeling of violaxanthin,
a large increase in the percentage of 18O incorporation into the 1’-
hydroxyl was observed (Table 7.2). However, very little [1°OIABA
was synthesized by the stressed tissue, only about 5 ng/g fr wt/4 h
(Table 7.2, "[‘°O]ABA" column). It can be deduced that the [1°OIABA
synthesized after water stress was nearly entirely labeled in the 1’-
hydroxyl position and unlabeled in the carboxyl group to effect the
drastic change in labeling patterns from a small amount of additional
[‘90]ABA A second conclusion which can be drawn from the results
in Table 7.2 is that the bulk of ABA synthesized (about 80%) was
unlabeled (compare "Total ABA" versus "[‘°O]ABA" columns). The

control experiment with 2 h water-stress in “’03 gave the expected

146

Table 7.2. Position and Extent of ABA Labeling in Spinach Leaves After Running

the Xanthophyll Cycle in ”O; for 4 h

Detached leaves were irradiated in N2 for 50 min to convert violaxanthin
to zeaxanthin, then placed in 20%:80% (v/v) 18Oz: N2 for 4 h, and then either
frozen or stressed in air for 2 h before freezing. An experiment showing stress-
induced ”O incorporation into ABA is included. Samples were quantified as

their methyl ester derivatives by GC-SIM and GC-MS/MS (Rock and Zeevaart,

 

 

 

 

1990).
Percent [“OIABA by position
Treatment Total ABA [‘°O]ABA
4’-keto 1’-hydroxyl carboxyl
ng/ gfw ng/ gfw/tfrne
Control 18.0
4 hr in "02,
no stress 33.0 18.8 2.7 19.8 77.5
4 hr in "Oz, 2 hr
stress in air 125.0 24.0 3.5 63.0 33.5
4 hr in N2, 2 hr

stress in "’02 224.0 152.0 1.4 1.0 97.6

 

147

high percentage incorporation into the side chain carboxyl group
and a high rate of [”OIABA biosynthesis (Table 7.2). Labeling of the
4’-l<eto function of ABA (Table 7.2) was consistent with low
incorporation into the ring-hydroxyl functions of violaxanthin
(Appendix D, Table D4).

We interpret the results of Tables 7.1 and 7.2 as evidence in
support of violaxanthin as a precursor to ABA. Both violaxanthin and
9’-cfs-neoxanthin are present in Spinach leaves at levels
approximately 100-fold greater (Khachik et a1., 1986) than ABA
(Table 7.2). The high percentage of l’-hydroxyl-labeled [”OIABA in
the pulse chase experiment (Table 7.2) provides evidence for a
biosynthetic link between violaxanthin and ABA. Antheraxanthin can
also be an ABA precursor based on ”O enrichment (Table 7.1).
However, the large, transient changes in antheraxanthin levels in
leaves exposed to a high photon flux density suggest it functions
solely as an intermediate between violaxanthin and zeaxanthin (data
not shown). The lack of ”O incorporation in 9'-cfs-neoxanthin
(Table 7.1) correlates with the lack of [”O]ABA biosynthesis two
hours later (Table 7.2). Therefore, 9’-cis-neoxanthin could be the

source of the unlabeled ABA. This interpretation is supported by

 

 

 

148
Similar results with bean leaves (Li and Walton, 1987). Alternatively,

unlabeled violaxanthin could be the source of de nova unlabeled
ABA.

Subcellular sequestering of xanthOphylls (Siefermann and
Yamamoto, 1974; Siefermann-Harms et a1., 1978; Siefermann-Harms,
1985) suggests that the ABA precursor pool of xanthophyus may not
be the same as the xanthOphyll cycle pool. However, it is clear that
these two pools are connected, based on ”O incorporation into the
1’-hydroxyl position of ABA after labeling violaxanthin through the
xanthophyll cycle (Tables 7.1 and 7.2). 9-Cis-violaxanthin is.
probably an intermediate in ABA biosynthesis from violaxanthin; an
isomerization reaction from trans-violaxanthin has been postulated
(Li and Walton, 1990). If the xanthophyll cycle is specific for all-
trans-violaxanthin (Yamarnoto and Higashi, 1978), then isomerization
of all-trans-violaxanthin to 9-c1's-violaxanthin must be rapid,
considering the [”OIABA data (Table 7.2). In order to provide
stronger evidence for ABA biosynthesis from ”O-labeled
xanthophylls, 9-cis-violaxanthin could be analysed by FAB-MS to
determine if the ”O enrichment is higher than for violaxanthin.

Chloroplast envelopes have a separate pool of violaxanthin from the

 

I F.

 

 

149
thylakoids (Siefermann-Harms et a1., 1978) and may exhibit different

”O labeling patterns in xanthophyll cycle pulse-chase experiments,
providing additional information on the subcellular location of ABA
biosynthesis. The rapid and specific ”O labeling of trans-
violaxanthin by the xanthophyll cycle allows examination of the
biosynthetic relationship of violaxanthin to ABA and epoxy-
carotenoids such as 9-cis-violaxanthin, neoxanthin and 9’-cfs-

neoxanthin.

7.4. LITERATURE CITED

Budzildewicz H, Brzezinka H, Johannes B (1970) Zur Photosynthese
griiner Pflanzen, 2. Mitt.: Massenspelctroskopische

Untersuchungen Carotinoiden. Monatshefte fiir Chernie 101:
579-609.

Creelman RA, Gage DA, Stults IT, Zeevaart IAD (1987) Abscisic acid
biosynthesis in leaves and roots of Xanthium struman’um. Plant
Physiol 85: 726-732

Cogdell R (1989) Chlor0phylls. In TW Goodwin, ed, Plant Pigments.
Academic, New York, pp 183-230

Demrnig B, Winter K, Kruger A, Czygan F-C (1987) Photoinhibition
- and zeaxanthin formation in intact leaves: a possible role of the

xanthophyll cycle in the dissipation of excess light energy.
Plant Physiol 84: 218-224

Demmig-Adams B, Adams WW III, Heber U, Neimanis S, Winter K,

 

150

Kriiger A, Czygan F-C, Bilger W, Bj6rkman O (1990) Inhibition
of zeaxanthin formation and of rapid changes in radiationless
energy dissipation by dithiothreitol in Spinach leaves and
chloroplasts. Plant Physiol 92: 293-301

Humbeck K, Réimer S, Senger H (1989) Evidence for an essential role
of carotenoids in the assembly of an active photosystem II.
Planta 179: 242-250

Khachik F, Beecher GR, Whittaker NP (1986) Separation,
identification, and quantification of the major carotenoid and
chlorophyll constituents in extracts of several green vegetables
by liquid chromatography. I Agric Food Chem 34: 603-616

Li Y, Walton DC (1987) Xanthophylls and abscisic acid biosynthesis
in water-Stressed bean leaves. Plant Physiol 85: 910-915

Li Y, Walton DC (1990) Violaxanthin is an abscisic acid precursor in
water-stressed dark-grown bean leaves. Plant Physiol 92: 551-
559

Plumley FG, Schmidt GW (1987) Reconstitution of chlorophyll a/b
light-harvesting complexes: xanthophyll-dependent assembly
and energy transfer. Proc Natl Acad Sci USA 84: 146-150

Rock CD, Zeevaart IAD (1990) Abscisic (ABA)-aldehyde is a
precursor to, and l’,4’-trans-diol a catabolite of, ABA in apple.
Plant Physiol 93: 915-923

Rock CD, Zeevaart IAD (1991) The aba mutant of Arabfdapsis
thalfana is impaired in epoxy-carotenoid biosynthesis. Proc
Natl Acad Sci USA, submitted

Sieferrnann D, Yamarnoto HY (1974) Light-induced de-epoxidation of
violaxanthin in lettuce chloroplasts. III. Reaction kinetics and
effect of light intensity on de-epoxidase activity and substrate
availability. Biochirn Bi0phys Acta 357: 144-150

Siefermann-Harms D (1977) The xanthophyll cycle in higher plants.
In M Teveni, I-IK Lichtenthaler, eds, Lipids and Lipid Polymers

151
in Higher Plants. Springer-Verlag, New York, pp 218-230

Siefermann-Harms D (1985) Carotenoids in photosynthesis. 1.
Location in photosynthetic membranes and light-harvesting
function. Biochim Biophys Acta 811: 325-355

Siefermann-Harms D, onard I, Douce R (1978) Light-induced
changes of the carotenoid levels in chlorOplast envelopes.
Plant Physiol 61: 530-533

 

Vetter W, Meister W (1985) Fast atom bombardment mass spectrum
of B-carotene. Org Mass Spectrom 20: 266-267

Yamamoto HY (1979) Biochemistry of the violaxanthin cycle in higher
plants. Pure Appl Chem 51: 639-648

Yamamoto HY, Higashi RM (1978) Violaxanthin de-epoxidase: lipid
composition and substrate Specificity. Arch Biochem Biophys
190: 514-522

Zeevaart, IAD (1971) Effects of photoperiod on growth rate and
endogenous gibberellins in the long-day rosette plant spinach.
Plant Physiol 47: 821-827

 

CHAPTER 8

CONCLUDING REMARKS

152

153

In viva ”O labeling and analysis of [”OJABA by tandem mass

Spectrometry has proven a powerful approach to elucidating the ABA

 

biosynthetic pathway, which is Shown in Figure 8.1. This approach
established that ABA-aldehyde iS a precursor to, and ABA-trans-diol
a catabolite of ABA. Feeding studies with the synthetic deuterium-
labeled ABA analogs ABA-alcohol, ABA-aldehyde, and ABA-trans-diol
also contributed to understanding the biosynthetic relationships of
these compounds in plants. ”O Labeling and FAB-MS analysis of
epoxy-carotenoids has provided additional support for the indirect
pathway of ABA biosynthesis.

By applying these tools to the study of ABA biosynthesis in
ABA-deficient mutants, the biosynthetic blocks in the tomato, potato,
barley, and Arabidapsis mutants have been identified. The flacca
and sitiens mutants of tomato and aba of Arabidapsis provide a clear
picture of two important components of the ABA pathway: a minor
Shunt pathway from ABA-aldehyde --> ABA-alcohol --> ABA, and the
indirect pathway of ABA biosynthesis from epoxy-carotenoids.

Biosynthesis of trans-ABA is less clear, but the aba mutant provides

154

support for a parallel pathway from all-trans-epoxycarotenoids, and
the tomato mutants Show that ABA-aldehyde functions as a precursor
to mans-ABA, albeit a minor pathway. The experimental approach
documented here is applicable to the characterization of novel ABA-
deficient mutants, which are clearly needed in order to further
characterize ABA biosynthesis.

What still remains as a large gap in our understanding of ABA
biosynthesis is the characterization of the "cleavage enzyme," which
specifically cleaves epoxy-carotenoids and is the most likely
biosynthetic step to be regulated by water Stress and other
environmental and deveIOpmental Signals. The xanthoxin and ABA-
aldehyde oxidase activities are not induced by water Stress (Sindhu
and Walton, 1987; Sindhu et a1., 1990). Another aspect of ABA
biosynthesis which warrants attention is the possible role of
xanthophyll isomerases (Li and Walton, 1990), which would convert
all-fians-epoxy-carotenoids to 9-c1s-epoxy-carotenoids. The
existence of a xanthoxin isomerase, which would convert the
prevalent 2-trans-xanthoxin to cis-xanthoxin, should not be
discounted.

The long term objective of research into ABA biosynthesis is to

 

 

155

understand the role of ABA in growth and development.
Characterization of the enzymes and genes involved in ABA
biosynthesis will allow analysis of the molecular mechanisms which
regulate plant responses to environmental and developmental
Signals, and provide a means to critically examine the role of ABA in

physiological processes.

 

8.1 LITERATURE CITED

Li Y, Walton DC (1990) Violaxanthin is an abscisic acid precursor in
water-stressed dark-grown bean leaves. Plant Physiol 92: 551-
559

Sindhu RK, Walton DC (1987) Conversion of xanthoxin to abscisic
acid by cell-free preparations from bean leaves. Plant Physiol
85: 916-921

Sindhu RK, Griffin DH, Walton DC (1990) Abscisic aldehyde is an
intermediate in the enzymatic conversion of xanthoxin to

abscisic acid in Phasealus vulgan's L. leaves. Plant Physiol 93:
689-694

APPENDICES

156

APPENDIX A
Table A1. Partition Coefficients of ABA-Aldehyde, ABA-trans-Diol and ABA in
Three Different Solvents as a Function of pH

K, = [soluteNMMJ/[solutemwm]. Compounds [10 ug of (3:)-ABA-aldehyde
and (:)-ABA-trans-diol, or 50,000 dpm (:)-ABA (Specific activity 110 Ci/mmol)]
were added to 10 mL 0.2 M potassium phosphate buffer having the specified pH
and partitioned once against an equal volume of solvent, which was collected
and dried. The aqueous phase was then adjusted with l N HCI to pH 3.0 (for
ABA and ABA-trans-diol) and partitioned five times with 5 mL ethyl acetate,
which was pooled and dried. ABA-trans-diol was quantified by CC-electron
capture detection, and ABA-aldehyde was quantified by GC-flame ionization
detection with eicosane as an internal standard. ABA was measured by
scintillation counting. The measured de for ABA correlate well with similar
determinations by Ciha et a1.II and Neill and Horgan”. The K, of ABA-aldehyde
was not appreciably affected by pH and is smaller (compound more
hydrophobic) than that of ABA. ABA-aldehyde was found to be labile in the
presence of strong acid or base (data not shown). The K, of ABA-trans-diol was
greater titan that of ABA (compound more hydrophyllic) and is strongly affected
by pH.

 

Solvent/pH Kd

 

ABA-aldehyde ABA-trans-diol ABA

 

Diethyl ether
pH 3.5 < 0.1 0.1 0.2
pH 5.0 0.1 0.4 0.4
pH 7.5 0.1 218.0 20.6
pH 8.2 0.1 00° 27.2

pH 10.0 0.1 00 28.6

 

157
Table A1 (cont’d.)

 

Solvent/ pH

M
pH 3.5

pH 5.0
pH 7.5
pH 8.2
pH 10.0
n-Hexanes
pH 3.5
pH 5.0
pH 7.5
pH 8.2
pH 10.0

 

ABA-aldehyde ABA-trans-diol ABA
< 0.1 < 0.1 0.1
< 0.1 0.6 0.1

0.1 242.0 6.8
0.1 00 13.0
< 0.1 00 27.7
NI).d 00 76.9
19.1 00 116.3
15.5 00 100.0
11.9 00 100.0
4.2 00 125.0

 

' Ciha AI, Brenner ML, Brun WA (1977) Rapid separation and quantification of

abscisic acid from plant tissues using high performance liquid
chromatography. Plant Physiol. 69: 821-826.

b Neill S], Horgan R (1987) Abscisic acid and related compounds. In L Rivier, A

Crozier, eds, The Principles and Practice of Plant Hormone Analysis.

London, Academic, pp 111-167.

° 00 > 1000

'3 Not determined, recovery too low.

158
APPENDIX B

Table 8.1. Stability of Deuterium Label in ABA-Aldehyde and ABA-trans-Diol at
Various pHs

Compounds (10 pg) were purified by reverse phase HPLC or incubated
at room temperature in the dark for Six d in 0.2 M potassium phosphate buffer.
The solution was adjusted to pH 7.0 for ABA-aldehdye, or pH 3.0 for ABA-trans-
diol, and the compounds partitioned into ethyl acetate. Samples were analyzed

by GC-negative chemical ionization-selected ion monitoring. Because little

 

exchange occurred in the labeled compounds over extended periods of time at
pH extremes, it was concluded that the labeled compounds are suitable for use

as internal standards for quantitation and as substrates for in viva feeding

 

 

 

 

experiments.
Treatment [2H6] ABA-Al de hyde [2H7] ABA- Trans-Dial
% m/z=248 % m/z=254 % m/z= 280 % m/z=287

Stack solution 0.1 49.7 0.2 56.5
HPLC-purified 0.7 43.5 0.1 57.3
pH 2.8 1.5 35.9 0.2 58.2
pH 5.0 0.5 39.3 0.1 55.0
pH 7.5 1.9 34.1 0.1 56.2

pH 8.2 1.0 45.3 0.1 55.6

 

1 59
APPENDIX C

Table C.1. Correction Factors for Carotenoid Integration Data

The integration data obtained from normal phase HPLC of carotenoid
extracts monitored at 450 nm were multiplied by these scalars (Beer’s Law is a
linear function) to correct for differences in Specific extinction coefficients and
absorbance at 450 nm. Differences in absorbance at [Mm were calculated
from absorbance spectra taken by a Hewlett-Packard 1040M photodiode array
detector (kindly provided by Professor Derek Lamport, Michigan State

University) which was online with the HPLC effluent.

 

 

Carotenoid Scalar
B—Carotene 1
Lutein and Isomers 1.0477
Zeaxanthin and Isomers 1.0378
Antheraxanthin 1.1313
Violaxanthin 1.2169
9- Cis-Wolaxanthin 1 .6576

Neoxanthin and 9’-Cis-Neoxanthin 1.8391

 

 

1 60
APPENDIX D

CALCULATIONS OF NATURAL ISOTOPE CONTRIBUTIONS
TO MASS SPECTRAL DATA

D.1. Mass Spectral and SIM Data

The standard curve of ABA-Me used to quantify unlabelled and
”O-labeled ABA from plant extracts by GC-SIM does not include
data on the naturally abundant heavy isotopes, Le. only the
molecular ion abundance (m/z = 278) is integrated over the
retention time of the ABA-Me peak. Therefore, to accurately quantify
[”O,,]ABA-Me (m/z = 280, 282, 284; n = 1-3) from in viva labeling
experiments, the theoretical contribution of naturally occurring 13C
and ”O isotOpes to the heavy classes of ABA-Me must be
subtracted. Conversely, when a known amount (determined
spectraphotometrlcally) of an ABA biosynthetic precursor is fed to
plant material and the resultant ABA quantified, the heavy isotope
contribution must be calculated and added to the measured quantity
of ABA in order to accurately determine the biosynthetic capacity of
the tissue.

The stable isotOpe l3C currenfly comprises 1.115% of the

atrnospheric CO2 which is fixed by photosynthesis. Because plants

 

161

discriminate against ”C fixation at a level less than 3 percent
(Farquhar et a1., 1989), the estimate of 1.115% ”C is satisfactory. The
probability that any ABA-Me molecule (16 carbon atoms) contains
two atoms of 1“C can be calculated from a binomial expansion of the
natural abundances of ”C and 13C. The percent contribution of ”C
at mass (M + 2)‘ relative to the nominal mass is approximately equal

to 0.0060 times the square of the number of carbon atoms present in

 

the molecule (Watson, 1985). Therefore, each heavy class of ABA-
Me [(M + 2)‘, (M + 4)’, (M + 6)‘; m/z = 280, 282, 284] has a ”C
contribution to its ion abundance equal to 1.536% [= 0.006(16)2] of
the correSponding ion abundance of the class which is two atomic
mass units lighter. By subtracting from m/z = 282 the ”C
contribution by m/z = 280, one eliminates bana fide ”O-labeled ABA-
Me that was in fact [”CzleOIABA-Me. When quantifying [”OIABA,
this underestimation of [”OIABA is small and negligible (< 1.536%).
The naturally abundant ”0 comprises 0.2% of atmospheric oxygen
and therefore contributes an additional 0.8% of the ion abundance at
m/z = 278 to the abundance of ABA-Me m/z = 280 (because there
are four oxygens in ABA-Me). Contributions to [”OZIABA and

[”OflABA due to naturally occurring ”O are not included because

162

the ”O labeling experiments preclude this effect.

As an example, Table D1 shows the measured and
transformed data of an [”OIABA sample from the experiment
reported in Chapter 7 (Table 7 .2). The individual classes of
[”OIABA (one, two, or three ”O atoms) can be expressed as a
percentage of total [”OIABA, which is an important parameter when
analyzing MS/MS data.

D.2. MS/‘MS Data

Interpretation of MS/MS data is also complicated by ”C
isotOpe effects. When parent ions are selected in a triple
quadrupole mass Spectrometer to determine the position of ”O
labeling, ABA-Me molecules containing two 1“C atoms are also
included in the analysis. Their relative abundance in the daughter
ion spectrum is a result of the abundance of ABA-Me molecules
which are two atomic mass units lighter than the parent ion. When
there is only Slight ”O enrichment in ABA-Me, the ”C effects are
proportionally large.

Quantification by MS/MS of [”OIABA labeling patterns is
confined to ”O incorporation in the side chain of ABA-Me because of

the lack of proton extraction effects and the abundance of ions

 

163

Table D.l. Example of SIM Data Correction for 13C and ”0 Natural Abundance

 

 

Contributions
m / 2 Measured Correction Transformed
Area (A) Term Area
278 59.420 none 59.420
280 14.034 Am(0.02336) 12.646
282 1.606 AW(0.01536) 1.390
284 0.034 Am(0.01536) 0.009

 

% ["OIABA of total = 19.12%
% m/z 280 of [”OIABA = 90.04%
% m/z 282 of [”OIABA = 9.90%

% m/z 284 of [”OIABA = 0.06%

 

164
produced (daughter fragment m/z = 141; see Heath et a1., 1990). If

”0 is incorporated in the side chain of [”OIABA-Me, the Side chain
daughter ion is Shifted to m/z = 143 or 145 for one or two ”O atoms,
respectively. The presence of ABA-Me molecules with two ”C
atoms produces side chain fragment ions with m/z = 141, 142, and
143 because there is a distribution (assumed random) of the two 1"C
atoms in the carbon skeleton of ABA-Me. Furthermore, ions at m/z
-— 143, 144, and 145 can be the result of one ”O atom and two 1"C
atoms in the ABA-Me molecule. This mixed labeling of ABA-Me by
experiment-derived ”O and naturally abundant ”C is subtracted from
the data to allow calculation of ”O labeling patterns of the nominal
mass of ABA-Me plus ”O. This method assumes the labeling
patterns of ”C-containing ABA-Me are identical to those of unlabeled
material (i.e., no isotOpe discrimination by ABA biosynthesis
enzymes).

The observed MS/MS data for m/z = 142 and 144 (one ”C in
the Side chain) can be used to calculate the theoretical ”C
contributions to the ion abundance at m/z = 141, 143, and 146. The
number of possible positions of two 13C atoms in ABA-Me is

determined by a "combinations formula" which gives the number of

 

165

different combinations of 22 objects taken r at a time (different orders

of the same r not counted separately).

C“: n!
(n - r)! r!

For ABA-Me (16 carbons, n = 16), there are 120 different ways to
have two 1“C atoms (r = 2) in the same molecule:
16! = 120 .

(16-2)!21
For the side chain of ABA-Me (n = 7), there are 21 possible ways to
have two ”C atoms. For the ring moiety of ABA-Me (n = 9), there
are 36 possible ways to have two ”C atoms. Therefore, the number
of possible ways to have one ”C atom in the ring and one in the
Side chain is the difference between the total possible ways to have
two ”C atoms in ABA-Me (120) and the possible ways to have both
”C atoms in either the ring or the Side chain (21 + 36), which equals
63. The probabilities for each possible side chain labeling pattern
(zero, one or two 13C atoms) are calculated as follows.

1) The probability that one 1“O atom is in the Side chain = 63/ 120.

166

2) The probability that two 13C atoms are in the Side chain =

21/120.

3) The probability that zero ”C atoms are in the Side chain (both in
the ring) = 36/120.

Now the ratio of probabilities can be used as a scalar of the

observed ion intensity at m/z = 142 (or 144) to calculate the

theoretical contribution of ion intensity at m/z = 141 and 143 (or 143

and 145). A simple linear equation is set up, where theoretical

probability equals the observations:

n 0 cont' ution to ion abundance t 14 =
Observed ion abundance at m/z 142

Probability of ”C contributing to mzz 141
Probability of 13C contributing to m/z 142

or, rearranged:

Unknown contribution to ion abundance at m/z 141 =

[Observed ion abundance at m/z 142] [(36/ 120)/(63/ 120)]

= 0.57l[Observed ion abundance at m/z 142)].

167
Similarly, the unknown contribution to ion abundance at m/z 143 =

[Observed ion abundance at m/z 142] [(21/ 120)/(63/ 120)]

= O.333[Ion abundance at m/z 142].

In Table D2 the measured data from Table 7.2 are given as an
example of the corrections made to allow calculation of the extent
and position of ”O labeling as a percentage of total [”OJABA. Note
that in this example m/z = 284 data are not included. The MS/MS
data of a given mass is ‘Weighted" by multiplying its percentage
contribution to the total [”OIABA (e. 9:, m/z 280 =90.04%, from Table
D1). The results are shown in Table D.3.

D.3. FAB-MS of Carotenoids

Because carotenoids have 40 carbon atoms and FAB-MS
generates a Significant amount of proton adducts (Watson, 1985), the
isotOpe effect is a large factor when quantifying ”O enrichment of
carotenoids. The proton adduct effect, which is unique for each
carotenoid structure (data not Shown), precludes an accurate
measurement of ”O enrichment from data on the molecular ion
cluster. Therefore, the assumption is made that loss of protons is

negligible and that adduct formation is not dependent on mass (no

168

Table D2. Example of MS/MS Data Correction for ”C Natural Abundance

 

 

 

 

 

Contributions
m/z Observed Transformed Percent
m/z= 141 m/z= 142 m/z= 143 m/z= 141 m/z= 143 side
ring chain
280 125242 871 1 74301 120268 71400 62.75 37.25
282 2022 N.D.' 2031 l 2022 2031 1 9.05 90.95

 

' Not detected.

169
Table D.3. Calculation of ABA Labeling Patterns from Corrected SIM and MS/MS

Data

 

Percent of Total [”OIABA

 

 

 

 

Unlabeled in One 1"0 Atom in
Carboxyl Carboxyl

one in two in zero in one in two in
ring ring ring ring ring

 

(0.9004)(6Z.75) (0.0990)(9.05) (0.9004)(37.26) (0.0990)(37.25) trace

= 56.50 0.90 33.54 3.69 trace

 

 

 

170

isotope effects); ”O incorporation can then be determined by
empirical methods with a precision of a few percent. This is
accomplished by analysis of unlabeled and ”O-labeled carotenoid
samples by FAB-MS under identical ionization conditions. The
measured data for the molecular ion cluster are normalized (9.9:, as
percentage of nominal mass abundance). Then two transformations
of the raw data are performed: a) the correction for proton adduct
forrnafion, and b) the correction for 13C isotope contributions to ”O-
labeled ions.

In order to correct the observed data for proton adduct
formation of each carotenoid, a scalar factor, acmmmu, representing
the fractional contribution of the molecular ion (M+) to the ion
abundance at (M + 1)+, is calculated from the observed ion
abundance at (M + 1)+ of a given unlabeled carotenoid, minus the

theoretical contribution ( 44.6% of M+) from ”C:

%(M + Human,“ - 44.6

acarotenoid =

 

100

This scalar is then used to transform the measured data so that an

accurate estimate of actual ”O-labeled and unlabeled compounds is

—-'

 

 

171

obtained. For the unlabeled nominal mass the proton adduct

contribution to (M + H)+ is added back to Mummy

+actual ._ + +
M _' M observed + acarotenoidlv‘t observed '

For labeled masses, only ”O-labeled ion abundances are desired.
Therefore, the observed ion abundance from the unlabeled control is

subtracted from the observed ion abundance of the labeled sample

 

and corrected for proton adduct affects:

(M ‘1' x)labeled corrected, rt: 2 «>8
= (M ‘1' X)labeled observed ' (M + x)unlabeled observed
' 3(M + X ' l)labeledcorracted -
The transformations proceed sequentially from (M + 2)+ to (M +
8)+, if the carotenoid has four oxygen atoms. Note that this
correction does not apply to the (M + H)+ ion; the (M + m+,,,,.,,,,
comm value is 44.6 in the (M + $33,313,, com,“ calculation.

After the data are transformed to corrected ”O ion
abundances, the ”C isotOpe contribution to the data can be
subtracted (step b). It is assumed that the ”O labeling patterns of
”C-labeled carotenoids are identical to the nominal mass ”O

labeling patterns (no isotope discrimination). Thus the calculation of

 

 

172
carotenoid labeling patterns is Simplified and comparable to the ABA

calculations. Based on the theoretical l3C contribution to (M + 2)+
[402(0.006) = 0.096], one can subtract 9.6% of the corrected ”0 ion
abundance at (M + 2)+, (M + 4)*, and (M + 6)+ from the ion
abundance at (M + 4)*, (M + 6)*, and (M + 8)+, respectively [the
(M + 2)+ data were already corrected empirically by subtraction].
Now the actual percent ”O incorporation can be calculated.

Only "actual" data for M+, (M + 2)+, (M + 4)+, (M + 6)+, and (M +
8)+ are neccessary, because only nominal masses plus experimental
”O masses are included.

% enrichment = lOOIIM + 2) +IM + 4) +IM + 6) + (M + 81L

M+(M+2)+(M+4)+(M+6)+(M+8).

Enrichment of individual masses can also be‘calculated as a
percentage of total ”O-labeled sample from these corrected data.

An example of these calculations is given in Table D4.

D4. LITERATURE CITED

Heath TG, Gage DA, Zeevaart IAD, Watson IT (1990) Role of
molecular oxygen in fragmentation processes of abscisic acid
methyl ester in electron capture negative ionization. Org Mass
Spectrom 25: 655-663

173

Table D4. Example of [”O]Violaxanthin FAB-MS Data Correction for ”C Natural
Abundance and Proton Adduct Contributions

 

 

Sample Observed Corrected" Actualb
Unlabeled Violaxanthin
M+ 100 124.1
(M + H)" 68.7 ---
(M + 2)+ 22.0 ---
(M + 3)+ 5.8
(M + 4)+ 1.5 «-
Labeled Violaxanthin
M+ 100 124.1 124.1
(M + H)+ 72.2 44.6
(M + 2)+ 46.4 13.6 13.6
(M + 3)+ 30.3 21.2
(M + 4)+ 93.4 86.8 86.6
(M + 5)+ 59.6 38.7
(M + 6)+ 21.6 12.3 4.0
(M + 7)+ 5.6 2.6
(M + 8)+ 3.6 3.0 1.8

 

‘ Calculated as: (M + x)hboledconoctod.x=2-->8 =

(M + x)laboledoburvod " (M '1' X)mu.beiedobsemd ’ 3(M ‘1' x " l)lnboled corrected Where aviohnnthin
= 0.687 - 0.446 = 0.241. See (a) in text.

b Calculated as (M + 1‘).ch , «as = (M + x)co,,,c,,d - 0.096(M + x - 2)com.d.
Actual percent enrichment for this sample = 45.8% (see step b, text).

 

 

174

Farquhar GD, Ehleringer IR, Hubick KT (1989) Carbon isotope
discrimination and photosynthesis. Annu Rev Plant Physiol
Plant Mol Biol 40: 503-537

Watson IT (1985) Introduction to Mass Spectrometry, 2nd ed. Raven
Press, New York

"‘llllllllllilllilEs