Development of radioimmunoassays for measurement of abscisic acid and gibberellins during floral inductive treatments in...

MISSING IMAGE

Material Information

Title:
Development of radioimmunoassays for measurement of abscisic acid and gibberellins during floral inductive treatments in Citrus latifolia Tan
Uncontrolled:
Citrus latifolia
Physical Description:
x, 147 leaves : ill. ; 28 cm.
Language:
English
Creator:
Southwick, Stephen Mark
Publication Date:

Subjects

Subjects / Keywords:
Citrus fruits -- Physiology   ( lcsh )
Citrus fruits -- Flowering time   ( lcsh )
Abscisic acid   ( lcsh )
Gibberellins   ( lcsh )
Horticultural Science thesis Ph. D
Dissertations, Academic -- Horticultural Science -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1986.
Bibliography:
Bibliography: leaves 137-146.
Statement of Responsibility:
by Stephen Mark Southwick.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029609085
oclc - 15338798
System ID:
AA00017642:00001


This item is only available as the following downloads:


Full Text










DEVELOPMENT OF RADIOIMMUNOASSAYS FOR MEASUREMENT OF
ABSCISIC ACID AND GIBBERELLINS DURING FLORAL INDUCTIVE
TREATMENTS IN Citrus latifolia Tan.





By

STEPHEN MARK SOUTHWICK

























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






UNIVERSITY OF FLORIDA

1986


































Copyright 1986

by

Stephen M. Southwick












ACKNOWLEDGMENTS


Mr. Southwick would like to thank Dr. Thomas L.

Davenport for serving as chairman and Drs. Frederick S.

Davies, Donald J. Huber, Richard C. Smith, and R. H. Biggs

for serving as doctoral committee members and reviewing this

dissertation work. Mr. Southwick would like to extend

especial thanks to Alfred Chung, Dr. Albert Castro, and Dr.

James Ryan at the University of Miami Medical School and Dr.

Ron Block at Mount Sinai Hospital in Miami Beach for

outstanding technical and moral support. Without their help

Mr. Southwick would not have been able to learn the

necessary skills required to prepare and complete the

thorough job required for this dissertation.

Finally, Mr. Southwick would like to thank Regina

Karner and his family because they have helped in every

possible way so that Mr. Southwick could reach his goals.









TABLE OF CONTENTS


Pace


ACKNOWLEDGMENTS . . .

LIST OF TABLES . . .

LIST OF FIGURES . . .


ABSTRACT . .


CHAPTER

I INTRODUCTION . 1

II LITERATURE REVIEW .. . 3
Citrus Flowering . 3
Environmental Control of Flowering 4
Internal Control of Flowering 5
Plant Growth Regulator Immunoassays 8
Principles of the Immunoassay 9
Gibberellin Immunoassay .. 11
Abscisic Acid Immunoassay .. 11
Auxin and Cytokinin Immunoassay ... 12


III CHARACTERIZATION OF WATER STRESS AND
LOW TEMPERATURE EFFECTS ON FLOWER
INDUCTION IN CITRUS . ... 14

Introduction . 14
Materials and Methods . .. 15
Results and Discussion ... 17

IV INVESTIGATIONS OF HORMONAL CONTROL OF
CITRUS FLOWERING: DEVELOPMENT OF A
RADIOIMMUNOASSAY FOR THE MEASUREMENT OF
ABSCISIC ACID IN LEAVES AND BUDS OF
TAHITI' LIME . ... 33

Introduction . 33
Materials and Methods . .. 35
Results . ... 41
Discussion .... .. ... .. .... 60


viI


ix











V INVESTIGATIONS OF HORMONAL CONTROL OF
CITRUS FLOWERING: DEVELOPMENT OF A
RADIOIMMUNOASSAY FOR THE MEASUREMENT OF
GIBBERELLIN LEVELS IN LEAVES AND BUDS OF
'TAHITI' LIME . .. 71

Introduction . 71


Materials and Methods ... 73

Results . .. 88
Discussion ............. 124

VI SUMMARY AND CONCLUSIONS .. 132

LITERATURE CITED . ... 137

BIOGRAPHICAL SKETCH . .. 147









LIST OF TABLES


Table Page

III.1 Effect of continuous or cyclical
water stress on flower induction
in containerized 'Tahiti' lime
trees . . 19

111.2 Effect of moderate water stress
over time on leaf xylem pressure
potential and flower induction in
Tahiti' lime . .... 21

111.3 Effect of severe water stress over
time on leaf xylem pressure
potential and flower induction in
'Tahiti' lime . ... 24

111.4 Effect of low temperature over time
on leaf xylem pressure potential in
'Tahiti' lime . ... 27

111.5 Effect of low temperature over time
on flower induction in 'Tahiti'
lime . . 28

111.6 Effect of leaves and misting on
flowering of immature 'Tahiti' lime
cuttings . . 30

IV.1 The specificity of antiserum to
abscisic acid . ... 47

V.1 The specificity of antiserum to
gibberellins . ... .115












LIST OF FIGURES

Figure Page

IV.1 Standard curve for the (+) ABA
radioimmunoassay constructed from
n=20 consecutive assays to show day-
to-day reproducibility. The bars
indicate standard deviations of
triplicate samples . ... 42

IV.2 Linearized logit-log plot of the
standard curve for the (+) ABA
radioimmunoassay . ... 44

IV.3 Extract dilution analysis ... 48

IV.4 Effect of duration of water and low
temperature stress on total ABA
levels in 'Tahiti' lime leaves as
measured by the (+) ABA
radioimmunoassay . ... 51

IV.5 Effect of duration of water and low
temperature stress on free ABA
levels in 'Tahiti' lime leaves as
measured by the (+) ABA
radioimmunoassay . ... 53

IV.6 Effect of duration of water and low
temperature stress on total ABA
levels in 'Tahiti' lime buds as
measured by the (+) ABA
radioimmunoassay . ... 56

IV.7 Effect of duration of water and low
temperature stress on free ABA
levels in 'Tahiti' lime buds as
measured by the (+) radioimmunoassay 58

IV.8 NMR spectra of B-D-glycopyranosyl
abscisate tetraacetate in deuterated
DMSO . . .. 63

IV.9 NMR spectra of B-D-glucopyranosyl
abscisate tetraacetate in deuterated
DMSO . . ... 65









V. 1 Standard curve of the concentration
of GA3 (M, GA dissolved in
concentrated A2SO ) versus
absorbance (nm) used to calculate
molar coupling ratios of GA3-BSA
conjugates . . 75

V.2 Scheme of synthesis of GA3-BSA
conjugate by the method of Weiler
and Wieczorek (109) . .. 89

V.3 Scheme of synthesis of GA3-BSA
conjugate by the symmetrical
anhydride procedure of Atzorn and
Weiler (4) . ... 91

V.4 Scheme of synthesis for the
preparation of GA3-BSA by use of
hydroxysuccinimide to create GA3-
active ester . 94

V.5 Scheme of synthesis for the
preparation of GA3-amino-n-caproic
acid-BSA . . 96

V.6 Scheme of synthesis for the
preparation of GA4-adipic acid-BSA
. . 105

V.7 Standard curve for the GA4
radioimmunoassay constructed from
n=20 consecutive assays to show day-
to-day reproducibility. . 111

V.8 Linearized logit-log plot of the
standard curve for the GA4
radioimmunoassay . .. 113

V.9 Extract dilution analysis ... 117

V.10 Effect of time of water and low
temperature stress on GA levels in
'Tahiti' lime leaves as measured by
the GA4 radioimmunoassay ... .120

V.11 Effect of time of water and low
temperature stress on GA levels in
'Tahiti' lime buds as measured by
the GA4 radioimmunoassay ... .122


viii








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


DEVELOPMENT OF RADIOIMMUNOASSAYS FOR MEASUREMENT OF
ABSCISIC ACID AND GIBBERELLINS DURING FLORAL INDUCTIVE
TREATMENTS IN Citrus latifolia Tan.

By

Stephen M. Southwick

August, 1986

Chairman: Dr. T.L. Davenport
Major Department: Horticultural Science (Fruit Crops)


The hormonal control of citrus flowering was investi-

gated by utilizing several approaches. Water stress for as

little as 2 weeks induced flowering in container grown

'Tahiti' lime (Citrus latifolia Tan.) trees. Low

temperature (180C day/100C night) induced a time dependent

flowering response like that of moderate water stress. Low

leaf xylem pressure potentials, as compared to controls were

found only under water stress treatment, suggesting that a

common stress-linked event, separate from low plant water

potential, is involved in floral induction. Leafless,

immature cuttings were induced to flower by water stress

treatment, suggesting that leaves are not essential for a

floral inductive response.

Total abscisic acid (ABA) levels in leaves and buds in-

creased as the duration of floral inductive stress increased

and were higher under water than low temperature stress

conditions as measured by radioimmunoassay (RIA). Total ABA

ix








levels decreased when water stress was alleviated by

rewatering. Free ABA in leaves changed in a similar pattern

to that noted for total ABA in trees grown under the above

stress conditions. Levels of free ABA were less after

water or low temperature than before stress and fell sharply

after alleviating water stress in trees. Levels of free ABA

were 4- to 35-fold lower than total ABA.

Antisera against gibberellins (GA) were produced by

immunizing rabbits with a GA4-adipic acid-bovine serum

albumin immunogen. The antisera significantly cross reacted

with GAI, GA7, and GA9. Gibberellin levels in leaves and

buds collected from trees grown under water stress were

dissimilar to those levels measured in low temperature

stressed tissues as measured by GA RIA. Gibberellin levels

decreased in leaves and increased in buds when water stress

was alleviated by rewatering. The increased GA levels in

buds corresponded to the alleviation of water stress.

Gibberellins and ABA may play a role in water stress-induced

bud dormancy and budbreak. Water and low temperature stress

treatments induce flowering in 'Tahiti' lime, but the levels

of ABA and GA do not change in a similar manner as a result

of stress treatment.










CHAPTER I
INTRODUCTION


Flowering in citrus trees is essential for the

production of citrus fruit and is an important event for the

maintenance of a world-wide commercial horticultural

enterprise. Unfortunately, the flowering process of citrus

trees is understood to a limited extent. Environmental

factors such as temperature and photoperiod as well as

cultural practices like gibberellic acid spray have gathered

a large share of the flowering research emphasis over the

last 30 years in citrus. Many of the citrus flowering

studies have been conducted with field grown citrus trees

where the floral inductive conditions were poorly controlled

or characterized. In addition, more than one floral

inductive treatment had not been used in these previous

studies to compare plant's physiological and biochemical

similarities and differences in the flowering process. The

perception and integration of these environmental or

cultural factors into internal biochemical control have not

been investigated to any significant extent.

In the past, it has been difficult to study many of the

internal biochemical controls governing the flowering

process. Plant growth regulators (PGRs) which were thought

to exert great control over flowering in citrus were

inaccesible in the chemically complex tissues of the citrus

tree and it was impossible to identify and quantify subtle

PGR changes in those tissues where the flowering process may







be subject to environmental control. The recent

developments of PGR immunoassays have allowed for the

investigation of changing PGR levels in limited quantities

of plant tissues with little purification of plant extracts

prior to assay. However, these techniques were limited to

only a few select laboratories. Our goal was to develop

immunoassays in order to quantify those PGRs which were

thought to play a controlling role in the flowering process

of citrus. The approach taken in the experiments reported

here was to develop and characterize several floral

inductive treatments in container grown 'Tahiti' lime trees.

Container grown trees were used because they allowed for

control of floral inductive conditions in the greenhouse and

growth chamber. Moreover, plant physiological and bio-

chemical similarities and/or differences among several

floral inductive treatments and PGR levels could be

correlated and a theory for PGR involvement in the flowering

process of citrus could be developed. The success of this

approach would be helpful toward gaining a better under-

standing of citrus flowering.










CHAPTER II
LITERATURE REVIEW


Citrus Flowering

Flowering is the first step toward annual citrus fruit

production. The general sequence of events that lead to

citrus flowering under subtropical conditions in the

northern hemisphere is flower induction, which generally

occurs in early winter (December through January);

differentiation, which occurs in late January; and the

uninterrupted development of floral organs, which leads to

the opening of flowers in March through April (1, 5, 72).

However, floral induction has recently been reported to

occur in Citrus sinensis as early as November (62). Under

the climatic conditions of the tropics, some citrus

cultivars produce flowers throughout the year (72). Lemon

is the most widely known example of this behavior. Flower

induction apparently occurs more than once a year. The

stimuli regulating this response are not clear; however,

water stress is commercially used to induce flowering in

lemon (73). There have been a substantial number of studies

conducted that relate to flowering of citrus and are

reviewed in Volume II of the Citrus Industry (13, 18). The

specific control of citrus flowering is not well understood;

however, it is generally believed that plant growth

regulators (PGRs) are involved. The relevant literature

concerning environmental and internal control of flowering

in citrus is reviewed here. Moreover, this review will







cover literature that is related to the development of

immunological assays that have been used in these studies to

help elucidate the role of PGRs in the internal control of

citrus flowering.


Environmental Control of Flowering

Environmental control of flowering in citrus is

achieved climatically or through the cultural practices of

the grower. The most commonly believed environmental

controller of flowering in citrus is thought to be exposure

to periods of reduced temperature during the winter months

(38, 59, 72, 74, 78, 79). These findings have resulted from

both greenhouse and growth chamber experiments as well as

from observations of trees growing in the field. The

temperatures required to induce flowering range from 15 to

200C (day) and 5 to 150C (night) (38, 59, 77), but the

critical amount of time needed to induce flowering under low

temperature conditions has not been reported. It was

reported that air temperature was more important for

regulation of flowering than root temperature in Citrus

sinensis (38, 79). Attempts were made to induce flowering

in Citrus sinensis by utilizing low root temperatures (11

and 150C) together with non-inducive air temperatures (22

and 270C). Flowering could not be induced by low root

temperature and the lack of flowering was not correlated

with a general inhibition of growth, which had been due to

low root temperature (79). High temperature above 300C

inhibited flower formation under the growing conditions of

the growth chamber (77). There is good evidence in Citrus







that flowering is day neutral (77) and that changing photo-

period does not affect flowering. Flowering has been shown

to be controlled by photoperiod only under moderate tempera-

ture (240C day/190C night) conditions (59).

There are several practices which growers can use to

regulate flowering in citrus. Certain lemon (Citrus limon

(L.) Burm.f.) cultivars were induced to flower because of an

extended period of water stress (11, 82). Although it is

well known that general water stress in citrus promotes

flowering, there are only 2 reports in the literature

addressing the topic. These reports do not determine

whether water stress is truly a floral inductive treatment.

Moreover, the research has been conducted in the field with

mature trees where little control or characterization of

water stress could be achieved. Practices such as autumn

pruning (6), excessive nitrogen fertilization (56), applica-

tion of gibberellic acid (17, 35, 36, 37, 69, 72, 73, 75),

and excessive or late cropping (42, 43, 48, 49, 50, 61, 71)

can reduce flower bud formation and often the subsequent

yield in the following season. Although a reasonable amount

of research data have been generated with regard to the

environmental control of flowering in citrus, the amount of

research data concerning the endogenous, internal chemical

control of flowering is sparse.


Internal Control of Flowering

The exogenous application of gibberellic acid (GA) and

the inhibition of citrus flowering have led to the specula-

tion that GA controls flowering. Although it is true that







GA application can inhibit the formation of flowers in

citrus, there has been only one report of the measurement of

GA in citrus as it relates to flowering (35). In that

report a negative correlation was reported between length of

flowering branches and their flower load. Vegetative

branches are the longest and the levels of native GA-like

substances were highest in the vegetative flowerless

branches (35). As the number of leaves decreased and the

number of flowers increased (from vegetative to generative

shoots), the levels of GA-like substances also decreased.

Other circumstantial evidence implicates GA in control of

citrus flowering. The use of GA synthesis inhibitors such

as (2-chloroethyl)trimethylammonium chloride (chlormequat),

or succinic acid-2,2-dimethylhydrazide daminozidee) can

induce larger flower numbers under certain conditions (74,

75). This fact lends more credence to the theory of GA

control of flower formation.

The way in which GA may control flowering at the

present time is unknown and little research has been done in

citrus to clarify the problem. However, one report showed

that synthesis and total protein levels decreased in buds of

"Shamouti' orange trees during the flower formation period.

The application of GA during the same period resulted in

additional protein bands as compared to untreated controls

in polyacrylamide gel electrophoresis (76). It was

speculated that these additional bands may be proteins that

inhibited flowering.







The levels of carbohydrates are also thought by some to

be internal controllers of flowering in citrus (33, 42, 43,

48, 49, 50). The levels of carbohydrates in leaves were

found to decrease as fruit were left on trees later into the

season. The decrease in total carbohydrate levels

correlated with a reduction in flower formation, fruit set,

and yield the following year (42, 43 48, 50). Carbohydrate

levels were found to be associated with alternate bearing of

'Wilking' mandarin trees. Apparently, the levels of carbo-

hydrates increased due to fruit thinning along with the rate

of flower bud differentiation (33). The correlation was

best between fruiting and starch levels and poor with sugar

or total carbohydrate levels. Starch was found to accumu-

late in the roots of 'Wilking' mandarin in the "off" year of

an alternately bearing tree. However, Lewis et al. (61)

rejected the hypothesis that carbohydrate levels are

responsible for the lack of flowering in 'Wilking' mandarin

during the "off" years because they found that thinning

changed the production cycle without significantly affecting

carbohydrate levels. Similarly, Jones et al. (49) found

that the effectiveness of thinning treatments was not

clearly correlated with carbohydrate levels in 'Valencia

orange. There is disagreement in the conclusions drawn from

these studies and the discrepancies may have resulted from

using different plant tissues as well as differing

techniques to measure carbohydrates.

It is well documented that carbohydrate levels do

change in leaves, roots, buds, etc. as a result of crop load







or during the floral inductive period in citrus; however, it

is unlikely that a carbohydrate can directly control a

complex process such as flowering. On the other hand,

gibberellic acid has been shown to inhibit flower formation

in citrus. It has been difficult to measure gibberellin

levels in the small amounts of tissue required for the

analysis of PGRs in citrus buds. The development of

immunoassays to quantify PGRs has the potential of over-

coming many of the problems associated with the measurement

of PGRs and the literature concerning the development of PGR

immunoassays is outlined below.


Plant Growth Regulator Immunoassays

The immunoassay was first developed by Berson and Yalow

(8) for the quantification of insulin in human plasma and

was soon recognized as a powerful tool in endocrinology,

clinical medicine and pharmacology. The technique allowed

for the quantification of exceedingly low levels of low and

high molecular weight compounds, after little sample

preparation, in small amounts of tissue in a short period of

time (8, 117, 118). Fuchs and coworkers (25, 26, 27, 28)

were the first to attempt immunological procedures for the

detection and quantification of PGRs. Since these early

studies, the scope and usefulness of PGR immunoassays have

increased and radioimmunoassays (RIAs) and enzyme-linked

immunoassays (EIA) to quantify levels of gibberellins (GA),

abscisic acid (ABA), cytokinins, and auxins (14, 108, 109,

110) have been developed.







Principles of the Immunoassay

The immunoassay is based upon the competition of a

known amount of labeled antigen and an unknown amount of

sample antigen for a finite number of high-affinity binding

sites (118). In order to develop a successful immunoassay

there must be a suitable availability of specific antiserum,

labeled antigen which exhibits a comparable affinity for the

antibody as the unlabeled antigen, and a convenient and

reliable procedure for separation of antibody-bound from

free antigen without interference of antibody equilibrium

(108, 109, 110).

Plant growth regulators are low molecular weight

compounds. Thus, PGRs must be chemically coupled to higher

molecular weight compounds such as albumins, keyhole limpet

haemocyananin, or thyroglobulins (proteins) to create PGR-

immunogens (19). The synthesis of the PGR-immunogens and

the choice of functional groups used in the synthetic

procedure depends upon the degree of antibody selectivity

required. A high degree of selectivity is obtained when the

PGR is most exposed after coupling to the larger molecular

weight compound (19). Examples of greater selectivity due

to the synthetic procedure used to create the PGR-immunogen

include the ABA carbon 4-coupled immunogens. These immuno-

gens have been created by inserting a chemical spacer

between the carbon 4 of ABA and the protein carrier bovine

serum albumin (104). This coupling procedure allowed anti-

sera to cross react only with free ABA rather than both free

and bound ABA which results from the coupling of ABA at







carbon 1 to albumin (101, 103). Another example of antisera

selectivity has been noted where immunogen synthesis via the

ribosyl moiety of cytokinin ribosides yields antisera

reactive with the free bases, 9-ribosides, 9-ribotides, and

9-glucosides, but which discriminate trans zeatin from cis

zeatin or dihydrozeatin (63, 105). Antisera with different

selectivities can be achieved by varying immunogen

synthesis. Each PGR has different functional groups through

which conjugation to large carrier molecules can occur.

Moreover, other techniques are available for introduction of

reactive functional groups to PGRs which lack them (19).

Once a suitable polyclonal or monoclonal antibody has

been obtained, it is important to select a suitable tracer

antigen. Radionuclides such as 3H (108, 109, 110) or 1251

106, 113) have been used previously for RIA. In the case of

ELISA the use of enzymes, notably alkaline phosphatase, have

been used to label PGRs (4, 14, 15, 107, 111, 112). After

the selection of the tracer has been made, it is important

to select the optimum assay conditions and characterize the

antibody and a number of reviews deal with these subjects

(55, 108, 109, 110). In some cases it has been necessary to

chemically modify the antigen prior to analysis in order to

maximize the sensitivity of the assay. When carboxyl groups

of the antigen have been used to synthesize PGR-immunogens

(e.g. auxin (IAA), GAs) it was necessary to methylate

antigens in order to restore immunoreactivity (3, 4, 111,

113). The reason methylation is necessary is not completely

clear; however, it has been speculated that the antigen







coupling site of the immunogen and the antigen-protein link

may constitute part of the overall binding area of the

antibodies (110) rather than only the antigen which does not

completely occupy the antibody binding site.


Gibberellin Immunoassay

The development of immunoassays for the quantification

of GAs in plant tissue extracts has been limited. Both an

RIA and ELISA have been developed for GA1, GA3, GA4, GA7,

GA9, and GA20 (3, 4, 112). These assays were reported to be

sensitive. The RIA could measure pmol amounts of GA3 in

unpurified plant extracts and the ELISA could measure fmol

amounts of either GA3, GA4, and GA7 methyl ester. Approxi-

mately 1 mg of plant tissue was required to measure GA by

the immunoassays. Immunogen synthesis was carried out

through the carbon 6, COOH functional group of GA, and the

immunogen resulted in the production of antisera that were

selective for specific GAs. Antisera against apolar GAs

proved to be the most selective (3, 4, 113). Only a limited

number of GAs have been used to produce GA-directed anti-

bodies. Future approaches to GA isolation and quanti-

fication by immunoassay could utilize group-selective assays

(109, 110) in combination with a suitable separation system

such as high pressure liquid chromatography.


Abscisic Acid Immunoassay

There have been several RIAs and ELISAs developed for

ABA (14, 15, 60, 67, 86, 101, 103, 104, 107, 112). In the

early assays developed against ABA, the immunogen was







prepared by coupling (+) ABA via carbon 1 to protein

carriers. This immunogen resulted in the production of

antisera that preferentially cross reacted with (-) ABA

rather than (+) ABA, which is the enantiomer found in

plants. This result rendered the assay imprecise. This

problem was overcome by using (+) ABA immunogens (104).

However, the C 1-coupled (+) ABA antisera detected both ABA

and ABA-conjugates (101, 103). In order to produce antisera

which only reacted with free ABA, the C 4 of ABA was coupled

to tyrosylhydrazone and the resulting compound was coupled

to albumin to form an immunogen. The antisera obtained

selectively recognized free ABA and not ABA-conjugates

(104). The same approach was used in the production of ABA

monoclonal antibodies and allowed for the detection of less

than 5 pg 2-cis(+) ABA (67). The antisera obtained against

ABA or ABA and its conjugates does not cross react with

phaseic acid, dihydrophaseic acid, xanthoxin, or other

compounds known to occur in plants and which have similar

chemical structure (67, 101, 103, 104).


Auxin and Cytokinin Immunoassay

Radioimmunoassays and ELISAs have been developed for

indoleacetic acid (IAA) (84, 85, 106, 111). The assays were

sensitive for IAA in the fmol range when IAA as a methyl

ester was used as a standard (111). The production of IAA

immunogens by the Mannich reaction (85) resulted in the

production of antisera that was selective for IAA, but low

serum titers were obtained. Antisera with higher titers

were obtained by coupling the IAA carboxyl to the protein






carrier. However, the antisera had less selectivity and

cross reacted with side chains similar to that of IAA, such

as indole-3-acetaldehyde (106, 111). Because of these

problems, plant extracts have required some prepurification

prior to quantification by IAA immunoassay (110).

The production of cytokinin immunogens is based on the

procedure of Erlanger and Beiser (20) where periodate-

oxidized ribosides are reacted with amino groups of the

carrier protein followed by borohydride reduction (20, 63,

98, 104). Using this type of immunogen, antisera have been

obtained that are capable of determining free cytokinin as

well as 9-substituted cytokinins (98, 105). Antisera have

also been produced that discriminate between trans-zeatin

type and isopentenyladenosine type cytokinins (98, 105).

These RIAs made use of 3H or 125I for tracers and were

sensitive for cytokinin derivatives in the fmol range. By

using alkaline phosphatase-labeled isopentenyladenosine and

trans-zeatin type cytokinins, an increase of 5- to 10-fold

in sensitivity could be achieved (110).

The development of immunoassays for PGRs has allowed

for sensitive and specific measurement of PGRs in small

amounts of plant tissue samples. In the future it is

apparent that many more assays for the quantification of

specific compounds in plants will become available. More-

over, it will be possible to use group selective antisera

together with high pressure liquid chromatography or other

chemical analytical separation techniques to quantify

closely related compounds such as cytokinins and

gibberellins.










CHAPTER III
CHARACTERIZATION OF WATER STRESS AND LOW TEMPERATURE
EFFECTS ON FLOWER INDUCTION IN CITRUS


Introduction

'Tahiti' lime, Citrus latifolia Tan., is a sterile

triploid that is vegetatively propagated thereby avoiding a

juvenile period that is common to many tree species.

Flowers are produced and fruit set and mature on leafy

rooted cuttings as well as on plants propagated by air-

layering. This unusual habit makes 'Tahiti' lime suitable

as a test plant to study flowering in trees (16, 17, 92).

Flowering in citrus can be induced by low temperature (16,

20, 59, 77, 78, 79, 92) or water stress (11, 66, 82, 92) and

inhibited by applied gibberellin (17, 35, 46, 70, 73, 82).

Regulation of flowering by water stress is not common in

trees and generally is reported to be effective in tropical

and subtropical species (2, 7, 11, 82). These studies,

however, have been conducted under varying field conditions

and are not descriptive (7) in the sense that the studies do

not define the quantitative relationship between imposed

stress and the flowering response. These same shortcomings

are true for studies involving low temperature regulation of

flower induction.

In our effort to understand the chemical control of

flowering in 'Tahiti' lime, we felt it important to

manipulate flowering in small container grown trees and to

have more than one method of flower induction available to





15
us. By utilizing several floral-inductive treatments, we

hope to ascertain whether a common regulatory event or

signal controls flower induction. In this chapter we

describe the quantitative relationship between low tempera-

ture and water stress on floral induction.


Materials and Methods

Plant Material. In most experiments we used 1- to 2-

year-old "Tahiti' lime trees propagated by air-layering or

by bud-grafting on to Citrus macrophylla Wester rootstock.

The trees ranged in height from 0.5 to 1.0 m and were grown

in the greenhouse under South Florida growing conditions

(80) in 16-cm black plastic pots in a mix of 1 peat:l

perlite:l sand, and fertilized regularly with a 20:20:20 NPK

soluble fertilizer plus micronutrients. In other experi-

ments, cuttings were obtained from 18-year-old 'Tahiti' lime

trees on rough lemon (Citrus jambhiri Lush.) rootstock in

Rockdale limestone soil at the Homestead, Tropical Research

and Education Center. Cuttings were selected by clipping

immature current season's growth, which bore mostly fully

expanded, but non-hardened, immature leaves, and were

reclipped near the base under water. Leaves were removed so

that 5 to 8 nodes and 2 to 4 or no leaves remained,

depending upon treatment. Cuttings with one end were stuck

in vermiculite and were placed either in the greenhouse or

on another bench under intermittent misting.

Water Stress Treatments. Water stress of trees was

either continuous or cyclical. Continuous water stress was

established by sealing the pot, soil and roots in a plastic







bag with only stem and leaves exposed to the environment and

withholding water. Transpirational water loss from each

container grown tree was determined by monitoring the daily

weight decrease. Lime trees were transpiring approximately

140 ml water/day. From these measurements 67 percent of the

amount of water lost per day was added back to the tree

daily so that stress could be gradually imposed and leaf

drop minimized. When all leaves became wilted and the

mature leaves had a xylem pressure potential of at least

-3.5 MPa (severe stress as defined by Syvertsen, 95), 100 ml

of water per day was added to each in order to approximately

replace transpirational water lost and to keep these trees

under constant stress. Addition of 100 ml of water per day

to soil briefly saturated the soil until it drained to

container field capacity. In absolute terms, continuous

water stress may not precisely define this sequence of

events, but continuous water stress best describes our

observations with regard to the water status of these trees.

Cyclical water stress was achieved by stressing each tree to

the point of wilting as above and then refilling the

container to the full capacity, which set the soil at field

capacity. The dry (wilting), wet (container soil at field

capacity) cycle or continual stress was continued for the

duration of each experiment. Control and treatment trees

were preconditioned for at least 1 month through maintenance

of container soil at field capacity by applying water twice

daily through automatic drip irrigation. Leaf xylem pres-

sure potentials were measured predawn and midday by the







pressure bomb technique (89). Leaves were removed at the

petiole-blade abscission zone of each leaf, and measurements

were made within 30 s of leaf removal. Leaf xylem pressure

potential measurements were made at weekly intervals in the

time course experiments, and 2 leaves per tree (10

leaves/treatment) were measured at predawn and midday. Only

2 leaves were used from each tree in order to maximize the

number of leaves remaining on stressed trees and maintain

uniformity of treatment.

Low Temperature Treatments. Growth chamber experiments

were conducted at 180C/100C (day/night) temperatures with

12-h photoperiods at a photon flux ranging from 350-850

uE/m2/s. Prior to placing each tree in the growth chamber,

approximately one-half of all branch apicies including 2 to

3 leaves and nodes were clipped off (16). Controls were

treated likewise and were grown in the greenhouse under

South Florida conditions (80) and 290C/240C (day/night)

temperatures. In these experiments, total shoots produced

represent the sum of vegetative, mixed, and generative

shoots. Those 3 shoot types are defined here as they have

been previously (35, 70, 77). Briefly, vegetative shoots

carry leaves only, mixed shoots carry both leaves and

flowers, and generative shoots carry flowers only. Tables

have been obtained from at least 2 replicate experiments in

all cases.


Results and Discussion

Continuous or cyclical water stress for 4 to 5 weeks

(from the initiation of reduced water application to







restoration of daily irrigation to container soil field

capacity) resulted in flower induction of 'Tahiti' lime

(Table III.1). Continuous and cyclical water stress

resulted in more total shoot production as well as a

significantly greater number of flowers than controls.

After these trees had completed flowering, the flowers and

fruitlets were removed. The trees were allowed to resume

vegetative growth for a period of 2 months under controlled

greenhouse growing conditions. The same trees were induced

to flower a second time by the above procedure. The

rationale for using the same trees in experiment 2 was that

flower reinduction in the same population of trees by the

same treatment should indicate that our treatments were

truly effective since heavy flowering in subsequent flushes

does not occur in greenhouse-grown citrus. The results of

the second experiment were similar to those of the first.

Continuous and cyclical water stress resulted in trees

producing more total shoots and flowers than controls, which

produced only random and insignificant numbers of flowering

shoots and flowers. There were more flowers/plant produced

in continuous than cyclical stress of experiment 1, but not

in experiment 2. No significant differences were found

between continuous or cyclical stress with regard to shoots

per plant, shoot type, or flowers per plant, but continuous

stress generally resulted in greater numbers of total shoots

per plant and flowers per plant. In fact, as long as the

severity of stress resulted in prolonged wilting or a leaf

xylem pressure potential of -3.5 MPa for 4 to 5 weeks, the












Table III.1 Effect of continuous or cyclical water stress on flower
induction in containerized 'Tahiti' lime trees.




Water Stress Shoots/ Shoot Type (%) Flowers/Plant
Treatment Plant Vegetative Mixed Generative



EXPERIMENT 12

Control 5.7Y+0.5 13.0 0 87.0 5.0+ 1.4
Continuous 41.7+8.0 23.3 17.4 59.3 145.7+48.5
Cyclical 29.5+4.4 44.1 16.1 39.8 44.2+14.3


EXPERIMENT 2
(Repeat)

Control 0.3+0.5 0 0 100.0 0.3+ 0.5
Continuous 37.7+5.0 11.3 53.6 35.1 75.7+18.1
Cyclical 29.0+5.0 19.0 48.3 32.7 78.5+51.9



ZExperiment #1 2/2/84 to 3/17/84.
Experiment #2 5/11/84 to 6/12/84.

These data represent 1 of 2 replicate experiments.
In this experiment, the same 4 tree replicates/treatment were used.

Values represent means + standard deviations.







flower inductive response was similar and significantly

different from controls. Therefore, the continuous stress

condition was used because it was easy and allowed for

uniformity of treatment.

The above experiment, which had been performed at two

different times of the year including that when 'Tahiti'

lime typically does not flower, indicated that flowering

could be induced in containerized lime trees by a period of

water stress lasting for a 4 to 5 week period. In order to

more clearly define the duration and severity of water

stress needed to induce flowering, leaf xylem pressure

potentials were measured at weekly intervals over a 4 to 5

week period in a population of trees that were stressed.

The level of moderate stress maintained above controls at

predawn and midday for each time interval measured (Table

III.2). We defined moderate levels of stress (-2.1 to -3.0

MPa) as those levels intermediately between control and

severe stress (-3.5 MPa) (95). Midday leaf xylem pressure

potentials were generally lower and measurements less

variable than those at predawn, except at week 5 where

predawn stress was as great as that at midday. Intertree

variability and daily climatic changes were presumably

responsible for the variability in the pressure potential

measurements. Control trees produced the least number of

shoots/plant and those shoots were vegetative. More

shoots/plant were produced as a result of water stress and

the numbers generally increased in trees exposed to greater

durations of water stress. Flowering was induced after 2















Table III.2 Effect of moderate water stress over time on leaf xylem
pressure potential and flower induction in 'Tahiti' lime.




Leaf Xylem
Duration of Pressure Potential Shoots/Plant
Water Stress Predawn Midday
(wks) (MPa)



Control -0.34+0.08Y -1.48+0.15 4.50x+1.9

2 -0.90+0.42 -2.25+0.08 6.25+2.2

3 -1.62+0.82 -2.21+0.25 8.00+2.6

4 -0.87+0.09 -2.89+0.23 9.75+3.0

5 -2.89+0.62 -2.83+0.19 9.75+1.5





z % flowering shoots = sum of mixed and generative shoot percentages.

Y Values represent the means of 10 leaf replicates/treatment + standard
deviations.

x Values represent the means of 5 tree replicates/treatment + standard
deviations.















Table III.2 -- Extended


(%)Z
Shoot Type (%) Flowers/Plant Flowering
Vegetative Mixed Generative Shoots




100.0 0 0 0 0

68.0 16.0 16.0 3.0+0.82 32.0

46.9 21.9 31.2 5.0+2.16 53.1

43.6 20.5 35.9 9.0+2.16 56.4

10.3 56.4 33.3 21.0+8.04 89.7







weeks of stress. The percent flowering shoots and number of

flowers per plant increased with time under stress. The

highest percentage of flowering shoots and flowers per plant

were found after 5 weeks of water stress. Apparently,

moderate levels of stress can induce flowering in a

relatively short period of time (2 weeks), but the inductive

response is much greater after an extended time period (5

weeks).

In a similar experiment, 'Tahiti' lime trees were

severely stressed as indicated by leaf xylem pressure

potentials ranging from -3.25 to 3.67 MPa (Table III.3).

Predawn and midday leaf xylem pressure potentials were

significantly different from controls at each measured time.

Predawn and midday stress measurements were significantly

different from one another at 2 weeks, but thereafter,

pressure potentials were not different from one another, and

a constant level of water stress prevailed in these trees

throughout the experiment. At these severe stress levels, a

less variable leaf xylem pressure potential was maintained

than those measured for moderate stress (Table III.2)

indicating that control of stress (water potential) was

obtained under severe water stress conditions. As in the

preceding experiment, control trees produced very few shoots

and those shoots produced were vegetative. On the other

hand, severe water stress when compared to moderate water

stress resulted in much greater numbers of shoots and

flowers per plant as well as increased percentages of

flowering shoots at all measured time intervals. Flowering















Table III.3 Effect of severe water stress over time on leaf xylem
pressure potential and flower induction in 'Tahiti' lime.




Leaf Xylem
Duration of Pressure Potential Shoots/Plant
Water Stress Predawn Midday
(wks) (MPa)



Control -0.24+0.05Y -1.38+0.29 3.25+2.0

2 -2.00+0.35 -3.25+0.07 70.50+18.0

3 -3.41+0.83 -3.67+0.24 45.00+12.0

4 -3.56+0.25 -3.66+0.21 49.00+25.0

5 -3.54+0.27 -3.58+0.23 49.80+10.2


x % flowering shoots = sum of mixed


and generative shoot percentages.


Y Values represent the means of 10 leaf replicates/treatment +
standard deviations.

Z Values represent the means of 5 tree replicates/treatment +
standard deviations.















Table 111.3 -- Extended


(%)z
Shoot Type (%) Flowers/Plants Flowering
Vegetative Mixed Generative Shoots


100.0 0 0 0 0

16.0 23.7 60.3 246.75+15.8 84.0

8.9 36.7 54.4 97.2+33.2 91.1

6.9 26.6 66.5 144.6+41.6 93.1

6.4 34.1 59.4 168.9+22.8 93.5







trends between moderate and severely water-stressed trees

were dissimilar over time, with severely water stressed

trees producing the same number of flowers and flowering

shoots at each measured time interval. Under moderately

water stressed conditions, water stress duration was a

factor regulating the flower inductive response. The

flowering response appears to be time dependent when

regulated by moderate levels of water stress, but at some

point floral induction is more immediately reached under

conditions of more severe water stress.

Low temperature stress 180C/100C (day/night) time

course experiments were conducted with containerized

'Tahiti' lime trees growing in the growth chamber as

previously reported (10, 59, 77). Leaf xylem pressure

potentials did not significantly differ from one another at

predawn except at the 4-week time interval (Table III.4).

At midday, however, control trees growing in the greenhouse

had significantly more negative pressure potentials at all

intervals measured. Control trees, as in the previous water

stress experiments, produced very few shoots per plant, and

those produced were vegetative (Tables III.3 and III.5).

Flowering was induced in trees after having been in the

growth chamber for as little as 2 weeks, but the response

was not as great as that of severely water stressed trees

within the same time period (Table III.5). The low tempera-

ture stress of the growth chamber resulted in a flower

inductive response like that of moderate water stress, but

apparently not through a common reduction in leaf xylem














Table III.4 Effect of low temperature over time on leaf xylem
pressure potential in 'Tahiti' lime.z



Duration of Continual Low Temperature Stress (wks)

0 2 CY 4 C 6 C 8 C


Leaf Xylem Pressure Potential (MPa)

Measured at predawn:

-0.360 -0.400 -0.166 -0.479 -0.373 -0.413 -0.426 -0.340 -0.326
+0.098x 0.105 0.041 0.055 0.101 0.069 0.043 0.092 0.064



Measured at midday:

-1.446 -0.613 -1.306 -0.623 -1.926 -0.633 -1.840 -0.500 -1.94
+0.203 0.109 0.118 0.072 0.086 0.052 0.149 0.066 0.08




Z Low temperature conditions utilized in these experiments were
12 hours each 180C day/100 night.

Y C = control treatments at each time measurement.

x Values represent means of 10 leaf replicates/treatment + standard
deviation.














Table III.5 Effect of low temperature over time on flower induction in
'Tahiti' lime.z



Duration of
Reduced Shootsy Flowers/Floweringx
Temperature /Plant Shoot Type (%) Plant Shoots
(wks) Vegetative Mixed Generative (%)


Control 5.20+0.4 100.00 0 0 0 0

2 6.20+3.8 54.84 12.90 32.26 5.60+4.03z 45.16

4 9.80+3.8 44.90 18.37 36.73 14.20+6.76 55.10

6 13.20+3.7 33.33 36.37 30.30 25.40+10.02 66.67

8 15.20+5.8 22.37 14.47 63.16 30.00+9.97 77.63




z Low temperature conditions utilized in these experiments were
12 hours each 180C day/100C night.

Y Values represent means from 5 tree replicates/treatment + standard
deviations.

x % flowering shoots = sum of mixed and generative shoot percentages.







pressure potential. A different signal, mediated through a

common mechanism, may be regulating floral induction.

Severe water stress rather than low temperature stress

consistently produced the greatest number of flowers and

flowering shoots. Floral induction best describes the

floral response observed after both water and low tempera-

ture stress. This belief is based upon results obtained

from other experiments which will be presented elsewhere (S.

M. Southwick and T. L. Davenport, submitted for publication

J. Amer. Soc. Hort. Sci.) indicating that lime trees forced

to produce shoots by branch pruning produced a greater

percentage of flowering shoots after imposing the above

stress treatments.

A final experiment was conducted to determine if water

stress would induce flowering on cuttings that had been

obtained from trees growing in the field. Cuttings (see

Materials and Methods) were separated into 2 populations:

those with leaves and those with leaves removed. From each

population, 1 set was placed in the greenhouse and allowed

to dehydrate by only occasionally irrigating (water stress),

and another set was put on a different bench under

intermittent mist (non-stressed) to alleviate dehydration

(40). During the period in the greenhouse prior to flower

production, all leaves, except for the most immature, wilted

and abscised. After 33 days had elapsed, those cuttings

that had been placed in the greenhouse dedicatedd), both

leafy and those with leaves removed, produced flowers (Table

III.6). The greatest number of shoots, flowers, and














Table III.6 Effect of leaves and misting on flowering of
immature 'Tahiti' lime cuttings.


Total
Treatment Shoots Shoot type (mean) Flowers Flowering
(no.) Vegetative Mixed Generative (no.) Shoots
(%)

Greenhouse
No Leaves 8.3z 5.6 0 2.7 3.0 32.0
+ 1.5 + 2.1 + 0.6 + 1.0

Leaves 16.7 5.0 2.7 9.0 22.7 70.0
+ 1.1 + 3.6 + 1.1 + 1.7 + 11.7



Mist Bed
No Leaves -

Leaves 14.7y 8.0 2.3 4.3 10.3 45.4
+ 6.5 + 4.0 + 1.5 + 2.1 + 1.5




z Each value represents the mean of 3 experiments where at least
10 replicates/treatment were used. + = standard deviation.

y Cuttings in the mist bed did not have shoots or flowers
after 33 days, therefore, those cuttings were removed from the
mist bed, placed on another bench in the greenhouse and data
recorded 5 weeks later.







flowering shoots were produced on cuttings that had

initially borne leaves. Nevertheless, those cuttings that

had their leaves manually removed produced both vegetative

and flowering shoots. Since immature, leafy cuttings were

selected for these experiments, it is improbable that a

previously stored floral message was present. Furthermore,

flowers produced on leafless cuttings suggest that it is not

essential for leaves to be present for floral induction and

that perception of flowering cues occurs within the shoot,

or as most likely in the bud itself. Although it has been

speculated that citrus roots may produce a substances)

which can be transported to shoots and exert control over

bud-break and flowering (38). These immature cuttings used

in the experiments reported here, never flower in the field

while attached to the tree until they go through a period of

maturation.

Cuttings that were placed in the mist bed did not

produce shoots or flowers. Therefore, after the same 33

day period, these cuttings were removed from the mist bed

and placed on an open bench in the greenhouse, and after

another 5 weeks those cuttings bearing leaves produced

vegetative and flowering shoots as a result of water stress

(Table III.6). Cuttings without leaves which had been

placed in the mist bed did not produce any new shoots and

eventually died.

In conclusion, 'Tahiti' lime trees preconditioned at

container soil field capacity for about 1 month can be

severely water stressed for a period of as little as 2 weeks





32

and consistent flower inductive responses obtained. The

floral response seems to be time dependent under conditions

of moderate water stress and low temperature. However,

floral induction from low temperature when compared to water

stress is not mediated through a common decrease in leaf

xylem pressure potential. Immature leafless cuttings can

produce flowering shoots under water-stress conditions,

indicating that leaves are not essential for flower

induction in 'Tahiti' lime.










CHAPTER IV
INVESTIGATIONS OF HORMONAL CONTROL OF CITRUS FLOWERING:
DEVELOPMENT OF A RADIOIMMUNOASSAY FOR THE MEASUREMENT
OF ABSCISIC ACID IN LEAVES AND BUDS OF 'TAHITI' LIME


Introduction

Abscisic acid (ABA) is a plant growth regulator which

is typically associated with general stress phenomena in

plants (65, 68, 100). The levels of ABA and ABA-conjugates

(9, 53, 81) increase in plant tissues during temperature

(100) and especially water stress (9, 34, 68, 100, 119, 120,

122) and a physiological role for ABA has been suggested

most often for stomatal control (65, 68, 100). It is

thought that free ABA produced in guard cells is responsible

for changing K+ levels, which changes cellular water

potential and alters cell turgor (65, 100, 112). The

regulation of stomatal physiology by ABA is thought to be

one way in which plants regulate water loss in order to

adapt to various environmental stresses (65, 68, 100).

Abscisic acid and ABA-conjugates have been found to

increase during citrus fruit maturation and have been

related to exocarp senescence and the transition of

chloroplasts to chromoplasts in this tissue (10, 31, 34).

The accumulation of ABA or possibly the 2, trans-isomer of

ABA in fruit has been purported to influence bud dormancy

and contribute to biennial-bearing in 'Wilking' mandarin

(30, 47). The specific identities of ABA-conjugates,

however, have not been described.







A speculative role for ABA flowering of selected short

day plants has been suggested (121). Even though an ABA

application alone cannot induce flowering under absolutely

non-inductive conditions, short day plants such as Pharbitis

(39), Lemna (41), and Chenopodium (54, 91) can be promoted

to flower by ABA application if test plants had been

slightly induced. Application of ABA (100 ng/ml) to the

culture medium stimulated flower formation in vegetative

stem explants of Torenia which otherwise had little flower

forming capacity (96). There are no reports, however, of any

link between ABA levels and flowering in citrus or in any

tree species for that matter.

Until recently, it has been difficult to measure ABA in

small amounts of plant tissue and correlate a role for ABA

with the flowering process of citrus. Advances in immuno-

logical methods (101, 103, 104, 107, 108, 109, 110, 112,

114) have minimized the need for plant tissue extract

purification and allow for analysis of many samples using a

minimal amount of tissue. In these studies reported here we

have developed a radioimmunoassay for measurement of total

ABA as described previously (60, 86, 103, 104). 'Tahiti'

lime trees were grown under the floral inductive conditions

of severe water and low temperature stress (Chapter III).

Since low temperature and water stress lead to the

accumulation of ABA, we wondered whether the accumulation

could be linked to control of flowering in lime. The

changes in ABA levels in leaf and bud tissues were measured







by radioimmunoassay and compared with floral inductive

treatment.


Materials and Methods

Plant Material and Extraction Procedure. In these

experiments, 1- to 2-year-old container grown 'Tahiti' lime

(Citrus latifolia tan.) trees were grown in the greenhouse

or growth chamber as described in Chapter III. Five tree

replicates/treatment were severely water stressed (at least

-3.5 MPa leaf xylem pressure potential, Chapter III) and

leaf and bud tissue were collected from trees after 2, 3,

and 4 weeks of continual water stress. Water stress was

alleviated by rewatering the container soil to field

capacity and leaf and bud tissue were collected 1 and 10

days later. In addition, 5 tree replicates/treatment were

placed in the growth chamber under low temperature

conditions as described previously (Chapter III). Leaf and

bud tissue were collected after 2, 4, and 6 weeks of growth

in the chamber. Immediately after each water or low

temperature stress interval had elapsed, leaves or buds from

each tree replicate were removed and washed with distilled

water, towel dried, pooled together, frozen in liquid N2 and

ground in a mortar with a pestle into a fine powder. The

tissue was stored in the freezer at -180C for less than 90

days until use. The extraction procedure of Weiler (103)

was followed with slight modification. Fifty to 100 mg of

frozen powdered tissue was extracted with 90% methanol

(MeOH) containing 10 mg/l 2, 6-di-t-butyl-4-methyl phenol

(BHT) (69) (200 ul/mg frozen tissue) for 2 days in the dark







at 40C. The tissue was shaken at regular intervals daily

and the supernatant collected after centrifugation at 7,000

rpm for 10 minutes. Aliquots of the extract were diluted

1:10 or 1:25 with phosphate buffered saline (PBS, 0.01 M Na

phosphate, 0.15 M NaC1, pH 7.4) prior to radioimmunoassay.

Preparation of (+) ABA-BSA Conjugates. The ABA-BSA

conjugate was prepared by the method of Weiler (104). How-

ever, instead of human serum albumin, bovine serum albumin

was substituted. The molar coupling ratio was ca. 7 mol ABA

bound per mol BSA as determined by isotopic recovery.

Preparation of B-D-glucopyranosyl abscisate. Abscisic

acid (+ABA ) (Fluka;264.32 mg, 1.0 mmoles) was gradually

added and stirred into 2.0 ml of dioxane in a 25 ml round

bottom flask at room temperature. After the ABA had

completely dissolved, 1.25 ml of absolute ethanol (EtOH) and

3.25 ml of distilled water were gradually added which

resulted in a clear yellowish solution. Cesium bicarbonate

(194.0 mg, 1.0 mmoles) was added and the pH of the reaction

mixture was lowered to 7.0 by adding several milligram more

ABA, after which it was gently stirred for 30 minutes. The

volume was reduced by one-half via rotary evaporation at

400C. Absolute EtOH (5 ml) and then benzene (5 ml) were

added, and the mixture was reduced to dryness. The

resulting yellow crystals were redissolved in EtOH (10 ml)

and then benzene (10 ml) was added, and the solution was

evaporated to one-third. Benzene (15 ml) was added and the

solution was reduced to one-third as before. Benzene was

added twice more to the capacity of the 25 ml round bottom







flask and evaporated as described above and the final

solution was reduced to dryness. The yellow-white ABA Cs

salt crystals were dried in vacuo over P205 for 3 hours.

One millilitre dry dimethylformamide (DMF) was added to

the ABA-Cs salt at room temperature while gently stirring.

Acetobromo-a-D-glucose tetraacetate (Sigma; 415.32 mg, 1.01

mmoles) in 0.75 ml dry DMF was added dropwise to the

solution. Another 0.25 ml DMF was used to rinse the

acetobromo-a-D-glucose tetraacetate from its weighing

container and also added to the solution. The resulting

solution was stirred overnight at room temperature. The

mixture was centrifuged at 7,000 rpm for 15 minutes and the

supernatant loaded onto a silica gel G-60 (EM Scientific

230-400 mesh) 1.2 x 100 cm column equilibrated in chloro-

form:methanol (96:4, v/v). The ABA-GE tAc was eluted with

300 ml of the above solvent mixture into 4 ml fractions

number 17-30. Those fractions which showed one spot, Rf

0.824 on TLC (Whatman reverse-phase MKFG silica gel 1 x 3

cm, 200 u thickness) plates developed in chloroform:methanol

(96:4, v/v and visualized by UV) were pooled. Other

fractions containing ABA-GE tAc and minor contaminants were

rechromatographed on a new silica gel G-60 column as

described previously after solvent evaporation, and those

fractions showing a single spot on TLC were collected again.

All fractions containing purified ABA-GE tAc were pooled,

and the solvent was evaporated. The white ABA-GE tAc

crystals were dried in vacuo over P205 and stored at -180C

(73-75mp; 247 mg).







The preparation of ABA-GE from ABA-GE tAc was adapted

from the procedure of Lehmann et al. (57) and was followed

with slight modification in purification. Briefly, 3 g of

dehusked, ripe, mature sunflower Helianthus annuus seeds

were ground in an ice cold mortar and pestle with 25 ml, 100

mM Na-phosphate buffer (pH 7.0) and sand until homogeneity.

The homogenate was centrifuged for 20 minutes at 20,000 rpm

which resulted in the formation of 3 layers. The middle

layer (17.0 ml) was collected. Eleven millitre of Na-

phosphate buffer was slowly added to a solution of ABA-GE

tAc (60 mg) in 3.5 ml EtOH, and gently stirred at room

temperature. The 17.0 ml of crude sunflower seed enzyme

preparation was added and allowed to react for 24 hours.

The reaction was stopped with 20 ml of EtOH and the mixture

was centrifuged for 20 minutes at 20,000 rpm. The super-

natant was collected and reduced in volume to 1 to 2 ml via

rotary evaporation, and the concentrated solution was

filtered through glass wool and loaded onto a silica gel G-

60 column (1.2 x 100 cm) equilibrated in chloroform:me-

thanol:water (75:22:3, v/v). The putative ABA-GE was eluted

in 4 ml (fractions 52-65) in the above solvent system. The

putative ABA-GE was identified by TLC on the aforementioned

Whatman reverse phase silica gel (Rf 0.45; ABA-GE tAc, Rf

0.95) with the above solvent system. The fractions

containing ABA-GE were pooled, and the solvent was removed

by evaporation. The resultant ABA-GE was dissolved in EtOH

(5 ml) and then benzene (10 ml) and these solvents were

reduced via rotary evaporation. Approximately 10 ml benzene







was added and reduced 2 more times. Anhydrous methanol was

added dropwise to the dry ABA-GE, and the solution was

filtered through glass wool to remove any traces of silica

gel. Methanol was removed by evaporation under a stream of

dry N2 gas. The residue dried in vacuo over P205 (24.81mg;

55.5%). The identities of ABA-GE tAc and ABA-GE were

verified by analysis of micro NMR spectra (Bruker AM-200

spectrometer). The spectra were obtained at 200 mHz.

Immunization Procedure. An antigen emulsion was

prepared by dissolving 1.5 mg lyophilized conjugate in 1.5

ml PBS (0.01 M Na phosphate, 0.15 M NaC1, pH 7.4) and

emulsified with 2.0 ml complete Freund's Adjuvant (Difco).

Randomly bred New Zealand White rabbits (12 to 16 weeks old)

were immunized with a total of 1.0 ml of freshly prepared

antigen emulsion by making 2, 0.25 ml intramuscular and 2,

0.25 ml intradermal neck injections. The first collection

of antisera began 10 days later. Boosters were administered

biweekly and the preparation (45) and titre (114) of anti-

serum were tested 1 week after each booster until suitable

for radioimmunoassay use. Antisera were stored frozen at

-180C.

Radioimmunoassay Procedure. The following reactants in

order of addition were pipetted into a test-tube in an

icebath: 200 ul of buffer (0.01 M Nal phosphate, 0.15 M

NaCd, pH 7.4), 100 ul of standard or appropriately diluted

plant extract, 100 ul of tracer (DL-cis, trans-G-3H abscisic

acid, 39 Ci/mmol, Amersham; 2000 cpm, 0.0513 pmol), 100 ul

of ABA-directed antiserum diluted 1:625 (final assay







dilution). The reactants were swirled by vortexing and

incubated in the icebath for 1 hour. Next, 0.5 ml of ice

cold saturated (NH4)2SO4 was added to each tube to separate

free from antibody bound ABA and vortexed. This solution

was incubated in the icebath for 30 minutes and then tubes

were centrifuged for 30 minutes at 7,000 rpm. From each

tube a 0.5 ml aliquot was drawn and pipetted into a

scintillation vial and 10 ml of scintillation cocktail added

and counted to 1% significance. Blanks for non-specific

binding were essentially prepared by procedures described

previously (103, 114) where water was substituted for

antiserum.

Separation of Free ABA From ABA-conjugates. Thin layer

chromatography plates (Slica gel G, Fisher Scientific) were

prewashed 2 times in MeOH prior to use. Leaf or bud

extracts (50 to 100 ul) were streaked and developed in

toluene:ethyl acetate:acetic acid (50:30:4, v/v/v; Rf 0.5

free ABA; 0.0 ABA-GE; 0.33 phaseic acid) (9). The silica

gel Rf corresponding to free ABA was collected and extracted

in a test-tube with 0.5 ml 95% EtOH. The tubes were centri-

fuged 30 minutes later at 7,000 rpm for 10 minutes. The

supernatant was collected and 100 ul aliquots were added to

each replicate radioimmunoassay tube. The EtOH was

evaporated under a stream of dry N2 gas and the radio-

immunoassay procedure was followed as outlined above, except

that, 300 ul rather than 200 ul of buffer was used. The

recovery of free ABA from 'Tahiti' lime leaf and bud







extracts after thin layer chromatography and radioimmuno-

assay was ca. 60%.


Results

Reaction Parameters of the ABA Radioimmunoassay. All

immunized rabbits responded to the treatment, but. in the

present study only antisera from 1 rabbit was selected.

When incubated under standard assay conditions, the antisera

bound 0.40 pmol of 3H (+) ABA at a final dilution of 1:625

and had an affinity binding constant of Ka=8.5 x 10-12M

(88). The binding of ABA to the antiserum was not pH

dependent over the range of 4.0 to 10.0 and the reaction

between antibody and ABA had come to completion after a

period of 1 hour in the icebath. Under the standard assay

conditions, 100 ul of 1:625 diluted serum (final assay

dilution) bound approximately 50% of tracer. Unspecific

binding was less than 3%.

Assay Sensitivity. A typical standard curve in the

non-linear plot which was prepared for each day's assay is

shown in Figure IV.1 and in the linear, logit/log, trans-

formation in Figure IV.2. The linearity of this plot

indicated that reaction equilibrium was attained under the

conditions employed. The measuring range of this assay

corresponding to the linear range of the logit/log plot was

from 25 to 1000 pg/assay and the detection limit at the 95%

confidence limit was 25 pg.

Assay Specificity and Accuracy. The specificity of the

ABA-directed antibody was tested in several ways. The

number of compounds that were either structurally or



























Figure IV.1 Standard curve for the (+) ABA
radioimmunoassay constructed from n=20 consecutive
assays to show day-to-day reproducibility. The bars
indicate standard deviations of triplicate samples.
Bo=amount of tracer bound in the absence of (+) ABA
standard; B=amount of tracer bound in the absence of
standard.









/00

so -
1' 2\

\
60 \

\4
o C \
\

\B

/ 0 /00 /000

ABA (pg /ossoy )





























Figure IV.2 Linearized logit-log plot of the standard
curve for the (+) ABA radioimmunoassay.
logit [(B/Bo)/(100-B/Bo) .


















S20




o





be
. -I

o

-3



-j


Sto0 o00 /000
SABA (pg/ ossay )







physiologically related to ABA were assayed for cross-

reactivity with antisera. Cross reactivities were

determined as previously described (101, 103, 104, 114).

The only compounds which cross-reacted with the ABA-directed

antisera were (+) ABA (50.8%) and methyl-(+)-abscisate

(75.2%) (Table IV.1). The other compounds tested did not

cross-react with the ABA-directed antiserum to any signifi-

cant extent.

An indication of the specificity of this antisera for

ABA is shown in Figure IV.3. Potentially interfering

compounds in lime extracts were detected by using extract

dilution curves. These curve were found to parallel the

standard curve. This parallelism indicates the absence of

interfering compounds cross-reacting with the antisera (104)

and having different affinity constants as compared to ABA

(103, 104). Abscisic acid (50 pg) added to leaf extracts

showed a 91% recovery over the dilution range tested and

paralleled the standard curve. The assay is precise and

reproducible as is evident from Figures IV.1 and IV.2. The

standard curve was constructed from n=20 consecutive assays

to show day-to-day reproducibility expressed as percent

coefficient of variation of 6.1%. Triplicate determinations

of unknown samples with readings throughout the measuring

range gave coefficients of variation of 5.6% and the

complete procedure (including extract processing and

immunoassay) is 7.5% for an average sample.

Measurement of Total and Free ABA in 'Tahiti' Lime

Leaves. The level of total ABA increased in leaves with














Table IV.1 The specificity of antiserum to abscisic acid.






Compound pmol Required to Cross Reactivity
Yield B/Bo=50%z (%)



(+)-abscisic acid 1.55 100.00

(+)-abscisic acid 3.05 50.80

methyl-(+)-abscisate 2.06 75.20

(+)-abscisyl-(2,3,4,6- 15.60 9.90
0-tetraacetyl)-B-D-glucose
ester

(+)-B-D-glucopyranosyl 19.40 8.00
abscisate

phaseic acid >178.60* 0.86

dihydrophaseic acid >355.80* 0.43

xanthoxin 2609.10 0.06


* Highest concentration assayed.

z B=% 3H (+) ABA binding in presence of compound; Bo=3H
(+) ABA binding in absence of compound.






























Figure IV.3 Extract dilution analysis. Diluted leaf
extracts of Citrus latifolia Tan. contained no (+) ABA
or 50 pg/assay (+) ABA.











A ABA saondord
Leaf extract of Cifrus Lo li Torn. ---o
Leaf exfroct withff( ABA added ---


/0 100 o100
(J) ABA (pg/ossoy)
ao' ol to
001 0. / o.0
Plant material assayed (mg)







time elapsed after water stress to a maximum of approxi-

mately 2.5 ng/mg frozen tissue (Figure IV.4). Immediately

(1 day) after water stress was alleviated by rewatering the

soil to field capacity, the level of total ABA fell and

continued to fall to a minimum measured at 10 days after

alleviating water stress. On the other hand, the level of

total ABA was lower in leaves collected from trees growing

under low temperature conditions than those leaves grown

under water stress conditions. The increasing trend of ABA

levels in leaves grown under low temperature conditions was

similar to that trend noted for leaves growing under water

stress conditions. The amount of ABA measured in leaves

growing under low temperature conditions increased as time

grown under low temperature increased. However, the

increases were not significantly different at each measured

time interval.

The levels of free ABA increased in the early part of

water stress treatment and reached a maximum (0.24 ng/mg

frozen tissue) after 2 weeks (Figure IV.5). After 3 and 4

weeks of stress, however, the level of free ABA decreased.

After water stress had been alleviated by rewatering the

soil to field capacity, the level of free ABA dropped to

non-detectable levels. The levels of free ABA in leaves

collected from trees grown under low temperature conditions

for 2, 4, and 6 weeks were less than that level measured

prior to the beginning of low temperature stress. The free

ABA level in leaves of low temperature grown lime trees did

not change greatly as a result of time elapsed in the growth



























Figure IV.4 Effect of duration of water and low
temperature stress on total ABA levels in 'Tahiti' lime
leaves as measured by the (+) ABA radioimmunoassay.
The bars indicate standard deviations of triplicate
samples. The control value at 0 time represents the
control level of total ABA throughout the measurement
time.









End water stress----
2.5

/ LEAVES

Severe water stress
(/ Low temperature stress ----



S2. 0









1 /.5








1.0 I I III II
0 2 3 4 5


TIME (wks )



























Figure IV.5 Effect of duration of water and low
temperature stress on free ABA levels in 'Tahiti' lime
leaves as measured by the (+) ABA radioimmunoassay.
The bars indicate standard deviations of triplicate
samples. The control value at 0 time represents the
control levels of free ABA throughout the measurement
time. nd = non-detectable levels of free ABA











LEAVES

Severe wafer stress
Low temperature stress o---











S-End water stress


0 1 2 3 4


TIME ( wks )





55

chamber. The free ABA levels in leaves growing under low

temperature conditions were less than those levels measured

in water stressed leaves at 2, 3, and 4 weeks.

Measurement of Total and Free ABA in 'Tahiti' Lime

Buds. The level of total ABA in lime buds increased as time

in which trees were grown under water stress progressed

(Figure IV.6). One day after water stress had been

alleviated, the total amount of ABA measured in buds

decreased to the same level found after 2 weeks of water

stress. The levels of total ABA continued to decrease with

time elapsed after rewatering and 10 days later the total

ABA level was equivalent to that measured in buds prior to

the beginning of water stress. The total ABA level measured

in buds during the low temperature stress conditions of the

growth chamber were less at 2, 4, and 6 weeks than that

level measured prior to the beginning of stress. The levels

of total ABA measured in buds of trees grown under low

temperature stress showed an increasing trend of ABA level

against time spent growing in low temperature conditions.

The levels of free ABA in buds decreased as the time

trees were maintained under water stress conditions

increased (Figure IV.7). The levels of free ABA dropped to

non-detectable levels after water stress had been alleviated

by rewatering. The free ABA levels measured in buds

collected from trees after 2 weeks of low temperature

conditions were less than those levels measured prior to the

beginning of stress. The levels of free ABA measured in

buds at 2 and 4 weeks of low temperature stress were not



























Figure IV.6 Effect of duration of water and low
temperature stress on total ABA levels in 'Tahiti' lime
buds as measured by the (+) ABA radioimmunoassay. The
bars indicate standard deviations of triplicate
samples. The control value at 0 time represents the
control level of free ABA throughout the measurement
time.










SBUDS
Q- 5-

V) Severe water stress -
ELnd woter stress- Low temperature stress --


i --





\\
41









l.


I I I I I I I
0 I 2 3 4 5 6


TIME ( wks )



























Figure IV.7 Effect of duration of water and low
temperature stress on free ABA levels in 'Tahiti' lime
buds as measured by the (+) ABA radioimmunoassay. The
bars indicate standard deviations of triplicate
samples. The control value at 0 time represents the
control level of the ABA throughout the measurement
time. nd = non-detectable levels of free ABA









0./4 BUDS

0. 13- Severe water stress ---
S 0.12-
/2) Low temperature stress 0-- 0


S0.9-


0.08-



006
SEnd water stress
0.05

Q 0.04-

003-

0.02-

'-/-.
nd and

0 2 3 4 5 6


TIME ( w ks )







different from one another. ABA levels measured at 6 weeks

were slightly less than those measured at 2 and 4 weeks of

low temperature stress.


Discussion

Antibodies were obtained that were specific for (+)

ABA. These antibodies can be used in a radioimmunoassay

which permits the direct quantification of total ABA between

25 and 1000 pg/assay in unpurified extracts from 'Tahiti'

lime. The antisera showed cross-reactivity with methyl-(+)-

abscisate (75.2%) and (+)-abscisic acid (50.8%) and these

results are similar to those reported previously (101, 103,

104). In those previous reports where radioimmunoassays had

been developed for (+) ABA coupled through the C 1, COOH

(101, 103, 104) functional group, the methyl ester of ABA

significantly cross-reacted with the antisera. The (+)

enantiomer of abscisic acid was not preferentially bound in

those previously reported assays. Rather, the (-)

enantiomer of ABA bound more preferentially than the (+)

enantiomer. The reason for this is not clear (104). In

the case of the assay reported here, no preferential binding

of the (-) enantiomer of ABA was apparent as evidenced by

50. 8% cross-reactivity of (+) ABA. Nevertheless, the fact

that (+)-abscisic acid is the naturally occurring form of

ABA found in plants (119, 120), and only cross-reacted 50%

with the antisera, may have decreased the sensitivity of

this assay when measuring ABA in plant extracts. In citrus,

however, it is not known whether the (+) enantiomer of ABA

is found exclusively. Xanthoxin and phaseic acid may occur







in amounts comparable to those of ABA in plants (103).

Dihydrophaseic acid may occur in hundredfold higher amounts

(100, 101, 103, 119). These compounds did not cross-react

with our antiserum to any significant extent and the assay

should not be affected at any concentration within the

physiological range.

Other potential cross-reactants against this ABA-

directed antisera, (+)-abscisyl-(2,3,4,6-0-tetraacetyl)-B-D-

glucose ester (ABA-GE-tAc) and (+)-B-D-glucopyranosyl

abscisate (ABA-GE) (53, 81), did not cross-react to the same

extent as that which had been reported previously for (+)

ABA-directed antisera (101, 103, 104). The reason for this

is not clear; however, it has been found that antibodies

developed by different persons through the same procedures

can have different biochemical characteristics (Dr. A.

Castro, March 17, 1985, personal communication).

These 2 ABA conjugates, ABA-GE tAc and ABA-GE, were

prepared by previously unpublished procedures. The prepara-

tion of esters of PGRs has been previously accomplished by

reaction of the free acids with 0-substituted halocarbo-

hydrates in the presence of triethylamine catalysts (58).

However, the use of amines in reaction with alkyl halides

can result in racemization of asymmetric carbons. Moreover,

the tertiary amine, triethylamine, reacting with an alkyl

halide can undergo alkylation to form an amine salt. This

reaction side product can reduce the formation of new

products and thus reaction yield, thereby requiring more

extensive purification. A need exists for a versatile and







easy procedure to prepare plant growth regulator esters

under mild conditions. Cesium carbonate or cesium bicar-

bonate has been used to form cesium salts of amino acid and

peptides (102) which can be reacted with alkyl halides to

form esters (44, 102).

The ABA-Cs salt is formed easily by adjusting the

reaction mixture to pH 7.0 as had been described for amino

acids and peptides (102). The purified ABA-Cs salt reacts

upon the addition of acetobromo-a-D-glucose in DMF. Thin

layer chromatography after 5, 15, and 30 minutes showed

progressively increasing amounts of ABA-GE tAc formed.

Yields of ABA-GE tAc (Figure IV.8) and ABA-GE (Figure IV.9)

reported were lower than those obtainable via workup of a

larger number of fractions because only those fractions

showing no evidence of cross contaminating compounds were

selected after column chromatography.

The free acid of ABA can be reacted with CsHCO3 to form

ABA-Cs salt, which can then be reacted with halocarbo-

hydrates to form esters of ABA. The preparation of ABA-GE

by this procedure and purification by silica gel column

chromatography are simple and easily scaled up.

The levels of total ABA measured in water and low

temperature stressed leaves were different from one another

(Figure IV.4). This observation is understandable in that 2

different stress conditions have been imposed on tree leaves

and the ABA levels may change as a result of leaves adapting

to different stress conditions. A dramatic change in total

ABA levels occurred when water stress was alleviated due to
































Figure IV.8 NMR spectra of B-D-glucopyranosyl
abscisate tetraacetate in deuterated DMSO.

















OAOAc




S-0 -glucopyronosyl obscisole letraoce/ote


I I I !
9.0 8.0 7.0 6.0 5.0
PPM


40 .O .0 0.0
































Figure IV.9 NMR spectra of B-D-glucopyranosyl
abscisate in deuterated DMSO.

















OH


0 0 OH

S- 0 glucopyrnosyl obscisale
-0-' g/ucopyronosy/ obscisole


I I I I I | I
9.0 8.0 7.0 6.0 5.0 4.0 J.0 20 .0 00 0 -2 0
PPM







rewatering the soil to field capacity. Even as trees

continued to recover from stress, the levels of total ABA

fell to a minimum of 10 days after water stress had been

alleviated. The levels of free ABA paralleled those of

total ABA, except those measurements made at the 2 week

interval. The level of free ABA was approximately 15 to 35

times less than that of total ABA, which indicated that ABA

conjugates, or other'cross-reacting compounds previously not

described, were responsible for the making up the total ABA

measurement. It was unlikely that non-described compounds

were cross-reacting with our antisera because extract dilu-

tion curves from leaves of 'Tahiti' lime were close to

parallel (Figure IV.3) indicating the lack of interference

of cross-reacting compounds in the extract with our antibody

(104). It is more likely that ABA-conjugates are very high

in lime leaves. This was also evident from visualizing TLC

plates (silica gel GF 254, Fisher Scientific) under UV after

development in toluene:ethyl acetate:acetic acid (50:30:4,

v/v/v). Only at the origin could ABA be visualized and that

compound visualized corresponded to ABA-GE (9). The

presence of bound ABA, which was presumed to be ABA-GE (34),

was found to attain 4- to 10-fold higher levels than free

ABA in senescent citrus fruit peel. Moreover, the levels of

ABA previously reported in citrus flavedo (2.5 ng/mg fresh

wt, [34]), and young fruitlets (0.1 ng/mg fresh wt., [29])

as well as previous measurements of total ABA by Walton

(101) are quantitively similar to those measurements made in

the radioimmunoassay for ABA reported here. Apparently, ABA







levels differ in leaves as a result of the type of stress

imposed. The question of whether these changing levels of

ABA have any role in citrus flowering cannot be answered by

these experiments reported here. However, it was shown in

Chapter III that leaves are not required in all cases for

the production of flowers in 'Tahiti' lime cuttings.

The levels of total ABA measured in buds during and

after water stress of 'Tahiti' lime were greater than those

measured at any time during low temperature stress

conditions of the growth chamber (Figure IV.6). The levels

of free ABA in buds were 10 to 40 times less than total ABA

levels, indicating that the levels of ABA-conjugates were

changing more so than free ABA. The level of free ABA

dropped sharply when measured after 4 weeks of water stress

and when measured at 1 and 10 days after alleviating water

stress by rewatering (Figure IV.7). This fact would

circumstantially suggest that the levels of ABA-conjugates

are accumulating during water stress to a more significant

extent than under the conditions of low temperature stress.

The levels of free ABA decreased after water stress was

alleviated suggesting that the decrease in free ABA may be

linked to the removal of buds from water stress imposed

dormancy. In reports from previous research, there have

been mixed results with regard to a generalized role for ABA

in bud dormancy and budbreak (87, 104). In species such as

Ribes nigrum L. and Fagus sylvatica L. there is a good

correlation between the ratio of free and conjugated ABA

with the physiological or morphological status of buds







(116), but in species such as Betula pubescens Ehrh., there

is not a good correlation. Significant research emphasis

has been expended with regard to ABA involvement in bud

dormancy and the release of buds from winter dormancy.

Seeley and Powell (90) made monthly measurements for 1 full

year of free ABA and hydrolyzable ABA in 'Golden Delicious'

apple buds. They found that free ABA increased during mid-

summer before and after entry into early dormancy and

increased to a maximum prior to leaf fall. Thereafter, the

free ABA level decreased to a minimum prior to bloom.

Hydrolyzable-ABA increased gradually during fall and winter,

reached a maximum during the early stages of bud develop-

ment, and decreased rapidly just prior to full bloom.

These researchers concluded from their work and after

summarizing the research of others, that ABA has a casual

role in the inception of winter dormancy in apple. Abscisic

acid was shown to diffuse from apple bud scales (94).

Moreover, extracts from bud scales, as well as ABA applica-

tion to buds which had scales removed, mimicked the inhibi-

tory effect of bud scales on bud growth inhibition during an

in vitro experiment with dormant apple buds in branch

cuttings (94). In the case of 'Tahiti' lime, there is a

good apparent correlation between the levels of free and

conjugated ABA and budbreak under conditions of water

stress, but not under the conditions of low temperature

stress. Apparently, the 2 methods used to induce flowering

in lime regulate flowering, but not by altering the levels

of ABA in similar fashion.







Total and free ABA in 'Tahiti' lime leaves and buds

were measured by radioimmunoassay in these experiments. The

levels of ABA-conjugates were present in much higher levels

than those of free ABA. In fact, the levels of ABA-

conjugates may have been even higher than indicated by RIA

because of the lower level of cross reactivity of ABA-GE in

our assay than those reported previously (101, 103, 104).

The levels of ABA changed in dissimilar fashion when

measured in leaves and buds of lime trees grown under either

the floral inductive conditions of water or low temperature

stress. Therefore, it appears that ABA does not show a

consistent pattern of change as a result of floral inductive

treatment. Abscisic acid seems to be actively synthesized

and metabolized under water stress conditions and the levels

of total and free ABA decrease once water stress is

alleviated. Abscisic acid may be involved in dormancy and

budbreak. The role of ABA in physiological aspects of

flowering, budbreak and dormancy of lime is still unclear.

Exogenous sprays of ABA as well as more detailed analysis of

ABA levels and ABA receptor studies may help elucidate the

physiological role for ABA in these processes.










CHAPTER V
DEVELOPMENT OF A GIBBERELLIN RADIOIMMUNOASSAY
FOR MEASUREMENT OF GIBBERELLIN LEVELS IN LEAVES
AND BUDS OF FLORAL INDUCED 'TAHITI' LIME


Introduction

Evidence from several lines of research suggest that

gibberellins (GAs) are the group of plant growth hormones

most likely to control flowering in citrus. One line of

research showed that GA3 inhibited flowering in citrus after

being applied to citrus as whole tree sprays (5, 17, 37, 74)

or as an application to buds (35). These exogenous

applications of GA to citrus have shown that GA inhibits

flower formation. However, GA applied after a floral

inductive water stress period can inhibit flowering (81) and

therefore, as a result of GA application, flower buds may

never form or revert to vegetative apicies (35, 82).

Another line of evidence showed that the use of GA synthesis

inhibitors like (2-chloroethyl)trimethyl ammoniumchloride

(chlormequat) or succinic acid-2,2-dimethylhydrazide

daminozidee) induced large numbers of flowers under certain

conditions (73, 74), thereby supporting the likelihood for

GA control of flowering.

Even though GA has been found to inhibit flowering when

applied to citrus trees, the endogenous GA levels and how

they might change as a result of floral inductive treatment

are poorly understood in citrus. Gibberellin A1 has been

found in water sprouts of citrus (51). Gibberellin A1 and

GA9 as well as an unknown gibberellin have been identified






in young 'Washington' navel oranges (52). For the most

part, however, only gibberellin-like substances have been

isolated from citrus tissues (29, 32, 35, 99, 115). These

substances are termed "GA-like" because they exhibit the

same activity as GA3 in GA-sensitive bioassays. Endogenous

GA-like compounds were found to be highest in vegetative

shoots of citrus (shoots bearing only leaves), next highest

in mixed shoots (shoots bearing leaves and flowers) and

least in generative shoots (shoots bearing only flowers)

(29, 35). These different GA levels correlated with the

amount of flowering in each of these 3 shoot types and

suggested that the endogenous levels of GA controlled the

flowering behavior of citrus.

Gram or killigram quantities of plant tissue are

usually required to make plant hormone measurements because

the amounts found in tissues are low and losses are great

during extraction and purification. Therefore, to

investigate GA changes in small quantities of citrus leaf

and especially bud tissue we felt it necessary to develop a

GA-sensitive immunoassay. The measurement of GA's by

immunoassay has been reported to be sensitive, rapid,

reproducible, inexpensive, and require only small quantities

of tissue for measurements (3, 4, 113).

In an effort to investigate the role of endogenous

gibberellins in citrus flowering, we have induced flowering

by low temperature and water stress as previously described

in Chapter III. Leaf and bud tissues were collected

throughout the stress periods and analyzed for GA in order







to correlate changes in GA levels with the floral inductive

treatments.


Materials and Methods

Plant Material and Extraction Procedure. In these

experiments, 1- to 2-year-old container grown 'Tahiti' lime

(Citrus latifolia Tan.) trees were grown in the greenhouse

or growth chamber as described in Chapters III and IV. Five

tree replicates/treatment were severely water stressed (at

least -3.5 MPa leaf xylem pressure potential, Chapter III)

and leaf and bud tissue were collected from trees after 2,

3, and 4 weeks. Water stress was alleviated by watering the

container soil to field capacity and leaf and bud tissue

were collected 1 and 10 days later. In addition, 5 tree

replicates/treatment were placed in the growth chamber under

low temperature conditions as described previously (Chapter

III). Leaf and bud tissue were collected after 2, 4, and 6

weeks of growth in the chamber. Immediately after each

water or low temperature stress interval, leaves or buds

from each tree replicate were removed and washed with

distilled water, towel dried, pooled together, frozen in

liquid N2, and ground in a mortar with a pestle into a fine

powder. The tissue was stored in the freezer at -180C for

less than 90 days until use. The extraction procedure of

Weiler (9103) was followed with slight modification. Fifty

to 100 mg of frozen powdered tissue was extracted with 90%

MeOH, containing 10 mg/l 2,6-di-t-butyl-4-methyl phenol

(BHT) (200 ul/mg frozen tissue) for 2 days in the dark at

40C. The tissue was shaken at regular intervals daily and







the supernatant collected after centrifugation at 7,000 rpm

for 10 minutes. Aliquots of the extract were diluted 1:5 or

1:10 with phosphate buffered saline (PBS, 0.01 M Na

phosphate, 0.15 M NaC1, pH 7.4) prior to radioimmunoassay.

Preparation of GA3-BSA Conjugates by Mixed Anhydride

Reaction. The GA conjugate was prepared by the method of

Weiler and Wieczorek (113). Thin layer chromatographic

analysis (12, 64, 113) revealed the absence of any free

uncoupled GA3 in the diasylate. The coupling of GA3 to BSA

was verified by spectroscopic analysis (21, 22) and the

coupling ration was ca. 1.2 to 1.7 mol GA3 coupled to 1 mol

protein. The molar coupling ratio was calculated by

developing a standard curve of GA3 concentration (Figure

V.1; GA3 dissolved in concentrated H2S04) versus the optical

density at the peak of the absorption spectrum 415 nm, and

then calculating the mol GA3 coupled to 1 mol protein (21).

The GA-BSA conjugate was assayed for the presence of bound

GA and no change in the absorption peak maximum was noted;

therefore, we concluded that GA was bound to the protein.

Preparation of GA3-BSA Conjugates by Symmetrical

Anhydride Reaction. Gibberellin A3 was coupled to BSA by

the method of Atzorn and Weiler (3, 4). The conjugate was

analyzed as above for the presence of free unreacted GA3 and

none was found. The coupling ratio was ca. 5.4 to 7.2 mol

GA3 coupled to 1 mol protein.

Effect of pH on Coupling of GA3 to BSA via the Mixed

Anhydride Reaction. A 0C solution of BSA (840 mg) in 44 ml

DMF/water (1:1, v/v) was prepared. This solution was split




























Figure V.1 Standard curve of the concentration of GA3
(M, GA3 dissolved in concentrated H2SO4) versus
absorbance (nm) used to calculate molar coupling ratios
of GA3-BSA conjugates.































r = O. 993
y =0.0001338(x) 0.0000


A BSORBANCE







into 6 equal portions and the pH was adjusted to 6.0, 7.0,

8.0, 9.0, and 10.0. The pH of the final portion was

unadjusted. The GA3 anhydride was prepared on the same

molar basis/portion as by the method of Weiler and Wieczorek

for preparation of GA3-BSA conjugates (113). Upon addition

of GA3 anhydride to the BSA mixture, the desired pH was

maintained by addition of 1 N NaOH. The BSA solution with

an unadjusted pH (control) was reacted by the procedure of

Weiler and Wieczorek (113).

Synthesis of GA3-BSA via Hydroxysuccinimide.

Gibberellin A3 (311.76 mg, 0.9 mmoles) and hydroxysuccinimde

(126.5 mg, 1.1 mmoles) were mixed at room temperature (250C)

with 1.8 ml of dry tetrahydrofuran (THF) until completely

dissolved. The solution was then cooled to 0C and stirred

at this temperature for 5 minutes. Dicyclohexylcarbodiimide

(DCC) (206.0, 1.0 mmoles) was dissolved in 200 ul dry THF.

The DCC solution was added to the THF-GA3 mixture in 2 equal

lots. The second lot was added 2 minutes after the first.

This mixture was stirred at 0C for 1 hour and then over-

night at 40C. The mixture was filtered with suction and

washed 3 times with dry THF. The filtrate was collected and

the THF removed via rotary evaporation. The resultant syrup

residue, GA3 active ester was dissolved in 1.0 ml dioxane

and this solution was filtered through glass wool. Another

0.5 ml of dioxane was used to rinse the flask clean of

residue and was filtered through the same glass wool. The

GA3 active ester solution was added dropwise in 0.5 ml

aliquots to a 0C solution of BSA (170 mg) in 1.5 ml







water/0.5 ml dioxane. While the mixture was stirred, the pH

was adjusted to 7.5 with 1 M NaHCO3 and kept at pH 7.5 for

the remainder of these reactions. Thirty minutes later

another 0.5 ml of GA3 active ester was added and the mixture

was stirred at 0C for another 30 minutes. This mixture was

stirred for another 4 hours at room temperature. Next, the

mixture was cooled to 0C and the final 0.5 ml of GA3 active

ester was added and stirred for 30 minutes. The mixture was

then stirred overnight at room temperature. The GA3-BSA

conjugate was dialyzed for 16 hours against 1 liter

dioxane/water (1:4, v/v), and for 5 days against daily

changes of 1 liter, pH 7.5 0.1M phosphate buffer. The

coupling ratio of GA3 bound to BSA was ca. 3.1 mol GA3 to 1

mol protein.

Carbobenzoxychloride Protection of Amino Groups of

Amino-n-Caproic Acid. Amino-n-caproic acid (26.34 g, 0.2

moles) was dissolved in 25 ml 4 N NaOH and while stirring

was chilled to 50C on ice. Alternately, 30 ml of 4 N NaOH

and 15.65 ml (18.7 gm) of carbobenzoxychloride (CBZ,

Chemical Dynamics Corp.) were added to the amino-n-caproic

acid mixture in 5 equal portions over a 30 minute period,

starting first with CBZ. After each addition, the solution

was vigorously shaken and cooled in an ice bath. After all

5 portions had been added to the amino-n-caproic acid

solution, the pH was adjusted to 8.0-9.0 with 1 N NaOH and

the solution stood overnight at room temperature. This

solution was transferred to a separatory funnel. Water (50

ml) and then 100 ml of diethyl ether were added and the







mixture shaken. After shaking, the aqueous layer was

separated and saved and another 50 ml of fresh diethyl ether

added to the separatory funnel containing the aqueous phase.

This mixture was shaken and the aqueous layer recovered as

previously described. The aqueous fraction was cooled on

ice to 0C. Once cooled, the aqueous solution was added to

a separatory funnel and 150 ml of ethyl acetate was added

and the pH adjusted to 2.0 with 6 N HC1. The mixture was

shaken and allowed to separate. The aqueous phase was

collected again and an equal volume of ethyl acetate was

added to the separatory funnel and the mixture shaken. The

aqueous phase was drained and the ethyl acetate fractions

pooled. The pooled ethyl acetate fractions were washed with

25 ml of cold 1 N HC1 and the aqueous portion removed. The

ethyl acetate fractions were next washed with 30 ml of water

3 times, and the aqueous portion was removed each time. The

ethyl acetate fraction containing CBZ-amino-n-caproic acid

was dried over MgSO4 and then filtered with suction. Next,

the ethyl acetate was removed via rotary evaporation and a

thick oil resulted. To this oily residue, 20 ml of ethyl

acetate was added and the solution warmed to 600C over

steam. To this solution, warm (600C) petroleum ether was

added dropwise until the ethyl acetate solution containing

CBZ-amino-n-caproic acid became cloudy and then drops of

ethyl acetate were added to clear the solution. The

solution was cooled to room temperature and once crystals

precipitated from the solution it was put on ice. (m.p. 52-

54;46.6 gm). Thin layer chromatography of this product on







silica gel G (Analtech, Inc., Newark, DE) plates developed

in ethyl acetate:pyridine:acetic acid:water (200:2:6:1,

v/v/v/v) revealed one spot, Rf 0.85 (Rf 0.41 amino-n-caproic

acid). Spots were visualized by 0.2% ninhydrin in acetone,

O-tolidine/KI stain (93).

Preparation of t-butyl ester of N--CBZ-amino-n-caproic

acid. N6-CBZ-amino-n-caproic acid (0.15 moles; 39.7915 g)

was mixed with 300 ml of methylene chloride and stirred on

ice to 0C in a round bottom flask. To this solution, 1.5

ml of concentrated H2SO4 was added and 10 minutes later 100

ml of 99% isobutylene was bubbled into the flask while

constantly stirring at 0C. This clear solution was removed

from ice after 15 minutes, capped with a stopper and then

allowed to stir at room temperature for 2 days. This

solution was recooled to 0C and then transferred to a

separatory funnel and 50 ml of water was added and the

mixture shaken vigorously. Next, 25 ml of diethyl ether was

added and the mixture shaken. The aqueous layer was removed

and the remaining solution was washed with a mixture of 20

ml of saturated NaCl and 50 ml, 1 M NaHCO3. After removing

the aqueous layer, the remaining solution was washed 3

times, each with 50 ml saturated NaCl and the aqueous layer

removed each time. The remaining solution containing the

putative N6-CBZ-amino-n-caproic acid t-butyl ester was dried

over MgSO4 and then filtered. Methylene chloride was

removed via rotary evaporation until a thick clear syrup

resulted (40.0 g). The product was identified by infrared

spectrometry. IR(CHC13) carbonyll bands of t-butyl ester and







urethane of benzyloxycarbonyl ranging from 1680 to 1750 cm

-1; and NH band at 3450 cm-1].

Hydrogenolysis of N6-CBZ-amino-n-caproic acid t-butyl

ester. One gram of 10% Pd on carbon was placed in a glass

vacuum bottle and 4 ml (48 mmoles) of concentrated HC1 was

added dropwise to the Pd. Immediately thereafter, the

mixture was diluted with 125 ml absolute EtOH. The fully

protected N6-CBZ-amino-n-caproic acid t-butyl ester (16.06

g, 50.0 mmoles) dissolved in 75 ml absolute EtOH was

gradually added to the vacuum bottle and the mixture placed

on a Parr hydrogenator. Air was removed from the flask and

H2 gas added to 25 psi. The flask was shaken for 2 hours at
room temperature. After careful removal from the hydro-

genator, approximately 150 ml of the MeOH was added and the

mixture filtered quickly under suction to remove Pd

catalyst. The filtrate was collected and refiltered through

Whatman #1 paper and rinsed with more MeOH to remove all Pd.

Methanol was removed via rotary evaporation at 30-350C. To

the residue, 50 ml water was added and the mixture was

transferred to a separatory funnel. The mixture was shaken

and the aqueous layer removed. The organic fraction was

dried over MgSO4, filtered with suction, and the solvent

removed via rotary evaporation. Residual traces of solvent

were removed from the orange-colored product via vacuum

distillation at less than 1 mm Hg. Thin layer chromato-

graphy was performed and visualized with ninhydrin, o-

tolidine/KI stain as described previously (Rf 0.89 N6-amino-

n-caproic acid t-butyl ester; Rf 0.6 amino-n-caproic acid t-





82

butyl ester, 2.22 gm, 30.9%). The product was identified

and confirmed by infrared spectrometry and nuclear magnetic

resonance. IR(CHCl3) carbonyll of t-butyl ester at 1720
cm-1 in the absence of benzyloxycarbonyl] NMR (CDC13) [A

singlet (-O-C(-CH3)3 overlapping with a series of multiplets

(-CH2-CH2-CH2-) from 1 to 2.2 delta (15H); a multiple from

2.2 to 2.55 delta (2H;-CH2-C=0); multiple from 2.6 to 3.0

(2H;-N-CH2-) and a broad band from 3.0 to 3.5 delta (2H;-

NH2)].

Coupling of GA3 to amino-n-caproic acid t-butyl ester.

Amino-n-caproic acid t-butyl ester (221.85 mg, 1.18 mmoles)

was added to a round bottom flask by pipette and then 250 ul

dry DMF was added. This solution was constantly stirred and

GA3 (346.4 mg, 1.0 mmoles) was slowly added. An additional

0.5 ml of dry DMF was added while the GA3 was dissolving so

that a clear mixture resulted. The mixture was then cooled

on ice to 0C. Dicyclohexylcarbodiimide dissolved in 250 ul

dry DMF was added dropwise over a 7 minute period to the

GA3, amino-n-caproic acid t-butyl ester mixture and the

reactants were stirred at 0C for 1 hour. Next, 1-hydroxy-

benzotriazole monohydrate (153.14 mg, 1.0 mmole, Aldrich

Chemical Co.) dissolved in 250 ul of dry DMF was added over

a 1 minute period to the above mixture. This mixture was

stirred overnight at 40C and then centrifuged at 7,000 rpm

for 15 minutes. The supernatant was collected into a

separatory funnel and 10 ml of ethyl acetate added. The

ethyl acetate fraction was washed 4 times, each time with 1

ml of saturated NaCl and the aqueous phase was removed after







each wash. Next, the ethyl acetate fraction was washed 3

times, each time with 1 ml of 1 M citrate buffer, then,

washed each of 4 times with 1 ml saturated NaC1, and finally

washed 3 times, each time with 1 ml of a 1:1 (v/v) mixture

of 0.1 M NaHCO3/saturated NaC1. The aqueous fraction in

every case was removed after each wash. The ethyl acetate

fraction was collected and dried over MgSO4, filtered with

suction, and the solvent evaporated via rotary evaporation

and resulted in a gummy, yellow-green product (509.2 mg,

89.6%). This product was redissolved 5 ml ethyl acetate and

centrifuged at 7,000 rpm for 10 minutes to remove residual

dicyclohexylurea. The supernatant was collected and the

solvent evaporated via rotary evaporation until white

crystals remained (439.5 mg, 77.3%). Thin layer chromato-

graphy (Whatman reverse-phase MKFG silica gel 1 x 3 cm, 200

u thickness) plates developed in benzene:water:acetic acid

(9:1:9, v/v/v), and prepared for visualization under UV as

described by Cavell et al. (12) and MacMillan and Suter (64)

and/or by the ninhydrin, o-tolidine/KI stain, showed Rf 0.71

GA3-amino-n-caproic acid t-butyl ester. The product was

identified by infrared spectrometry. IR (CHCl3)[showed

carbonyl bands of amide from 1630 to 1680 cm-1; t-butyl

ester bands from 1690-1730 cm-1; lactone bands from 1740 to

1790 cm-1].

Removal of t-butyl ester from GA3-amino-n-caproic acid.

Gibberellin A3-amino-n-caproic acid t-butyl ester (439.5 mg)

was cooled on ice to 0C and 1.0 ml of anisole (Aldrich

Chemical Co.) was added and the mixture slowly stirred. One







ml of anyhdrous trifluoroacetic acid (CF3HCO2) was added to

the mixture and it was stirred for 15 minutes at 0C and

then 40 minutes at room temperature. The solvents were

evaporated via rotary evaporation and residual solvents were

removed with reduced pressure of less than 1.0 mm Hg. The

resulting reddish, syrup-like product was dissolved in 1.0

ml MeOH and loaded onto a 90.0 x 2.2 cm LH-20 (Pharmacia)

column equilibrated in MeOH. The putative GA3-amino-n-

caproic acid was chromatographed with 200 ml of degassed

MeOH and monitored by absorbance at 280 nm. Fractions 16,

17, 18 (4ml) were pooled and the solvent evaporated via

rotary evaporation and a rose-colored product remained.

Thin layer chromatography plates (Whatman reverse phase, as

above) developed in benzene:water:acetic acid (9:1:9, v/v/v)

and visualized as described previously, revealed 1 spot Rf

0.68.

Coupling of GA3-amino-n-caproic acid to BSA. A

solution of GA3-amino-n-caproic acid (341.35 mg, 0.743

mmoles) was dissolved in 0.5 ml DMF and cooled to 0C. To

this solution, l-ethyl-3-(3-dimethylamino propyl) carbo-

diimide (EDC) (427.32 mg, 2.23 mmoles) dissolved in 1.0 ml

DMF was added dropwise to the cold GA3-amino-n-caproic acid

solution and stirred constantly for 30 minutes. This

solution was added very slowly in drops to a 0C solution of

BSA (300 mg) in 2.0 ml water/0.5 ml DMF, pH 7.7. This

mixture was stirred for 30 minutes at 0C and then overnight

at 4C. The GA3-amino-n-caproic acid-BSA conjugate was

dialyzed against 1 1, DMF/water (1:4, v/v) at 40C for 1 day







and then against 1 1, 0.1 M Na phosphate buffer pH 7.5 for 4

days. Daily changes of Na phosphate buffer were made. The

coupling ratio of GA3 bound to BSA was ca. approximately 10

mol GA3 bound to 1 mol protein.

Preparation of Monomeric Adipic Anhydride. Adipic acid

(50.0 gm, 0.342 moles) and 150 ml of acetic anhydride were

refluxed together for 2 hours. About 100 ml of acetic acid

and acetic anhydride were removed under reduced pressure and

then an additional 100 ml of acetic anhydride were added.

The mixture was refluxed for another 2 hours and then the

acetic acid and acetic anhydride were removed under reduced

pressure at a temperature slightly below 1000C. The crude

product was purified by vacuum distillation and fractions

were collected at boiling points between 105 to 1250C at

less than 1 mm Hg.

Preparation of Mono-n-succinimidyl Adipate. A solution

of 10.0 gm (0.02 moles) of N-hydroxysuccinimate in a minimal

amount of DMF and ethyl acetate was added to the crude

product (11.8 gm, 0.092 moles, adipic anhydride) in a round

bottom flask (200 ml) while stirring. The reaction flask

was placed in a desiccator in vacuo under P205. Heat was

generated immediately after the addition of hydroxysuc-

cinimide. The reaction was run for 2 days at room

temperature. The solvent was removed under reduced pressure

in a hot water bath at 400C. A white, crystalline product

was obtained. It was quenched with anhydrous ether,

filtered and washed with ether. The crude product was

recrystallized with ethyl acetate and a white crystalline







product was obtained (m.p. 99-1000C). A second crystal-

lization from isopropyl alcohol/isopropyl ether yielded

white needles (m.p. 101-1030C; Fwt.=243/131; C 10; H 13; N

1; O 6). Calc. C=49.4 H=5.55 N=5.76 0=39.48: Found

C=49.15 H=5.41 N=5.91 IR(KBr)[carbonyl band of carboxylic

acid at 1700 cm-1; carbonyl band of ONSu at 1742, 1788, 1812

cm-1].

Preparation of GA4-mono-N-succinimidyl adipate-BSA. A

solution of mono-n-succinimidyl adipate (60.75 mg, 0.25

mmoles) and CDI (44.6 mg, 0.275 mmoles) in 1.25 ml dry DMF

was cooled to -200C while gently stirring. To this

solution, DCC (51.5 mg, 0.25 mmoles) in 0.75 ml was added

dropwise over a 3-minute period. The mixture was stirred at

-200C for 30 minutes and then at -50C for 30 minutes, and

finally at 0C for an additional 1 hour. To this reaction

mixture, we added a solution of GA4 and 35 ul triethylamine

(83.1 mg, 0.25 mmoles) dissolved in 0.5 ml dry DMF with

constant stirring at room temperature. The mixture was

stirred for 2 hours and then was centrifuged at 7,000 rpm

for 10 minutes. The GA4-active ester was added dropwise

over a 30 minute period to BSA (85.0 mg) dissolved in

water/DMF (2.0 ml/0.5 ml) at 0C. This solution was then

stirred overnight at 40C. The following day, again, GA4

active ester was prepared as described above and added to

the same BSA solution which had been cooled to 0C and had

an additional 0.5 ml water added. The GA4-adipic acid-BSA

conjugate was dialyzed against 1 1, DMF/water (1:4, v/v) for

1 day and thereafter against Na phosphate buffer as







previously described for GA3-amino-n-caproic acid-BSA. The

coupling ratio of GA4 bound to BSA was ca. 59 mol of GA4

bound to 1 mol protein.

Immunization Procedure. An antigen emulsion was

prepared by dissolving 1.5 mg lyophilized conjugate in 1.5

ml PBS (0.01 M Na phosphate, 0.15 M NaC1, pH 7.4) and

emulsified with 2.0 ml complete Freund's Adjuvant (Difco).

Randomly bred New Zealand white rabbits (12 to 16 weeks old)

were immunized with a total of 1.0 ml of freshly prepared

antigen emulsion by making 2, 0.25 ml intramuscular and 2,

0.25 ml intradermal neck injections. The injections were

made at weekly intervals for 4 weeks and the first

collection of antisera began 10 days after the fourth

injection. Boosters were administered thereafter biweekly

and the preparation (45) and titre (110) of antiserum was

tested 1 week after each booster until suitable for radio-

immunoassay use. Antisera were stored frozen at -180C.

Radioimmunoassay Procedure. The following reactants in

order of addition were pipetted into a test-tube in an

icebath: 200 ul of buffer (0.01 M Na phosphate, 0.15 M

NaCl, pH 7.4), 100 ul of standard or appropriately diluted

plant extract, 100 ul of tracer (l,2(n)-3H gibberellin in

A4, 38.2 Ci/mmol, Amersham; 2,000 cpm, 0.0524 pmol), 100 ul

of GA4-directed antibody diluted 1:125 (final assay

dilution). The reactants were swirled by vortexing and

incubated in the icebath for 1 hour. Next, 0.5 ml of ice

cold saturated (NH4)2SO4 was added to each tube to separate

free and antibody-bound haptene (24, 114) and the contents







swirled by vortexing. This solution was incubated in the

icebath for 30 minutes and then tubes were centrifuged for

30 minutes at 7,000 rpm. From each tube a 0.5 ml aliquot

was drawn and pipetted into a scintillation vial and 10 ml

of scintillation cocktail added and counted to 1%

significance. Blanks for non-specific binding were

essentially prepared by procedures described previously

(103, 114) where water was substituted for antiserum.


Results

Preparation of GA-BSA Conjugates. The preparation of

GA3-BSA conjugates by the methods of Weiler and Wieczorek

(113) (Figure V.2) and Atzorn and Weiler (Figure V.3) (4)

yielded relatively low coupling ratios of GA bound to BSA

and did not yield antisera directed against GA when screened

by the method of Ochterlony (83). We speculated that low

molar coupling ratios of GA3 to BSA were responsible for the

lack of GA-directed antibody production and tried to modify

the mixed anhydride or symmetrical anhydride conjugation

techniques in order to improve the coupling ratios and

obtain a GA antibody. One alternative procedure was to

adjust the pH of the reaction mixture during the production

of GA3 anhydride. Analysis of these GA3-BSA conjugates

which had been prepared at different pHs yielded no improve-

ment in GA3 to BSA coupling ratios. In another procedure,

the previously prepared GA3-BSA conjugates were rereacted by

the mixed anhydride or symmetrical anhydride procedure as

described previously. However, the rereaction of a GA3-BSA

conjugate did not help to improve the molar coupling ratio






































Figure V.2 Scheme of synthesis of GA -BSA conjugate by
the method of Weiler and Wieczorek (113).






Mixed Anhydride





OH' '- H
CO2H

(CH3)CCHC2-- -CI [CH3-(CH2)3]3 N


0

OH "OH

OC
o
O=C -O-CH2-CH(CH3)2

+ NH2


0
OH -~OH
C.O

HN --- immunize
/"BSA I




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E1IV9L7ER_9V1R06 INGEST_TIME 2014-04-14T19:19:47Z PACKAGE AA00017642_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES