Planta (1989)179:89 96 Planta �9 Springer-Verlag 1989

Stamens and gibberellin in the regulation of corolla pigmentation and growth in Petunia hybrida David Weiss and Abraham H. Halevy* Department of Horticulture, Hebrew University of Jerusalem, P.O. Box 12, Rehovot 76100, Israel

Abstract. Removal of stamens, or even of only the anthers, at an early stage of corolla development, before the start of main anthocyanin production, inhibited both growth and pigmentation of at- tached corollas of Petunia. When only one or two stamens were removed from one side, the inhibi- tion was restricted to the corolla side adjacent to the detached stamens. Application of gibberellic acid (GA3) substituted for the stamens in its effect on both growth and pigmentation. In detached corollas, isolated at the early-green stage and grown in vitro in sucrose medium, GA3 promoted growth and was essential for anthocyanin synthe- sis. A marked enhancement of anthocyanin pro- duction was observed 48 h before the increase in corolla growth rate. Corollas detached at later stages were able to continue their growth and pig- mentation in sucrose without GA3. When Paclobu- trazol (fi-[(4-chlorophenyl)-ethyl]-~(1,1-dimethy- lethyl)-H-l,2,4-triazol-l-ethanol), an inhibitor of gibberellin biosynthesis, was added to the growth medium of in-vitro-grown corollas, pigmentation was inhibited but there was no effect on corolla growth. Low levels of GA3 counteracted the Paclo- butrazol effect on pigmentation but did not affect growth. The above results indicate that the effect of GA3 (and probably that of the stamens) on cor- olla growth is independent of its effect on pigmen- tation. Gibberellic acid and paclobutrazol had no effect on [14C]sucrose uptake by in-vitro-grown corollas. The activity of phenylalanine ammonia- lyase was correlated with the effect of stamens and GA3 on pigmentation in corollas grown in vivo and in vitro.

* To whom correspondence should be addressed

Abbreviations: GA = gibberellin; GA3 = gibberellic acid; PAC = Paclobutrazot; PAL = phenylalanine ammonia-lyase

Key words: Anthocyanin - Corolla - Gibberellin and corolla growth, pigmentation - Petunia - Sta- men - Phenylalanine ammonia-lyase

Introduction

Anthocyanins are the most abundant pigments in flowers (see e.g. Harborne 1976). In the early stages of development petals are mostly green and con- tain chlorophyll. Coloration normally appears at the later stages of flower development, after the degradation of most of the chlorophyll. We have found, however, that the chlorophyll content of petunia corollas increases during the early stages of flower development, when anthocyanins appear and coloration becomes apparent (Weiss et al. 1988).

The process of petal pigmentation seems to be an integral part of flower development (see e.g. Biran and Halevy 1974). Klein and Hagen (1961) demonstrated that the pigmentation of in-vitro- grown corollas of Impa t i ens balsamina was related to their growth. All factors promoting petal growth also promoted anthocyanin production. Corolla growth can be affected by external factors such as light, temperature and photoperiod, and internal factors such as the supply of metabolites and correlative effects of the developing flower parts (see Kinet et al. 1985, Chpts. 4-7). Several papers have been published demonstrating the in- fluence of stamens on corolla growth (Mohan Ram and Rao 1984; Bala et al. 1985; Raab and Koning 1987).

In a few cases, gibberellin (GA) application has been reported to substitute for the missing stamens in promoting corolla growth (Mohan Ram and Rao 1984; Bala et al. 1985). Nester and Zeevaart

90 D. Weiss and A.H. Halevy: Stamens, GA, pigmentation and growth of Petunia corollas

(1988) found that the corolla did not develop in a GA-deficient tomato mutant. However, no infor- mation seems to be available on the involvement of stamens in the control of anthocyanin synthesis, while the involvement of GA in this process is un- clear. In some systems, GA inhibited anthocyanin production (Heinzmann and Seitz 1977), while in others it promoted pigment synthesis (Zieslin et al. 1974) and the activity of phenylalanine ammo- nia-lyase (PAL) (Christina et al. 1968), which is known to be a key enzyme at an early stages of anthocyanin biosynthesis (see e.g. Wong 1976).

The purpose of our work was to investigate the controlling mechanisms of the transition from the green to the colored stage of the corolla and the interrelationship between corolla growth and pigmentation in petunia. Is anthocyanin synthesis directly related to corolla growth, or are these two independent processes occurring simultaneously in the same organ?

Material and methods

Plant material. Seeds from an F1 hybrid of Petunia hybrida (cv. Hit Parade Rosa, Benary, FRG) were sown in multipot plastic trays filled with a peat:vermiculite mixture (1:1 v/v), and germinated in a greenhouse (27/18 ~ C day/night tempera- ture). Uniform seedlings at the 10-leaf stage were transplanted into 15-cm pots filled with peat:volcanic gravel mixture (1:1, v/v) and grown in a greenhouse (29/18~ C day/night) under natural photoperiod. Plants were irrigated three times a day, with fertilization (20-20-20 NPK) being supplied through an irrigation system. Flowers not used for experiments were re- moved from the plants twice a week, and plants were trimmed monthly so as to keep them compact. They were discarded after one year.

Determination of anthocyanin. Anthocyanins were extracted from fresh Petunia corollas (100 mg) using 10 ml of 1% HC1 in methanol. Concentrations were determined by absorbance measurements (UV-Vis Spectrophot ometer D MS 100S; Varian, Mulgrave, Australia) at 530 nm and 657 rim, using the formula As30-0.25 x A657 to correct for the chlorophyll and its degra- dation products present in the extract (Mancinelli et al. 1975). This formula should be used for petunia flowers since chloro- phyll was found to be present in the corolla at all stages of development (Weiss et a1.1988).

Removal of stamens. For stamen removal at early stages of flower development, one sepal was detached and a small longi- tudinal incision was made at the base of the corolla; through this incision the stamens were detached and removed with thin tweezers. The incision was done so as not to wound the main vascular veins. In control flowers, one sepal was removed and corollas incised but stamens were left intact.

Application ofGA. Gibberellic acid (GA3; Sigma Chemical Co. St. Louis, Mo., USA) and GA4+7 (ICI, Bracknell, Berkshire, UK) was dissolved in methanol and mixed with lanolin to a final concentration of 3.10-3M, and 0.2 ml was applied with a needleless microsyringe to the sites of stamen removal.

In-vitro culture. Detached flower buds were immersed in 1% Ca-hypochlorite for 10 rain and rinsed several times in deion- ized water. Sepals were removed, the corolla cut transversely, and the upper part (the future limb) was used for culture. Corol- la limbs were placed in culture vials containing (unless otherwise stated) 10 ml of 150 mM sucrose in deionized water adjusted with KOH to pH 5.5. In some experiments GA3, Paclobutrazol (PAC) or both were included in the culture media. The cultured tissue was normally incubated for 72 h at 24 ~ C with constant light from cool-white fluorescent lamps (Tadiran, Tel Aviv, Israel) at 80 ~tmol. m - 2. s- t.

[14C]Sucrose uptake by excised corollas. Corolla sections taken at stage 2 (see Results) were used. Twenty milligrams of corolla tissue was placed in vials and [z4C]sucrose (Amersham, Bucks, UK) was added to each vial containing 5 ml of 150 mM cold sucrose, to give a final specific activity of 48.1 kBq/mmol. The vials were constantly shaken by a shaking device during incuba- tion, the latter was conducted under the conditions stated under In-vitro culture. After incubation the tissue was removed from the solution and blotted with filter paper. Free space was washed for 10 min with cold sucrose (150 mM) followed by a distilled-water wash; after this, the tissue was blotted dry and frozen in liquid nitrogen. Each sample was extracted in 0.5 ml in tissue solubilizer (Soluene 350; Hewlett-Packard In- struments Co., Palo Alto, Cal., USA) for 24 h, after which 3 ml of scintillation solution (Lumex [Lumar, Landgraat, The Netherlands]:xylene, 1:2.5, v/v) was added to each sample. Samples were counted in a Beckman Instruments (Palo Alto, Cal., USA) LS 7800 scintillation counter and the results calcu- lated as uptake (in ~g) per gram fresh weight (FW).

Extraction and assay of phenylalanine ammonia-lyase (PAL, EC 4.3.1.5). Phenylalanine ammonia-lyase was extracted at 4~ from 150 mg of corolla tissue by homogenizing with 5 ml of cold 25 mM Na-borate buffer, pH 8.8, containing 2 mM Na- bisulfite. A clear supernatant was obtained by centrifugation of the homogenate at 12000.g for 10 min. Samples were taken for protein determination according to Bradford (1976).

The enzyme was assayed according to Zucker (1969). The reaction mixture contained 30 mM of borate at pH 8.8, 2 mM of L-phenylalanine (saturated solution) and 0.2 ml of enzyme extract. The final volume was 3.7 ml. The rate of the enzymatic reaction was measured spectrophotometrically (see Determina- tion of anthocyanin content) at 35 ~ C by following the increase in absorbance at 290 nm over a period of 2 h. The activity was linear for 2 h with time and with protein concentration up to 60 ixg-ml - I .

Results

Flower development, and pigmentation. Develop- ment of petunia flowers was divided into seven stages (see Fig. I in Weiss et a1.1988). Figure 1 shows an apparent similarity in the courses of flow- er-bud development and pigmentation. Stage 1 is a small flower bud with the corolla just protruding from the clayx. Corollas at stages I through 3 are green: their anthocyanin content is negligibe and they exhibit a slow growth rate. The average bud lengths at stage 1 and stage 3 were 3 and 9 ram, respectively. By the end of stage 3 the stamens are already developed. The transition from stage 3 to

D. Weiss and A.H. Halevy: Stamens, GA, pigmentation and growth of Petunia corollas 91

F I I I I I I I ](Q) (o; J 1

4O0

1

30o%

kt-

200

u~ 1 0 0

180 150 120 90 60 30 0 30 60 TfME(h) 1 2 3 4 5 6 7 STAGE

Fig. 1. Changes in fresh weight and anthocyanin concentration of petunia corollas with time and developmental stage. Anthesis was designated as time 0. Means of 20 flowers; bars = SE

k" Lt_

O

v 1 z

32

<~

Table 1. Effect of removal of all five stamens at various develop- mental stages on growth and anthocyanin content of petunia corollas. Measurements were taken when untreated controls of each group reached anthesis. Mean of 20 flowers ,,1, SE

Stage at which Fresh weight Anthocyanin stamens were removed (rag) (A530/100 mg)

Intact 346 ,,1, 27 1.91 ,,1, 0.22

Incision (stage 2), stamens intact 357 ,,1, 31 1.82 _-t- 0,27

2 34_+45 0.02_+0.0 3 96_+21 0.05,,1,0.01 4 254_+ II 0.91 ,,1, 0.15 5 359,,1, I4 1.78-+0,19

stage 4 is characterized by the advent of antho- cyanin accumulation and by a sharp increase in the corolla growth rate. Bud length at stage 4 was 15 ram. The bud continues growing throughout stages 5 and 6 with a parallel increase in pigmenta- tion. Anthesis occurs at the end of stage 6: at this stage flower length was 50 mm. At stage 7 (70 mm) there is still additional growth, but almost no an- thocyanin increase, the anthocyanin level therefore decreasing. Accumulation of anthocyanin in stages 4 through 7 occurred almost exclusively in the upper part of the corolla (the limb), which be- came pink, while the lower part (the tube) was white with a light-greenish tint.

Effect of stamen removal on corolla growth and pig- mentation. Different numbers of stamens were re- moved from attached flower buds at various stages of development. When control flowers reached stage 6, all the corollas were detached, weighed and pigments were extracted. Removal of stamens

4 0 0 - STAMEN NO.

3oo []-3 i / / ~// / /

0 2-~ ] -~/ / / 2 3 4 5

STAGE AT WHICH STAMENS WERE REMOVED

Fig. 2. Effect of removal of various numbers of stamens at various developmental stages on corolla growth in Petunia. Measurements were taken when control flowers of each treat- ment group reached anthesis. The open column at left repre- sents the weight of intact control flowers. Means of 20 flowers; bars = SE

Table 2. The effects of excision of two stamens on growth and pigmentation of petunia corollas. The two stamens were re- moved from one side of young flower buds at stage 3. When untreated control flowers reached anthesis the treated flowers were dissected into two parts. Weight and anthocyanin content were determined for each side. Means of 30 flowers. Data in percent of untreated corollas. Fresh weight of control flowers was 374 mg and ar/thocyanin (A53o/100 mg FW) was 1.32

Treatment FW Anthocyauins (%) (%)

Side with intact stamens 98 104 Side with excised stamens 71 63

greatly inhibited corolla growth and anthocyanin formation (Table 1, Fig. 2). Inhibition was very pronounced when the removal was done at early stages. Removal of even one stamen at stage 2 strongly retarded corolla growth and pigmenta- tion. With advancing developmental stages inhibi- tion decreased and removal of more stamens was required to retard growth or pigmentation. At stage 5, stamen removal no longer had any effect on both processes (Fig. 2, Table 1).

When only one or two stamens were excised at an intermediate stage (stages 3 and 4), there was a local inhibition of growth and pigmentation.

This observation was further verified by removing two adjacent stamens at stage 3, which indeed in- hibited growth and pigmentation only on the side adjacent to the excised stamens (Table 2). Excision of only the anther (leaving the filament intact) had the same inhibiting effect as the removal of the entire stamen. The removal of stamens on one side also caused curving of the corolla. Incision without stamen removal caused neither growth inhibition nor curving.

92 D. Weiss and A.H. Halevy: Stamens, GA, pigmentation and growth of Petunia coroUas

Table 3. The effects of stamen removal and GA3 application on growth and pigmentation of petunia corollas in vivo. Two or five stamens were removed from flower buds at stage 2. Gibberellic acid (3.10-3 M in lanolin) was applied to the sites of the excised stamens. Corolla weight and anthocyanin content were measured when control flowers reached anthesis. Means of 30 flowers per treatment • SE

Treatment Weight increment Anthocyanin

mg % of A53o/ % of control 100rag FW control

Control 320+_20 100 1.67_+0.12 100

Removal of 30_+6 9 0.10_+0.05 6 2 stamens

Removal of 150_+8 46 1.36+0.14 81 2 stamens + GAa

Removal of 0 0 0.04-+0.01 2 5 stamens

Removal of 70 4- 7 22 0.61 4- 0.06 36 5 stamens + GA~

In further experiments we found that excision of the sepals and the pistil, without damaging the stamens, had no effect on corolla growth and pig- mentation (data not shown).

Effect of GA3. Different numbers of stamens were removed at stage 2 and various growth regulators were applied at the site of excision. Benzyladenine (N6-benzylaminopurine), indole-3-acetic acid (IAA) and naphthalene-l-acetic acid (NAA) at 10 - 6 to 10 -3 had no effect on corolla growth and antho- cyanin production (data not shown). In contrast, GA3 replaced the excised stamens with respect to promotion of both growth and pigmentation (Ta- ble 3). The GA effect was more pronounced when only two stamens were removed, as compared to flowers with five excised stamens. No additional growth and pigmentation was obtained by using higher concentrations (up to l0 -2 M) of GA. Gib- berellin A4+7 was less effective than GA3 in pro- moting growth and pigmentation (data not shown).

Anthocyanin production in excised corollas. Corolla limbs at various developmental stages were grown in vitro for 72 h. The results, presented in Fig. 3, show that corollas at stages 2 and 3 required su- crose for both growth and pigmentation. Adding GA3 to the growth medium more than doubled corolla growth and anthocyanin synthesis, versus sucrose alone. At stage 4, however, GA3 had no effect on pigmentation and somewhat inhibited corolla growth. Gibberellic acid without sucrose

200

t-- Z

f~ 100

I

0.6 !

kL

0.4 o o

_z 0.2 Z

~A O 22

< 0

[--1 WATER 17~ SUCROSE

- I ~ S U C R O S E + GA 3

FFF'n G A 3 x ~

A ~ x~ x x

x x

X N

5(" / b("A Dz.X IXX ?x 'x

X X r r n

7'T / /

/ / / /

Z / / / / / / / / / / / / / / / / / / / / / / / / / /

/ / / /

F - / /

x ; ~ x x ] x x l x x ] xx.J x x I x: Xl

xN ] x.-xJ x .x ] x, x l "K'x]

,%XFFFrl

B

x 'X

/ / X . N )cxq / / ~

NYX] / / g N

/ /

• 2 3 4 FLOWER DEVELOPMENTAL STAGE

Fig. 3A, B. Effect of sucrose (150 raM) and GA3 (5.10 5 M) on growth (A) and anthocyanin content (13) of petunia corollas grown in vitro. Corollas were detached from flowers at various developmental stages and incubated for 72 h in the various media. Means of 30 replications; bars = SE

had no effect on either growth or anthocyanin con- tent (Fig. 3).

The kinetics of growth and pigmentation was studied with corolla limbs of stage 3 grown in vitro in media containing sucrose or sucrose+GA3 (Fig. 4). In sucrose medium both growth and an- thocyanins advanced in a parallel manner. In su- crose + GA3 medium, the pigment production was advanced by GA3 much earlier than growth. The lag period for anthocyanins was only about 12 h, that for growth more than 48 h.

Effect of Paclobutrazol. When young corolla limbs were grown in a sucrose medium without GA, only slight pigmentation was observed. To test whether this was a consequence of GA synthesis in the cor- olla, we added to the medium the GA-biosynthesis inhibitor Paclobutrazol (PAC) at 10 mg.1-1. This concentration was found in preliminary studies to be effective in inhibiting pigmentation. When PAC was added to the growth medium, it totally inhib- ited further anthocyanin synthesis but had no ef- fect on corolla growth (Fig. 5). Adding GA3 at

D. Weiss and A.H. Halevy: Stamens, GA, pigmentation and growth of Petunia corollas 93

0.5

~ 0,4

E o0.3

zO,2 Z

{o

~

<[

- - 1 I I ] t

12 24 48 72 96 TIME (h)

30

E b-

20 Z ~E bJ CE: %; Z

�9

0 u

Fig. 4. Growth and anthocyanin content of petunia corollas detached at stage 3 and grown in vitro in medium containing sucrose or sucrose + GA3. Growth is expressed as FW increment per detached corolla. Means of 16 replications; bars= 2SE

0.5

[] OA--

L.

~D 0.1 / ~

~o ~ S

~7 / / / / / / / / / / / /

/ / / /

$+PAC

i~ I 11 ~/ //// 1"~0 z

fs

/ ' 4 / / / O

;-.q ~ r0 8 S+PAC+ S + G A 3 S+PAC +

GA3(5,10-7M) (5.10 -7 M) GA3(5.10-SM)

Fig. 5. Effects of PAC and GA3 on growth and anthocyanin synthesis in petunia corollas grown in vitro. Corollas detached at stage 2 were incubated for 72 h in PAC (10 nag/l) with or without GA3 (5.10 v M or 5-10 5 M). Growth is expressed as FW increment per detached corolla. Means of 20 replica- tions; bars = SE

a low concentration (5-10 - 7 M) to the sucrose or to the PAC-containing medium greatly promoted anthocyanin synthesis but had no effect on corolla growth. A higher concentration of GA3 (5. 10 -5 M) promoted both growth and pigmenta- tion.

When PAC was applied to corolla tissues at different stages of growth it was found (Fig. 6) that at the early stages, in which the tissue is able to produce small amounts of anthocyanin in sucrose media alone, but in which GA3 greatly promotes anthocyanin synthesis (see Fig. 3), PAC greatly in- hibited pigment production, while in stage 4, PAC (like GA3) had no effect on pigmentation. Similar results were obtained with 50 rag.11 of PAC (data not shown).

0,6

L~

'o.4 O O

z 0 , 2

p-

0

_ I ~ SUCROSE

I~ SUCROSE+PAC . [~ -

- T IIIHI - ITIT]] SUCROSE + PAC + GA 3 v / / /

_ / /

/ / - - / /

v ~

/ . / / / / / / / / / / / / / / / / /

2 3 4 DEVELOPMENTAL STAGE

Fig. 6. Effects of PAC (10 rag/l) alone or with GA3 (5-10-5 M) on anthocyanin production in petunia corollas in vitro. Corol- las were detached from flower buds at various developmental stages and incubated for 72 h. Means of 20 replications; bars = SE

Sucrose uptake in vitro. The results in Fig. 3 indi- cate that young corolla tissue which contains very little or no anthocyanin, can produce some antho- cyanins when supplied with sugar. Gibberellic acid enhances pigment production only in the presence of sucrose. One possible explanation is that GA3 works by promoting sucrose uptake by the tissue. This possibility was thus examined. The rate of [14C]sucrose uptake in vitro was linear in the first 48 h in both media, but although anthocyanin syn- thesis was promoted by GA3, there was no effect of GA3 on sucrose uptake by the corollas (data not shown).

Effect of GA3 on PAL activity. Phenylalanine am- monia-lyase activity is often related to anthocyanin

94 D. Weiss and A.H. Halevy: Stamens, GA, pigmentation and growth of Petunia corollas

0.8 (o)

_--- I I I I I k: (o)

o200 0.6 cn E

E o o

�9 c~ 0.4 q" Cl

100

'5 o

s . .J

0 - 0 I 2 3 4 5

DEVELOPMENTAL STAGE Fig. 7. Phenylalanine ammonia-lyase activity and anthocyanin content of petunia corollas at various developmental stages. Means of four different experiments; bars = SE

Table 4. Effects of stamens, GA3 and PAC on PAL activity in petunia corollas in vivo and in vitro. For in-vivo studies all stamens were removed from flower buds at stage 2. Some flowers received GA3 (3.10-3 M) at the site of the excised sta- mens. The PAL activity was measured 30 h after treatment. For in-vitro studies, corollas of stage 3 were incubated in a solution containing sucrose (150 raM) or sucrose (150 mM) and GA3 (3.10 3M) or PAC (10 mg.1-~). The PAL activity was measured after 30 h incubation. Means of four different experi- ments _ SE

Treatment PAL

Activity (nmol cinnamic acid. (rag protein)- ~ .h -1)

K~ (mM L-phenyl- alanine)

In vivo

Control 72_+ 6.1 0.153 Stamens removed 20_+ 2.3 Stamens removed 105 _+ 12.0 + GA3

In vitro

Water 33 __+ 6.0 - Sucrose 98 +_ 8.0 0.170 Sucrose + GAa 179 + 9.0 0.165 Sucrose + PAC 36 _+ 4.0 0.170

synthesis, but its product may also serve as a pre- cursor for other metabolites (Wong 1976). When PAL activity was measured in corolla tissue at var- ious developmental stages (Fig. 7), it was found to be closely related to that of anthocyanin produc- tion. When corolla limbs at stage 3 was incubated in a medium containing GA3 or PAC (Table 4),

GA3 promoted and PAC inhibited PAL activity in parallel with their effects on anthocyanin synthe- sis (see Figs. 3, 5, 6). However, the growth regula- tors had no effect on the K m of the enzyme. Gib- berellic acid had no effect on enzyme activity when added to a cell-free enzyme extract (data not shown).

Phenylalanine ammonia-lyase activity was also measured in flowers from which stamens had been removed, with and without the application of GA3 (Table 4). In all cases, PAL activity paralleled the effect of the various treatments on anthocyanin production.

Discussion

The questions posed at the start of this study were: What is the regulating mechanism of corolla pig- mentation, and is the control of corolla pigmenta- tion tied directly to its growth? Our work shows that although under normal conditions anthocyan- in synthesis closely parallels corolla growth (Fig. 1), the two processes seem to be regulated by the same factors but in two independent con- trolling mechanisms, operating simultaneously.

In the early stages of flower development (the slow-growth stages), both growth and pigmenta- tion seem to be controlled by the stamens, or more specifically the anthers (Fig. 2, Tables 1, 2). We have established by microscopic observations that the pollen mother-cells have already undergone meiosis by stage 2. Most of the corolla enlargement occurs after the stamens are developed.

The effect of the stamens seems to be local and restricted to the corolla part adjacent to the stamen (Table 2). It seems that the anthers produce gibber- ellin which is then transported to the corolla where it stimulates both growth and anthocyanin synthe- sis. Gibberellic acid could act as a partial substitute for the excised stamens; GA~ + 7 was less effective. The fact that these GAs did not completely substi- tute for the stamens may indicate either that the stamens produce an additional controlling factor, or that the endogenous GA is another gibberellin which is more active in the process than GA3 and GA~ + 7.

Promotion of growth and pigmentation by GA3 was also found in isolated corolla limbs grown in vitro. Here, too, GA is only required in the early developmental stages. Similar results with GA were obtained with corollas of lpomoea nil (Raab and Koning 1987). Adding PAC, an in- hibitor of GA biosynthesis, to the medium, inhib- ited pigmentation at these early stages. This indi- cates that some GA is also synthesized in the young

D. Weiss and A.H. Halevy: Stamens, GA, pigmentation and growth of Petunia corollas 95

corolla tissue, i.e. that not all of the GA present in the corolla is transported from the stamens.

When the flower reaches the rapid-growth stage (Fig. 1), the corolla is no longer dependent on the stamens or exogenous GA for growth and pigmentation. At these later stages PAC did not inhibit corolla growth and pigmentation in vitro. This seems to indicate either that GA is required only for the induction of corolla growth and an- thocyanin synthesis and not for the maintenance of these processes, or that by these later stages the corolla has accumulated sufficient GA, either from the anthers or from synthesis at earlier stages, to proceed autonomously with the processes of growth and pigmentation. Koning (1985) found a sharp increase in GA levels in Gaillardia petals at the start of the corolla fast growth stage, followed by a decrease later on. This may support the view that GA is mainly required for signalling the initial growth process, and that another, unidentified controlling factor is involved in the later stages of corolla growth.

Although GA promotes both corolla growth and pigmentation, our kinetic studies (Fig. 4) and the effect of PAC (Fig. 5) demonstrate that these are two independent processes. A marked increase in anthocyanin content in vitro in response to GA3 was observed 48 h before the increase in growth (Fig. 5), Paclobutrazol, which apparently inhibited GA biosynthesis in the corolla, totally inhibited anthocyanin synthesis but had no effect on corolla growth (Fig. 5). This inhibition was counteracted by a low concentration of GA3, again with no ef- fect on growth. These results may indicate that although both processes are regulated by GA, lower concentrations are required for anthocyanin production than for corolla growth, They also demonstrate that while a certain level of growth can proceed without GA, GA is essential for an- thocyanin synthesis.

The presence of sugar in the growth medium was found to be essential for anthocyanin produc- tion in vitro, as has been shown before in other tissues (Cordts et al. 1987). The possibility that GAa promotes anthocyanin synthesis by enhanc- ing sucrose uptake by cultured corollas has been ruled out. However, these results do not eliminate the possibility that, in the intact flower buds, the anthers and GA are involved in the mobilization of metabolites to the corolla, especially in the early stages of flower development (Halevy 1987).

Little is known about the involvement of GA in anthocyanin synthesis. Phenylalanine ammonia- lyase is a key enzyme in the synthesis of phenolic compounds, including anthocyanins, The activity

of PAL was found to be closely correlated with the various factors affecting anthocyanin produc- tion both in vivo and in vitro (Fig. 7, Table 4), The fact that GA3 and PAC applications, respec- tively, promoted and inhibited both anthocyanin synthesis and PAL activity, supports the hypothe- sis that at least part of the GA promotion of antho- cyanin synthesis is based upon promotion of PAL activity. The mode of action of GA on PAL activi- ty is not known. Gibberellic acid had no effect on the activity of the enzyme in a cell-free extract, or on the Km of the enzyme.

In summary, our work shows the existence of an interorgan system regulating both corolla growth and pigmentation in petunia flowers. In the early developmental stages of the flower, the anther produces a signal which promotes corolla pigmentation and growth. This signal appears to be gibberellin. Although growth and anthocyanin synthesis occur simultaneously in the same organ, and both are regulated by the same phytohormone, they are two independent processes. The fact that young corollas requires only sucrose and GA for pigment synthesis in vitro may indicate that these are the only external regulating factors controlling pigmentation of this tissue.

We are thankful to Dr. Meira Ziv (Rehovot) for adivse on the in-vitro culture of corollas. This study was supported by the Pearlstein Fund for Research in Ornamental Horticulture at the Hebrew University. We thank the donors for their kind help.

References

BaIa, R., Rao, I.V.R., Mohan Ram, H.Y. (1985) Influence of stamens, gibberellic acid and sucrose on corolla growth in Gladiolus. J. Plant Physiol. 122, 87-92

Biran, I., Halevy, A.H. (1974) Effect of short-term heat and shade treatments on petal colour of'Baccara ~ roses. Phys- iol. Plant. 31, 180-185

Bradford, M.M. 0976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254

Christina, K., Cheng, C., Marsh, H.V. (1968) Gibberellic acid promoted lignification and phenylalanine ammonia-lyase activity in dwarf pea (Pisum sativum). Plant Physiol. 43, 1755 1759

Cordts, J.M., Scorza, R., Bell, R.L. (t987) Effect of carbohy- drates and nitrogen on the development of anthocyanins of a red leaf peach (Prunus persica (L.). Batsch) in vitro. Plant Cell Tiss. Org. Cult. 9, 103-J10

Halevy, A.H. (1987) Assimilate allocation and flower develop- ment. In: Manipulation of flowering, pp. 363-378, Ather- ton, J.G., ed. Butterworths, London

Harborne, J.B, (t976) Function of flavonoids in plants. In: Chemistry and biochemistry of plant pigments, pp. 736-778, Goodwin, T.W., ed. Academic Press, London

Heinzmann, U., Seitz, U. (1977) Synthesis of phenylalanine am-

96 D. Weiss and A.H. Halevy: Stamens, GA, pigmentation and growth of Petunia corollas

monia-lyase in anthocyanin-containing and anthocyanin- free callus cells of Daucus carota L. Planta 135, 63-67

Kinet, J.M., Sachs, R.M., Bernier, G. (1985) The physiology of flowering, vol. III. CRC Press, Boca Raton, Fla., USA

Klein, A.O., Hagen, C.W. (1961) Anthocyanin production in detached petals of Impatiens balsamina. Plant Physiol. 36, 1-9

Koning, R.E, (1985) Gaillardia. In: Handbook of flowering, vol. 5, pp. 117-126, Halevy , A.H., ed. CRC Press, Boca Raton, Fla., USA

MancineLli, A.L., Yang, C.P.H., Lindquist, P., Anderson, O.R., Rabino, I. (1975) Photocontrol of anthocyanin synthesis. IIl. The action of streptomycin. Plant Physiol 55, 251 257

Mohan Ram, H.Y., Rao, I.V.R. (1984) Physiology of flower bud growth and opening. Proc. Indian Acad. Sci. 93, 353- 374

Nester, J.E., Zeevaart, J.A.D. (1988) Flower development in normal tomato and a gibberellin-deficient (GA-2) mutant. Am. J. Bot. 75, 45-55

Raab, M.M., Koning, R.E. (1987) Changes in responsiveness to ethylene and gibberellin during corolla expansion of Ipo- moea n//. J. Plant Growth Regul. 6, 121-131

Weiss, D,, Schonfeld, M., Halevy, A.H. (1988) Phytosynthetic activity in the Petunia corolla. Plant Physiol. 87, 660-670

Wong, E. (1976) Biosynthesis of flavanoids. In: Chemistry and biochemistry of plant pigments, pp. 464-526, Goodwin, T.W., ed. Academic Press, London

Zieslin, N., Biran, I., Halevy, A.H. (1974) The effect of growth regulators on the growth and pigmentation of 'Baccara' rose flowers. Plant Cell Physiol. 15, 341-349

Zucker, M. (1969) Induction of phenylalanine ammonia-lyase in Xanthium leaf disks. Photosynthetic requirement and ef- fect of day length. Plant Physiol. 44, 912-922

Received 29 January; accepted 14 April 1989