Gibberellins and ParthenocarpicAbility in Developing ... · incompatible Clementine mandarin may...

Plant Physiol. (1992) 99, 1575-15810032-0889/92/99/1 575/07/$01 .00/0

Received for publication November 11, 1991Accepted March 2, 1992

Gibberellins and Parthenocarpic Ability in DevelopingOvaries of Seedless Mandarins1

Manuel Talon*2, Lorenzo Zacarias, and Eduardo Primo-Millo

Michigan State University-Department of Energy Plant Research Laboratory, East Lansing,Michigan 48824 (M.T.); and Department of Citriculture, Instituto Valenciano de Investigaciones Agrarias,

46113 Moncada, Valencia, Spain (L.Z., E.P.-M.)

ABSTRACT

Satsuma (Citrus unshiu [Mak] Marc.) and Clementine (Citrusreticulata [Hort.] Ex. Tanaka, cv Oroval) are two species of seedlessmandarins differing in their tendency to develop parthenocarpicfruits. Satsuma is a male-sterile cultivar that shows a high degreeof natural parthenocarpy and a high fruit set. Seedless Clementinevarieties are self-incompatible, and in the absence of cross-polli-nation show a very low ability to set fruit. The gibberellins (GAs)GA53, putative 17-OH-GA53, GA44, GA17, GA19, GA20, GA29, GA,,3-epi-GAI, GA8, GA24, GA9, and GA4 have been identified fromdeveloping fruits of both species by full-scan combined gas chro-matography-mass spectrometry. Using selected ion monitoringwith [2H2]- and [13C]-labeled internal standards, the levels of GA53,GA44, GA19, GA20, GA1, GA8, GA4, and GA, were determined indeveloping ovaries at anthesis and 7 days before and after anthesis,from both species. Except for GA8, levels of the 13-hydroxy-GAswere higher in Satsuma than in Clementine, and these differenceswere more prominent for developing young fruits. At petal fall,Satsuma had, on a nanograms per gram dry weight basis, higherlevels of GA53 (10.4x), GA44 (1 3.9x), GA1, (3.0x), GA20 (1 1.2x), andGA1 (2.0x). By contrast, levels of GA8 were always higher inClementine, whereas levels of GA4 did not differ greatly. Levels ofGA, were very low in both species. At petal fall, fruitlets of Satsumaand Clementine contained 65 and 13 picograms of GA1, respec-tively. At this time, the application of 25 micrograms of paclobu-trazol to fruits increased fruit abscission in both varieties. Thiseffect was reversed by the simultaneous applications of 1 micro-gram of GA3. GA3 alone improved the set in Clementine (13x), buthad little influence on Satsuma. Thus, seedless fruits of the self-incompatible Clementine mandarin may not have adequate GAlevels for fruit set. Collectively, these results suggest that endoge-nous GA content in developing ovaries is the limiting factor con-trolling the parthenocarpic development of the fruits.

Initial indications that GAs may be involved in set anddevelopment of citrus fruits are associated with the obser-vation that exogenous applications of GA3 improve fruit setof certain citrus species and cultivars, such as Clementine

' Financial support, through fellowships to M.T. (Instituto Na-cional de Investigaciones Agrarias) and L.Z. (Ministerio de Educaciony Ciencia), and through research grant 8206 to E.P.-M. from theInstituto Nacional de Investigaciones Agrarias, Madrid, Spain, isgratefully acknowledged.

2 Present address: Department of Citriculture, Instituto Valencianode Investigaciones Agrarias, 46113 Moncada, Valencia, Spain.

mandarins (Citrus reticulata [Hort.] Ex. Tanaka) (20). SeedlessClementine mandarins are self-incompatible, and in the ab-sence of cross-pollination show a very low ability to set fruit.Conversely, many cultivated species of citrus set naturallyseedless fruit. Satsuma mandarin (Citrus unshiu [Mak] Marc.),a male-sterile cultivar, normally shows both a high degree ofparthenocarpy and a high fruit set.

Typically, GA applications to Satsuma cultivars have notbeen effective in increasing fruit set (2). Qualitative analysesof endogenous GAs performed by GC-MS indicated thatmembers of the 13-hydroxylation pathway (GA53, GA44,GA39, GA20, GA1, GA29, and GA,) are predominant in devel-oping ovaries of Satsuma (9, 23). Of these GAs, only the twoinactive end products of the early 13-hydroxylation pathway,GA29 and GA8 were detected in Clementine (23). This GApathway also appears to be present in both reproductive (16,21, 25) and vegetative tissues (17, 21, 25) of Citrus sinensis(L.) Osbeck.

Additionally, reproductive organs of citrus also contain, atlower levels, non- and 3,B-hydroxylated GAs such as GA24,GA25, GA9, and GA4 (9, 16, 21). Gibberellin A3 and/or iso-GA3 have been occasionally detected in vegetative shoots(17) and developing fruitlets (25) of C sinensis, as well as indeveloping ovaries of Satsuma and Clementine (23). How-ever, because commercial applications of GA3 in citrus areusual, the identification of these compounds as endogenouscomponents has been questioned (16, 17). In C sinensis, wehave shown that parthenocarpic mutants with a fruitingpotential similar to the seeded genotype also have the samepattern of GA accumulation and practically the sameamounts of GAs (21). We have also suggested that theinability of Clementine mandarins to set a high percentageof parthenocarpic fruits appears to be associated with lowerlevels of endogenous GA-like substances (23). In the presentinvestigation, we report the main GAs from developing fruitsof Satsuma and Clementine mandarins. We also note thelevels of endogenous GA53, GA44, GA59, GA20, GA,, GA8,GA4, and GA, in developing ovaries of these seedless culti-vars and examine the effects of both paclobutrazol and GA3on fruit set and abscission.

MATERIALS AND METHODS

Plant Material

Clementine (Citrus reticulata [Hort] Ex. Tanaka, cv Oroval)and Satsuma (Citrus unshiu [Mak] Marc., cv Owari) mandar-

1575

www.plantphysiol.orgon April 29, 2020 - Published by Downloaded from Copyright © 1992 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

Plant Physiol. Vol. 99, 1992

ins were selected for these studies because both yield exclu-sively seedless fruit (14, 23). For GA analyses, reproductiveorgans were harvested at random from mature trees at thefollowing reproductive stages (3): stage C, ovary at stage ofyoung pollen (-7 DAF); stage E, ovary at pollination (0 DAF);stage F, ovary at petal fall (7 DAF); and stage I, developingfruit (42 DAF). Because GA3 is frequently applied to citrusfor commercial reasons, samples were collected in experimen-tal orchards where GA3 treatments had never been made.After measurements of fruit growth, samples were frozen inliquid N2 and lyophilized. Identification and quantificationanalyses of endogenous GAs were performed on two inde-pendent harvests sampled in 1990 and 1991. Growth andabscission parameters were determined on batches of 300tagged fruits per treatment using four different trees of eachspecies.

GA and PP3333 Treatments

Gibberellin A3 was dissolved in 5% v/v aqueous ethanolcontaining 0.05% v/v Tween 20 (20 ,ugmL-1). PP333 [1-(4-chloroethyl) 4,4-dimethyl-2-(1,2,4-triazol-1-yl) pentan-3-0],an inhibitor of the GA biosynthesis (19), was dissolved in0.05% v/v Tween 20 (500 ug.mL-1). A mixture of GA3 andPP333 at the above concentrations was also prepared. De-veloping fruits at petal fall (10 DAF) from both varieties wereindividually treated at random with GA3, PP333, or with amixture of both. One single application per fruit was madeby dipping the developing fruit in the solution for 3 to 5 s.This technique delivers about 50 ,tL of solution per fruit.Control fruits were treated with water containing Tween 20and ethanol. After application, abscission was recordedweekly for a period of 42 d.

Identification of GAs by GC-MS

Forty grams dry weight of developing fruit at stage I fromSatsuma and Clementine mandarins were extracted for iden-tification of GAs. The samples were homogenized in 1 L ofcold methanol, the solvent was removed, and the residuewas resuspended in 300 mL of 80% aqueous methanol at40C overnight. The combined filtrates were reduced to theaqueous phase in vacuo, an equal volume of 0.1 M K-Pi bufferat pH 8.2 was added, and the pH was adjusted to 8.2. Theextract was centrifuged 10 min at low speed and the super-natant was partitioned against n-hexane (lx, 1:1, v:v) anddiethyl ether (5X, 2:1, v/v). The aqueous phase was acidifiedto pH 2.5 and partitioned against EtOAc (3X, 1:1, v:v). Theorganic phase was washed with water at pH 3.0 (2X, 20:1,v:v) and evaporated to dryness. The dried extract was dis-solved in 15 mL of 0.1 M K-Pi buffer at pH 8.2, the pH wasfully adjusted to 8.0, and the solution was purified by PVPPchromatography as described (21). The extract was absorbedto a 1.0-g column of PVPP and the column was eluted with

3 Abbreviations: PP333, paclobutrazol; EtOAc, ethyl acetate; KRI,Kovats retention index; MeTMSi, methyl ester trimethyl silyl ether;PVPP, polyvinylpolypyrrolidone; SIM, selected ion monitoring.

60 mL of the same buffer. The eluate was adjusted to pH 2.5and applied to a column of charcoal:Celite (1:2, w/w) ad-sorption chromatography (24). Ten milliliters of 0.1 M K-Pibuffer at pH 8.2 were added previously to the elution flaskand the GAs were eluted with 800 mL of 80% acetone. Theacetone was removed and the aqueous phase was adjustedto pH 2.5 and partitioned against EtOAc as described above.The organic fractions were reduced to a small volume andpurified by silicic acid:Celite (1:2, w/w) adsorption chroma-tography (24). The GAs were eluted with 200 mL of chloro-form:EtOAc (2:3, v/v). Five milliliters of water (pH 8.0) wereadded to the eluate and the organic solvents were removedunder reduced pressure. The pH of the extract was adjustedto 8.0 and the extract was further purified by anion exchangechromatography with a 10-cm long column of QAE-Sepha-dex-25 (21).

After application of the sample, the column was washedwith 3 volumes of water at pH 8.0 and the GAs were elutedwith 4 volumes of 1.0 N acetic acid. The pH of the eluate wasadjusted to 7.0 and the extract was evaporated to dryness.The residue was subjected to C18 reverse phase HPLC asdescribed previously (24). The HPLC fractions were collectedat 1-min intervals from 5 to 37 min. The samples weremethylated and trimethylsilylated as before (24). The Me-TMSi derivatives of the GAs were analyzed by GC-MS usinga JEOL AX 505 double-focusing mass spectrometer coupledto a Hewlett-Packard 5890 gas chromatograph. The sampleswere coinjected with an extract of Parafilm (7) to provide therelative retention times (KRI) into a Hewlett Packard Ultra 2fused silica capillary column (22 m x 0.32 mm, 0.52 ,um filmthickness). Gas chromatography and MS operating parame-ters were as described (24). Reference spectra and KRIs fromauthentic GAs (GA4, GA53, GA44, GA19, GA20, GA,, GA8,GA9, and 3-epi-GA1) were used for comparison.

Quantification of GAs by SIM

Extraction and purification procedures were essentially thesame as described above. Freeze-dried samples, 3.0 g dryweight of ovaries at stages C, E, and F from each mandarin,were used for the quantitation analyses. The material wasground and homogenized with a mortar and pestle in 100mL of cold methanol as described above. The following GAswere added to the combined filtrates as internal standards:[17,17-2H2]GA53 (99.0% enrichment), [20-2H]GA44 (95.7% en-richment), [17,1 7-2H2]GA19 (94.2% enrichment), [17,17-2H2]GA20 (99.3% enrichment), [17,17-2H2]GA1 (99.2% enrich-ment), [17-'3C]GA8 (92.3% enrichment), [17,17-2H2]GA9(97.1% enrichment), and [17,17-2H2]GA4 (91.7 enrichment)(see 'Acknowledgments' for the source of [2H2]-GAs). Theextracts were centrifuged and partitioned against n-hexaneand diethyl ether as above.The samples were then purified in a column of char-

coal:Celite and partitioned against EtOAc as described above.The samples were purified by silicic acid adsorption chro-matography and further by QAE-Sephadex-25 ion exchangechromatography as described above. The eluate of the QAE-Sephadex column was reduced to a small volume and ad-justed to pH 2.5. The extract was then passed through a Sep-

1576 TALON ET AL.



GIBBERELLINS AND PARTHENOCARPY IN MANDARINS

Pak cartridge. The cartridge was washed with 5 mL of 5%acetic acid in water and then with 5 mL of water. The GAswere eluted with 5 mL of 80% aqueous methanol (21). Thedried eluate from the Sep-Pak cartridge was subjected toreverse-phase HPLC (24). Residual contaminating com-

pounds were removed from the methylated samples as de-scribed (21). Quantitative GC-SIM analyses for GAs were

carried out essentially as described before (24). The deriva-tized samples were analyzed by GC-SIM with magnetic fieldswitching using the same instrument and column utilized inthe qualitative analyses.The concentrations of GA53, GA44, GA19, GA20, GA,, GA8,

GA4, and GA9 were calculated from the peak area ratios 450/448, 433/432, 436/434, 420/418, 508/506, 595/594, 420/418, and 286/284, respectively (4), taking into account theabundance of natural isotopes (24). The amounts of internalstandards to be added in the quantification analyses were

estimated in preliminary experiments. The SIM data obtainedin these experiments provided estimates of the GA levels inthe fruitlets. In a first experiment of quantification, variousamounts of internal standards based on the previous esti-mates were added to the extracts to quantify the levels ofGA53, GA44, GA19, GA20, and GA9. Based on the data obtainedin these experiments, a second experiment of quantitationwas performed (5). In this experiment, levels of GA53, GA44,GA19, GA20, GA1, GA8, GA4, and GA9 were quantified.

RESULTS

Fruit Set and Abscission in Satsuma andClementine Mandarins

The pattem of fruit growth in seedless Satsuma and Cle-mentine mandarins follows a sigmoid curve, as described forcitrus fruits (1). A comparison between the fruit growth andthe abscission rates in these two species has been publishedelsewhere (23). Clementine develops smaller and lighter ova-ries and fruits than Satsuma (Table I). Abscission of devel-

100

c GA3+ PP~~~~~'3133 T/70of 60th(GA +PP33) Te rrr brsshw +SE WerGA.+ PP333

bar ar omte,tesmoTs igrta h E

opin Triswseaie ytetn 0DFwoefutwit a sigl aplcto f 3 (1 Tg,P33(5,g,o

CONTR OL T

40 11

20 *0

0 7 14 21 28 35 42 0 7 14 21 28 35 42

Days after treatment

Figure 1. Fruit abscission in Satsuma (A) and Clementine (B) man-

darins. Fruitlets at petal fall (10 d after anthesis) were treated withone single application of PP333 (253g), GA3 pg), or with a mixture

of both (GA3 + PP333). The error bars show ± SE. Where the error

bars are omitted, the symbol is bigger than the St.

oping fruits was examined by treating 10 DAF whole fruits

with a single application of GA3 (1 gg), PP333 (25 Mg), or a

mixture of both. Whole fruits were dipped in the above

solutions for 3 to 5 s. Forty-two days after the treatment,Satsuma controls had a 50% abscission rate (Fig.btA), whereas

Clementine had a 95% abscission rate (Fig. SB). For both

varieties, PP333 treatment produced a large number of yel-

lowish fruits that did not develop properly and increased the

rate of fruit drop and the total number of abscised fruits. just

1 week after treatment, PP333 had induced a 40% abscission

rate (controls had a 20% abscission rate) in both species (Fig.

1). Four weeks after treatment, PP333 application had caused

100% abscission in Clementine and 80% in Satsuma. In

contrast, GA3 treatments in Clementine increased fruit set in

the absence (13X) or presence of PP333 (9X) (Fig. iB). In

Satsuma, GA3 had no effect on the abscission of fruitlets (Fig.1A). Exogenous GA3, however, suppressed PP333-inducedabscission in both varieties.

Table I. Growth Parameters of Developing Ovaries of Satsuma and Clementine Mandarins

Diameter Length' Fresh Weight Weight Ratio

mm mg

-7 DAFSatsuma 3.5 ± 0.5 10.9 ± 0.9 23 ± 3 0.29Clementine 2.6 ± 0.3 8.2 ± 0.5 10 ± 1 0.33

0 DAFSatsuma 4.1 ± 0.2 12.8 ± 0.6 34 ± 1 0.33Clementine 3.0 ± 0.2 8.9 ± 0.5 13 ± 1 0.35



a Values (mean ± SD) include the style attached to ovaries.

1577




Identification of GAs in Satsuma andClementine Mandarins

The following 13-hydroxy-GAs were identified in 42 DAFfruits from Satsuma and Clementine mandarins: GA53, Pu-tative 16,17-dihydro-17-hydroxy-GA53 (4, 10, 24), GA44,GA17, GA19, GA20, GA29, GA1, 3-epi-GA1, and GA8 (TableII). In addition, these fruits contained nonhydroxy-GAs, suchas GA24 and GA9, and 3fl-hydroxy-GAs, such as GA4. Theidentification is based in the comparison of mass spectra andKRIs with those from authentic GAs (GA53, GA44, GA19,GA20, GA1, 3-epi-GA1, GA8, GA4, and GA9) or from pub-lished data (GA17, GA29, GA24, and 17-OH-GA53) (22). Anal-yses were repeated twice on samples harvested in two con-

secutive seasons. None of these analyses showed the presenceof GA3, a compound previously detected in ovaries at anthesisof Satsuma and Clementine mandarins (23).

Quantification of GAs in Satsuma andClementine Mandarins

Quantification of GA levels was carried out on two inde-pendent harvests corresponding to the seasons of 1990 and1991. In the first harvest, the contents of GA53, GA44, GA19,GA20, and GA9 were measured in developing ovaries andfruits of Satsuma and Clementine mandarins at three differ-ent stages of development: 7 d before anthesis, at anthesis,and 7 d after anthesis. The following year, a second experi-ment was performed to quantify the levels of GA53, GA44,GA19, GA2 GA1, GA8, GA4, and GA9. Although the twoexperiments showed similar results, Figure 2 reports datafrom the second and more complete experiment. Of the GAsquantified, GA8 was generally the most abundant (concentra-tion up to 180 ng. g dry weight-'), with progressively lowerconcentrations of GA19, GA20, GA44, GA53, GA4, GA1, andGA9 (about 1 ng-g dry weight-'). It is clear in Figure 2 thaton a ng.g dry weight-' basis, levels of the 13-hydroxy-GAs,GA53, GA44, GA19, GA20, and GA1 were higher in Satsumathan in Clementine. The differences in the levels of theseGAs were more prominent as the young fruits were devel-

oping. During pre-anthesis and in comparison with Clemen-tine, developing ovaries of Satsuma contained higher levelsof GA53 (1.5x), GA44 (9.5x), GA19 (2.7x), GA20 (2.5X), andGA1 (1.1X). In ovaries at anthesis, the differences betweenSatsuma and Clementine in the levels of these GAs increased.At petal fall, these differences were the highest, with thefollowing increases of Satsuma over Clementine: GA53(10.4x), GA44 (13.9x), GA19 (3.0x), GA20 (11.2x), and GA,(2.0x). In contrast to these GAs, the levels of GA8, the lastmember of the 13-hydroxylation pathway, were higher inClementine than in Satsuma in the three stages studied (1.7-to 2.2-fold) (Fig. 2). Of the endogenous GAs detected inClementine, GA8 was by far the most abundant. In bothvarieties, GA8 levels increased progressively as the fruits weregrowing. The concentration of GA4 was similar in ovaries ofSatsuma and Clementine mandarins. Levels of GA9 were

very low in both varieties. However, two independent quan-

tifications showed that GA9 was higher in fruits at petal fallof Satsuma than in those of Clementine (Fig. 2). It is alsointeresting to examine the pattem of GA change when thelevels of GAs are expressed in absolute values (e.g. pg.

fruit-'). On this basis, the absolute amount of GAs in Satsumaincreased progressively (data not shown) as the young fruitswere developing (GA4 excepted). However, in Clementine,this increase only occurred up to anthesis. Furthermore,ovaries at petal fall of Clementine mandarins contained lowerlevels of GA53, GA20, GA1, GA4, and GA9 than ovaries atanthesis.

In all cases analyzed, the amounts of GAs per fruit were

also higher in Satsuma than in Clementine. Table III illus-trates this observation, showing the GA levels measuredduring two consecutive seasons (1990 and 1991) in de-veloping fruits at petal fall (7 DAF). In these analyses, it isclear that fruitlets of Satsuma had higher levels of GA53,GA44, GA19, GA20, GA1, GA8, GA4, and GA9 than those ofClementine.

DISCUSSION

Citrus species usually produce an excessive number offlowers, which is common for many tree fruit varieties. After

Table II. GAs Identified by Full-Scan GC-MS of Their MeTMSi or Methyl Derivatives in 42 DAF Developing Fruits from Satsuma andClementine Mandarins

GA HPLFC KRI Ion Mass/Charge and Relative AbundanceFraction

GA53 28-29 2466 M- 448 (65), 419 (15), 416 (26), 389 (32), 251 (35), 241 (34), 207 (100)GA44 23-25 2785 M+ 432 (85), 403 911), 373 (42), 357 (8), 238 (48), 208 (50), 207 (100)GA19 23-25 2581 M+ 462 (12), 447 (5), 434 (100), 402 (56), 375 (65), 374 (82), 208 (45)GA17 23-25 2557 M+ 492 (52), 460 (21), 433 (20), 401 (28), 373 (16), 208 (100), 207 (85)GA20 21-22 2468 M+ 418 (100), 403 (26), 375 (63), 359 (25), 345 (14), 301 (24), 207 (28)GA29 11-12 2630 M+ 506 (100), 491 (15), 447 (12), 389 (14), 375 (27), 303 (24), 207 (30)GA1 15-16 2629 M+ 506 (100), 491 (10), 448 (22), 377 (18), 313 (15), 208 (9), 207 (31),GA8 08-10 2746 M+ 594 (100), 579 (8), 535 (9), 504 (3), 448 (10), 379 (11), 375 (18)17-OH-GA53 23-25 2707 M+ 538 (17), 523 (4), 448 (46), 407 (59), 375 (100), 347 (31), 181 (37)3-epi-GA1 13-14 2747 M+ 506 (100), 491 (7), 459 (7), 448 (38), 376 (19), 208 (23), 207 (30)GA24 28-29 2445 M+ 374 (8), 346 (3), 342 (30), 314 (100), 310 (31), 286 (70), 226 (95)GA9 28-29 2321 M+ 330 (12), 298 (100), 286 (33), 270 (46), 243 (56), 227 (33), 226 (17)GA4 26-27 2489 M+ 418 (18), 386 (25), 358 (18), 328 (26), 289 (79), 284 (100), 225 (93)

1578 TALON ET AL.




10

01-

-8 -4 0 4 8 -8 -4

7.5

cr)

2!)

-0(3

C.D

0 4 8

fruit, the genetic factor responsible for parthenocarpy stim-ulates high contents of hormones in the ovaries at anthesisand immediately after it (13). Thus, in pears (8) and grapes

(12), parthenocarpic fruits contain higher levels of GA-likesubstances than seeded fruits. In C sinensis, however, theGA content of seedless and seeded genotypes, as examinedby GC-MS, follows the same pattern of change and does notdiffer qualitatively, although the parthenocarpic mutant hasslightly greater amounts of GAs (21).Mandarin-type Citrus, on the other hand, are often self-

incompatible and have an unstable set-drop mechanism. Thismay result in massive fruit drop, as in Clementine. Seedlesscvs of Clementine also develop smaller fruits (Table I) andshow lower parthenocarpic fruit set (Fig. 1) than seedlesscultivars of Satsuma. In contrast to Satsuma, Clementinemandarins will respond to application of GAs increasing fruitset (20 and Fig. 1), the GA being more effective when appliedbetween anthesis and petal fall (6). For both varieties, how-ever, application of PP333, an inhibitor of GA biosynthesis(19), increases fruit abscission, and this effect is completelysuppressed by exogenous GA3 (Fig. 1).These results suggest that a reduction in the endogenous

levels of GAs results in fruit abscission, whereas an increasein GAs (applied or endogenous) reduces abscission. Thus, itappears that fruits of Clementine mandarins probably do nothave endogenous GA levels adequate for fruit set. Develop-ing fruits of both Satsuma and Clementine contain GA53,putative 17-OH-GA53, GA44, GA17, GA19, GA20, GA29, GA,,3-epi-GA1, GA8, GA24, GA9, and GA4 (Table II). The identi-fication analyses were performed on two consecutive harvestsin experimental orchards that were never commerciallytreated with GA3. In contrast to previous results (23), no

evidence for the presence of GA3 or iso-GA3 was found. Theabove results may support the suggestion that young fruitsof Satsuma (9, 23) and Clementine (23) have both the early13-hydroxylation pathway (GA53, GA44, GA17, GA19, GA20,GA29, GA1, 3-epi-GAI, and GA8), which is the major path-way, and the nonhydroxylation pathway (GA24, GA25, andGA9).

In these mandarins, the logical precursor of GA4 might be

Days after flowering

Figure 2. Concentrations of GAs in developing ovaries from Sat-suma (0) and Clementine (0) mandarins as determined by GC-MS-SIM using stable isotope-labeled GAs as quantitative internalstandards.

pollination, and especially after petal fall, excess fruits absciseas the young fruits are developing. The drop of fruitletsappears to be a result of competition for assimilates andnutrients (18). In Citrus, it has been shown that applicationof GAs to the pistil results in a greater mobilization ofassimilates to young ovaries (18). However, GA applicationdid not improve fruit set of most cultivated citrus varieties(2). These species, including Satsuma mandarins, are mainlyseedless standard cultivars exhibiting moderate to high par-thenocarpic ability and without a requirement for pollinationfor adequate fruit set. It has been suggested that in seedless

Table Ill. Levels of GAs in Developing Fruits at Petal Fall (7 DAF) ofSatsuma and Clementine Mandarins

Quantification analyses were performed by GC-SIM on twoconsecutive harvests sampled in 1990 and 1991.

Satsuma ClementineGA

1990 1991 1990 1991

pg/fruitGA53 154 115 15 4GA44 262 271 NDa 7GA19 1425 1762 328 214GA20 616 531 84 18GA9 17 33 3 2GA, NAb 65 NA 13GA8 NA 1555 NA 1242GA4 NA 31 NA 6

ND, Not detected. b NA, Not analyzed.

:-

1579




GA9. The results show also that levels of the 13-hydroxy-GAs, GA53, GA44, GA19, GA20, and GA1, are higher in Sat-suma than in Clementine (Fig. 2). Interestingly, GA8 levelsare higher in Clementine in the three stages analyzed. The2,3-hydroxylation of GA, to GA8 is an inactivation reaction.It is possible that in Clementine the ability to catabolize GA8is reduced in comparison with Satsuma. Alternatively, Cle-mentine might have higher rates of 2#-hydroxylation thanSatsuma. Because the levels of all GA8 precursors are low,their turnover may be more rapid.

In both varieties, GA4 levels do not differ greatly and GA9levels are relatively very low (Fig. 2). The relative contributionof GA20, GA4, and GA9 in the synthesis of GA1 is not known,but it is of interest that the GA20 content of Satsuma ovariesis at least 1 order of magnitude higher than those of GA4 andGA9. During petal fall, a twofold increase in the content ofGA1 was observed in Satsuma in comparison with Clemen-tine. In other physiological processes better studied, such asstem elongation, it has been shown that GA1 plays a pivotalrole (11, 15). Tall plants in Arabidopsis, for example, contain34 ng GA1 g dry weight-', and a twofold reduction in thecontent of GA, is correlated with dwarfism (22). In Satsumaand Clementine, PP333 applications result in fruit abscission,but applied GA3 can restore normal fruit set (Fig. 1). There-fore, it should be concluded that GAs control the partheno-carpic development of fruits.

Application of GA3 does not enhance natural fruit set inSatsuma, although this treatment improves considerably fruitset in Clementine (Fig. 1). This observation suggests that incontrast to Satsuma, developing fruits of Clementine do nothave adequate levels of endogenous GAs for fruit set. Wehave also shown that levels of the 13-hydroxy-GAs are lowerin Clementine (Fig. 2).

In conclusion, (a) in Satsuma, high levels of endogenousGAs are associated with elevated parthenocarpic ability, andin Clementine, low levels of GAs are associated with reducedparthenocarpic ability; (b) applied GA3 improves fruit set inClementine, but has little influence in Satsuma; (c) inhibitionof GA biosynthesis induced by paclobutrazol results in fruitabscission in both varieties; and (d) exogenous GA3 counter-acts PP333-induced fruit abscission also in both varieties.Consequently, we propose that endogenous GAs control theparthenocarpic development of fruits in Satsuma and Cle-mentine. This suggestion implies the occurrence of a thresh-old level for GAs in the processes of fruit set.

ACKNOWLEDGMENTS

We thank Dr. J.A.D. Zeevaart of the Michigan State University-Department of Energy Plant Research Laboratory, Michigan StateUniversity, East Lansing, MI for advice and the gift of labeled GAs.(A wide range of labeled GAs can be purchased from ProfessorL. N. Mander, Research School of Chemistry, Australian NationalUniversity, P.O. Box 4, Canberra, A.C.T. 2601, Australia.) Wealso thank Dr. D.A. Gage of the Michigan State University-National Institutes of Health Mass Spectrometry for help with massspectrometry.

LITERATURE CITED

1. Bain JM (1958) Morphological, anatomical and physiologicalchanges in developing fruits of Valencia orange, Citrus sinensis(L.) Osbeck. Aust J Bot 6: 1-24

2. Coggins JrCW, Hield HZ (1968) Plant growth regulators. In WReuther, LD Batchelor, HJ Webber, eds, The Citrus Industry,Vol 2. University of California Press, Berkeley, CA, pp371-389

3. Frost HB, Soost RK (1968) Seed reproduction: Development ofgametes and embryos. In W Reuther, LD Batchelor, HJ Web-ber, eds, The Citrus Industry, Vol 2. University of CaliforniaPress, Berkeley, CA, pp 290-324

4. Fujioka S, Yamane H, Spray CR, Gaskin P, MacMillan J,Phinney BO, Takahashi N (1988) Qualitative and quantita-tive analyses of gibberellins in vegetative shoots of normal,dwarf-I, dwarf-2, dwarf-3 and dwarf-5 seedlings of Zea mays L.Plant Physiol 88: 1367-1372

5. Fujioka S, Yamane H, Spray CR, Katsumi M, Phinney BO,Gaskin P, MacMillan J, Takahashi N (1988) The dominantnon-gibberellin-responding dwarf mutant (D8) of maize ac-cumulates native gibberellins. Proc Natl Acad Sci USA 85:9031-9035

6. Garcia-Martinez JL, Garcia-Papi MA (1979) The influence ofgibberellic acid, 2,4-dichlorophenoxy-acetic acid and 6- ben-zylaminopurine on fruit set of Clementine mandarin. Sci Hor-tic (Amst) 10: 285-293

7. Gaskin P, MacMillan J, Firn RD, Pryce RJ (1971) "Parafilm': aconvenient source of n-alkane standards for the determinationof gas chromatographic retention indices. Phytochemistry 10:1155-1157

8. Gil GF, Martin GC, Griggs WH (1972) Fruit set and develop-ment in the pear: extractable endogenous hormones in par-thenocarpic and seeded fruits. J Am Soc Hortic Sci 97:731-735

9. Goto A, Yamane H, Takahashi N, Hirose, K (1989) Identifica-tion of nine gibberellins from young fruit of Satsuma mandarin(Citrus unshiu Marc.). Agric Biol Chem 53: 2817-2818

10. Hedden P, Phinney BO, Heupel R, Fujii D, Cohen H, GaskinP, MacMillan J, Graebe JE (1982) Hormones in young tasselsof Zea mays. Phytochemistry 21: 391-393

11. Ingram TJ, Reid JB, Murfet IC, Gaskin P, Willis CL, Mac-Millan J (1984) Internode length in Pisum. The Le gene con-trols the 33-hydroxylation of gibberellin A20 to gibberellin A1.Planta 160: 455-463

12. Iwahori S, Weaver RJ, Pool RM (1968) Gibberellin-like activityin berries of seeded and seedless tokay grapes. Plant Physiol43: 333-337

13. Nitsch JP (1971) Perennation through seeds and other structures:Fruit development. In FC Steward, ed, Plant Physiology, aTreatise, Vol 6A. Academic Press, London, pp 413-501

14. Ortiz JM, Zaragoza S, Bono R (1988) The major citrus cultivarsin Spain. Hort Science 23: 691-693

15. Phinney BO (1984) Gibberellin Al, dwarfism and the control ofshoot elongation in higher plants. In A Crozier, JR Hillman,eds, The Biosynthesis and Metabolism of Plant Hormones,Cambridge University Press, Cambridge, UK, pp 17-41

16. Poling SM (1991) Identification of endogenous gibberellins inimmature navel orange fruit. J Agric Food Chem 39: 677-680

17. Poling SM, Maier VP (1988) Identification of endogenous gib-berellins in Navel orange shoots. Plant Physiol 88: 639-642

18. Powell AA, Krezdorn AH (1977) Influence of fruit-setting treat-ment on translocation of 14C-metabolites in citrus during flow-ering and fruiting. J Am Soc Hortic Sci 102: 709-714

19. Rademacher W (1990) New types of plant growth retardants:additional perspectives for practical applications in agricultureand horticulture. In RP Pharis, SB Rood, eds, Plant Growth

1 580 TALON ET AL.




Substances 1988. Springer-Verlag, Berlin, pp 611-61820. Soost RK, Burnett RH (1961) Effect of gibberellin on yield and

fruit characteristics of Clementine mandarin. Proc Am HorticSci 77: 194-201

21. Talon M, Hedden P, Primo-Millo E (1990) Gibberellins in Citrussinensis: a comparison between seeded and seedless varieties.J Plant Growth Regul 9: 201-206

22. Talon M, Koornneef M, Zeevaart JAD (1990) Endogenousgibberellins in Arabidopsis thaliana and the possible stepsblocked in the biosynthetic pathways of the semi-dwarf ga4

and ga5 mutants. Proc Natl Acad Sci USA 87: 7983-798723. Talon M, Zacarias L, Primo-Millo E (1990) Hormonal changes

associated with fruit set and development in mandarins dif-fering in their parthenocarpic ability. Physiol Plant 79:400-406

24. Talon M, Zeevaart JAD (1990) Gibberellins and stem growthas related to photoperiod in Silene armeria L. Plant Physiol 92:1094-1100

25. Turnbull CGN (1989) Identification and quantitative analysisof gibberellins in Citrus. J Plant Growth Regul 8: 273-282

1581



Gibberellins and ParthenocarpicAbility in Developing ... · incompatible Clementine mandarin may...

Documents

Transcript of Gibberellins and ParthenocarpicAbility in Developing ... · incompatible Clementine mandarin may...