unshiu and their change in fruit development

Transcription

1 Plant & Cell Physiol. 16: (1975) Charaterization of plant growth substanes in Citrus unshiu and their hange in fruit development Nobutaka Takahashi, somaro Yamaguhi, Toshiaki Kono, Mihiharu goshil, Kazuyoshi Hirose and Kunihiko Suzuki2 1 Department of Agriultural Chemistry, The University of Tokyo, Bunkyo-ku, Tokyo and 2 Okitsu Branh, Tree Fruits Researh Station, Okitsu, Shimizu-shi, Shizuoka-ken, Japan (Reeived August 14, 1975) The hormonal hange in fruit development of Citrus unshiu was investigated. A remarkable hange in the ontent of auxin and ABA was observed, suggesting that the hormonal balane may affet the physiologial absission of the fruits. The ourrene of AA, AM, AA-Me and ABA in the young fruits was onfirmed, while itrus auxin ould not be deteted. Citrus unshiu is a popular Citrus speies ultivated in Japan. t blooms in the middle of May and ten days after full bloom, the fruit absission starts; the pattern (7) varies every year, two peaks in average daily drop (about thirty and fifty days after full bloom) being observed in 1966 and.one peak (about thirty days after full bloom) in These physiologial absissions are onsidered to be ontrolled by the endogenous hormonal balane. We have studied the hormonal hanges in the development of the young fruits and the role of endogenous plant hormones in the physiologial absission of Citrus unshiu. A group at the Citrus Researh Center in Riverside, California laimed (1) the ourrene of a new type of non-indoli auxin, "itrus auxin", and non-ourrene of indoli auxin in Citrus plants. On the other hand, physiologial and hemial evidene to warrant a reappraisal of the above laim has aumulated. A group at the Hebrew University, srael, deteted auxin-like ativity (2, 3, 6) in various Citrus tissues and further identified ( 4) AA in elongating shootlets of Citrus limon by paper hromatography using labelled AA. n 1971, we preliminarily reported ( 8) the isolation and identifiation of AA and lam in very young fruits of Citrus unshiu. Thu5 the proposal (5) to avoid the term "itrus auxin" has been made. Herein we wish to report not only the hormonal hanges in young fruits of Citrus unshiu during development but also details of the isolation and identifiation of plant growth regulators suh as AA, lam and AA-Me. Abbreviations: AA, indole-3-aeti aid;!am, indole-3-aetamide; AN, indole-3-aetonitrile; AA-Me, methyl indole-3-aetate. 111

2 112 N. Takahashi,. Yamaguhi, T. Kono, K. goshi, K. Hirose and K. Suzuki Plant growth substanes in Citrus unsltiu 113 Materials and methods Fruits Fruits were pluked from trees of Citrus unshiu ultivated in the field of Okitsu Branh, Tree Fruits Researh Station, Okitsu, Shimizu-shi, Japan, and immediately frozen with dry ie and stoked in a deep freezer at -2oC for 2 to 3 weeks until extration. E.rtimation of auxin and ABA ontent Auxin ontent in various samples was determined by Avena urvature test and ABA ontent by Milborrow's method (11). Thin layer and j;aper hromatograph:)' Thin layers of silia gel GF 254 (Merk).2 mm thik were used for qualitative analysis and those.5-mm thik for preparative work. For paper hromatography, filter paper Toyo-roshi No. 51 was used. Development was onduted in asending mode with the following solvent systems: a) sopropyl ether-aeti aid (95 : 5) b) Ethyl aetate-hloroform-aeti aid (2 : 8 : ) ) Ethyl aeta te-n-hexane ( : ) d) sopropyl alohol-one. ammonia-water ( 1 : : ) Gas hromatography A Hitahi K-53 gas hmmatograph with a hydrogen flame ionization detetor was used. A silanized glass olumn ( m x 3 mm i.d.) was paked with 2% SE-3 and XE-6 on Chromosorb W. Retention times were determined under the following onditions: SE-3, olumn temp. 16 C, N2 flow 38 mlfmin; XE-6, olumn temp C (4 /min), N2 flow 25 mlfmin. Mass sj;etra, GC-li1S and mass hromatography These were obtained with a Hitahi RMU 6L mass spetrometer equipped with a diret inlet probe and a ombined Hitahi 63 (GC)-RMU 6M (MS)-Hita 11 data analysis system. Preparation rif samples for estimation rif plant growth regulators in the fruits Fruits harvested five to seven days after full bloom were extrated with methanol or aetone in three fold volume of the fruits using a blender. The extration was repeated three times. The solvent was removed under redued pressure to give an aqueous residue, whih was extrated with n-hexane at ph 3 and then with the same volume of ethyl aetate as the aqueous solution three times. The ombined ethyl aetate solution was extrated with 3% sodium biarbonate solution in a half volume of the organi layer three times. The ethyl aetate layer, whih was dried by passing through an anhydrous sodium sulfate olumn, was evaporated under redued pressure to give a neutral fration. The sodium biarbonate layer was aidified to ph 3 with 3 N sulfuri aid and the aqueous solution was extrated with ethyl aetate. After drying the organi layer over anhydmus sodium sulfate, evaporation of the solvent afforded an aidi fration. solation and haraterization rif auxins and two peptides ) Extration and solvent frationation Fruits (1 kg) harvested seven days after full bloom in 1969 were extrated with ether (S liters) five times using a blender. The ake obtained by filtration of the homogenized mixtures was extrated with methanol (25 liters) two times. The ombined ether fration was onentrated to 3 liters under redued pressure and extrated with 3% sodium biarbonate solution (7 ml) three times. The ether layer dried over anhydrous sodium sulfate was evaporated to give Fration B (22.6 g). The sodium biarbonate solution was aidified to ph 3 with 3 N sulfuri aid with ether (7 ml) three times to afford Fration A (2.9 g). The methanol extrat was onentrated to give an aqueous residue whih wa separated into Fration C (an aidi fration, 3.5 g) and Fration D (a neutral fration, 56. g) with ethyl aetate in the same manner as in the ether extration. 2) Purifiation of Fration A Fration A (2.9 g) was subjeted to nine transfers ounterurrent distribution between 1M phosphate buffer of ph 6.5 (1 ml) and ethyl aetate ( 1 ml). Lesser fration numbers indiate the presene of polar substanes. Fration 4 to 7 showing strong auxin ativity were ombined to reover a resinous solid (1.9 g), whih was purified through.silii aid partition hromatography. Silii aid (Mallinrodt, 1 g) was throughly mixed with 1M phosphate buffer of ph 7.2 (7 ml) and paked into a olumn with n-hexane. The ative fration was dissolved in aetone (5 ml) and silii aid (15. g) was added to the solution, then the mixture was evaporated under redued pressure. The silii aid powder thus obtained was plaed on top of the olumn and elution was started with n hexane followed by n-hexane ontaining ethyl aetate inreasing in ontent by.5% steps in a!-liter fration. A rude ative material (14 mg) obtained fmm 2% ethyl aetate eluate was further purified by another partition olumn. The olumn was prepared from silii aid (15 g) pretreated with 1M phosphate buffer of ph 6.2 (1 ml) and benzene. The sample was applied to the olumn as in the ase of the first olumn. Elution was onduted with benzene ontaining butanol inreasing in ontent by. % steps in a 3-rnl fration. Eluate ontaining.2 to.4% butanol showed strong auxin ativity. Evaporation of the solvent afforded a solid. This was subjeted to preparative paper hromatography using solvent system d. The zone of Rf.25 to.35 was ut off and eluted with methanol. The residue of this eluate was dissolved in an aqueous sodium biarbonate solution and aidified with 3 N sulfuri aid to ph 3 then extrated with ethyl aetate. Evaporation of the solvent afforded a partially rystalline solid (3.5 mg), whih was identified a AA. 3) Purifiation of Fration C Fration C (3.5 g) was subjeted to ilii aid part1t1on hromatography using n-hexane-ethyl aetate solvent system as in the ase of Fration A. Evaporation of the eluate (n-hexane-ethyl aetate, 8 : 2-75 : 25 and 7 : 3) gave two frations showing positive Ehrlih test. A solid (9 mg) reovered from the fration eluted with n-hexane-ethyl aetate (7 : 3) was purified through

3 14 N. Takahashi,. Yamaguhi, T. Kiino, K. goshi, K. Hirose and K. Suzuki Plant growth substanes in Citms unshiu 15 a neutral alumina olumn, whih was prepared by paking aluminum oxide (l'derk, ative neutral ativity, 3 g) and a mixture of benzene and ethyl ateate (1 : 1). The olumn was eluted with the same solvent system (1 liter), then ethyl aetate ontaining methanol inreasing in ontent by 2.5% steps in a!-liter fration. Elution with ethyl aetate ontaining 2.5 to 5% methanol afforded a solid (4 mg), whih gave a positive Ehrlih test. This was subjeted to preparative paper hromatography using solvent system a. The zone of Rf.1-.2 was ut off and eluted with methanol. Evaporation of the solvent gave a rystalline solid (18 mg), whih was rerystallized into rods from ethyl aetate and n-hexane, and identified as JAM. The fration eluted with n-hexane-ethyl aetate (8 : 2-75 : 25) showed a spot positive to Ehrlih test with the same Rf value as AA. This fration was purified by preparative thin layer hromatography using solvent system a and the zone of Rf was eluted with methanol. After methylation, the presene of AA-Me in this fration was onfirmed by GC-MS. 4) Purifiation of Fration D Fration D (56.) was dissolved in ethyl aetate (5 ml) and poured into a olumn prepared from neutral alumina (Merk, ativity ) and ethyl aetate. Elution was onduted with ethyl aetate (41iters) and ethyl aetate ontaining methanol inreasing in ontent by 2.5% steps in a 2-liter fration. The eluates with ethyl aetate ontaining 2.5 to 1% methanol were ombined and evaporated to give a resinous solid (6.2 g). This was further subjeted to silii aid adsorption hromatography, in whih a olumn was prepared with silii aid (4 g) and ethyl aetate. The olumn was eluted with ethyl aetate ontaining methanol inreasing in ontent by 2.5% steps in a!-liter fration. The elutae with ethyl aetate ontaining 2.5% methanol showed a spot at the same Rf value as that of JAM. By repeated thin layer hromatography of this fration using solvent system and rerystallization from ethyl aetate and n-hexane, JAM (35 mg) was isolated. The eluate with ethyl aetate ontaining 5-7.5% methanol showed a spot positive to Ehrlih test and a faint lam spot on thin layer hromatography using solvent systems a and e, while another spot was revealed by spraying with 7% sulfuri aid and heating. Repeated thin layer hromatography and rerystallization from hot methanol afforded two rystals with m.p C and C, the yield of the former being 9 mg and the latter 52 mg. nfrared spetra of these rystals and preliminary hydrolysis experiments indiated that they were peptides, espeially the 1 ystal showing m.p C ontaining tryptophan. dentifiation rif neutral auxin Young fruits (1 kg) harvested seven days after full bloom in 1974 were extrated with aetone (2 liters) three times using a blender. Evaporation of solvent afforded an aqueous residue (5 liters), whih was extrated with ether (4liters) three times and then ethyl aetate (4liters) three times at ph 5.5. The ether-soluble and ethyl aetate-soluble frations were separately partitioned into a neutral and an aidi fration. The ether-soluble neutral fi ation (25. 7 g) was subjeted to silii aid adsorption hromatography. A olumn was prepared with silii aid (Mallinrodt CC-7 Speial, 2 g) and n-hexane. The oily sample was applied onto the top of the olumn and the olumn was eluted with n-hexane ontaining ethyl aetate inreasing in ontent by 2% steps in a 2-liter fration. A solid (656 mg) was reovered from the eluate with n-hexane ontaining 2% ethyl aetate, whih showed lear auxin ativity. This was subjeted to ten transfers ounterurrent between n-hexane ontaining 1% ethyl aetate (upper phase) and water ontaining 2% methanol (lower phase). Frations 7 to 9 (26.2 mg in total) showing auxin ativity as well as positive Ehrlih test was separately purified by preparative thin layer hromatography using solvent system d, monitored by the Ehrlih test. Zones giving a positive Ehrlih test were ombined and this proedure was repeated two times. Finally, the most purified fration showing a spot at Rf. 77 on thin layer hromatography using solvent sytem d was subjeted to mass hromatography. As shown in the text, the presene of AA-Me in this fration was onfirmed. Hormonal hange in fruit development Results and disussion The hormonal hange in the fruit development was surveyed in 196R to Herein the results for 1971 to 1973 are presented. The young fruits were harvested at different developmental stages and extrated as desribed in Materials and methods. The ativity of auxin in the methanol or aetone extrat and the aidi and neutral ethyl aetate frations from fruits in 1971 to 1973 was determined by Avena urvature test and that of ABA in the aidi fration of 1972 and 1973 by Milborrow's method (11). Sine in the preliminary experiments, the ontent of gibberellin was estimated to be below.5 ftg/kg by the rie seedling test, gibberellin was not tested. Table 1 shows the total weight and number of fruits harvested at eah stage in Similar results were obtained for the samples obtained in 1971 and Table l Properties qf young fruits harvested for bioassy of auxin and ABA in Stage of sampling Total weight of Number of fruits We1ght per!nut (days after full bloom) fruits (kg).. (g) (before full bloom)

4 116 N. Takahashi,. Yamaguhi, T. Kono, K. goshi, K. Hirose and K. Suzuki Plant growth substanes in Citrus unshiu :::; ::::1 g u 5 X / (May 24) o! - - Total (Crude) Auxin 33 ug/1 fruits Aidi Auxin _Neutral Auxin -'"\... /"123 ug/1 fruits / "\. o., o \ \ "-.._""... / _..:;:.:... \.. / _.-'/) \'. \(il.:_:..:.:--- : Fig.. Change rif auxin and ABA ontent (197 ). Fig. 1 shows the hange of auxin ativity during development in three frations from fruits harvested in The total auxin ativity reahed maximum (15 p,gfkg fresh fruitl) five days after full bloom and suddenly started dereasing to reah the minimum after twenty days and again started inreasing to another maximum peak after thirty days. The neutral auxin ativity showed its maximum peaks five (71! gfkg) and thirty-five days (26 p,gfkg) after full bloom, while the hange of the aidi auxin ativity was not very remarkable in omparison with those in 1972 and 1973, the maximum peaks being observed ten (6 p,gfkg) and thirty-five days ( 42 p,gfkg) after full bloom. As shown in Fig. 2, the highest total auxin ativity (15 p,gfkg) was observed ten days after full bloom in 1972, while the aidi auxin ontent hanged similarly to that of the total auxin and the neutral auxin showed a hanging pattern similar to that of The ontent of ABA reahed maximum ( 44 p,gfkg) seven days after full bloom and sharply dereased to minimum ontent (3 p,gfkg) eighteen days after fi11l bloom and then started inreasing. to reah to the other maximum ontent (12 p,gfkg). After the seond maximum, the ontent of ABA inreased in monotone. Fig. 3 shows the hange of auxin ontent in three frations and that of ABA in n this year, the total auxin ontent was remarkably high (8 p,gfkg), showing the maximum peak six days after full bloom. The hange of aidi auxin ontent was similar to that of the total auxin ontent, while that of the neutral auxin showed two peaks (7 pgfkg and 2 p,gfkg) six and thirty-two days after 15 " -u <C 1 > -5 /,3D ug/1 fruits - - Total (Crude) Auxin o, Aidi Aux1n,., t: ll- - Neutral Auxin,Jj V.-::---_ - - ABA /x \, o+..,.:.,. 7B ug/1 frults.r -- i./ \X ",. : :.. \ " """... ""'<( < _ ;--=--- -<:.::r-2:--. >< o 1 2 3o 4o 5o 6o ;j! (May 15 ) Fig. 2. Change rif auxin and ABA ontent (1972). 1 Unless otherwise stated, the ontent indiates weight equivalent to AA (or ABA) per kg of fresh fruits. Fig. 3. Change rif auxin and ABA ontent (1973). i 6 u BOO ] 4 > a \"---lbd ug/1 fruits r\\ : \... j 2.' \"\ Total (Crude) Auxin... Aidi Auxin - - e--- Neutral Auxin -x- ABA., \, i!/ "-, ',. 71 ug/1 fruits ), ){.,!/; / ' =-,""" /.. (May 11) /,/ Q-.Q- --.::----- )( ::./ u D...: full bloom, respetively. The ontent of ABA reahed maximum seven and thirtyfive days after full bloom and ontinued to gradually inrease after this period, this pattern being similar to that of The results in the above experiments summarized are as follows. The ontent of aidi auxin varies over the years, the highest ontent varying from 8 ftgjkg in 1973 to 6 p,gfkg in The pattern of hange in the aidi auxin ontent in fruit development is quite harateristi; the maximum peak is always observed in the very early developmental stage of fruits (five to ten days after full bloom) and after reahing the maximum, a sudden derease takes plae. On the other hand, the hange of the neutral auxin ontent is similar every year, two maximum peaks being observed about seven and thirty-five days after full bloom. Only trae amounts of endogenous auxins ould be deteted in two-month-old fruits. n a very early stage, a peak of ABA was also observed. nterestingly, a similar hange of auxin ontent has been reported (9) in the fruit development of tomato. n this ase, two maximum peaks of auxin ontent, seven and thirty days after anthesis, were observed. To orrelate hormonal balane to fruit absission, the physiologial absission was surveyed using trees in the similar growing stage and under irumstanes similar to those of trees from whih the young fruits were harvested for analyses of endogenous hormones. The results of the survey in 1973 are summarized in Fig. 4. The inrease of the absission of fruits started ten days after the maximum peak of auxin ontent, and the maximum peak of absission was attained ten to fifteen days after the minimum of auxin ontent. The seond peak of the absission, whih was not as lear as the first, was observed fifteen lays after the seond maximum peak. of auxin ontent. 6

5 118 N. Takahashi,. Yamaguhi, T. Kono, K. goshi, K. Hirose and K. Suzuki Plant growth substanes in Citrus wzshiu /''-...- /" <., _...- '\. '\. -- \ -...)(. )( Fig. 4. various Absission if fruits in developmental stages in ;:,. Total on Monitor... Retenti Time (t ).. * Over Flow AN (S.No.56) (May 11) Endogenous auxin has long been known to play a dominant role in leaf absission. That is, a high level of auxin in leaf blades retards absission, while its derease indues absission. On the other hand, ABA was isolated as an absissionaelerating substane from young otton fruits in vigorous absission. Further Davis and Aldiott (J) have reported the orrelation between the hange of ABA ontent and absission of developing otton fruits. The hanging pattern of ABA ontent is quite similar to that in young fruits of Citrus wzshiu, while the absission pattern is rather different in both ases. On the basis of this knowledge, we suggest that a sudden derease of the auxin ontent from its maximum peak as well as the appearane of the ABA peak play important roles in the absission of fruits of Citrus unslziu. solation of endogenous plant growth regulators in young fruits if Citrus unshiu J lentifiation of the hemial speies of auxin, the ontent of whih shows a dramati hange during fruit development, is important. Young fruits (1 kg) were harvested seven days after full bloom in They were extrated and purified as desribed in Materials and methods. From Fration A, 3.5 mg of an amorphous powder with strong auxin ativity on Avena urvature test was obtained. t showed a positive Ehrlih test, UV absorption (Amax 22, 272 and 288 nm) harateristi of an indole ring, and its Rf value was idential with that of AA in solvent systems a nad b. Treatment with etherial diazomethane of this solid afforded a methyl ester (oij) with a mass spetrum idential with that of AA-Me [mfe 189 (M+), 13, 13, 77] and the same retention time as that of AA-Me (tr 5. min on SE-3, 11.1 min on XE-6). Thus, it was identified as AA. ;:,. 1 A L L t Sannf ng Number (S. No.) ' m/e 13 m/e 189 m/e 77 m/e 156 m/e 155 Fig. 5. Upper: Total ion monitor hromatogram. A glass olumn (3 mm X m) or 2% XE-6 on Chromosorb \ was used. The sample was injeted at 16 C oven temperature and 2 min later the oven temperature was raised at the rate of 5 C a min. Sanning started at tr = 5.5 min with 5 se intervals. The retention time (trl or AA-Me was 7.2 min and orrelated with sanning number (S. No.) 19. TR of lam was 1.3 min and orrelated with S. No. 56. Lower: Mass hromatogram monitored at mfe 189, 156, 155, 13 and 77. As Fration B lost auxin ativity during storage, identifiation of the auxm m this fration was onduted using the sample prepared in The result will be desribed later. The presene of AA in Frtion C was onfirmed by GC-MS. The peak at tr 5. min on gas hromatography (SE-3 olumn) of the methylated sample from the most purified fration had a mass spetrum idential with that of AA-Me. The ontent of AA in Fration C was estimated to be a. 3 mg by Avena urvature test and gas hromatography. n the final paper hromatography, another rystalline ompound (13 mg), m.p. l83 C, showing a positive Ehrlih test, was isolated. t showed a mass spetrum [mfe 174 (M+), 13, 131 whih was idential '

6 111 N. Takahashi,. Yamaguhi, T. Kono, K. goshi, K. Hirose and K. Suzuki with that of authenti lam. Although it seems unusual that lam, whih is a neutral ompound, was isolated from the aidi fration, this may be explained by its very polar nature whih auses inomplete distribution in solvent frationation. This assumption is supported by the observation that the Rf value of lam on thin layer hromatography is lower than that of AA, and lam was eluted later than AA from the partition olumn. Fration D afforded 4 mg of lam and two kinds of rystal after several purifiation steps. One of the new rystals showed a positive Ehrlih test and the other a negative test. They were onsidered to be peptides on the basis of R spetrosopy and preliminary hydrolysis experiments. Strutural eluidation of these pptides is now under way. n the experiment using the sample obtained in 1969, identifiation of the neutral auxin was inomplete, beause Fration B lost its auxin ativity during storage. Furthermore, the amount of lam isolated annot explain the total neutral auxin ativity in young fruits beause of the very low ativity of lam. To establish the hemial speies in the neutral fration, young fruits were again harvested seven days after full bloom in They were extrated with aetone and the extrat was frationated as desribed in Materials and methods. The most purified fration thus obtained showed a spot positive to Ehrlih test on thin layer hromatography, the Rf value of whih was idential with that of AA-Me. For the final onfirmation, this fration was subjeted to mass hromatography. The total ion monitoring hromatogram and those monitored at m/e 189, 13 and 77, at whih prominent peaks were observed in the mass spetrum of AA-Me, are shown in Fig. 5. The retention times of these peaks were ompletely idential with those of AA-Me. The possibility of o-ourrene of AN was exluded by mass hromatograms monitored at m/e 13 (harateristi of the fragment indole-3-ch 2 +), 156 and 155 (the moleular ion and base peak, respetively). As shown in Fig. 5, only one peak of m/e 13 was observed at the retention time of AA-Me (7.2 min) but no peak for AN (1.3 min) was deteted in all hromatograms. lam and two peptides isolated from Fration D were identified in the neutral fration extrated with ethyl aetate. The aidi auxin ontent in fruits harvested in 1974 was estimated to be a. 3 1-lgfkg by Avena urvature test. The experiment desribed above onfirmed that young fruits harvested five lo seven days after full bloom ontain AA as an aidi auxin and AA-Me and lam as neutral auxins. The isolation and identifiation of AA, AA-Me and lam in onsiderable high ontents in young fruits are inonsistent with the laim by the Riverside group of the non-ourrene of indoli auxin. Further, we pointed out that a non-indoli auxin suh as "itrus auxin" was not deteted in our extensive survey of fruits of Citrus unshiu in various developmental stages. This observation together with that of the srael group asts serious doubt on the ourrene of "itrus auxin". We are grateful to The Naka vvorks, Hitahi Ltd., for determination of mass hromatograms. - J ij, :;/ ', l - ' : -,/ "... : :tl': -.,; Plant growth substanes in Citrus unshiu Referenes ( 1 ) Davis, L.A. and F. T. Addiott: Absisi aid: Correlations with absission a nd the development in the otton fruits. Plant Physiol. 49: ( 1972). ( 2 ) Goldshmidt, E. E. and S. P. Monselise : Extration of auxin and other growth regulators from itrus stems by entrifugal fore. Physiol. Plant. 21 : (1968). ( 3) Goldshmidt, E. E.: The auxin indued urvature of itrus petals. Plant Physio/. 43: (1968). ( 4 ) Goldshmidt, E. E. and R. Goren: On the identifiation of na tive auxin in itru' tissue. Can. J. Bot. 49: (1971 ). ( 5 ) Goldshmidt, E. E., R. Goren, S. P. Monselise, N. Takahashi, H. Tgoshi,. Ya maguhi a nd K. Hirose: Auxin in itrus. A reappraisal. Siene 174: (1971). ( 6) Goren, R. and E. E. Goldshmidt: Regulative systems in the development itrus fruit L The hormonal balane in orange fruits tissue. Physiol. Plant. 23: ( 1971). ( 7) Hirose, K., M. Yamamoto and T. Ohata: Studies on thinning of itrus fruits by appliation of hemials. Effets of 1-naphthaleneaetie aid (NAA) on fruit thinning of 'Satsuma' mandarin (Citrus unshiu Marovith). Bull. Hart. Re.<. Stat. Japa11. Ser. B: (1972). ( 8) goshi, M.,. Yamaguhi, N. Takahashi and K. Hirose: Plant growth substanes in the young fruits of Citrus unshiu. Agr. Bioi. Chem. 35: (1971). ( 9) wahori, S.: Auxin of tomato fruit at different stages on its development with a speial referene to high temperature injuries. Plant & Cell Physiol. 8: (1967). (1) Khalifah, R. A., L. N. Lewis and C. W. Conggins, Jr.: New natural growth promoting substanes in young itrus fruits. Siene 142: (1963). () Mil borrow, B. V. : dentifiatin and measurement of ( + )-absisi aid in plants. n Biohemistry and Physiology qf Plant Growth Substanes. Edited by F. Wightman and G. Setterfield. p The Runge Press, Ottawa, ll!