p. 585 to 589. Pergamon Press. Printed in (

NESIS IN OVARIAN TI~ A TELEOST, THE GUPPY ;4 RETICULATA

G. D. LAMBERT AND M. G. E. POT 'or Comparative Endocrinology, State Un echt, Utrecht, The Netherlands

(Received 8 April 1974)

Almtract--1. Homogenates of ovaries of 3- and 12-month-old Iguppies we 6th pregueno- lone-7a-SH and progesterone-4-a4C, and with androstenedione.mdrostenedione- 1,2-all, re~, 2. From the double-labeled experiment, 17~-hydroxypre17 ~-hydroxypregnenolone, ~progesterone, dehydroepiandrosterone, androstenedione and testosterone vwere isolate ied by paper- and thin-layer chromatography, derivative formation and b2by recrystall astant specific activity. 3. In the ovary of the guppy the delta-5 pathway in ste~steroidogenesi .~s, as may be concluded fr6m the 8H/a4C ratios of the identified metabolitesmetabolites. 4. From the androstenedione-l,2-3H incubation, ovaries of 12-month- ppeared to be able to synthesize steroids with functional groups in the 11-1position (1 •one and 11 fi- hydroxytestosterone). 5. The androgen synthesis in ovaries of older guppies may Ibe consider le explanation of the sex conversion occasionally observed.

INTRODUCTION Anoth~~ther problem in physiology of tlthe fish ovary tpacity for synthesizing ll-keto-11-ket and llfl- IN PREVIOUS~VlOUS studies (lambert et al., 1971), it was is its caI hydroxys~steroids. 11-Ketotestosterone has a strong shownL that the ovary of the viviparous teleost androgen,genie capacity (Idler et aL, 1961), and therefore Poeeiliaia reticulata contains aromatizing enzymes, might be considered as a male sex honhormone in fish. are able to catalyse the conversions of which , it has also been isolated ffrom ovarian stenedione* and testosterone to oestradiol. However androstenedio tissue (E3(Eylath & Eckstein, 1969); incubincubation experi- ver, direct evidence for complete steroid However ments pr,~roved ovaries to be capable of synthesizing ,'sis in the ovary of this teleost was not obtained. synthesis in the ovar 11-ketotestosterone (Eckstein, 1970; Eckstein & An incubationacubation experiment with the radioactive ~:,,lo*h 1 nd a4C-progesterone Eylath, 1970; Eckstein & Katz, 1971;197] Reinboth, precursors SH-preguenolone and hydroxytestosterone was considered necessary. Suchch a double-labeled 1972), and in some cases llfl-hydrox 1972). The experiment could also be used~ed to discriminate (Eckstein and Katz, 1971; Reinboth, could not be between a delta-4 and a delta-55 pathway in steroid synthesis of the mentioned steroids c demonstrated in ovaries of young guppguppies (Lambert synthesis. et aL, 1971). It is possible that in oldelolder animals the * The following trivial names aree used throughout this 11-hydroxylation is functioning. For thist reason it paper: progesterone, pregu-4-ene-3,20-diov.3,20-dione;~ione preguenc~ has been attempted to demonstrate thetl formation lone, 3fl-hydroxypregn-5-ene-20-oneone; 17a-hydroxypro- of such steroids in ovaries of virgin g~guppies which gesterone, 17 o~-hydroxypregn-4-ene-3,20-~me-3,20-dione;20-dione; 17~x- are at least 12 months old. hydroxypregnenolone, 3fl,17~x-dihydroxyprydroxypregu-5-ene-20-~ypreg one; dehydroepiandrosterone, 3fl-hydroxyhydroxyandrost-5-ene- 17-one; androstenedione, androst-4-e~trost-4-ene-3,17-dione;4-ene-3 MATERIALS AND METHOE[ETHODS testosterone, 17fl-hydroxyandrost-4-ene-3~t-4-ene-3-one;ae-3-one: adre,- Animals nosterone, androst-4-ene-3,11,17-trione; 11 fl-hydro- xyandrostenedione, 11 fl-hydroxyoxyandrost-4-ene-3,17- Virgin female guppies (Poecilia reticular,reticulata) were reared dione; ll-ketotestosterone, 17fl-hydroxyhydroxyandrost-4-ene- under laboratory conditions at 24°C and artificial light 3,11-dione; 11 fl-hydroxytestosteroneme, llfl,17fl-dihydro- during 12 consecutive hours per day. xyandrost-4-ene-3-one; eortisol, 11 fl,17~,21-trihydro- xypregu-4-ene-3,20-dione; cortisone, 17~,21-dihydro- Radioactive steroids xypregu-4-ene-3,11,20-trione; oestradiol,:radiol, 3,17fl-dihydro- xyoestra-l,3,5(10)-triene; oestrone, 3-hydroxyoestra- The following radioactive steroids, all pmrchased from 1,3,5(10)-triene-17-one. New England Nuclear Corp., Boston, Mas., U.S.A., 585 J. G. D. LAMBERTAND M. G. E. POT

alayer chromato- Saponification of Lates was carried out : 17 Ci/mMol), with sodium hydroxi to Bush (1961). Ci/mMol), pro- ~1ol), oestradiol- Measurement of radZ strone-4-14C (sp. = (sp. act.: 55 Samples were essa'. aclear Chicago Mark I 55"1 mCi/mMol), scintillation counter tillator of PPO (4 g) and adreno- and POPOP (40 mg) 1.). Radioactive areas rtisol-4-14C and on TLC plates and I e located by means of on with sodium a Berthold thin layer tiogram scanner. 4-14C and ll- ared from 11 fl- Incubation procedure en°ster°ne-4-~4C The ovaries were decapitated guppies, ively by reduction with sodium borohydride, pooled, weighedw and t at 0°C in a medium (1 ml/1001/100 mg tissue) per litre aqua dest.: ts NaC1 (11(11.69 g), K( aCl~ (0.44g), MgCl~ (0.41 g), NaHCOa q 2Os (0.14g), glucose :hemicals used were of analytical grade; organic (0"10 g), nicotinamidn fumaric acid (0.01 g). s were redistilled once just before use. Cofactors AppropriateAppropri~ quantit i precursors and co- a the incubation mixture were obtained from factors (/ATP, NAE ADH and NADPH; ager, Mannheim, Germany. 2/zMol of each) wer~ ~-5 ml incubation vials, to which tthe homoge ted. Incubations were ~tography carried otout at 25°C losphere or in a 95% O2 and 55% CO~ ga continuous shaking. •layer chromatography (TLC) was carried out on After 1 hihr, addition; 12 ~Mol) of cofactors :ed plates (10x20cm) with silica gel, F254 were adde~added and after r the enzyme reactions , A.G.) or cellulose MN 300, F 254 (Antec, A.G.) were terrrterminated by of dichloromethane :ated tanks using the following systems: system 1 : (10 ml). omethane-methanol (97 : 3); system 2: benzene- ,' (8:2); system 3: benzene-methanol (9: 1); Extraction and fractionation system 4: benzene--cyclohexane (1 : 1); system 5: benzene--ethylacetate (7 : 3); system 6: hexane- Labeled and unlabeled carriers were added to the ethylacetateetate (1 • 1); system 7: cyclohexane-ethylacetate incubation mixture before extraction with dichloro- (1 : 3); system 8: ethylacetate-cyclohexane-ethanol (45 : methane (3 × 10 ml). The combined dicdichloromethane 45 : 10);~; system 9: toluene-acetone (7:3); system 10: extracts were evaporated in vacuo and the residue petroleumam ether (40--60°C)-tert. butanol (1 : 1); system dissolved ini aqueous methanol (70%; 8 mll and stored at 11 : petroleumtroleum ether (40--60°C)-tert. butanol (3 : 1); - 18°C overnightov to precipitate triglyceridq'cerides. Following system 12: toluene saturated with propylene glycol, centrifugal~tion the supernatant (the steroid fraction) was Only cellulose plates were developed in system 12. evaporatec~orated to dryness, and then fractionatedfractio into a The spot3tted cellulose plates were dipped up to the origin phenoliclic aand a neutral fraction (see LambertLambel et al., 1971). in a mixtureixture of methanol-propylene glycol (5 : 1), and allowed to dry for 10 rain before development. After RESULTS chromatography, the 3-keto-A4-steroidsroids were located by u.v. absorption, whereas the other steroids or derivates I. Incubation with pregnenolone-7~-at~e-7~-SH and Pro- were detected by spraying with primuline-imuline according to gesterone-4-14C as precursors Wright (1971). (190 mg) of eight, Paper chromatography (PC) wasvas carried out on The homogenate of the ovaries (190 J Machery and Nagel M.N. 263ff. paper (Diiren, 4-month-old virgin guppies was incubaincubated with 3H- Germany) in the system petroleumam ether (40-60°C) - pregnenolone (4-99/zCi; 10.1/zg) and 1414-C-progeste- methanol-water (5 : 4 : I). Steroidsas were detected by rone (1.15/zCi; 6"3/zg) in the presence of cofactors absorption of u.v. light. in an atmosphere of 95700 O2 and 5% C(CO~. After the Gas-liquid chromatography (GLC)~) was applied in the incubation the following steroids (100 /zg of each) quantification of the steroids durin g the recrystallization were added as carriers: pregnenolone, progesterone,r procedure, by using a Hewlett-Packardlckard 402 gas chro- 17c~-hydroxypregnenolone, 17o~-hydrydroxyprogeste- matograph with a flame ionizationa detector and a 4 ft rone, dehydroepiandrosterone, androstenedione, 3% SE 30 column. testosterone, oestradiol and oestrone.estrone. Following the extraction and removal of triglglycerides the Microchemical reactions steroid extract was fractionated into a phenolic Oxidation was brought about withth chromium trioxide fraction and a neutral fraction. according to Eckstein (1970). Methylation was carried out withth dimethyl sulphate Phenolic fraction as described by Brown et al. (1957). Formylation was performed by¢ dissolving the dry Following TLC in system 3 the oeoestradiol and steroid in formic acid 98% (0'5 ml).p. After 2 hr at room oestrone fractions were methylated. After TLC in temperature the formylation is completed. system 8 it appeared that the SH as wellwe as the 14C genesis in ovarian tissue of Poecilia reticn

:ed carriers of identical to andros1 d dehydroepiandro- low to permit sterone formate res oreover, a constant specific activity wa~ er repeated crystal- lizations (Table 1). Androstenedione. em 1 of the andro- stenedione fraction R least three radio- 1 fraction was active areas, one of, ~onds with authentic only the areas androstenedione. ' enedione area was ~rs were eluted eluted and the acti' 1 appeared to resist 7a-hydroxypro- oxidation and forn rally, the substance aydroepiandro- showed a constant ivity after repeated :; (3) androstenedione; (4) pregnenolone; crystallizadlizations (T~ le purified andro- ogesterone. TLC in system 2 of the first stenedion,stenedione was ma labeled; the 8H/14C n affected a separation in a combined 17~- ratio is 33. ~ypregnenolone and testosterone fraction and Testost~tosterone. Fo: ! the fraction con- 7~-hydroxyprogesterone fraction• After TLC taining ththe carrier l delded a compound em 1, it was possible to separate 17cx-hydro- with the same mo thentic testosterone menolone from testosterone. The individual formate after TLC This formate was ! fractions were further purified by TLC, saponifiec~onified, and aft~ estem 10 the radio- tive formation and crystallization, active pn~roduct sho tr chromatographic .Hydroxyprogesterone. The fraction contain- behaviombehaviour as autl ;terone. After re- e carrier 17a-hydroxyprogesterone was sub- crystallizedlization to co ic activity (Table 1), to the procedure of formylation. The 8H,x'C- a 8HSH acti~activity was th be detected. ~und remained unaffected and showed an .nedione-l,2-SH as ntical to authentic 17a-hydroxyprogesterone II. Incubation wi~ TLC in system 1. Finally, the 8H, x4C- precursor rand showed a constant specific activity after Homoggenized ovarian tissue (400 mgrag) of six, 12- ed crystallizations (Table 1). month-ohmonth-old virgin guppies was incubat1bated with SH- 17 a-Hydroxypregnenolone. The fraction was androsten~stenedione (6.7 /zCi) in the presence of formylatedated and the radioactive material showed cofactors in an air atmosphere. Aftel!ter incubation the sameme mobililmobility on TLC in system 1 as authentic the folio,following labeled carriers were added: 11fl- 17~-hydroxypregnenolone formate. After sapon- hydroxytestosterone-4-1~C (8.93 nCi, 50/~g); ll- ification and a renewed TLC in the same ketotesto.,ketotestosterone-4-x'C (7.63 nCi, 50/~g)~g); 11 fl-hydro- system,t, the tritium active area appeared to corres- xyandrostenedione-4-x4C (7.89 nCi; 50 t~g); pond with~ith carrier 17a-hydroxypregnenolone. adrenosteadrenosterone-4-~4C (6.21 nCi, 50 pg); oestrone~- Dehydroepiandrosterone. The dehydroepiandro- x~C (6.73 nCi, 50/~g); oestradiol-4-~4(iol-4-14C (8-04 nCi, sterone fraction obtained after PC was separated 50/xg). into three radioactive areas after• a TLC in system 1. Following extraction and removal~val of 1triglycerides, One of these active areas corressponds with carrier the steroid extract was subjected to a phenolic dehydroepiandrosterone. Followin~ing a second TLC separation procedure. in system 7, the active dehydroeMroepiandrosterone fraction was eluted and divided into three portions Conversion to 3H-oestrogens for oxidation, formylation andmd crystallization. The phenolic fractions could be sel;,arated with After oxidation and formylation,1, and after TLC in TLC in system 3 into two major radicdioactive areas system 1 the tritium active products:lucts showed an R I corresponding with oestradiol and oestrstrone.

Table 1. The last three er~~,stallizations of steroids isolated from an incubation of ovarian tiss~issue of P. reticulata'data withwitl~ pregnenolone-7~-3H and progesterone-4-1~ Specific activity (dis/min per mg)

Steroid Crystallization 1 Crystallization 2 Crystallizationdlization 3 17a-Hydroxyprogesterone aH 4038 3304 3377 1~C 2603 2214 2282 Dehydroepiandrosterone SH 2347 1830 1804 14 C __ ~ -- Androstenedione SH 1478 1386 1391 x4C 53 47 42 Testosterone SH 712 643 625 14 c ~ ~ -- I. G. D. LAMBERT AND M. G. E. POT

:oids and their derivatives from an ineub an tissue of reticulata with androstenedione-l,2-ZH 3H/14C ratios

Free Formylated Oxidiz :thylated

7.98 7'83 7.61 0.66 0"57 0.55 2.13 1'96 2'04

:raalol. tart oi me ~ta---~ material,erial behavingoenavlng pounupound w:with the sal as authentic adreno- atographically like oestradiol, was methylated, sterone con TLC ii Formylation of the ~C in system 6 the SH-~4C product had the above-m~entioned hal and subsequent nobility as authentic oestradiol monomethyl TLC in ssystem 1 s 14C product with an Another portion of the "oestradiol" fraction Rs identkidentical to that : 11-ketotestosterone rmylated and after TLC in system 5 the 3H-~4C formate. From the Ps (Table 2) the total ;t showed an Rs value identical to authentic amount ¢of produce osterone was calcul- tiol diformate. The 8H/14C ratios of the two ated; 0.2 per cent c dione was converted Iives appeared to correspond (Table 2), and to l11 l-ket, -ketotestosterc he average of these ratios the total production 1 lfl-H)ydroxyand~ adrenosterone. These lculated. The percentage of androstenedione fractions were pu LC in system 11. ted to oestradiol (_+ 0.2Yo/100 mg tissue) was A secon,d chrom~ as carried out on cellulose plates ir • After these two tone. The "oestrone" fraction was also purificatiq9urification steps, t H activity remaining formylated and partly methylated, The with eacheacl: steroid a ~e too low to permit ctivity in the areas corresponding with the further examination. :ives after TLC in system 3 (oestrone formate) and in system 6 (oestrone-methyl ether) appeared to DISCUSSION consist: mainly of 14C activity. The amount of 3H- oestroneae was too low to allow further examination. In the,ovary of the guppy, P. reticulata,reticuk steroido- genesis tatakes place in the granulosa celcells of growing van Oordt, Conversion•sion to llfl-hydroxy- and ll-ketosteroids follicles I(Lambert, 1970; Lambert & 1974). TTwo different incubation expe~eriments were The neutral fraction was first subjected to TLC carried olout to investigate some pathwayrs in steroido- in systemem 4 in order to remove fatty material, and genesis. Firstly, homogenized ovarian tissue of then se;parated into four fractions by TLC in system 3-month-old virgin guppies was incubated with 1. The four fractions, correspondingnding with the four 3H-pregnenolone and x4C-progesterovesterone, to study added labeled carriers, were eluted and further which metabolic pathway is functioning the delta-5 purified as follows. route, or the delta-4 route. A seconecond incubation 11 fl-hydroxytestosterone. The fraction was was carried out with ovarian homogenates of purified first by TLC in systema 9 (running twice) 12-month-old virgin guppies, with 3H-andro- and then by TLC on cellulose I)lates in system 12. stenedione as the precursor, to deterndetermine whether The 3H,14C material, correspondinging in RI value with the ovary is able to synthesize steroids with authentic 11 fl-hydroxytestosteronene, was then partly functional groups in the 11-position. oxidized. After TLC in system 1,, the ZH,I4C product The results of the 3H-pregnenololmolone and 14C- had the same mobility as authentic 11 fl-hydro- progesterone incubation demonstrate tlthat the ovary xyandrostenedione. Another portionrtion was formylated of young guppies contains enzyme systems for and after TLC in system 1 thele chromatographic synthesizing steroids by the delta-5 as well as the behaviour was the same as authenticathentic 1 lfl-hydro- delta-4 pathway. The synthesis by the delta-5 xytestosterone formate. From thethe free steroid and route follows from the isolation of~f tritiatritiated dehydro- both derivatives, the 8H/14C ratiosins were established epiandrosterone. The isolation of 14C-labeled (Table 2); from the average of,f the ratios it was androstenedione indicates the possibilipossibility of synthe- calculated that + 0.02 per cent of androstenedione sizing steroids by the delta-4 route. The delta-4 was converted to 1 lfl-hydroxytestosteroistosterone, route does not seem to be the principa)rincipal pathway in 11-Ketotestosterone. This fraction;tion was chromato- the ovary of the guppy, for the am~amount of I~C- graphed first by TLC in systemm 10 and then by androstenedione is low in comparis9arlson with the partition chromatography on celluloselulose plates (system amount of 3H-androstenedione (the ~H/14C ratio 12). Oxidation of the 3H,~C materialmterial, eluted after is 33). Moreover, in the testosterc;tosterone fraction, partition chromatography, yieldedded a 8H,14C corn- tritium activity was the only one to be detected. ~enesis in ovarian tissue of Poecilia retic~

ratios of 17a- the ovaries of old g : be the explanation and andro- of sex conversion, observed in older :oncluded that female guppies. one to andro- ng step. Thus, Acknowledgement- s are grateful to genesis mainly Professor Dr. P. G. dt for his interest and :,ckstein (1970) his critical reading of pt. Tilapia aurea, three different nating delta-5 '~ :I route, where AXELgODL. R., MA~ GOLDZlEHER J. W. & terone may oe convertea toO testosterone PULLIAMPI.JLLIAI~ J. E. ( ive identification of it androstendione as an intermediate. In the microquuantities of ~teroids by recrystal- , there are no indications for this third lization to constaJ tivity. Acta. Endoc., ay. Although we have not succeeded in Copenh. 49, (Suppl tstrating a synthesis of oestradiol in this BROWNJ. B., BULB! GgEEmVOOD F. C. ment, it may be concluded that ovaries of (1957).~An addition step for a method for estimati~estimating oestriol, [ oestradiol in human th-old guppies should be able to synthesize urine. J.3 Endocr. 1 5iol with preguenolone as precursor, since Bum I. E. (1961) rography of Steroids. rely transforms the latter into androstenedione Pergam~~rnon Press, C is also able to convert androstenedione to EcKsa~m B. (1970) pathways of steroid dial (lambert et aL, 1971). biosynttrnthesis in ova teleost, Tilapia aurea. results of the androstenedione incubation Gen. & compar. E~ -312. to prove that the ovaries of old guppies Ecr~r~iN B. & EYI~ The occurrence and apable of synthesizing 1 lfl-hydroxy- and biosynttrnthesis in vitr .'stosterone in ovarian asteroids; tritiated llfl-hydroxytestosterone tissue ofo the mulk to, derived from two biotope~,ms.~es. uen.Gen. o~& compar, r.naocr.r. 14,1% 396-403.~ I1 as tritiated 11-ketotestosterone have been EcKsrEINEcrsrnNi B. & KATZ Y. (1971) Steroidoge~enesis in post- d. A synthesis of llfl-hydroxy-andro- and prespawnedpre., ovaries of a Cichlid fish,, Tilapia aurea. stenedionelone and adrenosteron, however, has not been Camp. Biochem. Physiol. 38A, 329-338. demonstrated. Although an interpretation of EVLAT~ UJ. & ECKS'rEn~B. (1969) Steroid I:biosynthesis in quantitativetative findings of in vitro studies is difficult, vitro by ovarian tissue of mullets (Mugil sp.) from sea the obvious9vious predominance of l l-ketotestosterone and fresgh-water. Gen. & compar. EndocrEndocr. 13, 504. indicatestes that it is the most important one from all IDLERD. R., SCHMIDT P. J. & BIELY JJ. (1961) The steroidsIS with a functional group in the 11-position androgemic activity of l l-ketotestostero;stosterone: a steroid synthesizedsized in the ovaries of 12-month-old guppies, in salmcsalmon plasma. Can. J. Biochera. Ph)psiol. 39, 317- A transformati~ 320. asformation of androstenedione to oestradiol LAM'aERT.1J. G. D. (1970) The ovary of the guppy Poecilia by homomogenates of the ovaries from older guppies reticularreticulata. The granulosa cells as sitqsites of steroid was alsoe,~ established;~etahli~h,~rl. pern~. 100mg1/3/3 ~.ag of~,e ovarian~,,.~;~ tissue,;~.,~ biosynthesis.. . . Gen. & compar. Endocr. 115, 464-476. 0.2 per cent of androstenedionene appeared to be LAMBERTJ. G. D., TnUSSENJ. H. H. & EYLEYLATH U. (1971) converted to oestradiol. The recent:ent resultsr, obtained Conversion of aH-androstenedione and 8t8H-testosterone with ovaries of 12-month-old guPl~ppies are different to aH-oestradiol-17fl in vitro by the ovaDovary of the guppy from the results of the androstenedi~~nedione incubation Poecilia reticulata. Proc. Kon. Ned. Akad.Ak, Wet. Ser. with homogenates of ovaries from 3-month-old C. 74, 52-59. animals (Lambert et al., 1971). None of the above- LAMBERTJ. G. D. & VAN OORDT P. G. W. J. (1974) mentioned steroids with a functional:tional group in the Ovarian hormones in teleosts. Fortshritt¢~rtshritte der Zoologie. 22, 340-349. 1 l-position has been demonstratedted after incubation REImmTaR. (1972) Hormonal control ofc the teleost of young ovaries with aH-androstemldrostenedione;tenedi~ the ovary. Am. Zoologist 12, 307-324. conversion of androstenedione into oestradiol was WRIGHTR. S. (1971) A reagent for the n~non-destructive as high as 0'5~o per 100 mg tissue.Le. These differences location of steroids and some other lipop~hilic materials seem to prove that steroid meta3.bolism in the ovary on silica gel thin-layer chromatograms. JJ. Chromatogr. changes during aging of virgin guppies. Oestrogen 59, 220-221. synthesis (oestradiol) decreases , whereas androgen synthesis (11-ketotestosterone) increase.,increases.ases. The pro- Key Word lndex--Teleost; ovary; steroidogenesis;sU duction of typical fish andro~gens, such as 11- Poecilia reticulata. ketotestosterone and 11fl-hydrox3:oxytestosterone, in