Fish Physiology and Biochemistry vol. 7 nos 1-4 pp 229-235 (1989) Kugler Publications, Amsterdam/Berkeley

Sex steroids in intersexual fishes S.T.H. Chan and W.S.B. Yeung* Department of Zoology, University of Hong Kong; *Department of Applied Biology and Chemical Technology, Hong Kong Polytechnic, Hong Kong. Keywords: Monopterus albus, Rhabdosargus sarba, sex reversal, protogyny, protandry, steroidogenesis, sex steroids

Abstract Studies of the in vitro gonadal steroidogenesis in intersexual fishes, using labelled testosterone as precursor, showed large species variation. The protogynous Monopterus albus produced predominantly 5Sr-reduced metabolites while the protandrous Rhabdosargussarba produced mainly 5S-reduced products. Both fishes synthesized 1I-oxotestosterone; the synthesis of which appeared to mediate mainly through adrenosterone in M. albus but via 11-hydroxytestosterone in R. sarba. When the plasma levels of androstenedione, testosterone, I1 -oxotestosterone, 11 3-hydroxytestosterone, estrone and 173-estradiol among the male, intersexual and female phase of the same species were compared, available data showed that either there was no obvious difference among the different sexual phases or the differences could be accounted for by the seasonal reproductive activities of the animal. Except for androstenedione, there are no marked changes in plasma testosterone, 11-oxotestosterone, 11 3-hydroxytestosterone, estrone and 173-estradiol levels in the intersexual phase compared with the female and male, it is unlikely that these classical sex steroids act as a primary trigger of natural sex reversal in these fishes; the role of androstenedione awaits further elucidation. Introduction The endocrinology in intersexual fishes has been investigated in a number of studies, but no generalization can be made on the pattern of steroidogenesis in these fishes (see reviews Reinboth 1979; Chan and Yeung 1983). We report here in vitro and in vivo steroidogenic studies on two intersexual fishes, fresh-water protogynous Monopterus albus and marine protandrous Rhabdosargussarba.

Materials and methods Animal samples M. albus was obtained from the wild through local

fish suppliers, and R. sarba was obtained both from the wild and from fish farms. The sexual status of both fishes were determined by histology based on the criteria used by Chan and Phillips (1967) and Yeung and Chan (1987) respectively. The males refer to those with a functional testicular tissue and an undeveloped (R. sarba) or spent (M. albus) ovarian tissue, and vice versa for a female phase. When both testicular and ovarian tissues were present in the intersexual gonad during the maturation of one and the regression of the other, usually only the male tissue remained functional in the intersexual phase in both species. The seasonal stages of the fishes were primarily grouped according to the fish yearly spawning time; however, as spawning seasons of these species are protracted, additional histological criteria were

230 CHCI3:Ether

(9:1)

-

Saturated C -ditnes 1 9 E1

(sitica

5A - -38- 17

CHCt3:Ether

gel)

(9:1)

50-A-3R-17 CRUDE BUSH 83 EXT RACT

(siLica

E 2

gel)

Saturated C -diones 1 9 Ad (RA) E (O,A) 1 (OA) (0,A)

5-A-3-17 5*A-3B-17 E (0,A) 110HAd

(O,R)

(0,A)

TO:CH 3 0H 17QO P - Adr 11OHAd Saturated

Adr

(siLica

B:EtO0

C19-diols

Saturated

(92:8)gel)

(24:1)

111KT

-

CHC:

3

C C 3

0H (84:6)

(O,A) 8:EtAC

(1:1)

(silica

gel)

110T

B:EtAc

50-A-3a,17B 58-A-3.,17B 58-A-

30,17

Ad 110HAd

Adr

5a-Androstane-3, 17-diane 59-Androstan-3-ol-17-one 5a-Androstan-38,178-diol 5-Androstan-3a,17A-dioL 58-Androstan-38, 178-di)o 58-Androstan-3a, 178-di Androstenedion. 1 1ydroayandrostenedione ndrenosterone BAne-.89.. Benzene

CB- a -f e

CaCl

(0,R,A)

(1:1) 110HT (,

(silica

50-A-3R.178

(O.A,E) (O,A,E)

gel) 10HT

Sa -A-3, 17 50- A-3X-17

5-A-3a, 17) 5S-A-3a.179

110T

110HT (si lic

50-A-3S,17B (O1A,E) 5-A-3B, 178 (0,A,E)

C -dil5 1 9 (alumina)

17IOHP

T (RA) Adr (RA)

E, E2 EtAc EtOH

Estrone Estradi o Ethyl Acetate Ethanol

CH30 P

Methanol

T 11OT

Testosterone 11 -xotestosterone

110HT

118-Hydroeytestosterone

To

Toluene

R,A)

gel)

Progesterone

Chloroform 3

Fig. 1. Flow sheet for chromatographic separation of various steroid metabolites. 0, R, A and E in brackets indicate isopolarity of oxidized, reduced, acetylated and enzymatic derived products of the metabolites with authentic derivatives.

also used. Gonads with mature gametes, and with little intermediate gametogenic stages were classified as 'spawning', while 'prespawning' or 'postspawning/inactive' were those with germ cells at different stages of gametogenesis or with mainly resting gonial cells respectively.

tography and thin-layer chromatography. The systems used and the abbreviations for various steroids were given in Fig. 1. The metabolites were identified by comparing the mobility of the metabolites and their derivatives with authentic steroids and steroid derivatives in different chromatographic systems, and by recrystallization to constant specific activity.

In vitro studies Finely chopped gonadal tissue (40-900mg) from freshly killed M. albus and R. sarba of different sexual status in the spawning season were incubated with 24Ci (14C)-testosterone (specific activity 58mCi/mmol) in Kreb's ringer bicarbonate solution (pH 7.4) and isotonic medium (pH 7.4; Idler and Macnab 1967) respectively for 2h at 250C with 95% oxygen and 5% carbon dioxide. No cofactor was added. Incubations were done in duplicates or triplicates. After the completion of incubation, steroid metabolites were extracted, purified by paper chroma-

In vivo studies Steroids in plasma samples of the fishes in different reproductive stages were extracted and the amount of adrostenedione (Ad), testosterone (T), 11-oxotestosterone (1lOT), 11 -hydroxytestosterone (I 1OHT), estrone (El) and 17/-estradiol (E2) were measured by radioimmunoassay. For the simultaneous determination of plasma Ad, T, lOT and 11OHT, the method by celite chromatography and RIA (Chan and Yeung 1987) was followed. E2

231 an u

80.

80.

80.

-Anorosteneohone

5f-Androstan-3.l7t-diol

5 -Androston -3,17P diL 5-Arndoston, -t7-one

_

Androstenedione

0 1-O.xotestosterone o] I13 -Hydroytestostne

*SI-And onsto dione 17-

0 Adrenosterone O 11-Ootestosterone

60

0 .2

U

/ 30

'In(C

C y

dxu

2,Q

xua[

Erly Lat O P

status

status

O EybL

P!'k

Ef

Sexual

status

Fig. 2. Conversion of radioactive testosterone to various steroids by M. albus gonadal tissue at various sexual phases. The percentages represent percentages of total activity recovered after organic extraction. Unidentified metabolites were not listed.

and E1l were assayed directly from plasma extract without any prior purification step. Results In vitro studies M. albus (Fig. 2) High 5Se-reduced activity and a low androgen producing ability were found in the female. In the early intersexual phase, the production of 5a-reduced products and androgens were low, and 82.8% of the metabolites was Ad. In the male phase, 31.0% of the metabolite was lIOT. A significant amount of Adr was also produced. No 53-reduced metabolites had been found in all incubations. R. sarba (Fig. 3) Sixty three percent of the metabolites from male phase was 5-androstan-3a,17-diol. On the other hand 59% of the metabolite from the female phase was Ad. The intersexual phase gave an intermediate amount of the products. All the sexual phases produced lOT and 11OHT but the ratio of 11OHT to lOT was highest in male phase (9.3) and lowest in the female phase (1.7); intersexual phase, 7.2. No 5a-saturated product had been identified in all incubations.

Fig. 3. Conversion of radioactive testosterone to various steroids by R. sarba gonadal tissue at various sexual phases. The percentages represent percentages of total activity recovered after organic extraction. Unidentified metabolites were not listed.

Plasma steroid levels M. albus (Table 1) In the postspawned/inactive period, the plasma levels of Ad was highest in the intersexual phases and lowest in the male phase. High level of Ad was found in both the female and the intersex in the prespawning period. There was no significant difference in the level of T among different sexual phases in the same reproductive stage except the mid-intersex which had a significantly higher level during the postspawned period and a lower level in the spawning period. The level of E2 was highest in the female during the spawning period. The female phase showed a prespawning increase in Ad and a spawning increase in E2 when compared with that in the inactive period. Ad level decreased in the mid-intersexual phase and remained constant in the male phase throughout the reproductive cycle. T levels decreased in the midintersexual phase and increased in the male phase during the reproductive cycle. For 1llOT and 1 OHT, no significant variations were found in different sexual phases throughout the reproductive cycle (Fig. 4A). No E1l was detected.

232 Table 1. Levels of plasma free androstenedione, testosterone and 170-estradiol in Monopterus albus during inactive, prespawning, and spawning periods Steroid and sexual status

Postspawning/ inactive

Prespawning

Spawning

Androstenedione Female Early intersex Mid intersex Male

667 2003 1867 409

(8) (7) (6) (16)

1117 +

607

111 ±

88 + 301c ± 282c + 35c ±

Testosterone Female Early intersex Mid intersex Male

164 184 381 90

+ 51 (8) + 40 (6) ± 76a (6) ± 10 (16)

17/-estradiol Female Early intersex Mid intersex Male

85 155 108 71

±

14 (8) + 23a (3) + 22 (5) ± 12 (16)

83 (9) 1251 ± 343 (4) 577 + 55d (8) 19 (9)

247 ± 105a (4) 136 + 13z (8) 25 ±

17 (9)

57 + 25 (4) 63 + 14 (9)

155 363 ± 46Z 487 ± 38

(12)

157 ±

(12)

28

(2) (9)

39 ± Ic,z (2) 216 + 34Z (9) 157 + 28x (11) 28 + 28C (2) 79 + 19a (8)

Data are shown as mean (pg/ml) ± SEM; - no sample available; Statistical comparisons between groups were made in accordance with (a) the seasonal phases of the reproductive cycle, with the postspawned/inactive phase chosen arbitarily as the control (xp < 0.05, Yp < 0.02, p < 0.01), and (b) the sexual status in sex reversal, with the female phase chosen arbitrarily as the control (ap < 0.05, bp < 0.02, Cp < 0.01, dp < 0.001).

R. sarba (Table 2) The 3 sexual phases showed no significant difference for all the steroids measured in the postspawned/inactive period. During the prespawning period, higher levels of Ad and E2 were found in the female than those in the intersex and the male. The levels of all the steroids measured in the male were similar to those in the intersex during the same reproductive period. In the spawning period, the level of Ad was lowest in the female. The levels of T in female were significantly lower than those in the intersex but not those in the male. The levels of Ad increased in the spawning period in both the male and the intersex. In the female, pre-spawning increases in Ad and E2 were found. There was a slight, but significant, increase in T levels during the spawning period in the intersexual phase. For I 1OHT, no significant variation was found on different sexual status. No El was detected and only 1 of the 77 samples investigated possessed a detectable amount of 1 lOT (Fig. 4B).

Discussion Comparing the in vitro steroidogenic capacity in the two intersexual fishes studied, there was obvious species variation. 5ao-reduced metabolites were found only in M. albus, but only 5-reduced products were evident in R. sarba. The present data, together with earlier results obtained from incubations with progesterone as precursor (Yeung et al. 1985; Yeung and Chan 1985), suggest a possible difference in the biosynthetic pathway of I IOT. The production of Ad and Adr by M. albus suggests T - Ad - IlOHAd - Adr - 11OT being the major biosynthetic pathway in this fish. On the other hand, because of the formation of OT ap11OHT but not Adr, T - 11OHT pears to be the major pathway in R. sarba. However, these data do not exclude the possible operation of alternative pathways for the biosynthesis of 1 lOT in both fishes. The present in vitro data show that the male and the female phases of the two intersexual fishes have

233 11 - OH-testosterone 0.7

11- Oxotestosterone

11P-OH-testosterone 11-Oxotestosterone

m

*2-5 *

a

15. 0

0

0

0 0 a'

CE

c -C 0

Il

n

16)

(4L(5) 14) -T"·i3

(6) 5)(140 (7)' '- *--T

07.

C

IUo E

d V)

P a

Ul

(.)

2

(9)

Pg o

15.

o

C

AC

C

- 00In

0· nu

=1

0

(10) (12) (15) 0---0-

0

C

c C

oC

Ie

0c

'25

o

0 !

1 0

4_- -o (6)_(lr (5) 0

-

0

-vl

(5)_0

S

(3)=

n

07'

_L

w25

n-

v

a.

·

2. a _L

n

-

(8) (4) (3) _._._._-

5.

0 c C

.c

0. )

(n

(7)

?

(2t (2L (/.! (1 El' Wf

(11)

C

(2)(17)

E M

.

d

(A) M. albus

0 0

I

I -

0

I

( )

n (M)._ 4 d' I'

6) ?

(10j6) (8) 0?

(B) R. sarba

Fig. 4. Levels of plasma free I 1-hydroxytestosterone and 1 -oxotestosterone in M. albus (A) and in R. Sarba (B) during inactive, prespawning, and spawning periods. Horizontal bars represent the means; numbers in parentheses represent numbers of zero readings.

different steroidogenic capacity. Whilst the gonad in the female phase of M. albus has a higher 5areductase activity, the gonad in the male phase has a higher ability in the production of androgens. In R. sarba, gonadal 5-reductase activity is high in the male phase and relatively low in the female phase. Ad production is, however, higher in the female than in the male. Sex reversal is, therefore, accompanied by a change in gonadal steroidogenic capacity of the fish. In M. albus, the initiation of protogynous sex reversal is coincided with a proliferation of interstitial Leydig cells at early intersexual phase. This is accompanied by an abrupt increase in Ad produc-

tion both in vitro and in vivo and a decrease in 5eareductase activity in vitro. The production of Ad is relatively low in both the male and the female phases. In R. sarba, the changes in steroidogenesis appear to be gradual as shown by the intermediate amount of metabolites produced in the intersexual phase. But it is possible that the lack of drastic structural changes in the intersexual gonad of R. sarba may be due to the difficulty in identification of its 'early intersexual phase' by normal histological criteria. Also the proliferation of female follicles in R. sarbaappeared not as obvious as that of the male lobules in M. albus. T, IIOT, E2, and El are the common steroids

234 Table 2. Levels of plasma free androstenedione, testosterone and 17-estradiol in Rhabdosargussarba during inactive, prespawning, and spawning periods Steroid and sexual status

Postspawning/ inactive

Prespawning

Spawning

Androstenedione Male Intersex Female

666 + 82 (10) 735 + 181 (12) 701 + 114 (15)

619 + 71 (8) 723 + 72 (4) 1625 ± 314,b (3)

1067 ± 89z (10) 1282 ± 121x (7) 654 + 61c (8)

Testosterone Male Intersex Female

360 + 21 (10) 272 + 54 (12) 278 ± 33 (15)

321 + 55 302 + 48 796 ± 305

(8) (4) (3)

332 + 55 494 + 82 232 + 31

(10) (7) (8)

17f3-estradiol Male Intersex Female

335 ± 37 (11) 319 ± 74 (12) 307 + 30 (15)

428 + 106 (6) 437 + 86 (4) 994 + 189Y,b (3)

490 + 93 484 + 72 336 ± 49

(10) (6) (3)

Data are shown as mean (pg/ml) + SEM; Statistical comparisons between groups were made in accordance with (a) the seasonal phases of the reproductive cycle, with the postspawned/inactive phase chosen arbitrarily as the control (xp < 0.05, Yp < 0.02, p < 0.01), and (b) the sexual status in sex reversal, with the female phase chosen arbitrarily as the control (ap < 0.05, bp < 0.02, Cp < 0.01).

measured in the study of fish reproductive endocrinology. In M. albus and R. sarba, 11OT and El show no marked variation throughout the reproductive cycle. Plasma testosterone increases in the functional male and E2 increases in the functional female as spawning is approaching. Similar changes have been reported in many gonochoristic fishes to be associated with reproductive activities. It is, therefore, unlikely that these common steroids serve as primary trigger for natural sex reversal in the fishes investigated. The present results on levels of androstenedione, a non-classical steroid measured in fish plasma showed obvious variation in relation to the seasonal reproductive activities of the female suggesting its possible involvement in the reproduction in these two species. The most interesting feature, as in agreement with the in vitro data, is the marked increase in Ad levels in early intersexual phase of M. albus. As natural sex reversal is found to be a post-spawning event (Chan and Phillips 1967), the rise in Ad levels in a presumably steroidogenic inactive period, and the constant and low level of the steroid in the male phase suggest that this steroid might bear some relation to natural sex reversal. However, whether this surge of Ad serves as the primary trigger or just reflects a secondary event

for natural sex reversal awaits further elucidation. The present study, and also in many previous investigations, has shown that fish gonads are capable of producing a variety of 'non-classical' steroids such as 5c and 5:-reduced steroids. Due to the lack of specific antibodies quantitative assays of these steroids by radioimmunoassay are at present impossible. As these steroids may be of importance in our understanding of the fish endocrinology, particularly that of intersexual species, it is essential that specific assays should be established for these non-classical steroids.

Acknowledgements We thank Dr. D.E. Kime for the generous gift of the antisera against I I OT and 1 OHT. This study was supported by research grants from the University of Hong Kong.

References cited Chan, S.T.H. and Phillips, J.G. 1967. The structure of the gonad during neutral sex reversal in Monopterus albus (Pisces; Teleostei). J. Zool. Lond. 151: 129-141.

235 Chan, S.T.H. and Yeung, W.S.B. 1983. Sex control and sex reversal in fish under natural conditions. In Fish Physiology Vol. IXB. pp. 171-222. Edited by W.S. Hoar, D.J. Randall and E.M. Donaldson. Academic Press, New York. Chan, S.T.H. and Yeung, W.S.B. 1986. A new method for the simultaneous determination of androstenedione, testosterone, I l-oxotestosterone and 113-hydroxytestosterone in fish plasma using combined technique of Celite chromatography and radioimmunoassay. J. Ster. Biochem. 25: 1013-1022. Idler, D.R. and Macnab, H.C. 1967. The biosynthesis of 1l-ketotestosterone and 1l3-hydroxytestosterone by Atlantic salmon tissue in vitro. Can. J. Biochem. 45: 581-589. Reinboth, R. 1979. On steroidogenic pathways in ambisexual fishes. Proc. Ind. Nat. Sci. Acad. Part B 45: 421-428.

Yeung, W.S.B., Adal, M.N., Hui, S.W.B. and Chan, S.T.H. 1985. The ultrastructural and biosynthetic characteristics of steroidogenic cells in the gonad of Monopterus albus (Teleostei) during natural sex reversal. Cell Tiss. Res. 239: 383394. Yeung, W.S.B. and Chan, S.T.H. 1985. The in vitro metabolism of radioactive progesterone and testosterone by gonads of the protandrous Rhabdosargus sarba at various sexual phases. Gen. Comp. Endocrinol. 59: 171-183. Yeung, W.S.B. and Chan, S.T.H. 1987. The gonadal anatomy and sexual pattern of the protandrous sex-reversing fish, Rhabdosargus sarba (Teleostei: Sparidae). J. Zool. Lond. 212: 521-532.