Identification of testosterone and progesterone in hemolymph of larvae of the fleshfly Sarcophaga bullata

GENERAL

AND

COMPARATIVE

Identification

ENDOCRINOLOGY

52, 368-378 (1983)

of Testosterone and Progesterone in Hemolymph Larvae of the Fleshfly Sarcophaga bullata

of

D. DE CLERCK,* W. EECHAUTE,? I. LEUSEN,? H. DIEDERIK,$ AND A. DE LOOF* *Catholic University of Leaven, Zoological Institute, Naamsestraat 59, B-3000 Leuven, Belgium, f University of Gent, Laboratory of Normal and Pathological Physiology, De Pintelaan 185,’ B-9000 Gent, Belgium, and fZoologica1 Laboratory, Section for Comparative Endocrinology, University of Vtreckt, Padualaan 8, 3584 CH Utrecht, The Netherlands

AcceptedDecember 9, 1982 Testosterone- and progesterone-like substances were detected by radioimmunoassay (RIA) in chromatographed extracts of hemolymph from larvae of Sarcophaga bullata (S.B.). Gas chromatographic (GC) analysis after heptafluorobutyric acid (HFBA) derivation of hemolymph extracts, purified by paper and silica gel column chromatography, showed a peak in the GC recordings with the same retention time as the HFBA derivative of pure testosterone. A testosterone concentration of 92 ng/lOO ml hemolymph was found by GC; the concentration of progesterone, calculated on the basis of the RIA, was about two times higher. After preparing the o-pentafluorobenzyloxime (OPFB)-heptafluorobutyryl ester (HFB) derivatives of the hemolymph extracts, negative ion chemical ionization capillary gas chromatography-mass spectrometry (NCFGC-MS) proved that hemolymph of larvae of the fleshfly Sarcophaga bullata indeed contains testosterone and progesterone. Several metabolites and precursors of the latter compounds could also be detected during the NCI/ GC-MS analyses. Estrogens could not be traced by any of the methods we used. This is the fast time that these steroids have been identified in insect hemolymph. These results add interesting perspectives for comparative endocrinology.

The immunocytochemical localization and chemical characterization of insulin (Duve and Thorpe, 1979; Duve et al., 1979), of pancreatic polypeptide (Duve and Thorpe, 1980; Duve et al., 1982), and of gastrin-cholecystokinin-like peptides (Dockray et al., 1981; Duve and Thorpe, 1981) in the heads of Calliphora blowfly species, together with the discovery that at least adult Diptera seem to use the steroid ecdysterone, their molting hormone, as the physiological equivalent of estrogens of egg-laying vertebrates in the induction of vitellogenin synthesis (Huybrechts and De Loof, 1981; De Loof et al., 1980) led to the recently introduced concept (De Loof, 1982) that the endocrine systems of vertebrates and insects may have more elements in common than generally assumed. The existence of an androgenic hormone was reported some time ago in several crustaceans, e.g., Orchestia gammarella (Char368 0016-6480/83 $1.50 Copyright All rights

@ 1983 by Academic press, Inc. of reproduction in any form reserved.

niaux-Cotton,

1957) and in the insect Lam1966a,b,c; 1969). As nothing is known about the chemical nature of such a hormone in arthropods and as steroid hormones are very conservative molecules in the course of evolution (Sandor and Mehdi, 1979), we investigated as a first approach whether perhaps biologically active steroids found in biological material from vertebrates, especially testosterone and progesterone, also occur in some insects and specifically in hemolymph from larvae of the flesMy Sarcophaga bullata.

pyris (Naisse,

MATERIALS

AND METHODS

Animals. S. bullata were raised as described elsewhere (Huybrechts and De Loof, 1981). Hemolymph from unsexed last instar larvae which had left the food and which had empty guts for at least 8 hr was used. By insertion of the tine tip of a capillary into the body cavity, hemolymph was collected without tissue fragments, transferred to ice-cooled l-ml vials, centrifuged, and stored at - 20” until use.

TESTOSTERONE

AND PROGESTERONE

Radioimmunoassay. In a fust type of experiment hemolymph (18.5 ml) from about 550 larvae was extracted with freshly distilled ether (200 ml). The ether extract was washed with 0.1 N NaOH solution and water and evaporated to dryness under a stream of nitrogen. The residue was then chromatographed on a Sephadex LH 20 column (i.d. 1.1 cm; 32 cm height) using benzene/dichloromethane/methanol (60/3.5/5) as elution fluid. The first 5 ml of eluate was discarded; then, 2.5fractions, 2 ml, were collected and evaporated to dryness and the residue was redissolved in 250 ).~l 10 mM sodium phosphate buffer. One hundred microliters was taken from each fraction for radioimmunoassay of testosterone and progesterone. Antiprogesterone-l 1S-BSA serum and anti-testosterone7~BSA serum from Miles were used. The anti-testosterone serum was highly specific: cross-reactivity of 34 C-19 steroids and of ecdysteroids, cw-ecdysone, P-ecdysone, makisterone, muristerone, and 22,25-dideoxyecdysone was less than 0.1%; cross reactivity of So-dihydrotestosterone, 5Bdihydrotestosterone, and 4-androstene-3,17dione was 45, 22, and 2%, respectively. Sixty-two steroids (4pregnenes and 5a- and SB-pregnanes) were controlled for reactivity with the anti-progesterone serum: cross reactivity was 6, 12, and 14%, respectively, for desoxycorticosterone, 20a-hydroxyprogesterone, and ZOBhydroxyprogesterone and less than 0.1% for other steroids Gas clavomatogrphy (CC). In a second type of experiment hemolymph (250 ml) was extracted with dichloromethane (2000 ml); after addition of 60,000 dpm [1,2,6,7-3H]testosterone (sp act 81 Ci/mmol) the extract was evaporated to dryness and delipidated by treatment with water/ethanol (3/l); after centrifugation the ethanol/water phase was reextracted with dichloromethane and the extract was chromatographed on Sephadex LH-20. The eluate fraction (20th to 30th ml), in which testosterone is eluted, was evaporated to dryness and the residue chromatographed on paper in the petroleum ether system (SO- lOO”)/methanol/water (100/70130). After radiochromatogram scanning the testosterone peak was excised and eluted with methanol; the methanolic eluate was evaporated to dryness and the residue chromatographed on a silica gel column (1.d. 1 cm; 5 cm height) using 2% ethanol in dichloromethane and the testosterone containing eluate (ml 14 to 23) evaporated in a test tube. The residue was redissolved in 1 ml ethanol; 200 JL~was taken for LSC (scintiilator: 5 g PPO + 300 mg POPOP + 20 ml ethanol/liter toluene). The remaining 800 ~1 was evaporated to dryness for GC analysis (with electron capture (EC) detection) after HFBA derivation. Preparation of HFBA derivatives was performed by the addition of 100 ~1 heptafluorobutyric acid anhydride and 100 pl benzene; the reaction occurred at 25”

IN Sarcophaga buNuta

3

for 18 hr. The reaction mixture was evaporated and redissolved in 100 ~1 n-heptane. Gas chromatography with EC detection was carried out on a 2% SE-30 on Diatoport S (2-m column) at 235” using a Carlo Erbs 2350 Series gas chromatograph. Negative ion chemical ionization gas chrotnatuography-mass spectrometry (NCIIGC-MS). Analyses were done with a HP 5985B-CC-MS-DS instrument. The procedure for extraction, ion-pair extraction, and capillary gas chromatographic separation are outlined by Diederik and Lambert (1982). The derivation of the ketosteroids to o-pentafluorobenzyloxime (GPFB) derivatives was carried out by a modification of the method of Nambara et al. (1975). The ketosteroids were converted to the corresponding oximes by heating the samples at 100” for 1 hr with 0.2 ml of a 0.2% solution of o-pentafluorobenzylhydroxylamineHCl in pyridine and 5 ~1 0.01 M HCl. After GPF oximation hydroxyl functions were converted to heptafluorobutyryl (HFB) esters by a modification of the method of Berthou et ai. (1974) using 100 )~l heptafluorobutyrylimidazole as reagent instead of the corresponding HFB anhydride. The samples were heated at 100” for 1 hr. The o-pentafmorobenzyloxime of the keto function increases the molecular weight by I95 Da per keto group and the heptatmorobutyryl ester increases the molecular weight by 196 Da per hydroxyl group. The OPFB-HFB derivatives of steroids possess high molecular weights; this increases the sensitivity and the specificity since the background noise is much lower in the higher mass range (Gleispach ef al., 1981). The electron capture negative ion chemical ionization mass spectrometric technique was chosen to record the spectra of these electrophylic substituted derivatives (Markey et al., 1978). This technique provides enhanced sample ion current and also a significantly reduced background noise. Furthermore, using this technique the probability is low that a sample ion at a particular mle ratio will be obscured by the presence of a fragment ion, derived from bigh molecular weight contaminant of the undissolved sample under the GC conditions employed in the analysis (Hunt and Crow, 1978). Steroids in hemolymph extract were identified by comparing the obtained normalized NCI mass spectra out of a total ion current of mle 400 - m/e 1000 with the spectra of standards at the expected GC retention times under negative ion conditions using methane as reagent gas. Background correction was performed for the analysis of testosterone in the hemolymph extract. This was done by subtracting from the spectrum recorded at retention time 31:38 min, which is supposed to correspond to testosterone-OPFB-HFB ions, the spectra recorded at retention time 31:43 and 31:58 min. Multiple ion recording of preselected ions (selected

DE CLERCK

370

ion monitoring (SIM) data) was used for reconstruction of mass fragmentograms to confirm the molecular anion or the (M - HF) ion of steroid derivatives of their expected GC retention times.

RESULTS In the first type of experiment an extract of hemolymph from larvae of S. bullata was chromatographed on a Sephadex LH-20 column and 2-ml fractions were collected for steroid radioimmunoassay. When the fractions were analyzed for the presence of substances which interfere with the antiserum against testosterone, an elution pattern with two peaks (Fig. 1.1) was constructed for the hemolymph extract: a small peak A (from the 14th to the 17th ml) and a much larger peak B (from the 20th to the 30th ml). Comparing the elution pattern with elution curves for pure steroids (Fig. 1.2) chromatographed on the same column, a complete correspondence between peak B and the elution curve for pure testosPg-@

/c/

-Ria-progesterone

(curveI1)

----Ria-testosterone

Icurve

I)

-Progesterone ----20 OH Prog -.-. Testosterone 15000 -

sooo,

Ji, IO

,l 1 Lf 20

I I LI-I i. 30

I 40

ELUATE

I

lmll

FIG. 1. Elution pattern of hemolymph extract (Fig. 1.1) on Sephadex LH-20: fractions of 2 ml were analyzed by RIA for testosterone (curve I) and progesterone (curve II). Fig. 1.2 shows elution curves of pure steroids.

ET AL.

terone is observed, while peak A occurs in the same zone in which androstenedione is eluted. Calculated in terms of testosterone peak A represents about 152 pg and peak B about 1095 pg of a compound which crossreacts with the antiserum against testosterone. An elution pattern with two peaks is also obtained when the 2-ml fractions are analyzed for material which interferes with progesterone for binding with the anti-progesterone serum. Indeed, an important peak C between the 13th and 17th ml is seen in the elution pattern (Fig. 1.1, curve II); this peak corresponds closely with the elution curve for progesterone and represents about 2.550 pg of material, calculated in terms of progesterone. This means that the progesterone concentration is approximately two times higher than that of testosterone. A second more flattened peak D occurs in elution pattern II between the 20th and 29th ml, where 20a-hydroxyprogesterone is eluted; peak D represents about 650 pg of material, calculated in terms of progesterone. In the second type experiment the testosterone containing fraction, obtained after Sephadex LH-20 chromatography of the extraction of 60 ml hemolymph, was further purified by paperand silica gel column chromatography and analyzed for testosterone by gas chromatography (GC at 235’) .after HFBA derivation. The GC recording for the fraction isolated in this way (Fig. 2) showed a peak with the same retention time, i.e., 5:08 (between 5:05 and 5:lO) as the HFBA derivative of 50 pg pure testosterone while the blank treated through the whole technical procedure lacked interfering peaks with comparable retention times. Analyses at lower temperatures, i.e., at 225” and 215”, resulted in retention times of 7:22 min (between 7:20 and 7:25) and lo:35 min (between lo:32 and 10:38), respectively, for the HFBA derivative of both the pure testosterone and the isolated fraction. After

TESTOSTERONE

AND PROGESTERONE

FIG. 2. Gas chromatogram (1) of the HFBA derivative of pure testosterone, (2) of the testosterone fraction isolated chromatographically from 250 ml hemolymph, and (3) of a blank treated through the whole isolation method. Gas chromatography was carried out on a 2% SE-30 on Diatoport S at 23.5” (see Methods). Arrows correspond to retention time of 5x08 min.

correction for experimental losses, a testosterone concentration of 92 ng per 100 ml hemolymph was calculated. Analysis of the mass spectra gave the following results. The typical fragment ions for the o-pentafluorobenzyloxime-heptafluorobutyryl derivative of testosterone (testosterone-OPFB-HFB: MW 679.3) are m/e of 659.3 (M - HF), 429.3 (M - (HF -t 30)), and 498.2 (M - PFB). The molecular anion or an (M - HF) ion carries most of the sample current under these NC1 conditions (Hunt and Crow, 1978). The retention time of these fragment ions is 29:32 min (Fig. 3A). In the hemolymph extract, the three typical ions are also found but at the 31:48 min retention time (Fig. 3B). Because of this slight difference in retention time as compared to the standard, additional parameters were introduced. The testosterone-OPFB-HFB peak is immediately preceded by the cholesterol-HFB present

IN

Sarcophaga

bulkm

371

in this extract and the latter has a well known retarding effect on these types of steroid derivatives resulting in slightly higher retention times, due to thermal deg-, radation of cholesterol-HFB ) eliminative heptafluorobutyric acid, during the gas chromatography (Francis et al., 1978). This was confirmed by adding to the hemolym~~ extract an internal standard which eluted in between that of cholesterol-HFB and testosterone-OPFB-HFB. The retention time of this standard 11 @hydroxyetiochoianolone-OPFB-di-HFB had also increased by nearly 2 min compared to its retention time in a standard mixture of steroids without cholesterol. The spectra with retention times of 3 I:43 and 3158 min (Figs. 4A, B) were recorded to determine the background spectra in the direct neighborhood of the compound which might correspond to testosteroneOPFB-HFB. No important peak at m/e of 659.2 was found in these spectra. tracting these background spectra ( and scan 195) from the spectrum at 3B:48 min a clean spectrum was obtained (Pig. 4C) which was remarkably similar to the standard spectrum of testosterone-QPFBHFB (Fig. 3A). The relative abundance ratio of the (M - HP) isotopes with m/e of 659.2, 660.2, and 661.2 is 100:33.9:7.0 the standard and 100:33.0:6.4 for the 3 min peak of the hemolymph extract. T values correspond to a molecular anion or (M - HF) ion with the same number of C atoms as in the testosterone-OPFB derivative (McLafferty, 1980). A reconstruction of the mass fragmentograms of the three characteris ions for testosterone-OPFB-H selected ion monitoring of nine 5) between retention times 25 and 44 min revealed that these ions only occur at the retention time typical for testosteroneOPFB-HFB. The large peaks at the #m/e of 629.3 at 35 ta 37 min are not related tosterone because the correspondin ment ions at mle 659.3 and 498.2 are

372 1

DE CLERCK

tl

149

RET.

TIME:

27.32

TOT

kBUttb=

tl-PFB

a

173

RET.

TIME:

31.43

19863. tw2

TOT

ET AL.

FiEIJND=

EFI~E

txrnmtn:

657.3

.TESTOSTEROt~E-OPFB-HF~

I 3:; . ‘3

656.3

BUSE

15478.

76?3.

M-HF

REL t3BUND

FK;‘RBUttD:

657.2,’

Ml~J=i579

493.2

.3

M- (HF+30)

1110.

‘:-HF 653.2

r7.17

629.3

REL t3BUND

M/2 792.2

330.6

877.6

715.5

659.2 ff;‘;

948.4

100.0 33.0 6.4

. 630

720

760

300

840

880

920

360

1008

B

FIG. 3. (A) Normalized NCI/CH4 mass spectrum of the anti-isomer of standard testosterone-OPFBHFB derivative at retention time 29:32 min. Fragment ions: m/e 659, 629, and 498. Isotopic ratio (M - HF) ion: 100.0:33.9:7.0. (B) Normalized NCIICH4 mass spectrum of hemolymph from larvae of Sarcophaga bullatn at retention time 31:48 min. Ions with significant abundance: m/e 659, 629, and 498. Isotopic ratio of the base peak (M - HF) ion: 100.0:33.0:6.4.

missing. Addition of the peak area of all three fragment ions at 3 1:48 min yields 84% of the peak area of the total ion current of nine selected ions monitored. This indicates that the background is very low and that we are dealing with a fairly pure substance. The absence at m/e of 639.3 (Fig. 3B) excludes the presence of dehydroepiandrosterone (DHA)-OPFB-HFB at the retention time of the fragment ions of testosterone OPFB-HFB, although both derivatives possess the same (M - HF) ion (Figs. 6 and 3A, respectively). Since DHA-OPFBHFB differs from testosterone-OPFBHFB by retention time-it elutes a few minutes earlier from the column with one peak-Fig. 5 shows only twice the (M -

HF) ion m/e of 659.3 of the syn- and antioxime isomers of testosterone OPFBHFB, indicating that DHA-OPFB-HFB could not be detected at all in this particular hemolymph sample. On capillary columns syn- and antioxime isomers of some 3-0x04-ene steroids are separated as MO-TMS derivatives (Maume et al., 1979) as was found here also for the OPFB-HFB derivative at the retention time typical for testosterone. For instance, the 17-ketosteroid DHA either as MO-TMS or as OPFB-HFB derivative gives only one GC peak (Fig. 6). For more detailed information concerning MO-TMS derivatives see Leunissen and Thyssen (1978) and Horning et aE. (1968). Finally, we also checked whether the hemolymph extract contained metabolites

TESTOSTERONE

tl

135

RET,

TIME:

31 .S$

AND PROGESTERONE

TOT

UBUND=

13898

IN Sarcophaga

.

BOSE

FR:‘WUND

bullata

:

3-u

344

.3.~’

345.

M-b&-

I

1

M-FFB

438 ..? 64.3 * ,> 568 .‘3

FIG. 4. (C) Normalized averaged NCIKH4 mass spectrum of the testosterone-OPFB-HFB tive represented in Fig. 3B after background spectra were subtracted (A and B).

of testosterone. For example, the (M HF) ion m/e of 661.2 in the above mentioned SIM data experiment could be detected at the proper retention time of 5~ dihydrotestosterone-OPFB-HFB which confirmed that Scr-dihydrotestosterone was present in the hemolymph extract, be it in

deriva-

small amounts (this will be described elsewhere). From the combination of all these data we may conclude that the ions with pve/eof 659.3, 629.3, and 498,2 are the fragment ions derived from testosterone-OPF HFB in the hemolymph extract.

374

DE CLERCK

~HUEMOLYMFH -NEG-CIdH4I

ET AL.

S.HULLRTb’OPFB-HFB-DER. .5TORRI 2X10-S i [FINTII

‘....,, -

2:

,

.

I

I

I

I

36

37

3: 4: 5: 6: 7: 8: 3: 10:

I

I Ii

““‘--. ...._._..,,,,-,,,,, -

I

34

35

1 38

3!3

4’0

489.3 498.2 623.3 659.3 661.3 673.3 703.3 753.4 TOTAL

to to to to to to to t.0 ION

4’1

.C 4’2 4’3

489.3 498.2 623.3 659.3 661.3 673.3 f03.3 753.4

-$‘-I

FIG. 5. A reconstruction of the mass fragmentograms of hemolymph from larvae of Sarcophaga bullata during 25 to 44 min of three characteristic fragment ions (m/e 498.2, 629.3, and 659.3) for testosterone-OPFB-HFB out of a selected ion monitoring data of nine ions under NCI/CHd conditions .

A similar approach was followed for the identification of the di-OPFB derivative of progesterone (MW 704.3). The typical fragment ions are m/e of 684.3 (M - HF), 654.3 (M - (HF + 30)), and 523.3 (M - PFB) (anti-isomer, Fig. 7A) with a retention time of 51:lO min. Exactly the same m/e ion 0

131

RET.

TIME:

25.5%

SOT

RBUND=

peaks were found in the hemolymph extract at the 51:12 min retention time (Fig. 7B). The background was very low, partially because of the excellent clean-up by the ionpair extraction procedure; background correction was not even necessary. The relative abundance ratio for the ions with m/e 3607

.

B&X

DEHYDRO-EPI-fiNDROSTEROtdE--OPFBHFB

100 50

509.2 426.2

FK~FIBIJNII

: 653.2.’

MW=679.3 REL

446.3461.3

10.~0,

REL

0 100 640.2

50 692.2

20.6

662.4

1.8

703.0

0 600

620

640

660

680

700

720

740

760

780

6. Normalized NCI/CH4 mass spectrum of standard dehydroepiandrosterone-OPFB-HFB derivative at retention time 25:58 min. Fragment ions: m/e 659, 639, and 509. Isotopic ratio (M - HF) ion: 100.0:34.7:7.3. FIG.

TESTOSTERONE

AND PROGESTERONE

IN Sarcophaga

bullata

375

II-HF

1g 0 5Q 1 445 cl I] 10040b

.z

( 48b

520

FROGESTERWE-D 58 1 @I,,

u-

486.3

440

721.:> , ,. 7 2 bj

RET.

753.5 , . (

,

760 TIME:

,

5’30 , . ..(

,

,

,

.3 600

640

I FlNT I :s , 830

8 40

TOT

FtBUHD=

4

cHF+3Bj 654.3

d-l

098.9

800 51.12

556.3 , ._, SE.0

I-OPFB Mld=784.3

797.7 * )

r ‘3 J . ‘3

E,8 .: . 3

M-PFB 523.3

*

*

I 920

-

68b

66S.3 669.3 604.3

937.2 t

t-l ./ 2

I

’ 635.3 686.3

H;:;D

2.0 1.2 lb0.b 40.8 8.0

1 2 .a, 1 Q = M-HF

M-PFE

t-l- (HF+SB)

b r:.8 *

PROGESTEROt~E-D

I -OPFB 84S.C

728

76.9

8b0

S4b

f. UNT I1

872.2

‘920.7 89b

9Zb

M..’2 684 .2 635.2 606.2 695.2

REL HBUND 1cw.b 33. e 8.9 3.9 -

FIG. 7. (A) Normalized NCI/CH4 mass spectrum of the anti-isomer of standard progesterone-diOPFB derivatives at retention time 51:lO min. Fragment ions: m/e 684, 654, and 523. Isotopic ratio (M - HF) ion: 100.0:40.8:8.0. (B) Normalized NCI/CH4 mass spectrum of hemolymph from !arvae of Surcophaga bufiatu at retention time 51:12 min. Ions with significant abundance: m/e 684, 654, and 523. Isotopic ratio of the base peak (M - HF) ion: 100.0:39.1:8.9.

of 684.3, 685.3, and 686.3 was 100.0:40.8:8.0 for the standard and 100.0:39.1:8.9 for the extract. This isotopic ratio is in excellent agreement with the number of C atoms present in the progesterone-di-OPFB derivative (McLafferty, 1980). The 3-o-pentafluorobenzyloxime-20-omethyloxime derivative of progesterone (progesterone-OPFB-MO: MW = 538.2) was also prepared and could be identified in the hemolymph extract by its fragment ions, namely, m/e of 518.2 (M - HF), 488.2 ((RI + I)-(HF + OCH3)), and 357.2 (M PFB) at the retention time of 36:20 min for the syn- and 37:30 min for the anti-isomer of this dioxime. The loss of 31 mass units under electron impact (EI) conditions of o-

methyloxime derived steroids is commonly achieved, giving a (M - 31) +‘ion (Adlercreutz et al., 1975; Aringer et al., 1971), The loss of 50 mass units under NCI conditions, giving the m!e of 488.2 anion, can be explained by the elimination of of the pentafluorobenzylic ring, th nation of OCHs from the ~-met~y~oxime function, and the protonation of the nitrogen by the reagent gas, which demonstrate the presence of at least two derived keto functions in the steroid isolated from the hemolymph extract. In our opionion, these data allow us to conclude that the hemolymph extract contains progesterone. During the NCVGC-MS analyses we also

376

DE CLERCK

obtained sufficient evidence to prove the presence of androstenedione, Sa-dihydrotestosterone, 1lB-hydroxytestosterone, 1 l-ketotestosterone; SP-androstane-3a,l7@diol; So-androstane-3B, 17l3-diol; androst-5-ene3p,17p-dial; 17a-hydroxyprogesterone; and 17~ - hydroxy,20B - dihydroproges terone. Estrogens could not be found, not even in trace amounts, neither by RIA nor by NCUGC-MS and not by positive ion chemical ionization GC-MS either. These results will be described separately. DISCUSSION Chromatography on the Sephadex LH-20 column combined with RIA demonstrated that the hemolymph extract from larvae of S. bullata contained substances which were bound by antibodies against testosterone as well as against progesterone (see Fig. 1). About 88% of testosterone antibody-positive material showed an elution peak B, which corresponded closely to the elution curve of pure testosterone; on the other hand, about 80% of progesterone antibodypositive material was eluted at the same position (peak C) as pure progesterone. Although both antisera, especially the one against testosterone, were highly specific, interference from other substances could not be excluded; nevertheless, the radioimmunoassays following the chromatographic step on Sephadex LH-20 suggested that both progesterone and testosterone (or dihydrotestosterone, which is also eluted with testosterone from Sephadex LH-20) were present in hemolymph from larvae of S. bullata. The small peaks A (curve I) and D (curve II) suggested the presence of other cross-reacting compounds, probably 4-androstenedione and 20~ or 20@hydroxyprogesterone, respectively, as these steroids are eluted at these positions. Radioimmunoassay gives only an indirect indication of the presence of a steroid in insufficiently purified extracts from biological material; other substances may indeed interfere or disturb the equilibrium be-

ET AL.

tween antigen and antibody. Therefore, physicochemical techniques such as GC and EC detection and NCI/GC-MS were used for the analysis and characterization of the isolated steroids. The GC analysis with EC detection has shown that after HFBA derivation of the testosterone-containing fraction, isolated from the hemolymph extract by different types of chromatography (Sephadex LH20, paper chromatography, silica gel column chromatography), a peak with an identical retention time as the HFBA derivative of pure testosterone was seen in the GC recordings. The same GC technique, however, was not sensitive enough to confirm the presence of other steroids (progesterone, 20~ or 20B-hydroxyprogesterone, 4-androstenedione) in hemolymph or to confirm our RIA findings. Final proof of the presence of testosterone, progesterone, and several metabolites has been given by the very powerful NCUGC-MS technique. All data concerning the m/e of major fragment ions, isotopic ratios, peak area ratios, selected ion monitoring data, and retention times were in agreement with those of the corresponding standards. In our opinion, the combination of all our data, obtained from different types of analytical approaches, constitutes conclusive evidence of the presence of both testosterone and progesterone in hemolymph from S. bullata larvae. This is the first time that these steroid hormones have been clearly demonstrated in biological material from insects. Although it cannot be completely ruled out, it is unlikely that the steroids we found originate from the liver, which the larvae were fed. Indeed, the liver very rapidly metabolises reproductive steroids (Lacroix and Eechaute, 1975). The liver used as food had disintegrated almost completely by the time the larvae left it, and furthermore the larvae had emptied their gut at least 8 hr before hemolymph was collected.

TESTOSTERONE

AND PROGESTERONE

A metabolization study, whereby larvae were injected with tritiated precursors of testosterone, has almost been completed; the data of this study indicate the existence of different enzymes, which are normally involved in progesterone and/or testosterone biosynthesis in vertebrates, so confirming the results described here. Some enzymatic reactions of steroid formation have already been demonstrated in invertebrates (Sandor and Mehdi, 1979), but the hormones themselves have never been demonstrated in hemolymph. In our experiments, however, the presence of the biologically active steroids progesterone and testosterone was shown clearly. The fact that no estrogens could be demonstrated by any of the analytical techniques used was not completely unexpected. Indeed, flies seem to use their moulting hormone ecdysterone as a female sex hormone, at least as related to vitellogenin synthesis (Huybrechts and De Loof, 1981; De Loof, 1982). The next goal of our research is to obtain more qualitative and quantitative data about the different steroids during the developmental cycle and to elucidate the functions of these steroids. Further research will be needed to find out whether testosterone or its metabolites might perhaps function as a male sex hormone in insects (De Loof, 1982). In our opinion, the concentrations of testosterone and progesterone we found in hemolymph are such that they should not be considered too low to have a physiological significance. These results constitute another argument in favor of the concept that the basic principles of the endocrine system might perhaps be universal in the animal kingdom (De Loof, 1982) and that some vertebrate reproductive steroids not only occur in chordates and echinoderms (see Schoenmakers, 1979), but may also be common among protostomians.

IN Sarcophaga

bullata

377

ACKNOWLEDGMENTS We thank the IWONL of Belgium for sponsoring this project. We also thank Mrs. Van der Eeken for typing, Mrs. Puttemans for the figures, Mr. Evans for critical evaluation of the text and applied mass spectrometrical, and the Toxicological Laboratory (AMTOL B.V.), Kleverparkweg 13 Rd, 2023 Ca Haarlem, The Netherlands, for the NCIIGC-MS analyses.

REFERENCES Adlercreutz, H., Martin, E, Wahlroos, 6, and Soini, E. (1975). Mass spectrometric and mass fragmentographic determination of natural and synthetic steroids in biological fluids. J. Steroid Biochem. 6, 247-259. Aringer, L., Eneroth, P., and Gustafson, J. A. (1971). Trimethylbromosilane catalyzed trimet~ylsiIy~ation of slow reacting hydroxy- and oxosteroids in gas chromatographic-mass spectrometris analysis.

Steroids

17(3),

377-398.

Berthou, F., Bardou, L., and Floch, H. H. (1974). Separation of urinary androstanediol and pregnandiol isomers by a combined gas-liquid chromatography-thin layer chromatography method. 9. Chromatogr. 93, 149-165. Charniaux-Cotton, N. (1957). Croissance, regeneration et determinisme endocrinien des caracteres sexuels d’orchestia gammarella Pallas (Crustace Amphipode). Ann. Sci. Nat. 19, 411-560. De Loof, A. (1982). New concepts in endocrine control of vitellogenesis and in functioning of the ovary in insects. In “Exogenous and Endogenous Influences on Metabolic and Neural Control” (A. D. F. Addink and N. Spronk, Eds.), pp. 145177. Pergamon, Oxford/New York. De Loof, A., Huybrechts, R., and Briers, T. (1980). Do insects have steroid sex hormones and do compounds with juvenile hormone activity occur in vertebrates? Ann. Sot. R. Zool. Belg. 110, 179-184. Diederik, H., and Lambert, .I. 6, D. (1982). Steroids in plasma of the female rainbow trout before and after ovulation by NCI-GUMS. In ‘Proceedings Intern. Symp. Reprod. Physiol. Fish” (C. J. 5. Richter and H. J. Th. Goos, Eds.), pp. 107-108. Pudoc , Wageningen. Dockray, 6. I., Duve, H., and Thorpe, A. (1981). Immunochemical characterisation of gastrlniehoiecystokinin-like peptides in the brain of the blowfly, Calliphora vomitaria. Gen. Camp. Endocrinol.

45, 492-496.

Duve, H., and Thorpe, A. (1979). Imrnu~o~~ores~~ut localisation of insulin-like material in the medial

378

DE CLERCK

neurosecretory cells of the blowfly, Calliphora vomitoria (Diptera). Cell Tissue Res. 200, 187-191. Duve, H., and Thorpe, A. (1980). Localisation of pancreatic polypeptide (PP)-like immunoreactive material in neurones of the brain of the blowfly, Calliphora erythrocephala (Diptera). Cell Tissue Res. 210, 101-109. Duve, H., and Thorpe, A. (1981). Gastrin/cholecystokinin (CCK)-like immunoreactive neurones in the brain of the blowfly, Calliphora erythrocephala (Diptera). Gen. Comp. Endocrinol. 43, 381-391. Duve, H., Thorpe, A., and Lazarus, N. R. (1979). Isolation of material displaying insulin-like immunological and biological activity from the brain of the blowfly Calliphora vomitoria. Biochem. J. 184, 221-227.

Duve, H., Thorpe, A., Lazarus, N. R., and Lowry, P. J. (1982). A neuropeptide of the blowfly Calliphora vomitoria with an amino acid composition homologous with vertebrate pancreatic polypeptide. Biochem. J. 201, 429-432. Francis, A. J., Morgan, E. D., and Poole, C. E (1978). Flophemesyl derivatives of alcohols, phenols, amines and carboxylic acids and their use in gas chromatography with electron-capture detection. J. Chromatogr. 161, 111-117. Gleispach, H., Wurz, E., and Mayer, B. (1981). Measurement of plasma steroids in the femtomole range using gas chromatography-mass spectrometry. In “Analytical Chemistry Symposia Series: Proc. Symp. on the Analysis of Steroids, Eger, Hungary, 1981,” Vol. 10, “Advances in Steroid Analysis” (S. Giiriig, Ed.), pp 307-314. Elsevier, Amsterdam/Oxford/New York, 1982. Horning, M. G., Moss, A. M., and Horning, E. C. (1968). Formation and gas-liquid chromatographic behavior of isometric steroid ketone methoxime derivatives. Anal. Biochem. 22, 284-294. Hunt, D. F., and Crow, E W. (1978). Electron capture negative ion chemical ionization mass spectrometry. Anal. Chem. 50(13), 1781-1784. Huybrechts, R., and De Loof, A. (1981). Effect of ecdysterone on vitellogenin concentration in haemolymph of male and female Sarcophaga bullata. Int. J. Znvertebr. Reprod. 3, 157-168. Lacroix, E., and Eechaute, W. (1975). The metabolism of oestradiol by perfused livers and liver slices of normal and serrotic rats. J. Med. 6, 411-431.

ET AL. Leunissen, W. J. J., and Thyssen, J. H. H. (1978). Quantitative analysis of steroid profiles from urine by capillary gas chromatography. J. Chromatogr. 146, 365-380. Markey, S. P., Lewy, A. J., and Colburn, R. W. (1978). Studies in biogenic amine metabolism by mass spectrometry. In “Quantitative Mass Spectrometry in Life Sciences II. (A. P. de Leenheer, R. R. Rocucci, C. van Peteghem, Eds.), pp. 1737. Elsevier Scientific, Amsterdam/Oxford/New York. Maume, B. E, Millot, C., Mesnier, D., Patouraux, D., Doumas, J., and Tomori, E. (1979). Quantitative analysis of corticosteroids in adrenal cell cultures by capillary column gas chromatography combined with mass spectrometry. J. Chromatogr. 186, 581-594. McLafferty, E W. (1980). Table 2.2, Isotopic contributions for carbon and hydrogen. In “Interpretation of Mass Spectra,” 3rd ed. Univ. Sci. Books, Mill Valley, Calif. Naisse, J. (1966a). Controle endocrinien de la differenciation sexuelle chez l’insecte Lampyris noctiluca. I. Role androgene des testicules. Arch. Biol. Liige 77, 139-201. Naisse, J. (1966b). Controle endocrinien de la differenciation sexuelle chez Lampyris noctiluca. II. Phenomenes neurosecretoires et endocrines au tours du developpement postembryonnaire chez le male et la femelle. Gen. Comp. Endocrinol. 7, 85-104.

Nalsse, J. (1966~). Controle endocrinien de la diierenciation sexuelle chez Lampyris noctiluca. III. Influence des hormones de la pars intercerebralis. Gen. Comp. Endocrinol. 7, 105-110. Naisse, J. (1969). Roles des neurohormones dans la differentiation sexuelle de Lampyris noctiluca. J. Insect Physiol. 15, 877-892. Nambara, T., Kigasawa, K., Iwata, T., and Ibuki, M. (1975). Studies on steroids. CIII. A new type of derivative for electron capture-gas chromatography of ketosteroids. J. Chromatogr. 114,81-86. Sandor, T., and Mehdi, A. E. (1979). Steroids and evolution. In “Hormones and Evolution” (J. E. W. Barrington, Ed.), pp. l-72. Academic Press, New York. Schoenmakers, H. (1979). “Steroids and Reproduction of the Female Asterias rubens L.” Ph.D. thesis, University of Utrecht.