Polypeptide composition and biosynthesis of the yolk proteins in the firebrat Thermobia domestica (insecta thysanura).код для вставкиСкачать
Archives of insect Biochemistry and Physiology 9:299-312 (1988) Polypeptide Composition and Biosynthesis of the Yolk Proteins in the Firebrat, Thermobia domestica (Insecta, Thysanura) Andre Rousset, Colette Bitsch, and JacquesBitsch Laboratoire d 'Entomologie, CNRS URA 108, Universitt! Paul-Sabatier, 31062 Toulouse, France ion-exchange chromatography of crude ovarian extracts of the primitive insect Thermobia domestica allowed the separation, in native conditions, of major and minor vitellins of molecular weights of 300,000 and 430,000, respectively. Their polypeptide subunits were determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunotransfer using an antiserum prepared against major vitellin. This protein was resolved into large (M, 166,000-212,000) and small (around M, 50,000) polypeptides. Minor vitellin, on the other hand, exclusively contained small polypeptides that are immunologically different from those of the major vitellin. Vitellogenin polypeptides from the hemolymph of mature females exhibited electrophoretic mobilities and immunological properties similar to vitellin polypeptides. Pulse-chase experiments showed that the female fat body synthesizes radioactive and immunoprecipitable proteins, whose polypeptide pattern is close to that of the major vitellogenin. However, part of the primary vitellogenic polypeptides, at M, 210,000 and 212,000, is rapidly processed to M, 176,000 and 182,000 subunits. These two polypeptides, as well as the precursors, enter into the composition of the major hemolymph vitellogenin. Finally, processing of the still uncleaved 210,000-212,000 polypeptides takes place in the ovary, which performs the same step of vitellogenin maturation as the fat body. Key words: polypeptides, immunology, pulse-chase, processing INTRODUCTION Many investigations have been performed on insects to characterize the hemolymph vitellogenins, which are mainly synthesized by the fat body, and the so-called vitellins, which accumulate in the maturing oocytes (recent reviews in [l-41). A comparison between various insect species of several Received May 23,1988; accepted September 28,1988. Address reprint requests to Andre Rousset, Laboratoire d'Entomologie, Universite PaulSabatier, 118 route de Narbonne, 31062 Toulouse Cedex, France. 0 1988 Alan R. Liss, Inc. 300 Rousset et al. orders has led to interesting considerations on the evolution of these proteins . Little information is as yet available for other groups, such as the primitive Thysanura (Apterygota). In adult Thermobiu dornesticu, previous investigations using native polyacrylamide gel electrophoresis demonstrated the presence of three vitellins in crude ovarian extracts . Three female-specific proteins of the same electrophoretic mobilities were also detected in the hemolymph and their fluctuations measured during the alternative previtellogenic and vitellogenic phases of the reproductive cycles that are synchronized with the permanent molting cycles. Further investigations were carried It was out using an antiserum prepared against the major vitellin fraction demonstrated that the vitellogenin assigned to M,430,000 is immunologically different from the group M, 240,000 + 300,000 (the major fraction). The purpose of the present study was to obtain a better characterization of the yolk proteins of T. domestics by determining their polypeptide composition and to investigate their biosynthesis using pulse-chase experiments with fat bodies incubated in vitro. [n. MATERIALS AND METHODS Animals Themzobiu dornesticu (Packard) were reared under constant temperature and humidity (37°C and 80% rh). Investigations were carried out on 8-month-old females, kept with males, and inseminated during the first 3 days of the 10day intermolt period . Most results were recorded on day 3 of the intermolt during the vitellogenic phase of the ovarian cycle . Some experiments were also performed on females during the previtellogenic phase (on days 7 and 10 of the intermolt period) and on males collected on days 3 or 10. Ion-exchange Chromatography Ovaries pooled from 200 inseminated females collected on day 3 were homogenized in phosphate-buffered saline (0.05 M sodium phosphate, pH 7.6, 0.01% Triton X-100, 2 pglml aprotinin, 0.3 M NaC1). After centrifugation, the supernatant was applied to a DEAE* Sepharose CL-6B column (Pharma- cia, Sweden), as previously described dry at -80°C. [q.Lyophilized fractions were stored Protein Determination Protein determination was performed according to the method of Bradford [lo] using BSA (Sigma Chemical Co.) as standard protein. Collection of Samples and PAGE Ovaries were obtained from C02-anesthetized females, rinsed in Ephrussi’s saline solution [ll] containing 1 mh4 PMSF, pooled and stored frozen *Abbreviations: BSA = bovine serum albumin; DEAE = diethylaminoethylcellulose;PAGE = polyacrylamide gel electrophoresis; PMSF = phenylmethylsulphonyl fluoride; SDS = sodium dodecyl sulfate; TCA = trichloracetic acid. Yolk Polypeptides in Thermobia domestica 301 at -80°C. Hemolymph samples (2 p1) were collected in glass micropipettes after puncturing the neck, immediately diluted with the Ephrussi's solution (1:2), centrifuged (15,OOOg, 10 min, 4"C), and then stored frozen at -80°C. Pore-limited electrophoresis was conducted in nondenaturing polyacrylamide gradient slab gels (6-20%) as previously described ; about 5 pg of proteins were run in each well. The electrophoresis system of Laemmli [l2] was used for separation of the peptides in SDS. About 8 pg of proteins were used in each well. Electrophoresis was carried out at 24 mA (constant current) for 5 h at 20°C using 16-cm-long, 1-mm-thick, 6.5% slab gels, with 3% stacking gel (SDS-discontinuous buffer system in Bio-Rad Protean apparatus). Several denaturing reagents were tested: 1 and 0.5 M 2-mercaptoethanol, and 0.1 and 0.05 M dithiothreitol solutions. The solution of 0.1 M dithiothreitol gave the best results and complete denaturation without artifacts. Molecular weights were obtained by comparing the electrophoretic mobilities of protein standards. Pharmacia molecular standards used were thyroglobulin ( M , 669,000), ferritin ( M , 440,000), catalase ( M , 232,000), lactate dehydrogenase ( M , 140,000), and BSA ( M , 66,000). SDS-PAGE Bio-Rad protein standards used were myosin ( M , 200,000), galactosidase ( M , 116,000), phosphorylase B ( M , 92,000), BSA ( M , 66,000), and ovalbumin ( M , 45,000). The absorbance of the Coomasie Brillant Blue stain of the protein bands was recorded with a microdensitometer (type PHI 3, Vernon, France, with a Wratten N.22 filter). Immunological Procedures Antiserum against the major vitellin fraction was prepared as previously described [q.For qualitative investigations, the Ouchterlony double-diffusion test was performed using 1%agarose gel in 0.02 M barbital buffer, pH 8.6, for 2 days at 4°C. Analysis was carried out using 4 pg of crude ovary extract and 1.5 pg of each fraction separated by DEAE chromatography. The gels were dried, stained for 5 min in 0.1% amido black 10B (Merck), and then destained in 7% acetic acid. For immunoblotting, 0.5 pg protein samples were first separated by 6.5% SDS-PAGE, equilibrated in methanollTris-glycine buffer (1:4), and then electrophoretically transferred onto a nitrocellulose membrane (0.125 A, 15 h, 4°C). The blot was quenched for 3 h at ambient temperature in a blocking solution (0.05 M sodium phosphate, pH 7.6, 0.15 M NaCI, 5% dried milk), incubated with the antiserum, and revealed. Immunoprecipitation of vitellogenic proteins in 0.2 ml samples was carried out by adding antiserum (25 pl), major vitellin (10 pg) as carrier, and 0.8 ml buffer (0.05 M sodium phosphate buffer, pH 7.8, 0.15 M sodium chloride, 1% Triton X-100, 1% Na-deoxycholate, 1mM PMSF, 2 pglml Aprotinin), according to the method of Bownes and Nothiger . Synthesis of Yolk Proteins by the Fat Body and Pulse-Chase Experiments Fat body attached to the dorsal part of the abdominal wall was carefully dissected out in cold saline. After rinsing three times, fat bodies of two vitellogenic females were incubated in 300 p1 of M 20 medium (Gibco, France) 302 Rousset et al. that was supplemented with [3H]amino acid mixture (Amersham, Buckingamshire, England). The concentration of the labeled amino acids in the medium was adjusted to 0.1 pCilp1. In further experiments, fat bodies were labeled for a 10-min pulse, then removed, immediately homogenized, or transferred into vials containing 300 p1 of unlabeled medium for up to a 90min chase. Fat bodies were homogenized in 3-4 vol of the immunoprecipitation buffer in an all-glass Potter-Elvehjem homogenizer and then sonicated, and the homogenate centrifuged at lO0,OOOg for 60 min at 4°C. Vitellogenic proteins were immunoprecipitated either in the supernatant of the fat body samples or in the incubation medium. Total proteins of the medium were precipitated with TCA (10%) in the presence of 0.3 mglml BSA as carrier. Measurement of Radioactivity The radioactivity of a precipitated aliquot was counted using a liquid scintillation system (Beckman LS 3801). The immunoprecipitate was also dissolved in an SDS-mercaptoethanol buffer  and the polypeptides separated on SDS-PAGE and subjected to fluorography, according to Chamberlain [El. The dried gels were exposed to a Kodak film at -80°C for an appropriate period (between 7 and 30 days) and routinely developed. The radioactivity of each polypeptide was calculated by correlating the total radioactivity of the immunoprecipitate with the optical density of the radiolabeled bands measured with a microdensitometer. RESULTS Purification of Vitellins Purification of vitellins was carried out using ion-exchange chromatography on ovarian homogenates (about 4,000 follicles). Elution with a NaCl gradient separated several different fractions (Fig. 1) with absorbance values from 0.2 to 0.8. The fractions were analyzed on native PAGE and their patterns compared with those from crude ovarian extract (Fig. 2). No protein was detected in fractions eluted at 0.10 or 0.15 M NaCl (data not shown). The fraction eluted at 0.22 M NaCl showed a single protein band, which exhibited the same electrophoretic mobility as the band at M , 430,000 found in the ovarian extract. Two protein bands were stained in each of the fractions eluted at 0.25 and 0.30 M NaCl; they corresponded to 240,000 and 300,000 vitellins. The 300,000 vitellin band represented 80% of the proteins eluted at 0.30 M NaCl, but only 40% of the proteins eluted at 0.25 M NaCI. Although separation between the 240,000 and 300,000 proteins was not complete, it appears that the fraction eluted at 0.30 M NaCl contained most of the 300,000 vitellin that, in the ovary, represents about 70% of the total soluble proteins. Fractions were submitted to a double-immunodiffusion test using antiserum prepared against the protein fraction eluted at 0.30 M NaCl [q.As shown in Figure 3, precipitation arcs were obtained against the crude ovarian extract and also against the 0.30 and 0.25 M fractions. The continuity between the precipitation arcs demonstrates the antigenic identity of the samples. Yolk Polypeptides in Thermobia domestica 303 0 3 0.8- 0 E 0.6- l 4 I C 0 N m 0) 0.4 - u C a a ;; 0.2ul n a 100 200 300 400 500 e l u t i o n volume 60 0 700 (mi) Fig. 1. Elution profile of crude ovary extract of Thermobia domestica after DEAE chromatography. Fractions were eluted with a linear gradient from 0.05 to 0.40 M NaCl in 0.05 M phosphate buffer, pH 7.6. Absorbance was measured at 280 nm, and 10-ml fractions were collected. Furthermore, the nonreactivity of the antiserum against the fractions eluted at 0.10,0.15, or 0.22 M NaCl was observed in serial dilution tests, demonstrating the absence of the 240,000 and 300,000 vitellins in these fractions. Thus, the chromatographic method used led to a suitable separation of two groups of vitellins from the ovary of T. durnestica: a major vitellin that is chiefly composed of the M, 300,000 protein, and a minor vitellin (M, 430,000). Polypeptide Composition of the Vitellins Protein fractions were analyzed by electrophoresis under denaturing conditions (Fig. 4) and the relative proportions of the polypeptide bands measured by densitometry (Table l).The major fraction, eluted at 0.30 M NaC1, was resolved into numerous bands forming two groups: the first (about 75%) composed of large polypeptides and the second (about 25%) composed of small polypeptides. Six large polypeptides were detected: two thin bands at M,210,000 and 212,000, forming less than 10% of the total polypeptides, then three bands at M , 172,000, 176,000, and 182,000, among which the 176,000 and 182,000 bands were predominant and a 166,000 band was seen only as traces. Two polypeptides of low molecular weight (46,000 and 57,000) were also present. The fraction eluted at 0.25 M NaCl showed a polypeptide pattern similar to that of the major fraction, with an increasing percentage of the 166,000 band and also with smaller polypeptides of M, 50,000 and 52,000. These three polypeptides are probably subunits of the 240,000 vitellin that is predominant in the 0.25 M fraction. Lastly, the fraction eluted at 0.22 M NaCl was resolved exclusively into small polypeptides with molecular weight lower than 60,000, among which the 46,000 subunit was predominant. These data can be compared with the results obtained from homogenates of vitellogenic ovaries (Table 1).More than 75% of the ovarian polypeptides were of 304 Rousset et al. Yolk Polypeptides in Tbermobia domestica 305 high molecular weight, between 166,000 and 212,000. The bands at M , 176,000 and 182,000 together corresponded to about 65% of the total vitellogenin polypeptides. Two additional bands of low molecular weight were present, at M , 46,000 and 57,000. Thus, the polypeptide pattern from crude ovarian extract is qualitatively and quantitatively similar to that of the fraction eluted at 0.30 M NaC1. Immunotransfer was carried out on these SDS-PAGE electrophoretograms (Fig. 5). The minor 210,000 and 212,000 bands that were present in the ovary extract and in .the 0.30 M fraction appeared to be highly immunoreactive. A strong reaction was also shown for the bands between 166,000 and 182,000 in both the ovarian extract and the 0.30 M fraction. A reduction in crossreactivity was observed for the small polypeptides, particularly for the 46,000 and 57,000 bands. In contrast, the small polypeptides of the 0.22 M fraction did not crossreact, indicating that the 430,000 vitellin and the 300,000 vitellin polypeptides are different. TABLE 1. Composition and Percentage of the Different Polypeptides From Female-Specific Proteins of Vitellogenic Female Hemolyph (VQ H), From a Crude Ovary Extract (OV), and From 0.30, 0.25, and 0.22 NaCl DEAE Chromatographic Fractions Mr x 10-3 212 210 182 176 172 166 57 52 50 48 46 44 VQH ov ("/.I ("/.I Large polypeptides 30 5 30 5 23 35 16 30 a a Small polypeptides a 14 0.30 ("/.) 4 5 20 32 12 a 0.25 0.22 ("/.I ("/.I 1 12 8 16 15 13 11 15 12 a 10 13 20 11 56 21 'Traces. Fig. 2. Native PAGE of crude ovary extract (OV) and of 0.30, 0.25, and 0.22 M NaCl DEAE chromatographic fractions. ST = standard proteins (M, x Fig. 3. Ouchterlony double-diffusion plate. Central well contained an antiserum against the major vitellin fraction. Peripheral wells contained an OV and 0.30, 0.25, 0.22, 0.15, and 0.10 M NaCl DEAE chromatographic fractions. Fig. 4. SDS-PACE of the hernolymph of 7-day-old previtellogenic females ( P V Q H), 3-day-old males (uH), and of vitellogenic females (V Q H), and also of O V and of 0.30, 0.25, and 0.22 M NaCl DEAE chromatographic fractions. ST = standard proteins. The molecular weights ( x of the hernolymphatic nonvitellogenic polypeptides are noted on the left; those of vitellogenins and vitellins, on the right. 306 Rousset et al. @ 3:4 1:4 FBP 3:4 1:4 FB15 3:4 1:4 M15 Figures 5 and 6 3:4 1:4 FB30 3:4 1 :4 M30 Yolk Polypeptides in Thermobia domestica 307 Polypeptide Composition of the Hemolymph Proteins The polypeptide composition of the hemolymph proteins was investigated by SDS-PAGE analysis from samples collected from males and previtellogenic and vitellogenic females. The results are shown in Figure 4. The polypeptide pattern of male hemolymph did not change throughout the molting cycle. Two main groups of bands were discernible on the electrophoretogram. The first, constituting only 13% of the total polypeptides, was composed of several closely spaced bands of high molecular weight (M, between 169,000 and 230,000). The second group was composed of three bands ( M , 75,000, 83,000, and 89,000), together representing about 80% of the total polypeptides. Traces of other polypeptides, having a molecular weight below 70,000, were also detected. While the polypeptide composition of the previtellogenic female hemolymph was similar to that of males, in vitellogenic female hemolymph, the band at M, 83,000 became very weak, the band at M , 182,000 was intensified, and additional bands appeared at M, 176,000, 210,000, and 212,000, so that the high molecular weight subunits were about 50% of the total hemolymph polypeptides. The majority of the large polypeptides correspond to subunits of proteins present in the hemolymph only during the vitellogenic phase of the reproductive cycle. The distinct bands at M , 210,000 and 212,000 represented about 60% of the vitellogenic subunits, whereas the bands at M, 176,000 and 182,000 represented 40% (Table 1). Immunotransfer did not detect reactivity against the polypeptides of the male hemolymph (data not shown). By contrast, several polypeptides in the hemolymph of vitellogenic females were highly immunoreactive, forming two groups of bands at M , 210,000-212,000 and 176,000-182,000 (Fig. 5). The immunoreactivity of the small polypeptides found in the hemolymph was very weak. Protein Synthesis by the Fat Body Abdominal fat bodies were dissected out from firebrats collected on day 3 of the intermolt. Preliminary investigations involving incubation of fat bodies for 3 h in the presence of labeled amino acids showed that male and female fat bodies synthesized radioactive proteins that were continuously released into the medium. Using SDS-PAGE and fluorography analysis, it was demonstrated that no radioactive polypeptide was detected in the immunoprecip- (Figs. 5, 6) Fig. 5. lmrnunoblotting after separation by SDS-PAGE of 0.25, 0.22, and 0.30 M NaCl DEAE chromatographic fractions of crude ovary extract (OV) and of vitellogenic female hemolymph (V Q H). LP = large polypeptides. SP = small polypeptides. Fig. 6. SDS-PAGE fluorographs of the imrnunoprecipitated proteins synthesized by the fat body after an in vitro 3H-pulse-chase. FBP: fat body samples after 10-min pulse-labeling. FB15: fat body samples after 15-min chase. M15: incubation medium samples after 15-rnin chase. FB30: fat bodies after 30-min chase. M30: incubation medium after 30-rnin chase. Two lanes are given for each sample: on the left, the lane obtained from 3/4 of the immunoprecipitate, and on the right, that from the remaining 1/4. 308 Rousset et al. itate of the incubation medium of male fat bodies, whereas several radiolabeled subunits were detected in the immunoprecipitate of female fat body medium: several large polypeptides (between M, 176,000 and 212,000) were heavily labeled, while the labeling of other small polypeptides (about M, 50,000) was very weak. Two groups of labeled polypeptides were synthesized after a 10-min pulse (Fig. 6): polypeptides ( M , 63,000, 225,000, and 280,000) whose radioactivity was lower than 1%of the pulse-labeling and polypeptides more strongly labeled whose molecular weight was similar to that of the major vitellogenic polypeptides of the hemolymph (M, 176,000, 182,000, 210,000, and 212,000). The total radioactivity of the immunoprecipitated proteins varied between 10,000 and 22,000 dpm depending on the samples, but the proportion between the different polypeptides was always similar: 75% for the doublet at M, 210,000-212,000, 20% for the 182,000 band, and 5% for the 176,000 band. The changes in the radioactivity of the different polypeptides were studied after various times of the chase, and the results are expressed in Figure 7A as a percentage of the total radioactivity of both fat body extracts and incubation medium. Polypeptides at M, 63,000, 225,000, and 280,000 disappeared after 15-30 min (cf. Fig. 6). The radioactivity of the 210,000-212,000 bands regularly decreased, whereas the 176,000 and 182,000 polypeptides progressively became heavily labeled. After 30 min, the 176,000 and 182,000 subunits carried half of the total radioactivity, and at 60 min, each attained a radioactivity over that of the doublet at M, 210,000-212,000. With regard to the immunoprecipitated proteins accumulating in the medium, it can be noted that the labeled 176,000, 182,000, and 210,000-212,000 bands were also present, whereas the 63,000, 225,000, and 280,000 polypeptides were absent (Fig. 6). A small radioactive polypeptide at M, 46,000 had appeared in the medium after 20 min of incubation. During the first 10 min of the chase, the radioactivity of the 176,000, 182,000, and 210,000-212,000 bands was distributed as in the fat body at time 0 (Fig. 78). Thereafter, the radioactivity of the doublet 210,000-212,000 decreased markedly, while that of the 176,000 and 182,000 bands increased. After 30 min of chase, the doublet was less radioactive than either the 176,000 or 182,000 bands. DISCUSSION Previous work on the primitive insect Thermobiu domesticu has shown the presence of several electrophoretically and immunologically distinct vitellins in the ovary of adult females during the vitellogenic phase of their reproductive cycle [6,7l. In the present study, the vitellins extracted from mature ovaries were separated into two main fractions: a major fraction in which the 300,000 vitellin was preponderant (80%) with a low percentage of the 240,000 vitellin, and a minor fraction in which the 430,000 vitellin alone was collected. The polypeptide substructure of both fractions was then analyzed by SDSPAGE, and the crossreactivity of the subunits was detected using an antiserum prepared against the major vitellin fraction. The major vitellin fraction contains both large and small subunits. The high molecular weight polypeptides, 75% of the total polypeptides, are Yolk Polypeptides in Thermobia domesfica 75 - c 309 i t\ 210,000 212,000 \t; \\ 50 - 8 > t > + u 25 - a 0 0 a Lz O - ! 0 I , , / 5 10 15 20 60 30 90 C H A S E D U R A T I O N , rnin 75- 0- 3, 210,000 ! 212,000 \ i\ \ 50- 182~000 s > t ? 25- V 'z 0 a II: 0- 0 5 10 15 20 30 90 60 CHASE DURAlION , rnin Fig. 7. Changes in the percentage of radioactivity of the main polypeptides from the vitellogenic proteins synthesized by in vitro cultured fat bodies at different times of the chase following a pulse-labeling of 10 min. Other polypeptides, whose total radioactivity was below I % , have not been represented. Time 0 denotes the beginning of the chase. A, Total vitellogenic proteins, both from fat body and from the incubation medium. B, Vitellogenic proteins secreted into the incubation medium. Subunits at M, 176,000 (A ), 182,000 ( O ) ,and 210,000 and 212,000 (0). composed of a "doublet" at M , 210,000-212,000 and of several bands from 166,000 to 182,000. Two small subunits of My46,000 and 57,000 constitute 25% of the total polypeptides. All the subunits are immunoreactive. The minor vitellin fraction exclusively contains small polypeptides, among which the most abundant ( M y46,000) did not crossreact with our antiserum. So the small subunits of the minor vitellin (M, 430,000) appear to be immunologically different from those contained in the major vitellin, although their electrophoretic mobility was found to be identical. The polypeptide pattern 310 Rousset et al. from crude ovarian extracts is qualitatively and quantitatively similar to that of the major vitellin fraction. The hemolymph of vitellogenic females contains polypeptides of the same electrophoretic mobility and of the same immunological properties as those of the major vitellin fraction, but with different proportions (Table 1). On the basis of the size of their vitellogenin polypeptides, pterygote insects have been distributed into three groups [1,4,5]. Although the major vitellin fraction of T. domesticu (bands at M , 240,000 + 300,000) contains two different classes of polypeptides, this insect belongs to group one, which comprises the majority of insect orders, namely, the Dictyoptera and Orthoptera. But, by its minor vitellin (at M, 430,000), which exclusively contains small polypeptides, it could also find a place in group three of Harnish and White (or class I1 of Kunkel and Nordin), which, until now, seemed to be restricted to higher Diptera. Thus, the primitive apterygote insect T. domestics, generally regarded as closely related to the hypothetical ancestor of pterygote insects, appears to constitute an original model, characterized by the coexistence of two distinct classes of vitellogenins. Incubation of fat bodies of T. domestica in the presence of tritiated amino acids demonstrated that the fat body synthesizes proteins that were precipitated by our antiserum. The polypeptide composition of the vitellogenic proteins secreted by female fat bodies is characterized by high incorporation of radioactivity in the large subunits. The pattern obtained after fluorography appears to be very similar to that of the hemolymph polypeptides after immunotransfer. Thus, the polypeptide composition of vitellogenic proteins secreted by cultured fat bodies is close to that of the hemolymph vitellogenins. Biosynthesis of yolk proteins by the fat body of 7: domesticu was also investigated using pulse-chase experiments in order to learn about the processing of the major vitellogenin. The radioactivity of the doublet at M , 210,000-2l2,000, which contains 75% of total radioactivity of fat body extracts after a 10-min pulse, progressively decreases during the chase whereas the radioactivity of the 176,000 and 182,000 subunits increases. In addition, a 46,000 polypeptide secreted into the medium becomes slightly radioactive. These changes involve a transfer of radioactivity from the larger to the smaller subunits and suggest that the 210,000-212,000 polypeptides are the precursors, about a third of which is cleaved in the fat body. The uncleaved precursors are then secreted into the hemolymph, along with their proteolytic products. Moreover, pulse experiments show that the female fat body synthesizes some other polypeptides, at M, 63,000, 225,000, and 280,000, which are not secreted into the medium. The very weak radioactivity of these bands (below 1%) prevents a clear insight into their possible role in the processing of the major vitellogenin. In the majority of insects, which fall into group one of Harnish and White [1,5], vitellogenin appears to be synthesized as a large precursor with a molecular weight generally over 200,000 . In Leucophuea maderue, for example, two large polypeptides of My179,000 and 260,000 have been suggested as the precursors of several smaller vitellogenins ranging from 57,000 to 118,000 However, different results have been obtained in the same cockroach species by Brookes , who considered that the large polypeptide (M, 250,000) is not a subunit of the vitellogenin, [la. Yolk Polypeptides in Thermobia domestica 31 1 although it precipitates with antiserum against the vitellin. In another recent study on Leucophaea maderae, Della-Cioppa and Engelmann [191, using pulselabeling of fat body with 32P, found a phospholabeled 215,000 polypeptide that precipitated with antiserum against vitellogenin. This pro-vitellogenin was proteolytically processed within the fat body, and the products were secreted into the hemolymph as mature vitellogenin subunits. In Locusfa migvatoria, two primary products at M, 225,000 and 235,000 have been identified by Chen  and considered as vitellogenin precursors, while the final polypeptides range from M, 52,000 to 126,000. According to Wyatt et al. , the primary translation product of L. migratoria fat body is a 200,000 peptide, whose glycosylation yields the two vitellogenic polypeptides of M, 225,000 and 235,000. In T. domesticu, a comparison of the different vitellogenic polypeptides from the hemolymph with those from mature ovaries reveals striking changes in their respective percentages: the 210,000-212,000 bands were much less abundant in the ovary (10 vs. 60%), the 176,000 and 182,000 bands much more abundant (65 vs. 40%), whereas polypeptides smaller than 60,000 reached a quarter of total vitellins. These findings indicate an additional processing phase of vitellogenins when they are sequestered by the ovaries. Ovarian processing has been considered in some other insect species, such as Leucuphaea maderae  or Blafella gemanica . In Locusfa migrabria, it has been suggested that the proteolytic cleavage of the vitellogenic proteins, which normally occurs in the fat body, may be continued in the hemolymph and in the ovary . A recent study on Aedes aegypti , comparing the peptide substructure of the vitellogenin and of the vitellin, also concludes that the final processing of the yolk proteins takes place in the ovary. In T. dornesticu, a similar step in the processing of the primary polypeptides occurs both in fat body and in ovary tissues. This situation appears to be unique among insects and therefore may have evolutionary significance. LITERATURE CITED 1. Harnish DG, White DN: Insect vitellins: Identification, purification and characterization from eight orders. J Exp Zoo1 220, 1 (1982). 2. Chen TT, Hillen LJ: Expression of the vitellogenin genes in insects. Gamete Res 7, 179 (1983). 3. Keeley LL: Physiology and biochemistry of the fat body. In: Comprehensive Insect Phys- iology, Biochemistry, and Pharmacology. Kerkut GA, Gilbert LI, eds. Pergamon Press, Oxford, Vol3, pp 211-248 (1985). 4. Kunkel JG, Nordin JH: Yolk proteins. In: ibid., vol 1, pp 83-111 (1985). 5. Harnish DG, White DN: An evolutionary model for the insect vitellins. J Mol Evol 18, 405 (1982). 6. Rousset A, Lapalisse J: Qualitative and quantitative changes in hemolymph proteins during an ovarian cycle and after insemination in Thermobiu dornesticu (Packard) (Thysanura, Lepismatidae). Arch Insect Biochem Physiol3, 277 (1986). 7. Rousset A, Bitsch C, Bitsch J: Vitellogenins in the firebrat, Thermobiu dornesticu (Packard): Immunological quantification during reproductive cycles and in relation to insemination (Thysanura, Lepismatidae). J Insect Physiol33, 593 (1987). 8. Bitsch C: Effects of ovarian maturation and insemination on the length of intermoult in Thermobiu dornesticu. Physiol Entomol 10, 15 (1985). 312 Rousset et al. 9. Bitsch C, Baehr JC, Bitsch J: Juvenile hormones in Themzobiu domesticu females: Identification and quantification during biological cycles and after precocene application. Experientia 41, 409 (1985). 10. Bradford MM: A rapid and sensitive method for the quantification of microgram quantities of progein utilising the principle of protein-dye binding. Anal Biochem 72, 248 (1976). 11. Ephrussi B, Beadle GW: A technique of transplantation for Drosophilu. Am Nat 70, 218 (1936). 12. Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 (1970). 13. Bownes M, Nothiger R: Sex determining genes and vitellogenin synthesis in Drosophilu melunoguster. Mol Gen Genet 182, 222 (1981). 14. Zhai QH, Postlethwait JH, Bodley JW: Vitellogenin synthesis in the lady beetle Coccinellu septernpunctutu. Insect Biochem 24, 299 (1984). 15. Chamberlain JP: Fluorographic detection of radioactivity in polyacrylamide gels with the water-soluble fluor, sodium salicylate. Anal Biochem 98, 132 (1979). 16. Harnish DG, Wyatt GR, White BN: Insect vitellins: Identification of primary products of translation. J Exp Zoo1 220, 11(1982). 17. Koeppe JK, Ofengand J: Juvenile hormone induced biosynthesis of vitellogenin in Leucophueu moderue from large precursor polypeptides. In: The Juvenile Hormones. Gilbert LI, ed. Plenum Press, New York, pp 486-504 (1976). 18. Brookes VJ: The polypeptide structure of vitellogenin and vitellin from the cockroach, Leucophuea maderue. Arch Insect Biochem Physiol3, 577 (1986). 19. Della-Cioppa G, Engelmann F: The vitellogenin of Leucophueu rnuderue. Synthesis as a large phosphorylated precursor. Insect Biochem 27, 401 (1987). 20. Chen 7T: Vitellogenin in locusts (Locustu rnigratoriu). Translation of vitellogenin mRNA in Xenopus oocytes and analysis of the polypeptides products. Arch Biochem Biophys 202, 266 (1980). 21. Wyatt GR, Locke J, Bradfield JY: The vitellogenin genes of Locustu migrutoriu and other insects. In: Advances in Invertebrate Reproduction. Engels W, ed., vol3, pp 73-80 (1984). 22. Wojchowski DM, Parsons P, Nordin JH, Kunkel JH: Processing of pro-vitellogenin in insects fat body: A role for high-mannose oligosaccharide. Dev Biol 176, 422 (1986). 23. Borovsky D, Whitney PL: Biosynthesis, purification, and characterization of Aedes uegypti vitellin and vitellogenin. Arch Insect Biochem Physiol4, 81 (1987).