THE JOURNAL OF EXPERIMENTAL ZOOLOGY 253~88-98(1990)

Differential Effects of Testosterone and Dibutyryl Cyclic AMP on the Meiotic Maturation of Mouse Oocytes In Vitro TOM W. ECAY AND R. DOUGLAS POWERS Department of Biology, Boston College, Chestnut Hill, Massachusetts 02167 ABSTRACT Fully grown, meiotically immature mouse oocytes were isolated and cultured under varying conditions with the aim of determining a) whether the inhibitory effects of testosterone on oocyte meiotic maturation require the synthesis of new oocyte proteins and b) if the meiosis-inhibiting effects of testosterone and dibutyryl cyclic AMP (dbcAMP) are distinct and can be differentiated. We found that the inclusion of puromycin in culture medium containing testosterone has no effect on the meiosis-inhibiting potency of testosterone or upon the reversibility of testosterone effects. We conclude that testosterone inhibits oocyte meiosis by a mechanism that is independent of protein synthesis. We also found that oocytes exposed to testosterone recover more rapidly, as evidenced by the timing of germinal vesicle breakdown (GVBD) following placement in a control medium, than do oocytes exposed to dbcAMP. Through further investigation of this phenomenon we have determined the sequence of testosterone and dbcAMP effects relative to the time course of GVBD. A testosterone-sensitive event occurs 20 min prior to GVBD, while the dbcAMP-sensitive event precedes GVBD by 41 min. The nature of this difference may involve the differential interaction of testosterone and dbcAMP with a set of puromycin-sensitive proteins that are required for GVBD. When oocytes were initially cultured in medium containing both puromycin and either testosterone or dbcAMP and then moved to medium containing puromycin alone the incidence of GVBD was reduced relative to oocytes never exposed to puromycin. This observation suggests that mouse oocytes contain proteins that are required for GVBD and that experience a high turnover rate. The degree of reduction in GVBD was a function of the length of puromycin exposure and was significantly greater in dbcAMP- than in testosterone-exposed oocytes. If oocytes were initially cultured in medium containing puromycin and dbcAMP, the rate of GVBD upon removal of dbcAMP was initially slow but increased with time. This observation is consistent with the hypothesis that dbcAMP inhibits oocytes at a point prior to the functioning of the puromycin-sensitive proteins. However, if oocytes were cultured in medium containing puromycin and testosterone the rate of GVBD following testosterone removal was not significantly reduced relative to oocytes that were not exposed to puromycin. This observation suggests that testosterone acts to inhibit meiosis at a site beyond the function of the puromycin-sensitive proteins or that testosterone causes a reduction in the turnover rate of these proteins. We conclude that the effects of testosterone and dbcAMP on the mouse oocytes are distinct and distinguishable and further that testosterone will be a useful tool in dissecting the molecular mechanisms controlling mouse oocyte meiotic maturation.

Shortly before birth, mammalian oocytes arrest development in early prophase of the first meiotic division. Meiotic arrest continues until sexual maturity when oocytes within mature follicles are stimulated to resume meiosis by the preovulatory surge of gonadotropins and possibly other agents of ovarian and follicular origin (Tsafriri, '78; Westergaard et al., '84; Downs et al., '88). In the absence of these stimuli, meiotic resumption can be initiated by placing fully grown oocytes into culture (Pincus and Enzmann, '35, Edwards, '65). Q 1990 WILEY-LISS, INC.

Meiotic resumption is marked morphologically by dissolution of the nuclear membrane (a process designated germinal vesicle breakdown or GVBD) and by extrusion of the first polar body, and this complete process is termed oocyte maturation (Masui and Clarke, '79). The phenomenon of spontaneous in vitro oocyte maturation has Received January 6, 1989; revision accepted June 28, 1989. Tom W. Ecay, Ph.D., is now at Department of Physiology and Cell Biology, University of Texas Medical School at Houston, P.O. Box 20708, Houston, TX 77225. Address reprint requests there.

TESTOSTERONE AND dbcAMP EFFECTS ON MOUSE OOCYTES

prompted the suggestion that the follicle exerts some inhibitory influence over the process of maturation. This is supported by the observations that oocytes become competent t o mature (i.e., will mature if removed from the ovary) long before the follicle has fully developed (Sorenson and Wasserman, '76) and that follicular fluid contains components that can inhibit maturation under the appropriate culture conditions (Tsafriri et al., '76; Downs et al., '85). Oocyte maturation is a complex biological process involving changes in metabolic, synthetic, and cytoskeletal activities as well as membrane and chromosome structure (reviewed in Masui and Clarke, '79). Because of the phenomenon of spontaneous in vitro maturation, mammalian oocytes are not readily amenable t o investigations of the meiosis-stimulating activity of various molecules. Investigations into the control of mammalian oocyte maturation have largely centered on the inhibition of spontaneous maturation by various physiologic and pharmacologic agents and the behavior of several in vitro model systems (Hillensjo et al., '79; Tsafriri et al., '76; Cho et al., '74; Smith and Tenney, '80.; Downs et al., '85; Eppig et al., '83). Membrane-permanent analogs of cAMP and agents that activate adenylate cyclase or inhibit phosphodiesterases inhibit oocyte maturation in several mammalian species (Cho et al., '74; Hillensjo, '77; Racowsky, '85; Rice and McGaughey, '81; Schultz et al., '83a,b). In the mouse, elevated oocyte cAMP alters the pattern of phosphorylated proteins and may maintain a maturation-stimulating protein in an inactive, phosphorylated form (Schultz et al., '83a; Bornslaeger et al., '86). Steroid hormones, and in particular testosterone, also inhibit the spontaneous maturation of mouse oocytes (Smith and Tenney, 'SO). Maturation inhibition by testosterone is dose dependent and reversible. Little work has been done on the effects of testosterone on mouse oocytes, and questions remain as t o the nature of testosterone inhibition and the possible interactions between the mechanisms of cAMP and testosterone inhibition. It has recently been suggested that testosterone and other steroids inhibit maturation by maintaining elevated cAMP levels within oocytes. This hypothesis is based on the observation that steroids inhibit a cAMP phosphodiesterase activity in oocyte extracts (Kaji et al., '87). However, measurements in intact oocytes have failed to demonstrate an effect of testosterone or other steroids on oocyte cAMP levels

89

(Schultz et al., '83a). Thus, the relationship between the mechanisms of testosterone and dbcAMP inhibition of mouse oocyte maturation remains unresolved. The objective of the investigations reported here was t o define possible differential effects of testosterone and dibutyryl cAMP (dbcAMP) on mouse oocyte maturation. We believe that an understanding of the effects of testosterone alone and as related t o dbcAMP effects will permit delineation of the various processes occurring in oocytes during the resumption and completion of meiotic maturation. Demonstration of these effects serves to define future investigations into the molecular events involved in the control of meiosis resumption.

MATERIALS AND METHODS Animals Swiss albino mice (CD-1)3-5 weeks of age were obtained from Charles River Breeding Laboratories, Inc., Wilmington, MA. The mice were kept at a constant room temperature of 20-24°C with controlled lighting (12 hr light:12 hr dark) and provided with food and water ad libitum. In general oocytes were obtained from randomly cycling mice 6-12 weeks old; however, we have found that 8-10 week old mice give the highest yield of oocytes and mice of this age range were used preferentially. In other studies (e.g., Eppig et al., '83; Schultz et al., '83a) hormonally primed mice have been used as a source of immature oocytes. In our hands, hormonally primed and unprimed mice produce oocytes with similar sensitivities t o steroid hormones (Barrett et al., 'S9), and unprimed mice were used for the present study. Culture medium Our standard egg culture medium (SECM) used throughout this study is a modification of Biggers ('71) medium for mammalian eggs and embryos. It contains 90.0 mM NaCl, 4.6 mM KCl, 1.71 mM Caelactate, 1.19 mM KH2P04, 1.19 mM MgS04, 25.0 mM NaHC03, 0.5 mM Na.pyruvate, 21.4 mM Na-lactate, 5.55 mM glucose, 4.0 g/liter crystalline bovine serum albumin (BSA), and 20.0 mg/ liter phenol red. SECM is equilibrated in an atmosphere of 5% C02 in air at 37°C for at least 6 hr before being adjusted to a pH of 7.4 with 1N HCl. It is then sterilized by filtration in a 0.45 km Nalgene filter unit and stored at 4°C. Medium dispensed to tissue culture dishes is equilibrated at

90

T.W.ECAY AND R.D. POWERS

37"C, 100%humidity, 5% C 0 2 in air for 2 hr prior t o use. The SECM is used within 3 days of preparation. When needed, 100 pg/ml dbcAMP was dissolved directly into SECM. This concentration of dbcAMP was used throughout this study and inhibits GVBD by greater than 98% (Cho et al., '74). Puromycin at 100 pg/ml was also dissolved directly into SECM. A 10 mM stock solution of testosterone in 100% ethanol was stored at -20°C. Dilutions of this stock were made directly into culture medium when needed. Unless otherwise stated testosterone was used at a concentration of 80 pM. Controls containing 1%ethanol did not differ from ethanol-free controls in any parameter of meiotic maturation measured in this study. In experiments comparing the effects of dbcAMP and testosterone on oocyte meiotic maturation the addition of 1%ethanol t o dbcAMP cultures did not alter the effect of dbcAMP on any measured parameter of oocyte maturation.

Collection, culture, and scoring of oocytes Mice were killed by cervical dislocation, and their ovaries were promptly removed. Four to six ovaries were used per experimental replicate depending on the number of oocytes required. With the aid of a dissecting microscope (Wild) oocytes were liberated by puncturing follicles with a 30 gauge needle. Cumulus-free oocytes and those with loosely adhering cumulus cells were collected with a mouth-operated pipet. Only those oocytes showing no signs of degeneration, in the size range of 75-85 pm, and completely free of cumulus cells after three washes were kept for further culture. Unless otherwise stated, oocytes were collected in SECM containing the additions in which they were t o be cultured. All oocytes were cultured in 3 ml of medium in Falcon 3008 dishes in a COz incubator at 37°C. At the end of an experiment oocytes were examined with a dissecting microscope (80 x magnification) and scored as immature (GV) if they contained an intact GV, as first metaphase (Ml) if GVBD had occurred but a polar body was not present, or as matured (PB) if a polar body had formed. Any oocytes showing signs of cytolysis, fragmentation, or nonspherical shape were considered necrotic and were subtracted from the total number of oocytes when determining the percentage of oocytes at a stage of maturation. In some experiments the process of GVBD was followed by inspecting oocytes at relatively short intervals (15, 30, or 60 min depending on appro-

I.

Control

100 pg/ml Purornycin

-

estosterone + Puromycin

Fig. 1. The effect of puromycin on the inhibition of GVBD by testosterone. Oocytes were cultured for 18 hr in SECM containing 80 pM testosterone or 100 pgiml puromycin as indicated. Oocytes were then scored for their stage of maturation. Values are expressed as the number of oocytes at a particular stage of maturation as a percentage of viable oocytes at the end of culture. Viability generally exceeded 98% in all replicates. Shown are the mean and SEM of four replicate experiments, each culture condition replicate containing at least 30 oocytes.

priateness). Oocytes were considered to have undergone GVBD when no sign of the nucleolus and the nuclear membrane could be distinguished with a dissecting microscope.

RESULTS Puromycin effects on testosterone inhibition of GVBD In our control SECM, a culture period of 18 hr permits the majority of fully grown mouse oocytes to complete meiotic maturation as evidenced by the presence of a PB. Nearly all of the fully grown oocytes placed into culture are competent t o resume meiosis, as few contain an intact GV at the end of culture (Fig. 1). The addition of testosterone (80 pM) t o SECM results in a block t o meiosis resumption, and nearly all oocytes cultured with testosterone retain an intact GV after 18 hr in culture (Fig. 1).The inhibitory effect of testosterone on the spontaneous in vitro maturation of mouse oocytes is completely reversible as most oocytes (>go%) will undergo GVBD within 3 hr of removal from SECM containing testosterone (Fig. 31, and if cultured for 16-20 hr most will form a PB (not shown). Our results confirm the similar observations reported by Smith and Tenney ('80) with one notable exception. We observe that 80 pM testosterone does not affect oocyte viability (Fig. l)while Smith and Tenney report that 33% of oocvtes cultured with 80 uM testosterone remain viable at the end of culture. We suggest

91

TESTOSTERONE AND dbcAMP EFFECTS ON MOUSE OOCYTES

two explanations t o account for this difference. First, the BSA in our culture medium may promote greater viability than the polyvinylpyrrolidone used by Smith and Tenney. Second, we obtained higher viability under all culture conditions (generally L 95%) than Smith and Tenney who report a range of 50-94% viability in control cultures. This may indicate that we have been more discrimination in selecting healthy oocytes at the initiation of culture. The most widely accepted mechanism by which steroid hormones evoke a response in target cells involves an interaction of the hormone-receptor complex with chromosomal elements and the subsequent synthesis of novel protein species (O'Malley and Means, '74). To determine whether the meiosis-inhibiting effect of testosterone is mediated by or dependent upon the synthesis of new oocyte proteins, we investigated the effects of the protein synthesis inhibitor puromycin on oocyte maturation in the presence and absence of testosterone. When cultured in SECM containing 100 kg/ml puromycin, oocytes reinitiate meiosis (as evidenced by GVBD) in numbers similar to oocytes cultured in control SECM (Fig. 1).However, puromycin inhibits PB extrusion, and oocytes remain blocked at the M1 stage of maturation. Similar effects were observed at puromycin concentrations as low as 1 pg/ml. Since 100 Kg/ml puromycin has been shown t o block oocyte protein synthesis completely (299%; Clark and Masui, '83) and rapidly (within 20 min; Wassarman et al., '76) these results suggest that upon isolation from the ovary, fully grown oocytes contain all of the proteins required to resume meiotic activity and that no new novel protein synthesis is required to complete GVBD. Similar observations and conclusions have been reported by others (Wassarman et al., '76; Clarke and Masui, '83). If the inhibitory effect of testosterone is mediated by a hormone-evoked synthesis of new proteins, then inhibition of protein synthesis by the addition of puromycin to the culture system should remove the testosterone block t o meiosis resumption. Oocytes so cultured would be expected to undergo GVBD but t o remain at the M1 stage due to the puromycin block to PB formation. As is shown in Figure 1, however, oocytes cultured in SECM containing both testosterone and puromycin remain blocked at the GV stage. During an 18 hr culture period, the inclusion of puromycin produced no lessening of the inhibitory potency of testosterone.

Imt

0 Testosterone dbcAMP

20

4

4

6

12

18

Duration of Preculture (hr.) Fig. 2. The time-delayed effect of puromycin on GVBD. Oocytes were isolated and cultured in SECM containing 100 pgiml puromycin and either 80 pM testosterone or 100 pg/ml dbcAMP. After a culture period of the length shown, oocytes were washed free of testosterone or dbcAMP and cultured in SECM containing puromycin. Oocytes were scored for GVBD 12 hr later. Shown are the means of two or three replicate experiments, and the total number of oocytes cultured is indicated.

The results of Figure 1 suggest that the action of testosterone is not dependent upon the synthesis of new oocyte proteins. However, Ekholm and Magnusson ('79) have reported that puromycin alone has an inhibitory effect on rat oocyte GVBD. This effect of puromycin was only observed after a short (4 hr) preculture of oocytes in medium containing both puromycin and dbcAMP. Upon the removal of dbcAMP, rat oocytes remained blocked at the GV stage. With this observation in mind it is plausible t o suggest that a rapid synthesis of proteins under the control of testosterone, occurring prior to the maximal effects of puromycin, could be of sufficient magnitude t o inhibit GVBD long enough t o allow the more delayed effect of puromycin to become evident. To test this possibility we cultured oocytes in SECM containing both testosterone and puromycin for varying periods. We then transferred these oocytes to SECM containing only puromycin and scored for GVBD after 12 hr. In these experiments, inhibition of GVBD significantly greater than in control cultures (i.e., no testosterone preculture) was only observed when the exposure to puromycin during preculture was 12 hr or longer (Fig. 2). Complete inhibition of GVBD required a puromycin preculture of 18 hr.

92

T.W. ECAY AND R.D. POWERS

]: l

n m 60>

; 8 5

n

40-

20 -

W 0 80 pM Testosterone 0 0 100 pgml dbcAMP W 0 100 pglml Puromycin

30

60

90

120

150

180

Time (min.)

Fig. 3. The time course of oocyte recovery from culture in dbcAMP and testosterone. Oocytes were cultured for 18 hr in SECM containing 100 pg/ml dbcAMP or 80 pM testosterone. Oocytes were then washed through three changes of SECM. Some oocytes were then placed in SECM containing 100 pg/ ml puromycin. Oocytes were scored for GVBD at 30 min intervals. The percent of oocytes having undergone GVBD is shown as a function of the time after removal from dbcAMP or testosterone. Each point is the mean and SEM of four replicate experiments, each replicate containing at least 40 oocytes.

Mouse oocytes are, therefore, sensitive to a delayed effect of puromycin on GVBD; however, this delayed effect develops slowly in the mouse oocyte. Together with the results of Figure 1, the observation of a slowly developing effect of puromycin on GVBD supports the conclusion that the inhibitory effect of testosterone on oocyte maturation occurs by a mechanism that is independent of the synthesis of new oocyte proteins. Experiments were conducted in parallel to those just described except that dbcAMP was substituted for testosterone during preculture (Fig. 2). Relative to testosterone, a time-dependent effect of puromycin on GVBD developed more rapidly in oocytes precultured with dbcAMP.

Oocyte recovery from testosterone and dbcAMP The in vitro inhibition of mouse oocyte maturation by both testosterone and dbcAMP has been shown to be reversible (Smith and Tenney, '80; Cho et al., '74). Reversibility has generally been demonstrated by determining the percentage of M1 or PB oocytes at some time (6-24 hr) following a n initial exposure of oocytes to testosterone or dbcAMP. To further investigate the process of oocyte recovery from testosterone and dbcAMP we have determined the time course of GVBD fol-

lowing oocyte removal from SECM containing these agents. These results are shown in Figure 3. Similar results were obtained when the initial culture period (containing testosterone or dbcAMP) was varied from 6 to 24 hr (not shown). The time course of oocyte recovery from testosterone and dbcAMP, as illustrated in Figure 3, demonstrates several properties of the effects of these agents on mouse oocytes. First, GVBD is completed by almost all oocytes within 180 min of their removal from testosterone or dbcAMP. A similar timing of GVBD has been observed for oocytes upon removal from the ovary (Paleos and Powers, '81; Powers '82). In addition, the curves in Figure 3 are displaced from each other along the time axis. The position of each curve can be compared quantitatively by the GVBD,, value (the time at which 50% of the oocytes are scored as GVBD). For the results shown in Figure 3, GVBDBovalues are 67 & 3 and 100 & 3 min for testosterone- and dbcAMP-exposed oocytes, respectively. Similar differences in GVBDSo were observed for oocytes cultured in SECM containing testosterone or dbcAMP for as short as 6 hr (not shown). For oocytes exposed to dbcAMP for 18 hr, a GVBDBoof 100 min is similar to that reported for oocytes following isolation from the ovary (GVBD,, = 90 min; Paleos and Powers, '81; Powers '82) and for oocytes recovering from 7.5 hr of dbcAMP inhibition (GVBDBo = 84 min; Clarke and Masui '85). For oocytes exposed to testosterone for 18 hr a GVBD50 of 67 rnin implies that oocytes recover more quickly from 18 hr of testosterone inhibition than from removal from the follicle. The rapid and near-complete recovery of oocytes maintained at the GV stage for 18 hr further demonstrates the nontoxic nature of the in vitro effects of testosterone and dbcAMP in this system. The results illustrated in Figure 3 and the differences noted in GVBD,, suggest the possibility that the sites of testosterone and dbcAMP action are temporally distinct within the time course of GVBD. The following experiments were designed to further define the nature of the observed differences in GVBD,, values. Oocytes were cultured for 18 hr in SECM containing either testosterone or dbcAMP. After this initial culture period, each group was washed through several changes of fresh SECM and placed into SECM for varying periods (0-60 min). Following this extended wash period each group of oocytes was split, half being placed into SECM containing testosterone and half into SECM containing dbcAMP. Subgroups

TESTOSTERONE AND dbcAMP EFFECTS ON MOUSE OOCYTES 100

B

I

80

93

a

h

8

-x6 0 .-c .-> c .-

E40 Q,

v)

20

45

60

20

30

40

Length of Wash (min.)

Fig. 4. The loss of oocyte sensitivity t o testosterone or dbcAMP as a function of the length of washout. Groups of 60100 oocytes were cultured for 18 hr in SECM containing either 100 pg/ml dbcAMP (A) or 80 pM testosterone (B). 00cytes were then washed free of inhibitor and cultured in control SECM for the indicated times. Each group of oocytes was then split, half being placed into medium containing testos-

terone (open bars) and half into medium containing dbcAMP (hatched bars). Oocytes were scored for GVBD at 15 rnin intervals. Sensitivity is defined as the percentage of oocytes that retained a GV by 180 min after the end of the wash period. Shown are the means of four replicate experiments (a and b are significantly different at P 5 .01).

We have used the data from Figures 3 and 4 to define the site of testosterone and dbcAMP action with respect to the timing of GVBD. This has Initial culture Secondary Length of culture wash (rnin) Testosterone dbcAMP been done in the following manner. Figure 4A illustrates that for oocytes exposed to dbcAMP, Testosterone 20 22" 40" washed for 45 min, and then reexposed t o 30 15 42 dbcAMP, 35% became insensitive to the second 45 19 43 exposure to dbcAMP (65% remained sensitive). dbcAMP 30 33 The insensitive oocytes had become committed to 45 21 39 60 24 50 undergo GVBD, at least as regards dbcAMP efAve. (min) 2 SEM 20 * 2.2 41 t 2.3 fects, during the 45 min wash. Reading off the "See text for an explanation of these values. dbcAMP curve in Figure 3, the time corresponding to 35% GVBD is 84 min. This time of 84 were then scored for GVBD at 15 min intervals. min includes the 45 min wash period and 39 min During the second exposure to maturation in- of dbcAMP exposure. In these oocytes, 39 min of hibitors, any oocytes scored as GVBD by 180 min dbcAMP exposure was sufficient to inhibit GVBD postwashing were termed insensitive t o that in- in 65% of the oocytes yet was insufficient to effect hibitor. the remaining 35% of oocytes. The time point of When the wash period was short (i.e., 5 10 min) dbcAMP action is then 39 min prior to GVBD. all oocytes remained sensitive to either dbcAMP Continuing this analysis for all of the data shown or testosterone independent of the inhibitor used in Figure 4 yields the results shown in Table 1. during the initial culture period. As the wash Under the conditions of our experiments, the tesperiod was lengthened, oocytes consistently lost tosterone-sensitive event precedes GVBD by 20 sensitivity t o dbcAMP more rapidly than t o tes- min, while the dbcAMP-sensitive event occurs 41 tosterone irrespective of which inhibitor min prior to GVBD.l (testosterone or dbcAMP) was present during the initial culture period (Fig. 4). The delayed loss of We consider these time points to be reflective of the relative temporal proximity of the testosterone- and dbcAMP-sensitive events in the sensitivity to testosterone relative to dbcAMP is pathway to GVBD. These time points should not necessarily be consistent with the hypothesis that the site of tes- viewed as the absolute positions of these events within the time of meiotic maturation as the timing of the GVBD process is in tosterone action is temporally more proximal to course part a function of the subjective judgment of the individual observer GVBD. scoring oocytes for GVBD. TABLE 1. The timing of testosterone and dbcAMP effects on mouse oocvte GVBD

94

T.W. ECAY AND R.D. POWERS

Another explanation for the observed difference in the timing of testosterone and dbcAMP action 00 dkAMP + Purornycin is that these two compounds are metabolized or 80 scondaw Culture washed out of oocytes at different rates. Informa.0 Puromycin 00 Control tion contained within Figure 3 argues against a g 60 6 0 1 significant difference in the rate of washout or metabolism. The rate of GVBD as represented by 5 the slope of the linear portion of each curve in Figure 3 is the same for testosterone (1.4% a GVBDimin) as for dbcAMP (1.3% GVBDlmin) 20 cultured oocytes. Assuming that the rate of progression through the events that follow the testosterone- and dbcAMP-sensitive steps and that lead M) 120 180 240 300 360 720 to GVBD is uniform and unaffected by GVBD inTime (min.) hibition, then the similar rates of GVBD implies that oocytes are released from the testosterone Fig. 5. The effect of puromycin on the time course of and dbcAMP block to GVBD at similar rates (i.e., GVBD following exposure to testosterone or dbcAMP. 00metabolism and wash out of testosterone and cytes were cultured in SECM containing 100 p,g/ml puromycin and either 80 p,M testosterone or 100 pg/ml dbcAMP for dbcAMP occur at similar rates). 12 hr. Oocytes were then washed and placed into control SECM or SECM containing puromycin and scored for GVBD The effect of puromycin on the time at 30-60 min intervals. Shown are the mean of two replicate course of GVBD experiments. The numbers in parentheses are the total numPuromycin has a delayed effect on GVBD, and ber of oocytes scored. the development of this effect is more rapid in dbcAMP-exposed than in testosterone-exposed initial GVBD events occurred at a relatively high oocytes (Fig. 2). It is possible that the differential rate (70% GVBD at 6 hr). These observations are effect of puromycin on dbcAMP- and testosterone- consistent with the results of Figure 2 and supcultured oocytes is the result of the demonstrated port the suggestion that puromycin sensitivity detemporal difference in testosterone and dbcAMP velops differentially in the presence of dbcAMP or sites of action. To investigate this question we testosterone. determined the effect of puromycin on the time As is shown in Figure 2, a 12 hr exposure to course of GVBD for oocytes exposed to testos- puromycin significantly reduces the proportion of terone and to dbcAMP. In initial experiments we both dbcAMP- and testosterone-precultured ooobserved that puromycin caused a small increase cytes capable of resuming meiosis in the presence in GVBDBo (15% for both testosterone and of puromycin. As expected, the rate of GVBD for dbcAMP) but did not effect the slopes of the time dbcAMP-precultured oocytes, placed into SECM course curves (Fig. 3). containing puromycin, was very low (10% by 12 A second set of experiments was designed to hr in Fig. 5). In contrast, testosterone-precultured determine the effect of a prolonged exposure to oocytes underwent GVBD in the presence of puromycin and testosterone or dbcAMP on the puromycin at a rate similar to oocytes placed in rate of GVBD (Fig. 5 ) . Oocytes cultured in SECM control SECM following preculture. containing dbcAMP and puromycin for 12 hr and DISCUSSION then maintained in control SECM underwent Since the initial observation that testosterone nearly complete GVBD by the end of the scoring period (78% GVBD at 12 hr). For these oocytes the is an effective and specific inhibitor of meiotic rate of GVBD was initially low (26% GVBD at 6 maturation in denuded mouse oocytes (Smith and hr) but then increased (52% over the final 6 hr). Tenney, '80) there have been few reports to eluciThis pattern of GVBD is consistent with the hy- date the mechanism by which testosterone acts in pothesis that puromycin-sensitive proteins must this system. The most generally accepted model of be resynthesized by oocytes prior to GVBD. 00- steroid hormone action involves the synthesis of cytes precultured in SECM containing testos- novel protein species under the direction of the terone and puromycin also underwent nearly receptor-hormone complex (O'Malley and Means, complete GVBD after 12 hr in control SECM (93% '74). These new proteins then alter cell behavior GVBD by 12 hr). In these groups of oocytes the and mediate the hormonal response. Several lines

TESTOSTERONE AND dbcAMP EFFECTS ON MOUSE OOCYTES

of evidence presented here serve to discount this model as accounting for the testosterone effect on the in vitro resumption of mouse oocyte meiosis. First, “classic” steroid effects are generally preceded by a lag period of 1 to several hours. This lag is the time in which steroid evoked protein synthesis occurs. We have identified a testosterone sensitive event that precedes GVBD by 20 min. I n addition, the effect of testosterone on mouse oocyte GVBD is rapidly reversible (GVBDBo= 67 m i d . In other systems, steroid effects mediated by newly synthesized proteins can be blocked by inhibitors of protein synthesis. However, inclusion of puromycin in our culture system did not lessen the inhibitory potency of testosterone. Therefore, we conclude that the fully grown mouse oocyte contains all of the proteins required to respond to testosterone and that any novel oocyte proteins synthesized as a result of testosterone exposure do not contribute to the process of maturation inhibition at the level of the control of GVBD. Furthermore, since oocytes in the continued presence of puromycin (for as long as 12 hr) are capable of undergoing GVBD upon the removal of testosterone (Figs. 2, 5), the mouse oocyte must contain all of the proteins necessary to recover from the meiosis-inhibiting effects of testosterone. The puromycin insensitivity of the testosterone effect suggests that in the control of GVBD, the site of action of testosterone is not at the level of the genome. Because testosterone does not inhibit the resumption of meiosis by acting through the “ c l a ~ s i c a steroid l~~ hormone mechanism, alternative mechanisms must be investigated. One possible site of testosterone action is a plasma membrane receptor. The existence of a surface receptor for progesterone in Xenopus oocytes has been well established (Sadler and Maller, ’81). Recently a glucocorticoid receptor-like antigen has been localized to the surface membrane of S-49 mouse lymphoma cells, and this putative receptor has been implicated as a mediator of glucocorticoid effects in this cell line (Gametchu, ’87). Also, steroids with structural similarities to testosterone and progesterone are known modulators of GABA-A receptors (Harrison et al., ’87), A general phenomenon of steroid action in many other systems is a permissive effect on CAMP-mediated hormonal responses. Indeed, low concentrations of testosterone have been observed to enhance the inhibitory action of elevated cAMP on the in vitro maturation of mouse (Eppig et al., ’83) and pig oocytes (Rice and McGaughey,

95

’81).In these instances, it has been suggested that testosterone enhances some event in the pathway from elevated cAMP to GVBD inhibition. Recently it has been reported that the activity of cAMP phosphodiesterases in oocyte extracts are partially inhibited by testosterone and other steroids (Kaji et al., ’871, suggesting that steroids elevate oocyte CAMP. However, the physiological relevance of these observations to intact oocytes remains unclear since direct measurements of oocyte cAMP levels have not demonstrated a significant difference in cAMP in the presence or absence of testosterone concentrations that markedly inhibit meiotic maturation (Schultz et al., ’83a). Whatever the nature of the testosterone enhancement of cAMP effects, it is independent of a direct testosterone effect upon oocyte CAMP levels, as testosterone does not alter the sensitivity of adenylate cyclase to exogenous stimulation (Schultz et al., ’83a). It may be hypothesized that testosterone mimics, at the molecular or physiologic level, the effects of dbcAMP or elevated oocyte cAMP on the resumption of meiosis. This possibility is unlikely for several reasons. First, the cellular effects of cAMP are thought to be mediated solely by the activity of CAMP-dependent protein kinases, and we know of no report demonstrating a direct interaction between steroids and protein kinases. In addition, we have demonstrated here a distinct temporal difference in the site of testosterone and dbcAMP action. Our observations that dbcAMP but not testosterone increases the rate of calcium efflux from GV stage oocytes (Ecay and Powers, ’85) (Ecay and Powers, in preparation), that lithium ions reverse testosterone but not dbcAMP inhibition of GVBD (Ecay and Powers, ’84) (Ecay and Powers, submitted), and that there is a delayed effect of puromycin on GVBD that develops more rapidly in dbcAMP-exposed oocytes (this paper) further support the assertion that the observed temporal difference is the result of a mechanistic difference between testosterone and been investigated. The inhibition of mouse oocyte GVBD by CAMP is mediated by a CAMP-dependent protein kinase (PK-A), which is hypothesized to maintain a stimulator protein in a phosphorylated and inactive form (Bornslaeger et al. , ’86). Microinjection of immature oocytes with the catalytic subunit of PK-A inhibits GVBD in the absence of elevated CAMP, while microinjection of a protein kinase inhibitor (PKI) stimulates GVBD in the presence of CAMP. Subsequent to PKI injection or during

96

T.W. ECAY AND R.D.POWERS

the period when oocyte cAMP levels are declining, a 60 kD protein is dephosphorylated (Bornslaeger et al., '86). This 60 kD protein is then a substrate for PK-A. Coincident with the dephosphorylation of the 60 kD protein is the phosphorylation of other proteins by CAMPindependent protein kinases. Whether the 60 kD protein is itself a protein kinase or a protein kinase regulator and what role this protein has in the subsequent events leading to GVBD has not yet been determined. The effect of testosterone on the pattern of phosphorylations and dephosphorylations that accompany oocyte maturation has not been investigated. A consequence of meiotic resumption is the appearance of a maturation-promoting factor (MPF) in the cytoplasm of maturing oocytes. MPF has been demonstrated in maturing mammalian (Sorensen et al., '851, amphibian (Masui and Markert, '7'1)' and starfish (Kishimoto and Kanatani, '76) oocytes as well as in mitotic yeast (Weintraub et al., '82) and cultured mammalian cells (Sunkara et al., '79). In oocytes and early embryos, MPF activity cycles in synchrony with the onset of metaphase (Gerhart et al., '84). MPF has been hypothesized t o be a universal regulator of nuclear membrane dissolution and chromatin condensation. In mouse oocytes the site of cAMP action is more distal to GVBD than the site of MPF action, and cAMP blocks the formation of MPF in immature oocytes. Cytoplasmic extracts of mature mouse oocytes promote GVBD when injected into immature Xenopus oocytes. Similar extracts from GV stage mouse oocytes maintained in dbcAMP do not promote GVBD following injection into Xenopus oocytes (Sorensen et al., '85). When mature mouse oocytes are fused with immature oocytes, GVBD occurs in the immature fusion partner in the presence of dbcAMP (Fulka et al., '88). These experimental systems can be used to test the relationship between testosterone inhibition and MPF action. If testosterone inhibits GVBD at a point prior to the action of MPF, we predict results similar to those cited above for dbcAMP. However, if testosterone inhibits maturation at a point beyond MPF we predict that the cytoplasm of testosterone-inhibited oocytes will promote GVBD in Xenopus oocytes and that the fusion of a mature and immature mouse oocyte in the presence of testosterone will not result in GVBD. MPF is present in immature Xenopus and starfish oocytes in an inactive form (preMPF).Experiments involving the serial dilution of oocyte

cytoplasm have demonstrated that the conversion of preMPF to MPF in these oocytes is an autocatalytic process (Wasserman and Masui, '75; Doree, '82). The initial activation of MPF in progesterone-stimulated Xenopus oocytes is dependent upon protein synthesis as cyclohexamide blocks GVBD in progesterone-stimulated Xenopus oocytes. However, the autocatalytic activation of MPF is independent of protein synthesis, as cyclohexamide does not block GVBD in oocytes injected with MPF. These results suggest that an initiator of MPF activation is sensitive t o protein synthesis inhibition. These results do not determine whether the MPF-initiator is synthesized de novo prior to MPF activation or whether it experiences a high rate of turnover. Mouse and rat oocytes undergo GVBD within 3 hr of removal from the ovary. In contrast, sheep and pig oocytes require an extended culture period prior to GVBD (8 hr, Moor and Crosby, '86; 16 hr, Fulka et al., '86). When protein synthesis is inhibited, mouse and rat oocytes progress beyond GVBD (Hillensjo, '77; Clarke and Masui, '83; Ecay and Powers, this paper), while sheep and pig oocytes remain at the GV stage (Moor and Crosby, '86; Fulka et al., '86). These results indicate that mouse and rat oocytes contain preMPF and the MPF-initiator, while sheep and pig oocytes must synthesize one or both of these proteins. By analogy to the Xenopus oocyte, we assume that it is the MPF-initiator that is protein synthesis dependent. In the sheep oocyte the synthesis of a novel 46 kD protein has been detected 1 hr prior t o GVBD (Moor and Crosby, '86). The role of this protein in the control of GVBD and its relationship t o the MPF-initiator and MPF are unknown. If mouse and rat oocytes are maintained at the GV stage by dbcAMP and protein synthesis is inhibited, the incidence of GVBD is reduced following dbcAMP removal (Ekholm and Magnusson, '79; Ecay and Powers, this paper). This indicates that mouse and rat oocytes contain proteins that regulate GVBD and that experience a high rate of turnover. We have demonstrated here that in the presence of testosterone mouse oocytes are relatively insensitive to protein synthesis inhibition. At this time we do not know whether this is due to a testosterone site of action, which is beyond the protein synthesis-dependent step or whether testosterone reduces the turnover rate of these proteins. If the latter case is true then testosterone may prove to be a useful tool in the isolation and identification of these proteins.

TESTOSTERONE AND dbcAMP EFFECTS ON MOUSE OOCYTES

In conclusion, we have shown here that testosterone and dbcAMP have distinct effects on the in vitro resumption of meiosis in mouse oocytes. The sites of action of dbcAMP and testosterone can be resolved temporally, and they are separated by a protein synthesis-dependent event. These observations will be useful in the course of dissecting the process of mammalian oocyte maturation at the biochemical and molecular levels.

ACKNOWLEDGMENTS This study was supported in part by NIH grant 16062.

LITERATURE CITED Barrett, C.B., E. Newber, and R.D. Powers (1989) Specificity of steroid inhibition of murine oocyte meiotic maturation in vitro. Biol. Reprod., in press. Biggers, J.D. (1971) New observations on the nutritional requirements of the mammalian oocyte and the preimplantation embryo. In: The Biology of the Blastocyst. R.J. Blandau, ed. Univ. of Chicago Press, Chicago, pp. 319-328. Bornslaeger, E.A., P. Mattei, and R.M. Schultz (1986) Involvement of CAMP-dependent protein kinase and protein phosphorylation in regulation of mouse oocyte maturation. Dev. Biol., 114:453-462. Cho, K.C., S. Stern, and J.D. Biggers (1974) Inhibitory effect of dibutyryl cAMP on mouse oocyte maturation in vitro. J. Exp. Zool., 187:383-386. Clarke, H.J., and Y. Masui (1983) The induction of reversible and irreversible chromosome decondensation by protein synthesis inhibition during meiotic maturation of mouse oocytes. Dev. Biol., 97:291-301. Clarke, H.J., and Y. Masui (1985) Inhibition by dibutyryl cyclic AMP of the transition to metaphase of mouse oocyte nuclei and its reversal by cell fusion to metaphase oocytes. Dev. Biol., 108:32:-37. Doree, M. (1982) Protein synthesis is not involved in initiation or amplification of the maturation-promoting factor (MPF) in starfish oocytes. Exp. Cell Res., 139:127-133. Downs, S.M., D.L. Coleman, P.F. Ward-Bailey, and J.J. Eppig (1985) Hypoxanthine is the principle inhibitor of murine oocyte maturation in a low molecular weight fraction of porcine follicular fluid. Proc. Natl. Acad. Sci. U.S.A., 82: 454-458. Downs, S.M., S.A.J. Daniel, and J.J. Eppig (1988) Induction of maturation in cumulus-cell enclosed mouse oocytes by follicle stimulating hormone and epidermal growth factor: Evidence for a positive stimulus of somatic cell origin. J. Exp. Zool., 245:86-96. Ecay, T.W., and R.D. Powers (1984) Possible calcium components in the inhibition of GVBD in the mammalian oocyte. J . Cell Biol., 99:55 (Abstract). Ecay, T.W., and R.D. Powers (1985) Changes in phosphatidylinositol metabolism may control germinal vesicle breakdown in the mouse oocyte. J. Cell Biol., 101:263 (Abstract). Edwards. R.G. (1965) Maturation in uitro of mouse, sheep, cow, pig, rhesus monkey, and human ovarian oocytes. Nature, 208:349-351.

97

Ekholm, C., and C. Magnusson (1979) Rat oocyte maturation: Effects of protein synthesis inhibitors. Biol. Reprod., 21: 1287-1293. Eppig, J.J., R. Freter, P.F. Ward-Bailey, and R.M. Schultz (1983) Inhibition of oocyte maturation in the mouse: Participation of CAMP,steroid hormones, and a putative maturation-inhibitory factor. Dev. Biol., 100:39-49. Fulka, J., Jr., J . Motlik, J. Fulka, and F. Jilek (1986) Effect of cyclohexamide on nuclear maturation of pig and mouse oocytes. J. Reprod. Fertil., 77:281-285. Fulka, J., Jr., J.-E. Flechon, J . Motlik, and J . Fulka (1988) Does autocatalytic amplification of maturation-promoting factor (MPF) exist in mammalian oocytes? Gamete Res., 21:185-192. Gametchu, B. (1987) Glucocorticoid receptor-like antigen in lymphoma cell membranes: Correlation to cell lysis. Science, 236:456-461. Gerhart, J., M. Wu, and M. Kirschner (1984) Cell cycle dynamics of an M-phase-specific cytoplasmic factor in Xenopus Zaeuis oocytes and eggs. J. Cell Biol., 98:1247-1255. Harrison, N.L., M.D. Majewska, J.W. Harrington, and J.L. Barker (1987) Structure-activity relationship for steroid interaction with the gamma-aminobutyric acid A receptor complex. J . Pharmacol. Exp. Ther., 241:346-353. Hillensjo, T. (1977) Dissociation of preovulatory maturational events in rat oocytes and cumuli in the presence of dibutyryl cyclic AMP. Acta Physiol. Scand., 100:261-263. Hillensjo, T., C.P. Channing, S.H. Pomerantz, and A. Schwartz-Kriporer (1979) Intrafollicular control of oocyte maturation in the pig. In Vitro, 15:39. Kaji, E., E.A. Bornslaeger, and R.M. Schultz (1987) Inhibition of mouse oocyte cyclic AMP phosphodiesterase by steroid hormones: A possible mechanism for steroid hormone inhibition of oocyte maturation. J . Exp. Zool., 243:489-493. Kishimoto, T., and H. Kanatani (1976) Cytoplasmic factor responsible for germinal vesicle breakdown and meiotic maturation in starfish oocytes. Nature, 260:321, 322. Masui, Y., and C.L. Markert (1971) Cytoplasmic control of nuclear behavior during meiotic maturation of frog oocytes. J. Exp. Zool., 177:129-146. Masui, Y., and H.J. Clarke (1979) Oocyte maturation. Int. Rev. Cytol., 57~185-282. Moor, R.M., and I.M. Crosby (1986) Protein requirements for germinal vesicle breakdown in ovine oocytes. J. Embryol. Exp. Morphol., 94:207-220. O’Malley, B.W., and A.R. Means (1974) Female steroid hormones and target cell nuclei. Science, 183:610-620. Paleos, G.A., and R.D. Powers (1981) The effect of calcium on the first meiotic division of the mammalian oocyte. J. Exp. Zool., 217:409-416. Pincus, G., and E.V. Enzmann (1935) The comparative behavior of mammalian eggs in uivo and in vitro. I. The activation of ovarian eggs. J. Exp. Med., 62:665-675. Powers, R.D. (1982) Changes in mouse oocyte membrane potential and permeability during meiotic maturation. J . Exp. Zool., 221:365-371. Racowsky, C. (1985) Effect of forskolin on the spontaneous maturation and cyclic AMP content of hamster oocytecumulus complexes. J . Exp. Zool., 234:87-96. Rice, C., and R.W. McGaughey (1981) Effect of testosterone and dibutyryl cAMP on the spontaneous maturation of pig oocytes. J. Reprod. Fertil., 62:245-256. Sadler, S.E., and J.L. Maller (1981) Progesterone inhibits adenylate cyclase in Xenopus oocytes. Action on the guanine

98

T.W. ECAY AND R.D. POWERS

nucleotide regulatory subunit. J. Biol. Chem., 256.63686373. Schultz, R.M., R.R. Montgomery, P.F. Ward-Bailey, and J.J. Eppig (1983a) Regulation of oocyte maturation in the mouse: Possible roles of intercellular communication, CAMPand testosterone. Dev. Biol., 95:294-304. Schultz, R.M., R.R. Montgomery, and J . Belanoff (198313) Regulation of mouse oocyte maturation: Implication of a decrease in oocyte CAMPand protein phosphorylation in commitment to resume maturation. Dev. Biol., 97:264-273. Smith, D.M., and D.Y. Tenney (1980) Effects of steroids on mouse oocyte maturation in vitro. J. Reprod. Fertil., 60: 331-338. Sorenson, R., and P. Wasserman (1976) Relationship between growth and meiotic maturation of the mouse oocyte. Dev. Biol., 50:531-536. Sorenson, R.A., M.S. Cyert, and R.A. Pedersen (1985) Active maturation-promoting factor is present in mature mouse oocytes. J. Cell Biol., 100:1637-1640. Sunkara, P.S., D.A. Wright, and P.N. Rao (1979) Mitotic factors from mammalian cells induce germinal vesicle breakdown and chromosome condensation in amphibian oocytes. Proc. Natl. Acad. Sci. U.S.A., 76:2799-2802.

Tsafriri, A. (1978) Oocyte maturation in mammals. In: The Vertebrate Ovary. R.E. Jones, ed. Plenum, New York, pp. 409-442. Tsafriri, A,, S.H. Pomerantz, and C.P. Channing (1976) Inhibition of oocyte maturation by porcine follicular fluid: Partial characterization of the inhibitor. Biol. Reprod., 14.511516. Wassarman, P.M., W.J. Josefowicz, and G.E. Letourneau (1976) Meiotic maturation of the mouse oocyte in uitro: Inhibition of maturation at specific stages of nuclear progression. J . Cell Sci., 22.531445. Wasserman, W.J., and Y. Masui (1976) Effect of cyclohexamide on a cytoplasmic factor initiating meiotic maturation in Xenopus oocytes. Exp. Cell Res., 91:381-388. Weintraub, H., M. Buscaglia, M. Ferrez, S. Weiller, A. Boulet, F. Fabre, and E.-E. Baulieu (1982) Mise en evidence d u n e activite “ M P F chez Saccharomyces cerevisiae. C. R. Acad. Sci. (Paris), 295:787-790. Westergaard, L., A.G. Byskov, C.Y. Anderson, J. Grinsted, and K.P. McNatty (1984) Is resumption of meiosis in the human preovulatory oocyte triggered by a meiosis-inducing substance (MIS) in the follicular fluid? Fertil. Steril., 41:377-384.