Synthesis and Accumulation of Hyaluronic Acid and Proteoglycans in the Mouse Cumulus Cell-Oocyte Complex during Follicle-stimulating Hormone-induced Mucification*

In most mammalian ovaries, the cumulus cell-oocyte complex (COC) expands at the time of ovulation by depositing an extensive extracellular matrix between the cumulus cells. This phenomenon can be reproduced in vitro by culturing COCs with follicle-stimulating hormone (FSH) and serum. Biosynthesis of hyaluronic acid (HA) and proteoglycans by mouse COCs in vitro was studied using [3H]glucosamine and [36S]sulfate as metabolic precursors. Radiolabeled complex carbohy- drates were analyzed by ion exchange chromatography, specific enzyme digestion followed by high per- formance liquid chromatography, and gel filtration. The specific activities of [SH]hexosamines in the la- beled molecules were determined by measuring the incorporation of 3H and 35S into chondroitin 4-sulfate disaccharides. When COCs were stimulated with FSH, HA biosynthesis increased 20-30-fold between 3-12 h later when expansion occurs, reaching a maximum rate of -780 pmol (as glucosamine)/COC/h compared with the unstimulated rate of -26 pmol/COC/h. The final concentration of HA in the expanded COC was calcu- lated to be -250 pg/ml. The effects of dibutyryl cyclic AMP (Bt2cAMP) on COC expansion and HA synthesis were similar to those of FSH, suggesting that the effects of FSH are mediated by CAMP. However, FSH significantly decreased the specific activity of the incorporated

In most mammalian ovaries, the cumulus cell-oocyte complex (COC) expands at the time of ovulation by depositing an extensive extracellular matrix between the cumulus cells. This phenomenon can be reproduced in vitro by culturing COCs with follicle-stimulating hormone (FSH) and serum. Biosynthesis of hyaluronic acid (HA) and proteoglycans by mouse COCs in vitro was studied using [3H]glucosamine and [36S]sulfate as metabolic precursors. Radiolabeled complex carbohydrates were analyzed by ion exchange chromatography, specific enzyme digestion followed by high performance liquid chromatography, and gel filtration. The specific activities of [SH]hexosamines in the labeled molecules were determined by measuring the incorporation of 3H and 35S into chondroitin 4-sulfate disaccharides. When COCs were stimulated with FSH, HA biosynthesis increased 20-30-fold between 3-12 h later when expansion occurs, reaching a maximum rate of -780 pmol (as glucosamine)/COC/h compared with the unstimulated rate of -26 pmol/COC/h. The final concentration of HA in the expanded COC was calculated to be -250 pg/ml. The effects of dibutyryl cyclic AMP (Bt2cAMP) on COC expansion and HA synthesis were similar to those of FSH, suggesting that the effects of FSH are mediated by CAMP. However, FSH significantly decreased the specific activity of the incorporated hexosamines while BtscAMP did not. Serum is necessary for the accumulation of HA in the COC matrix. HA synthesis in FSH-stimulated COCs was as high or higher in the absence of serum, but most was recovered in the medium and not in the COC matrix. The molecular size of the HA was >2 million dalton in either case, suggesting that the serum did not alter physical properties of HA. Stimulation of proteoglycan biosynthesis by either FSH or BtzcAMP was less pronounced (three to four times control) than for HA and was sustained throughout an 18-h culture period. A reduction of 80% in the deposition of newly synthesized PGs in the COC matrix by 0.5 mM 8-xyloside treatment did not affect the expansion of the cumulus.
* This work was supported in part by a grant from the Consiglio Nazionale delle Ricerche. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisernent" in accordance with 18 U. S In most mammals, oocytes in fully grown follicles are surrounded by compact layers of follicle cells to form the cumulus cell-oocyte complex (COC).' When the circulating level of gonadotropins increases during the preovulatory period, the compact organization of the COC expands with the deposition of a mucoid material around and between the cumulus cells, a process referred to as expansion or mucification (1,2). At ovulation, the expanded COC is released from the follicle as a viscous and elastic cell mass. Extracellular matrix components of the mucified COC appear to facilitate the pick-up of the COC by oviductal fimbria (3) and to induce changes in spermatozoa preceding the fertilization process (4). Presently, it is not known if mucification is induced directly in uiuo by the gonadotropins or is mediated indirectly by the mural granulosa cells (5-7). However, gonadotropins can induce expansion of the COC in vitro if serum is present in the culture medium (8,9). The expansion process appears to be mediated by CAMP. For example, cAMP analogues, adenylate cyclase activators, and phosphodiesterase inhibitors all stimulate mucification in vitro (10, ll), and FSH stimulates an increase in intracellular cAMP in the mouse COC in vitro (12, 13).
The ultrastructure of the extracellular matrix of the mucified COC contains a fibrillar network with a homogeneous, regular distribution that extends into the outer zona pellucida, the external coat of the oocyte (14). Specific hyaluronidases destroy this network (15) and dissociate the COC into individual cells (16). All the agents, including FSH, that induce mucification in vitro stimulate the incorporation of 3H, from [3H]glucosamine as a precursor, into newly synthesized macromolecules (10). The stimulation is primarily into hyaluronic acid (HA) as determined by precipitation with cetylpyridinium chloride and susceptibility to hyaluronidase (10). Thus, HA is an important component of the expanded matrix of the COC. However, little is known about the characteristics of proteoglycans synthesized by COCs nor about their possible function in mucification. These macromolecules contain glycosaminoglycan chains covalently bound to core proteins, and different types, such as those in cartilage which specifically bind to HA, contribute directly to the structure, organization, and physical properties of extracellular matrices (17). Further, PGs may influence fertilization processes directly. Addition of sulfated glycosaminoglycans to culture medium prevents physical changes in the zona pellucida which occur in isolated oocytes and which inhibit in vitro fertilization (18,19). Exogenous glycosaminoglycans also stimulate the acrosome reaction i n vitro for spermatozoa from several species (4, 20).
In the study described in this report, we define the temporal changes in synthesis of HA and PGs during mucification of the mouse COC following stimulation i n uitro with FSH or Bt,cAMP. We show that FSH and Bt2cAMP induce an increase of both HA and PG biosynthesis by COC cultured i n uitro, and that increased net synthesis of HA, but not of PGs, correlates closely with the expansion process. A model for mucification is proposed in which cumulus cells require FSH and a factor(s) derived from the oocyte to increase HA synthesis and require a factor(s) in serum to accumulate this HA in the COC matrix.  ) from Behring Diagnostics; pregnant mare's serum gonadotropin, hyaluronic acid (from human umbilical cord), Bt,cAMP and dimethyl polysiloxane (type 5X) from Sigma; p-nitrophenyl-P-D-xyloside from Koch-Light Laboratories (United Kingdom); Eagle's minimum essential medium with Earle's salt and HEPES from GIBCO; FCS from Flow Laboratories. Highly purified follicle stimulating hormone (NIDDK-rat-FSH-1-7) was kindly provided by the National Institute of Diabetes and Digestive and Kidney Diseases, and the National Hormone and Pituitary Program, University of Maryland School of Medicine. Female Swiss CD-1 mice (7-8-week old) were purchased from Charles River. The unsaturated hyaluronic acid disaccharide, Adi-HA, was prepared by digesting 10 mg of HA with 1 IU of chondroitinase ABC in 1 ml of 0.1 M Tris, 0.1 M acetate, pH 7.3, at 37 "C for 3 h. Digests were stored at -80 "C until further use.

Materials-Guanidine
Metabolic Labeling of COCs and Extraction of Radiolabeled Molecules-Adult, 2-month old mice were injected with 5 IU of pregnant mare's serum gonadotropin in 0.1 ml of physiological saline and killed by cervical dislocation 44-48 h later. Ovaries were removed and placed in minimal essential medium containing 1 mg/ml bovine serum albumin and buffered at pH 7.2 with 20 mM HEPES. COCs were released into the medium by puncturing large follicles. They were collected with a micropipette and transferred into droplets of culture medium (50-100 pl) covered with dimethyl polysiloxane to prevent evaporation of medium; between 25 and 100 COCs were used for each treatment, and at least three different experiments were done for each protocol. Unless specified in the text, COCs were cultured in minimal essential medium supplemented with 5% FCS as the basal control condition and with 5% FCS plus 1 pg/ml FSH or plus 2 mM BtZcAMP as the mucification induction condition (9). The cultures were incubated at 37 "C in a humidified atmosphere of 5% CO, in air for 18 h, a sufficient time for mucification to occur (9). For radiolabeling, [35S]sulfate (30 pCi/ml) and [3H]glucosamine (100 pCi/ml) were included in the medium. All manipulation steps of extraction of the COC were done under the dimethyl polysiloxane. After labeling, media samples were carefully removed from COCs with a micropipette. The COCs were washed twice with 25 pl of fresh medium and then extracted with 50 pl of 4 M guanidine HC1, 50 mM sodium acetate, pH 6.0, containing 2% (w/v) Triton X-100 and protease inhibitors (21, 22) at 37 "C for 5 h, then the extract was transferred to a tube and brought to appropriate volume with 4 M guanidine HC1. Media samples were diluted with 500 p1(-5 volumes) of 4 M guanidine HC1 buffer containing protease inhibitors. All extracts were stored frozen until further analyses.
Evaluation of Mucification Process-COCs were examined for cumulus mucification with a stereo microscope. Expansion, elasticity, and resistance to mechanical disruption of the COcs were criteria for a positive response.
Isolation of Proteoglycans-Guanidine HCl extracts of COCs and media samples were eluted on Sephadex G-50 columns (8 ml of bed volume for 2 ml of sample) equilibrated with 8 M urea, 50 mM sodium acetate, 0.15 M NaC1,0.5% Triton X-100, pH 6.0, to remove unincorporated isotopes and guanidine HC1 (23). Excluded fractions with labeled macromolecules were applied onto Q-Sepharose columns (0.7 X 4 cm) equilibrated with the same urea buffer. After sample application, each column was washed with 5 ml of equilibrating buffer and then eluted with a NaCl gradient (0.15-1.2 M) in the 8 M urea buffer (total volume of 46 ml) with a flow rate of 15 ml/h. Fractions of 1 ml were collected and aliquots were measured for radioactivity and conductivity.
Gel Filtration-A prepacked Superose 6 column (1 X 30 cm) was eluted with 4 M guanidine HC1, 50 mM sodium acetate, 50 mM Tris, 0.5% Triton X-100, pH 7 Quantitation of HA, PGs, and Glycoproteins-After adding 200 pg of HA and 100 pg of bovine serum albumin to each sample, aliquots of COC extracts and media samples were eluted on Sephadex G-50 columns (2 ml of bed volume for 0.5 ml of sample), equilibrated in 0.1 M Tris, 0.1 M acetate, 0.5% Triton X-100, pH 7.3, to remove unincorporated isotopes and guanidine HCl. Macromolecules in the excluded fraction were then digested with chondroitinase ABC as described above. A portion of each digest was applied to a Sephadex G-50 column (4 ml of bed volume) equilibrated in 8 M urea, 0.15 M NaC1, 50 mM sodium acetate, 0.5% Triton X-100, pH 6.0, and eluent fractions were analyzed to determine proportions of radioactivity that was enzyme susceptible (HA and dermatan sulfate disaccharides in included volume) and enzyme resistant (heparan sulfate and glycoproteins in excluded volume). Another portion of each digest was analyzed for the proportions of the disaccharide digestion products by an HPLC procedure using Partisil 5 PAC (24) after adding Adi-4S, Adi-6S, and Adi-OS disaccharide standards (5 pg each) as internal standards. The Adi-HA in each sample was derived from the enzyme digestion of the carrier HA.
For all experiments, radioactivities of samples were determined with a Beckman LS 5801. Differentiation of 35S and 3H activity was done by calculating 35S spill-over into the 3H channel using 36S standards prepared for each set of samples. Statistical Analysis-Differences between two groups of data were analyzed by Student's t test. p < 0.05 was considered significant.  sufficient time for expansion to occur (9). A COC extract and medium sample were prepared and eluted on Sephadex G-50 columns with an 8 M urea solvent. Labeled macromolecules in the excluded fractions were then analyzed by ion-exchange chromatography on Q-Sepharose. For the COC extract, Fig.   2a, -38% of "H-labeled macromolecules did not bind (peak l ) , -21% eluted as a narrow peak during the gradient at 0.28 M NaCl (peak 2), and -40% eluted as a broad peak between 0.55-0.85 M NaCl (peak 3 ) . Almost all of the incorporated "S activity eluted in peak 3. In a number of different experiments, 60-80% of the labeled molecules in peak 2 were digested to small fragments by Streptomyces hyaluronidase (data not shown). Further, authentic HA added to a similar sample eluted with peak 2. Therefore, this peak contained primarily labeled HA. For peak 3, -65% of the 'H (-70% of the "S) was susceptible to chondroitinase ABC and -25% of the 'H (-30% of the %) to heparitinase. Other experiments showed that most of the heparan sulfate eluted in the earlier fractions of peak 3 (data not shown). Peak 3, then, contained almost exclusively labeled dermatan sulfate and heparan sulfate PGs.

Cumulus Cell-Oocyte Complex Mucification in Vitro-
The medium sample gave a similar profile on Q-Sepharose, were recovered in the COC extract.
While recoveries of "S-labeled macromolecules from Q-Sepharose were nearly quantitative, averaging between 90-98%, recoveries of 'H-labeled macromolecules were poorer, 85-95% for media samples and only 45-70% for COC extracts. Most of the missing ' H in the COC extracts bound at the top of the Q-Sepharose columns. The ion-exchange procedure, then, is useful for purifying PGs synthesized by the COC; but the low recoveries of 3H, especially in the COC extracts, indicated that it gives poor yields of either HA (peak 2), or glycoproteins (peak 1) or both. Another procedure was therefore developed to address this problem.
Bovine serum albumin and HA were added to samples to facilitate recoveries, and the mixtures were eluted on Sephadex G-50 columns equilibrated with 0.1 M Tris, 0.1 M sodium acetate, 0.5% Triton X-100, pH 7. 3. The recoveries of macromolecular 'H and "S were quantitative. Aliquots of the excluded fractions were digested directly with chondroitinase ABC under conditions sufficient to digest all of the carrier and labeled HA as well as the labeled dermatan sulfate. Undigested macromolecules were removed by centrifugation after precipitation with 3 volumes of ethanol. Standard chondroitin sulfate disaccharides were added to the soluble digestion products, and the samples were then analyzed by HPLC on Partisil 5 PAC (24). For the COC extract, two ' H peaks were observed eluting with Adi-HA and Adi-4S, respectively (Fig. 3). The former contained 80-90% of the 'H in the analysis while the latter contained 10-20% as well as all of the ' 73.
The percent of 'H in heparan sulfate in the each sample was evaluated independently by digesting a separate aliquot with heparitinase followed by elution on Sephadex (3-50 to quantify the digestion products. For the COC extract, -5% of the total 'H and -30% of the 3sS was in heparan sulfate, the latter value in close agreement with the data described above for peak 3. The 3H and ?3 distributions for both the COC extracts and medium samples are summarized in Table I    with the results from the Q-Sepharose analyses. It is clear that HA accounts for virtually all of the 3H-labeled material which did not elute from Q-Sepharose; for example, labeled HA in peak 2 of the COC extract accounted for less than 20% of the labeled HA in the original extract. The distributions of radiolabel in HA, PGs, and glycoprotein fractions in all the following experiments were determined by the enzyme digestion and HPLC procedure. Specific Activity Corrections-When [3H]glucosamine is used as a metabolic precursor, it is diluted in the cytoplasm by endogenous pathways for synthesis of hexosamines. Thus, the specific activity of the UDP-N-acetylhexosamine precursors for glycosaminoglycan synthesis is much lower than that of the [3H]glucosamine in the medium, and it often varies depending upon the experimental conditions (25-28). For this reason, an indirect method was used to estimate the specific activity of the UDP-N-acetylhexosamine pool (28). The method uses the specific activity of [35S]sulfate in the medium,  Table 11. sa(S), combined with the 3H/35S-labeling ratio, Ir, in the monosulfated Adi-4S to calculate the specific activity of the galactosamine, sa(H), in this disaccharide derived from the dermatan sulfate synthesized during the labeling p e r i~d .~ The formula is (28): Because UDP-galNAc and UDP-glcNAc are in a rapid equilibrium (29), the specific activity of glucosamine in the newly synthesized HA is assumed to be equivalent to sa(H).
In the experiments reported in this paper the specific activity This is based on the fact that the specific activity of [36S]sulfate in the medium is not significantly diluted by endogenous sulfate sources in the intracellular phosphoadenosinephosphosulfate pool, the immediate metabolic precursor for the sulfate ester in glycosaminoglycan synthesis (28). For rat granulosa cells, less than 2% of the sulfate esters on the proteoglycans is derived from cysteine or methionine when environmental sulfate in the medium is not limiting, i.e., above 0.1 mM (M. Yanagishita  * Values represent averages of three experiments. Coefficients of variation were less than 25%. of the [35S]sulfate in the medium was determined to be -37. 5 pCi/pmol initially (30 pCi of [35S]sulfate/800 nmol sulfate/ ml), and corrections for radioisotope decay were made thereafter. The values for sa(H) and the total incorporation, then, can be used to calculate the masses of HA and PGs synthesized during the labeling period.
Changes in Net Synthesis-An example of such an analysis is shown in Fig. 4 for samples derived from COC extracts after incubation in labeling medium for 18 h without (control), or with stimulation of mucification by FSH or Bt2cAMP. The calculated results are given in Table I1 and summarized in Fig. 5. It is clear that FSH decreased the specific activity of the UDP-N-acetylhexosamine precursors relative to the control, whereas BtzcAMP did not. Thus, the values for the 3Hlabeled macromolecules must be corrected for these differences in specific activity. Both FSH and BtzcAMP increased net synthesis of all of the glycosylated components, but the largest increases occurred for HA.
Even though Bt2cAMP stimulated a larger incorporation of 3H into HA (7.2 x control) than FSH (6.1 X control), the correction for specific activity differences shows that FSH increased net HA synthesis (9.9 x control) more than did Bt2cAMP (6.5 x control).
The amounts of the newly synthesized glycosylated components released into the culture medium in this experiment were also determined by the same procedure. There were no significant changes between control and treated samples for any of the glycosylated components except the small increase of DS-PG by FSH treatment (Fig. 5). For this reason, in the following experiments only data for the COC extracts are presented.  Fig. 6. Maximal levels of stimulation of synthesis were obtained with 5 ng/ml FSH and 0.5 mM Bt2cAMP, with EDso values of -1 ng/ml and -0.2 mM, respectively. As noted above, HA synthesis increased much more than for PG synthesis. Both types of PGs, heparan sulfate and dermatan sulfate, showed about the same degree of stimulation. The specific activities of the Adi-4s galactosamine decreased for FSH but remained unchanged for Bt2cAMP (insets, Fig. 6).

-
Time Course of HA and PG Synthesis-HA and PG syn-Synthesis in Cumulus Cell-Oocyte Complexes 13845    6-12 h. The rate then returned toward control levels during 12-18 h. The maximal increases in both dermatan sulfate and heparan sulfate PG synthesis (three to four times control) were statistically significant but were much less than for HA. Additionally, the time course of stimulation was different, with elevated synthesis of PGs sustained through the 12-18h time period.
Characterization of PGs Synthesized by the COG-PG fractions (peak 3) were isolated from COC extracts and media samples of control and FSH stimulated COCs by chromatography on Q-Sepharose (see above). After concentration, aliquots were eluted on Superose 6 in 4 M guanidine HC1 with 0.5% Triton X-100. The elution profiles for the COC extract and medium PGs of the FSH-stimulated COCs are shown in Fig. 8. The medium sample showed two major 35S peaks, one excluded (Mr greater than 300 kDa) and the second centered at a K d of 0.21 (M, -150 kDa). Since dermatan sulfate is predominant in the medium, these two PG peaks probably correspond to DS-PG-I and DS-PG-I1 previously described as synthetic products of rat granulosa cells (30,31). The COC extract contained PG peaks with similar elution positions Hyaluronic Acid Synthesis in Cumulus Cell-Oocyte Complexes followed by a broad peak centered between Kd 0.2-0. 6. This latter peak probably contains intracellular glycosaminoglycan degradation products similar to those described for the rat granulosa cells (32, 33). The elution profiles for samples from the basal, unstimulated COCs were essentially identical. More complete analyses of the PGs synthesized by the COCs are underway. These preliminary results, however, indicate that there are no obvious qualitative changes in the PGs produced by COCs stimulated to mucify compared with unstimulated COCs.
Effects of Nitrophenyl-P-D-xyloside on HA and PG Synthesis and on COC Expansion-/3-Xylosides act as exogenous initiators for glycosaminoglycan synthesis, thereby competing with the endogenous core protein acceptors and decreasing mature PG synthesis, particularly for chondroitin sulfate/ dermatan sulfate PGs (34,35). FSH-stimulated COCs were continuously exposed to 0.5 mM nitrophenyl-P-D-xyloside during 18 h of culture to determine if inhibition of PG synthesis affects expansion. The xyloside treatment increased total 35S incorporation 9-10-fold. Almost all of the 35S-labeled molecules were recovered in the medium and eluted as free glycosaminoglycan chains on Superose 6, Fig. 8c (compare with 8b). The xyloside also inhibited the 3H incorporation into PGs accumulated in the COC matrix to -20% of FSHtreated COC, Table 111. In agreement with previous observations (36), HA synthesis was slightly inhibited (-75% of FSHtreated COC) by the xyloside treatment , Table 111. The xyloside treatment did not alter the FSH-stimulated COC mucification in any obvious way as assessed by light microscopy (data not shown). Thus, a significant reduction in the deposition of newly synthesized PGs in the COC matrix by the xyloside treatment did not affect the expansion of the cumulus, suggesting that PGs have little or no role in the mucification process.
Effect of the FCS on HA and PG Synthesis and Distribution-In the presence of FSH, but the absence of serum, cumulus expansion does not occur in vitro and the proportion of 3H-labeled macromolecules ( [3H]glucosamine as a precursor) in the medium increases (37). For this reason, the effects of FCS on net synthesis of HA and PG and on their distribution were studied in FSH stimulated COCs (Table 111). In the absence of FCS (BSA + FSH), net synthesis of PGs was unchanged while net synthesis of HA increased -60% compared to cultures with FCS (FCS + FSH). Further, the distribution of the PGs between the COC extract and the medium was unchanged. However, there was a dramatic redistribution of HA from the COC matrix into the medium, Table 111 and Fig. 9. After 18 h in the presence of FCS, -80% of the total labeled HA accumulated in the COC matrix. Conversely, in the absence of FCS, only -20% of the total labeled HA accumulated in the COC matrix with the remainder released into the medium, Table 111. Morphological observations confirmed that cumulus expansion did not occur for COCs incubated without FCS (data not shown).
Labeled macromolecules in the cell and medium fractions for both treatments were analyzed by Sephacryl S-1000 chromatography in 4 M guanidine HC1, 0.5% Triton X-100, Fig. 9. 3H-Labeled HA synthesized by the COCs in the presence of FCS and extracted from the COC matrix eluted in the excluded column volume, as indicated by the sensitivity of this peak to Streptomyces hyaluronidase digestion, Fig. 9a. Similarly, the labeled HA in the medium for COCs incubated without FCS was also excluded from the column, Fig. 9d. Thus, in both conditions, the hydrodynamic size of the labeled HA molecules indicates that they have M, values greater than 2 million (38), and it is unlikely that changes in their physical properties per se account for the observed redistribution.

DISCUSSION
The experiments described in this report use [3H]glucosamine and [35S]sulfate to study the synthesis and distribution of hyaluronic acid and proteoglycans in mouse cumulus oocyte complexes stimulated to mucify in vitro. While extraction of labeled macromolecules from the COC matrix with 4 M guanidine HC1 and detergent was efficient, there were significant losses of labeled HA in the subsequent ion-exchange step. Thus, this macromolecule was measured by an alternative procedure which involved chondroitinase digestion and analysis of the disaccharides derived from both the HA and the DS-CS PGs. The known specific activity of the radiosulfate in the medium and the ratio of 3H/35S in the chondroitin 4sulfate disaccharide were used to correct for changes in specific activity of hexosamine pools so that the amounts of HA and PGs synthesized/time could be estimated.
COCs stimulated to mucify by FSH in the presence of FCS synthesized HA at -10 times the rate for unstimulated COCs during 18 h of treatment, and most of this newly labeled HA accumulated in the COC matrix (Fig. 5 ) . The time course for the stimulation of HA synthesis and its accumulation correlates with COC expansion. After FSH stimulation, HA synthesis and accumulation was four to five times the control in the first 3 h, increased to 20-30 times the control during 3-12 h, and decreased back to four to six times the control during 12-18 h (Fig. 7 ) . COC expansion in these conditions becomes apparent 4-5 h after FSH stimulation and appears to reach completeness by morphological criteria by 12-14 h. The net increase in HA in the COC matrix during the 18 h was -2.0 ng/COC which would contribute a concentration of 250 pg of HA/ml in the matrix based on the estimated volume of the expanded COC (8 x mm3). COCs stimulated by FSH in the absence of FCS increased net HA synthesis as much or more as for COCs stimulated by FSH in the presence of FCS. However, the HA accumulates in the medium and not the COC matrix (Fig. 9), and these COCs do not expand. These results confirm and extend those reported by Eppig (37) who used cetylpyridinium chloride precipitation of 3Hlabeled macromolecules and treatment with Streptomyces hyaluronidase to show that HA was present in the COC matrix for cultures stimulated to mucify in the presence of FCS and in the medium when FCS was absent.
PGs bind to HA in some connective tissue matrices such as cartilage (17) and are important structural components in the organization of such matrices. PG synthesis and accumulation in the COC matrix did increase during mucification to levels two to three times the control values during the 18 h of treatment (Fig. 5 ) . However, this increase does not appear to contribute significantly to the expansion or organization of the COC matrix since xyloside treatment effectively prevented the accumulation of PGs in the COC matrix (Table  111) but did not prevent expansion. The PGs synthesized by the COC have molecular properties similar to those synthesized by rat ovarian granulosa cells. One HS PG, partially included on Superose 6, and two CS/DS PGs, one excluded and one partially included, were identified (Fig. 8). If the PGs are homologous, the observation that the PGs synthesized by mural granulosa cells do not bind to HA (39) would be consistent with the lack of their direct involvement in the organization of the HA matrix. However, PGs synthesized by cumulus cells may be an important component of the environment of the oocyte and for the fertilization process. It has been reported that the zonae pellucidae of isolated mouse oocytes cultured i n vitro become increasingly resistant to solubilization by chymotrypsin (40) thereby decreasing the success of i n vitro fertilization (19). These physicochemical changes in the zonae pellucidae are prevented by adding sulfated glycosaminoglycans (heparin and DS) to the culture medium (18). Moreover, these changes do not occur during oocyte maturation i n vivo nor when isolated oocytes are incubated with intact COC (40).
While our results and those of Eppig (37) have shown that factors in serum (or in follicular fluid) are required for accumulation of HA and expansion of COC matrix, their mechanism of action remains uncertain.
The Sephacryl S-1000 analyses ( Fig. 9) suggest that the absence of serum does not cause a decrease in the molecular size of the newly synthesized HA molecules, at least not below 2 million, which could have altered their physical properties. Serum contains HA-binding proteins (41), and these may be necessary for organizing the newly synthesized HA in the COC matrix. Protease digestion, for example, can dissociate the COC as can treatment with a specific hyaluronida~e.~ Alternatively, the serum factor(s) may be required for the cumulus cells to synthesize a critical component involved in organization of the HA in the matrix.
Cyclic AMP analogues have been shown to stimulate HA synthesis and mucification of COC i n vitro (9, lo), and it is thought that FSH exerts its influence on this process via CAMP pathways. Like FSH, BtzcAMP stimulates 3H incorporation from [3H]glucosamine as a precursor into all the glycosylated components of the COC matrix (Table 11). Dose response curves showed that FSH and Bt2cAMP were halfmaximally effective at 1 ng/ml and 0.2 mM, respectively ( Fig.  6), which, for FSH at least, is within physiological range. The COCs respond differently to these two factors in one respect. The specific activity of the UDP-hexNAc pools decreased significantly in FSH-stimulated cultures (-60% at maximal stimulation) but remained unchanged in BtzcAMP-stimulated cultures (Fig. 6). After correcting for the changes in specific activity, we found that BtzcAMP and FSH showed the same temporal patterns for stimulation of HA, PG, and GP synthesis. However, while levels of synthesis of PGs and GPs increased to the same extent with both treatments, FSH-stimulated cultures showed higher net synthesis of HA than obtained with Bt2cAMP. These observations are consistent with a direct role for CAMP in mediating FSH action, but they also suggest that other factors may be involved. In separate experiments, we have shown that FSH does not increase HA synthesis in cumulus cells separated from the oocyte unless medium conditioned by isolated oocytes is used during the hormonal stimulation.'j Thus, minimally, a soluble factor produced by the oocyte is essential in combination with FSH to stimulate HA production by cumulus cells. In summary, we have shown that HA and PG synthesis by COCs are increased during the FSH stimulation but that only HA synthesis and accumulation are directly involved in the mucification process. Studies are in progress to define roles of the oocyte and of serum on synthesis and organization of HA in the extracellular matrix of COC. Acknowledgment-We wish to thank Dr. Ronald J. Midura for his valuable advice during these studies.