Solubilization of Hyaluronic Acid Synthetic Activity from Streptococci and Its Activation with Phospholipids*

To date all hyaluronic acid synthetic systems have been of a particulate nature, and attempts at solubilization have been unsuccessful. This has hampered attempts to elucidate the mechanism by which hyaluronic acid is produced. In this paper we demonstrate that the hyaluronic acid synthetic activity from group C streptococcal membranes was solubilized using 2% digitonin and that the activity was optimized by reconstitution with cardiolipin at an optimum phospholipid/protein ratio (microgram/microgram) of 5:1. Furthermore, chromatography of the solubilized synthetase demonstrated that it eluted after the void volume of a Sepharose CL-6B column. CHAPSO, octyl glucopyranoside, sodium cholate, Triton X-100, and zwittergent 314 either inhibited or failed to solubilize the synthetic activity. Phospholipids other than cardiolipin also reconstituted the activity from the digitonin extract, particularly phosphatidylethanolamine and phosphatidylserine. In our system, the specific activity of hyaluronic acid synthetase was increased up to 63 times that of the system of the intact membrane. Furthermore, the total activity of the reconstituted system was 4.9 times greater than that of intact membranes. The soluble enzyme system showed similarities to the membrane-bound synthetase in the kinetics of production of trichloroacetic acid-soluble and -insoluble hyaluronic acid, and the hyaluronic acid produced was of comparable molecular weight.

Solubilization of Hyaluronic Acid Synthetic Activity from Streptococci and Its Activation with Phospholipids* (Received for publication, August 6, 1985) Mark X. Triscott and I. van  To date all hyaluronic acid synthetic systems have been of a particulate nature, and attempts at solubilization have been unsuccessful.
This has hampered attempts to elucidate the mechanism by which hyaluronic acid is produced.
In this paper we demonstrate that the hyaluronic acid synthetic activity from group C streptococcal membranes was solubilized using 2% digitonin and that the activity was optimized by reconstitution with cardiolipin at an optimum phospholipid/protein ratio (pg/pg) of 5:l. Furthermore, chromatography of the solubilized synthetase demonstrated that it eluted after the void volume of a Sepharose CL-6B column. CHAPSO, octyl glucopyranoside, sodium cholate, Triton X-100, and zwittergent 314 either inhibited or failed to solubilize the synthetic activity. Phospholipids other than cardiolipin also reconstituted the activity from the digitonin extract, particularly phosphatidylethanolamine and phosphatidylserine.
In our system, the specific activity of hyaluronic acid synthetase was increased up to 63 times that of the system of the intact membrane.
Furthermore, the total activity of the reconstituted system was 4.9 times greater than that of intact membranes. The soluble enzyme system showed similarities to the membrane-bound synthetase in the kinetics of production of trichloroacetic acid-soluble and -insoluble hyaluronic acid, and the hyaluronic acid produced was of comparable molecular weight.
Hyaluronic acid is a ubiquitous nonbranching acid mucopolysaccharide composed of @ 1-4-linked repeating disaccharide units of glucuronic acid p l-3 linked to N-acetylglucosamine. Even though the hyaluronic acid biosynthetic system was one of the first membrane heteropolysaccharide synthetic pathways studied (l), our understanding of the mechanism of synthesis of hyaluronic acid is still incomplete. This lack of knowledge is attributed to an inability to solubilize the synthetic system from either eucaryotic or procaryotic sources. Many investigators have taken advantage of the abundance and exclusive production of the glycosaminoglycan hyaluronic acid by groups A and C streptococci (2)(3)(4)(5). Other workers have investigated hyaluronic acid production in mammalian systems using intact membranes or whole cells (6-9). These studies, using particulate enzyme preparations, demonstrated that certain factors were essential for in vitro hyaluronic acid synthesis. These included the precursors UDP-glucuronic acid and UDP-N-acetylglucosamine as well as a requirement for the divalent cation magnesium or manganese (4). As far as the actual functioning of the system is concerned, it is not completely understood whether the alternating addition of the monosaccharides occurs at the reducing (10) or nonreducing (1) end of the growing hyaluronic acid polymer. In addition, the role of lipid intermediates in the synthesis of hyaluronic acid is unclear in eucaryotes (11,12) and has been discounted in streptococci. This is based on the demonstration that bacitracin and tunicamycin, which are both involved in the inhibition of dephosphorylation of lipid pyrophosphates, fail to inhibit hyaluronic acid synthesis. The role of phospholipids as necessary structural components of the enzyme system has not been investigated.
Other questions to be addressed include whether the nascent hyaluronic acid chains are linked to a protein, and the role, if any, of low molecular weight primer in hyaluronic acid biosynthesis. In order to resolve these basic questions and to determine whether eucaryotic production mechanisms parallel those of streptococci, it is necessary to solubilize the hyaluronic acid synthetic system. In the past, attempts to solubilize the enzyme system with detergents (6) or organic solvents (18) resulted in complete inactivation of the enzyme. In this report we demonstrate the first successful solubilization of the streptococcal hyaluronic acid synthetic system using the detergent, digitonin. In addition, phospholipids were required for the reconstitution of activity of this solubilized synthetic system, and their presence increased the total activity above that of intact membranes. When compared to intact membranes and the particulate enzyme systems previously described, the solublized synthetase produced hyaluronic acid at an equivalent rate and in the same molecular weight range, and a specific activity of 2406 nmol of glucuronic acid transferred/h/mg of protein was demonstrated for the reconsituted system. exponential phase (0.4 absorbance units at 650 nm), treated with hyaluronidase (bovine testes, Worthington; 3 mg/liter) and harvested in a continuous flow centrifuge (Sharples-Stokes Div., Pennwalt Corp., Warminster, PA.). The bacteria were washed in cold 0.15 M NaCl and resuspended to 10% (w/v) in protoplasting buffer (0.05 M Na2HP04/KH2P04, pH 6.1, 0.5 mM MgC12, 0.5 mM dithioerythritol containing 30% raffinose, phage lysin (10% v/v), and DNase (10 pg/ ml)) at 37 "C. The bacterial preparation was then incubated for 1 h at 37 "C, checked by Gram stain for complete protoplast formation, and sedimented at 7,500 X g for 8 min at 22 "C. The protoplasts were then resuspended to 30% (w/v) in protoplasting buffer without phage lysin and sedimented at 9000 X g for 10 min. The protoplasts were finally lysed in hypotonic buffer (0.05 M Na2HP04/KH2P04, pH 7.0, 10 mM MgC12, RNase (10 pg/ml), DNase (10 pg/ml)) and incubated at 37 "C for 1 h. The membranes were then washed 5 times in phosphate-buffered saline (0.05 M sodium phosphate, pH 6.9; 0.15 M NaCl) with 10 mM MgClz and 0.02% NaN3. The membranes were aliquoted and stored at 3-5 mg/ml of protein at -70 "C.
Detergent Extraction-Detergents were suspended or dissolved in phosphate-buffered saline (0.05 M Na2HP04/KHzP04, pH 6.9,0.15 M NaC1) containing 10 mM MgC12, 5 mM dithioerythritol, and 10% glycerol (buffer A). Bacterial membranes (3-5 mg/ml of protein) prepared as described above were sedimented by centrifugation at 10,000 X g and resuspended to their original volume in the appropriate detergent. The membranes were then extracted at 4 "C for 1 h with stirring. Once the membranes were extracted, the mixture was sedimented by centrifugation at 13,000 X g for 3 min, the supernatant was removed, and centrifuged for an additional 1 h at 196,000 X g. The 196,000 X g pellet was resuspended in buffer A to its original volume for comparison with the activity of 196,000 X g supernatant.
It should be noted that various lots of digitionin from Sigma gave reproducible results in our assays. However, highly variable results were obtained with different lots of digitonin from Calbiochem-Behring.
Phospholipid Reconstitution of Detergent Extracts of Membranes-Phospholipids were prepared as 10 mg/ml stock solutions in 0.45% CHAPSO in buffer A and sonicated until a uniform suspension was achieved. An aliquot of the 196,000 X g supernatant of detergent extracted membranes (125 pl, 0.20-1.25 mg/ml) was mixed with the appropriate concentration of phospholipid (125 pl). This mixture was then placed in dialysis tubing (12,000-14,000 M, cutoff) and dialysed against three changes of 200 volumes of buffer A over 48 h with stirring at 4 "C.
Assay of Hyaluronic Acid Synthetic System-The reaction mixture used was modified from that of Sugahara et al. (2) and Prehm (6). A final volume of 100 pl contained, 0.4 pmol of dithioerythritol, 10 pmol of MgC12, 30 nmol of UDP-GlcNac, 12.6 nmol of UDP-GlcUA (4.4 X lo4 cpm) and 2-150 pg of enzyme protein in 47 mM Na2HP04/KH2P04 buffer, pH 7.0. Amounts of enzyme protein, phospholipid, and detergent varied according to the experiments. All reactions were performed at 37 "C and were initiated by the addition of enzyme protein and terminated by the addition of 700 pl of cold 6.6% trichloroacetic acid solution. For the assays using whole membrane, solubilized synthetase, and reconstituted synthetase, the rate of the reaction was directly proportional to the enzyme concentration over the range used (10-fold). All reaction rates were taken from the initial linear portion of the product versus time curves.
For determination of the radioactivity incorporated into the trichloroacetic acid insoluble hyaluronic acid, the pellet was washed twice in 700 pl of cold 6.6% trichloroacetic acid, dissolved in 1 ml of Aquasol-2 (New England Nuclear), and counted using a scintillation counter (Packard Instrument Co., Tricarb 4640). For the determination of radioactivity incorporated into the trichloroacetic acid soluble hyaluronic acid, the trichloroacetic acid supernatant obtained in the assay above was dialysed against three changes of 500 volumes of distilled water over 96 h. The dialysed samples were then placed in 20 ml of Aquasol-2, and the radioactivity was determined as above. Zero time controls were run to correct for residual radioactivity.
Gel Filtration of Hyaluronic Acid-['4C]Glucuronic acid labeled hyaluronic acid was produced by phospholipid-reconstituted detergent extracts of streptococcal membranes incubated with reaction mixtures for 1 h at 37 "C. The polymer was released by incubating the reaction mixture at 37 "C with 1% sodium dodecyl sulfate for 1 h. This preparation was centrifuged at 196,000 X g, and the supernatant was dialyzed against three changes of 500 volumes of distilled water over 48 h to remove unincorporated UDP-['4C]glucuronic acid. The dialyzed sample was chromatographed on a Sepharose 2B column (1.5 x 48 cm) equilibrated with 0.5% sodium dodecyl sulfate and 0.15 M NaC1, run at a flow rate of 8 ml/h, and 2-ml fractions were collected. Void and total volumes were determined using latex beads (0.926 pm) and tritiated water, respectively.
Enzyme Treatments-Hyaluronidase (0.05 mg/ml) in 0.1 M sodium phosphate buffer, pH 5.3, with 0.15 M NaCl was added to hyaluronic acid and incubated at 37 "C for 1 h. Trypsin (2 mg/ml) treatment of detergent extracts (0.025 mg of protein/ml) was carried out in 0.05 M phosphate buffer, pH 7.4, with 0.15 M NaCl. The reaction mixture was incubated at 37 "C for 2 h and then terminated by the addition of soy bean trypsin inhibitor to a final concentration of 3 mg/ml.
Total Lipid Extract-Total lipid extracts were prepared according to the method of Bligh and Dyer (15). The extracts were reconstituted in buffer A containing 0.45% CHAPSO.
Protein Determinations-Protein determinations were carried out using the Lowry method as modified by Markwell et al. (16). The Markwell reagent was adjusted to 4% sodium dodecyl sulfate for assaying samples containing Triton X-100.

RESULTS
Detergent Extraction of Membranes-In preliminary experiments used to determine conditions for solubilization of the hyaluronic acid synthetic enzyme, streptococcal membranes were prepared and extracted with detergents as described under "Experimental Procedures." A range of concentrations were used for each detergent, and the concentration giving peak activity for hyaluronic acid synthesis is shown in Table   I. Analysis of the 13,000 x g pellets demonstrated that only 2.7-7.2% of the enzymatic activity remained associated with the membrane pellet after detergent extraction except for the pellets from CHAPSO and sodium cholate extractions (105 and 61.2%, respectively). Analysis of the 196,000 X g pellet fractions indicated that 3.1-163% of the activity was associated with the high speed pellets. No activity was found in any of the 196,000 X g supernatants indicating that the synthetase was not solubilized in an active form at the detergent concentrations tested.
The total recovery of enzymatic activity from each extraction protocol demonstrated that only CHAPSO enhanced the enzymatic activity as compared to intact membrane controls (290.3 nmoles/h, combined fractions; 162.2 nmoles/h, CHAPSO control; 108.3 nmoles/h, intact membrane), whereas the other detergents either destroyed or inhibited the enzymatic activity. Furthermore, the specific activity of the 196,000 X gpellet was more than 36-fold that of the membrane control with 163% of the original total activity. Attempts were made to solubilize the activity from the 196,000 x g pellets using the same detergent, different detergents, and detergent mixtures; however, no activity was found in any of the 196,000 x g supernatants obtained from these experiments.
Reconstitution of Hyaluronic Acid Synthetic Activity from Solubilized Membranes-Since less than 100% of the enzymatic activity was recovered from any of the detergent extracts other than the CHAPSO extract, the possibility existed that the detergents exhibited an inhibitory effect on the hyaluronic acid synthetic activity of the 196,000 x g supernatants. Therefore, all 196,000 X g supernatants were dialyzed against buffer A, and the hyaluronic acid synthetic activity of the dialyzed samples was assayed (Table 11). It should be pointed out that since Triton X-100 dialyzes poorly, the Triton levels were not decreased to the same levels as the other detergents before the supernatants were analyzed for enzymatic activity. Soluble hyaluronic acid synthesis activity was found for the first time in both the dialyzed digitonin and CHAPSO 196,000 x g supernatants. Dialyzed digitonin extracts gave the highest specific activity (41.7 nmol transferred/h/mg of protein) which, however, represented only 13.3 nmoles transferred/h (12.3% recovery). Therefore, still only TABLE I Detergent solubilization of hyaluronic acid synthesis enzyme(s) from group C streptococcal membranes Membranes were prepared as described under "Experimental Procedures" (3-4 mg of protein/ml). These intact membrane preparations were aliquoted (200 pl) and centrifuged for 3 min at 13,000 X g. The listed detergent solutions were prepared in buffer A. The pelleted membranes were resuspended to their original volumes in the detergent solutions and extracted at 4 "C for 1 h with stirring. Residual membrane was removed by centrifugation at 13,000 X g for 3 min, and the supernatant was then sedimented for an additional hour at 196,000 X g. Only the top half (100 pl) of the supernatant was removed and stored at 4 "C while the remainder was discarded. The pellet was next resuspended in 200 pl of buffer A. These two preparations were termed the supernatant and pellet, respectively. The preparations were then tested for hyaluronic acid synthetic activity as described under "Experimental Procedures." No activity was found in any of the supernatants, and the results below show the detergent concentration at which optimum activity was obtained in the pellet. The total synthetic activity of whole membrane was 108.3 nmol glucuronic acid transferred from UDP-glucuronic acid/h/mg of protein with a specific activity of 38.1. The values in parentheses are the percentage of activity as compared to intact membrane. TABLE I1 Hyaluronic acid synthesis activity enhancement through the addition of extracts from group C streptococcal membranes Detergent extracts were prepared as described in Table I. Addition experiments were performed by taking 125 pl of the respective extract, 125 pl of the additive, and dialyzing the mixture against three changes of buffer A over 48 h. After dialysis, 50-pl aliquots were analyzed for their hyaluronic acid synthesis activity as described under "Experimental Procedures." All assays were done in triplicate. e Trypsin; treated with trypsin (2 mg/ml) for 2 h at 37 "C at which time soybean trypsin inhibitor ( 3 mg/ml) was added to stop the reaction.

Digitonin
Total lipid extract was prepared by the method of Bligh and Dyer as described under "Experimental Procedures." 23.3% of the total activity was recovered when this activity was added to the total activity shown in Table I. Similar results were obtained for the CHAPSO supernatant after dialysis with a specific activity of 18.5 nmol transferred/h/mg of protein and a total activity of 6.6 nmol transferred/h (6.1% of intact membrane). After dialysis of the extracts from the other detergent solubilizations, no activity was observed. Therefore, the possibility existed that different components of the same system were being partially extracted from the membrane by each detergent and that the incomplete hyaluronic acid synthetic system was present in the supernatant. In order to test the hypothesis that the hyaluronic acid synthetic activity was dissociated by the extraction systems, the CHAPSO and digitonin extracts (196,000 X g supernatants) were added together, and dialyzed against buffer A, and then assayed for hyaluronic acid synthetic activity. An enhancement of the total activity (Table 11, 29.4 nmol transferred/h, 147.7%) was observed when compared to the total activity for the CHAPSO and digitonin extracts (6.6 and 13.3 nmol/h, respectively).
Each of the extracts was treated with trypsin to determine whether the active component was protein in nature. After the trypsin treatment of the digitonin extract, 71.8% (from 29.4 to 8.3) of the activity was lost. However only 35% (from 29.4 to 19.1) of the activity was lost when the CHAPSO supernatant was treated with trypsin. Since the data indicated that the CHAPSO extract contained a trypsin-resistant component of the hyaluronic acid synthetic system possibly of a lipid nature, a total lipid extract of the CHAPSO supernatant was prepared and added to the digitonin supernatant. A total activity of 287.9 nmol/h was observed for this preparation. In comparison to intact membranes, the lipid-reconstituted digitonin extract produced 266% of the total activity with a specific activity 12.6 times greater than the membrane, indicating that components of the lipid fraction of the membrane were required for total enzymatic activity. No synthetase activity was demonstrated associated with the lipid extract.
Next, in order to demonstrate that the hyaluronic acid synthetic activity was indeed soluble, a digitonin 196,000 X g supernatant was loaded onto a column of Sepharose CL-GB and eluted with sodium phosphate buffer containing 0.2% digitonin (Fig. 1). The enzymatic activity eluted after the void volume. Even though an accurate molecular size was not determined due to the presence of digitonin, it appears that the activity resides as an enzyme complex.
Reconstitution of Hyaluronic Acid Synthetic Activity with Phospholipids-Since a total lipid extract of the CHAPSO extract reconstituted the hyaluronic acid synthetic activity of the digitonin extract, attempts were made to reproduce the reconstitution using phospholipids. A range of phospholipid/

Effectiveness of different phospholipids in reconstituting hyaluronic acid synthesis activity in digitonin extracts from group C streptococcal
membranes Digitonin extracts were prepared as described in Table I. Reconstitution with phospholipids was carried out as described for addition experiments in Table 11. protein ratios (pglpg) from 0.1:l to 50:l in buffer A containing 0.45% CHAPSO were tested for each phospholipid, and the data are presented in Table 111. Cardiolipin, phosphatidylethanolamine, and phosphatidylserine reconstituted the hyaluronic acid activity of the digitonin supernatant (529.3, 207.6, and 237.7 nmol/h, respectively, as compared to the membrane control 108.3 nmol/h). Furthermore, cardiolipin enhanced the total activity of the digitonin supernatant 4.9-fold over the activity obtained from intact membranes with a 63.1-fold increase in specific activity. Phosphatidylcholine, phosphatidylglycerol, and phosphatidylinositol increased the total activity of the digitonin supernatant but to a more limited extent. Combinations of phospholipids did not enhance the activity over that of cardiolipin reconstituted system.

Optimum
In order to determine whether digitonin was the optimum detergent for extraction of the hyaluronic acid synthetic system, extraction of membranes with other detergents was repeated and the 196,000 X g supernatants were reconstituted with the optimum concentration of cardiolipin (Table IV).

Effectiveness of reconstitution of hyaluronic acid synthesis activity with cardiolipin for different detergent extracts of group C streptococcal membranes
The optimum cardiolipin/protein ratio (pg/pg) was 5:l for all detergent extracts. Detergent extracts were prepared as described in Table I. Reconstitution with cardiolipin was carried out as described for addition experiments in Table 11.

Detergent
Optimum Cardiolipin did not reconstitute any of the other 196,000 X g supernatants when various detergents and detergent concentrations were tested (2.5-24.4 nmol/h total activity versus 108.3 nmol/h for intact membrane). This experiment was repeated using phosphatidylethanolamine-reconstituted detergent extracts and a similar pattern of synthetic activity was observed.
Confirmation of Hyaluronic Acid Synthetic Activity Product-In order to determine the molecular weight range of the product of phospholipid-reconstituted hyaluronic acid synthetic activity, the sodium dodecyl sulfate-released product (see "Experimental Procedures") was chromatographed over Sepharose 2B (Fig. 2). The elution profile shows that high molecular weight product eluted at the void volume of the column and that a range of smaller molecular weight products were also produced. When the fractions containing [14C]glucuronic acid were pooled, treated with hyaluronidase (see "Experimental Procedures"), and rechromatographed, all of the radioactivity was found to elute at the total volume (Fig.  2). This demonstrated that [14C]glucuronic acid was incorporated into high molecular weight hyaluronic acid by the phospholipid-reconstituted system. Rate of Product Formation by Reconstituted Hyaluronic Acid Synthetic Activity-In the past Stoolmiller and Dorfman ( 5 ) and Sugahara et al. (2) demonstrated that the streptococcal hyaluronic acid synthetic system produced both trichloroacetic acid-insoluble and -soluble hyaluronic acid. Fig. 3 shows that this also is the case with the cardiolipin-reconstituted system. The production of trichloroacetic acid-insoluble material reaches a plateau at 30 min, whereas production of the soluble fraction reaches a plateau at 70 min. No appreciable soluble material was observed before 7 min.

DISCUSSION
The studies presented demonstrate that the hyaluronic acid synthetic system was solubilized as an enzyme complex by extraction of streptococcal membranes with 2% digitonin and reconstituted with cardiolipin for optimum activation. Previous investigators of this synthetic system have attempted solubilization with detergents (6, 17) and organic solvents (2, 18) but were unsuccessful. Because of this failure to solubilize the system, many facets of the production mechanism remain incompletely understood. For example the involvement of a lipid intermediate is uncertain in eucaryotes (11,12). In addition, the role of primer, the number of proteins involved in the synthesis, the regulation of synthesis, and the mecha- acid labeled hyaluronic acid produced by digitonin-extracted and phospholipid-reconstituted membranes from group C streptococci. Membranes were prepared, solubilized, and relipidated as described under "Experimental Procedures." Reaction mixture was added to the reconstituted extract and incubated at 37 "C for 1 h. The reaction was stopped by the addition of sodium dodecyl sulfate (l%, w/v) and incubated at 37 "C for an additional hour. This mixture was then centrifuged at 196,000 X g for 1 h, and the supernatant was dialyzed against three changes of distilled water over 48 h. The dialyzed preparation was then applied to a column (48 X 1.5 cm) of Sepharose 2B equilibrated with 0.5% sodium dodecyl sulfate. Fractions (2 ml) were collected at a flow rate of 8 ml/h, and aliquots (0.5 ml) were assayed for radioactivity. Fractions containing [14C]glucuronic acid were precipitated with acetone, resuspended in 0.1 M sodium phosphate buffer, pH 5.3, with 0.15 M NaCl, and digested at 37 "C for 1 h with bovine testes hyaluronidase at a final concentration of 0.05 mg/ml. The digest was then stopped with sodium dodecyl sulfate (1% w/v) and applied to the column as described above. Fractions (2 ml) were collected and tested for radioactivity. The elution profiles for intact ( U ) and digested (W) hyaluronic acid are shown.
nism of alternate addition of monosaccharides has not been totally resolved.
The initial experiments on detergent solubilization showed no activity in 196,000 X g supernatants from any of the detergents tested which confirmed previous reports (6). Analysis of our detergent extraction data indicated that octyl glucopyranoside, Triton X-100, and zwittergent 314 either denatured or inhibited the enzymatic activity of the hyaluronic acid biosynthetic system. CHAPSO did not extract any appreciable enzymatic activity, but it did enhance the activity of the 13,000 x g and 196,000 X g insoluble pellets 2.7-fold as compared to the activity of the intact membrane. The specific role of CHAPSO in this enhancement of synthetic activity remains to be determined even though it is likely that CHAPSO is opening up membrane vesicles, thereby allowing the enzyme access to the substrates.
The inhibitory effect of a number of detergents was still apparent upon dialysis of the detergent extracts against buffer A. The digitonin extract exhibited the greatest quantity of hyaluronic acid synthetase activity when compared to the other detergent extracts but was still low in comparison to intact membranes (12.3%). However, addition of cardiolipin reconstituted the system completely and showed an enhancement of 4.9-fold in total activity over that of intact membranes. When other detergents were used in an effort to solubilize the hyaluronic acid synthetic activity from the membrane and the extracts were reconstituted under optimal conditions, digitonin was found to be 21.7 times more effective than the next detergent, CHAPSO. The mechanism of digitonin solubilization of eucaryote membranes is through the removal of sterols. However, this class of molecules is not glucuronic acid into trichloroacetic acid-soluble and -insoluble hyaluronic acid using a cardiolipin-reconstituted digitonin extract of group C streptococcal membranes. Cardiolipin-reconstituted digitonin membrane extracts were incubated at 37 "C with reaction mixture, and the reaction was stopped by the addition of cold 6.6% trichloroacetic acid at the times indicated. Incorporation into the trichloroacetic acid-insoluble hyaluronic acid was determined as described under "Experimental Procedures." Incorporation of UDP-['4C]glucuronic acid into trichloroacetic acid-soluble hyaluronic acid was determined by the quantitation of the radioactivity remaining after dialysis of the trichloroacetic acid-soluble supernatant against five changes of water over 72 h. Polynomial least squares regression analysis was used to generate the curves from the experimental data. Total incorporation of [14C]glucuronic acid (c".)

as well as incorporation into trichloroacetic acid-insoluble (o"--o) and -soluble (A-A)
hyaluronic acid are shown.
Minutes present in streptococcal membranes, and, therefore, digitonin must solubilize the hyaluronic acid synthetic activity through another mechanism. Aloni et al. (19) recently reported that digitonin solubilized cellulose synthetase from Acetobacter xylinum, another microorganism lacking sterols, indicating that other mechanisms of membrane solubilization by digitonin occurred. The studies by Sugahara et al. (2) and Stoolmiller and Dorfman (1) demonstrated that a lipid intermediate was not involved in the biosynthesis of streptococcal hyaluronic acid, however, the stabilizing role of phospholipids was not recognized. Our studies indicated that phospholipids were necessary for the expression of synthetic activity by this system.
Cardiolipin, phosphatidylethanolamine, and phosphatidylserine reconstituted and enhanced the total activity of the digitonin extract over that of the intact membrane. Combinations of these phospholipids did not improve the enzymatic activity of the cardiolipin-reconstituted system. The reconstitution of the enzyme complex with these phospholipids is not unexpected when one considers that Gram-positive bacterial membranes usually contain phosphatidylglycerol, cardiolipin, traces of phosphatidylserine, and occasionally phosphatidylethanolamine (20,21). Phospholipids have previously been demonstrated to play a role in the activation and stabilization of membrane proteins. For example, in E. coli, the enzymes producing peptidoglycan (22), and in Salmonella typhimurium, the galactosyl transferase system (23) require phospholipid for activity. The phospholipid dependence of a guinea pig liver microsomal UDP-glucuronosyltransferase has also been demonstrated (24). Investigations into the size of the product of the reconstituted hyaluronic acid synthetic enzyme system showed that a variety of high molecular weight products were synthesized.
The variation in polymer sizes was consistent with the release of chains of polymer at different stages of completion. The size range also was similar to the hyaluronic acid produced in whole cells of group A and C streptococci and their membranes. The kinetic studies on the production of hyaluronic acid using the reconstituted soluble system were similar to results presented by Sugahara et al.
Finally, the ability to solubilize and reconstitute the hyaluronic acid synthetic system is an important step in the purification of this enzyme. The 63-fold increase in specific activity of the cardiolipin-reconstituted hyaluronic acid synthetic system over intact membranes indicates that this is a suitable starting point in purification of the complex. With a purified system the complete mechanism of hyaluronic acid biosynthesis in procaryotes and eucaryotes can be elucidated.