Isolation of chitin synthetase from Saccharomyces cerevisiae. Purification of an enzyme by entrapment in the reaction product.

Chitin synthetase, in the zymogen form, was extracted with digitonin from a particulate fraction from Saccharomyces cerevisiae and converted into active form by treatment with immobilized trypsin. When the activated enzyme was incubated with UDP-GlcNAc and other components of an assay mixture, a chitin precipitate formed, trapping a large portion of the synthetase. The enzyme was easily extracted frm the chitin gel with a recovery of approximately 50% and an enrichment of approximately 100-fold. Further purification was obtained by repeating the chitin step. After polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate, the purified synthetase showed a major band corresponding to Mr 63,000, a weaker band at Mr 74,000, and some other minor bands. Under nondenaturing conditions, an Mr of 570,000 was calculated for the enzyme from Stokes radius and sedimentation coefficient determinations. After electrophoresis in a nondenaturing gel and incubation with the components of the standard assay, chitin was formed and precipitated in the gel, yielding an opaque band. Soluble oligosaccharides were not precursors for insoluble chitin, suggesting that synthesis of chitin chains takes place by a processive mechanism. N-Acetylglucosamine stimulated the purified synthetase only slightly and did not participate as a primer in the reaction. The same chain length, somewhat more than 100 units of GlcNAc, was determined in samples of chitin that had been synthesized either in vivo, or with a membrane preparation or with purified synthetase. These results suggest that chitin synthetase itself is capable both of initiating chitin chains without a primer and of determining their length.

Chitin synthetase, in the zymogen form, was extracted with digitonin from a particulate fraction from Saccharomyces cerevisiae and converted into active form by treatment with immobilized trypsin. When the activated enzyme was incubated with UDP-GlcNAc and other components of an assay mixture, a chitin precipitate formed, trapping a large portion of the synthetase. The enzyme was easily extracted from the chitin gel with a recovery of approximately 50% and an enrichment of -100-fold. Further purification was obtained by repeating the chitin step. After polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate, the purified synthetase showed a major band corresponding to M, 63,000, a weaker band at M, 74,000, and some other minor bands. Under nondenaturing conditions, an M, of 670,000 was calculated for the enzyme from Stokes radius and sedimentation coefficient determinations. After electrophoresis in a nondenaturing gel and incubation with the components of the standard assay, chitin was formed and precipitated in the gel, yielding an opaque band.
Soluble oligosaccharides were not precursors for insoluble chitin, suggesting that synthesis of chitin chains takes place by a processive mechanism. N-Acetylglucosamine stimulated the purified synthetase only slightly and did not participate as a primer in the reaction. The same chain length, somewhat more than 100 units of GlcNAc, was determined in samples of chitin that had been synthesized either in vivo, or with a membrane preparation or with purified synthetase. These results suggest that chitin synthetase itself is capable both of initiating chitin chains without a primer and of determining their length.
Polysaccharides are major components of the cell wall in many procaryotic and eucaryotic organisms. Several enzymatic activities which participate in the synthesis of structural polysaccharides have been detected but little progress has been made in the purification of the enzymes (1). Consequently, the molecular mechanisms for the synthesis of these important biological products are not as well known as those of other cellular components. Among cell-wall polysaccha-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Scholar in Residence, Fogarty International Center, National Institutes of Health. Present address, Massachusetts Institute of 02139.
Technology, Center for Cancer Research, E17-233, Cambridge, MA § To whom correspondence should be addressed. rides, chitin is one of the most common because of its widespread occurrence in fungi (2). In Saccharomyces cereuisiae it has a special role as the major or sole component of the primary septum (3,4). This structural component of the cell has been used in our laboratory as a model for the study of morphogenesis (4). In the course of those investigations it was discovered that chitin synthetase, a plasmalemma-bound enzyme, is present in the cell as a zymogen and can be converted into an active form by partial proteolysis (5, 6). Recently, we found that chitin synthetase catalyzes a unidirectional synthesis of chitin in the plasma membrane (7). To elucidate the relationship between zymogen and active enzyme and to understand the functioning of the synthetase in the membrane, it was necessary to purify the enzyme. Although chitin synthetase was solubilized with digitonin from both Coprinus (8) and Saccharomyces (6) several years ago, efforts in our laboratory to further purify the enzyme with many conventional procedures met with failure. Finally, we were able to take advantage of the fact that the solubilized enzyme gives rise to an insoluble product (6). A considerable amount of synthetase is trapped in the chitin formed in the reaction and can be easily eluted from it. This is the crucial step in the procedure for isolation of the enzyme that we describe in this report.
Preparation of Immobilized Trypsin-Trypsin (400 mg) was coupled to activated CH-Sepharose 4B according to the manufacturer's directions. After coupling, no trypsin could be detected in the supernatant with the Hide Powder Azure assay (1 1). The gel was stored at 5 'C in 50 mM Tris-chloride, pH 7.5, containing 0.02% sodium azide. Before and after use, the gel was washed twice with 50 mM Trischloride, pH 7.5. The same preparation of trypsin-Sepharose could be used at least 10 times for activation of chitin synthetase zymogen without apparent loss of activity. In fact, aged preparations often performed better than fresh ones.

Methods
Yeast Strain and Culture Conditions-The organism was a diploid strain of S. cereuisiue, strain MatalMata prbl-I122/prbl-l122 (12), that lacks proteinase B. Cells were grown in 200 liters of minimal medium (2% glucose, 0.7% yeast nitrogen base (Difco)) in a fermentor at the pilot plant facility of the National Institutes of Health. Cells were harvested in late logarithmic to early stationary phase, when the absorbance at 660 nm, as measured with a Coleman Junior Spectrophotometer with a light path of 1 cm, was between 0.4 and 0.5. The yield was about 5 g of yeast, wet weight, per liter of culture.
Cell Disruption and Initial Purification Steps-Operations were performed at 0-5 "C, unless otherwise indicated. Yeast cells were processed in batches of approximately 0.5 kg. In a typical preparation, 557 g of cells, wet weight, were suspended in Buffer A (20 mM Trischloride, pH 7.5, containing 2 mM magnesium acetate) to a final volume of 1200 ml. A 200-ml portion of the suspension was added to 306 g of glass beads (0.5-mm diameter, B. Braun, Melsungen, West Germany) in the vessel of a Bead-Beater (Biospec Products, Bartlesville, OH), with ice cooling. The Bead-Beater was operated for two 4min periods, with a 1-min cooling period in between. Cell breakage was usually about 90%. The extract was aspirated from the glass beads with a long-tipped Pasteur pipette connected to an evacuated filter flask, and the beads were washed several times with small portions of Buffer A. Each portion of beads was used to disrupt three 200-ml batches of cell suspension. The final volume of the extract was 1570 ml.
The crude extract was centrifuged at 100,000 X g for 30 min. The supernatant liquid was discarded and the pellet was suspended with a Dounce homogenizer in Buffer A to a final volume of 1 liter. After centrifugation as above, the supernatant fluid was discarded, and the pellet was suspended in a solution containing 1% digitonin (w/v), 20 mM Tris-chloride, pH 7.5,5 mM magnesium acetate, and 0.2 M NaCl, to a final volume of 1600 ml. After homogenization with a Dounce homogenizer, the suspension was incubated at 30 "C for 45 min with shaking, then centrifuged as above. The supernatant liquid, 1480 ml, containing the solubilized chitin synthetase zymogen was stored at -80 "C. The next step, filtration through Sephadex G-75, was performed with 150 ml of solubilized zymogen each time. The column, 4 X 85 cm, was equilibrated with 20 mM Tris-chloride, pH 7.5, containing 0.1% digitonin; the same buffer was used for elution. The absorbance of the effluent at 280 nm was recorded with an LKB UV-monitor and the large peak emerging at the void volume was collected in a total volume of 220-250 ml.
Activation of Chitin Synthetase Zymogen with Immobilized Trypsin-The Sephadex G-75 eluate was mixed with 20-30 ml (settled gel volume) of trypsin-Sepharose 4B and the suspension was incubated with shaking for 30 min at 30 "C. The trypsin-Sepharose was pelleted in a clinical centrifuge and washed with 15 ml of 0.05 M Tris-chloride, pH 7.5. The wash fluid was added to the first supernatant liquid and the pooled solutions were concentrated to 80 ml in an ultrafiltration cell fitted with an Amicon XM-100 membrane.
The amount of immobilized trypsin required for optimal activation was ascertained in small scale trials for each batch of zymogen. With some batches, it was necessary to add more digitonin to the zymogen to obtain maximal activation with immobilized trypsin. No such requirement was found for activation with soluble trypsin, suggesting that the zymogen molecules had aggregated and had become partly inaccessible to the trypsin-Sepharose beads. The additional detergent may have been required to break up the aggregates. When additional digitonin was needed, the final concentration was -0.35%. Chitin-entrapment Step-To the bottom of each of four centrifuge tubes for the Beckman SW27 rotor were added 6 ml of a mixture containing 1 2 % glycerol, 6 mM UDP-GlcNAc, 40 mM N-acetylglucosamine, and 4 m M magnesium acetate. To 20 ml of concentrated enzyme from the previous step were added 3 ml of 50 mM UDP-GlcNAc, 1.2 ml of 0.8 M N-acetylglucosamine, and 0.1 ml of 1 M magnesium acetate. After mixing, the solution was immediately layered on the glycerol-containing cushion. The tubes were incubated for 15 min in a 30 "C bath, before being placed on ice for 2.5 h, A precipitate of chitin appeared almost immediately after mixing the components of the reaction and became heavy and flocculent subsequently. The tubes were centrifuged in an SW27 Beckman rotor for 25 min at 20,000 rpm (53,000 X g). The supernatant fluid was carefully aspirated and the pellets were stored overnight at -80 "C. For extraction, each of the four pellets was suspended in 10 ml of 20 mM Trischloride, pH 7.5, containing 0.1% digitonin and homogenized with a Dounce homogenizer. After incubation for 5 min at 30 "C, the suspension was centrifuged for 10 min at 16,000 X g and the supernatant fluid was saved.
When so desired, the chitin-purification step was repeated immediately in the same manner as the first one, but halving all quantities and using tubes for the Beckman SW40 rotor. The purified enzyme was stored in polypropylene tubes at -80 "C.
Chitin Synthetase Assay-The standard assay mixture contained 30 mM Tris chloride, pH 7.5, or imidazole acetate, pH 7, 32 mM Nacetylglucosamine, 4 mM magnesium acetate, 0.18 mg/ml of phosphatidylserine, 1 mM UDP-N-a~etyl['~C]glucosamine (4 X lo6 cpm/ pmol), and enzyme, in a total volume of 50 pl. With enzyme purified beyond the chitin-entrapment step, the mixture also contained 0.3 mg/ml of bovine serum albumin and 0.3 mg/ml of digitonin, and incubation was carried out in 1.5-ml Eppendorf polypropylene tubes. Incubation was at 30 "C for 30 min. Usually, the reaction was stopped with 10% trichloroacetic acid and the radioactivity in insoluble material was counted after filtration through glass-fiber filters, as described for P(1-3)glucan synthetase (13). With preparations that contained very little protein, activity was often determined by counting the material that remained at the origin after chromatography of the reaction mixture on Gelman silica-impregnated glass fiber strips (14). For determination of activity in zymogen preparations, the synthetase was activated with trypsin prior to assay (6).
The formation of soluble products (oligosaccharides) in chitin synthetase incubation mixtures was monitored by first adsorbing the remaining substrate on an anion-exchange resin and then determining the soluble radioactivity after filtration through glass-fiber filters Polyacrylamide Gel Electrophoresis-For SDS' gels, the procedure of Laemmli (16) was used. The concentration of acrylamide was 4% for the stacking gel and 12.5% for the running gel. The thickness of the gels was 0.75 mm for silver staining and 1.5 mm for Coomassie staining. Electrophoresis in nondenaturing gels was carried out at pH 8.9, with an acrylamide concentration of 3% for the stacking gel and 5% for the running gel, as described by Maize1 (17). Gels were stained overnight with a solution containing 0.5% Coomassie Brilliant Blue R, 35% methanol, and 10% acetic acid and destained with 35% methanol containing 10% acetic acid. For silver staining, the method of Merril et al. (18) was used (Silver Stain Kit from Bio-Rad). The Kodavue (Eastman Kodak) staining procedure was carried out according to the manufacturer's directions.
Preparation of Chitin in Vivo and in Vitro-For the isolation of chitin formed in uiuo, yeast cells were disrupted with glass beads in a Bead-Beater as described for enzyme purification but with a 30-ml adapter. From 7.5 g of yeast, wet weight, 40 ml of broken cell suspension in 50 mM Tris-chloride, pH 7.5, were obtained. The cell walls were sedimented by centrifugation at 3000 X g for 10 min and washed four times with water. To an aqueous suspension containing cell walls from 7 g of cells, wet weight, in a total volume of 7 ml, 340 ml of 50 mM Tris-chloride (pH 7.5), and 3 ml of Zymolyase 60,000 (7.5 mg/ml in 0.2 M potassium phosphate at pH 7) were added. The suspension was incubated at 37 "C with shaking and the absorbance at 660 nm was monitored. After 45 min, the absorbance had decreased to about 2% of the original value. Inspection under the phase-contrast microscope showed the presence of only septa and some cell-wall fragments. The suspension was centrifuged at 12,000 X g for 10 min. The pellet was washed twice with 40 ml of 1% SDS and three times with 40 ml of water, and finally suspended in water.
For the preparation of chitin by membranes, 5.6 g of yeast cells, wet weight, were disrupted with glass beads in the Bead-Beater (30ml adapter) as for the purification of chitin synthetase. Cell walls were sedimented by centrifugation of the crude extract for 10 min at 3000 X g, and membranes were sedimented from the supernatant fluid for 30 min at 100,000 X g. The membrane pellet was washed once in 20 mM Tris-chloride, pH 7.5, containing 2 mM magnesium acetate and resuspended in the same buffer to a final volume of 9.5 ml. Of this suspension, 2 ml were treated with trypsin to activate chitin synthetase (6) and then incubated for 15 h at 30 "C with UDP-GlcNAc in a scaled-up chitin synthetase assay mixture (see above; total volume, 40 ml, with sodium azide added to a final concentration of 0.02%). From a parallel incubation with UDP-["C]GlcNAc, 6.4 synthesized. After centrifuging for 10 min at 16,000 X g, the pellet pmol of chitin, as N-acetylglucosamine, were calculated to have been (15).
The abbreviation used is: SDS, sodium dodecyl sulfate.

Yeast Chitin Synthetase
was suspended in 20 ml of 1% SDS and the suspension was placed for 5 min in a boiling water bath, followed by centrifugation. The pellet was washed three times with water and resuspended in water.
Chitin synthesized by purified enzyme was that formed in the second chitin purification step of the synthetase. The polysaccharide was used after extraction of chitin synthetase and further washing with water.
Chitin Chin-length Measurement-The same technique was used for the different samples of chitin described in the previous section. In a typical experiment, 1 mg of chitin suspended in 100 pl of water was treated with 100 pl of NaB3H4 (1 mCi, 341 mCi/mmol) in 0.01 N NaOH at room temperature for 6 h. The reaction was terminated by addition of 1 N acetic acid (50 pl) and the sample washed by repeated centrifugation followed by resuspension in 0.05 M potassium phosphate at pH 6.3. The washed, reduced chitin was treated overnight at 30 "C with 100 pl (138 pg) of chitinase from Serratia (9) in a final volume of 300 pl. Any insoluble residue was removed by centrifugation. The supernatant fluid was applied to a Bio-Gel P-2 (200-400 mesh) column (1 X 90 cm), previously equilibrated with 0.2 N acetic acid, and eluted with 0.2 N acetic acid. Fractions of 0.45 ml were collected. Fig. 1 shows a typical separation of the Serratia chitinase digestion products. As can be seen, the Serratia enzyme generates an approximately equimolar mixture of reduced di-and trisaccharide from borotritide-reduced chains. Although the reason for this equimolar mixture is not completely clear, it probably reflects a requirement of the enzyme for an intact triacetylchitotriosyl group as minimum effective substrate. Although the Serratia preparation hydrolyzes the nonreduced disaccharide slowly, this reaction is probably the result of contamination by a specific diacetylchitobiase (9).
Aliquots of each fraction from the column were taken for measurement of radioactivity. For determination of N-acetylglucosamine, 15 pl of 1 M NaZHPO, was added to 100 pl of each fraction, followed by 5 p1 of Cytohelicase (diluted 3-fold) as a source of 8-N-acetylghcosaminidase. The mixture was incubated for 45 min at 30 "C and Nacetylglucosamine was measured colorimetrically (19). Since the reduced trisaccharide is hydrolyzed completely to a mixture of Nacetylglucosamine and reduced disaccharide by Cytohelicase (data not shown), the specific activity of the reduced trisaccharide, and, thus, the specific activity of N-acetylglucosaminitol could be determined directly. Chain lengths were calculated from the N-acetylglucosaminitol specific activity and a specific activity obtained by dividing total recovered radioactivity by total recovered N-acetylglucosamine. The latter was the sum of the free monosaccharide peak plus the N-acetylglucosamine liberated by Cytohelicase from the disaccharide peak. The same chain-length values could be obtained by direct measurement of radioactivity and N-acetylglucosamine (after Cytohelicase digestion) in the chitinase digest. In this case, it was necessary to determine the correct specific activity of sodium borotritide, which is known to contain radioactive impurities. For this purpose, triacetylchitotriose was reduced with excess borotritide. Following Bio-Gel P-2 chromatography, the specific activity of pure reduced trisaccharide was determined by measuring radioactivity and N-acetylglucosamine following Cytohelicase digestion. Since the same lot of borotritide was used for each experiment, the specific activity was found in each case to be the same.

RESULTS
Purification of Chitin Synthetase-In the present study, yeast cells were disrupted with glass beads, because conversion of cells to protoplasts, as used in previous experiments prior to lysis (6), is not suitable for large-scale preparations. The yield of enzymatic activity in a crude extract was somewhat greater than in protoplast lysates. Soluble proteins were removed by centrifugation and the particulate material was extracted with digitonin to solubilize chitin synthetase, (for details, see "Methods"). In early preparations, the crude extract was first centrifuged at low speed to remove cell walls, but this step was later discontinued because in some batches a considerable amount of membrane and of chitin synthetase sedimented with the walls. The digitonin extract was filtered through a Sephadex G-75 column, and the enzyme was recovered in the void volume fraction.
Activation of the zymogen form of the enzyme had usually been performed with trypsin solution followed by addition of soybean trypsin inhibitor. Recently, it was found possible to use immobilized trypsin attached to Sepharose 4B in the activation step. In this way, the activation is terminated by centrifuging the immobilized trypsin, thus avoiding addition of inhibitor. It seems also probable that the use of immobilized trypsin, because of steric constraints, will be less injurious to the enzyme molecules. Activation, however, takes place readily with the immobilized protease, presumably because the bond(s) broken in this process are in an exposed portion of the polypeptide chain.
When the activated enzyme was incubated with UDP-GlcNAc in a standard assay mixture, insoluble chitin appeared as a flocculent precipitate (6). On centrifugation, about half of the chitin synthetase activity sedimented with the chitin and could be recovered after extraction of the precipitate with Tris buffer. Sedimentation of the enzyme was not simply caused by affinity for chitin. Addition to the synthetase preparation of either "regenerated" chitin (20) or enzymatically prepared chitin, followed by centrifugation, did not lead to sedimentation of the enzyme (Table I).
We could not find conditions for washing the chitin pellet without extracting the synthetase. Therefore, the reaction mixture was layered, in a centrifuge tube, on a cushion containing 12% glycerol. During incubation, the flocculated chitin accumulated in the upper part of the tube. During subsequent centrifugation it was partially washed by passage through the glycerol layer. The best incubation conditions for maximal yield were found to be 15 min at 30 "C followed by storage for 2-3 h on ice, which corresponded to a conversion of about two-thirds of the substrate into product.
The chitin pellet obtained after centrifugation had a gellike appearance. After freezing at -80 "C overnight, thawing, suspending in buffer, and centrifuging, the pellet appeared to be much smaller. Freezing and thawing of the pellet before extraction increased the yield of chitin synthetase from -35% to -50% of the initial activity. The total recovery of activity (extracted pellet + supernatant) was usually about 75%.

Yeast Chitin Synthetase
Nevertheless, it was difficult to measure accurately chitin synthetase in the supernatant fluid, because of the presence of UDP, an inhibitor (21), and of remaining UDP-GlcNAc which diluted the radioactive substrate added.
Although 0.1% digitonin was routinely included in the buffer used to extract the enzyme from chitin, omission of the digitonin in some experiments did not lead to a substantial change either in the yield or in the purity of the synthetase.
A summary of one of the best preparations is given in Table  11. As shown in the table, the chitin-entrapment step can be repeated for further purification. This should be done immediately after extracting the enzyme from chitin in the first step. Storage of the enzyme a t -80 "C followed by repetition of the chitin step resulted in very poor recoveries, although the activity had not decreased during storage. In 29 preparations, the average purification in the first chitin-entrapment step was 97-fold. In the 12 preparations in which the second chitin step was carried out, the average purification in that step was &fold. Therefore, the average overall purification in both steps combined was about 500-fold, as compared with 630-fold for the preparation of Table 11.
The activity of purified enzyme was quite stable for several months at -80 " C . Upon lyophilization, about 70% of the activity was recovered.
Polyacrylamide Gel Electrophoresis of the Purified En-

TABLE I
Effect of chitin source on chitin synthetase precipitation The procedure was the same as described for the chitin-purification step under"Methods," except that the volume of activated and concentrated G-75-peak enzyme added was 7 ml (total activity, 500 milliunits) and all the other volumes were reduced in proportion. UDP-N-acetylglucosamine was omitted in tubes 2 and 3, which received preformed chitin. The amount of "regenerated" (20) or enzymatically obtained chitin (7.5 mg) was approximately that expected to be formed in reaction mixture 1. The "supernatant" resulted from centrifugation of the synthesized or preformed chitin. The "extract from pellet" was obtained after freezing the chitin pellets overnight, as outlined under "Methods." zyme-Electrophoresis in SDS gels of purified preparations gave rise to a major and a minor band after staining with Coomassie Blue (Fig. 2), corresponding to molecular weights of 63,000 and 74,000, respectively. Staining with silver salts led to the appearance of only the 63,000-dalton band after a short period of development. The higher-molecular-weight band, as well as several minor bands, appeared after further development (Fig. 2). Identical results were obtained with the Kodavue staining procedure. With all reagents used, the intensity of staining was much weaker than expected on the basis of the amount of protein added. We have no explanation for this observation, although it could be related to binding of digitonin to the hydrophobic proteins.
An interesting observation was made after gel electrophoresis of the purified enzyme in nondenaturing gels. When the gels were incubated in a chitin synthetase assay mixture, an opaque band appeared (Fig. 3A). The band became fluorescent after soaking the gel in a fluorescein isothiocyanate-wheat germ agglutinin solution (Fig. 3B). When I4C-labeled UDP-GlcNAc was used as substrate, and the gel was sliced and counted, a peak of radioactivity was detected in coincidence with the visible band (Fig. 3C). Furthermore, when a homogenate of the pertinent gel slices was treated with purified chitinase (9), the same proportion of radioactivity was solubilized as from a sample of enzymatically synthesized chitin (not shown). If substrate was omitted from the reaction mixture, no band appeared. We conclude that the opaque band consists of chitin, whose synthesis is catalyzed by the enzyme present in the gel.
Staining of nondenaturing gels with Coomassie Blue was very poor, usually yielding a wide band in coincidence with the chitin band. It seems, furthermore, that the proteins in the preparation do not separate well under these conditions. An enzyme purified through the first chitin step was submitted to electrophoresis in nondenaturing conditions, and an area corresponding to the chitin band was extracted with SDS-mercaptoethanol and electrophoresed again in an SDS gel. Several bands appeared after staining, an indication that the separation in the nondenaturing gel had been poor.
Hydrodynamic Properties of Purified Chitin Synthetme-The Stokes radius of the purified enzyme was measured by gel filtration and the sedimentation coefficient calculated after centrifugation in sucrose gradients (22). By the use of Svedberg's equation, a molecular weight of 570,000 was calculated (Table 111). The significance of this value is somewhat in doubt, because the enzyme appears to aggregate to a vari- In the first three steps, chitin synthetase activity was measured after activation with optimal amounts of An aliquot of the crude extract was separately centrifuged and the pellet was suspended in Tris buffer for Values in parentheses have been recalculated as if the whole preparation had been used in the chitin-trypsin (6). assay. entrapment steps. In fact, only about 10% of the total was used each time. c--. " "_. able extent, as a function of the concentration of digitonin (6).
Kinetic Properties of Purified Chitin Synthetase-The apparent K,,, for UDP-GlcNAc was 0.7 mM, compared to values between 0.6 and 1.5 mM, as obtained earlier with crude preparations of the enzyme (6, 21). Both phosphatidylserine and digitonin were required for maximal activity. In the presence of digitonin (0.5 mg/ml), a lower concentration of phosphatidylserine (0.05 mg/ml) was required than when the phospholipid was added alone; higher concentrations of phosphatidylserine were inhibitory. The dependence on these compounds varied from preparation to preparation. Usually, the enzymatic activity was increased about 3-fold in the presence of both phosphatidylserine and digitonin. In some cases, the enzyme showed practically no activity in the absence of both substances (data not shown).
A somewhat surprising result was that the purified enzyme had lost its dependence on addition of free N-acetylglucosamine almost completely (21). Omission of the acetylamino sugar from the reaction mixture decreased the activity by only 13%.
Formation of Oligosaccharides-At low concentrations of UDP-GlcNAc, (less than 30 p~) no insoluble chitin was formed (Fig. 4, and inset). Instead, N-acetylglucosamine was incorporated into water-soluble products. Chromatography on a Bio-Gel P-4 column and incubation with purified chitinase showed these Droducts to be a series of B(l-A)-linked chito-  * Determined by sedimentation in sucrose gradients (22). Standards, with sedimentation coefficient in parentheses, were: bovine liver catalase (1 1.2), apoferritin (17.6), urease (18.6). and a2-macroglobulin (19.5). The position ofchitin synthetase was ascertained by measuring the enzymatic activity in the fractionated gradients. Estimated from the Stokes radius and the sedimentation coefficient with Svedberg's equation (23). The partial specific volume of chitin synthetase was assumed to be 0.725. not shown). At higher concentrations of substrate, the amount of oligosaccharides sharply decreased and only represented a few per cent of the total reaction product. The formation of oligosaccharides may result from the presence of chitinase activity, either as a contaminant or intrinsic to chitin svntheoligosaccharides ranging from 2 to 8 or 9 sugar units (data tase itself. Because the hydrolytic enzyme is especially-effec- tive on the nascent chitin chains manufactured by the synthetase (15, 24) a very small amount might have been sufficient for the observed effect. Alternatively, the oiigosaccharides could represent prematurely terminated chitin chains. Addition of chitin to the enzymatic preparation in order to adsorb any putative chitinase (24) had no effect on the production of oligosaccharides. Nevertheless, when a mixture of the labeled oligosaccharides was incubated again with the chitin synthetase preparation in the absence of UDP-GlcNAc, some degradation appeared to occur, as judged from a shift in amount from the higher to the lower members of the series (data not shown). Therefore, a final conclusion about the origin of oligosaccharides cannot be made.
There is no indication that the oligosaccharides are capable of further elongation. Thus, incubation of radioactive tetraacetylchitotetraose with chitin synthetase and unlabeled UDP-GlcNAc did not result in a change in the amount or position of the oligosaccharide upon subsequent column chromatography. Furthermore, preliminary pulse-chase experiments failed to show incorporation of the oligosaccharides into larger compounds.
Length of Chitin Chains-Since preliminary experiments suggested that chitin chains made by purified enzyme carry a free reducing end group, chain lengths were estimated by reduction with borotritide, followed by digestion with purified Serratia chitinase (9) and separation of the products on a P-2 column. Because of the specificity of the chitinase, the end groups were converted into a mixture of reduced disaccharide and reduced trisaccharide. The latter was hydrolyzed by 0-Nacetylglucosaminidase to reduced disaccharide and free Nacetylglucosamine that could be measured colorimetrically, thereby allowing determination of the specific activity of the terminal N-acetylglucosaminitol (see "Methods"). The chain length, estimated from the ratio of total N-acetylglucosamine to N-acetylglucosaminitol, was a little above 100 (Table IV).
With the same procedure, chain length was determined in two other samples of chitin, one obtained from yeast cell septa and the other synthesized with membrane-bound synthetase. Surprisingly, the same chain length, within experimental error, was measured in all three cases (Table IV). The relative shortness of the chains might be artifactual, if chitin chains were broken under reduction conditions, i.e. 0.01 N alkali. This seems quite unlikely, because of the high stability of chitin to alkali. Furthermore, reduction of tetraacetylchitotriose with borotritide under the same conditions gave rise to a single product with the expected elution position in a Bio-Gel P-4 column. On the other hand, we cannot exclude the possibility that the chains are even shorter than determined.
The calculated values of the chain length are based on the assumption that sodium borohydride can react with all the reducing ends in insoluble chitin. This assumption has not been tested.

DISCUSSION
Chitin synthetase is the first enzyme, that has been extensively purified, capable of catalyzing the formation of a structural polysaccharide. Our main purification step takes advantage of a unique property of this type of enzyme, that of catalyzing the synthesis of an insoluble product with which the enzyme coprecipitates. Simple binding of the synthetase to chitin does not explain the precipitation of the enzyme, because external addition of the polysaccharide was without effect. Rather, it appears that the synthetase remains physically trapped in the highly hydrated and hydrogen-bonded network of chitin chains that it has spun, much as a silkworm is enclosed in its cocoon. This view is supported by the increase in enzyme recovery if the gel is disrupted and partially dehydrated by freezing and thawing. This methodology may be applicable to other polysaccharide synthetases when solubilized preparations of these enzymes become available.
Our purified preparations still contain contaminants, notably the band of 74,000 daltons that has been seen in all preparations subjected to electrophoresis. Some heterogeneity may be expected because of partial degradation of the enzyme by the trypsin used in the activation step. Conversely, some of the enzyme subunits may have escaped trypsin action and still be in the zymogen form. These subunits would remain together with those in the active form in the high-molecularweight aggregate found in nondenatured preparations and would accompany them during further purification. It is indeed possible that the 74,000-dalton band corresponds to enzyme subunits in the zymogen form. Another possibility, that chitin synthetase consists of different subunits, cannot be excluded at the present time.
Our tentative conclusion, that the 63,000-dalton band represents active chitin synthetase, is based on the fact that it is the major band and that it clearly shows enrichment when comparing first and second chitin steps. A final answer on this point will have to await the availability of antibodies to chitin synthetase. Recently, purification of chitin synthetase from Coprinus cinereus, a basidiomycete, was reported by Montgomery et al. (25). Their preparation also shows some heterogeneity, but the main band corresponds to a molecular weight of -67,000, which is not very different from our value.
There is no evidence that the enzyme from Coprinus can exist in a zymogen form (8).
It is of interest that the purified synthetase had almost completely lost dependence on free N-acetylglucosamine, which stimulated the activity severalfold in membrane preparations and might have been thought of as a possible primer. The only requirements of the purified enzyme were a divalent cation, phosphatidylserine, and digitonin. The last two may be favorable in that they tend to reconstruct an environment similar to that of the natural habitat of the enzyme, the plasma membrane. The synthetase was able to catalyze the formation of chitin after electrophoresis in nondenaturing gels and while still embedded in the gel. This provides a striking illustration of the autonomy of the enzyme. If a primer or a lipid intermediate, for which we have been unable to find evidence (6), were necessary for the reaction, they would have to be very strongly bound to the enzyme for the association to survive after electrophoresis.
The formation of oligosaccharides that takes place at low concentrations of substrate may be due to a trace of chitinase activity present in the preparation but it could also result from premature termination of growing chitin chains. If the first interpretation is correct, the decrease in production of oligosaccharide as the formation of insoluble chitin increases might be due to binding of the higher oligosaccharides to chitin by hydrogen bonding and consequent protection against chitinase action. Alternatively, premature termination of the chains may be prevented at higher concentrations of substrate by an increase in elongation rate. Whatever the interpretation, pulse-chase experiments proved conclusively that the free oligosaccharides are not intermediates in the synthesis of insoluble chitin. This suggests that each chitin chain is synthesized by a processive mechanism (26), i.e. it is not released from the enzyme until completed.
It is remarkable that samples of chitin synthesized in uiuo, by a membrane preparation, and by purified synthetase, all have the same chain length. It appears that the enzyme itself, and not other factors or organizational elements present in the membrane or in the intact cell, determines at what point a chitin chain should be terminated and a new one started.