Characterization of lecithin:Cholesterol acyltransferase from human plasma

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 222, No. 2, April 15, pp. 553-560, 1983

Characterization

of Lecithin:Cholesterol from Human Plasma:

II. Physical

KUI-SONG Department

Properties

Acyltransferase

of the Enzyme

CHONG, SHINICHI HARA, RICHARD AND ANDRAS G. LACKO’

E. THOMPSON,

of Chemistry, Division of Biochemistry, North Texas State University, and Department Biochemistry, Texas College of Osteopathic Medicine, Denton, Texas 76203 Received October 14, 1982, and in revised form December

of

9, 1982

The physical properties of purified human plasma 1ecithin:cholesterol acyltransferase (LCAT) were investigated by techniques including analytical ultracentrifugation, ultraviolet spectroscopy, electrofocusing, and circular dichroism. The partial specific volume of LCAT was determined by sedimentation equilibrium ultracentrifugation experiments in Hz0 and DzO solutions (0.702 ml/g). The M,. was 67,000 by sodium dodecyl sulfate (SDS)-gel electrophoresis and 60,000 by sedimentation equilibrium ultracentrifugation. The discrepancy between the two sets of data presumably arose from the glycoprotein nature of the enzyme. Studies of the ultraviolet spectrum indicated that LCAT contained 6.5% (w/w) tyrosine which corresponds to approximately 18 tyrosine residues/m01 of LCAT (polypeptide M, 45,000). Spectrophotometric titration of the ionizable phenolic side chains indicated that nearly all the tyrosine residues were buried at neutral pH while they became gradually exposed at higher pH. The apparent pK of this transition was about 12.0 contrasted with 9.8, the apparent pK of ionization of the free tyrosyl groups.

LCAT reaction mechanism has likely been due to the difficulties in the isolation of the enzyme in sufficient quantities for detailed characterization and to the unstable nature of the highly purified enzyme preparations (4). We have recently described a highly efficient and reproducible procedure for the purification of LCAT which allows the isolation of enzyme protein in the milligram range (5) and in a stable form (6). These recent developments now allow us to characterize the enzyme in detail. The present communication deals with the physical properties of LCAT which were investigated by techniques including analytical ultracentrifugation, ultraviolet spectrosdichroism, and electrocopy, circular focusing.

Lecithin:cholesterol acyltransferase (LCAT)’ catalyzes one of the three reactions that have been described for the esterification of cholesterol (1). The enzyme exhibits a number of unique features including the requirement for a macromolecular substrate complex (2) and a polypeptide factor (3). Despite the exciting potential for studies on the catalytic function and the nature of the enzyme substrate complex, the mechanism of action of LCAT remains largely unexplored. The relatively slow progress in the elucidation of the 1 To whom correspondence should be addressed: P.O. Box 13408, Department of Biochemistry, Denton, Tex. 76203 *Abbreviations used: LCAT, 1ecithin:cholesterol acyltransferase; SDS, sodium dodecyl sulfate; BSA, bovine serum albumin. 553

0003-9861/83 $3.00 Copyright All rights

0 1983 by Academic Press, Inc. of reproduction in any form reserved.

554

CHONG MATERIALS

AND

METHODS

Chemicals. Deuterium oxide (99.8%), calf serum fetuin, ovalbumin, aldolase, and chymotrypsinogen were purchased from Sigma Chemical Company, St. Louis, Missouri. Bisacrylamide, acrylamide, riboflavin, ammonium persulfate, N,N,N’,N’-tetramethylethylene diamine, P-mercaptoethanol and Coomassie blue R250 were purchased from Eastman Kodak, Rochester, New York. Ampholine gels were obtained from LKB. Amino acid standards were purchased from Pierce Chemical Company, Rockford, Illinois. All other chemicals were products of Fisher Scientific Company, Springfield, New Jersey, and were of reagent grade. PuriJicatiun of LCAT. LCAT was purified to homogeneity by a procedure reported earlier (4) consisting of four basic steps: (a) dodecyl amine-agarose chromatography; (b) DEAE-agarose chromatography; (c) delipidation; and (d) hydroxyl apatitejantibody-agarose chromatography. The final product was 20,000-fold purified over the starting material and was found to be homogeneous by several criteria including SDS-polyaerylamide gel electrophoresis, analytical ultracentrifugation, immunodiffusion, and immunoelectrophoresis. Protein determination. Bio-Rad protein assay kits were used for protein determination using a BSA standard according to the method described by Bradford (7). In addition, amino acid composition analyses were carried out (8) to determine protein concentration using norleucine as an internal standard. Polyacrylamide gel electrophoresti. Analytical gel electrophoresis in the presence of SDS was performed according to the method described by Maize1 (9). Determination of partial spec$ic volume. The partial specific volume (3 of the enzyme was determined according to Edelstein and Schachmann (10) employing sedimentation equilibrium analysis in the presence of Hz0 and DzO. The calculation of V is accomplished with the aid of the equation k - [(dlnc/dr2)D~0/(dlnc/dr2)H20] ’ = pD,O - pHzO[(dlnc/dr2)Dz0/(dlnc/drZ)HzO]

’

where k is the ratio of the molecular weight of the macromolecules in the deuterated solvent to that in the nondeuterated solvent, p is the density, and dlnc/ dr2 is the change in protein concentration with respect to the square of the distance from the center of the rotation. While the k value has been determined for some proteins (ll), no such data appears to be available for glycoproteins. Chung et al. (12) have recently reported that LCAT may contain up to 25% covalently linked carbohydrate thus necessitating the calculation of the k value for glycoprotein preparations. We have employed calf serum fetuin as a standard. Fetuin is available commercially in a

ET AL. highly purified form and has a M, (48,400) and carbohydrate content (25%) similar to LCAT (13). The purity of calf serum fetuin was assessed to he 9095% by SDS-polyacrylamide gel electrophoresis. One milliliter (75. mg/ml) of calf serum fetuin was dialyzed extensively against deionized HzO. The dialyzed sample was divided into two portions of equal volume, and they were pipetted into Pyrex test tubes and lyophilized. The first sample was redissolved in 6.5 ml of deionized water and the second one was redissolved in 0.5 ml of deuterium oxide (99.8% ). They were extensively dialyzed against deionized water and deuterium oxide, respectively. High-speed sedimentation equilibrium experiments, using a 12-mm double-sector cell, were carried out essentially as described by Yphantis (14) in a Beckman Model E analytical ultracentrifuge equipped with electronic speed control. dlnc vs dr’ was plotted and slopes (both in Hz0 and DzO) were determined by a linear regression computer program. The slopes of (dlnc/dr’)HzO and (dlnc/dr2)D20 were used to calculate the value of k. One milliliter of purified human LCAT (0.64 mg/ ml) was extensively dialyzed against deionized water. The dialysate was divided into two portions of equal volume and lyophilized. The two portions were redissolved in 0.5 ml of deionized water and in 0.5 ml of deuterium oxide, respectively, and then dialyzed against the respective buffers in deionized water and deuterium oxide. The analytical ultracentrifuge method was followed as described in the method portion (determination of k value using calf serum fetuin). Both of the slopes in (dlnc/dr’)H,O and (dlnc/ dr’)D,O in combination with k obtained from the fetuin experiment were used to calculate V of human LCAT using the following formula: k - [(dlnc/dr’)DzO, c=

LCAT/ (dlnc/dr’)H,O,

pDzO - pHzO[(dlnc/dr’)DzO, LCAT/ (dlnc/dr2)H20,

LCAT)] LCAT]

Molecular weight detminatima. The purified LCAT (0.8 mg/ml) was lyophilized and redissolved in either 1 mM sodium phosphate buffer, pH 7.2 or 6~ guanidine HCl containing 1 mM sodium phosphate buffer, pH 7.2, and 0.1 M /3-mercaptoethanol. The enzyme samples were then dialyzed to equilibrium against the respective buffers. Solvent densities were determined by weighing loml aliquots of the respective buffer solutions on a semimicro analytical balance and established to five significant figures. The slopes (dlnc/dr’) in the presence of 1 mM phosphate, pH 7.2, or in the presence of 6 M guanidine HCl/l mM phosphate buffer, pH 7.2, were determined and used to calculate the molecular weight by sedimentation equilibrium experiment. Molecular weight determination by SDS-polyacvyl-

PHYSICAL

PROPERTIES

gel electrophoresis. The molecular weight of human plasma LCAT was also determined by applying lo-20 fig of protein to SDS-gel electrophoresis (9) as described by Weber and Osborn (15) using phosphorylase B, ovalbumin, human serum albumin, aldolase, and chymotrypsinogen as standards. The distance of migration for each protein was established by the Quick-Quant gel scanning apparatus (Helena Laboratories). Spectrophotometric titration. One milliliter (0.2 mg/ ml) of purified human LCAT (mg/ml) was dialyzed extensively against 5 mM phosphate buffer, pH 7.2, with several changes of the dialysate. The enzyme solution was placed in a double-beam spectrophotometer in l-cm cells. Scanning was performed between 320 and 240 nm. The pH of the LCAT solution was gradually increased by the addition of small aliquots (l-30 ~1) of KOH (1-8 M) solutions. The concentration of tyrosine residues was calculated from the absorbance at 295 nm assuming a molar extinction eoefficient of 2300 (16). Circular dichroism (CD). The purified enzyme was subjected to CD study using the Jasco J-40A apparatus. The CD spectra were recorded at a sensitivity of 2 m”/cm in a strain-free quartz cuvette with a pathlength of 0.1 cm. The concentration of LCAT used for the CD study was determined by both amino acid composition analysis and Bio-Rad protein assay as described by Bradford (7). The pH dependence of the CD spectrum was carried out between 250 and 200 nm as described for the ultraviolet spectral studies above. CD data analvslsis. Spectral data were analyzed according to the method of Chen et al. (17) as modified by Thompson et al. (18) using the nonlinear regression analysis package HELIX 1.2 (available from the authors upon request). The three-component model describes data as fraction a-helix, P-pleated sheet, and remaining structures. Analytical electrofocusing. Analytical electrofocusing on LKB Ampholine polyacrylamide gel plates, pH range 3.5-9.5 (LKB Producter, Bromma, Sweden), was carried out by applying 20 ~1 of human plasma LCAT (1.0 mg/ml) according to LKB instruction using LKB 2117 Multiphor electrophoresis equipment. After electrofocusing was stopped after 2 h, the plate was removed and stained in 0.12% (w/v) Coomassie blue R-250 in 25% ethanol/8% acetic acid for 30 min at 65”C, and was destained in 25% ethanol/S% acetic acid. pH was determined from the calibrated plot of the distance traveled by proteins vs pH. amide

OF LCAT

555

of deuterium exchange and their partial specific volume is decreased by the same relative amount. The k value is the ratio of the molecular weight of the macromolecules in the deuterated solvent to that in the nondeuterated solvent. The k value calculated for calf serum fetuin (a glycoprotein) was found to be 1.016, which is similar to other k values reported for proteins containing no carbohydrate (11). The apparent partial specific volume of human plasma 1ecithin:cholesterol acyltransferase was determined to be 0.702 ml/g using a k value of 0.016. A partial specific volume of V = 0.707 ml/g was also calculated from polypeptide and carbohydrate composition analysis assuming 75% polypeptide and 25% carbohydrate content [hexoses 13%, hexosamines 6.2%, and sialic acid 5.4% (19)]. Molecular

Weight Determinations

The molecular weight of human plasma LCAT determined by sedimentation equilibrium experiments in 1 mM phosphate buffer, pH 7.2, using a V of 0.702 ml/g was 60,000 f 1400 (three determinations). Figure 1 shows the plot of dlnc (LCAT concentrations) vs dr’. The linearity of the plot is an indication of the homogenity of the enzyme preparation which had already been established by other methods (4). Molecular weight values obtained in the presence of 6 M guanidine HCl were 61,000 and 60,300 (two determinations). Such data indicate that the enzyme does not possess any quarternary structure. Molecular weight determinations were also carried out by polyacrylamide gel electrophoresis of 20 pg of LCAT in the presence of SDS yielding a value of 67,000 f 2000 (not shown). The discrepancy between the values from SDS-gel electrophoresis analysis as compared to those from sedimentation equilibrium experiments is likely to be due to the relatively high carbohydrate content of the enzyme (12, 19).

RESULTS

Partial

Spec$c

Volume (C)

When macromolecules are dissolved in D20, their weight is increased as a result

Ultraviolet

Absorption

Human plasma LCAT exhibited an ultraviolet spectrum with a maximum at 275

CHONG

556

ET AL.

r2

FIG. 1. Sedimentation equilibrium ultracentrifugation of purified human plasma lecithinxholesterol acyltransferase (0.8 mg/ml) in 1 mM phosphate buffer, pH 7.2. Ultracentrifugation was carried out at 18°C at 22,000 rpm in the An-D rotor in a 12-mm double-sector cell with sapphire windows. The molecular weight of LCAT was determined using V of 0.702 g/ml.

nm in 5 mM sodium phosphate, pH 7.2, and two maxima (at 280 and 291 nm) above pH 13 (data not shown). Figure 2 shows the spectrophotometric titration of LCAT be-

?

tween pH values 7 and 13. The large increase in absorbance at 295 nm indicates that nearly all the phenolic groups are unavailable to the solvent at neutral pH while

0 x

7

0

9

10

11

12

13

14

PH

FIG. 2. Spectrophotometric titration of LCAT (0) and N-acetyltyrosinamide (X) at 295 mM. One milliliter of LCAT (0.2mg/mI) was titrated by the addition of small increments (l-30 ~1) of KOH solutions of appropriate concentration (1-8 M).

PHYSICAL

200

210

PROPERTIES

220

557

OF LCAT

230

240

250

FIG. 3. Circular dichroism (CD) spectra of LCAT in the far ultraviolet region at pH 7.1 (0) and at pH 12.0 (X). The CD spectrum was obtained using selected samples of LCAT following the recording of’ the ultraviolet spectrum (Fig. 2).

they become fully exposed at higher pH values. The absorbance data allowed the calculation of the concentration of tyrosine residues per mole (eighteen) assuming a molar extinction coefficient of 2300. This value agrees well with the number of tyrosine residues per mole calculated from amino acid analysis (20). It appears that the exposure of the tyrosine residues to the aqueous environment requires some type of conformational transition of unfolding of the polypeptide chain with a characteristic apparent pK value of about 12.0 that is distinct from the apparent pK of ionization of the free tyrosine residues (Fig. 2). Circular

Dichroism

maining structure in 1 mM sodium phosphate buffer, pH 7.2. The CD spectrum of LCAT was also examined in a buffer of 0.2 M NaCl containing 1 mn! sodium phosphate, pH 7.2. Under these conditions, the enzyme appeared to have 23% a-helix, 32% P-pleated sheet, and 45% remaining structure. While the change to the high ionic strength medium was accompanied by a slight increase in P-pleated sheet structure, these data indicate no significant change in the overall conformation of the enzyme. Figure 4 shows the change in molar ellipticity at 222 nm as the function of pH. Similarly to the absorbance data shown on Fig. 2, no significant change in molar ellipticity occur between pH 9 and 11 while a relatively large increase was apparent between pH 11 and 12.

CD data obtained on the purified LCAT sample in the far ultraviolet region are Isoelectric Focusing shown in Figure 3. Human plasma LCAT (0.32 mg/ml) was found to have 24% CY- The electrofocusing of purified LCAT inhelix, 27% P-pleated sheet, and 49% re- dicated the presence of at least four bands

CHONG

558

I

ET AL.

I

I

I

I

I

0

9

10

11

12

PH

FIG. 4. The pH dependence of the molar ellipticity of LCAT at 222 nM. Conditions experiment were the same as described for the ultraviolet spectral studies (Fig. 2).

having p1 values from 4.2 to 4.5. An earlier experiment (not shown) performed with the preparative isoelectric focusing apparatus revealed a single zone of LCAT activity in the p1 region of 4.5-5.2. This discrepancy is probably due to the mixing of the enzyme with the Ampholine solution used in the preparative isoelectric focusing experiment. DISCUSSION

In this communication, several physical properties of human plasma LCAT are reported. Table I compares the physical parameters that resulted from this investigation and data published from other laboratories (12, 21-25). The partial specific volume (~7 of LCAT is lower than that of proteins containing only amino acid residues due to the presence of carbohydrate in the enzyme. The G calculated from poly-

for this

peptide and carbohydrate composition agreed very closely with the G value obtained from D20/H20 experiments (Table I). Chung et al. (12) calculated and reported a V value for LCAT (0.71 ml/g) calculated from the polypeptide and carbohydrate composition of the enzyme in an earlier communication. The molecular weight was determined by sedimentation equilibrium analysis (-60,000) both in the presence and absence of 6 M guanidine HCV0.1 M 2-mercaptoethanol indicating that LCAT is a single polypeptide. The different M, obtained by SDS-gel electrophoresis (67,000) is probably due to the carbohydrate content of LCAT. Similar discrepancies of molecular weights obtained by SDS-gel electrophoresis and sedimentation analysis were reported by Chung et al. (12) for LCAT and by other investigators for two other serum glycoproteins (26,27). Earlier

PHYSICAL

PROPERTIES TABLE

I

PHYSIOCHEMICAL PROPERTIESOFHUMANPLASMA Methods

and conditions

DrO-HZ0

in 0.2 M NaCl/O.OOl M phosphate, pH 7.2

Chung et al. (12) This communication This communication

4.28-4.37 4.1-5.5 4.2-4.5

Uterman et al. (22) Albers et al. (21) This communication

a-Helix @-Sheet Remainder a-Helix p-sheet Remainder

reports from our laboratory suggested that LCAT may have a dimeric structure (28). These earlier studies, however, were conducted with partially purified enzyme and radioactively labeled diisopropyl fluorophosphate was employed to establish the elution of the enzyme from agarose columns. Due to the very small amount of enzyme protein used in these experiments, in addition to the other factors listed above, the generation of artifacts was a likely possibility. Therefore, data presented in this communication (Table I) as well as from other laboratories (12, 21) strongly indicate that LCAT is a glycoprotein consisting of a single polypeptide chain with a M, of 60-61,000. Circular dichroism studies revealed that human LCAT has a relatively high content of p-pleated sheet structure. However, only a small increase in P-pleated sheet structure (from 27 to 32%) was observed when

This communication Albers et al. (21, 23) Chung et al. (12) Aron et al. (24) Uterman et al. (22) Kitabake et al. (25) This communication

0.71 ml/g 0.707 ml/g 0.702 ml/g

point

pH 7.2

Chung et al. (12) This communication

60,700 68,000; 66,000 69,000 66,000 67,000 65,000 67,000

Partial Specific Volume Composition analysis

Circular Dichroism in 0.001 M phosphate,

Sources

59,000 60,000

Sedimentation equilibrium in 6 M guanidine HCl SDS-gel electrophoresis

Isoelectric

LCAT

Observations

Molecular weight Sedimentation equilibrium in 0.001 M phosphate, pH 7.2

Parallel

559

OF LCAT

structure

structure

24% 27% 49% 23% 32% 45%

This This This This This This

communication communication communication communication communication communication

the CD spectrum was obtained with an LCAT sample in a buffer containing 0.2 M NaCl. Furukawa and Nishida (4) reported that LCAT activity was progressively lost when the enzyme was dissolved in a buffer containing higher ionic strength than 4 mM sodium phosphate. Since in the presence of 0.2 M NaCl only small changes resulted in overall structure as indicated by the CD analysis, the inactivation of LCAT at high ionic strength is more likely to be due to alterations in the active site region rather than to large conformational changes of the enzyme molecule. The near ultraviolet absorption spectra obtained, showed some apparent peak shifts as the function of pH. Thus the absorption spectrum due to tryptophan will dominate the near-uV wavelength range in LCAT at neutral pH (maximum at 275 nm) while a second maximum (at 280 nm) appears beyond pH 11. The spectropho-

560

CHONG

tometric titration of phenolic groups in the enzyme showed a large increase in absorbance at 295 nm with an apparent pK of about 12.0. In a control experiment, using N-acetyltyrosinamide, the apparent pK for the increase in absorbance was about 9.8 indicating that the ionization of phenolic side chains in the enzyme was delayed, perhaps due to a conformational change that allowed the exposure of the tyrosine residues to the solvent. Similarly, the largest change in molar ellipticity (at 222 nm) was observed above pH 11 supporting the view that the enzyme molecule begins to undergo a conformational transition at this pH. The state of the tyrosine residues (which are nearly totally buried at neutral pH) may be contrasted to the somewhat unusually exposed state (60%) of tryptophan residues in LCAT (20). The analytical isoelectric focusing pattern of LCAT (not shown) was found to be similar to that reported by Uterman et al. (22) and Doi and Nishida (29), but it is slightly different from the data of Albers (21). The latter investigators reported a higher pK and five instead of four components of the enzyme. Uterman et al. (22) suggested that the premixing of the enzyme protein with the Ampholine solution might cause alterations in the electrofocusing pattern. Doi and Nishida have also shown that the microheterogeneity of LCAT, as demonstrated by isoelectric focusing, was due to differences in sialic acid content, since the multiband pattern was abolished upon digestion with neuraminidase (29). ACKNOWLEDGMENTS This research was supported by grants from the National Institutes of Health (AC-03255 and BRSG SO7 RR07195), the American Heart Association (821043). the Robert A. Welch Foundation (B-935), and the Faculty Research Fund of the Texas College of Osteopathic Medicine. REFERENCES 1. GLOMSET, J. A. (1968) J. Lipid Res. 9, 155. 2. AKANUMA, Y., AND GLOMSET, J. A. (1968) J, Lipid Res. 9, 620. 3. FIELDING, C. J., SHORE, V. G., AND FIELDING, P. E. (1972) B&hem Biophys. Res. Commun 46, 1493.

ET AL. 4. FURUKAWA, Y., AND NISHIDA, T. (1979) J. BioL Chem. 254, 2’713. 5. CHONG, K. S., DAVIDSON, L., HUTTASH, R., AND LACKO, A. G. (1981) Arch. B&hem. Biophys. 211, 119. 6. JAHANI, M., AND LACKO, A. G. (1982) Biochim. Biophys. Acta, in press. 7. BRADFORD, M. M. (1976) Anal. Biochem. 73,248. 8. HARE, P. E. (1977) in Methods in Enzymology (Hirs, C. H. W., and Timasheff, S. N., eds.), Vol. 47, pp. 3-18, Academic Press, New York. 9. MAIZEL, J. V., JR. (1971) in Methods in Virology (Maramorsch, K., and Koprowski, V., (eds.), Vol. 5, p. 179, Academic Press, New York. 10. EDELSTEIN, S. J., AND SCHACHMANN, M. K. (1973) in Methods in Enzymology (Hirs, C. H. W., and Timasheff, S. N., eds.), Vol. 27, p. 82, Academic Press, New York. 11. MARTIN, W. G., WINKLER, C. A., AND COOK, W. H. (1959) Can&. J. Chem. 37, 1662. 12. CHUNG, J., ABANO, D. A., FLESS, G. M., AND SCANLI, A. M. (1980) J. BioL Chem. 254, 7456. 13. SPIRO, R. G. (1960) J. BioL Chem. 235. 2860. 14. YPHANTIS, D. (1964) Biochemistry 3, 297. 15. WEBER, K., PRINGLE, R. T., AND OSBORN,M. (1972) in Methods in Enzymology (O’Malley, B. W., and Hardman, J. G., eds.), Vol. 36, p. 3, Academic Press, New York. 16. BEAVEN, G. H., AND HOLIDAY, E. R. (1952) Advan Protein Chem. 7, 319. 17. CHEN, Y. H., YANG, J. T., AND CHAU, K. H. (1974) Biochemistry 13, 3550. 18. THOMPSON, R. E., SPIVEY, H. O., AND KATZ, A. J. (1976) BiochemisWy 15, 862. 19. CHONG, K. S., DAVIDSON, L., HUTTASH, R. G., AND LACKO, A. G. (1981) Fed Proc. 40, 1696. 20. CHONG, K. S., HARA, S., JAHANI, M., AND LACKO, A. G. (Unpublished observations). 21. ALBERS, J. G., CABANA, V. G., AND BARDEN-STAHL, Y. D. (1976) Biochemistry 15, 1084. 22. UTERMANN, G., MENZEL, H. J. ADLER, G., DIEKER, P., AND WEBER, W. (1980) Eur. J. Biochem 107, 225. 23. ALBERS, J. A., LIN, J. T., AND ROBERTS, G. P. (1979) Artery 5, 61. 24. ARON, L., JONES, S., AND FIELDING, C. J. (1978) J. BioL Chem. 253, 7220. 25. KITABAKE, K., PIRAN, U., KAMIO, Y., DOI, Y., AND NISHIDA, T. (1979) B&him. Biophys. Acta 573, 145. 26. MICKELSON, K. E., AND PETRA, P. H. (1978) J. BioL Chem 253, 5293. 27. KISIEL, W., AND DAVIE, E. W. (1975) Biochemistry 14, 4928. 28. LACKO, A. G., VARMA, K. G., RUTENBERG, H. L., AND SOLOFF, L. A. (1974) Stand. J. Clin Lab. Invest. SuppL 137, 29. 29. DOI, Y., AND NISHIDA, T. (1981) Fed. Proc.40,1695.