ARCHIVES
OF BIOCHEMISTRY
AND BIOPHYSICS
Vol. 198, No. 1, November, pp. 247-254, 1979
Purification
and Partial Characterization of a Carboxypeptidase from the Limpet (Pate/la vu/gata)l G. MICHAEL
Department
of Bacteriology
HASS University of Idaho,
and Biochemistry,
Idaho 838&’
Moscow,
Received May 8, 1979; revised June 14, 19’79 A carboxypeptidase A-like enzyme has been purified to apparent homogeneity from commercially available acetone powder from the visceral hump of the limpet, Patella vulgata. A two-step procedure involving affinity chromatography on l -amino-N-caproyl-Dphenylalanine-Sepharose and gel filtration resulted in a 3000-fold purification with an 80% yield. The enzyme is a single polypeptide chain of M, = 40,000 and exhibits both peptidase and esterase activities, which are characterized by dramatic excess substrate inhibition. Inhibition studies suggest that a metal ion is required for activity and demonstrate that the affinity label, N-bromoacetyl-N-methyl-L-phenylalanine, and a polypeptide carboxypeptidase inhibitor from potatoes (apparent K, approx. 2 nM) are effective against the limpet enzyme.
In contrast to the attention given to digestive enzymes of the “serine” endoproteinase family (e.g., trypsin, chymotrypsin) from a variety of species, relatively little information is available about the pancreatic carboxypeptidases and their evolutionary relatives. Most of our understanding of the latter class of enzymes comes from studies on the bovine and porcine carboxypeptidases A and B (for reviews see Refs. (l-3)). However, corresponding enzymes from the dogfish (4), lungflsh (5), shrimp (6), starfish (7), and others have been purified and partially characterized. These species possess pancreatic or hepatopancreatic systems and have evolved from the line leading to mammals at a relatively recent date. Not surprisingly, these carboxypeptidases are similar to the mammalian enzymes in size (M, approx. 35,000) and other properties. An investigation of the carboxypeptidase from the visceral hump of the limpet (Patella vulgata) was undertaken to allow a comparison of the mammalian carboxy-
Materials. Acetone powder from the visceral hump of the limpet (Patella v&g&a), standard proteins, substrates, and diisopropyl fluorophosphate were purchased from Sigma Chemical Company. The polypeptide carboxypeptidase inhibitor was prepared from Russet Burbank potatoes as described by Ryan et al. (8). N-Bromoacetyl-N-methyl-L-phenylalanine (LBAMP)* was synthesized as described previously (9). Sephadex G-100 SF and Sepharose 4B were purchased from Pharmacia Fine Chemicals. c-Amino-N-caproyln-phenylalanine- Sepharose was prepared as described for the arginine derivative reported earlier (10) using the procedures of Parikh and Cuatrecasas (11). The affinity resin contained approximately 2 pmol D-phenylalanine/ml of packed gel resin as determined by ammo acid analysis tier acid hydrolysis. Enzyme pw$cation. Limpet acetone powder (10 g)
1 This work was supported in part by a grant from the National Institutes of Health (GM22748) and is published with the approval of the Director of the Idaho Agricultural Experiment Station as Research Paper No. 7953.
2 Abbreviations used: Mes, 2[N-morpholino] ethanesulfonic acid; SDS, sodium dodecyl sulfate; HPLA, hippuryl-rn-phenyllaetic acid; BGP, hippuryl&phenylalanine; BGA, hippuryl-Larginine; L-BAMP, Nbromoacetyl-N-methyl-Lphenylalanine.
247
peptidases with their counterpart from an extremely ancient animal. The limpet was chosen not only because this animal evolved from the main line leading to mammals over 600 million years ago but also because its digestive system is quite primitive, being primarily intracellular. EXPERIMENTAL
PROCEDURES
0003-9861/79/130247-08$02.00/O Copyright 0 1979 by Academic Press, Inc. All rights
of reproduction
in any form reserved.
248
G. MICHAEL
was gently stirred in 100 ml of 0.5 M NaCl, 0.01 M Mes (pH 6.0) at 4°C for 6 h, and the resulting suspension was centrifuged at 16,000 rpm for 30 min at 4°C. The supernatant (approx. 100 ml) was applied at room temperature to a 1.5 x 6.5-cm column of c-amino-N-caproyl-n-phenylalanine-Sepharose 4B. Fractions of approximately 8.5 ml were collected during the sample application and subsequent wash with extraction buffer. After the protein concentration in the eluate dropped to negligible levels, the buffer was changed to 10 mM r,-phenylalanine, 0.1 M Na,C03 (pH 10.0) and the fraction size decreased to 2 ml each. Fractions exhibiting activity against HPLA (see below) were pooled, dialyzed at 4°C against 10 mM Tris-HCl (pH 7.5), and lyophilized. Partially purified carboxypeptidase was dissolved in 2 ml of H20, centrifuged, and chromatographed at 4°C on a 1.5 x 80-cm column of Sephadex G-100 SF. The eluting buffer was 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5) and fractions of approximately 1.2 ml were collected. At each stage of purification aliquots were retained for protein determinations by the microbiuret procedure (12) or by absorbance at 280 nm and for enzymatic activity measurements (see below). Enzyme assays and inhibition studies. The activity of the limpet carboxypeptidase was routinely measured using 1 mM concentrations of HPLA (13) or BGP (14) as substrate. The increase in absorbance at 254 nm at 25°C was monitored using 3-ml volumes of substrate in 0.5 M NaCl, 25 mM Tris-HCl (pH 7.5). Enzyme units were expressed as micromoles substrate hydrolyzed per minute, where the hydrolysis of 1 wmol resulted in an increase of 0.12 A at 254 nm using cuvettes with a l-cm pathlength. In trial experiments, activity was determined using 1 mM hippurylL-arginine as substrate (15). Solutions of purified limpet carboxypeptidase (pH 7.5) were incubated at 25°C for 8 h with the following inhibitors at the indicated concentrations: EDTA (10 mM), l,lO-phenanthroline (10 mM), HgCl, (10 mM), diisopropyl fluorophosphate (50 mM), and bromoacetate (50 mM). Inhibition was detected by assay against 1 mM HPLA. Solutions of purified limpet carboxypeptidase in 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5) were incubated at 25°C with L-BAMP (9). Aliquots were removed at various times and assayed with HPLA as substrate to detect inactivation. Apparent Ki values for the inhibition of the limpet carboxypeptidase by the polypeptide carboxypeptidase inhibitor from potatoes (8) were estimated by the method of Green and Work (16). Carboxypeptidase and varying amounts of the inhibitor were incubated for 15 min prior to activity measurements with HPLA as substrate. The concentration of stock solutions of the inhibitor were estimated using an A&h% value
HASS
of 3.0 and by titration carboxypeptidase A.
with
solutions
of bovine
Molecular weight estimation. The molecular weight of native limpet carboxypeptidase was estimated by gel filtration on a 1.5 x 80-cm column of Sephadex G-100 SF as described by Whittaker (17). Elution buffer was 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5) and fractions of 1.2 ml were collected. The ratio of the elution volume of carboxypeptidase (Ve) to that of blue dextran 200 (Vo) was compared with Ve/Vo values for standard proteins (bovine carboxypeptidase A, carbonic anhydrase, bovine serum albumin, chymotrypsinogen A, ovalbumin, and horse heart myoglobin). The subunit molecular weight was estimated by polyacrylamide gel electrophoresis in buffers containing SDS (18). The carboxypeptidase and protein standards listed above were treated with 2-mercaptoethanol prior to electrophoresis.
RESULTS
Purij?cation The purification of the carboxypeptidase found in extracts of acetone powder of the visceral hump of the limpet (P. vulgata) was achieved using two chromatographic steps (Table I). In the 6rst step approximately 1500 mg of protein, which was derived from 10 g of acetone powder, was applied at a high salt concentration (0.5 M) to a column of l -amino-N-caproyl-nphenylalanine-Sepharose. Almost all of the enzyme bound to the resin and most of the contaminants passed through (Fig. 1). The elution of the carboxypeptidase was effected at pH 10.0 using buffer containing 10 mM L-phenylalanine. This step resulted in a purification of over 600-fold with a yield of nearly 90%. Chromatography of partially purified carboxypeptidase on Sephadex G-100 (Fig. 2) resulted in an additional fivefold purification with little loss of enzymatic activity. Material at this stage of purification appeared to be homogeneous by gel electrophoresis in the presence of SDS (see below). Molecular
Weight and Subunit
Structure
The molecular weight of native limpet carboxypeptidase was estimated by gel filtration (17). A comparison of the Ve/Vo values of the carboxypeptidase with those of several standard proteins (Fig. 3A) suggested a molecular weight of approximately 40,000.
PURIFICATION
OF LIMPET TABLE
PURIFICATION
Fraction Extract Affinity chromatography G-100
OF
I
LIMPET CARBOXYPEPTIDASE
Protein bg)
Activity (units)
Specific activity” (units/mg)
Yield (%I
1610*
21.0
0.013
100
1
89 81
653 3270
2.2” 0.4”
18.7 17.0
8.5 42.5 (55?
LIHippuryl-DLphenyllactate (1 mM) used as substrate. * Determined by microbiuret procedure (12). c Estimated from Atso assuming A&f,% = 1.0. d Determined by titration with the polypeptide carboxypeptidase
The subunit molecular weight of limpet carboxypeptidase was estimated by polyacrylamide gel electrophoresis in the presence of SDS (18). The value of approximately 40,000 which was obtained by this procedure (Fig. 3B) was in good agreement with that estimated for the native protein indicating that limpet carboxypeptidase was a single polypeptide chain. Kinetics Limpet carboxypeptidase was shown to hydrolyze the typical carboxypeptidase A substrates, BGP and HPLA. No activity was observed with BGA, a substrate often used to assay carboxypeptidase B-like enzymes, under conditions such that a rate 1% of that observed for BGP would have been detectable. The dependence of initial velocity upon substrate concentration for BGP and HPLA is given in Figs. 4A and B, respectively. Although the K, values were much too low to be estimated accurately, approximate values of 10 ,uM were suggested by these data. Of particular interest was the dramatic excess substrate inhibition apparent for both the ester and peptide substrates. Inhibition
249
CARBOXYPEPTIDASE
Studies
The activity remaining after limpet carboxypeptidase was incubated for 8 h with a variety of inhibitors is given in Table II. Both chelating agents (l,lO-
Purification ( -fold)
inhibitor (see text for details).
phenanthroline and EDTA) completely eliminated activity. HgCl, which binds to cysteine residues as well as to other sites was partially inhibitory while bromoacetate, a reagent fairly specific for thiols under these conditions, had no significant effect. Similarly, high concentrations of diisopropyl fluorophosphate were not inhibitory, suggesting that limpet carboxypeptidase did not have an active site serine residue. Limpet carboxypeptidase was inactivated by L-BAMP, an affinity label for bovine carboxypeptidase A (9). At each reagent
"
40
i-'-'-'-i
FRACTION
N”hmER
FIG. 1. Chromatography of crude extract from visceral hump of P. vulguta on a column (0.9 x 15 cm) of l -amino-N-caproyl-~phenylalanine-Sepharose. The extract was applied in 0.5 M NaCl, 10 mM Mes (pH 6.0), and the carboxypeptidase eluted (J) with 10 mM L-phenylalanine, 100 mM sodium carbonate (pH 10.0). Fractions l-52 were approximately 8.5 ml each while 53-70 were 2 ml each. Absorbance at 286 nm (0) and activity against HPLA (A) were monitored.
250
G. MICHAEL
HASS
FIG. 2. Chromatography of partially purified limpet carboxypeptidase on a column (1.5 x 80 cm) of Sephadex G-100 SF. The buffer was 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5), and 1.2-ml fractions were collected. Absorbance at 280 nm (----) and activity against HPLA were monitored, and the fractions indicated by the bar were pooled.
concentration indicated on Fig. 5 inactivation was pseudo-first order as expected. The t,,z for inactivation at the highest concentration of reagent tested was 27 min, a value much greater than that of 3 min observed for bovine carboxypeptidase A in a similar experiment (9). The linear relationship between the pseudo-first-order constant and inhibitor concentration allowed calculation of an apparent second-order rate constant of 2.5 x 10m4 M-’ min-’ for the inactivation reaction.
Limpet carboxypeptidase was also inhibited by the polypeptide carboxypeptidase inhibitor from potatoes (8) (Fig. 6). The linearity observed at low inhibitor levels between percentage activity remaining and amount of inhibitor present permitted titration of the amount of enzyme in assay solutions. In this experiment inhibitor and enzyme were equimolar at 60 ~1 of inhibitor solution. Based upon the observed activity in the absence of inhibitor and a molecular weight of 40,000 for the limpet enzyme,
I
1.8z
*
‘.6
0.6-
I
83
1.4-
2
\
9
0.4-
--,
3
G
4
$
0.2-
\
h-2-
A
4.2 (A)
5.0
LOG
MOLECULAR
WEIGHT
42 (B)
4.4
LOG
MOLECULAR
4.6
4.8
WEIGHT
FIG. 3. Molecular weight of limpet carboxypeptidase. (A) The native molecular weight was estimated by gel flltration on a column (1.5 x 80 cm) of Sephadex G-100 SF (17). Protein standards were (1) carbonic anhydrase, (2) carboxypeptidase A, (3) ovalbumin, and (4) bovine serum albumin. Arrow indicates Ve/Vo value for limpet carboxypeptidase. (B) The subunit molecular weight was estimated by polyacrylamide gel electrophoresis in the presence of SDS (18). Protein standards were (1) horse heart myoglobin, (2) bovine carboxypeptidase A, (3) ovalbumin, and (4) bovine serum albumin. Arrow indicates relative mobility of limpet carboxypeptidase.
PURIFICATION
OF LIMPET
FIG. 4. Double-recim+ocal plot for the limpet carboxypeptidase-catalyzed and HPLA (B) at pH-7.5. -
specific activities of 55 and 30 units/mg were estimated (Table I) using 1 mM HPLA and 1 mM BGP, respectively. Apparent Ki values for the carboxypeptidase inhibitor estimated from free enzyme and inhibitor concentrations at equivalence were approximately 0.8-2 nM (6). DISCUSSION
251
CARBOXYPEPTIDASE
hydrolysis of BGP (A)
primitive basal gastropod and probably arose in the early Cambrian period approximately 600 million years ago (24). Among the group listed above the only animals that approach this early time of evolution are the starfish, an Asteroidian, which evolved in the Devonian period some 475 million years ago, and the shrimp, a Crustacean, which is dated at approximately 400 million years. The lungfish arose in the Devonian period about 350 million years ago, approximately 200-210 million years ago before modern shark (e.g., dogfish). The limpet and the shrimp are protostomates (i.e., mouth derived from blastospore) and are to be distinguished from such deuterostomates (i.e., anus derived from blastospore) as the dogfish, lungfish, and starfish. Protostomates and
The pancreatic carboxypeptidases comprise a particularly interesting class of enzymes from the standpoint of studying the evolution of structure and function. Not only is it intriguing to trace the development of carboxypeptidases in going from bacterial to mammalian sources, but the introduction of the zymogen function and the alteration in specificity from ancestral carboxypeptidase to carboxypepTABLE II tidase A-like and carboxypeptidase B-like enzymes are also worthy of investigation INHIBITION OF LIMPET CARBOXYPEPTIDASE (19). In attempting to elucidate these areas, pancreatic carboxypeptidases from a Concentration Percentage variety of mammals (e.g., human (ZO), Inhibitor activity” (M) bovine (l-3), porcine (21), guinea pig None 100 (22), goat (23), etc.) and similar enzymes EDTA 0.01 0 from such lower forms as dogfish (4), lungl,lO-Phenanthroline 0.01 0 fish (5), shrimp (6), and starfish (7) have 0.01 20 R&l, been purified and partially characterized. Diisopropyl Unfortunately, these enzymes appear to 0.05 fluorophosphate 99 exhibit few significant differences either in Bromoacetate 0.05 98 structure (e.g., all have M, ea. 34,000) or kinetic properties. Patella, a gastropod of a Enzyme was assayed with HPLA after an 8-h incuthe order Archaeogastropoda, is the most bation with inhibitors at the concentrations indicated.
252
G. MICHAEL
FIG. 5. Pseudo-first-order rate constant, k, for inactivation as a function of N-bromoacetyl-N-methylL-phenylalanine (L-BAMP) concentration.
deuterostomates have presumably evolved along separate lines. From an evolutionary standpoint a carboxypeptidase from Streptomyces griseus may well afford the most interesting comparison with its mammalian pancreatic counterparts (25). That this bacterial enzyme and carboxypeptidases A and B share a common ancestral protein is suggested by their similarities in metal ion requirement, in size, and in interaction with a polypeptide inhibitor from potatoes (26). The investigation of a digestive carboxypeptidase from the limpet, Patella vulgata, was initiated since this animal is extremely primitive, having evolved from the main line over 600 million years ago, during which period little anatomical change has occurred. It was thus hoped that a thorough investigation of the limpet enzyme would elucidate the evolution of the pancreatic carboxypeptidases and, perhaps, provide a link between bacterial and mammalian enzymes. The purification of the limpet enzyme was easily achieved in two steps with affinity chromatography on l -amino-N-caproyl-Dphenylalanine being unusually effective (Fig. 1, Table I). The utility of this resin is not unexpected in view of the low K, of the enzyme for its substrates and a low Ki value with respect to D-phenylalanine.3 It was anticipated that immobilized carboxypeptidase inhibitor from potatoes would be 3 G. M. Hass, unpublished observations.
HASS
useful in the purification of the limpet enzyme since a tight enzyme-inhibitor association was observed (Fig. 6) and since inhibitor- Sepharose had been successfully used to purify bovine and porcine carboxypeptidase (10). However, limpet carboxypeptidase bound to inhibitor-Sepharose could not be eluted without loss of activity. The first major difference between the limpet and mammalian carboxypeptidase became apparent when purified enzyme was subjected to polyacrylamide gel electrophoresis in the presence of SDS. Unlike the digestive carboxypeptidases listed above, all of which have molecular weights of 34,000-35,000 the limpet enzyme has a subunit molecular weight of approximately 40,000. This value was confirmed by gel filtration of the native protein. It was interesting to note that molecular weights of zymogens of the carboxypeptidases have been estimated at 40,000-45,000 (1,2). Some differences in kinetics were observed between the limpet and bovine carboxypeptidases. Although both enzymes exhibited dramatic excess substrate inhibition with respect to the ester, HPLA (Fig. 4B, (13)), bovine carboxypeptidase A was slightly activated by excess peptide substrate (27) while the limpet enzyme was inhibited under these conditions (Fig. 4A). Limpet carboxypeptidase was, like bovine carboxypeptidase A, highly selective toward BGP compared with the typical carboxypep30,
I
INHIBITOR
ADDED,
yL
FIG. 6. Titration of limpet carboxypeptidase with the carboxypeptidase inhibitor from potatoes (CPU. Enzyme and inhibitor (1.0 pM stock solution) were incubated prior to the addition of 3 ml of 1 mM HPLA.
PURIFICATION
OF LIMPET
tidase B substrate, BGA. Although no activity toward BGA was observed in our preparations, a careful study of extracts of fresh tissues should be performed prior to suggesting that the limpet diverged at a time before the introduction by gene duplication of separate carboxypeptidases A and B. Similarly, the failure to observe a zymogen form of the limpet enzyme is not, at this point, solid evidence that the zymogen function developed after divergence of the limpet family. Inhibition studies also have afforded a means of comparison. The limpet carboxypeptidase, like the mammalian carboxypeptidases A and B, required a metal ion for activity (Table II), and was neither a “serine” nor a “thiol” proteinase. Bovine carboxypeptidases A and B and the limpet enzyme were inactivated by L-BAMP (Fig. 5, Refs. (28, 29)). Bovine carboxypeptidases A and B have very similar amino acid sequences surrounding the amino acid residue (Glu-270 in carboxypeptidase A (28, 29)) which is modified by L-BAMP. Thus, a radically different sequence in the same vicinity of the limpet enzyme would suggest that separate carboxypeptidases A and B arose after divergence of the limpet. Both the bovine (8) and limpet enzymes (Fig. 6) were strongly inhibited by the polypeptide carboxypeptidase inhibitor from potatoes. Since a region of the inhibitor is believed to bind to a complementary region near the enzyme active site (30), limpet and bovine carboxypeptidases presumably share structural features in their respective contact zones. Although a convergent mechanism of evolution has not been ruled out by these studies, it is likely that the carboxypeptidase from the limpet, Patella uulgata, shares a common ancestor with the mammalian pancreatic carboxypeptidases. A thorough examination of the enzyme from the limpet or a closely related animal might, thus, provide considerable insight into the evolution of this class of enzymes. The relatively low amounts of enzyme present in acetone powder from the limpet dictates that fresh tissue and/or tissue from similar animals be tested for enzymatic activity prior to a fullscale investigation.
253
CARBOXYPEPTIDASE ACKNOWLEDGMENTS The author wishes to thank excellent technical assistance for his interest.
Judith E. Derr for her and Dr. A. W. Rourke
REFERENCES 1. PETRA, P. H. (1970) in Methods in Enzymology (Colowick, S. P., and Kaplan, N. O., eds.), Vol. 11, pp. 460-503, AcademicPress, NewYork. 2. HARTSUCK, J. A., AND LIPSCOMB, W. N. (1975) in The Enzymes (Boyer, P. D., ed.), Vol. III, pp. l-56, Academic Press, New York. 3. FOLK, J. E. (19’75) in The Enzymes (Boyer, P. D., ed.), Vol. III, pp. 57-79, Academic Press, New York. 4. LACKO, A. G., AND NEURATH, H. (1970) Biochemistry 9, 4680-4690. 5. REECK, G. R., AND NEURATH, H. (1972) Bioc&m&q 11, 3947-3955. 6. GATES, B. J., AND TRAVIS, J. (1973) Biochemistry 12, 1867- 1874. 7. FERRELL, R. E., CAMACHO, Z., AND KITTO, G. B. (1975) Biochem. Biophys. Acta 386, 260-269. 8. RYAN, C. A., HASS, G. M., AND KUHN, R. W. (1974) J. Biol. &em. 249, 5495-5499. 9. HASS, G. M., AND NEURATH, H. (1971) Biochemistry 10, 3535-3540. 10. AGER, S. P., AND HASS, G. M. (1977) Anal. B&hem. 83, 285-295. 11. PARIKH, I., AND CUATRECASAS, P. (1972) Biochemistry 11, 2291-2299. 12. GOA, J. (1953) Stand. J. Clin. Lab. Invest. 5, 218-222. 13. MCCLURE, W. O., NEURATH, H., AND WALSH, K. A. (1964) Biochemistry 3, 1897-1901. 14. FOLK, J. E., AND SCHIRMER, E. W. (1963) J. Biol. Chem. 238, 3884-3894. 15. WINTERSBERGER, E., Cox, D. J., AND NEURATH. H. (1962) Biochemistry 1, 1069-1078. 16. GREEN, N. M., AND WORK, E. (1953) Biochem. J. 54, 347-352. 17. WHITTAKER, J. R. (1963) Anal. Chem. 35, 1950-1953. 18. WEBER, K., AND OSBORN, M. (1969) J. Biol. Chem. 244, 4406-4411. 19. BRADSHAW, R. A., NEURATH, H., AND WALSH, K. A. (1969) Proc. Nut. Acad. Sci. USA 63, 406-411. 20. MARINKOVIC, D. V., AND MARINKOVIC, J. N. (1976) B&hem. Med. 14, 125- 134. 21. FOLK, J. E., PIE& K. A., CARROLL, W. R., AND GLADNER, J. (1960) J. Bill. Chem. 235, 2272-2277. 22. TARTAKOFF, A., GREENE, L. J., AND PALADE, G. E. (1974) J. Biol. Chem. 249, 7420-7431. 23. DIXIT, A., AND DUA, R. D. (1974) Indian J. B&hem. Biophys. 11, 227-229.
254 24.
G. MICHAEL RIJNNEGAR, B., AND POJETA, JR., J. (1974) Science 186, 311-317.
25. SEBER, J. F., TOOMEY, T. P., POWELL, J. T., BREW, K., AND AWAD, JR., W. M. (1976) J. Btil. Chem. 251,204-208. 26. GAGE-WHITE, L., HUNT, B. E., HASS, G. M., AND AWAD, JR., W. M. (1977) Fed. PTOC. 36, 892.
HASS 27. WHITAKER, J. R., MENGER, F., AND BENDER, M. L. (1966) Biochemistry 5,386-392. 28. HASS, G. M., AND NEURATH, H. (1971) Btichemistry 10,3541-3546. 29. HASS, G. M., GOVIER, M., GRAHN, D. T., AND NEURATH, H. (1972) Biochemistry l&3787-3792. 30. HASS, G. M., AKO, H., GRAHN, D. T., AND NEURATH, H. (1976) Biochemistry 15,93- 100.
OF BIOCHEMISTRY
AND BIOPHYSICS
Vol. 198, No. 1, November, pp. 247-254, 1979
Purification
and Partial Characterization of a Carboxypeptidase from the Limpet (Pate/la vu/gata)l G. MICHAEL
Department
of Bacteriology
HASS University of Idaho,
and Biochemistry,
Idaho 838&’
Moscow,
Received May 8, 1979; revised June 14, 19’79 A carboxypeptidase A-like enzyme has been purified to apparent homogeneity from commercially available acetone powder from the visceral hump of the limpet, Patella vulgata. A two-step procedure involving affinity chromatography on l -amino-N-caproyl-Dphenylalanine-Sepharose and gel filtration resulted in a 3000-fold purification with an 80% yield. The enzyme is a single polypeptide chain of M, = 40,000 and exhibits both peptidase and esterase activities, which are characterized by dramatic excess substrate inhibition. Inhibition studies suggest that a metal ion is required for activity and demonstrate that the affinity label, N-bromoacetyl-N-methyl-L-phenylalanine, and a polypeptide carboxypeptidase inhibitor from potatoes (apparent K, approx. 2 nM) are effective against the limpet enzyme.
In contrast to the attention given to digestive enzymes of the “serine” endoproteinase family (e.g., trypsin, chymotrypsin) from a variety of species, relatively little information is available about the pancreatic carboxypeptidases and their evolutionary relatives. Most of our understanding of the latter class of enzymes comes from studies on the bovine and porcine carboxypeptidases A and B (for reviews see Refs. (l-3)). However, corresponding enzymes from the dogfish (4), lungflsh (5), shrimp (6), starfish (7), and others have been purified and partially characterized. These species possess pancreatic or hepatopancreatic systems and have evolved from the line leading to mammals at a relatively recent date. Not surprisingly, these carboxypeptidases are similar to the mammalian enzymes in size (M, approx. 35,000) and other properties. An investigation of the carboxypeptidase from the visceral hump of the limpet (Patella vulgata) was undertaken to allow a comparison of the mammalian carboxy-
Materials. Acetone powder from the visceral hump of the limpet (Patella v&g&a), standard proteins, substrates, and diisopropyl fluorophosphate were purchased from Sigma Chemical Company. The polypeptide carboxypeptidase inhibitor was prepared from Russet Burbank potatoes as described by Ryan et al. (8). N-Bromoacetyl-N-methyl-L-phenylalanine (LBAMP)* was synthesized as described previously (9). Sephadex G-100 SF and Sepharose 4B were purchased from Pharmacia Fine Chemicals. c-Amino-N-caproyln-phenylalanine- Sepharose was prepared as described for the arginine derivative reported earlier (10) using the procedures of Parikh and Cuatrecasas (11). The affinity resin contained approximately 2 pmol D-phenylalanine/ml of packed gel resin as determined by ammo acid analysis tier acid hydrolysis. Enzyme pw$cation. Limpet acetone powder (10 g)
1 This work was supported in part by a grant from the National Institutes of Health (GM22748) and is published with the approval of the Director of the Idaho Agricultural Experiment Station as Research Paper No. 7953.
2 Abbreviations used: Mes, 2[N-morpholino] ethanesulfonic acid; SDS, sodium dodecyl sulfate; HPLA, hippuryl-rn-phenyllaetic acid; BGP, hippuryl&phenylalanine; BGA, hippuryl-Larginine; L-BAMP, Nbromoacetyl-N-methyl-Lphenylalanine.
247
peptidases with their counterpart from an extremely ancient animal. The limpet was chosen not only because this animal evolved from the main line leading to mammals over 600 million years ago but also because its digestive system is quite primitive, being primarily intracellular. EXPERIMENTAL
PROCEDURES
0003-9861/79/130247-08$02.00/O Copyright 0 1979 by Academic Press, Inc. All rights
of reproduction
in any form reserved.
248
G. MICHAEL
was gently stirred in 100 ml of 0.5 M NaCl, 0.01 M Mes (pH 6.0) at 4°C for 6 h, and the resulting suspension was centrifuged at 16,000 rpm for 30 min at 4°C. The supernatant (approx. 100 ml) was applied at room temperature to a 1.5 x 6.5-cm column of c-amino-N-caproyl-n-phenylalanine-Sepharose 4B. Fractions of approximately 8.5 ml were collected during the sample application and subsequent wash with extraction buffer. After the protein concentration in the eluate dropped to negligible levels, the buffer was changed to 10 mM r,-phenylalanine, 0.1 M Na,C03 (pH 10.0) and the fraction size decreased to 2 ml each. Fractions exhibiting activity against HPLA (see below) were pooled, dialyzed at 4°C against 10 mM Tris-HCl (pH 7.5), and lyophilized. Partially purified carboxypeptidase was dissolved in 2 ml of H20, centrifuged, and chromatographed at 4°C on a 1.5 x 80-cm column of Sephadex G-100 SF. The eluting buffer was 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5) and fractions of approximately 1.2 ml were collected. At each stage of purification aliquots were retained for protein determinations by the microbiuret procedure (12) or by absorbance at 280 nm and for enzymatic activity measurements (see below). Enzyme assays and inhibition studies. The activity of the limpet carboxypeptidase was routinely measured using 1 mM concentrations of HPLA (13) or BGP (14) as substrate. The increase in absorbance at 254 nm at 25°C was monitored using 3-ml volumes of substrate in 0.5 M NaCl, 25 mM Tris-HCl (pH 7.5). Enzyme units were expressed as micromoles substrate hydrolyzed per minute, where the hydrolysis of 1 wmol resulted in an increase of 0.12 A at 254 nm using cuvettes with a l-cm pathlength. In trial experiments, activity was determined using 1 mM hippurylL-arginine as substrate (15). Solutions of purified limpet carboxypeptidase (pH 7.5) were incubated at 25°C for 8 h with the following inhibitors at the indicated concentrations: EDTA (10 mM), l,lO-phenanthroline (10 mM), HgCl, (10 mM), diisopropyl fluorophosphate (50 mM), and bromoacetate (50 mM). Inhibition was detected by assay against 1 mM HPLA. Solutions of purified limpet carboxypeptidase in 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5) were incubated at 25°C with L-BAMP (9). Aliquots were removed at various times and assayed with HPLA as substrate to detect inactivation. Apparent Ki values for the inhibition of the limpet carboxypeptidase by the polypeptide carboxypeptidase inhibitor from potatoes (8) were estimated by the method of Green and Work (16). Carboxypeptidase and varying amounts of the inhibitor were incubated for 15 min prior to activity measurements with HPLA as substrate. The concentration of stock solutions of the inhibitor were estimated using an A&h% value
HASS
of 3.0 and by titration carboxypeptidase A.
with
solutions
of bovine
Molecular weight estimation. The molecular weight of native limpet carboxypeptidase was estimated by gel filtration on a 1.5 x 80-cm column of Sephadex G-100 SF as described by Whittaker (17). Elution buffer was 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5) and fractions of 1.2 ml were collected. The ratio of the elution volume of carboxypeptidase (Ve) to that of blue dextran 200 (Vo) was compared with Ve/Vo values for standard proteins (bovine carboxypeptidase A, carbonic anhydrase, bovine serum albumin, chymotrypsinogen A, ovalbumin, and horse heart myoglobin). The subunit molecular weight was estimated by polyacrylamide gel electrophoresis in buffers containing SDS (18). The carboxypeptidase and protein standards listed above were treated with 2-mercaptoethanol prior to electrophoresis.
RESULTS
Purij?cation The purification of the carboxypeptidase found in extracts of acetone powder of the visceral hump of the limpet (P. vulgata) was achieved using two chromatographic steps (Table I). In the 6rst step approximately 1500 mg of protein, which was derived from 10 g of acetone powder, was applied at a high salt concentration (0.5 M) to a column of l -amino-N-caproyl-nphenylalanine-Sepharose. Almost all of the enzyme bound to the resin and most of the contaminants passed through (Fig. 1). The elution of the carboxypeptidase was effected at pH 10.0 using buffer containing 10 mM L-phenylalanine. This step resulted in a purification of over 600-fold with a yield of nearly 90%. Chromatography of partially purified carboxypeptidase on Sephadex G-100 (Fig. 2) resulted in an additional fivefold purification with little loss of enzymatic activity. Material at this stage of purification appeared to be homogeneous by gel electrophoresis in the presence of SDS (see below). Molecular
Weight and Subunit
Structure
The molecular weight of native limpet carboxypeptidase was estimated by gel filtration (17). A comparison of the Ve/Vo values of the carboxypeptidase with those of several standard proteins (Fig. 3A) suggested a molecular weight of approximately 40,000.
PURIFICATION
OF LIMPET TABLE
PURIFICATION
Fraction Extract Affinity chromatography G-100
OF
I
LIMPET CARBOXYPEPTIDASE
Protein bg)
Activity (units)
Specific activity” (units/mg)
Yield (%I
1610*
21.0
0.013
100
1
89 81
653 3270
2.2” 0.4”
18.7 17.0
8.5 42.5 (55?
LIHippuryl-DLphenyllactate (1 mM) used as substrate. * Determined by microbiuret procedure (12). c Estimated from Atso assuming A&f,% = 1.0. d Determined by titration with the polypeptide carboxypeptidase
The subunit molecular weight of limpet carboxypeptidase was estimated by polyacrylamide gel electrophoresis in the presence of SDS (18). The value of approximately 40,000 which was obtained by this procedure (Fig. 3B) was in good agreement with that estimated for the native protein indicating that limpet carboxypeptidase was a single polypeptide chain. Kinetics Limpet carboxypeptidase was shown to hydrolyze the typical carboxypeptidase A substrates, BGP and HPLA. No activity was observed with BGA, a substrate often used to assay carboxypeptidase B-like enzymes, under conditions such that a rate 1% of that observed for BGP would have been detectable. The dependence of initial velocity upon substrate concentration for BGP and HPLA is given in Figs. 4A and B, respectively. Although the K, values were much too low to be estimated accurately, approximate values of 10 ,uM were suggested by these data. Of particular interest was the dramatic excess substrate inhibition apparent for both the ester and peptide substrates. Inhibition
249
CARBOXYPEPTIDASE
Studies
The activity remaining after limpet carboxypeptidase was incubated for 8 h with a variety of inhibitors is given in Table II. Both chelating agents (l,lO-
Purification ( -fold)
inhibitor (see text for details).
phenanthroline and EDTA) completely eliminated activity. HgCl, which binds to cysteine residues as well as to other sites was partially inhibitory while bromoacetate, a reagent fairly specific for thiols under these conditions, had no significant effect. Similarly, high concentrations of diisopropyl fluorophosphate were not inhibitory, suggesting that limpet carboxypeptidase did not have an active site serine residue. Limpet carboxypeptidase was inactivated by L-BAMP, an affinity label for bovine carboxypeptidase A (9). At each reagent
"
40
i-'-'-'-i
FRACTION
N”hmER
FIG. 1. Chromatography of crude extract from visceral hump of P. vulguta on a column (0.9 x 15 cm) of l -amino-N-caproyl-~phenylalanine-Sepharose. The extract was applied in 0.5 M NaCl, 10 mM Mes (pH 6.0), and the carboxypeptidase eluted (J) with 10 mM L-phenylalanine, 100 mM sodium carbonate (pH 10.0). Fractions l-52 were approximately 8.5 ml each while 53-70 were 2 ml each. Absorbance at 286 nm (0) and activity against HPLA (A) were monitored.
250
G. MICHAEL
HASS
FIG. 2. Chromatography of partially purified limpet carboxypeptidase on a column (1.5 x 80 cm) of Sephadex G-100 SF. The buffer was 0.1 M NaCl, 20 mM Tris-HCl (pH 7.5), and 1.2-ml fractions were collected. Absorbance at 280 nm (----) and activity against HPLA were monitored, and the fractions indicated by the bar were pooled.
concentration indicated on Fig. 5 inactivation was pseudo-first order as expected. The t,,z for inactivation at the highest concentration of reagent tested was 27 min, a value much greater than that of 3 min observed for bovine carboxypeptidase A in a similar experiment (9). The linear relationship between the pseudo-first-order constant and inhibitor concentration allowed calculation of an apparent second-order rate constant of 2.5 x 10m4 M-’ min-’ for the inactivation reaction.
Limpet carboxypeptidase was also inhibited by the polypeptide carboxypeptidase inhibitor from potatoes (8) (Fig. 6). The linearity observed at low inhibitor levels between percentage activity remaining and amount of inhibitor present permitted titration of the amount of enzyme in assay solutions. In this experiment inhibitor and enzyme were equimolar at 60 ~1 of inhibitor solution. Based upon the observed activity in the absence of inhibitor and a molecular weight of 40,000 for the limpet enzyme,
I
1.8z
*
‘.6
0.6-
I
83
1.4-
2
\
9
0.4-
--,
3
G
4
$
0.2-
\
h-2-
A
4.2 (A)
5.0
LOG
MOLECULAR
WEIGHT
42 (B)
4.4
LOG
MOLECULAR
4.6
4.8
WEIGHT
FIG. 3. Molecular weight of limpet carboxypeptidase. (A) The native molecular weight was estimated by gel flltration on a column (1.5 x 80 cm) of Sephadex G-100 SF (17). Protein standards were (1) carbonic anhydrase, (2) carboxypeptidase A, (3) ovalbumin, and (4) bovine serum albumin. Arrow indicates Ve/Vo value for limpet carboxypeptidase. (B) The subunit molecular weight was estimated by polyacrylamide gel electrophoresis in the presence of SDS (18). Protein standards were (1) horse heart myoglobin, (2) bovine carboxypeptidase A, (3) ovalbumin, and (4) bovine serum albumin. Arrow indicates relative mobility of limpet carboxypeptidase.
PURIFICATION
OF LIMPET
FIG. 4. Double-recim+ocal plot for the limpet carboxypeptidase-catalyzed and HPLA (B) at pH-7.5. -
specific activities of 55 and 30 units/mg were estimated (Table I) using 1 mM HPLA and 1 mM BGP, respectively. Apparent Ki values for the carboxypeptidase inhibitor estimated from free enzyme and inhibitor concentrations at equivalence were approximately 0.8-2 nM (6). DISCUSSION
251
CARBOXYPEPTIDASE
hydrolysis of BGP (A)
primitive basal gastropod and probably arose in the early Cambrian period approximately 600 million years ago (24). Among the group listed above the only animals that approach this early time of evolution are the starfish, an Asteroidian, which evolved in the Devonian period some 475 million years ago, and the shrimp, a Crustacean, which is dated at approximately 400 million years. The lungfish arose in the Devonian period about 350 million years ago, approximately 200-210 million years ago before modern shark (e.g., dogfish). The limpet and the shrimp are protostomates (i.e., mouth derived from blastospore) and are to be distinguished from such deuterostomates (i.e., anus derived from blastospore) as the dogfish, lungfish, and starfish. Protostomates and
The pancreatic carboxypeptidases comprise a particularly interesting class of enzymes from the standpoint of studying the evolution of structure and function. Not only is it intriguing to trace the development of carboxypeptidases in going from bacterial to mammalian sources, but the introduction of the zymogen function and the alteration in specificity from ancestral carboxypeptidase to carboxypepTABLE II tidase A-like and carboxypeptidase B-like enzymes are also worthy of investigation INHIBITION OF LIMPET CARBOXYPEPTIDASE (19). In attempting to elucidate these areas, pancreatic carboxypeptidases from a Concentration Percentage variety of mammals (e.g., human (ZO), Inhibitor activity” (M) bovine (l-3), porcine (21), guinea pig None 100 (22), goat (23), etc.) and similar enzymes EDTA 0.01 0 from such lower forms as dogfish (4), lungl,lO-Phenanthroline 0.01 0 fish (5), shrimp (6), and starfish (7) have 0.01 20 R&l, been purified and partially characterized. Diisopropyl Unfortunately, these enzymes appear to 0.05 fluorophosphate 99 exhibit few significant differences either in Bromoacetate 0.05 98 structure (e.g., all have M, ea. 34,000) or kinetic properties. Patella, a gastropod of a Enzyme was assayed with HPLA after an 8-h incuthe order Archaeogastropoda, is the most bation with inhibitors at the concentrations indicated.
252
G. MICHAEL
FIG. 5. Pseudo-first-order rate constant, k, for inactivation as a function of N-bromoacetyl-N-methylL-phenylalanine (L-BAMP) concentration.
deuterostomates have presumably evolved along separate lines. From an evolutionary standpoint a carboxypeptidase from Streptomyces griseus may well afford the most interesting comparison with its mammalian pancreatic counterparts (25). That this bacterial enzyme and carboxypeptidases A and B share a common ancestral protein is suggested by their similarities in metal ion requirement, in size, and in interaction with a polypeptide inhibitor from potatoes (26). The investigation of a digestive carboxypeptidase from the limpet, Patella vulgata, was initiated since this animal is extremely primitive, having evolved from the main line over 600 million years ago, during which period little anatomical change has occurred. It was thus hoped that a thorough investigation of the limpet enzyme would elucidate the evolution of the pancreatic carboxypeptidases and, perhaps, provide a link between bacterial and mammalian enzymes. The purification of the limpet enzyme was easily achieved in two steps with affinity chromatography on l -amino-N-caproyl-Dphenylalanine being unusually effective (Fig. 1, Table I). The utility of this resin is not unexpected in view of the low K, of the enzyme for its substrates and a low Ki value with respect to D-phenylalanine.3 It was anticipated that immobilized carboxypeptidase inhibitor from potatoes would be 3 G. M. Hass, unpublished observations.
HASS
useful in the purification of the limpet enzyme since a tight enzyme-inhibitor association was observed (Fig. 6) and since inhibitor- Sepharose had been successfully used to purify bovine and porcine carboxypeptidase (10). However, limpet carboxypeptidase bound to inhibitor-Sepharose could not be eluted without loss of activity. The first major difference between the limpet and mammalian carboxypeptidase became apparent when purified enzyme was subjected to polyacrylamide gel electrophoresis in the presence of SDS. Unlike the digestive carboxypeptidases listed above, all of which have molecular weights of 34,000-35,000 the limpet enzyme has a subunit molecular weight of approximately 40,000. This value was confirmed by gel filtration of the native protein. It was interesting to note that molecular weights of zymogens of the carboxypeptidases have been estimated at 40,000-45,000 (1,2). Some differences in kinetics were observed between the limpet and bovine carboxypeptidases. Although both enzymes exhibited dramatic excess substrate inhibition with respect to the ester, HPLA (Fig. 4B, (13)), bovine carboxypeptidase A was slightly activated by excess peptide substrate (27) while the limpet enzyme was inhibited under these conditions (Fig. 4A). Limpet carboxypeptidase was, like bovine carboxypeptidase A, highly selective toward BGP compared with the typical carboxypep30,
I
INHIBITOR
ADDED,
yL
FIG. 6. Titration of limpet carboxypeptidase with the carboxypeptidase inhibitor from potatoes (CPU. Enzyme and inhibitor (1.0 pM stock solution) were incubated prior to the addition of 3 ml of 1 mM HPLA.
PURIFICATION
OF LIMPET
tidase B substrate, BGA. Although no activity toward BGA was observed in our preparations, a careful study of extracts of fresh tissues should be performed prior to suggesting that the limpet diverged at a time before the introduction by gene duplication of separate carboxypeptidases A and B. Similarly, the failure to observe a zymogen form of the limpet enzyme is not, at this point, solid evidence that the zymogen function developed after divergence of the limpet family. Inhibition studies also have afforded a means of comparison. The limpet carboxypeptidase, like the mammalian carboxypeptidases A and B, required a metal ion for activity (Table II), and was neither a “serine” nor a “thiol” proteinase. Bovine carboxypeptidases A and B and the limpet enzyme were inactivated by L-BAMP (Fig. 5, Refs. (28, 29)). Bovine carboxypeptidases A and B have very similar amino acid sequences surrounding the amino acid residue (Glu-270 in carboxypeptidase A (28, 29)) which is modified by L-BAMP. Thus, a radically different sequence in the same vicinity of the limpet enzyme would suggest that separate carboxypeptidases A and B arose after divergence of the limpet. Both the bovine (8) and limpet enzymes (Fig. 6) were strongly inhibited by the polypeptide carboxypeptidase inhibitor from potatoes. Since a region of the inhibitor is believed to bind to a complementary region near the enzyme active site (30), limpet and bovine carboxypeptidases presumably share structural features in their respective contact zones. Although a convergent mechanism of evolution has not been ruled out by these studies, it is likely that the carboxypeptidase from the limpet, Patella uulgata, shares a common ancestor with the mammalian pancreatic carboxypeptidases. A thorough examination of the enzyme from the limpet or a closely related animal might, thus, provide considerable insight into the evolution of this class of enzymes. The relatively low amounts of enzyme present in acetone powder from the limpet dictates that fresh tissue and/or tissue from similar animals be tested for enzymatic activity prior to a fullscale investigation.
253
CARBOXYPEPTIDASE ACKNOWLEDGMENTS The author wishes to thank excellent technical assistance for his interest.
Judith E. Derr for her and Dr. A. W. Rourke
REFERENCES 1. PETRA, P. H. (1970) in Methods in Enzymology (Colowick, S. P., and Kaplan, N. O., eds.), Vol. 11, pp. 460-503, AcademicPress, NewYork. 2. HARTSUCK, J. A., AND LIPSCOMB, W. N. (1975) in The Enzymes (Boyer, P. D., ed.), Vol. III, pp. l-56, Academic Press, New York. 3. FOLK, J. E. (19’75) in The Enzymes (Boyer, P. D., ed.), Vol. III, pp. 57-79, Academic Press, New York. 4. LACKO, A. G., AND NEURATH, H. (1970) Biochemistry 9, 4680-4690. 5. REECK, G. R., AND NEURATH, H. (1972) Bioc&m&q 11, 3947-3955. 6. GATES, B. J., AND TRAVIS, J. (1973) Biochemistry 12, 1867- 1874. 7. FERRELL, R. E., CAMACHO, Z., AND KITTO, G. B. (1975) Biochem. Biophys. Acta 386, 260-269. 8. RYAN, C. A., HASS, G. M., AND KUHN, R. W. (1974) J. Biol. &em. 249, 5495-5499. 9. HASS, G. M., AND NEURATH, H. (1971) Biochemistry 10, 3535-3540. 10. AGER, S. P., AND HASS, G. M. (1977) Anal. B&hem. 83, 285-295. 11. PARIKH, I., AND CUATRECASAS, P. (1972) Biochemistry 11, 2291-2299. 12. GOA, J. (1953) Stand. J. Clin. Lab. Invest. 5, 218-222. 13. MCCLURE, W. O., NEURATH, H., AND WALSH, K. A. (1964) Biochemistry 3, 1897-1901. 14. FOLK, J. E., AND SCHIRMER, E. W. (1963) J. Biol. Chem. 238, 3884-3894. 15. WINTERSBERGER, E., Cox, D. J., AND NEURATH. H. (1962) Biochemistry 1, 1069-1078. 16. GREEN, N. M., AND WORK, E. (1953) Biochem. J. 54, 347-352. 17. WHITTAKER, J. R. (1963) Anal. Chem. 35, 1950-1953. 18. WEBER, K., AND OSBORN, M. (1969) J. Biol. Chem. 244, 4406-4411. 19. BRADSHAW, R. A., NEURATH, H., AND WALSH, K. A. (1969) Proc. Nut. Acad. Sci. USA 63, 406-411. 20. MARINKOVIC, D. V., AND MARINKOVIC, J. N. (1976) B&hem. Med. 14, 125- 134. 21. FOLK, J. E., PIE& K. A., CARROLL, W. R., AND GLADNER, J. (1960) J. Bill. Chem. 235, 2272-2277. 22. TARTAKOFF, A., GREENE, L. J., AND PALADE, G. E. (1974) J. Biol. Chem. 249, 7420-7431. 23. DIXIT, A., AND DUA, R. D. (1974) Indian J. B&hem. Biophys. 11, 227-229.
254 24.
G. MICHAEL RIJNNEGAR, B., AND POJETA, JR., J. (1974) Science 186, 311-317.
25. SEBER, J. F., TOOMEY, T. P., POWELL, J. T., BREW, K., AND AWAD, JR., W. M. (1976) J. Btil. Chem. 251,204-208. 26. GAGE-WHITE, L., HUNT, B. E., HASS, G. M., AND AWAD, JR., W. M. (1977) Fed. PTOC. 36, 892.
HASS 27. WHITAKER, J. R., MENGER, F., AND BENDER, M. L. (1966) Biochemistry 5,386-392. 28. HASS, G. M., AND NEURATH, H. (1971) Btichemistry 10,3541-3546. 29. HASS, G. M., GOVIER, M., GRAHN, D. T., AND NEURATH, H. (1972) Biochemistry l&3787-3792. 30. HASS, G. M., AKO, H., GRAHN, D. T., AND NEURATH, H. (1976) Biochemistry 15,93- 100.
