Characterization of a nonproteolytic arginine ester-hydrolyzing enzyme from snake venom.

Abstract An enzyme that catalyzes the hydrolysis of N-benzoyl-l-arginine ethyl ester and p-toluenesulfonyl-l-arginine methyl ester has been isolated from the venom of Agkistrodon contortrix laticinctus (broadbanded copperhead) by means of DEAE-cellulose chromatography. A high degree of homogeneity is suggested by sedimentation velocity, gel filtration, polyacetate electrophoresis, and isoelectric focusing. The purified enzyme has a sedimentation coefficient of 2.7 S, a diffusion coefficient of 8.3 x 10-7 cm2 per sec, and a molecular weight of 30,000. The isoelectric point, as determined by means of isoelectric focusing, was found to be 9.1. Enzymatic assays showed the preparation to be specific for arginine esters. The Km values determined with N-benzoyl-l-arginine ethyl ester and p-toluenesulfonyl-l-arginine methyl ester are 1.17 x 10-4 and 1.49 x10-3, respectively. The enzyme was inhibited by diisopropyl fluorophosphate, phenylmethylsulfonyl fluoride, and N-bromosuccinimide. Optimal rates of hydrolysis were observed from pH 7.5 to pH 8.5.

A high degree of homogeneity is suggested by sedimentation velocity, gel filtration, polyacetate electrophoresis, and isoelectric focusing. The purified enzyme has a sedimentation coefficient of 2.7 S, a diffusion coefficient of 8.3 X lo-' cm2 per set, and a molecular weight of 30,000. The isoelectric point, as determined by means of isoelectric focusing, was found to be 9.1. Enzymatic assays showed the preparation to be specific for arginine esters.
The K, values determined with N-benzoyl-L -arginine ethyl ester and p -toluenesulfonyl -L -arginine methyl ester are 1.17 X 10W4 and 1.49 x 10d3, respectively. The enzyme was inhibited by diisopropyl fluorophosphate, phenylmethylsulfonyl fluoride, and N-bromosuccinimide. Optimal rates of hydrolysis were observed from pH 7.5 to pH 8.5.
The ability of various venoms to hydrolyze amino acid esters has been well documented.
Recent evidence suggests that these esterases might play an important role in local tissue destruction (1,2). Tu, Passey, and Tu (3) and Tu, Chua, and James (4) tested the venoms of all four families of venomous snakes for their ability to hydrolyze p-toluenesulfonyl-L-arginine methyl ester and N-benzoyl-n-arginine methyl ester, and concluded that only venoms of the families Crotalidae and Viperidae possess enzymes capable of this hydrolysis.
In these same investigations, the substrates N-acetyl-n-tyrosine ethyl ester and Nbenzoyl-L-tyrosine ethyl ester were not hydrolyzed by most of the venoms, suggesting that the substrate specificities were similar to those of trypsin, rather than chymotrypsin.
However, venom enzymes were not inhibited by soybean or ovomucoid trypsin inhibitors, showing that venom enzymes were different from trypsin. * This work was supported by United States Public Health Service Grants 2ROl GM15591 and 5ROl FD-00014.
1 Recipient of Career Development Award l-K4 GM 41, 786-01 from National Institute of General Medical Sciences, National Institutes of Health. To whom requests for reprints should be sent.
To date, a study of the various physical and enzymatic parameters of a purified amino acid esterase has not been reported. This paper will describe a procedure for the isolation of an esterase shown to be homogeneous by sedimentation velocity, polyacetate electrophoresis, gel filtration, and isoelectric focusing, and report on some of its enzymatic and physical properties. Enzym.e Assays-Assays for proteolytic activity using casein as substrate were carried out following the modified method of Kunitz (5) as previously described (I).
Those proteolytic enzyme assays in which hemoglobin was used as substrate were performed by incubating 0.5 ml of enzyme solution (1 mg per ml) with 1.0 ml of 2% urea-denatured hemoglobin (in 0.01 M NaH2P04, pH 7.0) at 37", for 30 min (6). The reaction was terminated by the addition of 2.0 ml of 5% trichloracetic acid, and the absorbance of the acid-soluble products was determined at 280 mp. Proteolytic activity, as measured by the hydrolysis of the synthetic substrates Azocoll and Congocoll (Calbiochem) (7), was determined by incubating 5 mg of substrate in 3.0 ml of Tris-HCl buffer (0.1 M, pH 8.5) with 0.5 ml of enzyme solution (1 mg per ml). After 30 min, the solution was filtered and the absorbance of the liberated dye measured at 495 rnp (Congocoll) or 580 rnp (Azocoll).
Esterase activity toward BAEE' was routinely followed during the course of purification by the spectrophotometric method previously described (1). Enzymatic activities against casein, hemoglobin, Azocoll, Congocoll, and BAEE were expressed as specific activity = (absorbance change per min)/(milligrams of venom) X 1000. A titrimetric method was employed to determine the hydrolysis of the esters and amides listed in Table III The inhibition by N-bromosuccinimide was tested by adding 40 ~1 of a 25 mM solution of N-bromosuccinimide in sodium acetate buffer (50 mM, pH 4.5) to 1.5 ml of venom dissolved in the same buffer. After a 30.min incubation at 25", aliquots were removed and assayed for enzymatic activity.
Zsoelectric Fractionation-The electrofocusing column was filled with a linear gradient from Solution A to Solution B. The gradient was produced by means of a 1%ml divided box described by Svensson (8). Solution A consisted of 5 mg of purified enzyme and 0.70 g of carrier ampholyte dissolved in 55 ml of 50% sucrose (w/v).
Solution B consisted of 0.30 g of carrier ampho1yt.e dissolved in 55 ml of deionized water.
The applied voltage, initially 100 volts, was gradually increased to 500 volts by the end of the experiment, maintaining the power output at about 0.8 watts by manual adjustment.
After separation for 48 to 60 hours at 4", the column was drained and l&ml fractions were collected.
Chromatographic ProceduresDEAE-cellulose was suspended in the first buffer to be used in the elution procedure, and allowed to stand for several hours. After the fine particles had been decanted and the procedure repeated several times, 2 M NaCl dissolved in this buffer was added. The cellulose was then reequilibrated with buffer containin, 0 no NaCI, the column was poured and equilibrated for 24 hours with the first buffer to be used in the elution procedure.
Sephadex G-75 and G-IO were dispersed in the eluting solvent, and allowed to swell for 4 and 24 hours, respectively.
After the fine particles had been decanted several times, the gel was poured into the columns and washed for 24 hours with eluting solvent.
Column effluents were monitored with an ISCO model UA-2 ultraviolet analyzer and the elution patterns recorded with an ISCO B-inch chart recorder. Amino Acid Anal~sisA 2.5.mg enzyme sample was dissolved in 3.0 ml of constant boiling HCl and placed in a heavy walled ignition tube. The tube was sealed under vacuum and the protein was hydrolyzed for 24 hours at 110". Following hydrolysis, the sample tubes were opened and the HCl was removed by drying over NaOH pellets under vacuum.
After three washings, the residue was dissolved in 2.5 ml of 0.2 M sodium citrate buffer, pH 2.2. Amino acid analyses were performed on a Technicon amino acid analyzer, equipped with a column (1.6 x 40 cm) of Chromobeads B. and diffusion coefficients.
The schlieren patterns were recorded photographically on Eastman Kodak Metallographic plates. The plates were read using a microcomparator (Nikon Model 6C) equipped with a rotational stage. The sedimentation coefficient was calculated from the rate of movement of the maximum ordinate of the refractive index gradient.
No attempt was made to use the theoretically more correct second moment procedure of Goldberg (9) because of various experimental difficulties, particularly the low solubility of the enzyme near it.s isoelectric point.
The sedimentation velocity experiments were performed at 4" and 59,780 rpm using a double sector cell with an aluminumfilled Epon centerpiece.
The diffusion coefficient measurements were performed at 4" and 10,589 rpm using a capillary-type synthetic boundary cell according to the procedure described by Schachman (10 gation in the following manner. A saturated solution was prepared by adding excess lyophilized enzyme (1.2 mg) to 0.75 ml of buffer solution containing 0.1 M sodium glycinate at pH 9.0, followed by gentle stirring for 24 hours. The resulting suspension was centrifuged and dialyzed against two changes of the sodium glycinate buffer (2,000 volumes) for 18 hours. This solution was used in both the sedimentation velocity and diffusion experiments.
In an attempt to increase the solubility of the enzyme, a second saturated solution was prepared by addition of excess enzyme (1.2 mg) to 0.75 ml of a buffered solution of 0.1 M Tris-HCI, 0.1 M NaCl at pH 7.0. The dissolution and dialysis procedure were as stated above. The sedimentation patterns obtained using this solution are shown in Fig. 6. The dialyzate was used in the reference sector in each case. Purified bovine serum albumin was used as a standard to check on instrumental and technical errors in the evaluation of both sedimentation and diffusion.

M).
A representative chromatogram is shown in Fig. 1A. The highest esterase activity toward BAEE was found in Fraction I, with Fraction III also displaying some activity.
These two fractions also exhibited the highest proteolytic activity. Our primary interest was in the enzyme represented by Fraction I, and thus, no attempt was made to further purify the other fractions.

Purification of Enzyme Lyophilized
A. contort& Zaticinctus venom (1.0 g) was dissolved in 5.0 ml of 0.01 M Tris-HCI, pH 8.5, and dialyzed for 24 hours against this buffer. No loss of either proteolytic or esterase activity was observed.
The dialyzed venom sample was added to a column of DEAEcellulose previously equilibrated with 0.01 M Tris-HCI, pH 8.5. After all unadsorbed material had been completely eluted with the starting buffer, either a NaCl or a Tris-HCI salt gradient, or both, was annlied to the column (0 to 0.2 M. followed bv 0.2 to Fraction I from the preceding step was lyophilized to dryness, then redissolved in 20 ml of 0.01 M Tris-HCI, pH 9.5. After dialyzing for 24 hours against this same buffer, the solution was placed on a column previously equilibrated with 0.01 M Tris, pH 9.5. The sample was eluted by means of a 500-ml gradient of either NaCl or Tris-HCI, or both, of increasing concentration (0 to 0. The elution pattern obtained is presented in Fig. 1B. Esterase activity was highest in Fraction B, with Fractions A and C also displaying slight activity toward BAEE. Proteolytic activity was highest in Fraction C, but was also present in Fractions A and B. cellulose previously equilibrated with the same buffer. As can tryptophan were determined from data obtained using enzyme be seen in Fig. lC, all material was adsorbed on the column. samples hydrolyzed for 24 hours with constant boiling HCl. A 111-1 gradient from pH 9.5 to 8.8 (0.01 M glycine) was then al2-The tryptophan content of the protein was determined spectroplied to the column, and a single, symmetrical peak was obtained.
photometrically by the method of ljencze and Schmid (12). When 0.1 M glycine, pH 8.5, was passed through the column, a Cystine was determined from hydrolysates treated with persecond peak emerged.
Activity against UAEE was found in the formic acid prior to acid hydrolysis. 13ased on 1 residue of first peak, while proteolytic activity toward casein was found in methionine, the minimum molecular weight was calculated to be the second. A small amount of esterase was also found in the 10,300. Assuming that methionine occurs 3 times in the prosecond peak.
tein, the molecular weight would be 31,000. The purified esterase was then lyophilized to dryness, dissolved in 2.0 ml of deionized water, and passed through a Sephades G-10 column using deionized water as the eluting solvent. A summary of the purification steps is presented in Table I, where it c:ul be seen that the specific esterase activity was increased about 20.fold by the isolation procedure.
The recovery of protein (11 mg) represents a l'i;L recovery from unfractionated venom.

Physicochernical Properties
Electrophoresis-A high degree of purity was indicated by electrophoretic experiments on polyacetate strips. A number of experiments were performed at pH values ranging from 5.0 to 9.0. In each instance, only a single band could be detected. An example of such an experiment is presented in Fig. 2, in which the electrophoretic pattern obtained at pH 5.5 is presented.

Xedimentation
Coeficient-When the purified enzyme was sedimented at 59,780 rpm in the ultracentrifuge, a single peak was shown at both concentrations tested. A representative series of schlieren patterns is shown in Fig. 6. The sedimentation coefficient was calculated to be sZO,w = 2.7 S after appropriate corrections for viscosity, density, and temperature.
The sedimentation coefficients in the two buffers (glycine and Tris) agree to within 3%. The concentration of enzyme in the Tris buffer was 7y0 greater as determined by integration of the refractive index gradient curves from diffusion experiments.
The low solubility and limited quantity of enzyme, coupled with the sensitivity of the schlieren optical system, precluded an evaluation of s!&,,.
Isoelectric Focusing-As can be seen from Fig. 3, the preparation also appeared to be homogeneous by means of isoelectric focusing.
This technique also established the isoelectric point to be 9.1.
Gel Filtration- Fig.  4 presents the results obtained when the purified enzyme was passed through a column of Sephadex G-75. This technique also indicated homogeneity in the enzyme preparation.
The method of Andrews (11) was used to calculate the molecular weight of the purified enzyme. Fig. 5 shows the calibration curve obtained from the G-75 column with a number of proteins of known weight.
The purified enzyme gave an elution volume corresponding to a molecular weight of 31,000.

DiJLsion
Coe$cient-An apparent diffusion coefficient was calculated by the statistical method described by Schachman (10). The diffusion coefficient was calculated to be 8.3 X lop7 cm2 set after correction to a value corresponding to water at 20".
Partial Xpecific VoZume-The partial specific volume for the enzyme was estimated from the amino acid composition according to the procedure described by Schachman (10). The partial specific volume was calculated to be 0.71 ml per g using the relation, v = Cwivi/cwi, where wi and vi are weight percent and specific volume of Residue i. d1oZecular Weight-The molecular weight, M, of the enzyme as determined by sedimentation and diffusion measurements, ilnrino /l&d Composition-The amino acid composition of the enzyme is shown in Table II  Assay for esterase was by titration with NaOH at pH 8.35 as described in text.
b Assay for proteolytic activity was with casein as described in text. Venom concentration was 0.2 mg per ml. KaHP04 and the pH was adjusted to the desired pH immediately 0.998 g per cm3, and T = 20", the molecular weight was calculated prior to enzyme assay. Each point represents V,,, at that pH.
to be 27,000. This compares with the minimum molecular Snake Venom E&erase Vol. 245, No. 10

Enzymatic Properties
Effect of pH--Bs shown in Fig. 7, the enzyme exhibited a rather broad pH optimum, being most active at pH values from 7.0 to 9.0. Below pH 7 and above pH 9, the activity dropped off rather rapidly. E$ect of Divalent Cation-As can be seen from Table III, the cations Mn+2, Zn+2, and CO+~ approximately doubled the rate of hydrolysis of BAEE by the purified enzyme. Ni+2 and Mg+2 increased the rate of hydrolysis about 50% while Ca+Z and Cd+2 had little effect. None of the metals tested resulted in restoration of proteolytic activity. Substrate Spec$city-A rather large number of substrates known to be hydrolyzed by esterases of other sources as well as some common substrates for proteases were tested (Table IV).
The substrates BAEE, N-benzoyl-L-arginine methyl ester, ptoluenesulfonyl-L-arginine methyl ester, and p-nitrophenyl acetate were hydrolyzed by the venom enzyme. When the arginine residue was replaced with alanine, lysine, or tyrosine, no hydrolysis took place. No hydrolysis could be detected on the amide bonds of L-lysine-p-nitroanilide, N-benzoyl-L-arginine amide, p-toluene-n-arginine amide, or N-benzoyl-L-arginine p-nitroanilide, even with loo-fold excess of enzyme. The synthetic substrate, indophenyl acetate, readily hydrolyzed by acetyl cholinesterase (13), was not hydrolyzed by the venom esterase. In like manner, none of the proteolytic enzyme substrates (casein, hemoglobin, Congocoll, Azocoll) were hydrolyzed by the purified esterase.   Table IV. An example of these plots is shown in Fig. 8 where l/v is plotted against l/s using BAEE as substrate. DISCUSSION The results presented herein indicate that a reproducible procedure for the isolation of an esterase from the venom of A. contortrix laticinctus has been achieved. This procedure results in the isolation of an esterase of a high degree of purity as evidenced by four criteria: electrophoresis, ultracentrifugation, chromatography on Sephadex G-75, and isoelectric focusing. Three methods of determining molecular weight (combination of sedimentation velocity and diffusion coefficient, Sephadex gel filtration, and amino acid composition) gave molecular weights of 27,000, 31,000, and 31,000, respectively.
It can be concluded that t,he molecular weight of the purified esterase is near 30,000.
There has recently been some controversy as to whether BAEE and p-toluenesulfonyl-L-arginine methyl ester can be used as substrates for the assay of proteolytic enzymes in snake venoms (4,14,15) and whether the enzyme or enzymes responsible for the hydrolysis of BAEE and p-toluenesulfonyl-n-arginine methyl ester is actually proteolytic in nature.
The evidence from this investigation strongly supports the view that the hydrolysis of the synthetic substrates by snake venom is not caused by a proteolytic enzyme. Thus the conclusions drawn by Delpierre (15) from work on partially fractionated venom, and those of Wagner, Spiekerman, and Prescott (14) based on studies of a purified protease, are confirmed in this investigation.
From the results reported herein, it appears that an esterase quite different from any esterase yet reported in literature has been isolated and characterized from snake venom.
This amino acid esterase appears to be quite specific, as shown by the fact that only esters of arginine are acted upon. However, the alkyl group does not appear to play a major role, as shown by the fact that B=1EE and N-benzoyl-L-arginine methyl ester exhibited almost identical K, and V values.
This newly isolated esterase appears to be specific for ester bonds as illustrated by the fact that even with a loo-fold excess of enzyme, no hydrolysis could be detected on any of the synthetic substrates containing amide bonds, or with any of the peptide bonds of the various protein substrates tested. Like other esterases, serine appears to be at t.he active center as both diisopropyl fluorophosphate and phenylmethylsulfonyl fluoride inhibit the enzyme. Unlike enzymes which act either as proteases or esterases depending on the cation present (16), this enzyme showed no proteolytic activity when tested in the presence of a number of divalent cations.
Further chelating agents showed only slight inhibition of hydrolysis. Under the experimental conditions employed, the strong inhibition by N-bromosuccinimide suggests that tryptophan also plays an important role in the activity of the enzyme.