Oligo(methionyl) proteins - Wiley Online Library

1 downloads 0 Views 1015KB Size Report
the digestive products of Ac"(Met)E&LysNHz, (Met)E,,,Lys,. Met4 and Met5. ..... The two remaining endopeptidases of pancreatic juice, trypsin and elastase, were ...
Eur. J. Biochem. 145, 257-263 (1984)

0FEBS 1984

Oligo(methiony1) proteins Enzymatic hydrolysis of the model isopeptides N"-oligo(L-methiony1)-L-lysine Hubert GAERTNER and Antoine PUIGSERVER Centre de Biochimie et de Biologie Moleculaire du Centre National de la Recherche Scientifique, Marseille (Received April lO/July 11, 1984)- EJB 84 0444

A number of model isopeptides containing oligo(methionine) chains varying in length (2 - 5 residues) covalently linked to the &-aminogroup of lysine were synthesized by solid-phase procedures. Hydrolysis of these peptides by pepsin, chymotrypsin, cathepsin C (dipeptidyl peptidase IV) and intestinal aminopeptidase N was investigated using high-performance liquid chromatography to identify and quantify the hydrolysis products. Methionine oligomers grafted onto lysine were cleaved to tripeptides by pepsin. Chymotrypsin preferentially hydrolyzed the methionyl-methionine bond preceding the isopeptide bond. Cathepsin C released dimethionyl units from the covalently attached polymers. Intestinal aminopeptidase caused efficient hydrolysis of both peptides and isopeptide bonds although free methionine decreased the cleavage of the latter bond. Hydrophobic characteristics of oligo(methionine) chains promoted enzyme-catalyzed transpeptidations resulting probably from acyl-transfertype reactions. Complementary hydrolysis of the isopeptides by these digestive enzymes suggests that covalent attachment of oligo(amino acid)s to food proteins may improve their nutritional value. The nutritional value of dietary proteins has been known for a long time to be improved when foods and feeds are supplemented with free essential amino acids [l]. It has recently been shown, however, that a comparable result could also be obtained by covalent attachment of the limiting amino acid to the protein by chemical [2, 31 and enzymatic methods [4, 51. Although enzyme-catalyzed attachment of amino acids to proteins is certainly easier to use than the chemical method, the latter might nevertheless offer a decisive advantage in that it may prevent deteriorative changes during the processing or storage of food proteins. This could not be achieved when fortifications involved free amino acids. The chemical grafting of essential amino acids through their cr-carboxyl to the &-aminogroup of the lysyl residues of food proteins has resulted in the formation of a so-called isopeptide bond [6], which has further been shown to be readily hydrolyzable by intestinal brush border aminopeptidase N [7, 81. This finding is of great interest since membrane-bound aminopeptidase accounts for practically a11 the peptidase activity of the intestinal mucosa 19, lo], thus explaining how covalently attached amino acids happen to be available as well as the free forms. Using N-carboxy-L-methionine anhydride, it has recently been possible to polymerize methionine on the amino groups of casein and /3-lactoglobulin. Methionine, covalently attached to the a-amino group and about 40% of the €-amino groups of the lysyl residues of each protein has been found to

reach up to 30% of protein mass leading to oligo(methionine) side chains having an average length of six or seven residues [ l l , 121. Extensive digestion of the modified proteins by porcine pepsin, activated pancreatic juice and intestinal aminopeptidase N led to comparable proportional release of all the amino acids, including grafted methionine. However, despite the fact that hydrolysis of methionyl-methionine bonds could be predicted, at least to some extent, from our knowledge about the primary specificity of proteolytic enzymes, no data are at present available on enzymatic hydrolysis of long methionine polymers. The aim of this study was to investigate the hydrolysis of methionine oligomers covalently attached to the &-amino group of lysine by pepsin, a-chymotrypsin, elastase and aminopeptidase N in order to characterize the release and consequently the biological availability of methionine polymerized on proteins. Since under the polymerization conditions of N-carboxy-L-methionine anhydride onto the amino groups of proteins, the chain Iength of the oligomers may not be known, it has been found necessary to synhesize a few model peptides and isopeptides with a well-defined number of methionyl residues. In the present study it was found that high-performance liquid chromatography was a highly effective means of obtaining rapid and accurate identification of the digestive products of Ac"(Met)E&LysNHz, (Met)E,,,Lys, Met4 and Met5.

Abbreviations. Ac"(Met)SLysNH,, W-acetyl-N"-di(L-methiony1)L-lysinamide; Ac"(Met);LysNH,, W-acetyl-N'-tri(L-methiony1)-Llysinamide; (Met)iLys, W-tetra@-methiony1)-L-lysine; (MetELys, W-penta(1.-methiony1)-L-lysine;Boc, tert-butyloxycarbonyl; HCON(CH,),, N,N-dimethylformamide; F3AcOH, trifluoroacetic acid; HPLC, high-performance liquid chromatography. Enzymes. Aminopeptidase N (EC 3.4.1 1.2); cathepsin C (EC 3.4.14.1); a-chymotrypsin (EC 3.4.21.1); elastase (EC 3.4.21.11); pepsin (EC 3.4.23.1); trypsin (EC 3.4.21.4).

EXPERIMENTAL PROCEDURES Materials

Acetonitrile, orthophosphoric acid, methylene chloride, diisopropylethylamine and F3AcOH were obtained from Merck (Darmstadt, FRG). Bio-beads, 1% cross-linked polystyrene with 1.25 mmol chloromethyl substitution/g of dry resin, and the benzhydrylamine polymer, 1YOcross-linked

258

polystyrene with 0.24 mmol NH2/g of dry resin, were supplied by Bio-Rad Laboratories (Richmond, CA, USA) and UCB (Brussels, Belgium), respectively. HCON(CH3)2, from Fluka AG (Buchs, Switzerland), was stored over a Merk 0.4-nm molecular sieve. Boc-protected amino acids, as well as dimethionine and trimethionine, were purchased from Bachem Fine Chemicals (Bubendorf, Switzerland). Pepsin, trypsin (treated with tosylphenylalanyl chloromethane) and a-chymotrypsin were from Worthington Biochemical Corp. (Freehold, NJ, USA) while elastase, carboxypeptidase A and cathepsin were obtained from Boehringer (Mannheim, FRG). Pure hog aminopeptidase N was a gift from S. Maroux (Marseille, France). All other reagents and chemicals were of analytical grade.

as described by Robinson [17]. After four washings of the peptide-resin mixtures with anhydrous ether to remove anisole, the resulting crude peptides were extracted with dilute acetic acid and finally freeze-dried. They were then purified to homogeneity by successive filtration through three Sephadex G-10 or G-15 columns (2.5 x 100 cm) mounted sequentially and equilibrated with 100 mM acetic acid. Purity was assessed by reversed-phase high-performance liquid chromatography and amino acid analysis. The latter was performed on a Beckman model 120 C autoanalyzer equipped with an ICAP-10 computer. The free peptides were oxidized with performic acid prior to hydrolysis in distilled 5.6 M HC1 at 110°C for 24 h. Synthesis of the isopeptide (MetyLys has already been reported [7].

Synthesis of model peptides and isopeptides

Enzyme digestion studies

Peptide synthesis was carried out by the solid-phase procedure [13, 141 using manual addition of all reagents and solvents. Methylene chloride was redistilled from anhydrous potassium carbonate just prior to use. Fresh solutions of diisopropylethylamine and F3AcOH in CHzC12 were made up every day. Solid-phase synthesis of AcU(Met)",LysNHz and Ac" (Met);Lys NH2, which both contain a C-terminal amide group, required preliminary acetylation of Boc'-Lys with a 5-molar excess of acetic anhydride and diisopropylethylamine in HCON(CH3)2. The resulting lysine derivative was then covalently linked to the benzhydrylamine polymer (1.5 : 1.O amino acid :resin amino groups) with two successive additions of N,N'-dicyclohexylcarbodiimide, each containing a ratio of 1.O mol N,N'-dicyclohexylcarbodiimide/mol lysine. Under these experimental conditions, the degree of substitution of the solid support could reach 82% as estimated by ninhydrin analysis [I 51. Residual amino groups were then reacted with acetic anhydride. When the chloromethyl-resin was used in order to synthesize (Metfi Lys, (Met); Lys, Met4 and Met, attachment of the first Boc-protected amino acid (Boc-Xaa) was achieved by the caesium salt method [16] leading to a 62% substitution of the solid support. Once the first amino acid (C-terminal residue of the peptide) was linked either to the chloromethyl or benzhydrylamine resin, cleavage of the Boc group was carried out under mild acidic conditions. The usual procedure here was: CHzC12 (4 x 1.5 rnin), 30% F3AcOH in CH2Clz (1 x 1.5 rnin and 1 x 30 rnin), CHzClz (8 x 1.5 rnin), 5% diisopropylethylamine in CH2C12 (2 x 1.5 rnin and 1 x 10 rnin), CHzC12 (4 x 1.5 min). Incorporation of the second Boc-Xaa (penultimate residue of the peptide) into the growing peptide chain then resulted from the NJ"' dicyclohexylcarbodiimide-coupling reaction: Boc-Xaa in CH2Clz(2.5 molar excess, 1 x 10 rnin), N,N'-dicyclohexylcarbodiimide in CHzC12(2.5 molar excess, 1 x 120 rnin), CH2Clz (4 x 1.5 rnin), HCON(CH& (2 x 1.5 min). The same procedure, with the exception of HCON(CH& treatment, was repeated using a molar ration of both Boc-Xaa and N,Wdicyclohexylcarbodiimide in order to ensure complete acylation of the resin amino groups. This was checked on resin aliquots by the ninhydrin method [15]. When the reaction stopped far short of completion an additional coupling step was realized and termination was then carried out by acetic anhydride. Removal of completed peptides from solid support as well as the Boc group still protecting the last-incorporated amino acid was achieved with liquid anhydrous hydrogen fluoride in the presence of anisole in a Toho Kasei equipment,

Hydrolysis of the model peptides and isopeptides (2 mM solutions unless otherwise indicated) by digestive enzymes was performed at 37°C (see figure legends for enzyme/substrate mass ratios). Digestion by pepsin (in 0.01 M HCl) was stopped by incubating the reaction mixture in boiling water for 10 min, while digestion by trypsin, a-chymotrypsin, elastase, carboxypeptidase A and cathepsin (in a 50 mM phosphate buffer, pH 8.2) was stopped upon addition of 10% (v/v) 1 M HCl. The resulting digestion mixtures were analyzed with an HPLC system without any further treatment of the samples. Digestion by intestinal aminopeptidase (in a 50 mM phosphate buffer, pH 7.0) was stopped as indicated for pancreatic enzymes. All the enzymes used throughout this study were of the highest analytical grade since they were found to display their usual maximal activities towards their specific synthetic substrates. Determination of extent of hydrolysis of model substrates

Extent of hydrolyses of peptides and isopeptides by intestinal aminopeptidase N was determined by separating free methionine from the remaining peptides and quantifying it with the autoanalyzer. For the other enzyme digestion mixtures a Waters Associates HPLC system was used. This consisted of an M 6000 A and an M 45 solvent-delivery unit, a M 720 solvent programmer, a U6K universal liquid chromatograph injector coupled to an M 441 ultraviolet spectrophotometer and a two-channel chart recorder (M 730 data module), and a Merck Lichrosorb CI 8 reversed-phase column (7 pm, 4 x 250 mm). The absorbance detector was set to 214 nm with a full scale of A = 0.2. Peptides were eluted by a linear gradient, from 0.1 % orthophosphoric acid (solvent A) to acetonitrile (solvent B) at a constant flow rate of 1 mI/min. Orthophosphoric acid solution (pH 2.2), which was first filtered through a 0.45pm Millipore membrane, and acetonitrile were sonicated for 15 rnin for degassing. All runs were performed at room temperature. RESULTS Synthesis and characterization of model isopeptides

Table 1 shows that the overall yield, based on the starting Ac"(Boc)"Lys-resin, of both synthesis and purification through Sephadex columns of the two amide-group-containing isopeptides, by use of the benzhydrylamine-resin, was

259 Table 1. Solid-phase synthesis of model isopeptides and peptides Benzhydrylamine-resin (0.24 mmol NH2/g) and chloromethyl-resin (1.25 mmol/g) were used for synthesis of amide-group-containing isopeptides and the remaining peptides, respectively. Crude peptide was determined after removal from resins by liquid hydrogen fluoride Isopeptide and peptide

Resin

g 3.9 3.9 1.5 1.5 2.1 2.1

Ac"(MetELysNH Ac"(Met);LysNH, (MetfiLys (Met);Lys Met, Met5

First-linked amino acid

Crude peptide

Purified peptide

mmol

mg 120 140 860 805 850 1100

mmol

0.75 0.75 1.o 1.o 1.6 1.6

0.20 0.16 0.66 0.42 0.80 0.56

Table 2. Amino acid composition of methionine-containing isopeptides Isopeptides

Met

LYS

Met/Lys

25.8 19.3 20.8 15.3

1.98 2.96 3.96 4.94

nmol Ac"(Met)iLysNH2 Ac"(MetELysNH (MetfiLys (Met);Lys

51.1 57.2 82.2 75.6

Peptide injected (nrnoll

Fig. 2. Peptide calibration curves. Curves were obtained with the elution conditions described in Fig. 1 (computer-integrated peak area versus amount of peptides injected)

(Met);Lys, Met4 and Met, were eluted at 15.7 min, 20.9 min, 22.7 min, 27.1 min, 26.7 min and 30.8 min, respectively. The high purity of isopeptides was confirmed by amino acid composition as indicated in Table 2. HPLC separation and estimation of peptides Fig. 1. Reversed-phase HPLC ofmethionine oligomers. Chromatography was done on a 7-pm Lichrosorb RP18 column (4 mm x 25 cm) using a linear 30-min gradient from water/O.l% orthophosphoric acid to acetonitrile/water/O.l YOorthophosphoric acid (60-40%) with a flow rate of 1 mlimin

rather low (about 25%). This was mainly due to inefficient removal of the completed peptides from the solid support by liquid hydrogen fluoride. By contrast, with the chloromethylresin synthesis efficiency was higher and reached 66% with (MetELys. In all six cases, the material emerging from the Sephadex columns as a result of successive filtration was found by reversed-phase HPLC to be homogeneous. When the concentration of the mobile-phase modifier (acetonitrile) was increased linearly from 0 to 60% over a 40-min period (1.5%/min), A",Met)"2LysNHz, A:(Met)"3LysNHz, (MetfiLys,

The time course of enzymatic digestion of synthetic peptides and isopeptides could be easily followed by means of the HPLC system since Fig. 1 indicates that high resolution of methionine and its related oligomers up to pentamethionine was quite possible. Under the experimental conditions outlined in the figure legend, there was an increase in the retention time of methionine peptides as the number of methionyl residues increased. Calibration curves which compared the HPLC detector response at 214 nm, as computer-integrated peak areas, to the amount of methionine and related oligomers injected showed a linear relationship in the 2-40-nmol range (Fig. 2). The precise HPLC response for all the synthesized isopeptides was also determined. The isopeptides (MetELys and (MetELys resulting from enzymatic hydrolysis of (MetfiLys and (Met);Lys were clearly identified not only by their retention time on HPLC but also by quantitative amino acid analysis of the material eluted from the reversed-phase column. The small amounts of residual

260

Fig. 3. HPLC elution proJiles of peptic digests of (Met)\Lys ( A ) and Met4 ( B ) . Chromatographic conditions: (A) a 30-min and (B) an 18min linear gradient of 0.1% H 3 P 0 4 to acetonitrile/O.l% &Po4 (60-40%) with a flow rate of 1 ml/min. Aliquots (20 111) removed from the incubation mixture were denatured by heating and injected directly into the liquid chromatograph. Enzyme:peptide = 1 : 100 (w/w). (1) (Met)"Lys, (2) (MetELys, (3) Met3, (4) (MetELys, ( 5 ) Met4, (6) (Met);Lys, (7) MetI

Fig. 4. Time course of hydrolysis by pepsin of a 2 mM solution of (Met);Lys ( A ) and (Met)",ys ( B ) . Enzyme:peptide (.+----+) (Met)"Lys; (A----A) (MetELys; (0----0) (MetXLys; (0-0) Met3; (0-0) Met4

=

1 : 100 (w/w).

HPLC resolution was obtained with the peptic digests (MetELys and Met5 (data not shown). The former was essentially hydrolyzed to Met, and (Met)"Lys but the rate of Met, release was 5-times smaller than that obtained with (MetKLys, as indicated in Fig. 4. Moreover, minor digestion products such as Met, resulting from a transpeptidation reaction, (MetELys and (MetXLys could also be identified in Peptic and chymotryptic hydrolysis the incubation mixture. With the isopeptides containing a of synthetic peptides and isopeptides C-terminal amide group [Ac"(Met)",LysNH2 and All the peptides resulting from peptic digestion of (Met)&,- Ac"(Met)"zLysNHz]no hydrolysis was observed even when a Lys and Met4 were readily separated within 30 min under the high concentration of pepsin was used (enzyme: substrate experimental conditions indicated in Fig. 3. The isopeptide = 1 : 10, w/w). From pepsin digestion studies, it therefore emerged that (MetKLys was cleaved into Met, and (Met)qLys and to a lesser extent to Met4 and (Met)"Lys. Peak number (4) in the methionine oligomers covalently attached to the &-amino chromatogram was identified as (Met)6Lys and represented a group of lysine could be efficiently cleaved provided they contaminating peptide material (less than 2%) which was contained at least four residues. The non-hydrolysis of Met, initially present in the synthesized isopeptide. Comparable and the effective cleavage of Met4 and Met, are also consistent H3P04 contained in the collected samples have not been observed to produce any adverse effects during amino acid analyses. It was therefore possible to consider undertaking a study on quantitative hydrolysis of the synthesized peptides and isopeptides by enzymes of the digestive tract.

261

Fig. 5. Time course of hydrolysis by a-chymotrypsin of a 2mM solutioa of (Met);Lys ( A ) and (A4et)Z.w ( B ) . Enzyme:pePtide = 1:100 (w/w). (+----+) (MetYLys; (A----A)(MetELys; (0----0) (MetELys; (A-A) Met2; (0-0) Met3; (0--0) Met4

bond located either one or two residues ahead of the isopeptide bond by which methionine oligomers were covalently linked to the &-aminogroups of lysyl residues. The two remaining endopeptidases of pancreatic juice, trypsin and elastase, were quite unable to split off any of the methionylmethionine bonds of either the peptides or the isopeptides synthesized in this work.

Digestion studies with exopeptidases The other alimentary tract enzymes which might by involved in hydrolyzing methionine oligomers are carboxypeptidase A and the two membrane-bound peptidases of the enterocyte brush border, dipeptidylpeptidase IV and aminopeptidase N. As expected, carboxypeptidase A was found to hydrolyze Fig. 6. Time course of release by pig aminopeptidase N of the isodipep- only C-terminal-free methionine oligomers. Consequently, tide from Ac" (Met)%LysNH2( a ) . Ac" (Met)%LysNH2( b ) , (Met)$- this enzyme will not be directly involved in methionine release from the polymers covalently linked to the &-aminogroup of Lys ( c ) and (Met);Lys ( d ) . Enzyme:peptide = 1 :500 (wlw) lysine. Since we were short of dipeptidylpeptidase IV, we decided to use cathepsin C which is known to have the same substrate specificity and ability to release N-terminal with this idea, although they involve transpeptidation reac- dipeptides from polypeptides [18]. Moreover, this enzyme has tions. also been shown to catalyze polymerization of dipeptidyl units Methionine oligomers were also hydrolyzed by a-chymo- [19]. Hydrolysis of (Met)$Lys, Ac"(Met)"3LysNH2and Met4 trypsin. As shown in Fig. 5 , chymotryptic digestion of could only be achieved using a high enzyme concentration (MetELys led to Met4 and (Met)"Lys and lower amounts of (enzyme:substrate = 1 : 10, w/w). Considerably larger Met3 and (MetELys. It should be pointed out that amounts of (MetELys and Ac"(Met)"LysNH2were obtained chymotrypsin cleaved the bond which was only poorly with the first and second peptide, respectively, as compared hydrolyzed by pepsin although the latter must be considered to dimethionyl units thus indicating that polymerization of as a fivefold more efficient enzyme for hydrolyzing the isopep- dimethionine took place readily. Hexamethionine and tide. Time course hydrolysis of (MetfiLys by a-chymotrypsin octamethionine could be identified as a result of the enzyme showed comparable release of the two major peptides, digestion of Met4. Cathepsin was definitely unable to (Met);Lys and (Met)"Lysand at significantlylower rate, equal hydrolyze the isopeptide bond. Finally, the rate of hydrolysis release of (MetELys, Met3 and Met4. The presence of Met4 of methionine oligomers grafted onto the &-aminogroup of without any free lysine suggested that transpeptidation reac- lysine by intestinal aminopeptidase N decreased sharply as tions occurred. Dimethionine was the minor component of the chain length increased. As indicated in Fig. 6, the release the reaction mixture. Higher enzyme concentrations of Ac"(Met)"LysNH, was faster from Ac"(Met)"zLysNHzthan (enzyme:substrate = 1 : 10, w/w) were necessary for hydroly- from Ac"(Met&LysNHz. The same was true with (MetfiLys sis of Ac" (Met);LysNH2 as well as methionine pentamer and as compared to (MetfiLys. Although the isopeptide bond has tetramer, in which transpeptidation reactions took place too. already been shown to be effectively hydrolyzable using Finally, shorter peptide substrates such as A C " ( M ~ ~ ) " ~ L ~ Saminopeptidase NH~ N [8], the amounts of lysine and the correand Met3 were not cleaved at all. It could therefore be con- sponding amide were smaller than those of the isopeptides cluded that chymotrypsin was able to hydrolyze the peptide (Table 3).

262 Table 3. Hydrolysis of methionine model peptides by aminopeptidase N The extent of hydrolysis was determined with a ratio of hog aminopeptidase: substrate = 1:500 (w/w), incubation at 37°C for 30 min. Acetylated lysine collected from the HPLC system was estimated by amino acid analysis after a 24-h hydrolysis in 5.6 M HCI at 110°C Peptide

Ac"(Met)"2LysNH2 Ac"(Met)"3LysNH2 (MetKLys (Met);Lys

Hydrolysis products Aca(Met)"LysNH2 or (Met)'Lys

Ac"LysNH2 or lysine

1550 1500 1050 400

240 140 60 40

Fig. I . A proposed minimal scheme to explain pepsin-catalyzed hydrolysis and transpeptidation reactions with (MetjSLys as substrate

highly polymerized methionine. It is noteworthy that hydrolysis of (MetELys resulted in very small amounts of free Met2 compared to the isopeptide (Met)E2Lys.Finally, hydrolysis of the synthesized isopeptides by intestinal aminopeptidase to DISCUSSION free methionine and (Met)"Lys was the most noteworthy result By using solid-phase procedures to synthesize methionine achieved. The moderate hydrolysis of the isopeptide bond in oligomers, whether they be free or covalently linked to the this work probably resulted from enzyme inhibition by the &-aminogroup of lysine, it was possible to investigate in detail high concentration of free methionine in the medium since it hydrolysis of these polymers by the digestive enzymes of the has already been shown that hydrophobic amino acids are alimentary tract. This was achieved by reverse-phase potent inhibitors of aminopeptidase [S, 261. Because of more chromatography using an HPLC system since quantitative favourable kinetic parameters, hydrolysis of methionyl-meestimation of methionine and methionine oligomers was thionine bonds will be less affected by free methionine than found to be possible in a wide concentration range. Although that of the isopeptide bond. Another point of interest concerns transpeptidation reacHPLC has already been widely used for separating peptides and related derivatives on a qualitative analytical scale [20tions which occurred during pepsin-catalyzed and chymo231, only a few studies have dealt with its use from a quantita- trypsin-catalyzed hydrolysis of oligo(methionine) chains in tive standpoint [24]. which the oligomers were either free or covalently linked As shown in the present work, gastric pepsin, pancreatic to the &-aminogroup of lysine. Analysis of peptic digestion chymotrypsin and carboxypeptidase A as well as intestinal products of (MetELys indicated the presence of low amounts aminopeptidase N, and probably also dipeptidylpeptidase IV, of (MetELys, (MetELys and Met, while methionine, quite effectively hydrolyze in vitro methionyl-methionine dimethionine and lysine were lacking. Fig. 7 shows the bonds in oligo(methionine) chains which have been covalently transfer of the N-terminal residue of methionine from one linked to an &-aminogroup of lysine through an isopeptide isopeptide molecule to another, which may account for the bond. However, although the synthesized methionine identified transpeptidation products. Quantitative determinaisopeptides and peptides were considered ideal for studying tion of (MetELys in the reaction mixture indicated that 10% the release of methionine in the form of free amino acid or of the initial substrate underwent transpeptidation reaction. as dipeptides and tripeptides, it must be stressed that with Moreover, the total amount of (MetELys and Met, correoligo(methiony1)-proteins somewhat different results might sponding to the secondary hydrolysis products of transient (Met);Lys was roughly equal to that of (Met)",Lys. A compabe expected. The peptic digestion studies led to the noteworthy finding rable transpeptidation scheme might also be proposed for that the longer the oligo(methionine) chain, the more effective chymotrypsin hydrolysis of (MetfiLys although it was not the release of covalently bound methionine in the form of easy to account for some of the secondary reaction products. trimethionine. A fivefold increase in the rate of Met, release The presence of Met5, and Met, in pepsin-catalyzed hydrolywas observed with (Met)",Lys as compared to (MetELys while sis of Met, and of Met4 and Met, in that of Met5, as well as Ac"(Met)",LysNH, was not cleaved at all. Chymotryptic that of Met2, MetJ and Met, in chymotrypsin digestion of digestion of (Met)",Lys, (MetELys and Aca(Met)",LysNH2 the same substrates would be difficult to explain if acyltransfer resulted in preferential hydrolysis of the methionyl-methio- reactions were not involved. A number of studies on hydrolysis of small peptide nine bond preceding the isopeptide bond. By contrast, Ac" (MetELysNHz was not cleaved. Of interest also was the fact substrates by porcine pepsin have pointed out the existence that (MetELys was hydrolyzed to Met, and (Met)"Lys and of transpeptidation reactions involving the transfer of the N-terminal residue of a peptide to a second substrate molecule to a lesser extent to Met, and (Met)',Lys. Using bovine spleen cathepsin C instead of intestinal via an acyl-enzyme intermediate [27,28]. In view of our results, dipeptidylpeptidase IV we showed that N-terminal methio- the ratio of transpeptidation to hydrolysis was surprisingly nine dipeptides could be released from the synthesized model high, though peptic digestion was carried out at pH 2.0 with a peptides and isopeptides, although high concentrations of the very low concentration of enzyme as compared to the starting enzyme had to be used, as already reported [25]. The well isopeptides (1 : 100, w/w). The latter ratio has often been close known ability of cathepsin C to catalyze the polymerization to one and the pH has been up to 3.5 in typical of dipeptidyt units resulting from the substrate hydrolysis, transpeptidations. Fornation of transpeptidatianproducts at gave rise to growing soluble polymers and then to insoluble pH 2.0 was found to occur only when non-substrate activator

263

peptides were used to speed up either acyl-transfer or aminotransfer reactions [29]. Our results strongly support the idea that hydrophobic residues must precede the sensitive bond in order for pepsin efficiently to catalyze hydrolysis and acyl-transfer-type transpeptidations. Since multiple binding of peptide substrates is known to occur in A and B secondary sites of pepsin (each site containing four subsites numbered S, to S4 and S1, to Sc, respectively) as suggested by Berger and Schechter [30], a number of peptides were found to increase the activity of the enzyme towards small synthetic peptide substrates and to promote transpeptidation reactions. Thus, the high affinity of pepsin site A for leucine and methionine peptides may lead to effective hydrolysis of (Met)E4Lys into (MetyLys and Met3. Because of hydrophobic characteristics, the latter would therefore play the part of an activator [29]. Chymotrypsin-catalyzed hydrolysis of peptide substrates in the form of (MetELys, which are known to involve an acylenzyme intermediate [31], also resulted from multiple interactions between the enzyme active site and the substrate. The striking preference of the enzyme for large hydrophobic amino acids 132, 331 certainly argues in favour of such a mechanism. Finally, it is important to emphasize that, unlike other synthetic oligo (amino acids) [34], enzymatic hydrolysis of oligo(methionine) has never been studied before. All the problems relating to oligo(methionine) insolubility were overcome by grafting up to 7-8 residues of the amino acid on the 8amino group of lysyl residues of food proteins 1121. Under these conditions, the polymers were still soluble in aqueous buffers and could therefore undergo hydrolysis by the enzymes of the digestive tract [35]. The results reported here showed that endopeptidases and exopeptidases were suitable for hydrolyzing all the synthesized model peptides. Hydrolysis rates by pepsin and chymotrypsin increased with the length of methionine oligomers. The resulting methionine oligomers and isopeptides (Met)"l- ,Lys were further cleaved by intestinal aminopeptidase. Both this enzyme and carboxypeptidase allowed a complete breakdown of methionine oligomers into the free amino acid. The data presented in this paper are also relevant to the study of the mechanism by which proteases act on small peptide substrates, which also involves hydrolysis and transpeptidation reactions. This work was supported in part by grant 820085 from A.E.C./ Rh8ne-Poulenc. We are greatly indebted to Dr J. Van Rietschoten (Faculte de Medecine, HBpital Nord, Marseille) for his valuable help in synthesizing the model isopeptides. Our thanks are also due to Andrea Guidoni for amino acid analyses, Jessica Blanc for the English and Brigitte Videau for typing the manuscript.

REFERENCES 1. Altschul, A. M. (1974) Nature (Lond.) 248, 643-646. 2. Bjarnason-Baumann, B., Pfaender, P. & Siebert, G. (1977) Nutr. Metab. 21, (Suppl. I ) , 170- 173.

3. Puigserver, A. J., Sen, L. C., Clifford, A. J., Feeney, R. E. & Whitaker, J. R. (1978) Adv. Exp. Med. Biol. 105, 587-612. 4. Yamashita, M., Arai, S., Imaizumi, Y., Amano, Y. & Fujimaki, M. (1979) J. Agric. Food. Chem. 27, 52-56. 5. Arai, S., Yamashita, M. & Fujimaki, M. (1979) Agric. Biol. Chem. 43,1069-1074. 6. Puigserver, A. J., Sen, L. C., Gonzales-Flores, E., Feeney, R. E. & Whitaker, J. R. (1979) J. Agric. Food. Chem. 27, 1098-1104. 7. Puigserver, A. J., Sen, L. C., Clifford, A. J., Feeney, R. E. & Whitaker, J. R. (1979) J. Agric. Food. Chem. 27, 1286-1293. 8. Gaertner, H., Puigserver, A. & Maroux, S. (1981) FEBS Lett. 133, 135-138. 9. Maroux, S., Louvard, D. & Baratti, J. (1973) Biochim. Biophys. Acta 321, 282 - 292. 10. Louvard, D., Maroux, S., Baratti, J., Desnuelle, P. & Mutaftschiev, S. (1973) Biochim. Biophys. Acta 291, 747 763. 11. Puigserver, A. J., Gaertner, H. F., Sen, L. C., Feeney, R. E. & Whitaker, J. R. (1982) Adv. Chem. Ser. 198, 149-167. 12. Gaertner, H. F. & Puigserver, A. J. (1984) J. Agric. Food. Chem., in the press. 13. Merrifield, R. B. (1969) Adv. Enzymol. 32, 221 -296. 14. Barany, G. & Merrifield, R. B. (1979) in The Peptides: Analysis, Synthesis and Biology (Gross, E. & Meienhofer, J., eds) vol. 2, pp. 1-284, Academic Press, New York. 15. Sarin, V. K., Kent, S. B. H., Tam, J. P. & Merrifield, R. B. (1981) Anal. Biochem. 117, 147- 157. 16. Gisin, B. F. (1973) Helv. Chim. Acta 56, 1476- 1482. 17. Robinson, A. 0.(1969) in Solid-Phase Peptide Synthesis (Stewart, J. M. & Young, J. D., eds) pp. 41 -44, Freeman Publ. Co., San Francisco. 18. McDonald, J. K., Callahan, P. X., Ellis, S. & Smith, R. E. (1971) in Tissue Proteinases (Barrett, A. J. & Dingle, J. T., eds) pp. 69- 107, North-Holland Publishing Co., Amsterdam and London. 19. Voynick, I. M. & Fruton, J. S. (1968) Biochemistry 7,40-44. 20. Fullmer, C. S. & Wassermann, R. H. (1979) J. Biol. Chem. 254, 7208 - 7212. 21. Hearn, M. T. W. (1980) J . Liquid. Chromatogr. 3, 1255-1276. 22. Meek, J. L. (1980) Proc. Natl Acad. Sci. USA 77, 1632- 1636. 23. Wilk, S. & Orlowski, M. (1982) J . Chromatogr. 249, 121129. 24 Tischio, J. P. & Hetyei, N. (1982) J. Chromatogr. 236, 237-243. 25. Bella, A. M., Erickson, R. H. & Kim, Y. S. (1982) Arch. Biochem. Biophys. 218, 156-162. 26. Feracci, H., Benajiba, A., Gorvel, J.-P., Doumeng, C. & Maroux, S. (1981) Biochim. Biophys. Acta 658,148- 157. 27. Takahashi, M., Wang, T. T. & Hofmann, T. (1974) Biochem. Biophys. Res. Commun. 57, 39-46. 28. Wang, T. T. & Hofmann, T. (1976) Biochem. J. 153,691 -699. 29. Wang, T. T. & Hofmann, T. (1976) Biochem. J. 153,701 -712. 30. Berger, A. & Schechter, I. (1970) Philos. Trans. R. SOC.Lond. Ser. B 257,249 - 264. 31. Fastrez, J. & Fersht, A. R. (1973) Biochemistry 12, 2025-2034. 32. Morihara, K. & Oka, T. (1977) Biochem. J. 163, 531 -542. 33. Bizzozero, S. A., Baumann, W. K. & Dutler, H. (1982) Eur. J . Biochem. 122,251 -258. 34. Sela, M. & Katchalski, E. (1959) Adv. Protein Chem. 14, 39441 1 ....

35. Gaertner, H. & Puigserver, A. (1982) Le Lait 62, 607-614.

H. Gaertner and A. Puigserver, Centre de Biochimie et de Biologie Moleculaire du Centre National de la Recherche Scientifique, 31 Chemin Joseph-Aiguier, F-I 3402 Marseille-Cedex-9, Bouches-du-Rh8ne, France