Action on bovine αs1-casein of cardosins A and B, aspartic proteinases from the flowers of the cardoon Cynara cardunculus L.

BB

ELSEVIER

Biochimica et BiophysicaActa 1297 (1996) 83-89

Biochi~ic~a et BiophysicaA~ta

Action on bovine asi-casein of cardosins A and B, aspartic proteinases from the flowers of the cardoon Cynara cardunculus L. Miguel Ramalho-Santos, Paula Verfssimo, Carlos Faro *, Euclides Pires Departamento de BioqMmica, Faculdade de Ci~ncias e Tecnologia, Universidade de Coimbra, Apartado 3126, 3000 Coimbra, Portugal

Received 11 June 1996; accepted 21 June 1996

Abstract

The cleavage of purified bovine as~-casein separately by cardosin A and cardosin B, two distinct milk-clotting aspartic proteinases (APs) present in the stigmas of the plant Cynara cardunculus L., was studied. Casein digestion peptides were separated either by SDS-PAGE or by reverse-phase HPLC, and their N-terminal amino-acid sequences were subsequently determined by automated Edman degradation, thus identifying the cleavage sites. Results showed that both enzymes exert a similar but distinct action on bovine a~j-casein. In common they have the preference for the bond Phe23-Phe 24, and the cleavage of Trp164-Tyr165 and Phel53-Tyr 154. Cardosin A also cleaves the bond Tyr 165-Tyrlt'6, whereas Cardosin B cleaves an extra type of bond, Phe15°-Arg jSI, revealing a slightly broader specificity. A model for the action of both enzymes on bovine asl-casein is proposed and discussed. In comparison with the reported action of chymosin on bovine asi-casein, both cardosins proved to have a broader specificity towards this particular substrate due to a higher ability to cleave bonds between residues with large hydrophobic side-chains. Keywords: Cardosin; Aspartic proteinase; Proteolytic specificity; a~l-Casein; Chymosin;(Cynara cardunculus L.)

1. Introduction

Aspartic proteinases (APs) (EC 3.4.23) are a well-defined class of proteinases present in all major groups of the living world, from viruses to mammals [1-3]. Among the best characterized APs are mammalian chymosin, pepsin and renin, yeast proteinase A and the HIV-1 proteinase. Much less is known about: plant APs, but in the past few years knowledge has been building up rapidly in what concerns their purification, characterization, structure and function [4]. Certain APs, like chymosin, recombinant chymosin, pepsin and some fungal proteinases, possess milk-clotting activity, which accounts for their major importance in cheese technology [5-7]. We have purified and characterized two highly abundant APs present in stigmas of the Asteraceae plant Cynara cardunculus L. [8]. Aqueous extracts of these stigmas

Abbreviations: AP, aspartic proteinase; TFA, trifiuoroacetic acid; PVDF, polyvinylidenedifluoride. * Corresponding author. Fax: + 351 39 26798; e-mail: [email protected].

are traditionally used for the production of high quality sheep-milk cheese in Mediterranean countries, particularly in Portugal [9]. Milk-clotting activity is essentially characterized by the cleavage of the bond Phel°5-Met 1°6 in K-casein [10,11]. The milk-clotting activity identified in the stigmas of Cynara carduncuIus L. was initially thought to be due to a single AP [12-14], but recently we have shown that these flowers contain two APs, cardosin A and cardosin B, which are the products of two different, though related, genes [8,15]. Each cardosin consists of two subunits with apparent molecular masses of 31 and 15 kDa for cardosin A and 34 and 14 kDa for cardosin B. Both enzymes are synthesised in a precursor form and are believed to undergo proteolytic processing with the removal of the N-terminal prosegment and of an internal sequence located between the two subunits, known as the plant-specific insert [8]. This sequence, of about 100 amino acids, is present only in plant APs [4] and bears no sequence similarity with APs of mammalian or microbial origins. Cardosin A and cardosin B have acid pH optima and are specifically inhibited by pepstatin, indicating that they belong to the class of the aspartic proteinases [16], as do the other known milk-clotting enzymes used in cheese-

0167-4838/96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved. PII S01 67-4838(96)00103-3

84

M. Ramalho-Santos et al. / Biochimica et Biophysica Acta 1297 (1996) 83-89

making [17]. Specificity and kinetics studies were carried out using oxidised insulin B chain [12] and a chromophoric peptide [18] respectively, and the results suggested that in these aspects cardosin A appears to be similar to chymosin, whereas cardosin B closely resembles pepsin [19,20]. Another group has reported the purification of three milk clotting enzymes from dried flowers of Cynara cardunculus L. [21,22]. These enzymes were named 'cynarases' and subsequently 'cyprosins' and were assumed to have been derived from a common precursor by different processing [23]. Cyprosins have been shown to differ significantly from any of the cardosins [8]. The objective of this work was to study the action of the isolated cardosins on bovine asl-casein, in this way assessing the contribution of each enzyme for the overall proteolytic specificity of the cardoon rennet [14], and to compare the results with those previously reported for chymosin [24]. The results showed that cardosin B has a slightly broader specificity and an evident higher rate of proteolysis than cardosin A. In comparison with chymosin, both cardosins showed a broader specificity towards this particular substrate due to a higher ability to cleave bonds between residues with large hydrophobic side-chains. This is in agreement with the few reported results on the specificity of other plant APs [25].

2. Materials and methods

2.1. Materials Fresh flowers of Cynara cardunculus L. were collected from plants grown from seeds supplied by the Botanical Gardens of the University of Coimbra. Bovine as-casein (asl-casein with a small proportion of a~z-casein) and molecular mass standards for SDS-PAGE were obtained from Sigma (St. Louis, MO). Reagents used for sequencing were from Applied Biosystems (Warrington, U.K.). All other reagents were of analytical grade. 2.2. Preparation of asl-casein The method used to prepare asl-casein was based on the one described by Rasmussen et al. [26]. Commercial ces-casein (10 mg) was dissolved in 5 ml of 50 mM ammonium acetate, 8 M urea (pH 5.0) and purified by FPLC using a Mono S HR 5 / 5 column (Pharmacia, Uppsala, Sweden) equilibrated in the aforementioned solution. Elution was carried out with a linear gradient of 0 to 1 M ammonium acetate, 8 M urea (pH 5.0) at a flow rate of 0.5 ml/min. Each peak of absorbance (at 280 nm) was collected as a fraction and its purity assayed by SDS-PAGE. The as~-casein fraction was dialysed overnight against 100 mM sodium dihydrogen phosphate buffer (pH 6.2 or 6.8) at 4°C.

2.3. Preparation of cardosins Stigmas (1 g) from fresh flowers of Cynara cardunculus L. were ground in a mortar and pestle under liquid nitrogen. The ground tissue was then homogenized in 5 ml 0.1 M citric acid (pH 3.0) and centrifuged at 12000 X g for 10 min. The superuatant (4 ml) was applied to a Sephadex G-100 column (2.5 X 75 cm) equilibrated and eluted with 25 mM Tris-C1, pH 7.6 (buffer A) at a flow rate of 1.0 ml/min. Each peak of absorbance was collected as a fraction and assayed for activity by the casein hydrolysis method [27]. The active fraction was purified by FPLC using a Mono Q HR 5 / 5 column (Pharmacia, Uppsala, Sweden) also equilibrated in buffer A. The proteins were eluted with a linear gradient of NaC1 (0-0.5 M) in buffer A at a flow rate of 0.75 m l / m i n and the protein peaks were collected and assayed for activity. 2.4. Protein determination Protein concentration was determined by the method of Katzenellenbogen and Dobryszycka [28] using bovine serum albumin as standard. 2.5. Digestion of C~sl-casein by cardosins asl-Casein digestion was carried out at 30°C in 100 mM sodium dihydrogen phosphate buffer (pH 6.2 or 6.8). Cardosin A or B was added to the casein solution (enzyme/substrate mass ratio of 1,/500 or 1/1000) and the reaction allowed to proceed. At selected times (up to 2 h) aliquots were taken and the reaction stopped either with equal volume of SDS-PAGE denaturing solution, for electroblotting, or with 0.6% v / v TFA (final concentration), for HPLC. 2.6. SDS-PAGE and electroblotting of the casein peptides SDS-PAGE was performed in a Bio-Rad Mini Protean II electrophoresis apparatus according to the method of Laemmli [29]. After SDS-PAGE the peptides were transferred onto a PVDF membrane by electroblotting in 10 mM 3-(cyclohexylamino)-l-propanesulfonic acid (CAPS), 10% v / v methanol (pH 11) at 500 mA for 1 h. 2.7. Reverse-phase HPLC a~l-Casein digestion reactions were centrifuged for 2 min at 12 000 X g and the supernatant chromatographed by reverse-phase HPLC using a Techogel 15-300 C 8 0.46 x 25 cm column (HPLC Technology, UK). Elution was carried out with a linear gradient of acetonitrile in 0.1% v / v TFA; after a 5 min isocratic run a gradient of 0 - 8 0 % v / v acetonitrile was developed in 30 min. The elution rate was 1.0 ml/min. Blank assays were performed before and after

M. Ramalho-Santos et al. / Biochimica et Biophysica Acta 1297 (1996) 83-89

1

each assay. Peptides were detected at 215 nm and concentrated in a Speed-vacuum concentrator.

N-terminal amino-acid sequences of the protein bands from electroblotting and of the peptides isolated by reverse-phase HPLC were determined by Edman degradation using an Applied Biosystems 473-A sequencer.

The action of the two APs from cardoon on purified bovine Otsl-Casein was studied. A pH of 6.2 and a temperature of 30°C were the conditions used by Carles and Ribadeau-Dumas [24] on 1Eheir work with chymosin, and so these conditions were also used in the present work for further comparison. Enzyme/substrate mass ratios of 1:500 and 1:1000 and an additional pH value of 6.8 were investigated, but in all cases no significant differences were observed either on SDS-PAGE gels or on HPLC chromatograms of the digestion fragments (data not shown). When the enzyme/substrate mass ratio of 1:500 was used digestion occurred faster with both enzymes but no new peptides appeared. Fig. 1 depicts a gel with the time-course of digestion for both cardosin A (lanes 3 to 5, corresponding to 10, 60 and 120 min after enzyme addition, respectively) and cardosin B (lanes 6 to 8, same as cardosin A). Digestion by

34

5

6

78

66 45 36 29 24

2.8. Sequence determination of peptides

3. Results

2

85

20.1 14.2

Fig. 1. SDS-PAGE time-course of the digestion of otsl-casein by cardosin A and cardosin B. Purified bovine asl-casein was digested for up to 2 h with cardosin A or cardosin B at 30°C in 100 m M sodium dihydrogen phosphate buffer (pH 6.8). At selected times aliquots were taken, the reaction stopped with equal volume of SDS-PAGE denaturing solution and the samples loaded on a gel. The gel were stained for proteins using Coomasie brilliant blue. Lane 1 represents the molecular weight markers, indicated in kilodaltons. Lane 2 represents intact purified bovine aslcasein. Lanes 3 - 5 represent the time-course of digestion by cardosin A. Lanes 6 - 8 represent the time-course of digestion by cardosin B. Lanes 3 and 6 : 1 0 min; lanes 4 and 7 : 6 0 min; lanes 5 and 8 : 1 2 0 rain.

cardosin A yielded two major bands with apparent molecular masses of about 30 and 25 kDa. Digestion by cardosin B yielded three major bands with apparent molecular masses of about 30, 25 and 24 kDa. Cardosin A digested cesl-casein slower than cardosin B, which is in agreement with previously reported kinetic studies [18]. This difference is clearly shown by comparing lanes corresponding to the same digestion time, in particular the 10 min stage,

Table 1 Sites of cleavage by cardosin A and cardosin B on bovine o
Cleavage site

Fragment(s)

F-V-A- p- F- PF - V- A - P - F - P - E - V- F -

Phe23-Phe 24 Phe 23_Phe 24

Phe24-Trp 199 Phe 24_Trp 164 Phe24-Tyr 16s Tyr154-Trp164 Tyr J54-Tyr 165 Tyr 166-Trp 199 Tyr 165-Trp 199 Tyr 154_Trp199

Cardosin A SDS-PAGE bands

30 kDa 25 kDa

Reverse-phase

A

y - Q- r.- D-

Phe153_Tyr J54

HPLC peaks

B C D

Y - V - P - L - G- T - Q- YY- Y - V - P - L - Gy - Q- r.- D-

Tyr 165-Tyr 166 Trp164-Tyr t 65 Phe~53-Tyr 154

30 Id)a 25 1,Da 24 lcDa

F - V- A - P - F - P - E - V 'Y- V- A - P - F - P - E F - V- A - P - F -

Phe23 _Phe24 Phe23 _Phe24 Phe23_Phe24

Cardosin B SDS-PAGE bands

Reverse-phase HPLC peaks

E F G

R - Q- F ~ Y- Q- L - D - A - Y - P - S - G - A - N y - y - V - p - 1,- G- T -

a

Phe 150_Arg J51 Phet53_Tyr 154 Trp164-Tyr t 65

Phe24_Trp 199 Phe24_Trp 164 Phe24_Phe J50 Phe 24-Phe 153 Arg 15t -Phe 153 Tyr154_Trp164 Tyr165-Trp 199

Purified bovine otsl-casein was digested for up to 2 h with cardosin A or cardosin B at 30°C in 100 mM sodium dihydrogen phosphate buffer (pH 6.2 or 6.8). The casein peptides were separated by SDS-PAGE followed by electroblotting onto a PVDF membrane or by reverse-phase HPLC on a C 8 column. Their N-terminal amino-acid sequences were determined by automated Edman degradation, thus identifying the cleavage sites. The size of the corresponding fragments was determined by comparing the retention times of the various peaks and their respective N-terminal sequence. a The complete amino-acid sequence of the these fragments was determined by automated N-terminal Edman degradation.

86

M. Ramalho-Santos et al. / Biochimica et Biophysica Acta 1297 (1996) 83-89

when digestion by cardosin A (lane 3) presented only one fade 30 kDa band below intact % : c a s e i n , whereas digestion by cardosin B (lane 6) already led to the appearance

0.006

-75

0.004

50

0.002

25

II

tO

20

75

0.006

/. ./.:'"/

0.004-

O so ,~ .¢ .25

0.002-

III

lo

2o

0.006

75

/,./ // C

0.004 0.002

50 -25

lO

20

t (rain) B

175

t

0.006 0.004 2

5

50

0.002

II

1o

~ 7 5

0.006~m~ 0.004

50

0.002

25

III

io

20

0.006 -50

0.004 0.002

¸¸¸El 1o 20 t (min)

-25

of the three bands, in addition to intact as:casein. Despite this difference, the migration pattern of the major digestion peptides was roughly similar for both enzymes, and this could be an indication that the main cleavage sites of cardosin A and cardosin B did not differ substantially, and even that some of them were identical. The N-terminal amino-acid sequence analysis of electroblotted protein bands showed that for both enzymes all of the bands presented unequivocally a sequence corresponding to the cleavage of the bond Phe23-Phe 24 (see Table 1), which has been reported to be the bond in % : c a s e i n most susceptible to cleavage by chymosin [24,30,31]. Only the band with an apparent molecular mass of 30 kDa displays the expected mobility of the fragment Phe24-Trp 199, or Olsl-I, according to the nomenclature of Fox [32]. These results revealed the limitations of this first approach, as digestion peptides with apparent molecular masses of 25 or 24 kDa presented the same N-terminal sequence of the one with 30 kDa, these peptides no doubt being the product of subsequent proteolytic cleavage on the C-terminal region of % : c a s e i n . The small digestion peptides resulting from this cleavage were probably lost or not detected in SDS-PAGE, and so another approach for separation and N-terminal sequence determination was required. In this different approach the casein peptides were separated by reverse-phase HPLC on a Cs column. Fig. 2 shows the chromatograms obtained for cardosin A (A, I to III, corresponding to 10, 60 and 120 rain after enzyme addition, respectively) and for cardosin B (B, same as for cardosin A). After 2 h of digestion, when the peaks are more evident, cardosin A yielded four major peaks (A, B, C and D, chromatogram A.III), while cardosin B yielded three major peaks (E, F and G, chromatogram B.III), around the same values of the gradient (40-60% acetonitrile), but with quite different patterns. Again it is evident that the rate of proteolysis is higher for cardosin B, particularly when the 10 min stages are compared (chromatograms A.I and B.I), which is in agreement with the SDS-PAGE results shown before. The HPLC fractions corresponding to the peaks were concentrated and subsequently analysed by automated N-terminal amino-acid se-

.< Fig. 2. Reverse-phase HPLC chromatograms of the time course of the digestion of c~q-casein by cardosin A (A) and B (B). Purified bovine as:Casein was digested for up to 2 h with the enzyme at 30°C in 100 mM sodium dihydrogenphosphate buffer (pH 6.2). At selected times aliquots were taken, the reaction stopped with 0.6% v/v TFA (final concentration), the sample centrifuged for 2 min at 12000× g and the supernatant chromatographed by reverse-phase HPLC using a C 8 column. Elution was carried out with a linear gradient of acetonitrile in 0.1% v/v TFA; after a 5 min isocratic run a gradient of 0-80% v/v acetonitrile was developed in 30 min (dashed lines). The elution rate was 1.0 ml/min. Peptides were detected at 215 nm. Chromatograms I: 10 rain digestion; chromatograms II: 60 min digestion; chromatograms HI: 120 min digestion.

M. Ramalho-Santos et al. / Biochimica et Biophysica Acta 1297 (1996) 83-89

quence determination. In what concerns cardosin A, the corresponding cleavage sites identified were Phe153-Ty r ~54 (peak A), Tyr165-Tyr 166 (peak B), Trp164-Tyr165 (peak C) and again PheJ53-Tyr 154 (peak D). Although presenting the same N-terminal amino-acid sequence, peaks A and D correspond to two different casein fragments (see Table 1). In what concerns cardosin B, the cleavage sites identified were PheJS°-Arg jSl (peak E), Phel53-Tyr 154 (peak F) and Yrp164-Tyr 165 (peak G). Having determined the complete sequence of peaks E and F and by comparing the retention times of the various peaks and their corresponding N-terminal sequences, taking in account the primary structure of a~-casein, it was possible to determine the respective fragments, including for the bands in the SDS-polyacrylamide gel. Results obtained by the two different approaches are complementary and in agreement between themselves. These results are summarized in Table 1. The band with apparent molecular mass of about 24 kDa, corresponding to two of cardosin B's digestion peptides, Phe24-Phe ~5° and Phe 24Phe 153, was only faintly resolved in cardosin A's digestion runs because cardosin A does not cleave the bond Phe ~5°Arg jSI and may cleave the bond Phe~53-Tyr ~54 to a less extent than cardosin B.

4. Discussion

In the present work we have studied the action of cardosins A and B, APs from fresh stigmas of Cynara cardunculus L., on bovine c~l-casein. This substrate was chosen for two types of reasons. First, because it is a protein rich in hydrophobic residues (43% of the residues), allowing in this way a broad study of the specificity of APs, known for their preference for bonds between residues with hydrophobic side-chains [33]. There is also much useful information available about the primary [34,35], secondary and tertiary structure of bovine c~-casein [36]. Second, because it seemed appropriate to study the specificity of two milk-clotting APs towards one of the four major caseins. These proteins interact in a complex way to form the casein micelles [37], which undergo cleavage and destabilization during milk-clotting and cheese-making [38]. Action of milk-clotting APs on K-casein has been widely studied [11,39], including in our laboratory [13]. This interest is due to the initiating role that the cleavage of the bond Phel°5-Met ~°6 --- or, in one case, SerJ°4-Phe ~°5 - - plays in the destabilization of the casein micelles [10]. All these studies showed that this was the only bond cleaved in K-casein at this stage, c~-casein is also known to undergo cleavage during later stages of cheese-making [38], and the extent to which it is cleaved is an important factor for the yield and the textural and organoleptic properties of the cheese. In addition, the use of casein molecules as substrates might reveal itself fruitful, since these proteins are biosyn-

87

thesised specifically to be digested and serve as the major source of biologically active peptides to the young mammal [40]. Casein-derived peptides have been isolated that p o s s e s s opioid, immunostimulating, anti-thrombotic, antibacterial or anti-hypertensive activity, among others [4042]. Proteinases that cleave specific sequences in the casein molecules may be of interest in the identification and isolation of biologically active peptides carried within the protein, or in the modulation of their activity. Results showed that like the majority of the APs [33], both cardosins preferably cleave bonds between residues with hydrophobic side-chains. They particularly seem to be able to accommodate in their active site very large hydrophobic sequences, as is the case for Phe23-Phe 24, Phea53-Tyr 154, TrpJ64-Tyr165 and Tyr165-Tyr 166. This last bond is only cleaved by cardosin A, that does not seem to have a particular preference for any of the two bonds in the Trp164-Tyr166 area. Conversely, cardosin B cleaves only and very efficiently Trp164-Tyr165. Contrarily to the majority of the APs, we have previously detected exopeptidase activity in the cardosins [12]. However, such is not the case in this study, since the susceptible bond Tyr 165Tyr j66 in the Tyrl65-Trpl99 fragment was not cleaved by any of the enzymes. Cardosin B has a slightly broader specificity, cleaving a bond between a hydrophobic residue and a basic one, PhelS°-Arg 151. The cleavage of a bond of this kind had not been reported for the cardosins in previous studies [12]. Cardosin B also revealed a higher rate of proteolysis than cardosin A, and this is in agreement with previous kinetic studies [8,18]. Other bonds between residues with bulky hydrophobic side-chains are not cleaved by any of the cardosins probably because they are either not exposed (e.g., Ile71-Va172) [36] or located in an acidic region (e.g., Tyr91-Leu 92, Tyr94-Leu95), leading to electrostatic repulsion between the substrate and the active site of the enzymes, negatively charged [8]. These results suggest that cardoon rennets with a higher proportion of cardosin B relatively to cardosin A will display a higher proteolytic activity, not so much because of a broader specificity of cardosin B but rather due to its faster action. Taking in account the results in Table 1 and the evolution of the chromatographic profiles it is possible to conceive a model for the action of the cardosins on bovine Ce~l-casein (Fig. 3). The most susceptible bond is unequivocally Phe23-Phe 24. Next, cardosins exert cleavage in the Trp164-Tyr166 area, located in a region of bulky hydrophobic residues close to the C-terminal. Cardosin A cleaves both bonds, Trp164-Tyr165 and Tyrl65-Tyr 166, although either one or the other, whereas cardosin B cleaves only Trp164_Tyr165. They then proceed towards the N-terminal, first with the cleavage of Phel53-Tyr ~54 and then Phe tS°Arg jS~. Cardosin A exerts a slower cleavage on the region TrpJ64-Tyr166, leading to the appearance of the peptide Tyr154-Trp 199, not present in the digestion with cardosin B. The last bond, PheJS°-Arg ~51, is only cleaved to a small extent by cardosin B, due to its slightly broader specificity.

88

M. Ramalho-Santos et al. / Biochimica et Biophysica Acta 1297 (1996) 83-89 RI I

~

RI F23 ["-"7

F24 I

o.st-casein

W199 I

F24 I

~

~ W164/Y165 Y165/Y166 I I

Stage I

W199 I Stage n

W199 I Stage Ill

of the cardosins towards this particular substrate and in these particular conditions, in relation to chymosin. Nevertheless, and perhaps more important, these results are also an indication of a higher ability of the cardosins to cleave bonds between residues with large hydrophobic side-chains, as suggested by previous studies [12]. This seems to be consistent with the very few data available regarding the specificity of other plant APs [25].

Acknowledgements

F24

I

FIS0

I

RI51 Y154

fl

Stage V

Fig. 3. Proposed model for the action of the cardosins on bovine Ots~-casein. The enzymes exert cleavage on the casein molecule in a sequential manner probably due to conformational alterations that are a consequence of each particular cleavage. After cleaving the most susceptible bond, Phe23-Phe 24, the cardosins proceed from an area located in a region of bulky hydrophobic residues near the C-terminal, Trp 164-TyrJ66, towards the N-terminal. The extra peptide shown in Stage IV, marked ( * ), only occurs with cardosin A due to its slower cleavage of the region Trp164-Tyr166, in Stage III. Stage V only occurs with cardosin B, because of its slightly broader specificity.

This sequential order of cleavage occurs probably because the cleavage of a susceptible bond results in conformational changes in the casein molecule that improve the accessibility of the enzymes to another susceptible bond. Although much work has been done with other milkclotting enzymes ([6], and references therein), few data is available on their action on isolated caseins as the substrates. The pattern of hydrolysis of proteins in milk by the cyprosins has been studied [23]. In that work, the caseins from which a few of the digestion peptides originated were allegedly identified solely by the mobility of the peptides on IEF. Para-K-casein, the peptide released after the cleavage of the bond Phel°5-Met~°6 on K-casein, is assumed to be formed, but still the origin of the majority of the peptides was admittedly unknown [23]. In contrast with the approach presented here, the authors did not determine the nature of the digestion peptides, nor did they identify the specific sites of cleavage. In the conditions we used in this work chymosin only cleaved the bond Phe23-Phe24 [24]. This bond, or sometimes Phe24-Va125, appears to be the one in asl-casein most susceptible to cleavage by milkclotting APs [30,31]. Apart from this major cleavage, any of the cardosins cleaved three other bonds, located in the C-terminal of the casein molecule. However, other secondary sites of cleavage by chymosin have also been identified in other works [30,31,43], where different conditions were used. Depending on the pH and the salt concentration, chymosin cleaved more bonds in C~sl-Casein than just Phe23-Phe24, many of them located in the C-terminal. Taken together these results indicate a broader specificity

This work was mainly supported by the PAMAF (project 95.09.6547.5), administered by IFADAP, Portugal. Miguel Ramalho-Santos is the recipient of a fellowship from the PRAXIS XXI program (JNICT).

References [1] Tang, J. and Wong, R.N.S. (1987) J. Cell. Biochem. 33, 53-63. [2] Davies, D.R. (1990) Annu. Rev. Biophys. Chem. 19, 189-215. [3] Takahashi, K. (ed.) (1995) Aspartic Proteinases: Structure, Function, Biology and Biomedical Implications, pp. 1-550, Plenum Press~ New York. [4] Kervinen, J., Trrm~ikangas, K., Runeberg-Roos, P., Guruprasad, K., Blundell, T. and Teeri, T.H. (1995) in Aspartic Proteinases: Structure, Function, Biology and Biomedical Implications (Takahashi, K., ed.), pp. 241-254, Plenum Press, New York. [5] Foltmann, B. (1987) in Cheese: Chemistry, Physics and Microbiology (Fox, P.F., ed.), Vol. 1, pp. 33-61, Elsevier Applied Science, London. [6] Fox, P.F. and Stepaniak, L. (1993) Int. Dairy J. 3, 509-530. [7] Fox, P.F. (1993) J. Food Biochem. 17, 173-199. [8] Verlssimo, P., Faro, C., Molt, A.J.G., Lin, Y., Tang, J. and Pires, E. (1996) Eur. J. Biochem. 235, 762-768. [9] Sfi, F.V. and Barbosa, M. (1972) J. Dairy Res. 39, 335-343. [10] Delfour, A., Jollbs, J., Allais, C. and Joll~s, P. (1965) Biochem. Biophys. Res. Commun. 19, 452-455. [l 1] Drohse, H.B. and Foltmann, B. (1989) Biochim. Biophys. Acta 995, 221-224. [12] Faro, C.J., Moir, A.J.G. and Pires, E.V. (1992) Biotechnol. Lett. 14, 841-846. [13] Macedo, I.Q., Faro, C.J. and Pires, E.V. (1993) J. Agric. Food Chem. 41, 1537-1540. [14] Macedo, I.Q., Faro, C.J. and Pires, E.V. (1996) J. Agric. Food Chem. 44, 42-47. [15] Faro, C.J., Ver(ssimo, P., Lin, Y., Tang, J. and Pires, E.V. (1995) in Aspartic Proteinases: Structure, Function, Biology and Biomedical Implications (Takahashi, K., ed.), pp. 373-377, Plenum Press, New York. [16] Marciniszyn, J., Hartsuck, J.A. and Tang, J. (1977) in Acid Proteases: Structure, Function and Biology (Tang, J., ed.), pp. 199-210, Plenum Press, New York. [17] Green, M.L. (1977) J. Dairy Res. 44, 159-188. [18] Ver[ssimo, P., Esteves, C., Faro, C.J. and Pires, E.V. (1995) Biotechnol. Lett. 17, 621-626. [19] Martin, P., Trieu-Crot, P., Collin, J.C. and Ribadeau-Dumas, B. (1982) Eur. J. Biochem. 122, 31-39. [20] Martin, P. (1984) Biochim. Biophys. Acta 791, 28-36. [21] Heimgartner, U., Pietrzak, M., Geertsen, R., Brodelius, P., da Silva Figueiredo, A.C. and Pais, M.S.S. (1990)Phytochemistry 29, 14051410.

M. Ramalho-Santos et al. / Biochimica et Biophysica Acta 1297 (1996) 83-89

[22] Cordeiro, M., Jakob, E., Puhan, Z., Pais, M.S. and Brodelius, P.E. (1992) Milchwissenschaft 47, 683-387. [23] Cordeiro, M.C., Pais, M.S. and Brodelius, P.E. (1994) Physiol. Plant. 92, 645-653. [24] Caries, C. and Ribadeau-Duraas, B. (1985) FEBS Lett. 185, 282-286. [25] Kervinen, J., Sarkkinen, P., Kalkkinen, N., Mikola, L. and Saarma, M. (1993) Phytochemistry 32, 799-803. [26] Rasmussen, L.K., Hcjrup, P. and Petersen, T.E. (1992) Eur. J. Biochem. 207, 215-222. [27] Kunitz, M. (1947) J. Gen. Physiol. 30, 81-87. [28] Katzenellenbogen, W.M. and Dobryszycka, W.M. (1959) Clin. Chim. Acta 4, 515-522. [29] Laemmli, U.K. (1970) Natu~:e 227, 680-685. [30] P611isier, J.P., Mercier, J.C. and Ribadeau-Dumas, B. (1974) Ann. Biol. Anim. Biochem. Biophys. 14, 343-362. [31] McSweeney, P.L.H., Olson, N.F., Fox, P.F. and Healy, A. (1993) J. Dairy Res. 60, 401-412. [32] Fox, P.F. (1969) J. Dairy Sci. 52, 1214-1218. [33] Kay, J. and Dunn, B.M. (1992) Scand. J. Clin. Lab. Invest. 52 (Suppl. 210), 23-30.

89

[34] Mercier, J.C., Grosclaude, F. and Ribadeau-Dumas, B. (1971) Eur. J. Biochem. 23, 41-51. [35] Stewart, A.F., Willis, I.M. and McKinley, A.J. (1984) Nucleic Acid Res. 12, 3895-3907. [36] Kumosinski, T.F., Brown, E.M. and Farrell, H.M., Jr. (1991) J. Dairy Sci. 74, 2889-2895. [37] Schmidt, D.G. (1982) in Developments in Dairy Chemistry (Fox, P.F., ed.), Vol. 1, pp. 61-86, Elsevier Applied Science, London. [38] Dalgleish, D.G. (1982) in Developments in Dairy Chemistry (Fox, P.F., ed.), Vol. 1, pp. 157-187, Elsevier Applied Science, London. [39] Caries, C. and Martin, P. (1985) Arch. Biochem. Biophys~ 242, 411-416. [40] Meisel, H., Frister, H. and Schlimme E. (1989) Z. Ern~ihrungswissenchaft 28, 267-278. [41] Loukas, S., Varoucha, D., Zioudrou, C., Streaty, R.A. and Klee, W.A. (1983) Biochemistry 22, 4567-4573. [42] Fiat, A.M., Migliore-Samour, D., Joll~s, P., Dronet, L., Sollier, C.B.D. and Caen, J. (1993) J. Dairy Sci. 76, 301-310. [43] Mulvihill, D.M. and Fox, P.F. (1977) J. Dairy Res. 44, 533-540.