Application of the reaction of dithioesters with ε-amino groups in lysine to the chemical modification of proteins

Biochimica et Biophysica Acta, 957 (1988) 254-257

254

Elsevier BBA 33258

Application of the r e a c t i o n of dithioesters with c-amino groups in lysine to t h e chemical modification of proteins J 6 r 6 m e S o u p p e a, M a r t i n e U r r u t i g o i t y a a n d G u y L e v e s q u e b a Groupement de Recherches de Lacq, Artix and b Laboratoire de Chimie des Comp~$s Thioorganiques (associ~ au CNR$), ISMRA, Universit$ de Caen, Caen (France)

(Received 24 May 1988)

Key words: Dithioester; Lysine; Horseradish peroxidase; Papain

The reaction of lysine with dithieesters was applied to horseradish peroxidase donor:hydrogen-peroxide oxidoreductase, EC 1.11.1.7) using carboxymethyl dithiotridecanoate: three to four lysine residues were modified. The modified enzyme was soluble and active in diethyl ether. Papain (EC 3.4.22.2) was modified with carboxymethyi dithiobenzoate: two lysine residues were modified. The modified enzyme was soluble and active in dimethylsulfoxide. From these results it is concluded that dithioesters are efficient reagents for the modification of peripheral iysine residues of proteins. Aromatic dithioesters, les~ reactive but more selective, should be recommended for thiol-dependent enzymes such as papain.

Introduction The chemical modification of proteins has been extensively studied [1] either to change their physical properties or to determine the essential amino acids present in the active sites of enzymes [2]. The modification often consist of making amino acid residues react with specific reagents. The case of the e-NH 2 group of lysine has been widely fashioned, particularly with reagents such as anhydrides [3], aldehydes [4], TNBS [5] or other activated electrophiles [6]. Kjaer [7] demonstrated that carboxymethyl dithiocarboxylates are efficient thioacylating reagents of the e-amino group of lysine. In this paper we describe the use of dithioesters as re-

Al~breviations; DMSO, dimethylsulfoxide; TNBS, trinitrobenzenesulfonic acid; RZ, reinheitszahl. Correspondence: J. Souppe, Groupement de Recherches de Lacq, B.P. M, 64170 Arfix, France.

actants for the modification of the e-NH2 groups of lysine in horseradish peroxidase and papain as well as the properties of the newly modified enzymes. Experimental procedure Materials

Horseradish peraxidase (Sigma Type II) and papain (Sigma Type IV) were used without further purification. N-Benzoyl-D,L-arginine p-nitroanilide, L-cysteine and EDTA were purchased from Merck. 9-Methoxyellipticine was a gift of the Sanofi Co. Methods

Dithioesters were synthesized according to procedures described earlier [8], substituting tetrahydrofuran to ether as solvent for alipathic compounds. Enzyme activities were measured on a Hewlett Packard spectrophototometer HP 8450 A with the adapted temperature controller, 89100A. The per-

0167-4838/88/$03.50 © 1988 Elsevier Science Pubfishers B.V. (Biomedical Division)

255

oxidase activity was followed at 444 nm (e = 30 m M - 1 . cm-1) in 0.1 M citrate buffer (pH 5.0) at 30 ° C with o-dianisidine and H202 at concentrations of 0.13 and 2.0 mM, respectively. The horseradish peroxidase concentrations were determined at 402 nm, using a millimolar absorption coefficient of 102 mM - 1 . cm -1 [9]. The amidase activity of papain was measured in 50 mM Tris buffer (pH 7.5) at 50 °C with 1.0 mM N-benzoylD,L-arginine p-nitroanilide in the presence of Lcysteine and EDTA at a concentration of 0.60 m g / m l for both. The appearance of p-nitroaniline was followed at 410 nm (e ffi 8.8 m M - 1. c m - 1) over a 15 min period. The papain concentration was determined at 278 nm using a millimolar absorption coefficient of 56 raM- ]. cm- 1 [10]. Determination of the total -NH 2 groups in native and modified enzymes was performed with the TNBS method [5]. The concentration in thioamide was determined from the increase in absorbance at 280 nm using a millimolar absorption coefficient of 12.8 m M - ] . cm-!. The modification of horseradish peroxidase with carboxymethyl dith/otridecanoate was performed at room temperature in 3.0 ml 0.1 M phosphate buffer (pH 8.5) containing 6 mg horseradish peroxidase (0.15 /tmol) and 1.6 mg dithioester (5 pmol previously dissolved in 0.1 ml ethanol). After 18 h the reaction mixture was dialysed at 4 ° C against water, and then lyophilized. The yield in hemoprotein was 95~. Modification of papain with carboxymethyl dithiobenzoate was performed at 40 °C in 40 ml 0.1 M phosphate buffer (pH 8.5) containing 320 mg papain (8.6 #mol) and 100 mg dithioester (470 ~mol previously dissolved in 3 ml ethanol). After 2 h the reaction mixture was dialysed at 4 ° C against water and finally lyophilized. 260 mg were recovered corresponding to 100 mg protein (4.3 /~mol, i.e., 50~ yield). The activity was 11 mU per mg lyophilisate, where,~s that of the native enzyme was 28 mU per mg lyophilisate. The oxidation of 9-methoxyellipticine to 9oxoellipticine by the modified horseradish peroxidase was performed at 20 ° C in a quartz cuvette filled with 2.0 ml diethyl ether containing 0.25 mM 9-methoxyellipticine, 0.25 mg/ml lyophilisate of modified horseradish peroxidase. The test was initiated with 50/tl 0.2 M aqueous H202 and the

appearance of 9-oxoellipticine was followed at 515 nm (E = 9.2 raM- 1. cm- 1) [11]. The amidase activity of papain in DMSO was performed at 50 °C in a quartz cuvette filled with 1.5 ml DMSO containing 0.6 mg/ml L-cysteine, 0.6 mg/ml EDTA, 1.0 mM N-benzoyl-D,L-arginine p-nitroanilide and 15/~1 water. The reaction was initiated with the modified papain (0.3 ml of a 10 mg/ml solution of the lyophilisate in DMSO). Results

The reaction of free lysine with various dithioesters was shown to be nearly quantitative under mild experimental conditions: H3 N+ OCH-CH 2CH 2CH 2CH 2- N H 2 + R - C - SX -

OOC

ISI

H3N + /~C H - C H 2CH2CH2CH 2 - N H - C - R + HSX

ooc

-

II s

This result was applied to the modification of lysines born by enzymes. Horseradish peroxidase was chosen as an example of oxidoreductase. The enzyme is known to bear six accessible lysines [12]. The reaction of the native protein with carboxymethyl dithiotridecanoate at pH 8.5 lead to the modification of three to four lysines as determined by the TNBS method. The UV spectrum of the modified enzyme is shown on Fig. 1. Comparison with that of the native enzyme shows that the RZ value (Reinheitszahl) has decreased: this is due to the formation of thioamide bonds which absorb in the range 270-290 nm. (RZ = A4oz/A2so).

t

05C 0.4C~

1

03o ¢1

0.20

i

< 010 O0

200 2~0

360

350 do

450 ~o

Wavelength(nm)

1

550 600

Fig. 1. UV spectrum of horseradish peroxidase modified by carboxymethyl dithiotridecanoate.

256 TABLE I COMPARISON O F T H E PROPERTIES OF NATIVE A N D MODIFIED H O R S E R A D I S H PEROXIDASE (HRP) Native H R P Specific activity ( U / m g ) a 880 Optimal temperature ( ° C) b 34 Optimal pH c 5.1 K H2°2 (raM) d 1.30 Ko-dianisidinc (mM) d 0.13 m

Modified H R P 850 36 4.9 1.25 0.18

a The concentration in protein was determined by the method of Lowry et al. [20]. b Over a 1 rain period needed for the measurement of enzyme activities. c In 0.1 M acetate buffer in the pH range 3.0-6.0. d 30 o C, 0.1 M acetate buffer ( p R 5.0).

As for the peroxidative activity of the modified enzyme, kinetical data are given in Table I. The yield in specific activity in 96%. No important changes are observed regarding the K m values of H202 and o-dianisidine. Optimum pH and temperature do not greatly differ from that of the native enzyme. The most striking differences are in the solubilities in organic solvents. Whereas the native enzyme is soluble only in water, the modified sample is soluble at more than 0.5 jaM in acetone, diethyl ether, chloroform, ethanol, toluene and of course water. It remains quite insoluble in dioxane and tributyl phosphate. Horseradish peroxidase (HRP) is known to catalyse the oxidation of 9-methoxyellipticine to 9-oxoellipticine in the presence of H202 [11]. We could perform this reaction in d~thyl ether. As described in the experimental procedure the reaction medium is saturated with water (1.2~, v/v, at 20°C [13]). Under these conditions the modified enzyme respects a Michaelis behaviour: varying the concentration of H202 leads to: K~2Oz ffi 2.0 mM Vm -- 1.25 p mol- m i n - ~. (nmol HRP) - t (0.96 in water [11 ])

Papain is our second example of modification. The enzyme is known to bear 10 lysines in its primary structure [14]. The reaction of the native enzyme was carried out with various dithioesters: only aromatic dithioesters lead to active samples

of enzyme after modification. With carboxymethyl dithiobenzoate, two lysines out of 10 have reacted. The yield in protein is about 50~ whereas the specific activity was 64% that of the native enzyme. Like horseradish peroxidase, the modified papain showed interesting solubility properties in organic solvents: more than 2.5 mg protein per ml could be dissolved in benzene. Moreover, the native enzyme is completely inactive in DMSO, whereas the modified one has an activity of 3.1 mU per mg protein. (The specific activity of native papain in water was nevertheless 32 mU per mg protein.) The modification realized with pure papain (Sigma Type IV) could be repeated with the commercially available crude powder (Sigma Type II) dialysed before use. The yield in activity was 68~ and 27~ amino groups present in the dialysed preparation had reacted. As for solubilities and activities in organic solvents, the results were the same as those described for pure papain. Discussion

From what is known about the catalytically important amino acids of horseradish peroxidase [15], one can assume that no lysine plays any essential role in the mechanism of H202 activation [16]. Our results completely agree with this assumption, since no significant changes were observed in the values of K m, optimal pH and optimal temperature. The observed solubilities in organic solvents confirm that presumably peripheral lysine residues have been modified. The ability of horseradish peroxidase to catalyse the oxidation of 9-methoxyellipticine has been described by Meunier [11] and the reaction mechanism has been completely studied. The very low K m value of 9-methoxyellipticine in water (3.5 /~M) enables the reaction to proceed in spite of its very low solubility in water. The solubility of the modified protein in diethyl ether is of great interest, since 9-methoxyellipticine is also freely soluble in this solvent. As for the kinetical results, our Vm in diethyl ether is not very different from that found by Meunier in water. All these results argue for a modification of the peripheral lysines of horseradish peroxidase without an important change in the active site of the enzyme.

257

Papain is known as a thiol-proteinase. Poor yields in enzymatic activities when using aliphatic carboxymethyldithioesters indicate that the essential cysteine has been blocked according to the following reaction described earlier by Leon for non-proteic materials [17]: E-SH + R-C-SCH2COOH ~ E - S - C - R + HSCH 2COOH II II S S

This problem does not exist with horseradish peroxidase which has no cysteine in its primary structure [12]. The assay with horseradish peroxidase using an aliphatic dithioester leads to the modifio cation of a greater number of lysines than that with papain using an aromatic dithioester, though the temperature was higher in the latter example. As expected, aromatic dithioesters are less reactive than aliphatic ones. This is the reason why the above reaction probably proceeds only partially with carboxymethyl dithiobenzoate, leading to an acceptable yield in enzymatic activity. The solubility of the new enzyme in benzene indicates that, like for horseradish peroxidase, peripheral lysines were mofidied. The activity in DMSO with a minimum amount of water is of great interest for peptide synthesis even though biphasic or multisolvent systems have also been proposed [18-19]. The results given in this work show that dilhioesters are interesting mild reagents for the modification of lysines born by enzymes. Aliphatic dithioesters are good electrophiles and may be quite efficient for proteins which have no essential cysteines. Aromatic dithioesters are more selective so that lysine residues of thiol-dependent enzymes may also be modified under suitable experimental conditions. We are currently investigating the use of various functionafised dithioesters in order to vary the

isoelectric point of proteins and also to fashion aimed modifications at the artive sites of other enzymes.

References 1 Lundblad, R.L. and Noyes, C.M. (1984) in Chemical Reagents for Protein Modifications, Vol. I and II. CRC Press, Boca Raton, FL. 2 Kaiser, E.T., Lawrence, D. and Rokita, S.E. (1985) Annuo Rev. Biochem. 54, 565-595. 3 Klotz, I.M. (1967) Methods Enzym. 11, 576-580. 4 Geoghegan, K.F., Ybarra, D.M. and Feeney, R.E. (1979) Biochemistry 18, 5392-5396. 5 Drozdovskaya, N.R. (1982) FEBS Lett. 150, 385-389. 6 Kubota, I. and Tsugita, A. (1980) Eur. J. Biochem., 106, 263-273. 7 Kjaer, A. (1952) Acta Chem. Scand., 6, 327-332. 8 Kurzer, F. and Lawson, A. (1953) in Organic Synthesis Coll. Vol. V, pp. 1046-1049, John Wiley & Sons, New York. 9 Ohlsson, P.I. and Paul, K.G. (1976) Acta Chem. Scand. Ser. B Org. Chem. 30, 373-375. 10 Glazer, A.N. and Smith, E.L. (1961) J. Biol. Chem. 236, 2948-2954. 11 Meunier, G. and Meunier, B. (1985) J. Biol. Chem. 260, 10576-10582. 12 Ugarova, N.N. and Rozhkova, G.D. (1978) Biokhimiya 43, 793-797. 13 Merck Index, 10th Edn. (1983) p. 3751. 14 Nolan, C. and Smith, E.L. (1960) Proc. Soc. Exptl. Biol. Med. 105, 287-290. 15 Thanabai, V., De Ropp, J.S. and La Mar, G.N. (1987) J. Am. Chem. Soo. 109, 7516-7525. 16 Rozkova, U. (1978) Biokhimiya 43, 1242-1250. 17 Leon, N.H. and Asquith, R.S. (1970) Tetrahedron 26, 1719-1725. 18 Barbas, C.F. and Wong, C.-H. (1987) J.C.S. Chem. Commun. 533-534. 19 Cantacuzene, D. and Guerreiero, L. (1987) Tetr. Lett. 5153-5156. 20 Lowry, D.H., Rosebrough, N.A., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-270.