Subunit level crosslinking of rabbit muscle pyruvate kinase by o-phthaldialdehyde

ARCHIVES

OF BIOCHEMISTRY

AND

BIOPHYSICS

Vol. 279, No. 1, May 15, pp. 32-36, 1990

Subunit Level Crosslinking of Rabbit Muscle Pyruvate Kinase by o-Phthaldialdehyde Selma Yilmaz and Inci 6zerl Department

of Biochemistry,

School of Pharmacy,

Hacettepe University,

Ankara,

Turkey

Received September 19,1989, and in revised form December 26,1989

o-Phthaldialdehyde caused irreversible inhibition of rabbit muscle pyruvate kinase following preliminary formation of an enzyme-reagent complex. At pH 7.5, 35”C, the dissociation constant for the complex and the maximal pseudo-first-order rate constant for covalent modification were 0.32 & 0.08 mM and 2.54 ? 0.23 mini’, respectively. The inactivation was accompanied by uv-spectral changes pointing to isoindole formation, with a limiting stoichiometry of 1 isoindole linkage per enzyme subunit. Phosphoenolpyruvate, ADP, and ATP effectively protected the enzyme against inactivation, suggesting that the active site is the target of o-phthaldialdehyde action. As native and modified enzymes were indistinguishable with respect to mobility of the major band in sodium dodecyl sulfate-polyacrylamide gel electrophoresis, it was concluded that the crosslinkage was intrasubunit in character, and that the amino acid residues involved must be closely positioned in the polypeptide backbone. Lysine 366, previously shown to be selectively reactive toward 2’,3’-dialdehyde ADP (Bezares et al., 1987, Arch. Biochem. Biophys. 253, 133-137), and cysteine 325 or 357 are implicated. fc 1990

Academic

Press,

Inc.

Pyruvate kinase (EC 2.7.1.40) has four tissue-specific isozymes in mammals, distinguishable by their catalytic, regulatory, and immunological properties (l-4). Studies with rat cDNA and genomic clones indicate that the hepatic (L) and the erythrocyte (R) isozymes are produced from a single gene by use of different promoters (5). ACcordingly, these isozymes show distinct primary structural elements only in their N-terminal regions (5, 6). Sequencing of cDNA clones for the muscle (M,) and kidney (MJ type pyruvate kinases has shown these two isozymes to differ in an internal amino acid sequence localized in the domain of intersubunit contact, and analysis i To whom correspondence

should he addressed.

of genomic clones has led to the conclusion that Mi and M, pyruvate kinases are produced from the same gene by differential RNA splicing (7). An intergenic comparison, using pyruvate kinases Mi and L as representatives of their class, reveals nonidentity at approximately 30% of all amino acid positions, mostly accountable for by single base substitutions in the DNA sequence. In addition, the L isozyme has an N-terminal extension containing the residue phosphorylated by CAMP-dependent protein kinase (s-12). The active site sequences in all isozymes, on the other hand, have been strongly conserved (5, 7, 13), barring two substitutions (Leu 122 (M,) + Val (L); Lys 223 (M,) + Gln (L)). Primary structural domains defining the active site contour of cat muscle M, pyruvate kinase have been deduced from crystallographic evidence (13,14); studies (mainly on the M, isozyme) using affinity labels and less selective reagents have identified critical lysine (15-17), cysteine (18, 19), histidine (20, 21), and arginine (22,23) residues at or in the vicinity of the active site. A peripheral tyrosine residue at position 147 in the M, sequence also appears to be essential for activity (16, 24). The following study was prompted by the recent use of o-phthaldialdehyde to incorporate nearby lysyl and cysteinyl residues in various enzymes into an isoindole ring system (25-27). The potential of this reagent to introduce internal crosslinks at the active site of pyruvate kinase was tested to provide a comparative probe into the active site topologies of the various isozymes in solution and into conformational changes accompanying substrate and/or effector binding. The results obtained with rabbit muscle pyruvate kinase indicate that this isozyme is irreversibly inhibited by o-phthaldialdehyde with spectroscopic evidence for isoindole formation. The covalent modification, occurring at subunit level, involves at least one amino acid residue at the active site. MATERIALS

AND

METHODS

Rabbit muscle pyruvate kinase (crystalline suspension in ammonium sulfate, sp act > 200 U/mg) was obtained from Boehringer-

32 All

Copyright Cl 1990 rights of reproduction

by Academic Press, Inc. in any form reserved.

CROSSLINKING

OF RABBIT

MUSCLE

PYRUVATE

33

KINASE

1.0,

w” 0.5 w

0.1 0

r,

2 t,

of rabbit FIG. 1. Inactivation 0, Control; 0, 0.05 mM o-PDA.

I 6

b

min.

muscle pyruvate

kinase by o-PDA.

25

50

o-P&, Mannheim and dialyzed against, 50 mM Te.s’/O.2 mM EDTA, pH 7.5, prior to use. All other reagents were obtained from Sigma. o-PDA stocks were prepared daily in methanol. Inactivation was carried out at 35°C in 100 mM Tes, pH 7.5, containing 10 mM MgCl,, 100 mM KCl, and ca. 1 unit of enzyme/ml. The reaction was initiated by the addition of o-PDA. (The final concentration of methanol was 2.5”C.) Periodically, aliquots were quenched by dilution and residual activity was assayed at PEP = ADP = 2 mM, by the method of Kimberg and Yielding (28). Isoindole formation was monitored at 337 nm (Z-27) in an inactivation mixture containing 0.1-0.2 mM o-PDA and 0.8-2 mg protein/ml. Enzyme concentration was estimated from E’!‘lf, = 0.54 and M,, 237,000 per tetramer (29); the concentration of LX,1 “1” isoindole formed was calculated using cnlii = 7.66 mM ’ cm-’ (25). SDS-PAGE in 10% gels was performed according to Laemmli (30) after 24-h dialysis of samples against 2000 vol of 50 mM Tes/O.2 mM EDTA, pH 7.5. Protein concentration in dialyzed samples was determined by the method of Bradford (31). using bovine serum albumin as standard. RESULTS

The effect of o-phthaldialdehyde on pyruvate activity. Treatment of rabbit muscle pyruvate

kinase

kinase with o-PDA resulted in a first-order loss in activity (Fig. 1). The inactivation was irreversible by dilution (and by removal of the reagent by extensive dialysis). Quenching of the reaction by 1:l dilution with 40 mM cysteine also failed to restore activity. The dependence of the firstorder rate constants for inactivation on o-PDA concentration (Fig. 2, line a) indicated that the covalent change was preceded by the formation of an enzyme-reagent (E +R) complex (Scheme 1). KI and k were estimated to be 0.32 & 0.08 mM and 2.54 t 0.23 min-‘, respectively. Protection by substrates and effecters. Inclusion of Mg.ADP, PEP, and Mg.ATP in the inactivation medium provided significant protection against inactivation (Table I). The combined effect of PEP and ’ Abbreviations used: o-PDA, o-phthaldialdehyde; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; Tes, N-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid,

KIM-’

FIG. 2. Dependence of inactivation rate on reagent concentration. Line a, inactivation in the absence of substrate (E, 1 U/ml); line b, inactivation in the presence of PEP = ADP = 2 mM (E, 0.1-0.3 IJ/ ml). Pseudo-first-order rate constants for inactivation due to o-PDA (k’) were calculated by taking the difference between the observed rate constant at each reagent concentration and the rate constant for spontaneous inactivation (Fig. 1, line a).

Mg . ADP was studied by adding o-PDA to the assay system and following the time course of product formation. At PEP = Mg. ADP = 2 mM, in the absence of o-PDA, the enzymatic rate (determined by quantitating pyruvate as its 2,4-dinitrophenylhydrazone (28)) was constant up to 50% conversion to products. (Aan > 1.0, Fig. 3, curve a). When o-PDA was included, AszO showed a nonlinear variation with time (Fig. 3, b and c). The nonlinear progress curve was analyzed according to Eq. [ 11, relating to enzyme inactivation in the presence of saturating levels of substrate (32), substituting AszO for the process product (P) and assuming a pseudo-first-order with k’ = (k. o-PDA)/(K, + o-PDA). P,

- P, = P,7. emk”

ill

A representative semilogarithmic plot obtained with 0.1 mM o-PDA is given in Fig. 4, yielding k’ = 0.13 min- ‘. Data obtained at different reagent concentrations (with PEP and Mg eADP fixed at 2 mM) have been incorporated into Fig. 2 (line b). It may be seen that o-PDA and substrates effectively compete for the same site, as lines a and b intersect on the ordinate.

E + R 2 E . R & E’ (inactive) SCHEME

1.

34

YILMAZ

AND iizE~

TABLE I Protection

of Rabbit Muscle Pyruvate Kinase against Inactivation by o-Phthaldialdehyde

Substance added

Relative rate of inactivation”

None MgADP b

1.00 0.70 0.43

2mM 6mM

PEP 0.53 0.25

‘i’mM

6mM MgATP b

0.65 0.39 1.00 0.16

2mM

6mM Fructose 1,6-bisphosphate, 2 mM PEP, MgADP, 2 mM

I

’ At 0.05 mM o-PDA. * The inactivation mixture was supplemented with equimolar amounts of MgCl, and the nucleotide.

Chemical nature and stoichiometry of the reaction between pyruvate kinase and o-phthaldialdehyde. The modification ofpyruvate kinase with o-PDA was accompanied by an increase in the absorbance at 337 nm and the appearance of spectral properties (Fig. 5) suggestive of isoindole formation. The stoichiometry of the reaction was found to depend on the relative concentrations of reagent and enzyme: at 0.1 mM o-PDA and 34 PM subunit, the stoichiometry was 0.34 isoindoles formed/sub-

unit, and the enzyme retained its activity. A fivefold ratio (0.2 mM o-PDA;

a significant

fraction

J

0

of

Semilogarithmic plot for inactivation of pyruvate kinase by o-PDA added to the assay system. o-PDA, 0.1 mM (See text).

FIG. 4.

chiometry to 0.84 + 0.08 isoindoles per subunit, yielding totally inactivated enzyme. The limited modification of the enzyme at low o-PDA concentration was ascribed to reagent depletion by side reactions at peripheral sites. Electrophoretic behavior of the modified enzyme. SDS-PAGE patterns obtained with native, partially modified and extensively modified enzyme are shown in Fig. 6. The electrophoretic mobility of the major protein band (M, 58,000-60,000, depending on the protein load per lane) was found to remain unchanged, regardless of

A

increase in the reagent to enzyme 15 ELMsubunit) increased the stoi0.12 A

A

337 006

520 10

I

b

II I

250

25

FIG. 3.

l ,0.05

Progress curve for product formation. 0, No o-PDA added; and QO.10 mM o-PDA.

mM;

10

5

0

t ,min

20

10 t , min.

300

350

15 min.

400 nm

Spectral changes accompanying modification of pyruvate kinase by o-PDA. E, 2 mg/ml; o-PDA, 0.1 mM. (A) Time-course for the increase in absorbance at 337 nm. (B) Ultraviolet spectrum of enzyme containing 0.34 isoindole per subunit.

FIG.

5.

CROSSLINKING

a

b

OF RABBIT

C

FIG. 6. SDSPAGE of native and modified pyruvate kinase. (a) Enzyme modified to a final stoichiometry of 0.34 isoindole/subunit (3 pg protein), (b) totally inactivated enzyme (8 wg protein), (c) unmodified enzyme (3 pg protein).

the degree of modification. Thus the crosslinking effected by o-PDA must be confined to the isolated subunit. Minor bands at M, 77,000 and 119,000 most likely correspond to subunits modified at multiple sites and to dimeric species, respectively. DISCUSSION o-Phthaldialdehyde was found to cause irreversible inhibition of rabbit muscle pyruvate kinase. Accompanying spectral changes, taken to arise from isoindole formation, pointed to a stoichiometry of one isoindole ring per subunit. The modification occurred at the level of the individual subunit, as evidenced by the electrophoretic patterns, which showed an insignificant population of higher molecular weight bands, even in samples derived from inactivation mixtures containing a high reagent: protein ratio (Fig. 6, lane b). Although several routes for the degradation of isoindoles have been proposed (33), it is unlikely that the major protein band at 58-60 kDa is

MUSCLE

PYRUVATE

KINASE

35

an electrophoretic artifact generated by cleavage of intersubunit crosslinks (with the accessory bands in Fig. 6, lane b being remnants reflecting the original modified protein population), since higher molecular weight bands are totally absent in the sample modified with a limiting amount of o-PDA (Fig. 6, lane a), even at high protein load. The competitive protection afforded by substrates suggests that the crosslinking effected by o-PDA occurs at the active site. Previous studies with various enzymes have implicated lysine and cysteine residues as reactive centers (25-27). The active site contour defined by Xray crystallography (13) reveals multiple possibilities as to the residues which may actually be involved in the crosslinking of pyruvate kinase: Lys 114, 223, 229, 246, 269, 366; Cys 325, 357. Covalent inhibition by trinitrobenzene sulfonate (15) and 2’,3’-dialdehyde ADP (17) has shown Lys 366 to be selectively modified. It is likely that the reaction with o-PDA also involves this residue (and Cys 325 or 357). Crosslinking with more distant residues would be expected to affect electrophoretic mobility in the presence of SDS, enough to cause a splitting or widening of the major protein band in mixtures of native and o-PDA-modified enzyme, which was not observed. In conclusion, o-PDA appears to be an effective crosslinking agent modifying the active site of rabbit muscle pyruvate kinase. Extension of the present study to other pyruvate kinases should yield comparative information regarding the structural basis for differential isozyme function. REFERENCES I. Tanaka, T., Harano, Y., Sue, F., and Morimura, H. (1967) J. Riothem. (Tokyo) 62,71-91. 2. Imamura, K., and Tanaka, T. (1972) J. Riochem. (Tokyo) 71, 1043~1051. 3. Cardenas, J. M., Strandholm, J. J., and Miller,
36

YILMAZ

14. Stuart,

D. I., Levine,

M., Muirhead,

H., and Stammers,

D. K.

15. .Johnson, S. C., Bailey, T., Becker, R. R., and Cardenas, (1979) Biochem. Riophys. Rex Commun. 90,525-530.

J. M.

(1979) J. Mol. Biol. 134,109-142.

16. DeCamp, D. L., and Colman, 4449-4503.

R. F. (1986) J. Biol. Chem. 261,

17. Bezares, G., Eyzaguirre, J., Hinrichs, M. V., Heinrikson, R. L., Reardon, I., Kemp, R. G., Latshaw, S. P., and Bazaes, S. (1987) Arch. Biochem. Biophys. 253,133-137. 18. Bloxham, D. P., Coghlin, Biophys. Acta 525,61-73.

S. J., and Sharma, R. P. (1978) Biochim.

19. Annamalai, A. E., Tomich, J. M., Mas, M. T., and Colman, R. F. (1982) Arch. Biochem. Biophys. 219,45%57. 20. Bornmann, L., and Hess, B. (1974) Hoppe-Seyler’s Chem. 355,1073-1076. 21. Dann, L. G., andBritton, 22. Cardemil,

H. G. (1974) Biochem. J. 137,405-407.

E., and Eyzaguirre,

192,533-538.

Z. Physiol.

J. (1979) Arch. B&hem.

Biophys.

AND

ijZER

23. Kilinc, K., and&er, N. (1984) Arch. Biochem. Biophys. 230,321326. 24. DeCamp, D. L., and Colman, R. F. (1989) J. Biol. Chem. 264, 8430-8441. 25. Puri, R. N., Bhatnagar, D., and Roskoski, R., Jr. (1985) Riochemistry 24,6499-6508. 26. Puri, R. N., and Roskoski, R., Jr. (1988) Biochem. Biophys. Res. Commun. 150,1088-1095. 27. Puri, R. N., Bhatnagar, D., and Roskoski, R., Jr. (1988) Biochim. Biophys. Acta 957,34-46. 28. Kimberg, D. V., and Yielding, L. (1962) J. Biol. Chem. 237,32333239. 29. Cottam, G. L., Hollenberg, P., and Coon, M. J. (1969) J. Biol. Chem. 244,1481-1486. 30. Laemmli, U. K. (1970) Nature (London) 227,680-685. 31. Bradford, M. M. (1976) Anal. Biochem. 72,248-254. 32. Tian, W. X., and Tsou, C. L. (1982) Biochemistry 21, 1028p1032. 33. Alvarez-Coque, M. C. G., Hernandez, M. J. M., Camanas, R. M. V., and Fernandez, C. M. (1989) Anal. Biochem. 178, l-7.