The availability of essentially pure preparations of pig creatine kinase isoenzymes I and III and creatine kinase IV reported in this thesis has enabled establishment of the molecular authenticity of creatine kinase IV as a unique isoenzyme species. A comparison of their properties is detailed.
Creatine kinase isoenzyme I was isolated and purified from chloroform-acetone powder extracts of pig brain. A procedure has been developed involving ammonium sulfate fractionation, DEAE-cellulose chromatography and chromatography on Sephadex G-200 which yields 80 ± 4 mg of essentially homogeneous protein from chloroform-acetone powders (144 g) prepared from one kg of fresh tissue. The procedure is highly reproducible and simpler than the previously reported procedures for creatine kinase I from other sources. The enzyme [specific activity = 4,435 ±20 (µkat/l.)/A280; A280/A260 = 1.83 ±0.02] was essentially homogeneous on polyacrylamide-gel electrophoresis at pH 8.8 and 5.7 and polyacrylamide-gel electrophoresis in the presence of 0.1% (w/v) SDS.
Extensive modifications to the original purification procedure for creatine kinase IV (Keto and Doherty, 1968; Keto, 1970) involving conditions of DEAE-cellulose chromatography and crystallisation, and the inclusion of gel filtration on Ultrogel AcA22 (Method A), have permitted significant improvements in the purity of the enzyme [specific activity = 3,100 ±50 (µkat/l.)/A280; A280/A260 = 2.05; A280/A410 = 188 compared with specific activity = 2,850 ± 50 (µkat/l.)/A280; A280/A260 = 1.76; A280/A410 = 100 (Keto, 1970)]. However, yields were low,being of the order of 1-2% of the original activity extractable with 0.1 M phosphate buffer (~8 mg protein/kg fresh tissue).
An alternative procedure was developed (Method B) which is inexpensive, rapid (5 days compared with three to four weeks for Method A), simple (eliminates lengthy chromatographic steps) and yields moderate quantities of highly pure protein [51 mg protein per kg of fresh tissue; specific activity = 3,350 ±30 (µkat/1.)/A280; A280/A260 = 2.00; A280/A410 = 440 ±60 (uncorrected for Rayleigh scattering)]. The procedure is identical to to that in Method A as far as the step involving DEAE-cellulose chromatography, but subsequently uses to advantage the unique solubility properties.of this enzyme. Enzyme is precipitated at low ionic strengths (µ = 0.01) at pH ~7 and differentially solubilized from contaminating proteins by repeated extractions at higher ionic strengths (µ~0.3, pH ~7). After crystallisation of enzyme from concentrated protein solutions, remaining protein contaminants are removed by precipitation at pH 5.0 (µ = 0.1). The enzyme [A280/A410 = 501 (uncorrected) or 1,363 (corrected for Rayleigh scattering)] was found by atomic absorption spectroscopy to contain negligible iron (~0.002 g atom of Fe per subunit) and copper (<0.0006 g atom of Cu per subunit) representing ≤0.1% contamination with haem protein (assuming a MW of 20,000 with 1 g atom of Fe per mole). Electrophoretic homogeneity was demonstrated on polyacrylamide-gel electrophoresis at pH 4.65 and at pH 8.8 in the presence of 0.1% sodium dodecylsulfate. Heterogeneity detectable on polyacrylamide-gel electrophoresis at pH 8.8 (6 to 7 hr at 3 to 5°) was most likely artefactual. Parallel Ferguson plots and the values for Kr (retardation coefficient) indicate that the enzyme probably exists as the monomer under the experimental conditions and heterogeneity arises from charge differences between the species detectable. [Significant irreversible loss of enzyme activity at pH 8.8 and 30° has previously been demonstrated (Keto, 1970).] Results of isoelectric focusing on polyacryl-amide-gels were dependent on the experimental conditions used. Provided gels were prefocused in the presence of thioglycolate to remove contaminating free-radicals, additives such as 6 M ethylene glycol or 22% (w/v) glycerol were used to prevent incipient and isoelectric precipitation of protein and loading of sample was conducted from the cathode end of gels, the enzyme focused essentially as a single band with a pIapp at 25° of 8.13 ± 0.01.
Studies on the purified isoenzymes have included determination of A1%1 cm at 280 nm (CPK I: 8.523; CPK IV: 9.733) and analysis of the amino acid composition of isoenzymes I and III, and of isoenzyme IV prepared by methods A and B. Values for v̅ calculated from the amino acid composition were 0.732 cm3 g-1 (isoenzyme I and IV) and 0.734 cm3 g-1 (isoenzyme III).
The subunit molecular weight of creatine kinase IV (44,980) calculated from the amino acid composition data using an error function (Hoy et al., 1974) was approximately half-way between the values obtained by polyacrylamide-gel electrophoresis in the presence of SDS (42,000) and by high-speed equilibrium, sedimentation in 6 M guanidinium chloride (49,500) or by titration with DTNB (52,300). The minimum molecular weight of isoenzyme I determined from amino acid composition (41,900) was fully consistent with that derived from polyacrylamide-gel electrophoresis in the presence of SDS (41,500) and with the equivalent weight obtained by titration with DTNB (41,890 ±90).
Low recoveries for isoenzyme IV on amino acid analysis (90 ± 1%) compared with ~105% for isoenzyme I and ~99% for isoenzyme III, together with the detection of significant amounts (8% by weight) of phospholipid (phosphatidyl glycerol) non-covalently associated with creatine kinase IV have enabled the reconciliation of the apparent discrepancies between estimates of minimum molecular weight by the various methods.
Creatine kinase IV was shown to be unique amongst currently studied creatine kinases from mammalian sources in possessing more than one thiol group per subunit reacting readily with DTNB under oxygen-free conditions at pH 7.0 (or pH 7.27) . One thiol per subunit is titrated virtually instantaneously with complete loss of activity and a further three thiols react more slowly at an identical rate (k ~ 1 M-1 s-1 at pH 7.0 and 25.0 ± 0.1°). The total number of thiols per subunit titratable with DTNB in the presence of 6 M guanidinium chloride (8.52 ± 0.1 per subunit; equivalent weight = 52,300 by DTNB titration in the absence of denaturant) was commensurate with the number of cysteic acid residues per subunit (9.0 ± 0.2 per 45,000-dalton subunit by amino acid analysis) estimated for oxidised samples of protein. At most this represents ~5.3% of the total enzyme oxidised under the denaturing conditions of 6 M guanidinium chloride (oxygen-free solutions containing 1 mM EDTA).
The biphasic kinetics of titration of isoenzyme I with DTNB (90% of the reaction completed virtually instantaneously; 10% proceeding to completion with k ~ 9 M-1 s-1 at pH 7.5 and 25°) probably indicated equilibrium proportions of monomer and dimer present under reaction conditions.
Current assumptions regarding the non-essentiality of thiol groups in catalysis by creatine kinase have been extensively reexamined in the light of the stoichiometry, specificity and kinetics of reaction of 2-chloromercuri- 4-nitrophenol with rabbit skeletal muscle creatine kinase III and its derivatives, as well as with pig heart creatine kinase IV. It was found that two moles of 2-chloromercuri-4-nitrophenol react per mole of native rabbit skeletal muscle creatine kinase III subunit with biphasic kinetics. Reaction of the first mole is virtually instantaneous (k of the order of 106 M-1 s-1). The second-order rate constant for incorporation of the second mole of 2-chloromercuri-4- nitrophenol was found to be 464 ± 16 M-1 s-1 at pH 8.01 and 25.0 ±0.1°. S-Carboxamidomethyl-creatine kinase III reacted with a single mole of 2-chloromercuri-4- nitrophenol per mole of modified enzyme subunit with a rate constant (k = 480 ± 10 M-1 s-1 at pH 8.01, 25.0 ± 0.1°) almost identical to that for the second slow reaction with the native enzyme (k = 464 ± 16 M-1 s-1 at pH 8.01 and 25.0 ± 0.1°). The results indicated that carboxamidomethylation of the "essential" thiol prevents the extremely fast reaction of 2-chloromercuri- 4-nitrophenol with native creatine kinase III. Evidence for the migration of a mercurinitrophenol substituent from the "essential" thiol of creatine kinase III (cysteine1) to an adjacent thiol (cysteine2) was obtained from an analysis of the kinetic order of the reaction of DTNB with CPK-(MNP)1, the mono-substituted mercurial derivative of creatine kinase III. (Parallel studies of the reaction of 2-chloromercuri-4-nitrophenol with pig heart creatine kinase IV are also reported but were limited by the large magnitude of the rate constants and complications arising from precipitation of the enzyme.) The results reported in these studies indicate the need for extreme caution in interpretation of residual activities of derivatives of creatine kinase involving covalent modification of thiol groups. Most importantly, the results indicate that current assumptions regarding the non-essentiality of active-site thiols to catalysis by creatine kinase are yet open to question.