Dissimilarity in the folding of human cytosolic creatine kinase isoenzymes

Abstract

Creatine kinase (CK, EC 2.7.3.2) plays a key role in the energy homeostasis of excitable cells. The cytosolic human CK isoenzymes exist as homodimers (HMCK and HBCK) or a heterodimer (MBCK) formed by the muscle CK subunit (M) and/or brain CK subunit (B) with highly conserved three-dimensional structures composed of a small N-terminal domain (NTD) and a large C-terminal domain (CTD). The isoforms of CK provide a novel system to investigate the sequence/structural determinants of multimeric/multidomain protein folding. In this research, the role of NTD and CTD as well as the domain interactions in CK folding was investigated by comparing the equilibrium and kinetic folding parameters of HMCK, HBCK, MBCK and two domain-swapped chimeric forms (BnMc and MnBc). Spectroscopic results indicated that the five proteins had distinct structural features depending on the domain organizations. MBCK BnMc had the smallest CD signals and the lowest stability against guanidine chloride-induced denaturation. During the biphasic kinetic refolding, three proteins (HMCK, BnMc and MnBc), which contained either the NTD or CTD of the M subunit and similar microenvironments of the Trp fluorophores, refolded about 10-fold faster than HBCK for both the fast and slow phase. The fast folding of these three proteins led to an accumulation of the aggregation-prone intermediate and slowed down the reactivation rate thereby during the kinetic refolding. Our results suggested that the intra- and inter-subunit domain interactions modified the behavior of kinetic refolding. The alternation of domain interactions based on isoenzymes also provides a valuable strategy to improve the properties of multidomain enzymes in biotechnology.

Figure 1. Domain and subunit organizations of the homodimeric CKs (HMCK and HBCK), the heterodimeric form (MBCK) and the two domain-swapped chimeric forms (BnMc and MnBc).

Figure 2. Biophysical characterization of the proteins.

(A) Far-UV CD. (B) Intrinsic Trp fluorescence. (C) Extrinsic ANS fluorescence. (D) SEC analysis. The proteins were dissolved in 20 mM Tris-HCl buffer, pH 8.0. The protein concentration was 1 mg/ml for SEC analysis, and 0.2 mg/ml for spectroscopic experiments. The far-UV CD and Trp fluorescence spectra were obtained by subtracting the spectra of the buffer.

Figure 3. Inactivation of CKs by GdnHCl.

(A) Concentration-dependence of HBCK inactivation by GdnHCl. The other four enzymes had similar inactivation curves independent on enzyme concentration (data not shown). (B) Inactivation of the five enzymes by GdnHCl.

Figure 4. Equilibrium folding of CKs.

(A) Mean residue molar ellipticity at 222 nm during unfolding. (B) Parameter A (I ₃₆₅/I ₃₂₀) of intrinsic fluorescence during unfolding. The data were fitted by the three-state model shown in Scheme 1, and the thermodynamic parameters were listed in Table 2. The fitted curves are presented as solid lines. (C) ANS fluorescence during unfolding. (D) ANS fluorescence during refolding. For clarity, the transition curves of refolding monitored by CD and Trp fluorescence are not shown. Protein unfolding were carried out by incubating 0.2 mg/ml proteins in 20 mM Tris-HCl buffer (pH 8.0) containing various concentrations of GdnHCl at 25°C overnight. The fully-unfolded proteins were prepared by denaturing 24 mg/ml proteins in 6 M GdnHCl for 1 h, and the refolding was conducted by a manual dilution (1∶120) of the denatured proteins in 20 mM Tris-HCl buffer (pH 8.0) containing various concentration of GdnHCl and incubated at 25°C overnight.

Figure 5. Reactivation, refolding and aggregation kinetics of CKs.

(A) Refolding kinetics of CKs monitored by the intrinsic fluorescence emission at 350 nm (excited at 295 nm) versus time. (B) Reactivation kinetics monitored by the recovered activity of the samples quenched at suitable time intervals. The activity data were normalized by taking the activity of the native enzymes as 100%. (C) The time-course aggregation during refolding was monitored by the turbidity at 400 nm, and the data were recorded every 2 s. The refolding was initiated by manually diluting the fully denatured proteins in the refolding buffer with a mixing ratio of 1∶200 for refolding and reactivation and 1∶120 for aggregation. The reactivation and aggregation data were fitted by a single exponential kinetics, and the refolding data were fitted by the biphasic kinetics. The fitted curves are presented as solid lines.

Figure 6. Correlation of the extent of aggregation (A ₄₀₀) with the refolding and reactivation rate constants.

(A) A ₄₀₀ versus the reactivation rate constant. (B) A ₄₀₀ versus the fast-phase rate constant of refolding.