Mechanism of Protein Stabilization by Disulfide Bridges: Calorimetric Unfolding Studies on Disulfide-deficient Mutants of the α-Amylase Inhibitor Tendamistat

Abstract The present differential scanning calorimetry and circular dichroism studies on the mechanism of protein stabilization by disulfide bonds were concerned with two questions: is the increase in unfolding entropy upon removal of disulfide links sufficient for the explanation of the general stability decrease of disulfide-deficient mutants? Is it immaterial by which residue cysteine residues are replaced when disulfide bridges are to be opened? To answer these questions we investigated two disulfide bridge mutants of the α-amylase inhibitor Tendamistat where the large loop (C45A/C73A) or the small loop (C11A/C27A) had been opened by recombinant DNA techniques, and we compared the stability of the mutated proteins with that of wild-type Tendamistat published previously. To elucidate the significance of the nature of the group that replaces Cys we introduced in position 27 of the small loop four different amino acids instead of Cys: Ala, Leu, Ser and Thr. Surprisingly, opening of the small loop (17 residues) causes larger destabilization than opening of the large loop comprising 29 residues. The thermodynamic parameters at pH 7.0 are: wild-type: t 1/2 =81.6°C, Δ H cal =296 kJ mol -1 , large loop mutant (C45A/C73A): t 1/2 =58.6°C, Δ H cal =225 kJ mol −1 and small loop mutant (C11A/C27A): t 1/2 =42.7°C, Δ H cal =135 kJ mol −1 . This finding is at variance with the entropy hypothesis. The relative contributions to stability of enthalpic and entropic terms can be varied by a proper choice of substitutions. While the destabilization originating from C45A/C73A exchanges in the large loop turns out to be purely entropic, the stability decreases of the small loop mutants are caused by changes in both enthalpic and entropic terms. Leu or Ser in position 27 leads to an overall enthalpic destabilization. Thr in position 27 increases the transition enthalpy of this mutant to the value of the wild-type protein but increases at the same time the value of the transition entropy with the result of an overall entropic destabilization. Finally, in the C11A/C27A small loop mutant of lowest stability a very large enthalpic destabilization occurs, which is, however, partly counterbalanced by a reduction in the transition entropy. The preferential perturbation of the native state by the mutations is manifest in the increase of the native state heat capacity relative to that of the wild-type protein and the identity of the heat capacity of the unfolded state. Inspection of the structure of Tendamistat leads to a plausible explanation of the differences in the thermodynamic stability parameters. The N-terminal arm is likely to stabilize the native-like structure even in the absence of the 45 – 73 disulfide bridge, whereas no comparable mechanism is operative when the 11 – 27 bridge is open.