The Thermodynamics of Cd(II) Binding to Human Cardiac Troponin C: Using Mutants to Determine the Location of Cd(II) Binding

Abstract

Historically, cadmium (Cd) has been used in many different applications from cadmium paint to nickel-cadmium batteries. However, Cd is a nonessential, naturally occurring toxic element due to its low excretion rate and half-life of 10-30 years in humans (1). Due to similar properties, such as atomic radii and charge/radius ratios, Cd(II) has been found to replace the essential divalent calcium (Ca) ion in many proteins, including calmodulin, parvalbumin, and troponin C (1). A specific protein of interest is human cardiac troponin C (HcTnC), a Ca(II) binding protein in the EF hand family. This protein has four EF hand binding loops, where loops I and II are found in the regulatory domain of the protein, and loops III and IV play a more structural role. Interestingly, loop I is a "defunct" loop and does not bind Ca(II). Previous crystal structures of Cd(II) bound to the regulatory domain of HcTnC reveal two Cd(II) ions bound to EF hand loops I and II (2). While isothermal titration calorimetry (ITC) data from the Spuches lab reveals two Cd(II) ion binding to the N-domain (as verified in this work), displacement studies strongly suggest that Cd(II) is not binding to loop II as noted in the crystal structure. Cd(II) is known to bind strongly to cysteine residues in proteins, therefore our hypothesis is that Cd(II) is binding to loop I of the protein due to the cysteine residue present. To test this hypothesis, the following mutants have been overexpressed and purified: C35A, C84A, and the double-mutant C35A/C84A. ITC studies of both Ca(II) and Cd(II) binding to HcTnC mutants have been performed and compared to wild-type. It was expected that Ca(II) binding to the mutants would not be altered, but Cd(II) binding to both the C35A and double-mutant would not display a heat event corresponding to Cd(II) binding to the regulatory domain. It was also hypothesized that binding of Cd(II) to the C84 mutant protein would remain unchanged as compared to the Ca(II) binding. Thermodynamic studies revealed very unexpected results. All three mutants displayed a total of 3 Cd(II) ions binding to the protein, with 2 Cd(II) ions binding to the C-domain and one additional Cd(II) binding to the N-domain. This suggested that cysteine, a residue important to Cd(II) binding, was not necessary since the dual mutant still displayed favorable binding of 1 Cd(II) ion to the protein. In fact, Cd(II) binding to the dual protein does so with more favorable Gibbs free energy of binding than the other 2 mutants. Back titrations of Ca(II) revealed that Cd(II) is not binding to Loop II of the protein as was the case with wild-type. This data allowed us to propose a new model where Cd(II) binds to Loop I of the protein since this is a prearranged binding domain, and that Cys84 may play a role in metal loading. Future studies include NMR and Mass Spectrometry to look at metal location and speciation.