18 G. D. Cymes and C. Grosman
does, and hence, its pKa need not remain the same. This phenomenon, which is not
predicted by the mathematical expressions in Fig. 9 , has been well-documented in
other proteins and is usually referred to as a “non-Henderson–Hasselbalch” behav-
ior (see Supplementary Fig. 3 in Cymes and Grosman 2012 , for example). Finally,
the dependence of the pKa on the concentration of pH-buffer, which is not predicted
by Fig. 9 either, remains most puzzling to us.
Fig. 10 pH dependence of the rates and equilibrium constant of proton transfer. The data (out-
side-out configuration; − 100 mV; 10-mM pH-buffer; 1-μM ACh) were recorded from HEK-293
cells transiently expressing the δS12ʹK mutant. The indicated pH values are those of the bath and
pipette solutions. The data points corresponding to the rates (a, b) were not fitted with any func-
tion because we used different pH-buffers to cover the examined pH-range, and hence, the values
of k−B− and k + BH in Fig. 9 were not the same at all pH values. The data points corresponding to the
equilibrium constant (c) were not fitted with any function, either, because a “Henderson–Hassel-
balch-type” titration predicts a pH-independent pKa. Although deviations from this simple type of
titration can be modeled (see Supplementary Fig. 3 in Cymes and Grosman 2012 ), we do not have
enough information to do this, here