Engineered Ionizable Side Chains 17
we reported pertain to those estimated from the bursts of long openings. As for
the number of exponential components observed for the distribution of the two
open-channel current-level dwell times, small deviations from two-state kinetics
were often observed. Fitting the data with models assuming two-state kinetics, how-
ever, never changed the conclusions. Also, in the muscle AChR, the protonation
and deprotonation rates turned out to be slow enough for the accuracy of their es-
timates not to depend heavily on whether the correction for missed events is exact
or, simply, a good approximation. These rates were as slow as ~ 5–10 s−1 for both
the protonation and deprotonation rates of a lysine engineered in the M3 α-helix of
the δ subunit (estimated at ~–150 mV and external 10-mM Hepes, pH 7.4, in the
cell-attached patch-clamp configuration; Cymes and Grosman 2008 ) and as fast as
~ 20,000–25,000 s−1 for the deprotonation of a histidine engineered in the middle
of the M2 α-helix and the protonation of a histidine in the most intracellular turn of
M2 (both estimated at ~ – 100 mV and external 10-mM Hepes, pH 7.4, in the cell-
attached patch-clamp configuration; Cymes et al. 2005 ).
When comparing the values of protonation and deprotonation rates estimated
under different experimental conditions, it is important to bear in mind that both
the association and dissociation rates of protons depend on the solutions’ pH. This
is quite different from the kinetics of association and dissociation of any other mol-
ecule or ion, in which case only the association rate depends on the concentration
of the ligand. This unique behavior of protons is due to the fact that aqueous solu-
tions contain hydroxide anions, which are proton acceptors. Hence, proton-disso-
ciation rates become faster (and proton-association rates become slower) as the pH
increases (Fig. 10 ). More generally, aqueous pH-buffered solutions contain three
types of proton acceptor (hydroxide anion, water and the deprotonated form of the
pH-buffer) and three types of proton donor (water, hydronium cations and the pro-
tonated form of the pH-buffer), which is why proton dissociation and association
rates in these solutions are each the sum of three rates (Fig. 9 ). This also explains
the dependence of the proton association and dissociation rates on the concentration
of pH-buffer observed experimentally (Fig. 11 ).
As for the effect of the membrane potential, we only tested negative values
(negative on the intracellular side) and found that both the proton-association and
proton-dissociation rates become faster with hyperpolarization (Fig. 12 ), consistent
with the notion that protons enter the channel from the extracellular side, pause on
the ionizable side chain, and exit to the intracellular side.
We found that the way the rates of proton transfer vary as a function of pH, con-
centration of pH-buffer and membrane potential is such that the pKa also changes
with these variables (Figs. 10c, 11 c and 12c, respectively). The voltage dependence
of a pKa is reminiscent of the well-described voltage dependence of the dissociation
equilibrium constant of pore blockers in a variety of channels; here, the proton acts
as the pore blocker. On the other hand, the pH dependence of a pKa is very likely
the result of the protonation state of not only the engineered side chain, but also,
of nearby (naturally occurring) ionizable side chains changing as a function of pH.
Thus, the electrostatic environment around the mutant side chain changes as the pH