Molecular Events behind the Selectivity and Inactivation Properties of Model NaK-Derived Ion Channels

Scheme 1. Panel (A), partial sequence alignment among KcsA (PDB 1K4C), NaK2K (PDB 3T4Z), and NaK (PDB 3E8H) channels. Residues forming the selectivity filter are shaded violet. The mutated residues that turn NaK into NaK2K are boxed. Panel (B), zoom of the selectivity filter and pore helix structures of KcsA, NaK2K, and NaK channels. The position of the fluorescent reporter used in this work (W55) at the pore helix is highlighted in green. K+ ions bound either to a 4-binding site or a 2-binding site SF are represented by violet spheres.

Figure 1. Cation permeation through NaK and NaK2K channels reconstituted in asolectin giant liposomes. Left panel shows representative patch-clamp recordings at +150 mV and −150 mV attained under symmetrical conditions using 10 mM Hepes buffer, pH 7.0, containing either 200 mM NaCl (upper pairs of recordings) or 200 mM KCl (lower pairs of recordings). The discontinuous line indicates the zero current level in each case. In parallel experiments, 100 µM tetrapentyl ammonium (TPeA) was added to the bath solution to ensure that recorded currents are from channels oriented with their external side facing the bath solution. Right panel shows the averaged I/V curves from three seals obtained under bi-ionic conditions using 10 mM Hepes buffer, pH 7.0, containing 200 mM K+ in the bath solution, and 10 mM Hepes buffer, pH 7, containing 180 mM NaCl and 20 mM KCl in the pipette solution. Symbols represent the average value of the current (±SD) measured at the different holding potentials. Straight lines are just a mere guide to the eye. Notice that the I/V relationship within the voltage range used to determine the reversal potential is fairly ohmic for both channels.

Figure 2. Steady-state fluorescence characterization of NaK and NaK2K channels. Panels (A,B) show the fluorescence emission spectra of both proteins taken in 20 mM Hepes buffer, pH 7.0, containing 1 mM DDM (n-dodecyl β-D-maltoside) and either 200 mM NaCl (pink traces) or 100 mM KCl (blue traces). The excitation wavelength was 295 nm. The shifts in these spectra were quantitated through the fluorescence spectral center of mass (intensity-weighted average emission wavelength, <λ>) (Panel (C)). While the emission spectra taken in Na+ are very much alike in both channels, there is a clear difference in the W55 environment in the presence of K+, where the spectra displayed a blue shift of ~2 nm in the NaK2K channel and only ~0.5 nm in the NaK channel. Panel (D) shows the steady-state anisotropy values (<r>) (λex: 300 nm), which also differentiate between the K+- and Na+-bound forms of each channel. The higher difference between these two ionic conditions was detected in the NaK2K channel, where the <r> value in the presence of K+ is particularly low, while in the presence of Na+, the anisotropy is almost identical to that of the NaK channel. Each bar corresponds to the average ± S.D. of three…

Figure 3. Time-resolved anisotropy characterization of the NaK and NaK2K channels solubilized in DDM micelles. Panel (A), top view of the NaK2K channel (PDB: 3T4Z), highlighting in blue sticks, the symmetric position of the four W55 residues, one in each subunit (numbered from 1 to 4), within a square frame (red discontinuous line). The lower scheme in this panel illustrates the reversible homo-FRET process that takes place among these tryptophan residues within a square geometry with a lateral distance R. Assuming that the initially excited W55 residue occurs at subunit 1, k1 and k−1 are the lateral Förster′s transfer rates and k2 and k−2 the diagonal transfer rates from this residue. Panels (B,C) show representative time-resolved anisotropy decays (λex: 300 nm, λem: 345 nm) of the Y55W NaK and NaK2K channels. Such decays are always faster in the presence of K+ than in Na+. Furthermore, as in the steady-state measurements, the larger differences between the two ionic conditions were found in the NaK2K channel, where the homo-FRET process is more relevant. Straight lines represent the best fit of Equation (3) to the data.

Figure 4. Effects of cations on the thermal denaturation of the NaK (panels (A,B)) and NaK2K (panels (C,D)) mutant channels. The intrinsic fluorescence intensity (F.I.) of the channel proteins was monitored at increasing temperatures upon the addition of increasing concentrations of K+ (panels (A,C)) or Na+ (panels (B,D)). The figure shows that in both mutant channels, there is a clear K+ concentration dependence of the thermal denaturation, while such a process is unaffected by the presence of Na+.

Figure 5. Effect of cations on the midpoint temperature (tm) of thermal denaturation of the NaK (black squares) and W55 (red circles) mutant channels. The effects on the tm of increasing K+ (panel (A)) or Na+ (panel (B)) concentrations, respectively, are shown for both channels. Each experimental point is the average tm (in Celsius) ± S.D. (n = 3), calculated from three thermal denaturation titrations as those shown in Figure 4. The insert to panel A shows the fitting of the K+ titration data from both channels, as described in detail in reference [45]. The goodness of fit indicates that the binding of K+ to a single set of binding sites on both channels suffices to explain the experimental data. The apparent dissociation constants (KD) estimated for such single binding events are given in Table 2.

Figure 6. Representative experiments of binding of K+ to the NaK (left panels) and NaK2K channels (right panels) monitored by the changes in the fluorescence spectral shift <λ> (Panels (A,D)), steady-state anisotropy <r> (Panels (B,E)), and W55-W55 intersubunit lateral distances (Panels (C,F)). The measurements were performed as indicated in Figure 2 and Figure 3. The straight lines represent the best fit to an empirical single binding event model, as explained in the Section 4. KD values from such fits are summarized in Table 2 as the average from two independent experiments.