|
Figure 1. Wild-type subunits cannot rescue the lethality of the G466A subunits. (A) Whole cell Cs+ currents in tandem dimers. Holding potential, −120 mV, depolarizations from −80 to +60 mV in 10-mV increments. WT-WT tandem produces currents comparable to those of WT monomers. The arrow shows gating current. A tandem dimer of WT and G466A, with the mutant downstream of the WT, is completely nonfunctional. Cotransfected tsA201 cells were identified by 4-μm beads coated with CD8 antibody (Dynal Biotech). The tandem construct included the extra linker residues -NNNNNNAMN- between the two protomers. (B) Dominant-negative suppression. Coexpression of WT and G466A monomers reduces peak WT current amplitude at +70 mV in oocytes (n = 15–18 oocytes for each data point). The same molar amount of cRNA for WT was injected in each condition, with the remaining volume adjusted appropriately with water or mutant cRNA. The abscissa represents the molar fraction of WT cRNA. The dashed line is the suppression predicted by the binomial equation. The solid line is a single-parameter fit of a model in which there is an assembly penalty of 1.65 ± 0.14 kcal/mol for each mutant–WT contact in a tetramer (APPENDIX).
|
|
Figure 2. Western blots show that all constructs containing G466A, including the functional mutant G466A/V467G, produce roughly comparable amounts of protein, but with restricted glycosylation, as shown by the absence of the high molecular weight band observed in the WT.
|
|
Figure 3. Functionality tested in Gly mutants. (A) Functional Gly scan using FLAG-tagged double mutants expressed in Xenopus oocytes. Peak currents at +50 mV. All mutants contained G466A and cRNAs were injected at equimolar concentrations. Recordings were obtained 2 d after injection, n = 3–6 eggs. Similar results were obtained for all double mutants using mammalian cell expression (not depicted). (B) Cs+ currents obtained in whole cell recording as described in Fig. 1 A. (C) G-V relationships for Cs+ currents in WT, V467G, and G466A/V467G constructs (n = 4 in each case) in mammalian cells. These relationships were determined from isochronal tail current measurements. The G-V relations were fitted to the Boltzmann equation:GV=,where G(V) is normalized conductance, V1/2 is the half activation voltage, q is the unitless slope, and RT/F is 25 mV at room temperature. For WT: V1/2, q was −40.5 ± 0.8 mV, 4.99 ± 0.13; for V467G: V1/2, q was −55.2 ± 2.4 mV, 3.50 ± 0.23; for G466A/V467G: V1/2, q was −22.0 ± 1.2 mV, 2.16 ± 0.11. Free energy differences were calculated as ΔΔG = Δ(qFV1/2) for any two constructs. Although this is only a crude measure of the open–closed equilibrium energy, an improved approach based on the assumption that S6 mutants affect only the concerted opening transition (Yifrach and MacKinnon, 2002) is inconsistent with the differences between WT and the V467G mutant, because the mutation causes the G-V relationship to be both left shifted and shallower than that of the WT.
|
|
Figure 4. Constructs containing G466A, whether functional or not, are expressed on the plasma membrane. The top/bottom panel is blotted with anti-FLAG/anti-ERK antibody, respectively. The left side is protein from whole cell lysates of tsA201 cells transfected with the indicated constructs (see Fig. 2) and immunoprecipitated surface protein from the same cells on the right. The FLAG western shows comparable expression levels for the three Shaker constructs, both in whole cell lysates and from surface protein. Western blots from the same material shows the presence of the cytoplasmic proteins ERK1 (top band) and ERK2 in whole cell lysates but not from surface protein.
|
|
Figure 5. The G466A/V467G mutations have no effect on Rb+/K+ selectivity. (A) Currents shown from either an outside-out patch (WT on left) or whole cell recording (G466A/V467G on the right). Voltage protocols described in the text. (B) Reversal potentials of isochronal I-V plots were obtained by linear interpolation. Rb+ solutions caused a −6.9 ± 0.7 mV shift in WT (n = 4 cells) and a −4.7 ± 0.9 mV shift in mutant (n = 6) channels.
|
|
Figure 6. The G466A/V467G mutations have little effect on single channel conductance. (A) Examples of single channel openings in response to steps to +80 mV from a holding potential of −70 mV in cell-attached patches. Low-pass filter, 2 kHz. Data from these two patches were fit by amplitude histograms producing estimates of 2.01 pA for WT and 2.22 pA for G466A/V467G. (B) Nonstationary noise analysis for depolarizations to +80 mV in two cell-attached patches, 10 kHz low-pass filter. Estimates of single channel current i amplitude and Popen,max are shown. Estimates of i and Popen,max in nine WT patches were 1.84 ± 0.07 pA and 0.78 ± 0.01, respectively. In six G466A/V467G patches the estimates of i and Popen,max were 2.30 ± 0.09 pA and 0.68 ± 0.02, respectively.
|
|
Figure 7. The G466A/V467G mutations reduce block of intracellular tetraalkylammonium cations. Data shown from inside-out patches with reagents at indicated concentrations. Currents at +80 mV before and after exposure to blocker are superimposed. Dose–response relationships are shown in the bottom panels with molecular models of the blockers on the right. Data were fit using a single-site blocking model with Kd values of 3.77 μM (TBA) and 1.72 μM (TPentA) for WT channels, and Kd values of 730 μM (TBA) and 27.6 μM (TPentA) for G466A/V467G (n = 3–5).
|
|
Figure 8. The G466A/V467G mutations have little effect on the voltage dependence of TPentA block. Steady-state block at voltages between +20 and +80 mV from inside-out patch currents as in Fig. 7. The data were fit by the equation Fun/(1 − Fun) = C · exp(−δVF/RT), where Fun is the steady-state fraction of unblocked current, C is a constant, and δ is a unitless parameter representing the fraction of the electric field at the blocking site felt by the blocker and any ions that move with it. The estimates of δ were 0.15 ± 0.01 and 0.18 ± 0.03 for WT and G466A/V467G, respectively, for n = 3 patches each. These values are not significantly different (P > 0.05, t test).
|
|
Figure 9. Estimates of TBA blocking constants using stationary noise. Inside-out patches for G466A/V467G mutants in the absence (A and B) and presence (C and D) of 1 mM TBA. 160 depolarizations to +80 mV were applied, and stationary noise was measured over the intervals shown by the double arrows. A and C show mean and variance for these currents, with spectra shown in B and D. The spectrum of background noise at the −80 mV is shown in B. Data were fit by a weighted sum of a lorentzian and 1/f noise. The arrow indicates the corner frequency ƒc of the lorentzian in the patch exposed to 1 mM TBA.
|
|
Figure 10. Tests of steric clash. (A) Two opposing subunits of KcsA from a high-resolution crystal structure (1K4C; Zhou et al., 2001b), highlighting the position of the pore-lining residue homologous to Val467 in Shaker (red). Drawn with Swiss-PdbViewer (http://us.expasy.org) and POV-Ray (www.povray.org). (B) Single KcsA subunits showing the homologue to Shaker's Met440 (orange) and Gly466 (green) on the left, and the expected clash due to the G466A mutation on the right. K+ ions are depicted as gray spheres. (C) Western blot as in Fig. 2 showing poor glycosylation of all mutants containing G466A.
|
|
Figure .
|