|
Figure 1. The D50N KID mutation exhibits substantially impaired regulation by extracellular Ca2+ over a wide voltage range. (A) Representative hemichannel currents in Xenopus oocytes obtained using a voltage step protocol that consisted of 10-s steps, from 60 to −110 mV in intervals of 10 mV, followed by a 5-s step to −110 mV. Oocytes were voltage clamped to −20 mV between steps. Currents shown in each case are from single oocytes exposed to all three Ca2+ concentrations. WT Cx26, A40V, and G45E all showed substantial reductions in current magnitude at the higher Ca2+ concentrations, whereas D50N was nearly insensitive. (B) Bar graph showing ratios of the values of the macroscopic conductances measured in 2 and 0.2 mM Ca2+, G2Ca2+/G0.2Ca2+. The conductances were calculated from currents elicited by brief (1 s) voltage steps of ±10 and 20 mV from a holding potential of −20 mV. D50N exhibits near insensitivity to Ca2+. Each bar represents the mean ratio ± SE. n = 12 for WT Cx26, 8 for A40V, 12 for G45E, and 10 for D50N. (C) Plots showing mean ratios of macroscopic conductances measured in 2 and 0.2 mM Ca2+, G2Ca2+/G0.2Ca2+, over a voltage range of −80 to 40 mV. G-V curves were obtained from voltage ramps applied from 40 to −100 mV, 600 s in duration. Oocytes were held at a holding potential of −20 mV. Before initiating the ramps, oocytes were stepped to 40 mV for 30 s to allow conductance to reach steady-state. Ca2+ regulation shows some voltage dependence, with more effective inhibition of currents at large negative membrane potentials. D50N shows poor Ca2+ regulation over the entire voltage range.
|
|
Figure 2. The D50N mutation substantially alters open-state conductance and rectification of Cx26 hemichannels. Representative examples of patch-clamp recordings from inside-out patches containing single WT Cx26 (left) and D50N (right) hemichannels are shown. Currents shown were obtained in response to 8-s voltage ramps applied between ±70 mV and leak subtracted (see Materials and methods). WT Cx26 shows slight outward rectification, whereas D50N exhibits a reduced slope conductance (measured at 0 mV) and strong outward rectification. Bath and pipette solutions consisted of IPS (see Materials and methods).
|
|
Figure 3. SCAM experiments using MTS reagents indicate D50 is a pore-lining residue. (A and B) Shown are effects of MTSES (A) and MTSET (B) on macroscopic D50C hemichannel currents. Oocytes were held at −40 mV throughout. Oocytes were exposed to 100 mM NaCl salt solutions with Ca2+ concentrations as indicated in the top bar; black segments indicate oocytes were initially bathed in MND96. Because of the poor sensitivity to Ca2+, 200 µM Zn2+ (yellow segments) was used to periodically close D50C hemichannels. Application of 2 mM MTSES and 0.2 mM MTSET is indicated by arrows above the orange and blue segments, respectively. (C) Summary of the percent change in current (mean ± SE) after application of each reagent. n = 43 for MTSES, 18 for MTSET, and 4 for MTS-biotin X. (D–F) Representative examples of patch-clamp recordings obtained from cell-attached patches containing single D50C hemichannels recorded from oocytes that were not exposed to MTS reagents (D), were exposed to 2 mM MTSES (E), and were exposed to 0.2 mM MTSET (F). All unitary currents were obtained in response to 8-s voltage ramps between ±70 mV and leak subtracted. Bath and pipette solutions consisted of IPS.
|
|
Figure 4. Loop gating remains strong in D50N, but activation is shifted compared with WT Cx26. (A and B) Shown are normalized G-V relationships of WT Cx26 (A) and D50N (B) hemichannels expressed in Xenopus oocytes at two different external Ca2+ concentrations, 0.2 and 2 mM. Data were obtained by applying slow (600 s) voltage ramps from 40 to â100 mV from a holding potential of â20 mV. Ramps were obtained for each oocyte in 0.2 and 2 mM Ca2+ and normalized to the maximum value measured in 0.2 mM Ca2+. In WT Cx26, increasing Ca2+ shifted activation in the depolarizing direction and suppressed current magnitude. D50N was nearly insensitive to Ca2+. Each symbol represents the mean value. n = 11 for WT Cx26 and 7 for D50N. For clarity, only error bars for +SE are shown for 0.2 mM Ca2+ and âSE for 2 mM Ca2+. (C) Superimposition of G-V curves for WT and D50N with error bars omitted for clarity. Compared with WT Cx26, activation of D50N is shifted positive in low Ca2+ (0.2 mM) conditions and negative in high Ca2+ conditions (2 mM). Data are the same as in A and B.
|
|
Figure 5. Single hemichannel recordings demonstrate that D50N affects loop gating. Multichannel cell-attached patch recordings of D50N and Cx26 WT hemichannels are shown. The patch pipette and bath solutions consisted of IPS. For the D50N, shown are 2â3-min consecutive segments of current recorded at membrane potentials of â20, 20, 40, and 50 mV. The solid lines designated C represent the leak conductance of the patch when no hemichannels were open. Solid lines designated On represent current levels of single or multiple fully open hemichannels. Dashed lines represent intermediate conductances (substates). All point histograms obtained from these recordings are shown to the left of each trace. At voltages of â20, 20, and 40 mV, gating events are uniform in size and represent transitions between closed and fully open hemichannels. The inset below the 40 mV trace shows an open and closing transition at an expanded time scale that is characteristic of loop gating. At 50 mV, many subconducting transitions become evident and are consistent with the onset of Vj or fast gating. Note that D50N hemichannels exhibit substantial closure even at â20 mV as expected because of the shift in loop gating caused by the mutation. For the Cx26 WT hemichannel, the currents shown were recorded at â30 mV and 30 mV. Note that open probability for WT Cx26 is similar at both voltages. All currents were filtered at 1 kHz, and data were acquired at 5 kHz.
|
|
Figure 6. The presence of a negative charge at position 50 confers WT open hemichannel properties but a variable degree of Ca2+ regulation. (A and B) Values of unitary conductance (A) and open hemichannel rectification (B) for WT Cx26 and several D50 mutant hemichannels. Each bar represents the mean value ± SD; n = 18 for WT Cx26, 16 for D50N, 14 for D50A, 10 for D50E, 11 for D50C, 8 for D50C + MTSES, and 12 for D50C + MTSET. Open hemichannel conductance was measured as the slope conductance at Vm = 0. Rectification was measured as the ratio of the open hemichannel current at 70 and â70 mV. (C and D) Plots showing mean ratios of the macroscopic conductances measured in 2 and 0.2 mM Ca2+, G2Ca2+/G0.2Ca2+ over a voltage range of â80 to 40 mV. G-V curves were obtained from slow voltage ramps as described in Fig. 1 C. (C) Plots superimposing WT Cx26 and mutants with different substitutions at position 50. (D) Plots superimposing WT Cx26 and the D50C mutant, unmodified and modified with MTSET or MTSES.
|
|
Figure 7. Possible intra- and intersubunit interactions at the border of TM1 and E1 of Cx26. (A) Illustration depicting three of six connexin subunits around a central aqueous pore. Each subunit is represented by a different color. Cylinders inside the subunits represent the transmembrane domains. TM1 denotes the first transmembrane domain that is exposed to the pore. E1 and E2 denote the first and second extracellular loop domains, respectively. D50 and D46 residues in E1 of each subunit are shown projecting into the pore. For the green (left) and cyan (middle) subunits, the D50 side chains come into proximity with K61 (situated on a short helix in E1) and Q48 (in E1) from the adjacent subunits. D46 also comes into close proximity to Q48. The illustration is not drawn to scale or strictly according to the atomic structure to emphasize the arrangement of interacting residues at the extracellular end of the pore. (B) Representation of the same three connexin subunits that constitute a Cx26 hemichannel viewed from the pore using the atomic structure published by Maeda et al. (2009). The superimposed cone in the left panel depicts the gross shape of the pore. The boxed region is shown in two expanded views to highlight the interacting residues between the green and cyan subunits. Only segments A39-R75 and C180-E187 are displayed in the expanded views for clarity. In the originally published structure (Protein Data Bank accession no. 2ZW3; middle), K61 in the cyan subunit is shown in close proximity to the backbone carbonyl of F51 in the same subunit. D50 and Q46 in the green subunit are in close proximity to Q48 from the adjacent (cyan) subunit. Other putative interactions shown are D46 with R184 in the same (green) subunit and R184 with D47 of next adjacent (cyan) subunit. Segments of E1 and TM1 are indicated. In the alternative structure (2ZW3rev; right) K61 in the cyan subunit is positioned in closer proximity to D50 in the adjacent (green) subunit. Structures were generated using PyMOL software.
|
|
Figure 8. K61C functions as a heteromer and confers Ca2+ regulation. (A) Plots showing mean ratios of the macroscopic conductances measured in 2 and 0.2 mM Ca2+, G2Ca2+/G0.2Ca2+, over voltage range from â80 and 40 mV. G-V curves were obtained in oocytes injected with mRNA for WT Cx26, D50N mutant, and a 1:4 mixture of D50N:K61C. Oocytes injected with K61C alone did not exhibit measureable currents (not depicted). (B) Examples of single hemichannel currents obtained from an oocyte expressing D50N:K61C in a 1:4 ratio (gray lines). Currents are in response to voltage ramps applied from 70 to â70 mV. Each trace represents a separate patch with selected examples of hemichannels with intermediate conductance and rectification properties. Representative I-V relations of open WT Cx26 and D50N hemichannels are displayed as dashed lines. (C) Values of unitary hemichannel conductance measured as the slope conductance at Vm = 0 for WT Cx26 (â¾), D50N:K61C (gray circles), and D50N (â´). Values for D50N:K61C include those after exposure to MTSET (â¡) or MTSES (â). Each point represents data from a separate patch. WT Cx26 and D50N hemichannels each clustered tightly about their mean values, whereas D50N:K61C was spread out over a wider range, indicative of hemichannels with different conductances.
|
|
Figure 9. Mutant cycle analysis suggests that D50 and Q48 residues interact. (A) G-V relationships for WT Cx26, D50A, Q48A, and double mutant D50A + Q48A hemichannels at two different external Ca2+ concentrations, 0 (or nominal) and 2 mM. Data were obtained by applying slow (600 s) voltage ramps from 40 to â100 mV from a holding potential of â20 mV. For each oocyte, conductances at both Ca2+ concentrations were normalized to the maximum value in 0 mM Ca2+. n = 11 for WT Cx26, 17 for D50A, 14 for Q48A, and 6 for D50A + Q48A. The G-V curves are arranged as a mutant cycle where residues D50 and Q48 are mutated separately and together. The arrows connecting each pair of hemichannel phenotypes refer to mutation of either the D50 residue (horizontal arrows) or the Q48 residue (vertical arrows). The double D50A + Q48A mutant resembles D50A alone, suggestive of an interaction between D50 and Q48. (B) Fitted G-Vm relations normalized to their maximum conductance obtained by fitting data to a stochastic gating model developed for connexin channels (Paulauskas et al., 2009, 2012; see Materials and methods). These G-V curves represent voltage-dependent gating independent of rectification of open hemichannels. G-V curves for WT and mutant hemichannels are superimposed in 0 mM Ca2+ (left) and 2 mM Ca2+. Note the similarity of phenotypes displayed by D50A and double mutant D50A + Q48A in 0 and 2 mM Ca2+.
|
|
Figure 10. Effects of DTT on Cys-substituted hemichannels. (A) Representative records of membrane currents in oocytes voltage clamped at −40 mV. Voltage was occasionally stepped to 40 mV for 10 s to test for hemichannel activation. Application of 1 mM DTT (yellow) produced a large increase in current in oocytes expressing Q48C and the double mutant D50C + Q48C (bottom two panels). The effect of DTT was readily reversible after wash-out only for Q48C. DTT had essentially no effect on WT and D50C hemichannels (top two panels). (B) Bar graph summarizing the effects of DTT on WT Cx26, D50C, Q48C, and Q48C + D50C (C-C). Ala substitution at Q48 is also included. For each oocyte, current magnitude was measured at −40 and 40 mV before and after DTT. Data are plotted as fold increase in current after DTT. n = 13 for WT Cx26, 11 for D50C, 14 for Q48C, 7 for Q48A, and 7 for D50C + Q48C. Error bars represent mean ± SE.
|
|
Figure 11. Q48C and D50C can form disulfide bonds between adjacent subunits. Western blots generated from samples of unpaired oocytes injected with mRNA for WT Cx26, D50C, Q48C, and D50C + Q48C. Samples were run in nonreducing (â) and reducing conditions (+) in which DTT and ME were added. Single Cys substitutions (D50C and Q48C) presented monomers and dimers; the latter was sensitive to reducing agents. However, double Cys mutant (D50C + Q48C) presented multimers, also sensitive to reducing agents, consistent with disulfide bond formation between multiple subunits. Molecular mass is indicated in kilodaltons.
|
