Y. Ri et al., The role of a conserved proline residue in mediating conformational changes associated with voltage gating of Cx32 gap junctions, BIOPHYS J, 76(6), 1999, pp. 2887-2898
We have explored the role of a proline residue located at position 87 in th
e second transmembrane segment (TM2) of gap junctions in the mechanism of v
oltage-dependent gating of connexin32 (Cx32). Substitution of this proline
(denoted Cx32P87) with residues G, A, or V affects channel function in a pr
ogressive manner consistent with the expectation that a proline kink (PK) m
otif exists in the second transmembrane segment (TM2) of this connexin. Mut
ations of the preceding threonine residue T86 to S, A, C, V, N, or L shift
the conductance-voltage relation of wild-type Cx32, such that the mutated c
hannels close at smaller transjunctional voltages. The observed shift in vo
ltage dependence is consistent with a reduction in the open probability of
the mutant hemichannels at a transjunctional voltage (V-j) of 0 mV, In both
cases in which kinetics were examined, the time constants for reaching ste
ady state were faster for T86N and T86A than for wild type at comparable vo
ltages, suggesting that the T86 mutations cause the energetic destabilizati
on of the open state relative to the other states of the channel protein. T
he structural underpinnings of the observed effects were explored with Mont
e Carte simulations, The conformational space of TM2 helices was found to d
iffer for the T86A, V, N, and L mutants, which produce a less bent helix (s
imilar to 20 degrees bend angle) compared to the wild type, which has a sim
ilar to 37 degrees bend angle. The greater bend angle of the wild-type heli
x reflects the propensity of the T86 residue to hydrogen bond with the back
bone carbonyl of amino acid residue 182, The relative differences in propen
sity for hydrogen bonding of the mutants relative to the wild-type threonin
e residue in the constructs we studied (T86A, V, N, L, S, and C) correlate
with the shift in the conductance-voltage relation observed for T86 mutatio
ns. The data are consistent with a structural model in which the open confo
rmation of the Cx32 channel corresponds to a more bent TM2 helix, and the c
losed conformation corresponds to a less bent helix. We propose that the mo
dulation of the hydrogen-bonding potential of the T86 residue alters the be
nd angle of the PK motif and mediates conformational changes between open a
nd closed channel states.