Tetrameric, voltage-gated, potassium-selective K V 1.2 channels are encoded by KCNA2 and regulate axonal action-potential repolarization. Both gain and loss of KCNA2 function cause severe neurological disorders and, to our knowledge, loss-of-function variants are also dominant-negative. Here, we studied two aspartate substitutions at adjacent positions highly conserved throughout the K V -channel superfamily: p.H310D (ClinVar #1801703, associated with global developmental delay) and p.G318D: a de novo variant in a female patient with febrile seizures, global developmental delay and other symptoms. We performed electrophysiological recordings in Xenopus laevis oocytes and surface-trafficking assays in COS-7 cells using immunocytochemistry and flow cytometry. Oocytes injected with human K V 1.2(H310D) or K V 1.2(G318D) cRNA exhibited current comparable to that of uninjected oocytes ( p > 0.05, n > 5). Accordingly, neither variant-subunits could reach the cell surface ( n >5). Since both variants were loss-of-function, we evaluated their capacity for negative dominance by comparing the homozygous-wild-type and heterozygous conditions. “Heterozygous” oocytes injected with 1×K V 1.2 (wild-type) and 1×K V 1.2(H310D) cRNA (“1×” = 3 ng cRNA per oocyte) exhibited 7.3 ± 2.5% ( n = 12) macroscopic conductance compared to homozygous (2×) wild-type (100 ± 19%, n = 13). In cells co-expressing K V 1.2(H310D) subunits, only 49 ± 1.5% K V 1.2 (wild-type) subunits trafficked to the surface ( n = 8), relative to cells expressing only K V 1.2 (wild-type) subunits (100 ± 0.84%, n = 9). These results show that p.H310D is a strongly dominant-negative, loss-of-function variant. By contrast, oocytes “heterozygous” for p.G318D demonstrated 31 ± 7.3% ( n = 18) macroscopic conductance relative to homozygous-wild-type (100 ± 19%, n = 18). This fraction was further reduced in oocytes injected with 1×wild-type:2×G318D (6.1 ± 2.0%, n = 9). In cells co-expressing K V 1.2(H310D) subunits, 86 ± 2.5% K V 1.2 (wild-type) subunits reached the cell surface ( n = 9), relative to cells expressing only K V 1.2 (wild-type) (100 ± 1.9%, n = 8). The strong negative dominance by p.H310D is likely produced by the association of wild-type with H310D subunits, and their sequestration in non-trafficking hetero-tetramers. On the other hand, G318D likely weakens intersubunit interactions, producing less negative dominance. Given the patient’s severe neurological symptoms, this finding suggests that KCNA2 is haploinsufficient.
Boon et al. (Sun,) studied this question.