Abstract
Ion channels of the DEG/ENaC family can induce neurodegeneration under conditions in which they become hyperactivated. The C. elegans DEG/ENaC channel MEC-4(d) encodes a mutant channel with a substitution in the pore domain that causes swelling and death of the six touch neurons in which it is expressed. Dominant mutations in homolog channel UNC-8 result in uncoordinated movement. Here we show that this unc-8(d) movement defect is correlated with the selective death of cholinergic motor neurons in the ventral nerve cord, a milder toxicity compared with MEC-4(d). UNC-8(G387E), denoted UNC-8(d), and UNC-8(A586T) mutants encode hyperactivated channels that are strongly inhibited by extracellular calcium and magnesium in Xenopus oocytes. Reduction of extracellular divalent cations exacerbates UNC-8(d) toxicity in oocytes. We suggest that inhibition by extracellular divalent cations limits UNC-8 toxicity and may contribute to the selective death of neurons that express UNC-8 in vivo (Chapter 3). It was previously shown that neurons expressing hyperactive DEG/ENaC channels die by necrosis mediated by intracellular Ca2+ overload. Hyperactive mammalian ASIC1a and C. elegans MEC-4(d) conduct both Na+ and Ca2+ raising the possibility that direct Ca2+ influx through these channels contributes to the intracellular Ca2+ overload. However, we showed that UNC-8(d) is not Ca2+ permeable but still causes neuronal death suggesting that Na+ influx is sufficient to induce cell death. MEC-4(d) and UNC-8(d) differ not only in Ca2+ permeability but also in current amplitude, UNC-8 being strongly blocked by physiologic Ca2+ concentration. Given that these two channels show a striking difference in toxicity in vivo, we asked what is the contribution of Na+ conductance and Ca2+permeability to cell death. To address this question we built an UNC-8/ MEC-4 chimeric channel that retains the calcium permeability of MEC-4 and characterized its properties in oocytes. Our data support the hypothesis that for Ca2+ permeable DEG/ENaC channels, such as MEC-4, both Ca2+ permeability and Na+ conductance contribute to toxicity. However, for Ca2+ impermeable DEG/ENaCs, such as UNC-8, constitutive Na+ conductance is sufficient to induce toxicity and this effect is enhanced as current amplitude increases (Chapter 4). UNC-8(d) is blocked by extracellular Ca2+ in the non-physiologic micromolar range but causes neuronal death. This suggests that homolog DEG/ENaC subunits and accessory proteins that are co-expressed with UNC-8 in cholinergic motorneurons in vivo may modulate its Ca2+ sensitivity. I found that none of the selected proteins changed UNC-8 Ca2+ sensitivity when co-expressed with UNC-8 in oocytes (Chapter 5). Reduction of extracellular divalent cations exacerbates UNC-8(d) toxicity in oocytes suggesting that inhibition by extracellular divalent cations limits UNC-8 toxicity and may contribute to the selective death of neurons that express UNC-8 in vivo. To test this hypothesis I proposed to investigate a transgenic C. elegans with mutant UNC-8(d) channel in which the calcium binding site has been altered, to see if removing the calcium block will change the level of channel toxicity in vivo. Using a chimeric approach we swapped extracellular domains of UNC-8 with the corresponding domains of the less calcium sensitive MEC-4. We find that residues in the extracellular “finger” domain are responsible for UNC-8(d) divalent cation sensitivity. Furthermore, our results show for the first time that residues in this extracellular “finger” domain are involved in the interaction of MEC-4 channel with accessory protein MEC-6 (Chapter 5). This work furthers refines the contribution of different channel properties to cellular toxicity induced by hyperactive DEG/ENaC channels and expands on previous structure-function studies by establishing properties and functions of DEG/ENaC channels specific domains.