EAG鉀離子通道之次家族聚合專一性的結構基礎

Abstract

EAG (ether-a-go-go)鉀離子通道屬於電位控制鉀離子通道（voltage-gated potassium channels；KV channels）。一個具功能的電位控制鉀離子通道是由四個α-次單元組成，而α-次單元間須經由特殊的蛋白質區域進行專一性的聚合才能形成四聚體。EAG鉀離子通道家族可分為三個不同次家族，分別為Eag、Erg及Elk。這些離子通道構造相當類似，但因為特殊蛋白質區域控制聚合的專一性的差異，所以不同次家族的α-次單元無法聚合成同一個四聚體。過去的研究顯示Eag 鉀離子通道羧基端Carboxyl terminus (C端)的carboxyl assembly domain (CAD)以及Erg鉀離子通道C端的tetramerizing coiled coil (TCC)可能與聚合作用專一性的控制有關。但是是否還有其他影響聚合專一性的區域，例如胺基端Aminal terminus (N端)的PAS構造以及C端的C-linker及CNBHD，則尚未有進一步的研究。 本論文的目的是要建構rat Eag1 (rEag1)及human Erg (hErg)之間的chimera，來進一步探討EAG鉀離子通道次家族聚合專一性結構基礎。首先，我們利用rEag1-G440C和hErg-G628S這兩種無法通透鉀離子的孔道點突變(pore mutant)進行電生理實驗。這兩個孔道點突變分別與rEag1-WT和hErg-WT進行共同表現，我們觀察到rEag1-G440C會對rEag1-WT造成dominant-negative effect的效果，，而hErg-G628S則只對hErg-WT造成dominant-negative effect；這個結果再次確認屬從EAG家族不同次家族成員rEag1及hErg離子通道的α-次單元之間並不能互相聚合。接著，我們建構一系列rEag1及hErg C端(包括S6、C-linker、CNBHD和含有CAD、TCC之遠C端)蛋白質區域的chimera突變(chimera-A、B、C、D、E、F、G)。我們先觀察這些chimera突變的離子通道功能，若所表現的電流較大者則可以進行與無功能的孔道點突變做共同表現，而電流表現較小者則進一步結合孔道點突變後再分別與WT共同表現。我們再根據是否有dominant-negative effect，以判斷不同α-次單元間是否有聚合作用的發生。我們發現，只有更換含有CAD/TCC區域(遠C端)之chimera C與G才會有rEag1與hErg互相產生dominant-negative effect的現象。令我們驚訝的是，這四個chimera (rEag1-chimera C/G；hErg-chimera C/G) 能跨越次家族藩籬產生作用，而且還能繼續保有對原屬次家族的dominant-negative effect。這暗示了CAD/TCC可能不是唯一決定聚合專一性的蛋白質區域。我們也利用不具有CAD的rEag1-K848X突變以及去除TCC的hErg-R1032X突變進行相同實驗，結果發現這兩種去除CAD/TCC突變皆只能與原屬次家族聚合，但不能跨越次家族界線。這些結果證實，除了CAD與TCC外，rEag1和hErg尚有其他蛋白質區域決定其聚合專一性。 接下來，我們交換rEag1及hErg的cytoplasmic N terminus (chimera N)，發現chiemera N與chimera C/G一樣皆可跨越次家族間的界線而與另一次家族離子通道聚合，並保留與原屬次家族離子通道產生相互作用的能力。但如果只突變rEag1及hErg的PAS結構(chimera P)，則不具有次家族跨界聚合的能力。這個結果顯示N端上連接著PAS結構與S1穿膜段落的N-linker可能也對聚合的專一性有所貢獻。當我們進一步同時交換rEag1及hErg的N與遠C端(chimera NC)，發現rEag1-chimera-NC只能與hErg進行聚合作用，而不能與rEag1相互作用，然而hErg-chimera-NC卻還是能同時與rEag1及hErg聚合。這結果顯示，rEag1離子通道在N-linker以及遠C端(CAD)可能具有辨認次家族專一性的功能。控制聚合專一性類似的現象很可能也存在hErg離子通道，但是我們必須進一步針對N-linker區域進行研究才可能驗證此假說。 有趣的是，我們還發現rEag1-chimera-P和N會產生類似hErg的不活化 (inactivation) 現象。但是與hErg不同的是，這兩種chimera的不活化後尾巴電流(tail current)會出現次第減少的現象。我們提出證據顯示尾巴電流減少是因為處於不活化態的rEag1離子通道有一部分不經開啟態而直接由不活化態回到關閉態，並且這種由不活化態直接至關閉態的現象會隨著尾巴電位越過極化而更明顯。

EAG (ether-a-go-go) channel family, comprising Eag, Elk and Erg, are voltage-gated potassium channels. A functional voltage-gated potassium channel is composed of four α-subunits via the assembly of specific protein regions. Each member in the EAG channel family owns a specific protein region determinating its heteroteramerization and selectivity; therefore, Eag and Erg channel subunits can not form herteroteramers mutually. Crystal structures and preceded reaserches showed that there are several protein regions involved in the assembly of α-subunits, such as the PAS domain, S-6 segment and cytoplasmic carboxyl terminus. In this study, we suggest that the S-6 segment, C-linker and CNBHD are not important for specific recognition between sub-families of the Eag channel famly. Chimera mutants that exchange the carboxyl assembly domain (CAD) of rEag1 and the tetramerizing coiled-coil domain (TCC) of hErg can form not only homoteramers but also heterotetramers with the wild-type of each other. The rEag1 K848X and hErg R1032X mutants, which lack the CAD and TCC domain respectively, can not break the specific assembly between rEag1 and hErg. This data provided evidence suggesting that the coiled-coil structure is not the only recognition site between these α-subunits. In this study, we firstly found the fact that chimera mutants exchanging the N-terminus of rEag1 and hErg channels could also form heterotetramers with the wild-type of each other; however, PAS chimera mutants could not assemble with other species. These results indicated that the proximal N-linker residing in the proximal N-terminal also plays an important role in the recognition of species. In this study, we also introduced inactivation into noninactivating rEag1 by inserting the hErg PAS domain. This inactivation results in the reduction of tail currents with voltage dependence. In other words, the amplitude of the tail current will be reduced at a more negative tail potential. To interprete this phenomenon, we suggest that some inactivated channels can be deactivated directly from the inactivated state to the closed state without recovering to the open state.