BACE1 and presenilin/γ-secretase regulate proteolytic processing of KCNE1 and 2, auxiliary subunits of voltage-gated potassium channels

Abstract

BACE1 and presenilin (PS)/γ-secretase play a major role in Alzheimer's disease pathogenesis by regulating amyloid-β peptide generation. We recently showed that these secretases also regulate the processing of voltage-gated sodium channel auxiliary β-subunits and thereby modulate membrane excitability. Here, we report that KCNE1 and KCNE2, auxiliary subunits of voltage-gated potassium channels, undergo sequential cleavage mediated by either α-secretase and PS/γ-secretase or BACE1 and PS/γ-secretase in cells. Elevated α-secretase or BACE1 activities increased C-terminal fragment (CTF) levels of KCNE1 and 2 in human embryonic kidney (HEK293T) and rat neuroblastoma (B104) cells. KCNE-CTFs were then further processed by PS/γ-secretase to KCNE intracellular domains. These KCNE cleavages were specifically blocked by chemical inhibitors of the secretases in the same cell models. We also verified our results in mouse cardiomyocytes and cultured primary neurons. Endogenous KCNE1- and KCNE2-CTF levels increased by 2- to 4-fold on PS/γ-secretase inhibition or BACE1 overexpression in these cells. Furthermore, the elevated BACE1 activity increased KCNE1 processing and shifted KCNE1/KCNQ1 channel activation curve to more positive potentials in HEK cells. KCNE1/KCNQ1 channel is a cardiac potassium channel complex, and the positive shift would lead to a decrease in membrane repolarization during cardiac action potential. Together, these results clearly showed that KCNE1 and KCNE2 cleavages are regulated by BACE1 and PS/γ-secretase activities under physiological conditions. Our results also suggest a functional role of KCNE cleavage in regulating voltage-gated potassium channels.

Keywords: Alzheimer's disease; KCNQ1; MiRP; MinK.

Figure 1.

Putative BACE1 and PS/γ-secretase cleavage sites of human KCNE1 and KCNE2.

Figure 2.

KCNE1 and KCNE2 undergo sequential cleavage mediated by α- and PS/γ-secretases. A) Schematic diagram showing sequential cleavage of KCNE1 and KCNE2 by BACE1, α-, and PS/γ-secretases. α-Secretase inhibitor (TAPI-1) or BACE1 inhibitor (_DR9) decreases generation of KCNE C-terminal fragments (KCNE-CTFs), while PMA, an α-secretase activator, increases KCNE-CTF levels. KCNE-CTFs are then cleaved by PS/γ-secretases to generate KCNE intracellular domains (KCNE-ICDs). B) Western blot analysis of human KCNE1 full-length (F.L.) and its CTF expressed in B104 cells. KCNE1-CTF levels were increased by treatment with DAPT and further elevated by cotreatment with PMA, while partially decreased by TAPI-1. C) Western blot analysis of human KCNE2 F.L. and its CTF expressed in B104 cells. Similar to KCNE1, KCNE2-CTF levels are also increased by DAPT treatment and further elevated by cotreatment with PMA.

Figure 3.

PS/γ-secretase activity regulates generation of KCNE1- and KCNE2-ICDs. A) Western blot analysis showed that KCNE1-ICD is specifically regulated by PS/γ-secretase activity in B104 cells stably expressing full-length KCNE1. Epoxomicin, a proteasome inhibitor, prevented KCNE1-ICD degradation. B) KCNE2-ICD is detected in the similar conditions.

Figure 4.

Elevated BACE1 activity increases KCNE1- and KCNE2-CTF levels in B104 rat neuroblastoma cells. A) Overexpression of human BACE1 increased KCNE1-CTF levels in B104 cells. DAPT treatment further enhanced KCNE1-CTF levels in control and BACE1-overexpressing cells. B) KCNE2-CTF levels were also elevated by BACE1 overexpression in B104 cells.

Figure 5.

BACE1 and PS/γ-secretase activities regulate endogenous KCNE1-CTF levels. A) Detection of endogenous KCNE1 full-length (F.L.) and CTF bands in cultured mouse cardiomyocytes. DAPT (5 μM) treatment increased endogenous KCNE1-CTF levels. B) Endogenous KCNE1-CTF levels were also increased by BACE1 overexpression in mouse cardiomyocytes.

Figure 6.

α-Secretase, BACE1, and PS/γ-secretase activities regulate KCNE2-CTF levels in cultured mouse primary neurons. A) Detection of endogenous KCNE2 full-length (F.L.) and KCNE2-CTF bands in cultured mouse primary cortical/hippocampal neurons (DIV14). DAPT treatment specifically increased 10-kDa KCNE2-CTF levels. B) Relative levels of KCNE2-CTF in panel A were quantitated (n=3/condition). *P < 0.05. C) Endogenous KCNE2-CTF levels were also elevated by BACE1 overexpression, as compared to the primary neuronal cells transfected with control GFP vectors. D) Quantitative analysis of KCNE2-CTF levels shown in panel B (n=3/condition). **P < 0.01. E–G) Mouse primary neurons were also treated with 1 μM BACE inhibitor IV (E) or 1 μM GM6001 (a pan-α-secretase inhibitor; G) for 3 d, and KCNE2-CTF levels were analyzed by Western blotting. F) Quantitative analysis of KCNE2-CTF levels in neurons treated with BACE1 inhibitor IV or DMSO control. Relative KCNE2-CTF levels were normalized by full-length KCNE2 to avoid blot-to-blot variations (n=3/condition). *P < 0.05.

Figure 7.

BACE1 activity modulates KCNQ1/KCNE1 current activation in HEK293T cells. Whole-cell recordings of KCNQ1 and KCNE1 transiently transfected in HEK293T cells with or without overexpressing wild-type BACE1. A) Western blot analysis detects human KCNQ1 and KCNE1 coexpression in HEK293T cells. As shown in Figs. 2 and 3, either BACE1 expression and/or DAPT treatment increases KCNE1-CTF levels. Asterisks indicate nonspecific bands that are detected when transiently expressing KCNE1, not in stable B104 cells. B) Activation of KCNQ1/E1 currents by depolarizing voltage steps from −40 to +80 mV (inset). Prepulse duration at −80 mV was 5 s. C) Normalized conductance was plotted against voltage from experiments as in panel B. Curves were fitted with a Boltzmann equation. Conductance of the cells expressing KCNQ1/KCNE1 (KCNQ1/E1, black) was compared with the cells coexpressing wild-type BACE1 (red) and inactive mutant BACE1 (D289N; green). Significance was tested for the half-maximal activation (V_mid). *P < 0.05, ***P < 0.001.