Kv1.5 channels are regulated by PKC-mediated endocytic degradation

Abstract

The voltage-gated potassium channel Kv1.5 plays important roles in the repolarization of atrial action potentials and regulation of the vascular tone. While the modulation of Kv1.5 function has been well studied, less is known about how the protein levels of Kv1.5 on the cell membrane are regulated. Here, through electrophysiological and biochemical analyses of Kv1.5 channels heterologously expressed in HEK293 cells and neonatal rat ventricular myocytes, as well as native Kv1.5 in human induced pluripotent stem cell (iPSC)-derived atrial cardiomyocytes, we found that activation of protein kinase C (PKC) with phorbol 12-myristate 13-acetate (PMA, 10 nM) diminished Kv1.5 current (I_Kv1.5) and protein levels of Kv1.5 in the plasma membrane. Mechanistically, PKC activation led to monoubiquitination and degradation of the mature Kv1.5 proteins. Overexpression of Vps24, a protein that sorts transmembrane proteins into lysosomes via the multivesicular body (MVB) pathway, accelerated, whereas the lysosome inhibitor bafilomycin A1 completely prevented PKC-mediated Kv1.5 degradation. Kv1.5, but not Kv1.1, Kv1.2, Kv1.3, or Kv1.4, was uniquely sensitive to PMA treatment. Sequence alignments suggested that residues within the N terminus of Kv1.5 are essential for PKC-mediated Kv1.5 reduction. Using N-terminal truncation as well as site-directed mutagenesis, we identified that Thr15 is the target site for PKC that mediates endocytic degradation of Kv1.5 channels. These findings indicate that alteration of protein levels in the plasma membrane represents an important regulatory mechanism of Kv1.5 channel function under PKC activation conditions.

Keywords: Kv1.5; electrophysiology; endocytosis; ion channel; molecular biology; patch clamp; protein kinase C; structure–function relationships; ubiquitination; voltage-gated potassium channel.

Figure 1

PMA treatment decreases expression and current of Kv1.5 channels in HEK293 cells.A and B, concentration-dependent effects of PMA treatment for 3 h on the Kv1.5 expression (A, n = 5) and I_Kv1.5 (B, n = 6–13 cells for each concentration). C and D, time-dependent effects of PMA (10 nM) on the Kv1.5 expression (C, n = 5) and I_Kv1.5 (D, n = 7–14 cells for each time point). For western blot images, the actual molecular markers (the BLUeye Prestained Protein Ladder) are indicated beside the bands run on the same gels throughout the study. The 75-kDa band represents mature, fully glycosylated plasma membrane-located Kv1.5 channels, whereas the 68-kDa band represents core-glycosylated ER-located Kv1.5 channels. For western blot analysis, ∗p < 0.05, ∗∗p < 0.01 versus control (CTL). Rel, relative values. For current recordings, the voltage protocol shown in the inset of B is used throughout the study, pulse interval was 2 s ∗∗p < 0.01 versus CTL for current amplitudes upon >10 mV depolarizing steps. For western blots, data are presented as box plots and mean ± SD. For patch clamp, data are presented as mean + SD.

Figure 2

PKC inhibitors abolish PMA-induced reduction in Kv1.5 expression and I_Kv1.5in HEK293 cells.A, western blots result showing effects of 10 nM PMA treatment for 3 h on Kv1.5 expression with or without 10 μM BIM-1. Representative western blot images are displayed along with summarized data (n = 4). ∗∗p< 0.01 versus CTL; ^##p < 0.01 versus PMA. B, confocal images showing effects of 10 nM PMA treatment for 3 h on Kv1.5 expression with or without 10 μM BIM-1. Kv1.5 was labeled with an anti-Kv1.5 primary antibody and Alexa Fluor 488-conjugated (green) secondary antibody. The cell membrane was labeled using a Texas Red X-conjugated WGA (red) (n = 4). C, effects of 10 nM PMA treatment for 3 h on I_Kv1.5 with or without 10 μM BIM-1. Representative current traces are shown along with summarized I-V relationship. ∗∗p < 0.01 versus CTL; ^##p < 0.01 versus PMA+BIM-1 for current amplitudes upon >10 mV depolarizing steps. n = 6–13 cells. D, western blots image showing the effects of 200 nM Sotrastaurin or 200 nM Staurosporine on Kv1.5 expression of Kv1.5-HEK cells treated with or without 10 nM PMA for 3 h (n = 3). ∗∗p< 0.01 versus CTL; ^##p < 0.01 versus PMA. For western blots, data are presented as box plots and mean ± SD. For patch clamp, data are presented as mean + SD. Rel, relative values.

Figure 3

PMA treatment (10 nM) for 3 h has no effect on Kv1.1, Kv1.2, Kv1.3, and Kv1.4 channels in HEK293 cells.A, western blot images showing the effect of PMA (10 nM) for 3 h on expressions of Kv1.1, Kv1.2, Kv1.3, and Kv1.4 channels stably expressed in HEK293 cells (n = 3–4). B, I-V relationships showing the effect of PMA treatment for 3 h on currents of Kv1.1, Kv1.2, Kv1.3, and Kv1.4 channels stably expressed in HEK293 cells (n = 7–18 cells for each channel type). For western blots, data are presented as box plots and mean ± SD. For patch clamp, data are presented as mean + SD. C, sequence alignment of the N terminus of Kv channels. Amino acid residues marked in blue are identical or similar among all channels; amino acid resides marked in light blue are different among channels; amino acid resides marked in black are segments that are absent in some channels. Threonine (T) at amino acid 15 in the N terminus of Kv1.5 is marked as red.

Figure 4

Threonine at amino acid 15 in the N terminus of Kv1.5 is required for PMA-induced reduction in Kv1.5 expression and I_Kv1.5in HEK293 cells. For Kv1.5-WT (A), Kv1.5-ΔN209 (B), Kv1.5-Δ2-19 (C), and Kv1.5-T15A channels (D), schematic diagrams of channel structure are shown on the top row, western blot images are shown in the middle row, representative current traces along with summarized I-V relationships are shown in the bottom row. HEK293 cells stably expressing various channels were treated with 10 nM PMA for 3 h, and experiments were then performed. For western blots, n = 5 for WT, n = 3 for ΔN209, n = 3 for Δ2-19, n= 5 for T15A. ∗∗p < 0.01 versus CTL. For current recordings, n = 7–16 cells from three independent treatments for each channel. ∗∗p < 0.01 versus CTL for current amplitudes upon >10 mV depolarizing steps. For western blots, data are presented as box plots and mean ± SD. For patch clamp, data are presented as mean + SD.

Figure 5

PKC activation by PMA treatment induces monoubiquitination of Kv1.5 channels in HEK293 cells.A, effects of 10 nM PMA treatment on KCNA5 mRNA levels detected using real-time PCR (n = 4). B, the western blot image showing that PMA treatment (10 nM, 0.5 h) induces an extra band, close to 83 kDa (n = 4). C, overexpression of UbKO has no effect on the PMA-induced reduction in expression of the 75-kDa form of Kv1.5 channels (n = 6). ∗∗p < 0.01 versus CTL. Data are presented as box plots and mean ± SD. N.S., not significant; Rel, relative values.

Figure 6

Vps24 mediates PKC activation-induced Kv1.5 reduction in HEK293 cells.A, overexpression of Vps24 accelerates PMA (10 nM, 1 h)-induced Kv1.5 reduction. ∗p < 0.05, ∗∗p < 0.01 versus CTL (non-PMA treatment); ^#p < 0.05 versus PMA in empty pcDNA3 transfected cells (n = 5). B, knockdown of endogenous Vps24 using siRNA slows down PMA (10 nM, 3 h)-induced Kv1.5 reduction. ∗p < 0.05, ∗∗p < 0.01 versus CTL (non-PMA treatment), ^##p < 0.01 versus PMA in scrambled siRNA transfected cells (n = 6). Data are presented as box plots and mean ± SD. Rel, relative values.

Figure 7

Lysosomal inhibitor bafilomycin A1, but not proteasomal inhibitor MG132, completely abolishes PMA-induced reduction in Kv1.5 expression and I_Kv1.5in HEK293 cells.A, effects of 10 nM PMA (3 h) on Kv1.5 expression with or without bafilomycin A1 (Baf, 1 μM) or MG132 (10 μM). Representative western blot images are displayed along with summarized data. n = 3. ∗∗p< 0.01 versus CTL; ^#p < 0.05, ^##p< 0.01 versus PMA. B, effects of 10 nM PMA (3 h) on I_Kv1.5 with or without Baf (1 μM) or MG132 (10 μM). Representative current traces are displayed along with summarized I-V relationships (n = 7–26 cells). ∗∗p < 0.01 versus CTL; ^##p < 0.01 versus PMA, for current amplitudes upon >0 mV depolarizing steps. For western blots, data are presented as box plots and mean ± SD. For patch clamp, data are presented as mean + SD. Rel, relative values.

Figure 8

PKC activation by PMA treatment decreases currents of Kv1.5 channels expressed in neonatal rat ventricular myocytes. Representative current traces (A) and summarized current–voltage relationships (B) of I_Kv1.5 recorded from Kv1.5-transfected neonatal rat ventricular myocytes treated with PMA (10 nM) for 3 h in the absence and presence of BIM-1 (10 μM) or Baf (1 μM) are shown. ∗∗p < 0.01 versus CTL; ^##p < 0.01 versus PMA for current amplitudes upon >0 mV depolarizing steps. n = 9–12 cells in each group. Data are presented as mean + SD.

Figure 9

PKC activation by PMA treatment decreases I_Kv1.5in human iPSC-derived atrial cardiomyocytes.A, representative current traces and summarized current–voltage relationships of I_Kv1.5 before and after 0.1 mM 4-AP application. After currents were recorded in control conditions (before 4-AP), 0.1 mM 4-AP was applied to the bath solution for 2 min to achieve stead-state block, and currents were recorded in the same cell in the presence of 4-AP (after 4-AP). n = 5. ∗∗p < 0.01 versus before 4-AP for current amplitudes upon >0 mV depolarizing steps. B, representative current traces and summarized current–voltage relationships of I_Kv1.5 in control cells (n = 15 cells) and cells treated with 10 nM PMA for 3 h (n = 13 cells). After treatment, cells were transferred to the recording chamber superfused with the external solution. The voltage protocol is the same as shown in Figure 1B. ∗∗p < 0.01 versus CTL for current amplitudes upon >0 mV depolarizing steps. For A, a two-tailed paired Student's t-test was used. For B, a two-tailed unpaired Student's t-test was used. Data are presented as mean + SD.

Figure 10

Schematic illustrating the process of PKC activation-induced endocytic degradation of Kv1.5 channel. PKC activation targets T15 at the N terminus of Kv1.5 channel on the surface membrane and induces monoubiquitination of the channel. Ubiquitinated channel undergoes internalization through MVB/Vps24 into lysosome to degradation.