Chronic Propafenone Application Increases Functional KIR2.1 Expression In Vitro

Expression and activity of inwardly rectifying potassium (KIR) channels within the heart are strictly regulated. KIR channels have an important role in shaping cardiac action potentials, having a limited conductance at depolarized potentials but contributing to the final stage of repolarization and resting membrane stability. Impaired KIR2.1 function causes Andersen-Tawil Syndrome (ATS) and is associated with heart failure. Restoring KIR2.1 function by agonists of KIR2.1 (AgoKirs) would be beneficial. The class 1c antiarrhythmic drug propafenone is identified as an AgoKir; however, its long-term effects on KIR2.1 protein expression, subcellular localization, and function are unknown. Propafenone’s long-term effect on KIR2.1 expression and its underlying mechanisms in vitro were investigated. KIR2.1-carried currents were measured by single-cell patch-clamp electrophysiology. KIR2.1 protein expression levels were determined by Western blot analysis, whereas conventional immunofluorescence and advanced live-imaging microscopy were used to assess the subcellular localization of KIR2.1 proteins. Acute propafenone treatment at low concentrations supports the ability of propafenone to function as an AgoKir without disturbing KIR2.1 protein handling. Chronic propafenone treatment (at 25–100 times higher concentrations than in the acute treatment) increases KIR2.1 protein expression and KIR2.1 current densities in vitro, which are potentially associated with pre-lysosomal trafficking inhibition.


Introduction
Inward rectification was first detected in 1949 by Bernard Katz [1]. At first, it was called "anomalous rectification" to distinguish it from voltage-gated K (Kv) channel current in squid giant axons [2,3]. Since then, significant progress has been made in this area of research. These developments include cloning the members of inward rectifier channels, discovering the molecular mechanism of inward rectification, implementing genetic studies in experimental animals, constructing mutations in inward rectifier genes, and so on [4][5][6][7][8][9][10]. The K IR 2.1 channel protein is encoded by KCNJ2, and K IR 2.1 current is a major part of the inward rectifying potassium current (IK 1 ) in cardiomyocytes, as no detectable current was observed in ventricular myocytes from K IR 2.1 knockout mice [2,11,12]. IK 1 generated by K IR 2.x channels (K IR 2.1, K IR 2.2, and K IR 2.3 homo-and heterotetramers) is responsible for controlling the resting membrane potential and accelerating the final repolarization phase in cardiomyocytes [2,13]. Therefore, regulating the IK 1 current can greatly affect the excitability and arrhythmogenesis of cardiomyocytes [13,14]. Recent findings and the development of new compounds may yield new pharmacological IK 1 modulators [10,[15][16][17].
The drug propafenone-an orally active sodium channel blocking agent-was identified as an AgoKir [29,30]. It is currently used as a class Ic antiarrhythmic drug [31,32]. Previous experiments revealed that low concentrations (0.5-1 µM) of propafenone acutely increase IK 1 by binding with Cys311 of the K IR 2.1 protein [29][30][31]33]. Since the effect of propafenone on IK 1 generated by K IR 2.1 is acute, it implies that chronic treatment will be needed in patients to obtain a permanent benefit. It has been established that drugs influencing ion channel activity directly could also alter ion channel expression and function in the long-term [34,35]. However, the long-term influence of propafenone on K IR 2.1 channels is unknown. Therefore, we explored whether propafenone affects K IR 2.1 expression, its subcellular localization, and its functional consequence in the long-term.
In the current work, we confirm that propafenone can act as an AgoKir. We explored the long-term effects of propafenone on IK 1 and K IR 2.1, but also its close homologue K IR 2.2, channel expressions. Live imaging was also performed to determine the subcellular localization of K IR 2.1 proteins. As K IR 2.1 channel proteins degrade via the lysosomal pathway [36], the half-life of the protein in the presence of propafenone was determined [36][37][38]. This study provides a basis for further research on propafenone-based AgoKirs, thus contributing to the development of a therapy for diseases in which the function of K IR 2.1 is reduced.

As an AgoKir, Propafenone Can Increase Both Channel Expression and IK 1 Density
To confirm that propafenone could increase K IR 2.1 channel generated IK 1, the acute effects of low doses of propafenone on IK 1 were investigated by whole-cell patch-clamping in HEK-KWGF cells. Analyses were performed separately for each measured voltage point. Statistical analysis showed that the outward component of (−60 to −20 mV) IK 1 was increased, especially at −40 mV, when compared with the control upon perfusion with 1 µM propafenone ( Figure S1). Whereas, at a concentration of 25 µM, both the inward and outward components are decreased ( Figure S1). Similar results had been found previously [30].
Next, the long-term effects of propafenone on IK 1 and K IR 2.1 and K IR 2.2 protein expression were explored. Figure 1A,B show that propafenone can increase the expression levels of K IR 2.1 and K IR 2.2 proteins dose-dependently. To see whether the increased protein levels observed by western blot would also result in increased expression levels on the membrane and K IR 2.1-carried current, we treated HEK-KWGF cells with 50 µM propafenone for 24 h and analyzed them immediately for IK 1 current densities in the absence of propafenone ( Figure 1C). Long-term treatment of the cells with propafenone significantly increased both the inward component of IK 1 at membrane potentials between −120 and −100 mV and the outward component at membrane potentials between −70 and 30 mV.

Propafenone Specifically Works on KIR2.1 Channels and Shows a Long Residence Time
After determining the long-term effect of propafenone on KIR2.1 channel expression, we next investigated hERG (Kv11.1) and sodium channels (Nav1.5) to test if propafenone has a similar effect on other channel protein types over the same time period. HEK-hERG cells and HEK-Nav1.5 cells were used. As the HEK-Nav1.5 cell line stably expresses both Nav1.5 and KIR2.1 channels, we determined Nav1.5 protein expression alongside KIR2.1 expression levels from the same samples. As shown in Figure 2A,B, the expression levels of Nav1.5 and Kv11.1 channel proteins did not change in response to propafenone treatment. The expression level of KIR2.1 protein was increased similarly, as shown in Figure 1A, which indicates that propafenone can specifically work on KIR channels, even in the presence of another ectopically expressed ion channel (i.e., Nav1.5).

Propafenone Specifically Works on K IR 2.1 Channels and Shows a Long Residence Time
After determining the long-term effect of propafenone on K IR 2.1 channel expression, we next investigated hERG (K v 11.1) and sodium channels (Na v 1.5) to test if propafenone has a similar effect on other channel protein types over the same time period. HEK-hERG cells and HEK-Na v 1.5 cells were used. As the HEK-Na v 1.5 cell line stably expresses both Na v 1.5 and K IR 2.1 channels, we determined Na v 1.5 protein expression alongside K IR 2.1 expression levels from the same samples. As shown in Figure 2A,B, the expression levels of Na v 1.5 and K v 11.1 channel proteins did not change in response to propafenone treatment. The expression level of K IR 2.1 protein was increased similarly, as shown in Figure 1A, which indicates that propafenone can specifically work on K IR channels, even in the presence of another ectopically expressed ion channel (i.e., Na v 1.5).
To determine the retention time of propafenone's effect on K IR 2.1 expression levels following drug removal, a washout experiment was performed. The expression levels of K IR 2.1 proteins decreased significantly after washout, showing propafenone's washout effect ( Figure 3A,B). However, for the cells that received treatment with 50 µM propafenone, the expression of K IR 2.1 remained high after washout for 24 h ( Figure 3A [39]. Nav1.5 is a voltage-activated sodium channel expressed as a single band (~250 kDa) in the HEK-Nav1.5 cells [40]. Cells were treated with different concentrations of propafenone (1, 10, 25, and 50 µM) for 24 h (n = 5). Non-transfected cells (NT) were used as a negative control. (B) Summarized results of (A) (n = 5). The control protein level was set at 1 after correction for loading. Ponceau staining was used as a loading control. *** p < 0.001 vs. control.
To determine the retention time of propafenone's effect on KIR2.1 expression levels following drug removal, a washout experiment was performed. The expression levels of KIR2.1 proteins decreased significantly after washout, showing propafenone's washout effect ( Figure 3A,B). However, for the cells that received treatment with 50 µM propafenone, the expression of KIR2.1 remained high after washout for 24 h ( Figure 3A,B). This long residence time (24 h) indicates the persistence of propafenone's chronic effect. Propafenone specificity for K IR channels. (A) Western blot analysis of hERG channel protein, sodium channel protein, and K IR 2.1 channel protein expression levels. K v 11.1 is a voltage-activated potassium channel expressed as a core N-glycosylated immature form (~135 kDa) and a fully Nglycosylated mature form (~155 kDa) in HEK-hERG cells [39]. Na v 1.5 is a voltage-activated sodium channel expressed as a single band (~250 kDa) in the HEK-Na v 1.5 cells [40]. Cells were treated with different concentrations of propafenone (1, 10, 25, and 50 µM) for 24 h (n = 5). Non-transfected cells (NT) were used as a negative control. (B) Summarized results of (A) (n = 5). The control protein level was set at 1 after correction for loading. Ponceau staining was used as a loading control. *** p < 0.001 vs. control. . The control protein level was set at 1 after correction for loading. Ponceau staining was used as a loading control. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. control; # p < 0.05, ### p < 0.001 vs. propafenone-25 µM group; † p < 0.05, † † † p < 0.001 vs. propafenone 50 µM group.

Channel Function, Polyamine Binding Sites, and the Drug-Channel Interaction Location Do Not Interfere with the Long-Term Effect of Propafenone on KIR2.1 Expression
As IK1 current plays an important role in regulating various physiological processes, it is essential to know whether channel functions were involved in the long-term effect of propafenone. We used a KIR2.1-AAA non-conducting channel protein. In addition, BaCl2 was used as a channel-blocking agent. It demonstrated that the expression of KIR2.1 proteins was increased similarly as compared to WT at higher concentrations of propafenone (50 µM), while IK1 was inactivated by the KIR2.1-AAA mutation or blocked by BaCl2 (Figure 4A,B). Since polyamine binding sites (D172, E299, and E244) play important roles in the inward rectification of KIR2.1 channels and direct drug-channel interference is indispensable in the acute reactions of propafenone [2,29], we further investigated the roles of these factors. Western blots showed similar effects for these mutant proteins when compared with WT ( Figure 5). . The control protein level was set at 1 after correction for loading. Ponceau staining was used as a loading control. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. control; # p < 0.05, ### p < 0.001 vs. propafenone-25 µM group; † p < 0.05, † † † p < 0.001 vs. propafenone 50 µM group.

Channel Function, Polyamine Binding Sites, and the Drug-Channel Interaction Location Do Not Interfere with the Long-Term Effect of Propafenone on K IR 2.1 Expression
As IK 1 current plays an important role in regulating various physiological processes, it is essential to know whether channel functions were involved in the long-term effect of propafenone. We used a K IR 2.1-AAA non-conducting channel protein. In addition, BaCl 2 was used as a channel-blocking agent. It demonstrated that the expression of K IR 2.1 proteins was increased similarly as compared to WT at higher concentrations of propafenone (50 µM), while IK 1 was inactivated by the K IR 2.1-AAA mutation or blocked by BaCl 2 ( Figure 4A,B). Since polyamine binding sites (D172, E299, and E244) play important roles in the inward rectification of K IR 2.1 channels and direct drug-channel interference is indispensable in the acute reactions of propafenone [2,29], we further investigated the roles of these factors. Western blots showed similar effects for these mutant proteins when compared with WT ( Figure 5).   Live imaging showed that propafenone and chloroquine (CQ) can both cause intracellular K IR 2.1 accumulation but in a different pattern ( Figure 6A). CQ causes scattered clusters of K IR 2.1 on the edge of the cell, while propafenone causes brighter clusters both in the edge and center of the cell at 50 µM, but not at 10 µM. As shown before, CQ increases the expression of K IR 2.1 protein by inhibiting its degradation via the lysosomal pathway [36]. The different appearance of clusters in response to propafenone may point to a different mechanism for cell trafficking in comparison to CQ.
CHO-KD cells were used to investigate whether propafenone's long-term effect is late endosome/lysosome-related. K IR 2.1-Dendra2 is present in round clusters in the interior of the cells, which accumulate inside cells upon administration of 25 µM propafenone (3-48 h) ( Figure 6B). Images of DMSO-treated cells ( Figure 6B) show that K IR 2.1-Dendra2 is present in the plasma membrane and the cell's interior as many little clusters. Time-lapse imaging (Video S1, CHO-KD cells treated with DMSO) shows that these clusters move fast in all directions. K IR 2.1-Dendra2 moves slower after being treated with propafenone, and big clusters appeared in the cells' interiors (Video S1). The number of small clusters decreased (Video S1, CHO-KD cells treated with 25 µM propafenone for 3 h, 6 h, 24 h, and 48 h), and more and larger clusters became visible in the cells. Video S1 shows that the larger clusters of K IR 2.1-Dendra2 display less movement, while the remaining small clusters move faster. For the cells treated for 48 h with propafenone, multivesicular bodies (MVBs) appeared among the clusters of proteins ( Figure 6B, enlarged picture), indicating that K IR 2.1-Dendra2 might accumulate in late endosome-like structures. Similarly, as seen for protein expression levels ( Figure 5), intracellular K IR 2.1 protein accumulation was independent of D172H, D172R, E244A, E299A, or R312H mutations ( Figure 6C).
MVBs will deliver cargo destined for degradation to the lysosome [41,42]. Therefore, we tested the half-life of the K IR 2.1 proteins in HEK-KWGF cells in the presence of the translation inhibitor cycloheximide (CHX) (200 µg/mL). In the CQ treated group, the T 1/2 was significantly increased compared to the control (T 1/2 = 9.494 h vs. 4.774 h, Figure 7C,D). In contrast, no significant difference in T 1/2 was found following propafenone treatment (T 1/2 of 4.774 h in the control group vs. 5.247 h after propafenone treatment) ( Figure 7A,B). Thus, propafenone does not impair the degradation of K IR 2.1 proteins in lysosomes.  MVBs will deliver cargo destined for degradation to the lysosome [41,42]. Therefore, we tested the half-life of the KIR2.1 proteins in HEK-KWGF cells in the presence of the translation inhibitor cycloheximide (CHX) (200 µg/mL). In the CQ treated group, the T1/2 was significantly increased compared to the control (T1/2 = 9.494 h vs. 4.774 h, Figure 7C,D). In contrast, no significant difference in T1/2 was found following propafenone treatment (T1/2 of 4.774 h in the control group vs. 5.247 h after propafenone treatment) ( Figure 7A,B). Thus, propafenone does not impair the degradation of KIR2.1 proteins in lysosomes.

Discussion
We confirmed that acute administration of propafenone at low concentrations increases KIR2.1 currents in HEK-KWGF cells, which is similar to the results obtained in CHO cells transiently transfected with WT KIR2.1 [30]. Therefore, propafenone was shown to act as a KIR2.1 agonist, which we named "Agokir" [29]. However, since the effect of propafenone on KIR2.1 carried current is acute, chronic treatment will be required for longterm IK1 enhancement. Therefore, we investigated propafenone's long-term effect on KIR2.1 channels.
Propafenone is commonly administered in the clinic to treat atrial fibrillation because of its sodium channel blocking activity [43,44]. This sodium channel blocking property

Discussion
We confirmed that acute administration of propafenone at low concentrations increases K IR 2.1 currents in HEK-KWGF cells, which is similar to the results obtained in CHO cells transiently transfected with WT K IR 2.1 [30]. Therefore, propafenone was shown to act as a K IR 2.1 agonist, which we named "Agokir" [29]. However, since the effect of propafenone on K IR 2.1 carried current is acute, chronic treatment will be required for long-term IK 1 enhancement. Therefore, we investigated propafenone's long-term effect on K IR 2.1 channels.
Propafenone is commonly administered in the clinic to treat atrial fibrillation because of its sodium channel blocking activity [43,44]. This sodium channel blocking property results in a markedly depressed depolarization phase of the action potential and a widening of the QRS complex [45][46][47]. Some studies also showed that QRS duration was increased with or without QT interval prolongation in humans after treatment with propafenone [48][49][50][51][52]. Therapeutic plasma levels of propafenone in humans were estimated to range from 0.53 to 5.28 µM [33,53,54]. Furthermore, propafenone concentrations in human atrial tissues were on average ten times higher than those found in the plasma [53]. Such concentrations approach or are even similar to the concentrations found in our work, in which the chronic effect on K IR 2.1 protein results in an increase in expression levels.
In the present work, propafenone increases the outward component acutely at low concentrations (0.5 and 1 µM). Such an increase was shown to be achieved by the propafenone-K IR 2.1 channel binding-mediated decrease of channel affinity for polyamines and thus current rectification [30]. In contrast, propafenone at higher concentrations (25 and 50 µM) shows a strong acute blocking effect on both the inward and outward components. This block is caused by a propafenone-mediated decrease of the negative charge of the channel pore and channel affinity for phosphatidylinositol 4,5-bisphosphate (PIP 2 ), which is a lipid critical for channel activation [12,43]. For the long-term effect, however, propafenone treatment at high concentrations results in a significant increase in IK 1 densities. These latter electrophysiological measurements were performed in the absence of propafenone, thereby excluding its acute effect on K IR 2.1 channels. This significant increase likely occurred because propafenone inhibits channel backward trafficking, thereby indirectly increasing the K IR 2.1 channel expression level on the cell membrane.
In order to test propafenone's specificity for increasing K IR 2.1 channel expression, we investigated its effects on hERG channel expression, which is a voltage-gated potassium channel and thereby different from the non-voltage-gated K IR 2.1 channel. Sodium channel expression was also tested because propafenone is used as its blocking agent in the clinic [33,55]. Propafenone did not interfere with the expression levels of these two channels, revealing that propafenone has at least some specificity towards K IR channels. Some studies showed that there is reciprocal regulation between Na v 1.5 and K IR 2.1 channels [56][57][58][59][60][61][62]. K IR 2.1 overexpression increases the expression levels of Na v 1.5 in mouse hearts [58]. As propafenone increases the expression level of K IR 2.1 proteins significantly but does not interfere with that of Na v 1.5 in the same cells and with the same treatment in our study. We may thus hypothesize that propafenone might either affect cooperation between the two channels or act on a part of the trafficking pathway in which both channels do not cooperate. The previous study also showed that Na v 1.5 protein reduces K IR 2.1 protein internalization and promote its presence at the cell surface [58]. As the expression level of Na v 1.5 was not changed, the increased expression of K IR 2.1 protein was only affected by propafenone.
Ba 2+ is an efficient IK 1 blocking ion [10,63]. The atomic radius of Ba 2+ is close to that of K + ; therefore, it will fit into the K + selectivity filter and remain in that position due to its larger charge, effectively blocking the inward and outward K + flow [64][65][66]. At the same time, we also investigated a non-conducting K IR 2.1-AAA mutant to test the influence of channel functions on the long-term effect of propafenone. The results showed that the expression of K IR 2.1 channel proteins displayed no significant differences when compared with WT or not inhibited channels, which suggests that channel function (i.e., K + conduction) is not involved in the long-term effect of propafenone.
Polyamines, responsible for naturally occurring inward rectification, occupy two positions in the K IR 2.x channels: the cytoplasmic pore domain at K IR 2.1 E224 and E299 and the transmembrane pore domain at D172 [2,22]. Western blotting pointed out that propafenone dose-dependently increases WT, E224A, E299A, D172H, and D172R-K IR 2.1 protein expression in HEK-293 cells. Therefore, polyamine binding sites, and most likely polyamine binding too, appear not to be involved in the propafenone-mediated increase in K IR 2.1 expression levels.
Dynamics simulations predicted that propafenone interacts with K IR 2.1 by forming a hydrogen bond with the cysteine residue C311, which is identified as a direct channeldrug binding site [29,30,67,68]. Because of the proximity of C311 to the R312 residue [69], it is possible that the mutation R312H allosterically modifies the 310-QCRSSY-315 C-terminus domain, thereby precluding propafenone channel interaction. Increased K IR 2.1-R312H expression showed a similar result as WT, revealing that drug-channel interaction is most likely not involved in the chronic response to propafenone. In conclusion, propafenone specifically works on K IR 2 channels, but neither K + conduction, polyamine binding sites, nor direct drug-channel interactions are involved in the long-term effects of propafenone.
HEK-KWGF cells showed an intracellular accumulation of K IR 2.1 proteins after being treated with propafenone. The intracellular patterns, however ( Figure 6A), were distinct from cells treated with CQ, which is known to inhibit lysosomes. A potential explanation for this difference is that propafenone acts on both late endosomes and lysosomes, or on endosomes only. Live imaging on CHO-KD cells supported the observations reported above and revealed that after incubating 25 µM of propafenone for only 3 h, K IR 2.1-Dendra2 proteins started to accumulate in the cytoplasm ( Figure 6B). More and bigger clusters appeared at the following time points. The large clusters of K IR 2.1-Dendra2 proteins show less movement, which is in line with earlier research in which lysosomal diameter was increased using sucrose; enlarged lysosomes were correlated to a reduced diffusion rate [70]. Furthermore, MVBs were shown in the protein clusters after 48 h of incubation with propafenone, indicating that K IR 2.1-Dendra2 may accumulate within the late endosome compartment [71]. Protein also accumulates in late endosomes when lysosomes do not function well [71,72]. However, propafenone treatment did not increase T 1/2 as CQ, indicating that propafenone does not inhibit the function of lysosomes per se. From all these results, we conclude that chronic propafenone treatment increases K IR 2.1 protein expression and K IR 2.1 current densities in vitro following a 24 h treatment, which persists after washout and is potentially associated with pre-lysosomal trafficking inhibition. Moreover, additional proteins following a similar degradation route as K IR 2.1 could be affected similarly to propafenone.
Our data shows that propafenone can function as AgoKir at low concentrations without disturbing K IR 2.1 protein handling; however, it also shows long-term effects at higher concentrations. It is worthwhile to search for more potent propafenone analogs that should increase IK 1 without affecting K IR 2.1 channel expression, thus contributing to the development of new therapeutic avenues to address diseases related to dysfunctional K IR 2.1.

K IR 2.1-Mutant Expression Constructs and Transfection
K IR 2.1-E224A and E299A mutant constructs were obtained from Dr. Tristani-Firouzi (University of Utah School of Medicine, USA), and functional characteristics have been described previously by others [76] and us [34,77]. K IR 2.1-D172R and D172H mutant constructs were obtained from Dr. So (Seoul National University, College of Medicine, Republic of Korea), and functional characteristics have been described before [78]. A K IR 2.1-R312H mutant construct was obtained from Dr. Bendahhou (Université Côte d'Azur, France), and functional characteristics have been described recently [79]. Cell transfection was performed with linear polyethylenimine (PEI) with a molecular weight of 25,000 (Polysciences, Hirschberg an der Bergstrasse, Germany) as described previously by

Drugs
Chloroquine (CQ) (Sigma, St. Louis, MO, United States, cat. No. C6628) was dissolved in sterile water at a concentration of 10 mM and stored at −20 • C. Propafenone (Sigma, cat. No. C7698) was dissolved in DMSO at a concentration of 100 mM and stored at −20 • C until use. Cycloheximide (CHX, Sigma, cat. No. C7698) was dissolved in sterile water at a concentration of 5 mg/mL, stored, and aliquoted at −20 • C until use. All drugs were diluted on the day they were used.

Cloning
CHO-KD single-cell suspension was counted in a Brand™ Bürker counting chamber (Fisher Scientific, Landsmeer, The Netherlands). The cell suspension was diluted to obtain a concentration of 10 cells per mL, and cells were cultured in 96-well plates (100 µL/well). Cells were examined under a Nikon TMS inverted microscope (Nikon Instruments Europe B.V., Amsterdam, The Netherlands) after forming a single clone. The clones were expanded and then imaged by a Nikon Eclipse 80i epifluorescence microscope (Nikon Instruments Europe B.V.).

Live Imaging with Confocal Microscopy
HEK-KWGF cells were treated with propafenone (10, 50 µM) for 24 h, and then 488 nm laser light was used to visualize K IR 2.1-GFP. We cloned the existing CHO-KD cell line to create a pool of cells with high K IR 2.1-Dendra2 expression. CHO-KD cells were treated with propafenone at 25 µM for different time courses (3, 6, 24, and 48 h

Washout Experiment
HEK-KWGF cells were seeded in Ø 60 mm dishes overnight. 24 h after treatment with propafenone (25 or 50 µM), the medium of the cells was replaced by fresh supplemented DMEM. Protein lysates were harvested 24 or 48 h after changing the medium. The K IR 2.1 expression level was detected by Western blot.

Cycloheximide Assay
HEK-KWGF cells were seeded in Ø 35 mm dishes. After 24 h, cells were left untreated (control) or treated with 50 µM propafenone or 10 µM CQ for 24 h. 200 µg/mL cycloheximide was added during the last phases (2, 4, 6, 8, 10, or 12 h) of the 24 h treatment period. Cell lysates were prepared and processed as indicated in Section 4.4. Since propafenone and CQ increased the expression levels of K IR 2.1-GFP protein compared to control conditions before the start of CHX application, twice the amount of control lysate loaded on SDS-PAGE to enable visualization of the K IR 2.1 protein under control conditions at CHX t = 10 and t = 12 using similar ECL exposure times as for propafenone and CQ conditions.

Patch Clamp Electrophysiology
Whole-cell clamp measurements were performed using an AxoPatch 200B amplifier controlled by pClamp10.4 software (Molecular Devices, LLC, San Jose, CA, USA). The K IR 2.1 current in HEK-KWGF cells was measured at room temperature. Patch pipettes were made with a Sutter P-2000 puller (HEKA Elektronik, Lambrecht, Germany) and had resistances of 1-3 MΩ.

Statistics
Data are expressed as mean ± S.D. Differences between group averages were tested using a one-way ANOVA with a post-hoc test (Tukey's HSD), or an unpaired T-test. Data were considered significant when the p-value was less than 0.05. Statistical analysis was performed using GraphPad Prism version 9 (GraphPad Software, San Diego, CA, USA).