Kv1.3 voltage-gated potassium channels link cellular respiration to proliferation through a non-conducting mechanism

Cellular energy metabolism is fundamental for all biological functions. Cellular proliferation requires extensive metabolic reprogramming and has a high energy demand. The Kv1.3 voltage-gated potassium channel drives cellular proliferation. Kv1.3 channels localise to mitochondria. Using high-resolution respirometry, we show Kv1.3 channels increase oxidative phosphorylation, independently of redox balance, mitochondrial membrane potential or calcium signalling. Kv1.3-induced respiration increased reactive oxygen species production. Reducing reactive oxygen concentrations inhibited Kv1.3-induced proliferation. Selective Kv1.3 mutation identified that channel-induced respiration required an intact voltage sensor and C-terminal ERK1/2 phosphorylation site, but is channel pore independent. We show Kv1.3 channels regulate respiration through a non-conducting mechanism to generate reactive oxygen species which drive proliferation. This study identifies a Kv1.3-mediated mechanism underlying the metabolic regulation of proliferation, which may provide a therapeutic target for diseases characterised by dysfunctional proliferation and cell growth.


Introduction
Cellular energy metabolism is key to all biological processes. Biosynthesis of the macromolecular components of cells requires energy, generated in the mitochondria, in the form of ATP. Therefore redox regulation, and balancing the metabolic needs of oxidative phosphorylation (OXPHOS) required for ATP production and the generation of new lipids and nuclear material, is essential for cell proliferation. The mechanisms through which cells regulate their metabolism to drive cell growth and proliferation is not fully understood. These processes involve integrated environmental sensing and intracellular signalling, coupling the initiators of proliferation with the regulation of energy metabolism 1,2 . Here, we propose a role for a voltage-gated ion channel in the process linking proliferation and energy metabolism.
Voltage-gated potassium (Kv) channels are transmembrane proteins that facilitate K + movement through membranes via their intrinsic pores. K + influx into a cell or organelle activates signalling cascades that can regulate proliferation 3 , cell volume 4 , apoptosis, migration 5 and energetics 6,7 . The Kv1.3 channel is present in a widerange of tissues including the brain 8,9 , epithelial cells 10 , adipose tissue 6 and both skeletal and smooth muscle cells 11,12 . Kv1.3 channels stimulate cellular proliferation 3,13-15 with mechanisms implicated including ion conducting effects on plasma membrane potential 16 , Ca 2+ influx 15 and cell volume regulation 17 . Non-conducting signalling mechanisms, independent of ion influx/efflux, are emerging. These arise due to interactions between the channels and signalling/scaffolding proteins that either alter the channel's cellular location or couple the channel to intracellular second messengers 18,19 .
Cellular proliferation has a high energy demand, requiring ATP for processes ranging from signal transduction to DNA, RNA and protein synthesis 20 . Besides the plasma membrane, Kv1.3 has been reported at intracellular sites, including the nucleus 21 and the ER 22 . Plasma membrane and intracellular Kv1.3 channels may have distinct roles in cellular homoeostasis 18,23,24 . Kv1.3 has also been found in the inner mitochondrial membrane, the site of ATP synthesis 23,25,26 .
We hypothesised that the Kv1.3 channel may coordinate the energy demands required for Kv1.3-stimulated proliferation. We identify that intracellular Kv1.3 channels function through a non-conducting mechanism to regulate cellular energetics and demonstrate that regulation of respiration by the channel is a stimulatory mechanism for Kv1.3-mediated proliferation.

Materials and Methods
Cell culture and generation of Kv1. 3
Citrate synthase assay Citrate synthase assay followed described methods 34 .

Statistical analysis
Sample sizes were calculated using power calculations. Where sample sizes were not calculated sample size was considered adequate based on the size and reproducibility of between group differences. Samples were randomly assigned to experimental groups. Group variance was analysed with an F-test. Data were analysed using either a paired, or unpaired two-tailed Student's t-test and/or oneway and two-way ANOVA with a Sidak's or Tukey's post hoc test. Significance was determined when p ≤ 0.05.
In contrast to Kv1.3, expression of the voltage-gated K + channel Kv1.5 reduces proliferation 35 . Respirometry was used to determine whether Kv1.5 also has contrasting effects on respiration in HEK293 cells expressing Kv1.5. HEK293/Kv1.5 cells exhibited reduced proliferation (Fig. 1F) and had lower Routine/basal and Leak respiration compared to control HEK293 or HEK293Kv1.3 cells, independent of ROX (Fig. 1G, H). The opposing effects on proliferation by Kv1.3 and Kv1.5 channel expression are mirrored by effects on mitochondrial respiration.
Kv1.3 channels localise to the mitochondria but do not increase cellular mitochondrial content The Kv1.3-induced increase in mitochondrial respiration was not activated by K+ transport through plasma membrane Kv1.3 channels. Kv1.3 has been reported to localise to the mitochondrial membrane 23,25,26 . We investigated whether Kv1.3 channels localise to the mitochondria in our model. Immunocytochemistry was employed to investigate Kv1.3 cellular localisation in control HEK293 and HEK293/Kv1.3 cells. Cells were stained for Kv1.3 (anti-Kv1.3 antibody and fluorescent secondary reporter; green), and mitochondria (MitoTracker CMXRos; red) and imaged using confocal microscopy (Fig. 3A). HEK293 cells do not exhibit autofluorescence within the MitoTracker detection wavelengths in cells not stained with MitoTracker ( Supplementary Fig. S6A, B). As a secondary approach, HEK293 cells expressing mCherry-tagged Kv1.3 channels (HEK293/Kv1.3-P118; green) were immunofluorescently stained for mitochondrial complex IV (red) (Fig. 3A). In both cases Kv1.3 and mitochondria colocalisation was visualised in yellow and observed using a mitochondrial and Kv1.3 florescence intensity plotted line profile through a cross section of the images. HEK293/ Kv1.3 cells had greater Kv1.3 staining. The Kv1.3 channels were observed to co-localise with mitochondria. Interestingly, Kv1.3 expression is not ubiquitously distributed among mitochondria. Consistent with findings in the literature, our data suggest that the majority of the Kv1.3 channel protein is intracellular 13 .
Next we investigated whether increased respiration in Kv1.3-expressing cells was driven by increased mitochondrial content. Canonical biochemical markers of mitochondrial content, complex IV activity using respirometry (Fig. 3B) and citrate synthase activity (Fig. 3C) were determined in control HEK293 and HEK293/Kv1.3 cells. Expression of the mitochondrial protein ATP synthase was also examined using immunoblotting in HEK293/Kv.13 cells (Fig. 3D and Supplementary Fig. S6C). Kv1.3 expression did not increase cellular mitochondrial content.
We reasoned that Kv1.3-induced increases in mitochondrial respiration may occur via alterations to the mitochondrial network. Confocal microscopy was used to investigate the mitochondrial network of HEK293/ Kv1.3 cells. Cells were either stained with Mitotracker (red) (Fig. 3E) or immunostained for mitochondrial complex IV (red) (Fig. 3F). The gross structure of the mitochondrial network was unchanged by Kv1.3 expression. Kv1.3 does not exert its effects on cellular control HEK293 (n = 8) cells measured by high-resolution respirometry. Data expressed as the mean ± SEM. Data were analysed using either Student's t-test or two-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. respiration through mitochondrial biogenesis or network reorganisation.
The effect of mitochondrially-targeted Kv1.3 channel inhibition on respiration rate was examined in HEK293/ Kv1.3 cells using PAPTP (Fig. 4B). PAPTP reduced Routine and maximal ETS respiration, whilst increasing proton Leak respiration. PAPTP also reduced both ATP turnover rate and spare respiratory capacity in HEK293/ Kv1.3 cells (Fig. 4C, D). PAPTP did not affect proliferation, or Routine, maximal ETS and Leak respiration in control HEK293 cells ( Supplementary Fig. S9), indicating that PAPTP's effects in HEK293/Kv1.3 cells are not due to off-target activity. These data suggest that mitochondrial Kv1.3 channels may regulate proliferation and respiration.

Reactive oxygen species generation links Kv1.3-induced respiration to cellular proliferation
Mitochondria are the primary site of cellular reactive oxygen species (ROS) production. Kv1.3 expression increases mitochondrial respiration. HEK293/Kv1.3 cells displayed greater ROS concentrations than control HEK293 cells, determined using the cell permeable fluorescent ROS dye CellROX Deep Red (Fig. 5A, B). Menadione induces cellular ROS generation through redox cycling. ROS production in response to 100 μM menadione was greater in HEK293/Kv1.3 cells than in control HEK293 cells (Fig. 5C). These data suggest that increased mitochondrial respiration in HEK293/Kv1.3 cells leads to a greater capacity for ROS production. ROS can enhance cellular proliferation 39 . We investigated whether mitochondrial ROS in HEK293/Kv1.3 cells was driving the proliferative phenotype. MitoQ, a mitochondrially-targeted ROS scavenger, had no effect on cell viability (Fig. 5D), and significantly reduced proliferation in HEK293/Kv1.3 cells without affecting proliferation in control HEK293 cells (Fig. 5E). The increased respiration in Kv1.3 expressing cells increases capacity for ROS generation, which in turn drives the proliferative phenotype.

Kv1.3 regulates respiration independently of mitochondrial membrane potential, Ca 2+ or NADH redox status
We explored the mechanisms through which Kv1.3 regulates respiration. Mitochondrial ion channels can regulate the MMP to control respiration 40 . The MMP of HEK293/Kv1.3 and control HEK293 cells was investigated using the Cairn Photometry system. TMRM is a cellpermeant dye attracted to the negative charge of the mitochondrial membrane. TMRM accumulation is greater in HEK293/Kv1.3 cells, indicating greater MMP hyperpolarisation (Supplementary Fig. S10A). HEK293/ Kv1.3 and control HEK293 cells were sorted by flow cytometry following either treatment with TMRM, or TMRM and the OXPHOS uncoupler FCCP. Flow cytometry confirmed Kv1.3-induced increases in HEK293 cells MMP (Supplementary Fig. S10B). FCCP dissipates the MMP releasing mitochondrial TMRM. There was a greater difference in the mean fluorescent intensity  ). E Control HEK293 and HEK293/Kv1.3 cell counts following 3 days of proliferation, normalised to control, with and without 5 μM MitoQ (n = 9). Data were expressed as mean ± SEM and was analysed using either a Student's t-test or two-Way ANOVA with Tukey's post hoc test, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. between TMRM and TMRM with FCCP treated HEK293/ Kv1.3 cells compared with control HEK293 cells, indicating a more hyperpolarised MMP (Supplementary Fig.  S10C). To investigate the role of Kv1.3 channels in the MMP phenotype, the mitochondrially-targeted Kv1.3 inhibitor PAPTP was added during TMRM incubation in both HEK293/Kv1.3 and control HEK293 cells. PAPTP did not affect the MMP, suggesting that the increased MMP observed in HEK293/Kv1.3 cells is not directly linked to mitochondrial Kv1.3 or the proliferative phenotype ( Supplementary Fig. S10D).
Increased mitochondrial Ca 2+ concentrations can drive dehydrogenase enzyme activity enhancing NADH availability (the principal OXPHOS electron donor) and increasing respiration. Both Ca 2+ and NADH were measured in control HEK293 and HEK293/Kv1.3 cells as an indicator of mitochondrial redox status 41 . Cells were stained with the fluorescent Ca 2+ indicator Rhod2-AM and analysed by flow cytometry. Cellular and mitochondrial Ca 2+ levels were examined under two conditions; baseline Rhod2 fluorescence and Rhod2 fluorescence following addition of FCCP (20 µM). The mitochondrial Ca 2+ range can be estimated by examining the difference in Rhod2 fluorescence between baseline and FCCP treatment. Kv1.3 expression did not affect cellular Ca 2+ concentrations ( Supplementary Fig.  S10E) or mean mitochondrial Ca 2+ range in HEK293 cells ( Supplementary Fig. S10F). NADH concentrations measured by autofluorescence detection using flow cytometry were greater in HEK293/ Kv1.3 compared with control HEK293 cells (Supplementary Fig S10G). NADH autofluorescence was measured following 20 μM FCCP treatment to maximally oxidise the NAD + /NADH pool, and both 0.5 µM rotenone plus 2.5 µM Antimycin A treatment to maximally reduce the NAD + /NADH pool ( Supplementary Fig. S10G). NADH concentrations were higher in HEK293/Kv1.3 cells compared with control HEK293 cells in all conditions. The difference in fluorescence between the maximally oxidised and maximally reduced NAD + /NADH pools (NADH range) estimates total mitochondrial NADH and NAD +42 . HEK293/Kv1.3 cells had an increased NADH autofluorescence range compared to control HEK293 cells (Supplementary Fig. S10H). HEK293/Kv1.3 cells were treated with 100 nM PAPTP to inhibit Kv1.3 channels. PAPTP had no effect on NADH autofluorescence in HEK293/Kv1.3 cells (Supplementary Fig. S10I), suggesting that Kv1.3-induced increases in NADH do not mediate channel-enhanced respiration.

Kv1.3 channels regulate proliferation and respiration via a non-conducting mechanism
We interrogated the structural and functional properties of Kv1.3 that may regulate respiration. HEK293 cells were transfected to express mutant Kv1.3. Kv1.3-P89 is a voltage sensitive but non-conducting Kv1.3 channel with intact gating properties. It has a single point mutation (W389F) in the channel pore region (S5-S6) 13 . Kv1.3-P93 is a voltage insensitive and non-conducting Kv1.3 channel which shares the Kv1.3-P89 (W389F) point mutation and has three additional point mutations in the voltage sensor (S4) 13 . This triple mutation (R320N, L321A and R326I) shifts the channel activation to potentials below −170 mV. The channel is inactive at physiological voltages and functions as an inward rectifier 43 . Kv1.3 phosphorylation by intracellular kinases is a common posttranslational modification 44,45 . Kv1.3-P121 is voltage sensitive and conductance competent. However Kv1.3-P121 is phosphorylation-defective due to a Y447A substitution in its C-terminus which prevents phosphorylation of this site by the extracellular signal-regulated kinases 1/2 (ERK1/2) 13 .
We then determined the mechanism through which Kv1.3 channels regulate proliferation (Fig. 6B). Cells expressing Kv1.3 with a non-functioning pore (Kv1.3-P89) displayed increased proliferation when compared to control HEK293 cells. This suggests that ion conductance through the Kv1.3 channel is not essential for Kv1.3induced proliferation. Disruption of the Kv1.3 voltage sensor (Kv1.3-P93) ablated the channels effect on proliferation. HEK293/Kv1.3-P121 cells also did not exhibit increased proliferation compared to control HEK293 cells. These data indicate that Kv1.3-induced proliferation is regulated via a non-conducting mechanism requiring both an intact voltage sensor and the C-terminal ERK1/2 phosphorylation site.

Discussion
The Kv1.3 channel enhances cellular proliferation and has been implicated in diseases of pathological proliferation including cancer and cardiovascular disease 26,46 . Our data suggests Kv1.3 located in the mitochondria may be important to this phenotype. Proliferation is a highly energy demanding process requiring increases in ATP turnover. Kv1.3 channels induced an increase in cellular respiration and ATP turnover, independent of effects on mitochondrial content. However, Kv1.3 expression was not ubiquitously distributed among mitochondria. Whether the ion channel is enriched in a specific mitochondrial subpopulation remains to be determined.
Kv1.3-induced respiration had a direct mechanistic link to the proliferative phenotype beyond ATP generation. Kv1.3-driven OXPHOS increased the production of ROS which were required for proliferation. Using mutant Kv1.3 channels, we identified that Kv1.3-induced proliferation and respiration was independent of the channel pore. Mutation of the channel's voltage sensor ablated Kv1.3's effects on respiration and proliferation. Therefore the non-conducting process required to drive respiration remained linked to the channels' ability to sense the extracellular potential. The activity of intracellular ion channels can be regulated by kinase-mediated signalling cascades 13,44,45 . Several pro-proliferative growth factors, including platelet-derived growth factor, signal via ERK to induce their cellular effects 35 . Kv1.3 has been identified as an ERK1/2 substrate with a tyrosine 447 phosphorylation site in the C-terminal region 13 . ERK signalling is required for Kv1.3-mediated proliferation 13 . Mutation of the phosphorylation site ablated the Kv1.3-induced increase in respiration and proliferation. This study proposes a mechanism for Kv1.3-mediated cellular proliferation whereby Kv1.3 channels require phosphorylation by ERK1/2 and the capacity to sense the extracellular potential to increase respiration, driving ROS production and, subsequently, proliferation (Fig. 7). The reliance on both voltage sensing and ERK phosphorylation for the regulation of respiration is consistent with a model in which, upon voltage sensing, Kv1.3 undergoes a conformational change exposing the ERK phosphorylation site in the C-terminus 3,13,47 . This is also consistent with the observation that PAPTP, which locks the channel in the inactive state, inhibits Kv1.3-induced respiration and proliferation. Kv1.3-induced respiration is not dependent on K + ion conductance; therefore PAPTPs inhibitory effects may instead relate to preventing the conformational change in Kv1.3 that exposes the ERK phosphorylation site at residue 447. Fig. 7 Diagrammatic representation of the proposed mechanism for Kv1.3-induced respiration and proliferation. Kv1.3 channels potentially located at the mitochondria require voltage sensing and an intact ERK1/2 phosphorylation site (Y447) to stimulate mitochondrial oxidative phosphorylation. Kv1.3-induced oxidative phosphorylation drives increased ATP turnover, meeting the increased energy demands needed for proliferation. Simultaneously, Kv1.3induced oxidative phosphorylation generates mitochondrial reactive oxygen species (ROS) which stimulate the proliferative phenotype of the cells. This process is independent of plasma membrane Kv1.3 channel ion conductance. However, it may be speculated that growth factor receptors at the cell plasma membrane may signal to Kv1.3 channels via downstream ERK1/2-mediated phosphorylation.
We acknowledge limitations to our study. Our data using the mitochondrially-targeted Kv1.3 inhibitor PAPTP suggest that mitochondrial Kv1.3 channels may be important to the Kv1.3-induced proliferative and respiration phenotype. However, future studies confirming this association and linking the Kv1.3 organelle location to the non-conducting mechanism, potentially through mitochondrially-targeted versions of the channel pore, voltage sensor and ERK phosphorylation-defective Kv1.3 channels will be important.
We identify that Kv1.3 coordinates cellular respiration to meet the energy requirements needed for proliferation. We suggest a mechanism linking ERK1/2 signalling to Kv1.3-mediated increased respiration, ROS generation and proliferation. Our study has inferences for growth factor-mediated mechanisms that converge on ERK to drive physiological and pathological proliferation.