G protein-coupled and ATP-sensitive inwardly rectifying potassium ion channels are essential for HIV entry

The high genetic diversity of Human Immunodeficiency virus (HIV), has hindered the development of effective vaccines or antiviral drugs against it. Hence, there is a continuous need for identification of new antiviral targets. HIV exploits specific host proteins also known as HIV-dependency factors during its replication inside the cell. Potassium channels play a crucial role in the life cycle of several viruses by modulating ion homeostasis, cell signaling, cell cycle, and cell death. In this study, using pharmacological tools, we have identified that HIV utilizes distinct cellular potassium channels at various steps in its life cycle. Members of inwardly rectifying potassium (Kir) channel family, G protein-coupled (GIRK), and ATP-sensitive (KATP) are involved in HIV entry. Blocking these channels using specific inhibitors reduces HIV entry. Another member, Kir 1.1 plays a role post entry as inhibiting this channel inhibits virus production and release. These inhibitors are not toxic to the cells at the concentration used in the study. We have further identified the possible mechanism through which these potassium channels regulate HIV entry by using a slow-response potential-sensitive probe DIBAC4(3) and have observed that blocking these potassium channels inhibits membrane depolarization which then inhibits HIV entry and virus release as well. These results demonstrate for the first time, the important role of Kir channel members in HIV-1 infection and suggest that these K+ channels could serve as a safe therapeutic target for treatment of HIV/AIDS.

whose function is essential for their life cycle making them ideal drug targets 18,19 . The HIV viral protein Vpu displays K + channel activity to enhance virus release 20,21 . Additionally, the p13 protein encoded by HTLV-1 targets mitochondrial membrane potential that results in increased production of reactive oxygen species (ROS) by mitochondria 22,23 . The 6k protein encoded by Ross River virus forms a cation-selective ion channel which plays a role in its release 24,25 . Other viral proteins which display K + channel activity to regulate virus infection include the 6 K protein of Sindbis and Semliki forest virus 26 and Dengue virus C terminal peptide 27 . Modulation of potassium channels inhibits entry of Ebola virus 28 and replication of Bunyavirus 29 . The potassium channels are also utilized by viruses to control cell death pathways. For example, the Hepatitis C virus non-structural protein NS5A modulates the function of Kv2.1, a voltage-gated K + channel 27 and regulates cell apoptosis. HIV-1 protein Nef alters the intracellular K + ion concentration 30 by targeting large-conductance Ca 2+ -dependent K + channels (BK Ca ) 31 whereas viral Env protein, gp120 inhibits the voltage-gated K + channel (BEC1) activity resulting in decreased virus release 32 . HIV gp120 induces hippocampal neuronal apoptosis by enhancement of Kv channel functions through p38 MAPK phosphorylation in HIV associated neurocognitive disorder 33 .
In this study, we have systematically analyzed the role of K + channels in the entry, replication, and release of HIV-1 virus using pharmacological tools. We have observed that members of the K ir , GIRK and K ATP are involved in HIV entry whereas K ir 1.1 plays a role in the release of HIV. We have also elucidated the mechanism of action of these channels and observed that blocking these channels inhibited membrane depolarization which reduced HIV entry. We propose that potassium channels may be further explored as new, pharmacologically safe HIV therapeutics.

Results
potassium Channels in HIV production. HIV-1 induces an increase in intracellular K + concentration for efficient viral replication 34 . To determine this, we measured virus production in the presence of two potassium salts: KCl and K 2 SO 4 . MT4 cells were infected with NL4-3 (HIV-1) and incubated in the presence of increasing concentration of KCl or K 2 SO 4 for 48 hrs. Virus production was measured by detecting expression of the HIV-1 capsid protein, p24 by western blotting. An increase in salt concentration inhibited the virus production in a dose-dependent manner (Fig. 1a-d). Nearly 65-70% reduction in p24 was observed at 30 mM concentration of KCl or 20 mM K 2 SO 4 relative to β-actin. TEA, a broad-spectrum potassium channel blocker 35,36 , reduced virus production by 90% (Fig. 2a,b). These observations suggest that potassium channels may play a vital role in the life cycle of HIV.
potassium Channels in HIV entry. There are three major steps in the virus life cycle: virus entry, replication, and release. We proposed to investigate the role of potassium channels in modulating these three significant steps of HIV life cycle using specific inhibitors as mentioned in Table 1. 4-Aminopyridine (4-AP) blocks K V channel activity by delaying the burst time of the channel 37 . The activity of Kir 1.1 channels can be blocked by using BaCl 2. Ba + ions blocks cation movement by replacing the ion at the internal side of the central cavity of the www.nature.com/scientificreports www.nature.com/scientificreports/ channel 38,39 . BK channel activity can be inhibited by CoCl 2 . Cobalt inhibits the Ca ++ current which reduces the K + current through BK channels 38,39 . K ATP and GIRK can be targeted by Glibenclamide and Ifenprodil respectively. Glibenclamide blocks the channel activity from either side of the membrane 40 while Ifenprodil reversibly reduces the inward current by targeting GIRK from the extracellular side 40 .
We first examined the effect of family-specific K + channel blockers on HIV entry in the host cells. Each blocker was used at a concentration which was not toxic to the cells (data not shown). For our experiments, we have used the most conventional system, the TZMbl assay which is based on the expression of the luciferase gene under the HIV-1 long terminal repeat (HIV-LTR) promoter. These cells have been widely used to study HIV entry and infection cycle. The TZMbl cells were pre-incubated with the indicated concentration of each blocker for 1 hr followed by HIV infection in the presence of the blockers for additional 2 hrs. The cells were washed post infection and further incubated for 48 hrs in the absence of blockers. Virus production was estimated by measuring the luciferase gene expression. Increased K + ion concentration in media reduced virus entry by nearly 50%. The decrease in virus entry was not due to changes in osmolarity of the media as the addition of sucrose did not produce the same effect (Fig. 3a). TEA, a broad-spectrum K + channel blocker inhibited HIV entry by 50%. Interestingly, among all K + family blockers, only Ifenprodil, and glibenclamide reduced HIV entry by 50% which was reflected by decreased relative luminescence units (RLUs) (Fig. 3b). A similar decrease was observed even without pre-incubation with these compounds (Fig. 3c). No effect on virus entry was observed with other K + channel blockers, 4-AP, BaCl 2 , and CoCl 2 , indicating that these families are not involved in HIV entry. Taken together, these results suggest that K + channel subfamilies: K ATP and GIRK actively participate in HIV entry.  www.nature.com/scientificreports www.nature.com/scientificreports/ potassium Channels in HIV Replication. Next, we wanted to determine if potassium channels play a role at other stages of HIV life cycle post-entry. TZMbl cells were infected with the virus in the absence of various blockers. The cells were washed post infection and incubated in the presence of KCl, sucrose or specific blockers for 48 hrs. We observed a dose-dependent reduction in cell-associated luminescence after addition of KCl which confirmed that K + ions play a role post-entry of the virus. This effect is not due to changes in osmolarity as the addition of sucrose did not affect cell-associated luminescence (Fig. 4a). We further observed that addition of inhibitors 4-AP (K v channel), and BaCl 2 (K ir 1.1 channel) reduced virus production by 25% as evident from the decrease in cell-associated luminescence (Fig. 4b). This could be due to a reduction in Tat production due to the inhibition of virus replication inside the cells. Interestingly, other inhibitors such as CoCl 2 (B K channels), Ifenprodil (GIRK) and glibenclamide (K ATP ) did not inhibit virus replication when added post virus entry (Fig. 4b). Together, these results suggest that different K + channel families are involved at various stages of the HIV life cycle.
potassium Channels in HIV Release. Next, we were interested in determining if K + channels participate in HIV release. We transfected HEK293T cells with pNL4-3 DNA and incubated the cells post-transfection with various inhibitors for 24 hrs. Virus release was measured by detecting the expression of p24 in the culture supernatant by immunoblotting (Fig. 5a,b). We observed a 25% decrease in virus release in the presence of BaCl 2 indicating that Kir 1.1 family of potassium channels might be involved in the virus release. We observed a slight increase in cellular p24 with BaCl 2 which could either be due to a direct effect of BaCl 2 on increasing virus production or could be due to an accumulation of p24 in the cell as a result of a block in virus release. A slight decrease in virus release was also observed in the presence of CoCl 2 and 4-AP, but it was statistically insignificant. It is important to note that although TEA is a broad spectrum K + channel blocker, it is not highly sensitive for Inwardly rectifying K + channels as compared to BaCl 2 which is highly specific 38,39,41 Hence, we did not observe a similar effect on virus release with TEA. www.nature.com/scientificreports www.nature.com/scientificreports/ Membrane potential and HIV entry. We further investigated if membrane potential plays any role in modulating HIV entry, using slow-response potential-sensitive probe DIBAC4(3). This probe enters depolarized cells where it binds to intracellular proteins or membrane, and exhibits enhanced fluorescence on binding 40 . TZMbl cells were treated with valinomycin (a K + ionophore) before HIV infection. Valinomycin at 10 nM concentration impeded virus entry by nearly 50% (Fig. 6a) and depolarised the membrane as evident from increased fluorescence obtained with DIBAC4(3) (Fig. 6b,c). These results suggested that depolarization of membrane inhibits HIV entry.
Role of Potassium channel blockers in modulating membrane potential. Next, we hypothesized that potassium channel blockers might also reduce virus entry by increased depolarization of cells. To test this hypothesis, we have measured membrane depolarization in the presence of various potassium channel blockers. We have observed that the inhibitors Ifenprodil and Glibenclamide depolarised the cells by reducing K + efflux as evident from the increased fluorescence obtained with DIBAC4(3) (Fig. 7a,b) in the absence of HIV-1. Presence of Ifenprodil during HIV infection further enhanced fluorescence. These results suggested that K ATP and GIRK channel family may play a role in hyperpolarization of cells which is essential for virus entry. Similarly, presence of 4-AP also slightly increased membrane depolarisation suggesting its role during virus replication inside the cells (Fig. 7c,d). Presence of TEA and other K + family blockers had no effect on membrane potential. (Fig. 7c,d).

Discussion
Potassium channels play a key role in maintaining K + ion homeostasis and membrane potential and are involved in many cellular processes. These ion channels serve as an ideal therapeutic target in several diseases 14 . It has been previously reported that HIV virus production, in RH9 cells infected with HIV-1 strain LAI, decreases with  www.nature.com/scientificreports www.nature.com/scientificreports/ increasing KCl concentration in media suggesting that changes in extracellular potassium ions modulate HIV gene expression 34 , but the specific role of different potassium channel families during HIV life cycle is not known. In the current study, we have analyzed the role of specific K + channel families at different steps in the HIV life cycle using available pharmacological inhibitors.
We have observed that elevated concentration of both the K + salts KCl and K 2 SO 4 reduced HIV virus production (Fig. 1a-d). Addition of broad-spectrum K + channel blocker, TEA also diminished virus production suggesting that the effect of K + ions on virus production is specific and not due to changes in osmolarity which may also decrease cellular protein synthesis 42 . Hence, these results imply that HIV may utilize multiple K + channels during its life cycle (Fig. 2a,b).
The GIRK channels are present on the plasma membrane, and their opening is mediated by G protein-coupled receptors (GPCR) 43 . The GPCRs associated with GIRK are not involved in signal transduction pathways but directly activate these ion channels using effector proteins or the G protein subunits themselves. Several GPCRs are members of the chemokine receptor subfamily. The HIV coreceptor CCR5 is an example of GPCR which is involved in HIV entry 44 . Our study demonstrates that GIRK subfamily of potassium channels are involved in virus entry. Blocking these channels with Ifenprodil impeded HIV entry by 50% (Fig. 3b).
The K ATP channels are widely distributed and present in several tissues. Their activity is regulated by the intracellular pool of Adenine nucleotides linking cellular metabolism with membrane excitability. In the early steps of HIV infection, viral protein p2 increases the cellular ATP pool which facilitates reverse transcription 45 . Our results highlight that these channels play a significant role in HIV entry. Blocking the K ATP channel with glibenclamide also inhibits viral entry (Fig. 3b).
Changes in membrane potential regulate cell viability and cell cycle. Role of membrane potential in the fusion and entry of semliki forest virus has been demonstrated 15 . HIV Nef protein alters the membrane potential of the HIV infected cell. Overexpression of Nef hampers the activity of Ca +2 dependent K + channels 45 . Another HIV protein, Vpu, interacts with a cellular weak inward K + rectifier TASK1 and plays a crucial role in the enhancement of virus release from the host cell by causing membrane potential depolarization 46 . HIV envelop protein, gp120, induces cell death in neuronal cells by targeting K + channels [46][47][48] .
To understand the role of membrane potential in HIV entry, we used a fluorescent membrane potential probe DIBAC 4(3) and observed that a K + ionophore Valinomycin inhibited HIV entry by increasing membrane depolarization. This suggests that membrane hyperpolarisation is essential for HIV entry. We observed a similar increase in membrane depolarization in the presence of both Ifenprodil and Glibenclamide implying that these compounds may also block virus entry by depolarization of the membrane (Fig. 7a,b). This is the first report of a possible mechanism of action of Ifenprodil. Glibenclamide is already being used in the treatment of type 2 diabetes and works by causing membrane depolarization in pancreatic cells leading to release of insulin [47][48][49] . However, other mechanisms of action of these compounds through cell signaling or any other pathway needs to be further investigated.
The K V channels play a role in the progression of HIV associated dementia (HAD) 33,50 . Blocking the Kv channels with 4-AP inhibited intracellular virus production (Fig. 4b).
It is interesting to note that a non-specific K + channel blocker TEA, inhibited HIV production without altering membrane potential inferring that there might be other mechanism involved in this process also (Figs 3b and 7c,d).
Together, our study highlights the important role played by K ir channel families at different steps in the HIV life cycle and suggest these K ir channels may be pursued as new therapeutic targets for HIV infection.

Conclusion
To address the challenge of drug resistance, it is necessary to unravel new targets that could be developed as alternate HIV therapeutics. We have demonstrated that K ir potassium channels play a vital role at different steps of HIV life cycle and could serve as novel targets for the better management of HIV/AIDS.
Western Immunoblotting. MT4 cell were infected with NL4-3 for 48 hrs and lysed in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl pH 8.0, 150 mM sodium chloride, 1.0% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) containing 1× protease inhibitor cocktail (Roche, Germany). The cell lysate was resolved by running on a 12% SDS-polyacrylamide gel. The proteins were transferred to polyvinylidene difluoride (PVDF, Millipore, USA) membrane by electrophoretic transfer method and probed with human HIV serum (NIH AIDS Reagent Program cat: 3957) followed by labeling with anti-human HRP conjugated secondary antibodies (GE Healthcare, UK). The western blots were visualized with ECL (Pierce, USA) system followed by exposure to X-ray films. Bands on the X-ray films were analyzed with ImageJ software (http://imagej.nih.gov/ij/).