Rise and Fall of Kir2.2 Current by TLR4 Signaling in Human Monocytes: PKC-Dependent Trafficking and PI3K-Mediated PIP₂ Decrease

Monocytes and macrophages are the major components of innate immunity, mediating antimicrobial defense and inflammatory responses (1). They are activated by pathogen-associated molecules, for instance LPSs. LPSs activate the TLR4 complex, resulting in the production of proinflammatory cytokines such as TNF-α, IL-1, IL-6, and IL-8 (2, 3). There is evidence that ion channels are involved in the responses of monocytes/macrophages. For example, nonselective cation channels and the Orai family of Ca²⁺ channels are required for the migration, differentiation, and cytokine production of monocytes through changes in intracellular Ca²⁺ concentration ([Ca²⁺]_i) (4–7). Additionally, K⁺ channels have also been suggested to regulate a variety of cellular responses (8–13) via generating the negatively polarized membrane potential. In this study, inwardly rectifying K⁺ channels (Kir) attracted our attention because Kir has been reported in myeloid cells such as monocytes/macrophages rather than in lymphocytes. Owing to their high activity at membrane voltages close to the Nernst equilibrium potential of K⁺, Kir plays a relatively effective role in establishing the negative membrane potential. This hyperpolarization would drive Ca²⁺ influx once the Ca²⁺-permeable channels are activated in the immune cells (11, 14, 15).

Blood monocytes are generally considered to be the precursors for tissue macrophages and dendritic cells. However, recent studies have indicated that monocytes are also effectors of the inflammatory response to microbes via producing cytokines and chemokines, recruiting leukocytes, phagocytosis, and producing reactive oxygen species (1, 16, 17). THP-1 cells closely resemble human monocytes in their surface receptors and functional responses. When treated with phorbol ester (protein kinase C [PKC] activator), THP-1 cells differentiate into macrophage-like cells with a gradual emergence of Kir current (I_Kir) (13). However, the effects of the pathogen-recognizing receptors (e.g., TLR4) on K⁺ channel activity are currently unknown. In a pilot experiment with THP-1 cells, we found that treatment with LPS induces Kir current (I_Kir,LPS) from null current. Such a dramatic change drew our attention to reveal the activation mechanisms as well as the molecular identity.

Among the Kir subfamily, the Kir2.x group shows strong inward rectification. A previous study in human macrophages differentiated by PMA (PKC agonist) treatment suggested that Kir2.1 was responsible for the observed I_Kir (13). However, the molecular identity of PMA-induced I_Kir was simply based on the Ba²⁺ sensitivity and strong rectification that are shared by other Kir2.x members (15). Furthermore, the unitary slope conductance of PMA-induced Kir2 (IKR1) analyzed in the previous study ranged from 30 to 35 pS, which is slightly larger than the known conductance of Kir2.1 (23–30 pS) (15, 18). Therefore, precise investigation is still necessary to identify the functional Kir2.x in the human monocytes.

The activity of Kir2.x specifically requires phosphatidylinositol(4,5)-bisphosphate (PIP₂) in the inner leaflet of plasma membrane (15, 18–20). PIP₂ is the substrate for phospholipase C (PLC) and/or converted to phosphatidylinositol(3,4,5)-trisphosphate (PIP₃) by PI3K. Although PI3K is a key regulatory enzyme for a variety of signals, including the TLR4-mediated one (21, 22), the regulation of ion channels by PIP₂ decrease through conversion into PIP₃ has not been investigated except an artificial translocation of PI3K in NIH-3T3 cells transfected with KCNQ K⁺ channels (23).

In this study, we address the following matters regarding human monocytes: 1) the molecular nature of Kir channels induced by LPS and PMA, 2) the mechanism of the LPS-induced increase in I_Kir, and 3) PI3K-dependent feedback regulation of I_Kir after LPS treatment. Our results demonstrate that TLR4 stimulation induces membrane trafficking of Kir2.2 in a PKC-dependent manner. Furthermore, as a delayed response, we found that PI3K-dependent conversion of PIP₂ to PIP₃ reverses the initial I_Kir,LPS.

Undifferentiated THP-1 monocytes were cultured in suspension with RPMI 1640 medium supplemented with 10% (v/v) FBS and 1% penicillin/streptomycin at 37°C in 5% CO₂ in a humidified incubator. Cells were subcultured two to three times per week. THP-1 cells within 25 passages were used in this study. Human PBMCs were isolated by density gradient centrifugation with Histopaque (GE Healthcare). Monocytes were then purified from PBMCs by SpinSep human monocyte enrichment kits (StemCell Technologies) according to the manufacturer’s procedures. This work was approved by the Institutional Review Board of Seoul National University Hospital (Seoul, Korea).

Electrophysiological measurements were performed in the conventional whole-cell recording mode and cell-attached configuration at room temperature (22–25°C). Patch pipettes with a free-tip resistance of 3.5–5.0 MΩ were connected to the head stage of an Axopatch 200B patch-clamp amplifier (Molecular Devices, Sunnyvale, CA). pCLAMP software v.9.0 and Digidata 1332A (Molecular Devices) were used to acquire data and apply command pulses. Cells were transferred into a bath (∼0.15 ml) mounted on the stage of an inverted microscope (IX50; Olympus) and perfused with HEPES-buffered normal Tyrode (NT) solution at 5 ml/min. Single-channel activities were recorded at 10 kHz in the cell-attached configuration. Voltage and current data were low pass filtered at 5 kHz. Current traces were stored and analyzed using Clamp v.10.2 and Origin v.7.0 software (OriginLab).

The pipette solution for whole-cell experiments contained 140 mM KCl, 10 mM HEPES, 5 mM NaCl, 5 mM EGTA, 0.5 mM MgCl₂, and 2 mM MgATP, adjusted to pH 7.2 with KOH. The NT bath solution was composed of 145 mM NaCl, 10 mM HEPES, 5 mM glucose, 3.6 mM KCl, 1.3 mM CaCl₂, 1 mM MgCl₂, and 10 mM sucrose, adjusted to pH 7.4 with NaOH. To selectively evaluate the I_Kir, a 40K bath solution was used that contained 40 mM KCl with an equimolar decrease of NaCl in NT solution. The 140K bath solution for cell-attached mode contained 140 mM KCl, 5 mM NaCl, 1 mM MgCl₂, 1.3 mM CaCl₂, 5 mM glucose, and 10 mM HEPES (pH 7.4 with KOH); the pipette solution was composed of 140 mM KCl, 5 mM NaCl, and 10 mM HEPES (pH 7.4 with KOH). The internal pipette solution used for recording Ca²⁺ release–activated Ca²⁺ channel current (I_CRAC) contained 125 mM Cs-glutamate, 20 mM CsCl, 10 mM BAPTA, 3 mM MgATP, 1 mM MgCl₂, 10 mM HEPES, 0.02 mM IP₃, and 0.002 mM sodium pyruvate, and the pH was adjusted to 7.2 with CsOH. The CaCl₂ concentration of the bath solution was raised to 10 mM to measure I_CRAC. Stimulation of monocytes was performed in cell culture for 2, 4, 6, and 24 h at 37°C in an atmosphere containing 5% CO₂ in cell culture medium in the presence of 1 μg/ml LPS (Sigma-Aldrich), 500 nM PMA (Sigma-Aldrich), and 1–10 μg/ml lipoteichoic acid (Sigma-Aldrich). PI3K inhibitors LY294002 and wortmannin (Sigma-Aldrich), phosphatase and tensin homolog (PTEN) inhibitor bpV(phen) (Calbiochem), Akt signaling inhibitors MK-2206 (Santa Cruz Biotechnology) and SC-66 (Sigma-Aldrich), and tyrosine kinase inhibitor genistein (Sigma-Aldrich) were included in the intracellular pipette solution during whole-cell patch clamp recordings. Other chemicals were from Sigma-Aldrich.

THP-1 cells were pretreated with LPS for different periods (2–24 h) and washed with PBS. Cells were lysed in ice-cold lysis buffer at pH 7.5 containing 50 mM Tris-HCl, 5 mM EGTA, 2 mM EDTA, 5 mM DTT, 0.02% digitonin, and a protease/phosphatase inhibitor mixture (Roche Diagnostics). The samples were then frozen by floating the 1.5-ml tubes on a volume of liquid N₂ and thawed at room temperature. Cell lysates were then centrifuged at 14,000 × g for 30 min at 4°C, and the supernatant, which comprised the cytosolic fraction, was isolated. The pellet was then solubilized in an equal volume of the digitonin-based lysis buffer containing 1% Triton X-100 and centrifuged at 14,000 × g for 30 min at 4°C, and the supernatant, which comprised the membrane fraction, was isolated.

To obtain total protein, cells were harvested and suspended in homogenization buffer containing 50 mM Tris-HCl (pH 7.4), 100 mM NaCl, 5 mM EDTA, 1% Triton X-100, and a protease/phosphatase inhibitor mixture (Roche Diagnostics) for 1 h at 4°C. The samples were centrifuged at 13,000 × g for 15 min at 4°C. Protein concentration was determined by the Bradford assay.

The protein samples were mixed with Laemmli sample buffer, resolved by 10% SDS-PAGE, and transferred to polyvinylidene difluoride membranes in 25 mM Tris, 192 mM glycine, 0.01% SDS, and 20% methanol. Membranes were blocked in 1× TBS containing 1% Tween 20 and 5% BSA (blocking solution) for 1 h at room temperature with gentle rocking, and incubated overnight at 4°C with anti-Kir2.1, anti-Kir2.2 (Alomone Labs), and anti–Na⁺-K⁺ ATPase (Abcam) primary Abs followed by relevant secondary Abs after washing. The signals were determined using ECL Plus Western blotting detection reagents (Amersham Biosciences) and detected by film exposure. The intensity of each band was measured using with ImageJ analysis software program.

Prior to cDNA amplification, total RNA was isolated using TRizol (Invitrogen) from THP-1 cells and human primary monocytes. Human Kir2.1–4 and β-actin mRNAs were analyzed using an established RT-PCR method. Briefly, 1 μg total RNA was reverse transcribed and the produced cDNA was amplified with 35 PCR cycles (55°C for 0.5 min, 72°C for 1 min, and 95°C for 2.5 min). The nucleotide sequences of the primers used for amplification were (forward/reverse): Kir2.1, 5′-GGCTGTGTGTTTTGGTTGATAGC-3′ and 5′-ATAAGAGCTACGGCACTGTGTCG-3′; Kir2.2, 5′-ACCCCTACAGCATCGTGTCA-3′ and 5′-AGATGAGCAGCATGTACCGC-3′; Kir2.3, 5′-GACTTTGAGATCGTGGTCAT-3′ and 5′-CAGCACGGTGATCTTACTCT-3′; Kir2.4, 5′-ACGTGCGTTTCGTAAACCTG-3′ and 5′-GAGAAGAGCAGGCACATCCA-3′; and β-actin, 5′-GTCACCTTCACCGTTCCAGTT-3′ and 5′ -TTAGTTGCGTTACACCCTTTC-3′. The PCR products were electrophoresed in a 2% agarose gel at 120 V in 1× Tris acetate–EDTA buffer and visualized using ethidium bromide.

Total RNA was extracted using TRizol reagent (Invitrogen), and its purified concentration was determined NanoDrop at 260 nm. cDNA was produced using a SuperScript III first-strand synthesis system (Invitrogen) according to the manufacturer’s instructions. Primers were: β-actin (forward, 5′-GAGAAAATCTGGCACCACACC-3′; reverse, 5′-GGATAGCACAGCCTGGATAGCAA-3′); Kir2.2 (forward, 5′-CTCCTACCTGGCCAATGAGA-3′; reverse, 5′-GGTCTTGTGGAAGTGCGAGT-3′). cDNAs were amplified with SYBR Green (TOPreal qPCR 2× PreMIX) using a real-time PCR system (Applied Biosystems 7500 real-time PCR) under the conditions 95°C for 10 min, 40 cycles at 95°C for 10 s, 60°C for 15 s, and 72°C for 15 s. Specificity of amplification products was assessed by melting curve analysis. Relative gene expression was calculated using the comparative threshold (Ct) method (2^−ΔΔCt).

For immunofluorescence experiments, THP-1 cells were fixed with 4% paraformaldehyde in PBS for 30 min at room temperature. For Kir2 channel staining, cells were permeabilized with 0.1% Triton X-100 in PBS for 10 min, blocked in PBS with 10% FBS for 1 h at room temperature, and incubated with anti-Kir2.1 or anti-Kir2.2 (respectively, 1:200) primary Abs overnight at 4°C. The subcellular location of Kir2 channels was assessed using Alexa Fluor 488–conjugated (488 nm excitation, 1:500, Invitrogen) secondary Abs. Images were acquired using an Olympus FluoView FV1000 confocal microscope with a ×100 oil-immersion objective.

For suppressing PKCε expression, small interfering RNA (siRNA)-specific (si-)PKCε, si-Kir2.1, and si-Kir2.2 were purchased from Cell Signaling Technology and OriGene Technology. siRNAs and GFP were transiently transfected using Nucleofector kits and the corresponding protocols (Amaxa Biosystems), according to the manufacturer’s instructions for patch clamp recording and subfractionation.

THP-1 cells were transiently transfected with plasmids expressing the GFP fused tagged pleckstrin homology (PH) domain of either PLCδ (PH-PLCδ-GFP) or Akt (PH-Akt-GFP). The next day, the cells were treated with or without LPS for 24 h and then plated onto poly-l-lysine–coated glass-bottom dishes. Laser scanning confocal microscopy with the Olympus FluoView FV1000 and a ×60 oil-immersion objective was used to visualize PH-GFPs (excitation 488 nm) in single cells treated with NT solution or 30 μM LY294002 for 30 min.

[Ca²⁺]_i was measured using the fluorescent Ca²⁺ indicator fura 2-AM. THP-1 cells were loaded with fura 2-AM (5 μM, 30 min, 25°C) and washed twice with HEPES-buffered physiological salt solution. Fluorescence was monitored in a quartz microcuvette (1 ml) with stirring in a temperature-controlled (37°C) cell holder of a fluorescence spectrophotometer (Photon Technology International) at excitation wavelengths of 340 and 380 nm and an emission wavelength of 510 nm. At the end of each experiment, 5 μM ionomycin was applied to produce a maximum fluorescence ratio (R_max, 340/380 nm). Subsequently, 10 mM EGTA was added to confirm a minimum value of fluorescence ratio (R_min). The [Ca²⁺]_i values were calculated using the equation [Ca²⁺]_i = K_d × b × (R − R_min)/(R_max − R), where K_d is the dissociation constant (224 nM) of fura 2-AM and b is the ratio of the fluorescence excitation intensities of fura 2-AM at 380 nm under Ca²⁺-free and Ca²⁺-saturated conditions.

THP-1 cells preincubated in the absence or presence of the Kir2 antagonist ML-133 for 30 min were stimulated with LPS for 4 and 8 h in a humidified atmosphere (5% CO₂, 37°C). Thereafter, culture supernatants were aspirated, centrifuged at 200 × g for 10 min at 4°C, and stored at −80°C until quantification of cytokines. Concentrations of IL-8 and TNF-α were determined using the Bio-Plex 200 system (Bio-Rad).

Data were managed and analyzed using Microcal Origin v.7.0 software (Malvern Instruments). Statistical results are presented as the mean ± SEM. Student t tests were used where appropriate to evaluate for significance, which was accepted for p values <0.05.

Under whole-cell mode voltage clamp conditions, current/voltage relationships (I/V curves) were obtained in either NT or moderately high [K⁺] (40K) solution. In the unstimulated control THP-1 cells, only voltage-gated K⁺ channel current was observed (Fig. 1A) without acute effect of LPS up to 15 min of treatment (Supplemental Fig. 1). Hereafter, the whole-cell currents and I/V curves were evaluated using the 40K condition, and the presented inward I_Kir amplitudes were measured at −80 mV. The I/V curves of THP-1 cells exposed to LPS for different periods were compared: LPS-2h (1–3 h), LPS-4h (3–5 h), LPS-6h (5–7 h), and LPS-24h (23–25 h) (Fig. 1B). Inward K⁺ current was practically absent in the LPS-untreated control (−2 ± 0.4 pA/pF at −80 mV in 40K, Fig. 1A). Interestingly, a large I_Kir was observed in LPS-2h (−55 ± 8.1 pA/pF) that reversed at around −30 mV, close to the calculated Nernst equilibrium potential for K⁺ at 40K. I_Kir was further increased in LPS-4h (−119 ± 8.6 pA/pF) and spontaneously decayed to −74 ± 13.5 and −32 ± 6.7 pA/pF in LPS-6h and LPS-24h, respectively (Fig. 1B, 1C). The primary monocytes from human peripheral blood also showed no significant I_Kir before stimulation, while I_Kir,LPS was consistently observed from 2 to 6 h of LPS application and decayed spontaneously at 24 h (Fig. 1C). The appearance of I_Kir,LPS was accompanied by hyperpolarization of membrane potential (V_m) in THP-1 cells. The V_m of control and LPS-4h measured under the zero-current clamp mode were −31 ± 0.4 mV (n = 9) and −72 ± 1.7 mV (n = 5), respectively. Application of Ba²⁺ (100 μM), a representative blocker of Kir2 channel, depolarized the V_m of LPS-4h to the level of control cells (Supplemental Fig. 2). We also tested whether stimulation of TLR2 induces similar change in the membrane currents of THP-1. Different from LPS, a treatment with TLR2 agonist, lipoteichoic acid, did not induce I_Kir at all up to 24 h (Supplemental Fig. 3).

To characterize the molecular identity of I_Kir,LPS, unitary conductance was analyzed by the cell-attached patch clamp recordings. Both in THP-1 and human blood monocytes, inwardly rectifying channel activities were commonly observed, and their unitary conductance was almost the same (37.7 and 38.1 pS, Fig. 1D), which was closer to Kir2.2 (35–40 pS) than the smaller ones such as Kir2.1 and Kir2.3 (15, 18). According to the reported pharmacology of Kir2.2, the IC₅₀ of Ba²⁺ for Kir2.2 ranges from 0.5 to 2.3 μM whereas the IC₅₀ of Ba²⁺ for Kir2.1 ranges from 3 to 16 μM (15, 18, 24). Application of Ba²⁺ inhibited I_Kir,LPS with IC₅₀ of 1.42 μM at −100 mV in THP-1 cells (Fig. 1E, 1F).

The unitary conductance and Ba²⁺ sensitivity of I_Kir,LPS were consistent with the properties of Kir2.2. However, the expressions of Kir2.1 and Kir2.2 transcripts and proteins were commonly observed in THP-1 and primary monocytes (Fig. 2A, 2B). The amount of Kir2.2 and Kir2.1 proteins from total cell preparation was not changed in the period of I_Kir,LPS observation (Fig. 2B). The molecular identity of I_Kir,LPS was further proven by si-Kir2.1 and si-Kir2.2 transfection. Scrambed siRNA (scRNA), si-Kir2.1, or si-Kir2.2 was cotransfected with GFP vector in THP-1 cells. Twenty-four hours after the transfection, the cells were treated with LPS for another 4 h. Because the efficiency of siRNA transfection was generally low in the suspended culture cells such as THP-1, we could not apply the immunoblot assay but used RT-PCR analysis for the GFP⁺ cells sorted by flow cytometry. Although the RT-PCR analysis is semiquantitative, we confirmed the decreased amounts of mRNAs for Kir2.1 and Kir.2.2 in each population of cells transfected by si-Kir2.1 and si-Ki2.2, respectively (Fig. 2C). We then performed whole-cell patch clamp in the GFP⁺ cells under fluorescence microscopy examination. The amplitudes of I_Kir,LPS were indistinguishable between scRNA- and si-Kir2.1–transfected cells. In contrast, I_Kir,LPS was significantly decreased in the si-Kir2.2–transfected than the scRNA- and si-Kir2.1–transfected cells (Fig. 2D). Thus, Kir2.2 was suggested to be responsible for I_Kir,LPS. Although the total amount of Kir2.2 protein was not significantly changed, real-time PCR analysis showed increased level of Kir2.2 transcripts 4 h after the LPS treatment (Fig. 2E).

Recently, a selective inhibitor for Kir2, ML-133, was reported (25), which was confirmed in LPS-4h cells (Fig. 3A). Next, we compared the production of cytokines upon LPS stimulation between control and 30 μM ML-133–treated THP-1 cells. Both TNF-α and IL-8 secretion induced by LPS was significantly decreased in the ML-133–treated cells (Fig. 3B, 3C). To evaluate the plausible augmentation of Ca²⁺ influx by I_Kir,LPS-induced hyperpolarization, [Ca²⁺]_i was measured in LPS-treated THP-1 cells. Ca²⁺ store depletion-activated Ca²⁺ entry was induced by thapsigargin (2 μM), a SERCA inhibitor. Thapsigargin in Ca²⁺-free condition induced a transient increase in [Ca²⁺]_i (Δ[Ca²⁺]_i,TG) reflecting the spontaneous release of Ca²⁺ from endoplasmic reticulum stores (Fig. 3D). The amplitude of Δ[Ca²⁺]_i,TG was indistinguishable between control and LPS-4h groups. Subsequently, extracellular Ca²⁺ was replenished to 5 mM, which increased [Ca²⁺]_i above the initial level (Fig. 3D). The Ca²⁺ add-back after thapsigargin is a commonly used procedure to assess the store-operated Ca²⁺ entry (SOCE) in the immune cells (7). We also compared the I_CRAC in whole-cell patch clamp recordings. With inositol 1,4,5-trisphosphate (20 μM) and 10 mM BAPTA in the pipette solution, a spontaneous increase in inward Ca²⁺ current was observed in THP-1 cells. The reversal potential of I_CRAC was 40 mV, consistent with Ca²⁺-selective permeability. I_CRAC density was unchanged in LPS-4h (Fig. 3E). Despite the unaltered I_CRAC, SOCE levels were higher in LPS-4h than in control (Fig. 3D, green versus black trace). The pretreatment with 30 μM ML-133 lowered the SOCE of LPS-4h but not the SOCE of control cells (Fig. 3D, blue versus red trace). The peak amplitudes of SOCE from the tested groups were also directly compared as bar graphs (Fig. 3F).

Because the total protein levels of Kir2.1 and Kir2.2 were not significantly increased (Fig. 2B), we hypothesized that plasma membrane translocation of Kir2.2 was responsible for the I_Kir,LPS. Because extracellular lysine residue is absent in human Kir2.2, surface biotinylation assay could not be applied. Instead, we adopted immunoblotting after membrane fractionation (Fig. 4A). Note that the non–cytosolic membrane fraction also contained organellar membranes as indicated by the presence of calnexin and syntaxin 6. Nevertheless, there was insignificant Kir2.1 and Kir2.2 proteins in the membrane fraction isolated from control cells. The Kir2.2 signal in the membrane fraction continuously increased up to 24 h of LPS treatment whereas the constitutive plasma membrane protein, α subunit of Na⁺/K⁺-ATPase, was unchanged. In contrast, there was no appearance of Kir2.1 in the membrane fraction of LPS-treated THP-1 cells (Fig. 4A).

The pharmacological inhibitors of vesicle translocation from endoplasmic reticulum to Golgi and plasma membrane, such as brefeldin A and Exo-1 (26), largely suppressed the I_Kir induction in LPS-4h (Fig. 4B). Confocal immunofluorescence microscopy showed that Kir2.2 was homogeneously distributed in the cytoplasm of control cells. After LPS treatment, Kir2.2 was also observed in the periphery of THP-1 cells (Fig. 4C). In contrast, Kir2.1-specific fluorescence signals were localized at intracellular clusters and were unchanged by LPS treatment up to 24h (Fig. 4D).

A previous report showed that PKC activation induces I_Kir in THP-1 cells (13). Thus, we hypothesized that PKC might also be involved in I_Kir,LPS. Three different types of PKC inhibitors (Go6976, Go6983, and GF109203X [GFX]) were applied 30 min before LPS treatment, and I_Kir,LPS amplitudes at 4 h were compared. Among them, GFX effectively suppressed I_Kir,LPS (Fig. 5A, 5B). According to the known relative potency of PKC inhibitors, the above results suggest the involvement of PKCε isoform (see 17Discussion), which is also known as a component of the TLR4 signaling pathway in monocytes/macrophages and dendritic cells (27, 28). Consistently, in THP-1 cells transfected with si-PKCε, the I_Kir,LPS was significantly smaller than in control GFP-transfected cells (Fig. 5C). Furthermore, the membrane translocation of Kir2.2 was also inhibited by GFX or si-PKCε transfection (Fig. 5D).

We also assessed whether PKC activation alone induces I_Kir in monocytes. Treatment with PMA (0.5 μM), a chemical activator of multiple isotypes of PKC, also induced significant I_Kir. Different from I_Kir,LPS, the PMA-induced Kir current (I_Kir,PMA) did not show spontaneous decay but increased from 6 h up to 24 h after the PMA treatment (Fig. 5E). Interestingly, immunoblot assay of membrane fraction showed increases of Kir2.1 as well as Kir2.2 in the PMA-treated THP-1 cells (Fig. 5F). Consistently, I_Kir,PMA was attenuated by transfection with si-Kir2.1 as well as si-Kir2.2 (Fig. 5G). Finally, the single channel analysis showed two types of slope conductance in the PMA-treated cells, that is, 24.0 ± 2.1 pS (n = 4) and 39.5 ± 1.8 pS (n = 6) (Fig. 5H) (15, 18).

The spontaneous decay of I_Kir,LPS after the peak amplitude at 4–6 h of LPS treatment (Fig. 1B, 1C) was inconsistent with the continuous increase in Kir2.2 protein in the membrane fraction (Fig. 4A). In fact, the decay of I_Kir,LPS in LPS-24h was not affected by cotreatment with dynasore, an inhibitor of dynamin that is required for clathrin-mediated endocytosis (29, 30), (Fig. 6A, 6B). The increase in membrane capacitance by LPS, an indicator of vesicle fusion to plasma membrane, was also persistently observed up to 24 h (Fig. 6C). Therefore phosphorylation of channel proteins or modulation of Kir2.2 by membrane phospholipid were suggested as a plausible mechanism of the later decay of I_Kir,LPS.

It has been reported that Kir2.1 activity is modulated by phosphorylation at tyrosine residue (T242) in the C-terminal domain that is commonly found in the Kir2.x group (31, 32). However, application of the tyrosine kinase inhibitor genistein (100 μM) did not reverse the decay of I_Kir,LPS in LPS-24h (Fig. 6D). We also observed that the mean amplitudes of I_Kir in LPS-24h still showed spontaneous decay when cotreated with genistein (10 μM, 24 h) (Supplemental Fig. 4). It is known that TLR4 signals lead to activation of PI3K- and Akt-dependent signaling pathways (21, 22). However, treatment with the Akt inhibitor MK-2206 did not change the amplitude of I_Kir in LPS-24h (Fig. 6D). Another Akt inhibitor, SC-66, also had no effect on I_Kir in LPS-24h (Fig. 6F).

Although the total electrical negativity of PIP₃ is higher than that of PIP₂, it is known that Kir2.2 specifically requires PIP₂, not PIP₃, for activity via conformation-dependent interaction (33). Therefore, we hypothesized that the conversion of PIP₂ into PIP₃ by PI3K might explain the decay of I_Kir,LPS, which can be reversed by PIP₃ phosphatase, PTEN (Fig. 6G). Interestingly, an acute treatment with PI3K inhibitors LY294002 and wortmannin increased the amplitude of I_Kir,LPS, whereas the PTEN inhibitor bpV(phen) further decreased the I_Kir,LPS in LPS-24h (Fig. 6E, 6F).

To examine whether significant conversion of PIP₂ into PIP₃ by PI3K indeed occurs in LPS-24h, PH domain–containing fluorescence markers for PIP₂ (PH-PLCδ-GFP) and PIP₃ (PH-Akt-GFP) (34) were expressed in THP-1 cells to be analyzed by confocal microscopy. At 30 min after applying LY294002, a significant increase in the PH-PLCδ-GFP signal was observed in the peripheral region of LPS-24h (Fig. 7A) whereas no significant change was observed in the untreated control THP-1 cells (Fig. 7B). In contrast, the PH-Akt-GFP signal showed a reciprocal decrease by LY294002 (Fig. 7C). No change of PH-Akt-GFP signal was also confirmed in control cells (Fig. 7D). To summarize the data, the fluorescence intensity of each PH-GFP in the membrane region was normalized to the initial level (F_Mem,30/F_Mem,ini). In comparison with the relatively constant intensity of PH-PLCδ and PH-Akt in control cells (0.90 ± 0.04 and 0.95 ± 0.039, respectively), PH-PLCδ-GFP and PH-Akt-GFP were significantly increased (1.21 ± 0.088) and decreased (0.71 ± 0.056), respectively, in LY294002-treated cells (Fig. 7E).

In comparison with the previous studies of I_Kir upregulation in PMA- or M-CSF–stimulated macrophages and microglia (9, 13), our study provides several novel findings regarding the TLR4-stimulated human monocytes: 1) the molecular identity of I_Kir,LPS as Kir2.2, 2) PKC-dependent translocation to plasma membrane as the major mechanism of I_Kir,LPS, and 3) inhibition of I_Kir,LPS by PI3K-dependent decrease of PIP₂. Additionally, both Kir2.1 and Kir2.2 are induced by PMA treatment in THP-1 cells.

Although the specific values are variable depending on recording conditions, members of Kir2 (Kir2.1–2.4) show generally distinguishable unitary conductance: 20–30 pS for Kir2.1, 35–40 pS for Kir2.2, 10–15 pS for Kir2.3, and 14–15 pS for Kir2.4 (15, 18). Despite the presence of Kir2.1 and Kir2.2 mRNAs and proteins in the monocytes, the unitary slope conductance, the sensitivity to Ba²⁺ (relatively low IC₅₀), and the selective suppression by si-Kir2.2 transfection suggested that Kir2.2 is the molecular identity of I_Kir,LPS. In contrast to the numerous investigations regarding Kir2.1, the physiological role of Kir2.2 has been rarely investigated, except in the context of its coexpression with Kir2.1 in cardiomyocytes and aortic endothelial cells (24, 35). The unitary conductance of the Kir2.1/Kir2.2 heterotetramer is known to be <30 pS (15), smaller than that of Kir2.2. Additionally, plasma translocation of Kir2.1 was not observed in the LPS-treated THP-1 cells. Taken together, it is highly likely that a homotetramer of Kir2.2 is functioning in human monocytes stimulated by LPS. To our knowledge, this is the first functional finding of endogenous Kir2.2 in mammalian cells. Interestingly, the functional activity and the membrane translocation of Kir2.1 were also observed in PMA-treated THP-1 cells (Fig. 5F, 5H), which is consistent with the KIR1 described in a previous study (13). Although Kir2.1 was induced only by a chemical PKC agonist, further investigation of intrinsic receptors signaling the Kir2.1 regulation in monocytes/macrophages is worth pursuing.

Both confocal microscopy and immunoblot assay indicate that plasma membrane translocation of Kir2.2 is responsible for the I_Kir,LPS (Fig. 4A, 4C). Consistently, pharmacological trafficking inhibitors effectively inhibited the I_Kir,LPS (Fig. 4B). Although not rigorously investigated in the present study, the increase in plasma membrane electrical capacitance in the LPS-treated monocyte membrane might also reflect the trafficking of Kir2.2-containing vesicles to the plasma membrane (Fig. 6C). However, the increasing Kir2.2 in membrane fraction was not clearly reflected as a corresponding decrease in the cytosolic Kir2.2 proteins (Fig. 4A). Despite the multiple experimental clues that suggest the membrane translocation of Kir2.2 in the LPS-treated THP-1 cells, we could not exclude the possibility of a translational upregulation mechanism. In fact, the amount of Kir2.2 transcript was increased by LPS treatment (Fig. 2E). Although the total amount of Kir2.2 was not significantly increased up to 24 h after LPS, a marginal level of translational upregulation might have partly compensated the translocation from cytosol to membrane.

As for the intracellular signal triggering the putative trafficking of Kir2.2, the PKC pathway has attracted our attention because of the previous report in THP-1 cells (13). Also, treatment with the PKC activator PMA alone induced I_Kir, albeit with different time course, in THP-1 cells (Fig. 6E, 6F). Binding of LPS to the CD14/TLR4/MD2 complex activates a number of signaling cascades, namely, NF-κB, MAPK, and multiple isoforms of PKC (27, 28, 36). Among the three kinds of PKC inhibitors tested here, only GFX suppressed I_Kir,LPS. According to recent studies, Go6976 discriminates between conventional PKCs (α, β, γ) and novel PKCs (δ, ε). Go6983 preferentially inhibits conventional PKCs and PKCδ and PKCζ at nanomolar concentrations. GFX preferentially inhibits conventional PKCs (α, β) at concentrations below 1 μM, and also inhibits novel PKCs (δ, ε) at concentrations <5 μM (37, 38). Based on the pharmacological selectivity of the agents described above, it is likely that a PKCε-dependent signaling mechanism is at least partly responsible for the Kir2.2 trafficking. The involvement of PKCε in the membrane trafficking of an ion channel (NMDA receptor) and a transporter (pendrin) has also been suggested elsewhere (39, 40). Because the pharmacological selectivity of the PKC inhibitors are generally not highly reliable, careful interpretation is required. Therefore, the potential involvement of PKCε was further confirmed by si-PKCε transfection of THP-1 (Fig. 5C, 5D). Nevertheless, the partial inhibitory effect of si-PKCε transfection (Fig. 5C) suggested that PKCε may not be the sole pathway of I_Kir,LPS in human monocytes. The complexity of PKC isotypes regarding the Kir channel regulation was also suggested from the different pattern of Kir upregulation in PMA-treated THP-1; both Kir2.1 and Kir2.2 were induced as confirmed by immunoblot, genetic knockdown, and single-channel patch clamp experiments (Fig. 5F–H). Taken together, different isotypes of PKC might be responsible for the functional upregulation of Kir2.1 and Kir2.2 in human monocytes.

After the impressive increase at 4 h, the phenomenon of decaying I_Kir,LPS raised a mechanistic question. After excluding the endocytosis- or kinase-dependent mechanisms, we hypothesized that PIP₂ might be changed by PI3K, which is well known to be activated by LPS/TLR4 signaling (21, 22, 41). The reciprocal changes in I_Kir,LPS by PI3K inhibitors (LY294002 and wortmannin) and PTEN inhibitor (bpV[phen]) in LPS-24h suggest that the enhanced conversion of PIP₂ into PIP₃ (i.e., decreased PIP₂ availability) might have attenuated the I_Kir,LPS despite the sustained membrane localization of Kir2.2. The confocal microscopy of THP-1 cells transfected with PH-PLCδ-GFP and PH-Akt-GFP demonstrates the dynamic changes in PIP₂ levels in LPS-24h (Fig. 7). Although the anionic charge of PIP₃ is greater than that of PIP₂, its efficiency in Kir2.2 activation is far lower than that of PIP₂ (33), which might explain the spontaneous decrease in I_Kir,LPS.

The KCNQ type of K⁺ channel is also known to be dependent on the PIP₂ concentration for its activity. In a previous study investigating the chemical translocation of PI3K, however, acute PIP₃ formation was not accompanied by significant decrease in PIP₂ or in the KCNQ activity (23). Because the relative amount of PIP₃ is very small among the anionic phospholipids in the plasma membrane (42, 43), the formation of PIP₃ for the conventional signaling cascades (e.g., Akt pathway) may not necessarily induce a detectable decrease in the PIP₂ concentration. We interpret that the high ratio of endogenous PIP₂ over PIP₃ might be the reason for several hours of delay to induce noticeable reverse of I_Kir,LPS. Another possibility of the spontaneous decay mechanism was a putative inhibition of I_Kir by the downstream signaling from PIP₃. However, the pharmacological inhibition of the Akt pathway had no effect on the I_Kir,LPS in LPS-24h (Fig. 6F).

The endogenous PIP₃-dependent regulation of ion channel activity is generally thought to be less likely, although TRPC1 and CNG channels may be the exceptions (44, 45). On these backgrounds, our present results suggesting the altered PIP₂ availability for Kir2.2 regulation via the reciprocal activity of PI3K and PTEN are intriguing. The persistent increase in I_Kir,PMA might be due to the absence of PI3K recruitment. It would be worth investigating whether similar regulation of PIP₂-dependent Kir2.x by PI3K occurs in other cell types.

What is the physiological implication of the delayed reverse of I_Kir,LPS? Excessive activation of monocytes/macrophages in vivo by endotoxins can lead to septic shock with multiorgan failure and death. Although cytokines are important for the efficient control of pathogens, overproduction of cytokines can be harmful for the host (3). Activation of the PI3K/Akt pathway via LPS stimulation has been shown to negatively regulate NF-κB activation and the expression of inflammatory genes (21). Thus, spontaneous decay of I_Kir in the later phases of the LPS/TLR4 response might be an example of a negative feedback mechanism via the PI3K-dependent pathway.

In summary, we suggest that Kir2.2 is functionally upregulated by LPS stimulation via signaling mechanisms including PKCε-dependent membrane trafficking in human monocytes. Subsequent hyperpolarization enhanced Ca²⁺ influx and cytokine secretion. Additionally, spontaneous decay of Kir2.2 currents accounts for the decreased availability of PIP₂ due to PI3K activation (Fig. 7F). Note also that Kir2.1 and Kir2.2 might be differentially recruited depending on the PKC isotypes activated in human monocytes. Considering the critical roles of monocytes as a natural defense system, Kir2.2 in monocytes could serve as an important regulator for the innate immune mechanism.

We thank Prof. Joo Hong Jeon (Department of Physiology, Seoul National University College of Medicine) for helpful discussions and comments.

The authors have no financial conflicts of interest.