Association of Kv1.5 and Kv1.3 Contributes to the Major Voltage-dependent K+ Channel in Macrophages*

Voltage-dependent K+ (Kv) currents in macrophages are mainly mediated by Kv1.3, but biophysical properties indicate that the channel composition could be different from that of T-lymphocytes. K+ currents in mouse bone marrow-derived and Raw-264.7 macrophages are sensitive to Kv1.3 blockers, but unlike T-cells, macrophages express Kv1.5. Because Shaker subunits (Kv1) may form heterotetrameric complexes, we investigated whether Kv1.5 has a function in Kv currents in macrophages. Kv1.3 and Kv1.5 co-localize at the membrane, and half-activation voltages and pharmacology indicate that K+ currents may be accounted for by various Kv complexes in macrophages. Co-expression of Kv1.3 and Kv1.5 in human embryonic kidney 293 cells showed that the presence of Kv1.5 leads to a positive shift in K+ current half-activation voltages and that, like Kv1.3, Kv1.3/Kv1.5 heteromers are sensitive to r-margatoxin. In addition, both proteins co-immunoprecipitate and co-localize. Fluorescence resonance energy transfer studies further demonstrated that Kv1.5 and Kv1.3 form heterotetramers. Electrophysiological and pharmacological studies of different ratios of Kv1.3 and Kv1.5 co-expressed in Xenopus oocytes suggest that various hybrids might be responsible for K+ currents in macrophages. Tumor necrosis factor-α-induced activation of macrophages increased Kv1.3 with no changes in Kv.1.5, which is consistent with a hyperpolarized shift in half-activation voltage and a lower IC50 for margatoxin. Taken together, our results demonstrate that Kv1.5 co-associates with Kv1.3, generating functional heterotetramers in macrophages. Changes in the oligomeric composition of functional Kv channels would give rise to different biophysical and pharmacological properties, which could determine specific cellular responses.


ramers in macrophages. Changes in the oligomeric composition of functional Kv channels would give rise to different biophysical and pharmacological properties, which could determine specific cellular responses.
Voltage-dependent potassium channels (Kv) 10 have a crucial function in excitable cells of determining resting membrane potential and controlling action potentials (1). In addition, they are involved in the activation and proliferation of leukocytes (2). Functional Kv complexes are formed by four transmembrane ␣ subunits and up to four cytoplasmic ␤ subunits (3). The mammalian Shaker family (Kv1) contains at least eight different genes (Kv1.1-Kv1.8), coding for ␣ subunits, which form functional homo-and heterotetrameric complexes. Thus, Kv1 proteins can assemble promiscuously, yielding a wide variety of biophysically and pharmacologically distinct channels (4,5). However, although a number of studies have demonstrated that specific Kv heteromeric complexes predominate in nerve and muscle, many other possible combinations go undetected (6,7). Therefore, this mechanism of channel assembly may underlie some of the functional diversity of potassium currents found in the brain and the cardiovascular system.
Bone marrow-derived macrophages (BMDM) are fully differentiated cells. In response to different growth factors and cytokines, macrophages can proliferate, become activated, or differentiate. These cells have a key function at inflammatory loci, where they arrive 24 -48 h after lesion and remain until inflammation disappears. However, the persistence of macrophages at inflammatory loci is associated with the pathogenesis of a wide range of inflammatory diseases. Kv are tightly regulated during proliferation and activation in macrophages, and their functional activity is important for cellular responses (8). Proliferation and activation trigger an induction of the outward K ϩ current that is under transcriptional and translational control (8). Several lines of evidence indicate that post-translational events are involved in Kv regulation. In this context, assigning specific K ϩ channel clones to native currents is difficult, since this complexity is further enhanced by heteromultimeric assembly of different Kv subunits. Lymphocytes express several voltage-dependent K ϩ currents (n, nЈ, and l-type channels). Although Kv1.3, the major Kv channel in leukocytes, is associated with the n-type channel and Kv3.1 accounts for the l-type, the proteins responsible for the nЈ-type are unknown (2). Electrophysiological properties such as activation and inactivation of Kv1.3 expressed in T-cells and heterologous expression systems (Refs. 9 and 10 and references herein) are significantly different from those described in macrophages (8,10,11). In addition, unlike T-lymphocytes, brain and bone marrow macrophages also express Kv1.5 (8,10,(11)(12)(13)(14)(15). Kv1.3 and Kv1.5 differ in their biophysical and pharmacological properties and show distinct regulation in a number of cell types (8, 16 -18). Thus, different K ϩ channel subunit composition could lead to specific alteration of cellular excitability, thus determining specific cell responses.
The aim of the present study was to explore whether Kv1.5 has a function in the major voltage-dependent K ϩ current in macrophages. Our results suggest that Kv1.5 co-associates with Kv1.3, generating functional Kv1.3/Kv1.5 heterotetrameric channels. Upon different physiological stimuli, changes in the oligomeric composition of functional Kv could have a crucial effect on intracellular signals, determining the specific macrophage response.

EXPERIMENTAL PROCEDURES
Animals and Cell Culture-BMDM and Raw 264.7 macrophages, human embryonic kidney 293 (HEK-293) cells, EL-4 T cell line, and Xenopus laevis oocytes were used. BMDM from 6 -10-week-old BALB/c mice (Charles River Laboratories) were isolated and cultured as described elsewhere (8). Briefly, animals were killed by cervical dislocation, and both femurs were dissected with adherent tissue removed. The ends of bones were cut off, and the marrow tissue was flushed by irrigation with medium. The marrow plugs were passed through a 25-gauge needle for dispersion. The cells were cultured in plastic dishes (150 mm) in Dulbecco's modified Eagle's medium containing 20% fetal bovine serum and 30% L-cell-conditioned media as a source of macrophage-colony stimulating factor. Macrophages were obtained as a homogeneous population of adherent cells after 7 days of culture and maintained at 37°C in a humidified 5% CO 2 atmosphere. Raw 264.7 macrophages and EL-4 and HEK-293 cells were cultured in Dulbecco's modified Eagle's medium culture media containing 10% fetal bovine serum supplemented with 10 units/ml penicillin and streptomycin and 2 mM L-glutamine. Cells were grown in 100-mm tissue culture dishes for sample collection and on non-coated glass coverslips for electrophysiology and confocal imaging. In some experiments Raw 264.7 cells were incubated with 100 ng/ml recombinant TNF-␣ (PrepoTech E) for 24 h. All animal handling was approved by the ethics committee of the University of Barcelona and was in accordance with European Union regulations.
Protein Extracts, Immunoprecipitation, and Western Blot-Cells were washed twice in cold phosphate-buffered saline (PBS) and lysed on ice with lysis solution (1% Nonidet P-40, 10% glycerol, 50 mmol/liter HEPES, pH 7.5, 150 mmol/liter NaCl) supplemented with 1 g/ml aprotinin, 1 g/ml leupeptin, 86 g/ml iodoacetamide, and 1 mM phenylmethylsulfonyl fluoride as protease inhibitors. To obtain enriched membrane preparations, homogenates were centrifuged at 3,000 ϫ g for 10 min, and the supernatant was further centrifuged at ϳ150,000 ϫ g for 90 min. The pellet was resuspended in 30 mM HEPES, pH 7.4, and protein content was determined by Bio-Rad protein assay. Samples were separated into aliquots and stored at Ϫ80°C.
For immunoprecipitation studies, membrane pellets were resuspended in 1% Triton, 10% glycerol, 50 mM HEPES, pH 7.2, 150 mM NaCl incubated in the presence of protein A-Sepharose, and pelleted. Supernatants incubated overnight with anti-Kv1.3 antibody (Alomone) were further incubated with protein A-Sepharose, washed with 0.1% Triton/PBS, and centrifuged. Pellets were re-suspended in Laemmli SDS-loading buffer, boiled at 95°C, and separated on 8% SDS-PAGE. To detect immunoprecipitated Kv1.5, an anti-Kv1.5 antibody (1/250) produced and characterized in the Tamkun laboratory was used. Densitometric analysis of the filters was performed by Phoretix software (Nonlinear Dynamics). Results are the mean Ϯ S.E. of each experimental group.
DNA Constructs, Cell Transfection, and Microinjection-Rat Kv1.3 cDNA, kindly donated by T. C. Holmes (New York University), was amplified by PCR and ligated into pEYFP-C1 (Clontech) by using BglII and HinDIII restriction sites. Human Kv1.5 cDNA was amplified by PCR and ligated into pECFP-C1 (Clontech) by using BglII and EcoRI. Both constructs were verified by sequencing.
Raw 264.7 and HEK cells were grown on glass coverslips in 35-mm dishes, and transient transfection was performed using Lipofectamine 2000 (Invitrogen) to near 80% confluency. Twenty-four hours after transfection, cells were washed with PBS, fixed, and mounted with Aqua Poly/Mount from Polysciences, Inc.
X. laevis oocytes were prepared and injected by standard methods. Mature female frogs were purchased from the Centre d'Elevage de Xenopes (Montpellier, France). Animals were anesthetized in cold distilled water containing 1.7 g/liter tricaine(ethyl 3-aminobenzoate methanesulfonic acid; Sigma). Ovarian sacs were extracted by sterile surgical procedures and placed in sterile Barth's solution (88 mM NaCl, 1 mM KCl, 0.33 mM Ca(NO 3 ) 2 , 0.41 mM CaCl 2 , 0.82 mM MgSO 4 , 2.40 mM NaHCO 3 , and 20 mM HEPES at pH 7.5 supplemented with penicillin 100 IU/ml and streptomycin 0.1 mg/ml). Oocytes at stages V and VI were dissected out and kept at 15-16°C in sterile Barth's solution.
Full-length Kv1.3 and Kv1.5 were subcloned into pcDNA3. cRNAs were generated with T7 RNA polymerase by using the mMESSAGE mMACHINE kit according to manufacturer's instructions (Ambion). Oocytes were injected with 10 ng of Kv1.3 or Kv1.5 of in vitro transcribed cRNAs to form homotetramers. Heterotetramers were expressed in oocytes injected with 10 ng of total cRNA of Kv1.3 and Kv1.5 in the proportions of 1:1, 3:1, and 1:3 using a Variable Nanoject (Drummond Scientific Co). Twenty-four hours after injection, the follicular cell layer was partially removed by incubation for 30 min with 0.25 mg/ml type-1A collagenase (Sigma). Oocytes were maintained at 15-16°C in sterile Barth's solution, and recordings were made 3 days later.
Confocal Imaging and Fluorescence Resonance Energy Transfer (FRET) Experiment-Transient transfected cells were fixed with 4% paraformaldehyde, PBS for 10 min. Acceptor photobleaching method was used to measure the FRET. Fluorescent proteins from fixed cells were excited with the 458-nm or the 514-nm lines by low excitation intensities and 475-495-nm bandpass and Ͼ530-nm longpass emission filters, respectively. Subsequently, YFP protein was bleached by using maximum laser power, obtaining around 80% of acceptor intensity bleaching. After photobleaching images of the donor and acceptor were taken, FRET efficiency was calculated as ((F CFPafter Ϫ F CFPbefore )/F CFPafter ) ϫ 100, where F CFPafter is the intensity of fluorescence of donor after bleaching, and F CFPbefore is before bleaching. Loss of fluorescence intensity was corrected by measuring CFP intensity in the non-bleached part of the cell. The FRET values are expressed as the means Ϯ S.E. of n Ͼ 10 cells for each population.
Electron Microscopy-Cells were fixed in 4% paraformaldehyde, 0.2% glutaraldehyde in PBS for 30 min at room temperature and then replaced with 2% paraformaldehyde, 0.1% glutaraldehyde in PBS overnight at 4°C. Fixation was removed by washing with 0.02 M glycine in PBS. Cells were scraped and collected in Eppendorf tubes with 12% gelatin. After solidifying on ice, gelatin blocks were infiltrated in 2.3 M sucrose in PBS overnight at 4°C. Blocks were frozen in liquid nitrogen. Ultrathin cryosections were obtained using Leica ULTRACUT EM FCS at Ϫ118°C and subjected to immunogold labeling. Cryosections were incubated at room temperature on drops of 2% gelatin in phosphate buffer for 20 min at 37°C followed by 50 mM glycine in PBS for 15 min and 5% normal goat serum in PBS for 10 min. Then they were incubated with anti-Kv1.5 or and anti-Kv1.3 polyclonal antibodies in 5% normal goat serum in PBS for 30 min. After 3 washes with drops of 5% normal goat serum in PBS for 20 min, sections were incubated for 60 min using IgG anti-rabbit coupled to 10-nm or to 15-nm diameter colloidal gold particles (Aurion) using a 1:60 dilution in 5% normal goat serum in PBS. This was followed by 3 washes with drops of PBS for 10 min and 2 washes with distilled water. As a control for nonspecific binding of the colloidal gold-conjugated antibody, the primary polyclonal antibody was omitted. Finally, the cryosections were contrasted and embedded in a mixture of methylcellulose and uranyl acetate. At the double immunogold, an immunolabeling with one of the primary antibodies was done. In the first immunolabeling procedure we used the secondary antibody conjugated with the small gold (10 nm). Then the inactivation of the anti-IgG binding sites, using 3% paraformaldehyde, 2% glutaraldehyde in PBS for 2 h, was performed (20). After inactivation the immunolabeling with the other primary antibody was done using in this case the secondary antibody conjugated to 15-nm gold particles. Primary antibody dilutions were 1:50 in both cases. Samples were viewed with a Jeol 1010 electron microscope.
Electrophysiological Recordings-Whole-cell currents were measured with the patch clamp technique. An EPC-9 (HEKA) amplifier with the appropriate software was used for data recording and analysis. Currents were filtered at 2.9 kHz. Series resistance compensation was always above 70%. Patch electrodes of 2-4 megaohms were fabricated in a P-97 puller (Sutter Instruments Co.) from borosilicate glass (outer diameter of 1.2 mm and inner diameter of 0.94 mm; Clark Electromedical Instruments Co). Electrodes were filled with the 120 mM KCl, 1 mM CaCl 2 , 2 mM MgCl 2 , 10 mM HEPES, 11 mM EGTA, 20 mM D-glucose adjusted to pH 7.3 with KOH. The extracellular solution contained 120 mM NaCl, 5.4 mM KCl, 2 mM CaCl 2 , 1 mM MgCl 2 , 10 mM HEPES, 25 mM D-glucose adjusted to pH 7.4 with NaOH. After establishment of a whole-cell configuration, macrophages were clamped to a holding potential of Ϫ60 mV, with seal resistances of at least 2.5 gigaohms. All recordings were routinely subtracted for leak currents at Ϫ50 mV online. Only cells with a series resistance compensation of 80 -90% were selected for analysis. Uncompensated series resistances were 4 -8 megaohms, as currents evoked were less than 1nA; voltage errors from uncompensated series resistance were less than 2 mV.
To determine voltage dependence of steady-state activation, currents were elicited by 200-ms voltage pulses (Ϫ50 to ϩ50 mV, V h ) from a holding potential V h of Ϫ60 mV with 10-mV increments. To ensure complete recovery from inactivation, cells were repolarized for 45 s at Ϫ60 mV. After converting the steady-state peak outward currents (I k ) into conductances, G k (G k ϭ I k /(V h Ϫ E k ); E k , Nernst potential of K ϩ Ϫ79 mV), conductances at various membrane potentials were normalized to DECEMBER  Ϫ V)/k) ), where V 1 ⁄2 is the voltage at which the current is half-activated, and k is the slope factor of the activation curve. To calculate inactivation time, constants () cells were held at Ϫ60 mV, and pulse potentials of 4 s were applied. Inactivation adjustment was calculated from the peak of the current at 50 mV to the steadystate inactivation, and traces were fitted with Sigma Plot (SPSS Inc.). To analyze the cumulative inactivation, currents were elicited by a train of 8 depolarizing voltage steps of 200 ms to ϩ 50 mV once every 400 ms.

Role of Kv1.5 in Macrophages
Oocytes were voltage-clamped with a two-electrode system, Gene Clamp 500 (Axon Instruments). The voltage and the current microelectrodes were filled with KCl (3 M) and had resistances ranging from 0.5 to 1 megaohms. The volume of the oocyte recording chamber was 200 l. Recordings were done under constant bath perfusion (1 ml/min). The bath electrode was an Ag-AgCl pellet that made contact with the recording solution (115 mM NaCl, 2 mM KCl, 1.8 mM CaCl 2 , and 10 mM HEPES at pH 7.4) through an agar bridge. Membrane potential and current were digitized through a PCI-MIO-16E-4 Multifunction I/O Board and NI-DAQ (National Instruments). The board was controlled by the Whole Cell Analysis Program (kindly provided by John Dempster, University of Strathclyde). The signal was filtered at twice the acquisition frequency. Oocytes were clamped to a holding potential of Ϫ60 mV. To evoke voltage-gated currents, all oocytes were stimulated with 1-s square pulses ranging from Ϫ80 mV to ϩ80 mV in 10-mV steps. To ensure complete recovery from inactivation, oocytes were repolarized for 60 s at Ϫ60 mV. To calculate inactivation time constants (), pulse potentials of 5 s were applied. Inactivation adjustment was calculated from the peak of the current at ϩ 80 mV to the steady-state inactivation, and traces were fitted with Sigma Plot (SPSS Inc.). Data were leak-subtracted using hyperpolarizing P/4 subtraction pulses. Data were analyzed by the Whole Cell Analysis program.
Statistics-Values are expressed as the mean Ϯ S.E. The significance of differences was established by Student's t test or by analysis of variance (GraphPad, PRISM 4.0) where indicated. A value of p Ͻ 0.05 was considered significant.

Macrophages Express Kv1.3 and Kv1
.5 Channels-Voltagedependent potassium currents were evoked in macrophages by depolarizing pulses. Fig. 1, A and B, show representative potassium currents in BMDM and Raw 264.7 macrophages, respectively. The current density (picoamperes/picofarads)/voltage relationships depicted in Fig. 1C indicate that currents in BMDM were 3-fold higher than in Raw cells. Fig. 1D shows the normalized conductance against the test potential. Although the threshold for activation was about Ϫ40 mV in BMDM, channels opened at Ϫ20 mV in Raw macrophages. Whereas k slopes were similar for both groups (18.3 Ϯ 3 and 11.9 Ϯ 2 for BMDM and Raw, respectively, n ϭ 10), V 1 ⁄2 values were significantly different (Ϫ7.2 Ϯ 2 and 6.3 Ϯ 2 mV for BMDM and Raw, respectively, p Ͻ 0.001, n ϭ 10).
Kv1.3 is the main voltage-dependent K ϩ channel in leukocytes (2). This was confirmed by RT-PCR analysis in our cells. Fig. 1E shows that not only EL-4, a murine T-cell line, but also macrophages (BMDM and Raw) expressed Kv1.3. However, the presence of Kv1.5 was only observed in macrophages. Mouse brain and liver RNAs were used as positive and negative controls, respectively. The presence of Kv1.5 protein in both myeloid cell lines was further confirmed by Western blot analysis performed in crude membrane preparations (Fig. 1F).
Contrary to Kv1.5, Kv1.3 is highly sensible to specific toxins (21). K ϩ currents in BMDM are blocked by Kv1.3 inhibitors such as MgTx and ShK-Dap 22 (8,11). However, the presence of Kv1.5 indicates that either Kv1.5 forms homotetrameric complexes or Kv1.5 subunits are assembled with Kv1.3 in heterotetrameric structures that are sensitive to Kv1.3 toxins. Fig. 1G shows that K ϩ currents in both macrophage cell lines are inhibited by the presence of MgTx. The IC 50 values were 50 Ϯ 5.1 and 772 Ϯ 40 pM for BMDM and Raw, respectively (p Ͻ 0.001, n ϭ 10). MgTx does not block Kv1.5 (21). Therefore, these results indicate that, although Kv1.5 is expressed, this subunit does not form homomeric channels in macrophages. However, Raw and BMDM may differ in the K ϩ channel complex composition since biophysical and pharmacological differences were evident.
Co-localization experiments supported the pharmacological and biophysical data. Unfortunately, immunocytochemistry studies with anti-Kv1.3 and anti-Kv1.5 antibodies were unsuccessful, probably due to high levels of Fc receptors expressed in macrophages. However, since Raw cells have the different Kv1.3 and Kv1.5 intracellular processing and trafficking programs, we transfected these cells with Kv1.3-YFP and Kv1.5-CFP. In general, cells were poorly transfected, but confocal analysis of double-transfected Raw cells demonstrated that Kv1.5 co-localized with Kv1.3 at the macrophage plasma mem-  DECEMBER 8, 2006 • VOLUME 281 • NUMBER 49 brane (Fig. 2, A-D). In this context heteromeric channels could be immuno-detected by electron microscopy. Using this approach we detected oligomeric complexes both at the membrane and inside the cell (Fig. 2, E-H). Interestingly, some immunogold staining was also present at mitochondria (Fig.  2H), in which Kv1.3 has been previously described (22). Although a similar heteromeric pattern was obtained in dou-ble-transfected HEK cells (Fig. 2, I and J), no Kv1.5 immunogold labeling was observed in HEK cells expressing homomeric Kv1.3 channels (data not shown).
Although Kv1.5 is resistant to MgTx, both homomeric Kv1.3 and heterotetrameric Kv1.3/Kv1.5 channels are highly sensitive to this toxin (Fig. 3, L and M). However, although Kv1.3 was inhibited with an IC 50 of ϳ336 pM, hybrid Kv1.3/Kv1.5 channels were fully blocked but were much less sensitive to MgTx (IC 50 ϳ 25 nM). As observed in macrophages (see above), sensitivity to the toxin indicates that Kv1.5 is not expressed as a homomultimeric channel in HEK cells rather than forming heteromeric complexes with Kv1.3. The characteristic C-type inactivation of Kv1.3 is absent in Kv1.5 (21, 23). Thus, the presence of the latter in MgTx-sensible hybrid channels increased the time constant of inactivation in HEK cells (555 Ϯ 40 ms and 832 Ϯ 100 ms for Kv1.3 and Kv1.3/Kv1.5 at ϩ50 mV, respectively, p Ͻ 0.05), further demonstrating heteromeric Kv1.3/Kv1.5 association. In addition, panels on Fig. 3N show that, although the immunoprecipitation with anti-Kv1.3 antibody was not fully effective, Kv1.5 co-immunoprecipitated with Kv1.3 in double-transfected HEK cells. Thus, confocal imaging, immunoprecipitation, and immunogold detection by electron microscope (see Fig. 2J) indicated that Kv1.3 and Kv1.5 Yellow signifies co-localization. J, K ϩ currents generated in doubly transfected HEK cells. To evoke K ϩ currents, HEK cells were held at Ϫ60mV, and pulse potentials were applied as described in Fig. 1. K, plot of normalized conductance versus test potential. Conductance was normalized to the peak current at ϩ50 mV. L, dose-dependent inhibition curves of the K ϩ current by MgTx. Peak currents were evoked at ϩ50 mV, and the percentage inhibition was calculated as described in Fig. 1  co-localized at the membrane and co-assembled, forming a functional heteromeric complex. However, to further confirm this physical association, we undertook FRET analysis (Fig. 4). A confocal section of doubly labeled cells before photobleaching is shown in panels A and B. Once the long-wavelength fluorescence of the acceptor was eliminated (panel C), the donor intensity was measured again (panel D), and FRET efficiency was calculated (panels E and F). The mean FRET efficiency of homotetrameric Kv1.3 channels (Kv1.3-CFP/Kv1.3-YFP) indicated an increment significantly greater than 0% (8%). Similar results were obtained with hybrid Kv1.3/Kv1.5 channels. However, neither Kv1.3-YFP with CFP nor Kv1.5-CFP gave FRET efficiency different from 0%. Our results confirm molecular proximity between Kv1.3 and Kv1.5, as reported by co-immunoprecipitation, co-localization, and electron microscopy.

DISCUSSION
This study demonstrates that Kv1.3 and Kv1.5 can form heterotetrameric structures and suggests that Kv1.3/Kv1.5 hybrid K ϩ channels contribute to the major Kv channel in macrophages. Furthermore, our data also show that the Kv1.3/Kv1.5 ratio may vary in the Kv complex, leading to biophysically and pharmacologically distinct channels. Although heteromeric structures between either Kv1.3 or Kv1.5 and other Kv1 subunits had been described (7, 26 -29), our study demonstrates for the first time that Kv1.3 assembles with Kv1.5 to generate functional Kv complexes. Thus, similar to nerve and muscle, in macrophages it is difficult to assign currents to specific channels. Tissue localization and the absence of currents other than those generated by homomeric channels argue against certain Kv complexes (30 -32). Thus, the assembly of Kv1.3 with Kv1.1 and Kv1.2 was questioned by immunohistochemical localization in mouse rod bipolar cells (30). In addition, a differential distribution also argues against a heteropolymer between Kv1.1 and Kv1.5 in Schwann cells (31). In contrast, Kv1.5 co-assembles with Kv1.2 and Kv1.4 in GH3 pituitary cells, brain cortex, and the cardiovascular system (7,17,26). However, in GH3 cells, up to 70% of the channels were pharmacologically identified as Kv1.5 homomers (17). This is not the case in murine macrophages, since the presence of Kv1.5 in heterotetrameric Kv1.3/Kv1.5 channels induced a shift to positive potentials concomitantly with a loss of sensitivity to MgTx. Unlike Kv1.5, Kv1.3 and hybrids Kv1.3/Kv1.5 are highly sensitive to MgTx. Our data argue against homotetrameric channels consisting of Kv1.5, but the presence of Kv complexes generated only by Kv1.3 cannot be ruled out.
Kv1.3 subunit is the main responsible for outward-delayed rectifier potassium currents in leukocytes (2). Nonetheless, although half-activation voltages of Kv1.3 currents in T-cells range from Ϫ14 to Ϫ35 mV, V 1 ⁄2 values are more depolarized in BMDM and Raw macrophages (Refs. 8 and 11 and this study).
Unlike T-cells, macrophages express Kv1.5, which shows more positive potentials when expressed in heterologous expression systems (10). In brain macrophages, a switch from Kv1.5 to Kv1.3 during cell growth was described, and the half-activation voltage of the Kv currents was shifted ϳ17 mV to hyperpolarized values in proliferating cells (10). This is similar to what we found in BMDM and TNF-␣-activated Raw cells, where negative shifts could be caused by higher Kv1. 3 A change in the Kv1.3/Kv1.5 ratio in the Kv complex would affect kinetic parameters and pharmacology. Because TNF-␣ induces differential regulation of K ϩ channels (8,11,25), our results are consistent with the possibility that the cytokine alters the stoichiometry of the subunits in the channel. In addition, a new composition could also affect the interaction with the Kv␤ subunits in macrophages, thus modifying the gating kinetics (11). Association or lack of association with endogenous ␤-subunits could also affect the phenotype of the resulting current. Our data show that potassium channels from activated cells are more sensitive to small changes in the membrane potential. Thus, these cells would be more excitable at negative potentials. In this context it has been described that a depolarizing change of ϩ25 mV modifies IL-2 production, thus reducing the activation and antibody production of lymphocytes (34). Hormones and cytokines may produce long-term effects on excitability by regulating K ϩ channel gene expression. Although Kv1.3 expression is induced during activation and apoptosis, down-regulation is associated with immunosuppression (18). Dexamethasone, a glucocorticoid antagonist, induces Kv1.5 expression in pituitary cells and cardiomyocytes and inhibits Kv1.3 in T-cells (16 -18). Panyi et al. (35,36) located Kv1.3 in the immunological synapse between cytotoxic and target cells. Activating cytokines would increase the amount of Kv1.3 subunits, forming either homo-or heterotetrameric structures, and changes in Kv composition would modify membrane excitability within the immunological synapse between T-lymphocytes and macrophages. Thus, hormones B, half-activation voltage of K ϩ currents in macrophages. Values were calculated from steady-state activation curves of the outward current. Conductance was normalized to the peak current at ϩ50 mV. C, time constant of inactivation (). Cells were held at Ϫ60 mV, and a ϩ 50 mV pulse potential was applied during 4 s. D, percentage of cumulative inactivation. Currents were elicited by a train of 8 depolarizing voltage steps of 200 ms to ϩ50 mV once every 400 ms. The percentage was calculated as a result of the difference between the peak current at the first pulse and the remaining current at the last. E, IC 50 values of the K ϩ current in the presence of MgTx. Currents were evoked, and the percentage inhibition was calculated as described in Fig. 1. Open bars, BMDM; closed bars, control Raw 264.7 cells; gray bars, TNF-␣-activated Raw 264.7 cells. **, p Ͻ 0.01; ***, p Ͻ 0.001 versus control (Student's t test). and cytokines can affect the expression of cell surface homomeric and heteromeric channels in different ways, leading to specific alteration of excitability.
In summary, our results demonstrate that Kv1.3 and Kv1.5 form heterotetrameric K ϩ channels in macrophages. In addition, the physiological regulation of the Kv channel subunit stoichiometry may be an important mechanism triggering the specific immune response. The findings of the present study are of interest since K ϩ channels in leukocytes are considered pharmacological targets in autoimmune diseases (33), and as we demonstrate, different channel composition may change biophysical properties and alter the use of potential drug therapies. Macrophages present antigens to infiltrating T-lymphocytes, and the predominance of Kv1.3 in activated macrophages could generate more specific drug-sensitive complexes leading to more effective therapies.