|Home | About | Journals | Submit | Contact Us | Français|
A delayed rectifier voltage-gated K+ channel (Kv) represents the largest ionic conductance of platelets and megakaryocytes, but is undefined at the molecular level. Quantitative RT-PCR of all known Kv α and ancillary subunits showed that only Kv1.3 (KCNA3) is substantially expressed in human platelets. Furthermore, megakaryocytes from Kv1.3−/− mice or from wild-type mice exposed to the Kv1.3 blocker margatoxin completely lacked Kv currents and displayed substantially depolarised resting membrane potentials. In human platelets, margatoxin reduced the P2X1- and thromboxaneA2 receptor-evoked [Ca2+]i increases and delayed the onset of store-operated Ca2+ influx. Megakaryocyte development was normal in Kv1.3−/− mice, but the platelet count was increased, consistent with a role of Kv1.3 in apoptosis or decreased platelet activation. We conclude that Kv1.3 forms the Kv channel of the platelet and megakaryocyte, which sets the resting membrane potential, regulates agonist-evoked Ca2+ increases and influences circulating platelet numbers.
Ion channels are a large and diverse family of transmembrane proteins that play important roles in all cell types. Their functions in the platelet are poorly understood (reviewed in Mahaut-Smith, 2004), although it is clear from patch clamp studies of platelets and megakaryocytes that the largest amplitude ionic currents are conducted through voltage-gated K+-selective (Kv) channels (Maruyama, 1987; Kawa, 1990; Kapural et al. 1995; Romero & Sullivan, 1997). These channels activate on depolarisation to potentials positive to −60 mV, are steeply voltage dependent over the range −40 to −10 mV and are maximally activated at potentials above 0 mV (Maruyama, 1987). They open and close with a relatively slow timecourse and are therefore comparable to the ‘delayed rectifier’ K+ channels of excitable tissues that contribute to action potential repolarisation. In the platelet, this K+ conductance could serve to stabilise the membrane potential at rest or following influx of Ca2+ or Na+ through agonist-evoked channels such as Orai1, P2X1 and TRPC6 (Varga-Szabo et al. 2009). Voltage-dependent K+ channels also play crucial roles in volume regulation and cell proliferation of lymphocytes (Lewis & Cahalan, 1995; Chandy et al. 2004). However, the molecular composition of the voltage-gated K+ channel(s) in platelets, and their precursor cell the megakaryocyte, is unknown. Pharmacological studies (Maruyama, 1987; Kawa, 1990; Romero & Sullivan, 1997) indicate that one or more members of the Kv1 or Kv3 families could contribute, as reported for lymphocytes (Grissmer et al. 1992; Lewis & Cahalan, 1995). Here we show for the first time that the voltage-gated K+ channel of the platelet and megakaryocyte is formed by Kv1.3 subunits, with no evidence for a significant contribution from K+ channel subunits of other Kv families. We also show that the channel is not essential for megakaryocyte development, but that it influences the number of circulating platelets and promotes agonist-evoked increases in intracellular Ca2+, a key second messenger during platelet-dependent thrombosis.
Standard external saline contained (in mm): 145 NaCl, 5 KCl, 1 CaCl2, 1 MgCl2, 10 Hepes, 10 d-glucose, pH 7.35 with NaOH. CaCl2 was omitted for nominally Ca2+-free saline. The patch pipette saline contained (in mm): 150 KCl, 2 MgCl2, 10 Hepes, 10 d-glucose, pH 7.2 with KOH. Acid citrate dextrose (ACD) contained (in mm): 85 trisodium citrate, 78 citric acid, 111 glucose. Fura-2 was from Molecular Probes-Invitrogen (the Netherlands). Margatoxin and apyrase (type VII) were from Sigma (Poole, Dorset, UK).
Marrow was removed from the femoral and tibial marrow of C57/bl6 or Kv1.3−/− mice by flushing with standard saline containing 0.32 U ml−1 apyrase. For immunohistochemical studies, clumps of marrow were immediately frozen in Tissue-Tek® O.C.T.™ compound (Sakura, the Netherlands). For electrophysiological recordings, marrow was gently triturated to disperse the cells and maintained on a rotor at room temperature for use within 12 h. The generation of Kv1.3−/− mice (C57bl/6 background) has been described previously (Koni et al. 2003); control mice (bred in-house at the University of Leicester or from Charles River, UK) were matched for age and sex. For studies of human platelets, standard phlebotomy techniques were used to draw blood from informed, consenting donors according to a protocol approved by the local ethics committees of the University of Leicester and the University of Lund. For intracellular Ca2+ measurements, blood was anti-coagulated with ACD, platelet-rich plasma (PRP) prepared by centrifugation at 700 g for 5 min and washed platelet suspensions prepared by centrifugation at 350 g for 20 min. Platelets were treated with aspirin (100 μm) and apyrase (0.32 U ml−1). For cDNA preparation, whole human blood was collected into ACD supplemented with 2 mm EDTA, 0.1 μm PGE1 and 300 μm aspirin as described elsewhere (Amisten et al. 2008). Mouse blood was withdrawn into ACD by cardiac puncture under terminal gaseous anaesthesia according to UK Home Office guidelines. All procedures within this study were performed in accordance with ethical standards as outlined in Drummond (2009).
Ratiometric Ca2+ measurements from washed suspensions of fura-2-loaded platelets were conducted as described in detail previously (Rolf et al. 2001) using a Cairn cuvette spectrophotometer (Cairn Research Ltd, Faversham, UK). Platelets were loaded with fura-2 by incubation of PRP with 2 μm fura-2 AM for 45 min at 37°C and initially resuspended in nominally Ca2+-free saline in the presence of apyrase (0.32 U ml−1). CaCl2 was added to obtain external Ca2+ as required by the specific protocol. Fura-2 was calibrated extracellularly using a Kd of 224 nm following release by digitonin (50 μm) and determination of Rmin and Rmax values in Ca2+-free (10 mm EGTA) or Ca2+-containing (2 mm CaCl2) salines at neutral pH. All fura-2 recordings were corrected for background fluorescence, determined by quenching with MnCl2.
Conventional whole-cell patch clamp recordings were conducted using an Axopatch 200B, as described in detail elsewhere (Mahaut-Smith, 2004). Megakaryocytes were identified by their large size and multilobed nucleus (Mahaut-Smith, 2004). For voltage clamp recordings, series resistance (Rs) compensation of ≥70% was applied. Whole-cell capacitance, which is a quantitative measurement of the megakaryocyte platelet-generating demarcation membrane system (Mahaut-Smith et al. 2003), was read directly from the patch clamp amplifier following compensation of the current transients evoked by a 5 mV voltage step from −80 mV, a potential range that does not activate voltage-gated channels in these cells (Mahaut-Smith et al. 2003; Mahaut-Smith, 2004).
Murine platelets in ACD-anticoagulated whole blood were counted by flow cytometry according to the method developed by Alugupalli et al. (2001). Briefly, platelets were labelled with a FITC-conjugated anti-CD41 antibody (BD Pharmingen; 1 in 300 dilution) for 1 h and counted at a final dilution of >1000 in Ca2+-free (1 mm EGTA) standard platelet saline in the presence of 5.5 μm diameter fluorescent beads of known density (SPHERO ACFP-50-5, Spherotech, Lake Forest, IL, USA). Fluorescence was measured at low rate in a single channel (488 nm excitation) of a FACSCalibur flow cytometer (BD Biosciences), which provided clear separation of platelets, beads and unlabelled cells (see online Supplementary figure). The platelet count was calculated from the ratio of beads to platelets, the dilution factor, and the density of beads.
Platelets were purified and cDNA was generated as described elsewhere (Amisten et al. 2008). qPCR was performed using Quantifast SYBR Green PCR kit and QuantiTect Primer Assays (Qiagen, Venlo, the Netherlands) in a Roto-Gene 2000 thermal cycler (Corbett Life Science, NSW, Australia) according to the manufacturers’ instructions. Gene expression was calculated according to the ΔΔCt method with GAPDH as a reference (Pfaffl, 2009).
Sections 12 μm thick were cut from clumps of marrow frozen in Tissue-Tek® O.C.T.™ compound and thaw-mounted on slides. Sections were incubated in FITC-conjugated rat antimouse CD41 monoclonal antibody (BD Pharmingen) for 1 h and washed prior to analysis on an Olympus IX81 FV1000 confocal microscope. Fluorescence images (488 nm excitation, >500 nm emission) were acquired of randomly selected 0.1 mm2 fields of view (six from each of 7 WT and 7 Kv1.3−/− mice). Olympus FV1000 analysis software was used to draw around the periphery of stained cells and thus to compute megakaryocyte area.
Statistical significance was assessed using either Student's t test (paired for intracellular Ca2+ recordings and unpaired for membrane capacitance, megakaryocyte size distribution and platelet count) or one-way analysis of variance (membrane potential). Significance is indicated at levels of 0.05 (*), 0.01 (**), 0.005 (***) or 0.001 (****).
To identify components of the voltage-dependent K+ conductance of the human platelet, we used real-time PCR to screen purified platelet cDNA for transcripts of all known α (pore-forming) and β or other ancillary subunits of the Kv family of ion channels (51 targets, see Supplementary Table 1). A single α subunit transcript was detected, for the gene KCNA3, which encodes Kv1.3. The only other subunits detected were for KCNAB2 and KCNE3 (which encode Kvβ2 and Mirp2, respectively), but both were expressed at extremely low levels (<1% of Kv1.3; Fig. 1A). The primary megakaryocyte is an authentic surrogate for electrophysiological studies of the small and fragile platelet (Tolhurst et al. 2005), therefore we used whole-cell patch clamp to investigate Kv currents of megakaryocytes from wild-type and Kv1.3-deficient mice. Using pseudophysiological internal and external salines, voltage steps from a holding potential of −80 mV activated a large transient outward current at potentials positive to about −40 mV in wild-type megakaryocytes (Fig. 1B top panel and Fig. 1C), with a mean magnitude of 11.8 ± 1.0 nA at 0 mV (n= 14). This voltage-gated outward current was totally absent in Kv1.3-deficient megakaryocytes (Fig. 1B centre panel and Fig. 1C), consistent with the quantitative PCR screen and implies that this channel is a homomultimer of Kv1.3. In contrast to murine lymphocytes (Koni et al. 2003), we failed to detect any compensatory anion currents in more than 30 recordings from Kv1.3−/− megakaryocytes. Margatoxin (10 nm) is a relatively selective inhibitor of Kv1.3 (Chandy et al. 2004; Gutman et al. 2005) and this blocked the outward currents (Fig. 1B lower panel and Fig. 1C), further supporting the conclusion that Kv1.3 channels mediate the voltage-gated K+ conductance of the megakaryocyte and platelet.
In whole-cell current clamp recordings, the average membrane potential (Vm) of unstimulated megakaryocytes from wild-type mice was −46.6 ± 2 mV, n= 7 (Fig. 2B), around which regular spontaneous fluctuations of ~2–3 mV were observed (see sample recording in Fig. 2A). Margatoxin (10 nm) slowly depolarised this resting potential over a timecourse that varied between cells (see, for example, Fig. 2A), taking between 4 and 8 min to reach a stable Vm of −14.7 ± 3 mV (n= 7; P < 0.001; Fig. 2B). A similar range of timecourses was observed for the block of voltage-dependent outward currents by 10 nm margatoxin in voltage clamp experiments (Fig. 1C). The margatoxin-induced depolarisation was slowly, and only partially, reversible (not shown), consistent with the slow wash-off (>20 min) reported previously for this toxin (Garcia-Calvo et al. 1993). However, the depolarisation caused by margatoxin was specifically related to block of Kv1.3 as there was no significant loss of membrane potential during the first 8 min of whole-cell recording in untreated wild-type megakaryocytes (−43.9 ± 0.9 mV, P > 0.05; Fig. 2B). The Vm of Kv1.3-deficient megakaryocytes was also substantially depolarised, on average to −10 ± 2 mV (n= 5; P < 0.001, Fig. 2B), further supporting a major role for this channel in the resting potential.
Kv1.3 may influence platelet activation by increased agonist-evoked Ca2+ influx, through maintenance of the initial driving force for Ca2+ entry; furthermore, influx of cations will substantially depolarise the cell if counter-ion movement is not provided. In fura-2-loaded human platelets, margatoxin significantly reduced the peak Ca2+ increase following stimulation of P2X1 (1 μmαβmeATP) and thromboxaneA2 (500 nm U46619) receptors to 62 ± 10% and 76 ± 8% (P < 0.05) of control, respectively (Fig. 3A). In addition, margatoxin delayed the initial phase of store-operated Ca2+ influx, tested by addition of external Ca2+ after 10 min exposure to thapsigargin. The [Ca2+]i increase stimulated after 10 s was reduced to 69 ± 8% (P < 0.05) of control; however, there was no significant effect on the peak Ca2+ increase (Fig. 3B). Together, these results suggest that Kv1.3 promotes early Ca2+ influx through multiple pathways stimulated during platelet activation, including P2X1 ionotropic receptors and store-operated Ca2+ channels. An increase in [Ca2+]i is a key signal during platelet activation (Varga-Szabo et al. 2009) and in vivo studies have demonstrated important roles for P2X1 and Orai1 store-operated Ca2+ channel subunits in arterial thrombosis. Thus, Kv1.3 may also facilitate haemostasis and thrombosis; however, further work is required to address this issue. Direct interactions between Kv channels and integrins have been described in several cell types, including lymphocytes (Levite et al. 2000); thus, Kv1.3 may influence platelet aggregation and adhesion independently of effects on intracellular Ca2+. Additional studies are also required to determine which other channels act as counter-ion pathways for Ca2+ influx in the absence of Kv1.3; however, intermediate conductance Ca2+-activated K+ channels previously reported in human platelets are a likely candidate (Mahaut-Smith, 1995).
A role for Kv channels in thymocyte development has been described (Freedman et al. 1995), but no differences could be detected between Kv1.3−/− and age/sex-matched wild-type mice in the size distribution (Fig. 4A; P > 0.05) of bone marrow megakaryocytes, suggesting that megakaryocyte development is normal. In addition, the high specific membrane capacitance of the megakaryocyte, which reflects the amount of demarcation membrane system (Mahaut-Smith et al. 2003), was unaffected by loss of Kv1.3 (8.19 ± 0.24 μF cm−2 for WT and 7.98 ± 0.42 μF cm−2 for Kv1.3−/−, P > 0.05; Fig. 4B). This plasma membrane invagination system provides the additional membrane required for the process of thrombopoiesis (Schulze et al. 2006); therefore these data indicate that the platelet-generating capacity of the megakaryocyte is normal in Kv1.3−/− mice. In contrast, platelet counts were significantly increased in Kv1.3−/− mice compared to control mice (1.47 × 106vs. 1.12 × 106μl−1; n= 7, P < 0.005; Fig. 4C). The underlying basis of this increased platelet count is unknown and currently under investigation. One possibility is the role of mitochondrial Kv1.3 channels in apoptosis, as recently described in lymphocytes (Szabo et al. 2008), since platelet lifespan is controlled by an intrinsic programme of apoptosis (Mason et al. 2007). Alternatively, the reduced Ca2+ responses of Kv1.3-deficient platelets may decrease activation and thereby increase survival of circulating platelets. This is opposite to the situation in mice that express a constitutively active Stim1 mutant, whose platelets have elevated cytosolic Ca2+ levels and shorter circulation life-time leading to thrombocytopenia (Grosse et al. 2007).
In conclusion, this study demonstrates for the first time that Kv1.3 channels are responsible for the major K+ conductance and the resting potential of the platelet. Consequently, blockade of Kv1.3 reduces Ca2+ entry through a number of different agonist-stimulated pathways. Kv1.3 is known to play roles in immune responses, olfaction and glucose homeostasis (Fadool et al. 2004; Xu et al. 2004; Cahalan & Chandy, 2009) and proposed as a target for treatment of multiple sclerosis (Rangaraju et al. 2009). The present study extends the roles of Kv1.3 to the platelet where it influences the membrane potential, Ca2+ responses and circulating platelet numbers.
This work was funded by the British Heart Foundation (PG 06/028 and 05/014) and the Medical Research Council. We thank Gwen Tolhurst and Richard Carter for their participation in preliminary studies.
Experiments were carried out in the Department of Cell Physiology and Pharmacology and the MRC Toxicology Unit, University of Leicester. C.M. conducted and analysed the patch clamp, intracellular Ca2+ and megakaryocyte size distribution measurements; C.M., S.J., M.P.M.-S., A.H.G. and R.T.S. conducted and analysed the murine platelet counts; S.J. prepared all platelet samples except those used in qPCR studies; S.A. and D.E. conducted and analysed the platelet qPCR studies; I.D.F., S.J. and A.H.G. contributed essential discussion; I.D.F. and L.K. contributed essential materials; M.P.M.-S. designed the research and wrote the paper, which was approved by all authors.
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors