An elevation in cytosolic calcium concentration ([Ca2+]i) plays a central role in the physiologic activation of platelets. Although a number of calcium entry pathways are believed to exist in human platelets,1 information on the identity of the channels concerned is limited. The recent paper by Hassock and colleagues,2 which attempts to characterize the expression of homologues of the Drosophila transient receptor potential channel (TRPC) mutant in platelets, is thus valuable in this respect. However, the data reported by Hassock et al conflict with other published data in several respects. Until these discrepancies are resolved much uncertainty will remain.
Hassock et al provide evidence for at least 2 Ca2+ entry pathways in the platelet plasma membrane, one dependent and another independent of depletion of the intracellular Ca2+stores.2 The store-independent entry is suggested to be activated by diacylglycerol (DAG) and mediated by TRPC6. Activation of this nonselective pathway by the DAG analog 1-oleoyl-2-sn-glycerol (OAG) and by thrombin was demonstrated, and was suggested to be independent of protein kinase C (PKC) because the PKC inhibitor bisindolylmaleimide was without effect. The divalent cation entry evoked by OAG was modest and slow. We have also reported the activation of Ca2+ entry in platelets independently of Ca2+ store depletion.3 As in the experiments of Hassock et al, this pathway was nonselective and could be activated by OAG, phorbol-12-myristate-13-acetate, and thrombin. However, in contrast to Hassock and colleagues, we found that the above agents evoked rapid divalent cation entry that was blocked by the PKC inhibitor Ro-31-8220. Although the presence of a PKC-stimulated Ca2+ entry pathway seems at odds with the well-established inhibitory (negative feedback) effects of PKC on platelet Ca2+ signal generation,4 we have shown that the effect of PKC stimulation on platelet [Ca2+]i is time dependent. Ca2+entry was observed on initial PKC stimulation, but inhibition was observed after longer treatments.3
Hassock and colleagues conclude that the store-independent pathway suggested to involve TRPC6 is distinct from the store-dependent pathway on the basis of selectivity experiments. They suggest that the store-dependent pathway is selective for Ca2+ and does not admit Ba2+, in contrast to the store-independent route. This conflicts with other reports that store depletion using low concentrations of ionomycin5 or thapsigargin6stimulated the entry of Ba2+, Sr2+, and Ca2+ across the plasma membrane. The use of Ba2+ alone in such studies is inadvisable since it has relatively little effect on fura-2 fluorescence compared with Ca2+.7 Also, Ba2+ blocks many types of potassium channels,8 so membrane depolarization may account for reduced cation influx.
As well as TRPC6 expression, Hassock and colleagues demonstrate the expression of TRPC1 in human platelets. However, they report a low level of detection and suggest that TRPC1 is located in the inner-membrane rather than plasma-membrane fraction after membrane separation by free-flow electrophoresis. Thus they suggest against a role for TRPC1 in store-mediated Ca2+ entry. These results contrast with our own in at least 2 respects. First, we can readily detect TRPC1 in human platelets using an anti-TRPC1 antibody from Alomone Laboratories (Jerusalem, Israel)9,10 and have confirmed the detection of the same protein of about 100 kDa using the T1E3 anti-TRPC1 antibody characterized by Xu and Beech.11 Furthermore, we have demonstrated de novo coupling of TRPC1 and the type II inositol trisphosphate receptor (InsP3RII) when the intracellular Ca2+ stores are depleted.9,10 This has led us to propose that the activation of store-mediated Ca2+entry (SMCE) in platelets occurs by a secretionlike coupling mechanism involving TRPC1 and InsP3RII.9,10 The anti-TRPC1 antibody from Alomone is raised against the extracellular amino acid sequence 557 to 571, which is predicted to lie in the pore-forming region of the protein. In accordance with this, we have reported that this anti-TRPC1 antibody blocks both Ca2+ and Mn2+ entry evoked following Ca2+ store depletion using thapsigargin.12 This strongly supports a role for TRPC1 in SMCE in human platelets and furthermore suggests that TRPC1 must be located at least in part in the plasma membrane.
It is difficult to explain the conflicts between our data and those of Hassock et al. One factor may be the age of the cells. We used freshly isolated platelets prepared with minimal handling and conducted all experiments within a few hours of venipuncture. Hassock et al prepared membranes by free-flow electrophoresis using older cells obtained via blood banks, which necessarily have to be subjected to chemical treatment. Another factor may be the specificity of the antibodies concerned. In our hands, the anti-TRPC1 from Alomone and the T1E3 antibody (a gift from Prof D. J. Beech, University of Leeds, United Kingdom) recognize the same single band of about 100 kDa. The anti-Xenopus TRPC1 antibody used by Hassock et al to assess TRPC1 distribution between the inner and plasma membranes detected multiple protein bands, many more strongly than that of the predicted size of TRPC1.
Hence we believe much remains to be done to characterize Ca2+ entry pathways in platelets.
TRPC channels and Ca2+ entry in human platelets
Ca2+ entry is an important event in platelet activation but little is established regarding the details of the molecular components involved. Store-operated Ca2+ entry (SOCE) represents a major pathway, in addition to SOCE-independent mechanisms, that may involve direct activation by surface receptors or the involvement of second messengers.1-1 However, the identities of the entry channels are not known, and from the known properties of the SOCE pathway the molecular composition of the SOCE channels may differ between cells. Members of the transient receptor potential (TRP) family, of which there are 3 subfamilies (TRPC, TRPM, and TRPV), have been suggested as candidates for SOCE and non-SOCE channels.1-2 In our recent study1-3 we examined the expression and role of the TRPC family in human platelets. We reported the strong expression of TRPC6 located in the plasma membrane (PM) and suggested that it forms a SOCE-independent Ca2+entry channel. We also reported the low expression of TRPC1 that was found in intracellular membranes (IMs). Our results contrast with data published by Rosado and Sage1-4,1-5 who, using an anti-TRPC1 antibody from Alomone Laboratories (Jerusalem, Israel), reported strong expression of TRPC1 and its coupling to the type II inositol 1,4,5-trisphosphate receptor (IP3R) upon store depletion. In the preceding letter, Sage and colleagues highlighted these discrepancies and suggested that age of cells and specificity of antibodies used may explain the differences. We believe that our data are totally in line with the known properties of TRPC proteins. We reject the idea that age of cells could contribute to differences in the results obtained. All of our studies concerning Ca2+ measurements, phosphorylation, immunoprecipitation, and overexpression are carried out with freshly isolated cells. Only the data on purified PM and IM preparations are obtained using one-day-old platelets from the blood bank (with the delay arising because of the compulsory testing of pathogens), because it is simply not ethical to take one liter of blood from laboratory colleagues. On closer examination of the studies involved, we suggest that the high doses and poorly established specificity of the reagents used by the Sage group, in addition to methodological differences, may better explain the discrepancies involved.
In our study we presented evidence that, in line with the expression of TRPC6 in the PM, 60 μM 1-oleoyl-2-acetyl-sn-glycerol (OAG) stimulated Ca2+/Ba2+ entry that was essentially independent of protein kinase C (PKC), as the inhibitor bisindolylmaleimide I (Bis I) had no effect on the OAG-induced entry. Further, in line with the known negative-feedback properties of PKC,1-6 Bis I enhanced 1 U/mL thrombin-stimulated Ba2+ entry. Bis I totally inhibited OAG-induced platelet aggregation confirming this response to be PKC dependent. Under control conditions, a faster entry of Ba2+ by thrombin compared with OAG may be explained by the correct diacylglycerol (DAG) produced by thrombin in a tightly coupled system. Activation of Ba2+entry was independent of store depletion and thapsigargin (Tg) was a poor stimulator for Ba2+ entry, under conditions where it induced Ca2+ entry.1-3 Our results are entirely in agreement with the published properties of TRPC6,1-7 where TRPC6 is described as a store-independent, nonselective channel, activated by DAG in a membrane-delimiting manner. In contrast, Rosado and Sage1-8 reported that high concentrations of OAG and of phorbol-12-myristate-13-acetate (PMA) stimulated Ca2+ and Sr2+ entry in a PKC-dependent manner as the entry was blocked by the inhibitor Ro-31-8220. The PMA concentration used was 1 μM at a cell density of 1 × 108 platelets/mL. In our hands 100 nM PMA is sufficient to maximally activate PKC even with normal cell counts of 2 ×108/mL to 3 × 108/mL. It is possible that overstimulation of PKC may induce cation entry. However, the possibility that the higher concentrations of these agents used with low cell numbers may alter the cell integrity and activate entry of cations needs to be addressed. There is no doubt that PKC provides a negative-feedback role for the control of Ca2+ elevation as confirmed in our study. Sage and colleagues caution against the use of Ba2+, which may possibly block certain potassium channels. On the contrary, measurement of Ba2+ entry under defined conditions has proved very useful in monitoring the gating properties of many TRPC proteins, for example.1-9 While the role of K+ channels in Ca2+ fluxes in platelets needs to be better clarified, our experiments under identical conditions indicate that thrombin is a powerful mediator of Ba2+ entry but Tg is not. Sage and colleagues further suggest that store depletion by low concentrations of ionomycin and Tg stimulates the entry of Ca2+, Ba2+, and Sr2+, citing their study by Jenner et al.1-10 However, closer examination of this study reveals a vastly different methodology from that which we used. Very high levels (500 μM) of ionomycin were added to platelets for 20 minutes, followed by 2 centrifugations, incubation with cations, 2 further centrifugations, and analysis of cation release.1-10Our studies were carried out in spectrofluorimeter cuvettes monitoring fluorescence changes over 9 minutes1-3 and clearly showed a marked selectivity for Ca2+ more than Ba2+ upon store depletion.
Perhaps the largest difference concerns the expression and role of TRPC1. We have used 2 antibodies to TRPC1 (Ank and anti-XTRP1), which we show recognize hTRPC1 overexpressed in QBI-293A cells. These antibodies show a low level of expression of TRPC1 in platelet IM. These data have received support from 2 important studies recently published. Firstly, Mori and colleagues1-11 have shown that the knock-out of TRP1 (avian TRPC1) in the hematopoietic DT40 cell line resulted in a reduction of agonist-mediated release of Ca2+from the stores and a reduction of IP3-induced Ca2+ release from the endoplasmic reticulum. These results would be incompatible if all of the TRP1 was located in the PM. Further, Hofmann et al1-12 have defined the subunit composition of TRPC channels. They report that TRPC1 can exist as a heterotetramer with TRPC4 or TRPC5 but not with TRPC3, TRPC6, or TRPC7. Additionally, they report that TRPC1 stayed in membrane compartments in the cytoplasm unless it was coexpressed with either TRPC4 or TRPC5, in which case it located to the PM. We have shown that platelets express little or no TRPC4 or TRPC5, hence the location of TRPC1 in the IM entirely supports the findings of Hofmann et al.1-12 Our results do not rule out a more important role for TRPC1 in the SOCE activity of other cells (Hassock et al1-3) where the subunit composition of the TRP channels need to be determined. Rosado and Sage1-4,1-5 have used the anti-TRPC1 antibody from Alomone. The manufacturer's website (http://www.alomone.com/Site/p_home/home.htm) states that in rat brain it recognizes 2 products: a protein larger than 250 kDa and more faintly a protein at approximately 120 kDa. The molecular size of hTRPC1β is 80 kDa and the full length isoform is 34 amino acids longer.1-3 Recently Ong et al1-13 showed that the Alomone antibody did recognize a 120 kDa protein in mouse liver and mouse brain. However, it did not recognize overexpressed hTRPC1 under conditions in which a number of other antibodies did, including the anti-XTRP1 antibody used in our study. Therefore the identity of the proteins recognized by the Alomone antibody needs to be clarified. Using the Alomone anti-TRPC1 antibody, we have been unable to reproduce the findings of Rosado and Sage,1-4,1-5 although we cannot rule out variations of antibody specificity between different batches supplied by the manufacturer. We therefore suggest that results obtained with the Alomone anti-TRPC1 antibody should be verified with other better-established antibodies.
In conclusion, we feel that our work represents an important advance in our knowledge of the expression and role of TRPC proteins in platelet Ca2+ homeostasis. We have demonstrated the expression of TRPC6 and its role as a SOCE-independent Ca2+ entry channel in platelets. Clearly much remains to be determined regarding the molecules and mechanisms involved with the SOCE pathway in platelets.