Erythropoietin modulates calcium influx through TRPC2.

Mammalian isoforms of calcium-permeable Drosophila transient receptor potential channels (TRPC) are involved in the sustained phase of calcium entry in nonexcitable cells. Erythropoietin (Epo) stimulates a rise in intracellular calcium ([Ca](i)) via activation of voltage-independent calcium channel(s) in erythroid cells. Here, involvement of murine orthologs of classical TRPC in the Epo-modulated increase in [Ca](i) was examined. RT-PCR of TRPC 1-6 revealed high expression of only TRPC2 in Epo-dependent cell lines HCD-57 and Ba/F3 Epo-R, in which Epo stimulates a rise in [Ca](i). Using RT-PCR, Western blotting, and immunolocalization, expression of the longest isoform of mTRPC2, clone 14, was demonstrated in HCD-57 cells, Ba/F3 Epo-R cells, and primary murine erythroblasts. To determine whether erythropoietin is capable of modulating calcium influx through TRPC2, CHO cells were cotransfected with Epo-R subcloned into pTracer-CMV and either murine TRPC2 clone 14 or TRPC6, a negative control, into pQBI50. Successful transfection of Epo-R was verified in single cells by detection of green fluorescent protein from pTracer-CMV using digital video imaging, and successful transfection of TRPC was confirmed by detection of blue fluorescent protein fused through a flexible linker to TRPC. [Ca](i) changes were simultaneously monitored in cells loaded with Rhod-2 or Fura Red. Epo stimulation of CHO cells cotransfected with Epo-R and TRPC2 resulted in a rise in [Ca](i) above base line (372 +/- 71%), which was significantly greater (p < or = 0.0007) than that seen in cells transfected with TRPC6 or empty pQBI50 vector. This rise in [Ca](i) required Epo and extracellular calcium. These results identify a calcium-permeable channel, TRPC2, in erythroid cells and demonstrate modulation of calcium influx through this channel by erythropoietin.

Erythropoietin (Epo) 1 is a hematopoietic growth factor that regulates proliferation, differentiation, and viability of eryth-roid progenitors and precursors (1)(2)(3). Regulation of intracellular calcium ([Ca] i ) by erythropoietin is one of the signaling mechanisms controlling proliferation and differentiation of erythroid cells (4 -10). Evidence implicating calcium in control of erythroid growth and differentiation includes: (a) enhancement of Epo-induced murine erythroid colony growth by the ionophore A23187 and inhibition by treatment with EGTA, a nonspecific chelator of calcium (7); (b) demonstration that an increase in Ca 2ϩ influx is an early and necessary step in the commitment to differentiation of murine erythroleukemia cells (8 -10); and (c) the significant rise in [Ca] i stimulated by Epo observed at specific stages of human BFU-E differentiation (5). Substantial evidence supports the conclusion that erythropoietin stimulates calcium influx in erythroid cells through voltage-independent calcium-permeable channel(s) (8,10,(11)(12)(13). In patch clamp studies of human erythroid progenitor-derived cells, Epo stimulation increased calcium channel mean open time 2.5-fold and open probability 10-fold (13).
Recently, the ability of erythropoietin to activate calcium influx and influence cell proliferation and viability via stimulation of its receptor (Epo-R) in nonerythroid cells has also been demonstrated. Myoblasts have been shown to express Epo-R, and Epo stimulates myoblast proliferation to expand the progenitor population during differentiation (14). In these cells, Epo stimulated an increase in [Ca] i that was entirely dependent on extracellular calcium influx. Erythropoietin receptors have also been identified on neuronal cell lines and Epo stimulated calcium influx in these cells as well (15). Other studies have demonstrated an important neuroprotective and neurotropic effect of erythropoietin on brain tissue (16 -20). Epo stimulated an increase in cell viability in nerve growth factordeprived cells, as well as increases in 45 Ca 2ϩ uptake and [Ca] i . These effects were inhibited by nicardipine, suggesting that Epo may stimulate neuronal function and viability through activation of calcium channels (19). These studies suggest a broader role for Epo as a growth factor capable of maintaining proliferation and preventing apoptosis during differentiation and emphasize the importance of understanding the mechanism of erythropoietin regulation of calcium influx.
A major impediment in determining the mechanisms through which erythropoietin modulates calcium entry and understanding the impact of this on cell growth and differentiation has been the difficulty in identifying and cloning the calcium-permeable channel(s) involved. Recently, a transient receptor potential (TRP) protein superfamily was identified, consisting of a diverse group of calcium-permeable cation channels expressed in nonexcitable cells, based on the archetypal TRP cloned in Drosophila (21,22). Drosophila TRP is predominantly expressed in the visual system, is required for phototransduction, and is coupled via a G protein to phospholipase C. Based on sequences from Drosophila TRP, a large number of mammalian isoforms have been cloned, which have been divided into six subfamilies (22). All mammalian isoforms share six putative transmembrane domains similar to the core structure of many pore-forming subunits of voltage-gated channels except that they lack positive charged residues necessary for the voltage sensor. The classical (22) or short (21) TRP channels (TRPC) were selected for our initial studies, reported here, because these channels have many characteristics similar to the voltage-independent, Epo-modulated calcium-permeable channels identified in human erythroblasts. These TRPC proteins all contain ankyrin repeat domains, which may be important in protein/protein interactions, and amino acid sequence identity greater than 30% in their N terminus but high variability in their carboxyl-terminal region beyond the conserved TRP domain. This diverse family of channels has both storeoperated and receptor-modulated members, which function through intracellular second messenger systems (21,22).
The mechanism of regulation of [Ca] i by erythropoietin has previously been examined at the single cell level using fluorescence microscopy coupled to digital video imaging (4,5,12,13,23). In this report, we used RT-PCR to determine the expression pattern of TRPC on murine erythroid cell lines HCD-57 and Ba/F3 Epo-R, in which Epo stimulates a rise in [Ca] i (23). TRPC2 was the only TRPC detected in these cells under conditions that identified the presence of all TRPC in brain. Two of four TRPC2 splice variants were expressed in these hematopoietic cells. The expression and function of the longest isoform, TRPC2 clone 14 (24), was examined. Cell fractionation and immunolocalization using an antibody specific for TRPC2 clone 14 demonstrated plasma membrane expression. The ability of erythropoietin to modulate calcium influx through TRPC2 was determined using a digital video imaging system in which single cells that expressed transfected Epo-R were identified by detection of green fluorescent protein (GFP), cells that express transfected TRPC were identified by detection of blue fluorescent protein (BFP), and [Ca] i changes were simultaneously measured by Rhod-2 or Fura Red fluorescence.
Generation of Antibody Specific to mTRPC Clone 14 -A rabbit polyclonal antibody was generated to an epitope in the first 100 amino acids unique to the N terminus of mTRPC2 clone 14 (LNQNSTDVLESDPRP-WLTN, aa 79 -98) and affinity-purified. Specificity of the antibody was confirmed using in vitro translation products prepared with cDNAs for mTRPC2 clone 14 (24), mTRPC2 clone 17 (24), and mTRPC6 (28), cloned into pcDNA3. The in vitro products were prepared with the TNT quick coupled transcription/translation system (Promega, Madison, WI).
Immunolocalization of TRPC2 Channels in HCD-57 Erythroleukemia Cells with Anti-mTRPC2 Clone 14 Antibody-HCD-57 cells (3 ϫ 10 5 cells/chamber) were placed in each well of Lab-Tek Permanox Chamber Slides precoated with fibronectin. After 30 min, cells were washed twice with PBS, fixed in methanol at Ϫ20°C for 10 min, and permeabilized in 0.5% Triton X-100 in PBS for 5 min. Incubation for 10 min in 20% normal goat serum preceded staining with primary antibody (anti-TRPC2 clone 14) for 20 min at room temperature followed by secondary antibody (FITC donkey anti-rabbit IgG, Jackson Laboratories, West Grove, PA) for 20 min in the dark. Slides were stained with DAPI in Vectashield mounting medium (Vector Laboratories, Burlingame, CA) to visualize DNA. Cells were viewed using a Nikon Optiphot-2 microscope equipped for epifluorescence. Images were acquired with an air-cooled CCD SenSys digital camera from Photometrics (Tucson, AZ) and processed using IPLab and Enhanced Photon Reassignment software programs obtained from Scanalytics, Inc. (Fairfax, VA).
Immunoblotting of Crude Membrane Preparations-Cell pellets from CHO cells nontransfected or transfected with mTRPC2 clone 14 in pcDNA3 (see below), Ba/F3 Epo-R cells, HCD-57 cells, and Ter-119 ϩ splenic erythroblasts were removed from storage at Ϫ80°C, and 1 ml of Buffer I (10 mM Tris-HCl, pH 7.4, 1ϫ protease inhibitor mixture) was added. Fresh murine brain, heart, kidney, and spleen were homogenized with a Dounce homogenizer on ice to create a cell suspension retaining intact cells (confirmed by microscopy) and centrifuged, and Buffer I was added to each pellet. Lysates were sonicated, and an equal volume of Buffer II (10 mM Tris-HCl, pH 7.4, 300 mM KCl, 20% sucrose, 1ϫ protease inhibitor mixture) was added. Cells were then centrifuged at 10,000 ϫ g for 10 min at 4°C, and the supernatant was spun at 100,000 ϫ g for 1 h at 4°C. Crude membranes were solubilized in buffer containing 62 mM Tris-HCl (pH 6.8), 2% SDS, 10% glycerol. Protein was quantified using the DC Bio-Rad Protein Assay to analyze diluted samples. Conditions for SDS-PAGE and Western blotting with ECL were as described previously (23,29). Electrophoresis was performed on 8% polyacrylamide gels. After transfer, nitrocellulose membranes were incubated with anti-TRPC2 clone 14 (1:500) or anti-mEpo-R (sc697; diluted 1:100; Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Donkey anti-rabbit horseradish peroxidase-conjugated antibody (1:2000) was used as the secondary antibody.
Transfection of mEpo-R and mTRPC into CHO Cells-The pTracer-CMV vector (Invitrogen) containing an SV40 promoter driving expression of a GFP gene and a CMV promoter driving expression of mEpo-R was reported previously (23). TRPC2 clone 14 (24), TRPC2 clone 17 (24), and mTRPC6 (28) in pcDNA3 were subcloned into pQBI50 (QbioGene, Carlsbad, CA). The pQBI50 vector contains a CMV promoter, which drives expression of SuperGlo BFP fused through a flexible linker to TRPC2 or TRPC6. CHO cells at 50% confluence were transfected with pTracer-CMV vector (3 g/ml), pQBI50 vector (3 g/ml), and Lipo-fectAMINE (8 l/ml; Invitrogen) in Opti-MEM 1 for 5 h at 37°C. One ml was added to each 35-mm dish. At 5 h, an equal volume of Dulbecco's modified Eagle's medium with 20% FCS was added, and 18 h later this medium was replaced with Dulbecco's modified Eagle's medium with 10% FCS. Successful transfection of CHO cells with Epo-R and TRPC was verified by detection of GFP (excitation, 478 nm; emission, 535 nm) and BFP (excitation, 380 nm; emission, 435 nm), respectively, in the cells with digital video imaging (4,5,12,13,23). The optimal time for expression of pTracer CMV Epo-R and pQBI50 TRPC2 was 48 -72 h after transfection, and this time interval was selected to examine the response of transfected CHO cells to Epo. At this time, 20 -40% of individual CHO cells expressed both GFP and BFP. Successful transfection was also confirmed by Western blotting using whole cell lysates of nontransfected and transfected CHO cells (23). Anti-TRPC6 was obtained from Alomone Laboratories (Jerusalem, Israel).

Measurement of [Ca] i with Digital Video
Imaging-A fluorescence microscopy-coupled digital video imaging system was used to measure [Ca] i (4,5,12,13,23). To study changes in [Ca] i in transfected cells, we were not able to use Fura-2 as the detection fluorophore because its excitation and emission wavelengths overlap with those of GFP. Instead, we used the fluorescent Ca 2ϩ indicators Rhod-2 (Molecular Probes, Inc., Eugene, OR) (23,30,31) and, in later experiments, Fura Red (32,33). Rhod-2 is a single wavelength excitation Ca 2ϩ fluorophore (excitation, 540 nm; emission, 600 nm), and its fluorescence intensity is proportional to [Ca] i , fluorophore concentration, optical path, and excitation light intensity. The ratio F t (fluorescence at time t) divided by F o (fluorescence at base line) was used to reflect changes in [Ca] i in Rhod-2-loaded CHO cells. CHO cells were loaded with Rhod-2 (2 M, 20 min, 37°C) and stimulated with recombinant erythropoietin (2000 units/ml; Amgen). Rhod-2 fluorescence was measured at base line and at 1, 5, 10, 15, and 20 min after Epo stimulation. In later experiments, to minimize errors associated with fluorphore leakage and variation in lamp intensity, we used Fura Red (excitation, 460 and 490 nm; emission, 600-nm long pass) (32, 33), a dual wavelength excitation probe whose fluorescence intensity ratio is related to [Ca] i . In these experiments, transfected CHO cells were loaded with 5 M Fura Red-AM, in the presence of Pluronic F-127 to enhance loading, for 30 min at 37°C. Epifluorescence collected at 460-nm excitation was divided by that collected at 490-nm excitation to obtain the fluorescence intensity ratio, which was measured at base line and over a 20-min interval as described for Rhod-2. In some experiments, cells were incubated immediately prior to and during Epo stimulation with PBS containing 0.5 mM probenecid (Sigma) to block fluorophore exit from the cell.
[Ca] i measurements were performed in PBS either with (0.7 mM) or without external calcium (2 mM EGTA).

Expression of TRPC in Murine Tissues and Erythroid Cell
Lines Using RT-PCR-The ability of erythropoietin to stimulate calcium influx in murine erythroid cells and erythroleukemia cell lines has been demonstrated previously (7)(8)(9)(10)(11)23). Here, to explore whether this influx occurred through the classical TRP channels, the expression of TRPC1 to -6 was examined in HCD-57 murine erythroleukemia cells and in Ba/F3 Epo-R cells, a hematopoietic cell line stably transfected with murine Epo-R and previously shown to respond to Epo with a rise in [Ca] i (23). Expression was compared with that found in several other murine tissues, and brain was used as a positive control, since most TRPC are expressed in the brain. RT-PCR was performed using RNA isolated from murine brain, heart, kidney, spleen, Ba/F3 Epo-R, and HCD-57 cells. Results are shown in Fig. 1. TRPC2 mRNA was expressed in Ba/F3 Epo-R cells and in HCD-57 cells. No TRPC2 bands were observed when PCR was performed without the reverse transcriptase step, demonstrating that these products did not result from contaminating DNA. The identity of PCR bands was confirmed by sequencing. In contrast, no expression of other classical TRPC was detected in these hematopoietic cell lines.
Four splice variants of murine TRPC2 have been cloned: mTRPC2 clone 14 (24) (GenBank TM accession number AF111108), mTRPC2 clone 17 (24) (accession number AF111107), mTRPC2 ␣ (34) (accession number AF230802), and mTRPC2 ␤ (34) (accession number AF230803). All of these isoforms would be detected by the mTRPC2 primer set used for RT-PCR in Fig. 1. To determine which of these four isoforms are expressed in Ba/F3 and HCD-57 cells, we performed RT-PCR on RNA from brain, Ba/F3 Epo-R, and HCD-57 cells using primers that are capable of distinguishing them based either on sequence differences or size of the PCR product. A schema comparing the cDNAs of the different isoforms and illustrating the strategy of primer selection is shown in Fig. 2A. The primer nucleotide sequences are provided under "Experimental Procedures." Results of RT-PCR are shown in Fig. 2B. RNA for mTRPC2 ␣ and for mTRPC2 clone 14 were found in these hematopoietic cell lines. The identity of PCR bands was confirmed by DNA sequencing. In contrast, mTRPC2 ␤ and mTRPC2 clone 17 were not detected. An appropriately sized PCR product for TRPC2 clone 17 was observed with the clone 17 primers when cDNA for TRPC2 clone 17 (24) was used as the template, demonstrating the ability of this primer set to produce a product when template was available.
Generation of Antibody Specific to mTRPC2 Clone 14 -To further study the expression and function of the longest mTRPC2 isoform, clone 14, an affinity-purified antibody was generated to an epitope unique to the N terminus of mTRPC2 clone 14. To characterize antibody specificity, in vitro translation was performed using cDNAs for mTRPC2 clone 14, mTRPC2 clone 17, and mTRPC6 cloned into pcDNA3. Expression of the appropriate proteins was first documented with 35 S incorporation. These results are shown in Fig. 3A. Despite several attempts, translation of mTRPC2 clone 17 could not be improved, possibly because of the location of the translation start site of this isoform ( Fig. 2A) and/or the absence of a perfect Kozak sequence. In vitro translation reactions were then prepared without 35 S. Western blotting was performed with each of these in vitro translation products with antibody generated to mTRPC2 clone 14 (Fig. 3B) or with antibody to mTRPC6 (Fig. 3C). These results demonstrate the specificity of antibody generated to clone 14.
Membrane Localization of mTRPC2 Clone 14 in Hematopoietic Cell Lines-To determine whether mTRPC2 is expressed in the plasma membrane of hematopoietic cell lines, immunolocalization studies were performed with HCD-57 cells using anti-mTRPC2 clone 14. DAPI staining was used to localize DNA. Nonimmune rabbit serum was used as a control for specificity. Cell staining was visualized by fluorescence microscopy (Fig. 4). Images at different planes through the cell were deconvolved (Scanalytics software) to remove out-of-focus contaminating light to generate high resolution images. Representative results (Fig. 4, c and d) shown here demonstrate that endogeneous mTRPC2 clone 14 protein is localized at or in close proximity to the plasma membrane in these cells.
To confirm the localization of mTRPC2 to the plasma membrane, crude membrane preparations were prepared from nontransfected CHO cells, mTRPC2 clone 14-transfected CHO cells, Ba/F3 Epo-R cells, and HCD-57 cells. To examine the physiological relevance of mTRPC2 clone 14 expression in the Epo-modulated calcium increase in primary erythroid cells, membranes were also prepared from Ter-119 ϩ erythroblasts isolated from the spleens of phenylhydrazine-treated mice. Preliminary studies revealed that Epo stimulates a rise in calcium in these cells (data not shown). Western blotting was performed with protein isolated in the 100,000 ϫ g pellet or supernatant. A protein band of ϳ135 kDa was observed in membrane pellets from mTRPC2 clone 14-transfected CHO cells, Ter-119 ϩ erythroblasts, Ba/F3 Epo-R cells, and HCD-57 cells (Fig. 5A), whereas no equivalent band was observed in nontransfected CHO cells. The quality of our preparations was demonstrated by reprobing blots with antibody to murine Epo-R. Epo-R was predominantly observed in the membrane fraction in Ba/F3 Epo-R cells, HCD-57 cells, and Ter-119 ϩ erythroblasts (Fig.   FIG. 2. RT-PCR of mTRPC2 isoforms in murine hematopoietic cells. A, schema of cDNA for the four mTRPC2 splice variants ␣, ␤, clone 14, and clone 17. Primer locations (see "Experimental Procedures") that distinguish each isoform are indicated by the arrows. Hatched regions indicate DNA sequences unique to mTRPC2 ␣ or clone 17. Presumed ATG start sites, calcium pore regions, and the conserved TRP motif are indicated. B, RT-PCR was performed on RNA isolated from murine brain, Ba/F3 Epo-R cells, and HCD-57 cells using primers specific for each of the four mTRPC2 splice variants. RT-PCR with ␤-actin primers was performed as a control for RNA quality.  (d and h). After deconvolution to remove out-of-focus contaminating light (Scanalytics software), images clearly demonstrate endogenous mTRPC2 expression at or in close proximity to the plasma membrane. 5A). The different molecular weight bands observed for mEpo-R probably result from differential phosphorylation, since HCD-57 cells were grown in the presence of Epo, which would result in a phosphorylated Epo-R, but Ba/F3 Epo-R cells were cultured in IL-3 and Ter-119 ϩ cells were removed from Epo for several hours during the separation procedure. These studies confirm the membrane localization of mTRPC2 clone 14 and Epo-R in erythroid cells.
To determine whether mTRPC2 clone 14 protein expression is restricted to specific cell types, we performed Western blotting on membrane preparations from murine brain, heart, kidney, and spleen (Fig. 5B). No band was detected at 135 kDa in heart, kidney, or spleen, confirming the restricted expression pattern of this isoform. In brain, a band was observed at 135 kDa, consistent with RT-PCR results showing that clone 14 protein is expressed in brain. Two other bands were consistently observed adjacent to the fainter band at 135 kDa and may represent protein modifications or alternative clone 14 splice variants expressed only in brain. A light exposure is shown in Fig. 5B to clearly show the distribution of bands found in the membrane of murine brain.
Transfection of CHO Cells with pTracer-CMV Epo-R and pQBI50 mTRPC2 Clone 14 -To determine whether Epo is capable of regulating calcium influx through mTRPC2, we established a system in which single cells transfected with wild type Epo-R could be identified by GFP fluorescence, cells transfected with mTRPC2 could be identified by BFP fluorescence, and [Ca] i could be measured simultaneously in the identical cells with digital video imaging. CHO cells were used for these transfections, because they lack endogenous Epo-R and have been shown to contain all of the necessary transducers required for a modest growth factor-and Epo-induced [Ca] i increase (23,35). In Fig. 6, nine CHO cells are shown 48 h after cotransfection with Epo-R and full-length mTRPC2 clone 14 (Fig. 6A). Three CHO cells successfully transfected with Epo-R were detected by GFP fluorescence (Fig. 6B). The same three CHO cells also showed BFP fluorescence, indicating successful transfection with mTRPC2 (Fig. 6C). [Ca] i was measured in the same representative cells (indicated by the arrows) with Rhod-2 fluorescence (Fig. 6D). No interference by GFP/BFP was detected under conditions for Rhod-2 or Fura Red (see below) fluorescence measurements.
Expression of transfected mTRPC in CHO cells was further confirmed by immunoblotting. Cell lysates from nontransfected CHO cells or CHO cells transfected with mTRPC2 clone 14 or mTRPC6 were prepared. Western blotting was performed with antibody to mTRPC2 clone 14, and blots were stripped and reprobed with antibody to mTRPC6. Results confirming expression are shown in Fig. 7. The higher molecular weight of both proteins shown here compared with that in reticulocyte lysates (prepared from pcDNA3; Fig. 3) or crude membrane fractions (Fig. 5) is a result of linkage to BFP. These results also confirm that CHO cells express undetectable amounts of TRPC2 clone 14 or TRPC6 orthologs or that the TRPC antibodies fail to cross-react with the hamster TRPC proteins.  5. Membrane localization of mTRPC2 clone 14. A, Western blotting was performed on membranes prepared from nontransfected CHO, CHO cells transfected with mTRPC2 clone 14 in pcDNA3, Ter-119 ϩ splenic erythroblasts, Ba/F3 Epo-R cells, and HCD-57 cells as described under "Experimental Procedures." From the 100,000 ϫ g pellet, 50 g of membrane prepared from CHO cells and Ter-119 ϩ erythroblasts (limited quantity) and 150 g of membrane from Ba/F3 Epo-R and HCD-57 cells were loaded on each lane. Western blotting was performed with anti-TRPC2 clone 14 antibody, and blots were then stripped and reprobed with antibody to Epo-R. The approximate molecular masses (kDa) of mTRPC2 clone 14 and mEPO-R bands are shown on the right of the blot. B, Western blotting was performed on membranes prepared from CHO cells transfected with mTRPC2 clone 14; murine brain, heart, kidney, spleen, and Ba/F3 Epo-R; and HCD-57 cells as controls. Fifty g of membrane prepared from CHO cells and 150 g of membrane prepared from other tissues were loaded on each lane, and the blots were detected with antibody to mTRPC2 clone 14.

Response of CHO Cells Transfected with Epo-R and mTRPC2
to Epo-CHO cells were cotransfected with Epo-R and mTRPC2 clone 14 for 48 -72 h. Other CHO cells were cotransfected with Epo-R and empty pQBI50 vector or vector subcloned with mTRPC6 as controls.
[Ca] i in cells loaded with Rhod-2 was measured before and at intervals for 20 min after Epo stimulation. Results are shown in Table I. Epo stimulation of CHO cells transfected with Epo-R and empty pQBI50 vector demonstrated a modest increase in [Ca] i above base line (137 Ϯ 18%), consistent with our previous observations (23). Epo stimulation of CHO cells cotransfected with Epo-R and mTRPC2 clone 14 resulted in a much larger rise in [Ca] i above base line (372 Ϯ 71%), which was significantly (p Յ 0.0007) greater than that seen in cells cotransfected with mTRPC6 or empty pQBI50 vector. No increase in [Ca] i was seen when CHO cells cotransfected with Epo-R and TRPC2 were stimulated with diluent (PBS), demonstrating the specificity of the Epo response. In cells cotransfected with Epo-R and mTRPC6, which was not found in murine hematopoietic cell lines, the increase in [Ca] i in response to Epo was not statistically different from that observed in cells transfected with Epo-R and empty pQBI50 vector.
To further confirm results, experiments were performed in the presence of 0.5 mM probenecid to minimize fluorophore exit from cells. CHO cells cotransfected with Epo-R and mTRPC2 clone 14 or mTRPC6 were loaded with Rhod-2.
[Ca] i values before and after Epo stimulation are shown in Table II. The significant increase in [Ca] i observed in mTRPC2-transfected cells was maintained and amplified.
To confirm that the [Ca] i increase in response to erythropoietin in transfected CHO cells originated from external calcium influx rather than internal store release, CHO cells transfected with Epo-R and TRPC2 clone 14 were stimulated by Epo in the presence of calcium (0.7 mM) or its absence (2 mM EGTA).
[Ca] i was measured over 20 min in Fura Red-loaded cells (Fig. 8

DISCUSSION
In this report, CHO cells were transfected with wild-type Epo-R and with specific TRPC channels to study erythropoietin signal transduction involving calcium channel activation. This is the first study of receptor-mediated calcium signaling in which successful transfection of both receptor and putative calcium channel were authenticated at the single cell level by fluorescence from different fluorophores, GFP and BFP, and [Ca] i modulation was studied with a third fluorophore, Rhod-2   or Fura Red. This method can be generalized to many receptor/ channel systems.
The first major finding of the present study is that TRPC2 mRNA and protein are expressed on hematopoietic cells. Using three independent approaches, RT-PCR, Western blotting, and immunolocalization, we demonstrated that a classical TRP channel, mTRPC2, is expressed in murine hematopoietic cells. Four TRPC2 splice variants have been reported: mTRPC2 clone 14 (131 kDa) (24), mTRPC2 clone 17 (116 kDa) (24), mTRPC2 ␣ (99 kDa) (34), and mTRPC2 ␤ (100 kDa) (34). Their tissue distribution and function has been controversial (24, 34, 36 -38). High expression of mTRPC2 in testis, sperm, and VNO has been reported (24,36,38), but the presence of expressed sequence tags derived from TRPC2 mRNA in kidney (Gen-Bank TM accession number AA473022) and spleen (GenBank TM accession number AA145678) suggests that mTRPC2 may be present in many other tissues but at low levels. Here, using RT-PCR specific for each of the known mTRPC2 isoforms, we identified expression of mTRPC2 ␣ and mTRPC2 clone 14 but not mTRPC2 ␤ or clone 17 in murine hematopoietic cells. In addition, using a newly characterized antibody specific to the longest TRPC2 isoform, clone 14, we were able to confirm the presence of TRPC2 protein. Our results may differ from previous reports because of the technique (RT-PCR) utilized. Our studies support the possibility that mTRPC2 clone 14 protein expression is restricted to certain cell types including hematopoietic cells and brain. The plasma membrane localization of mTRPC2 in hematopoietic cells is consistent with current and previous functional results (24,36), which require membrane localization of mTRPC2 for calcium influx. Liman et al. (38) also localized rat TRPC2 to the microvillar plasma membrane of VNO receptor neurons.
Erythropoietin has previously been shown to modulate voltage-independent calcium channel(s) (8,12,13,23). The second major finding is that Epo is capable of modulating calcium influx through TRPC2. Epo, external calcium, TRPC2, and Epo-R are all required for a calcium rise in our CHO cell model system (Fig. 8, Tables I and II) (23) comparable with that seen in human BFU-E-derived cells stimulated with Epo, demonstrating the specificity of this response. Because Epo stimulates an increase in calcium in Ba/F3 Epo-R cells (23) and in Ter-119 ϩ erythroblasts (data not shown) in which both TRPC2 and Epo-R are expressed on the plasma membrane, these data demonstrate that TRPC2 is a physiologically relevant candidate for mediating the Epo-stimulated [Ca] i increase (4 -11). In nontransfected CHO cells, we were unable to detect expression of TRPC2 or TRPC6 with either RT-PCR (data not shown) or Western blotting, indicating that CHO cells constitute a good model system to study the interaction between Epo-R and TRPC channels. In addition, the large Epo-stimulated increase in [Ca] i (Ͼ300%) observed in TRPC2-transfected CHO cells, when compared with that observed in cells transfected with TRPC6 or BFP vector alone, is reminiscent of the magnitude (300 -400%) of the Epo-induced [Ca] i increase in human BFU-E-derived erythroblasts (5,12). In previous studies involving CHO cells transfected with Epo-R only (23), the magnitude of Epo-stimulated increase in [Ca] i was quite modest (Ͻ150%), suggesting that either TRPC2 in CHO cells is present in very small amounts not detectable by our RT-PCR conditions or that endogenous Ca 2ϩ -permeable channels in CHO cells and erythroid cells are different.
TRPC2 has been reported to be activated by both calcium store release (24,36) and receptor-operated mechanisms, including activation by the M5 muscarinic receptor (24) and in sperm by the glycoprotein ZP3 in the egg's extracellular matrix (36). TRPC2 has also been shown to have a very important function in rodent pheromone receptor activation in the VNO (37), through a mechanism postulated to involve phospholipase C but not calcium store release, since VNO lack calcium stores (38). Although the mechanisms by which Epo regulates TRPC2 activity were not addressed in this exploratory study, in erythroid cells, we were unable to demonstrate an increase in [Ca] i in the first 2 min after Epo stimulation, arguing against activation of TRPC by depletion of calcium stores in these cells. Of note, two other TRP family channels have also been shown to be regulated by growth factors: (a) GRC, a mouse homologue of VRL-1, by insulin-like growth factor-1 through regulation of membrane trafficking (39) and (b) TRPC3 by brain-derived nerve growth factor through activation of the neurotrophin receptor TrkB and phospholipase C (40).
In summary, we have shown plasma membrane expression of both mTRPC2 clone 14 and Epo-R in erythroid cells. Using CHO cells doubly transfected with Epo-R and TRPC channels, we demonstrated the ability of Epo-R to regulate calcium influx through mTRPC2 but not mTRPC6. Whereas our data clearly demonstrate a role for mTRPC2, it is probably not the only TRP channel of importance in hematopoietic cells. LTRPC2 is expressed on hematopoietic cells, has an important role in calcium influx in immunocytes, and is involved in tumor necrosis factor-␣-induced cell death (41,42). Sequence homology analysis using the Kimura pairwise near neighbor approach has placed mTRPC2 near the TRPM family channels LTRPC1 and LTRPC2 (24). Several TRP channels have been shown to form heteromultimers in vivo (22,43,44). LTRPC2 is a candidate calcium channel for regulation by hematopoietic growth factors and for interaction with mTRPC2. Important challenges will be to identify the other TRP channels expressed on hematopoietic cells, including ones that have not yet been cloned, and to determine the regulation and function of TRP homo-and heteromultimeric channels in hematopoietic growth factor-regulated proliferation, differentiation, and cell survival.