Molecular and Functional Characterization of a Novel Mouse Transient Receptor Potential Protein Homologue TRP7

Characterization of mammalian homologues ofDrosophila transient receptor potential protein (TRP) is an important clue to understand molecular mechanisms underlying Ca2+ influx activated in response to stimulation of Gq protein-coupled receptors in vertebrate cells. Here we have isolated cDNA encoding a novel seventh mammalian TRP homologue, TRP7, from mouse brain. TRP7 showed abundant RNA expression in the heart, lung, and eye and moderate expression in the brain, spleen, and testis. TRP7 recombinantly expressed in human embryonic kidney cells exhibited distinctive functional features, compared with other TRP homologues. Basal influx activity accompanied by reduction in Ca2+ release from internal stores was characteristic of TRP7-expressing cells but was by far less significant in cells expressing TRP3, which is structurally the closest to TRP7 in the TRP family. TRP7 induced Ca2+ influx in response to ATP receptor stimulation at ATP concentrations lower than those necessary for activation of TRP3 and for Ca2+ release from the intracellular store, which suggests that the TRP7 channel is activated independently of Ca2+ release. In fact, TRP7 expression did not affect capacitative Ca2+ entry induced by thapsigargin, whereas TRP7 greatly potentiated Mn2+ influx induced by diacylglycerols without involvement of protein kinase C. Nystatin-perforated and conventional whole-cell patch clamp recordings from TRP7-expressing cells demonstrated the constitutively activated and ATP-enhanced inward cation currents, both of which were initially blocked and then subsequently facilitated by extracellular Ca2+ at a physiological concentration. Impairment of TRP7 currents by internal perfusion of the Ca2+ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid revealed an essential role of intracellular Ca2+ in activation of TRP7, and their potent activation by the diacylglycerol analogue suggests that the TRP7 channel is a new member of diacylglycerol-activated cation channels. Relative permeabilities indicate that TRP7 is slightly selective to divalent cations. Thus, our findings reveal an interesting correspondence of TRP7 to the background and receptor stimulation-induced cation currents in various native systems.


Characterization of mammalian homologues of Drosophila transient receptor potential protein (TRP) is an important clue to understand molecular mechanisms underlying Ca 2؉ influx activated in response to stimulation of G q protein-coupled receptors in vertebrate cells.
Here we have isolated cDNA encoding a novel seventh mammalian TRP homologue, TRP7, from mouse brain. TRP7 showed abundant RNA expression in the heart, lung, and eye and moderate expression in the brain, spleen, and testis. TRP7 recombinantly expressed in human embryonic kidney cells exhibited distinctive functional features, compared with other TRP homologues. Basal influx activity accompanied by reduction in Ca 2؉ release from internal stores was characteristic of TRP7-expressing cells but was by far less significant in cells expressing TRP3, which is structurally the closest to TRP7 in the TRP family. TRP7 induced Ca 2؉ influx in response to ATP receptor stimulation at ATP concentrations lower than those necessary for activation of TRP3 and for Ca 2؉ release from the intracellular store, which suggests that the TRP7 channel is activated independently of Ca 2؉ release. In fact, TRP7 expression did not affect capacitative Ca 2؉ entry induced by thapsigargin, whereas TRP7 greatly potentiated Mn 2؉ influx induced by diacylglycerols without involvement of protein kinase C. Nystatin-perforated and conventional whole-cell patch clamp recordings from TRP7-expressing cells demonstrated the constitutively activated and ATP-enhanced inward cation currents, both of which were initially blocked and then subsequently facilitated by extracellular Ca 2؉ at a physiological concentration. Impairment of TRP7 currents by internal perfusion of the Ca 2؉ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N,N-tetraacetic acid revealed an essential role of intracellular Ca 2؉ in activation of TRP7, and their potent activation by the diacylglycerol analogue suggests that the TRP7 channel is a new member of diacylglycerol-activated cation channels. Relative permeabilities indicate that TRP7 is slightly selective to divalent cations. Thus, our findings reveal an interesting correspondence of TRP7 to the background and receptor stimulation-induced cation currents in various native systems.
Receptor stimulation by hormones, neurotransmitters, autacoids, Ca 2ϩ , and antigens leads to phospholipase C activation via G q protein or protein kinases. It has been recognized that subsequent hydrolysis of phosphoinositides is linked to Ca 2ϩ influx across the plasma membrane (1). Diverse ion channels activated by various triggers have been reported to be responsible for this type of Ca 2ϩ influx (1). Among members of the group, recent attention was particularly directed to capacitative Ca 2ϩ entry (CCE 1 ; in other words, Ca 2ϩ release-activated current or store-operated channel) that is activated through Ca 2ϩ release from the intracellular Ca 2ϩ store, endoplasmic reticulum, induced by inositol 1,4,5-trisphosphate (IP 3 ), and consequent depletion of Ca 2ϩ from the store (2)(3)(4)(5)(6)(7)(8). In addition, other plasma membrane ion channels directly activated by second messengers such as Ca 2ϩ , IP 3 , inositol 1,3,4,5-tetraphosphate, and arachidonic acid metabolites are also categorized as Ca 2ϩ -permeable channels activated in response to receptor stimulation (1, 9 -13).
Drosophila transient receptor potential protein (TRP), which was discovered through the genetic studies of the Drosophila visual transduction mutation (14), and the invertebrate and vertebrate TRP homologues so far have been the only molecular entities functionally characterized as channels responsible * This work was supported by research grants from the Ministry of Education, Science, Sports, and Culture and by the Research for the Future Program of the Japan Society for the Promotion of Science. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF139923.
In Drosophila photoreceptors, two homologues, Drosophila TRP and TRPL, are responsible for the light-sensitive conductances (15,17,35,38). By contrast, in native mammalian cells, cationic currents induced by receptor stimulation have yet to be resolved into components mediated by individual TRP homologues and other unidentified entities. Here we found a novel member of the mammalian TRP family, TRP7, from mouse brain. TRP7 was widely distributed, being the most abundant in the heart, lung, and eye. Functional properties of TRP7 recombinantly expressed in human embryonic kidney (HEK) cells were distinct from those of TRP3, which is structurally the most homologous to TRP7 in the TRP family. TRP7 was constitutively activated to the level much higher than that of TRP3. Ca 2ϩ influx mediated by TRP7 was enhanced by ATP receptor stimulation at the concentrations of ATP lower than those necessary for Ca 2ϩ release from the store and Ca 2ϩ influx via TRP3. The experiments also showed an essential involvement of diacylglycerol and intracellular Ca 2ϩ in activation of TRP7. In addition, the content of the intracellular Ca 2ϩ store was reduced by TRP7 expression but not by TRP3 expression. Nystatin-perforated patch clamp recordings from TRP7-expressing cells demonstrated the constitutive and ATP-enhanced nonselective cation currents, which were suppressed by a physiological concentration of extracellular Ca 2ϩ . Our findings demonstrate that TRP7 encodes a cation channel with distinctive functional properties among mammalian TRP homologues, suggesting that TRP7 is responsible for spontaneous and receptor stimulation-induced non-selective cationic currents known for important physiological roles in native tissues.
In Situ Hybridization-Brains were dissected from anesthetized 8-week-old B6 mice after perfusion with 4% paraformaldehyde in 0.1 M sodium phosphate buffer and were postfixed overnight at 4°C in the same fixative and subsequently dehydrated, embedded in paraffin, sliced at 8 m, and mounted onto silicon-coated glass slides.
Washed slides were incubated in 1.5% blocking reagent (Roche Molecular Biochemicals) in 100 mM Tris-HCl (pH 7.5), and 150 mM NaCl (DIG buffer) at room temperature for 1 h. Immunochemical detection of the hybridized probes was performed using alkaline phosphatase-conjugated anti-DIG antibody (Roche Molecular Biochemicals) at 500 times dilution in DIG buffer for 60 min. The alkaline phosphatase activity was visualized by 12-16 h incubation with nitro blue tetrazolium/5bromo-4-chloro-3-indolyl phosphate alkaline phosphatase detection system (Roche Molecular Biochemicals) in 100 mM Tris-HCl (pH 9.5), 50 mM MgCl 2 , and 100 mM NaCl under light protection.
Measurement of Changes in [Ca 2ϩ ] i -Cells on coverslips were loaded with fura-2 by incubation in Dulbecco's modified Eagle's medium containing 5 M fura-2/AM (Dojindo Laboratories, Kumamoto, Japan) and 10% fetal bovine serum at 37°C for 30 min and washed with the HEPES-buffered saline (HBS) containing (in mM) 107 NaCl, 6 KCl, 1.2 MgSO 4 , 2 CaCl 2 , 1.2 KH 2 PO 4 , 11.5 glucose, 20 HEPES, adjusted to pH 7.4 with NaOH. The coverslips were then placed in a perfusion chamber mounted on the stage of the microscope. Fluorescence images of the cells were recorded and analyzed with a video image analysis system (ARGUS-20/CA, Hamamatsu Photonics, Hamamatsu, Japan). The fura-2 fluorescence at an emission wavelength of 510 nm (bandwidth, 20 nm) was observed at room temperature by exciting fura-2 alternately at 340 and 380 nm (bandwidth, 11 nm). The 340:380 nm ratio images were obtained on a pixel by pixel basis and were converted to Ca 2ϩ concentrations by in vitro calibration. The calibration procedure was performed according to Ueda and Okada (43). All the reagents, except LaCl 3 and GdCl 3 , dissolved in water, ethanol, or dimethyl sulfoxide were diluted to their final concentrations in HBS or Ca 2ϩ -free HBS containing (in mM) 107 NaCl, 6 KCl, 1.2 MgSO 4 , 1.2 KH 2 PO 4 , 0.5 EGTA, 11.5 glucose, 20 HEPES, adjusted to pH 7.4 with NaOH, and applied to the cells by perfusion. LaCl 3 and GdCl 3 (100 mM in water) were diluted in HBS or Ca 2ϩ -free HBS from which KH 2 PO 4 was omitted. Data were accumulated under each condition from 2 to 4 experiments using cells prepared through 2-3 transfections.
Measurement of Mn 2ϩ Influx-Mn 2ϩ entry was measured through monitoring the decline of fluorescence of fura-2 induced by binding of Mn 2ϩ . Fluorescence quenching was studied using the fura-2 isosbestic excitation wavelength at 360 nm and recording emitted fluorescence at 510 nm. The system used was the same as in the [Ca 2ϩ ] i measurement. The fluorescence intensity relative to the initial values were obtained on a pixel by pixel basis. MnCl 2 and ATP dissolved in water were diluted to their final concentrations in nominally Ca 2ϩ -free HBS from which KH 2 PO 4 was omitted.
Electrophysiology-For electrophysiological measurements, coverslips with cells were placed in a perfusion chamber. Cells prepared in this manner had membrane capacitance of 8 -26 picofarads (n ϭ 13). Patch electrodes having a resistance of 4 -6 megohms (when filled with the pipette solution described below) were fabricated from 1.5-mm Pyrex capillaries, using an automatic multiple-stage micropipette puller (P-97, Sutter Instruments, USA), and heat-polished in a microforge bearing a thin indium-platinum heating wire (Narishige, Tokyo, Japan). Voltage generation and current signal acquisition were implemented through a high-impedance low-noise patch clamp amplifier (EPC-7, List Electronics, Darmstadt, Germany). Ramp currents were sampled at 2 kHz after low pass filtering at 1 kHz and analyzed with pCLAMP 6.02 software (Axon Instruments, Foster City, CA) for Fig. 8C. For long term recordings such as illustrated in Fig. 8A, current signals were stored on a computer hard disc (filtered at 50 Hz), using an A/D, D/A converter, MacLab/4 (AD Instruments, New South Wales, Australia). The details for the nystatin-perforated recording have been described previously (44). The pipette solution contained (in mM) 140 CsCl, 2 MgCl 2 , 10 HEPES, 20 glucose, adjusted to pH 7.2 with Tris base. For the conventional whole-cell recording, the pipette solution contained (in mM) 140 CsCl, 2 MgCl 2 , 10 BAPTA, and 0, 3, or 5 CaCl 2 , 2 ATP (free acid), 0.1 GTP (lithium salt), 10 HEPES, adjusted to pH 7.2 with Tris base. The Ca 2ϩ -free external solution contained (in mM) 140 NaCl, 0.5 EGTA, 10 HEPES, 20 glucose, adjusted to pH 7.4 with Tris base. For the nominally Ca 2ϩ -free external solution, EGTA was simply omitted from this solution. For 2 mM Ca 2ϩ -containing external solution, 2 mM CaCl 2 was added to the nominally Ca 2ϩ -free solution. For the cesium external solution, Na ϩ in the Ca 2ϩ -free solution was simply replaced by 140 mM Cs ϩ . The 10 mM divalent cation-containing external solution contained (in mM) either of 10 CaCl 2 , BaCl 2 , or MnCl 2 , 126 N-methyl-D-glucamine-Cl (NMDG-Cl), 20 glucose, 10 HEPES, adjusted to pH 7.4 with NMDG.

RESULTS
Primary Structure of TRP7- Fig. 1A shows the amino acid sequence of the mouse TRP7 deduced from the cDNA sequence in comparison with that of mouse TRP3 (41) which is structurally the most similar to TRP7 among the TRP homologues (see below). TRP7 is composed of 862 amino acid residues. Deletions of nucleotides encoding the amino acid sequences 322-376 and 261-376 in the clones p7␤ and p7␥/a11, respectively (see above under "Experimental Procedures"), are presumably resulting from alternative RNA splicing. A 382-base pair deletion in the clone p7␦ causes a frameshift that changes the residue 322 from phenylalanine to glutamate and terminates translation of TRP7 there. A hydropathy profile of TRP7 reveals eight hydrophobic segments and the hydrophilic N and C termini ( Fig. 1B), similar to those of other TRP subtypes (14, 16, 19 -25, 41). Sufficient length of hydrophobic regions to span the membrane, together with the lack of a hydrophobic N-terminal sequence indicative of the signal sequence, suggests that TRP7 is a membrane protein with a core of transmembrane segments and the flanking N-and C-terminal regions disposed on the cytoplasmic side. Deleted amino acid sequences in the spliced variants include the first hydrophobic region (H1) and the adjacent N-terminal portion. Fig. 1C depicts a phylogenetic tree of the TRP family constructed by the neighbor-joining method (45), based on the sequence alignment carried out by the ClustalW program (46). The vanilloid receptor (rVR1) that forms a non-selective cation channel activated by capsaicin is structurally related to members of the TRP family (47). The amino acid sequence of mouse TRP7 has identities/similarities of 34:56, 81:89, 40:60, 40:60, 75:84, 32:53, 33:53, and 13:34% with mTRP1 (48), mTRP3 (41), mTRP4 (41), mTRP5 (25), mTRP6 (24), dTRP (14), dTRPL (16), and rVR1 (47), respectively.
Tissue Distribution of TRP7 RNA-RNA preparations from different mouse tissues were subjected to Northern blot analysis using TRP7 cDNA as a specific probe. Hybridizable RNA signals of ϳ4.0 and Ͼ5.0 kb were detected at the highest level in the heart, lung, and eye and at lower levels in the brain, spleen, and testis ( Fig. 2A). To determine which cell types express TRP7 RNA in the central nervous system, para-sagittal sections of 8-week-old B6 mouse brains were subjected to in situ hybridization histochemistry using cRNA probes specific for TRP7 (Fig. 2, B and C). Cerebellar Purkinje cells were the most prominent site of TRP7 RNA expression in the brain; intense TRP7 hybridization signals were concentrated along the Purkinje cell layer (Fig. 2, B and C). TRP7 signals were also found intensely in the mitral layer of olfactory bulb and hippocampal neurons and diffusely in other regions including the cerebellar nuclei, pons, and cerebral cortex.
Functional Characterization of TRP7, Fluorescence Measurements-HEK293 cells serve as an excellent heterologous expression system to study functional properties of recombinant TRP channels, since stimulation of an endogenous P 2y2 purinoceptor induces intracellular phosphoinositide signaling but generates low endogenous activity of store-operated CCE channels (25) (see also Fig. 5). TRP7␣ cDNA, which is the longest form of the cloned TRP7 variants, together with a marker FIG. 1. Primary structure of mouse TRP7. A, the amino acid sequence (in single letter code) of the mouse TRP7 deduced from the cDNA sequence is aligned with that of mouse TRP3 (41) by the ClustalW program (46). Pairs of identical residues at one position are enclosed with solid protein CD8 cDNA was transiently expressed in HEK cells, and intracellular Ca 2ϩ concentration in TRP7␣-transfected cells was monitored by using fura-2 as an indicator and compared with that in vector-transfected control cells and cells transfected with TRP3, which is structurally the closest to TRP7. In the presence of 2 mM extracellular Ca 2ϩ , application of 100 M ATP to vector-transfected, CD8-positive control cells induced a rapid rise in [Ca 2ϩ ] i that peaked within 20 s and gradually decreased to the resting level within 300 s (Fig. 3A, open  circles). This transient rise in [Ca 2ϩ ] i was presumed to be due mainly to release from the intracellular Ca 2ϩ store, because omission of extracellular Ca 2ϩ did not significantly affect the peak level (Fig. 3B, open circles). The decay phase was accelerated by omission of extracellular Ca 2ϩ . When 100 M ATP was applied to TRP7␣-transfected, CD8-positive cells in the presence of extracellular Ca 2ϩ , the [Ca 2ϩ ] i rise greatly increased and showed a sustained phase after the initial transient phase (Fig. 3A, filled circles). Under the constant stimulation with ATP, [Ca 2ϩ ] i levels in TRP7-transfected cells did not return completely to the resting level even 30 min after the start of ATP stimulation (⌬[Ca 2ϩ ] i ϭ 19 Ϯ 4 nM, n ϭ 65). ATP-induced [Ca 2ϩ ] i increases were enhanced and prolonged also by TRP3 expression (Fig. 3A, crosses), but [Ca 2ϩ ] i levels in TRP3-transfected cells were almost the resting level after 30 lines and pairs of conservative residues at one position with broken lines. The hydrophobic regions H1-H8 are shown. The bracketing arrows indicate where deletion occurs through alternative RNA splicing in TRP7. B, the Kyte-Doolittle hydrophobicity profile of TRP7 was generated with the window size of 10 amino acids (72). C, the phylogenetic tree for the TRP family was generated by the neighbor-joining method (45), based on the sequence alignment carried out by the ClustalW program (46). Members of the TRP family are as follows: dTRP (14), dTRPL (16), mTRP1 (48), mTRP3 (41), mTRP4 (41), mTRP5 (25), mTRP6 (24), ceTRP (73), and rVR1 (47). min of ATP stimulation (⌬[Ca 2ϩ ] i ϭ 8 Ϯ 4 nM, n ϭ 32). In TRP7␣-and TRP3-expressing cells, Ca 2ϩ influx across the plasma membrane was likely to be a major cause of the [Ca 2ϩ ] i rise, because the ATP-induced [Ca 2ϩ ] i increase was much smaller in amplitude and much more transient in the absence of extracellular Ca 2ϩ (Fig. 3, B and D, filled circles and crosses) than in the presence of extracellular Ca 2ϩ at ATP concentration above 0.1 M (Fig. 3, A and C, filled circles and crosses). Interestingly, [Ca 2ϩ ] i increase due to ATP-induced Ca 2ϩ release from the internal Ca 2ϩ store in TRP7␣-transfected cells was smaller than in control and TRP3-transfected cells (Fig. 3, B and D).
The ATP-induced Ca 2ϩ influx in TRP7␣-and TRP3-expressing cells was observed separately from Ca 2ϩ release using the protocol as in Fig. 3B, where ATP was first applied in the absence of extracellular Ca 2ϩ , and 2 mM Ca 2ϩ was then added to the extracellular solution after [Ca 2ϩ ] i returned to the level before ATP application. Addition of Ca 2ϩ to the extracellular solution only slightly raised [Ca 2ϩ ] i above the resting level in control cells, whereas in TRP7␣-and TRP3-transfected cells, it elicited the large and sustained [Ca 2ϩ ] i rises. This sustained Ca 2ϩ influx contrasts with the transient nature of TRP5-dependent Ca 2ϩ influx (25). Importantly, ATP concentration required to induce the extracellular Ca 2ϩ -dependent [Ca 2ϩ ] i rise in TRP7␣-expressing cells was significantly lower than that needed to trigger Ca 2ϩ release from the intracellular Ca 2ϩ store; 1 M ATP induced almost the maximal level of TRP7␣dependent Ca 2ϩ influx but did not evoke detectable Ca 2ϩ release from the store (Fig. 3, D and E, filled circles). This indicates that activation of TRP7 by receptor stimulation is not coupled with Ca 2ϩ release from the intracellular Ca 2ϩ store or consequent depletion of the Ca 2ϩ store. TRP3-dependent Ca 2ϩ influx required relatively higher ATP concentrations for activation and was only half-activated by 1 M ATP (Fig. 3E,  crosses).
We have also examined the cells transfected with other alternatively spliced forms of TRP7, TRP7␤, or TRP7␥, which lack the first hydrophobic region (H1) and the adjacent Nterminal region. However, basal and ATP receptor stimulationinduced Ca 2ϩ influx in these transfectants were not significantly different from vector-transfected cells, although Ca 2ϩ release from the intracellular Ca 2ϩ store seemed to be a little reduced by expression of TRP7␤ and TRP7␥ in the manner similar to that in TRP7␣-transfected cells (data not shown).
TRP7␣ showed a significant basal activity without any stimulation of G protein-coupled receptor by agonists such as ATP.
In the presence of extracellular Ca 2ϩ , the resting [Ca 2ϩ ] i level in TRP7␣-expressing cells was higher than in control and TRP3-expressing cells (Fig. 3, A and B) (Fig. 3B). Re-addition of 2 mM extracellular Ca 2ϩ raised [Ca 2ϩ ] i back to the previous level before the Ca 2ϩ removal (data not shown). This high basal activity is an important characteristic that distinguishes TRP7␣ from other mammalian TRP homologues. TRP3 was also reported to have a basal activity (34), whereas this activity of TRP3 was considerably low compared with that of TRP7␣. Basal activities of TRP channels were clearly observed by monitoring Mn 2ϩ influx. In the presence of 1 mM Mn 2ϩ in the nominally Ca 2ϩ -free external solution, basal Mn 2ϩ influx was observed in TRP7␣-transfected cells at levels higher than in TRP3-transfected cells (Fig. 3F). Small basal Mn 2ϩ influx in TRP3-expressing cells did not derive from low expression of TRP3, since TRP7␣ and TRP3 elicited Mn 2ϩ influx to a similar extent, when they were maximally activated by 100 M ATP (data not shown).
Lanthanides La 3ϩ and Gd 3ϩ and the imidazole derivative SK&F96365 have been previously reported to block Ca 2ϩ -permeable channels including the TRP homologues (25). We tested these agents on the TRP7␣-dependent Ca 2ϩ influx. In Fig. 4, 100 M La 3ϩ or Gd 3ϩ (Fig. 4, A and B, filled circles) or 25 M SK&F96365 (Fig. 4 C, filled circles) were added to the extracellular solution 1 min prior to the addition of extracellular Ca 2ϩ . 25 M SK&F96365 and 100 M La 3ϩ significantly suppressed the [Ca 2ϩ ] i increase due to Ca 2ϩ influx, whereas the effect of 100 M Gd 3ϩ on Ca 2ϩ influx was not significant (Fig.  4D). These effects of the agents on TRP7␣ are similar to the effects on TRP5 reported previously (25).
To examine whether TRP7 induces Ca 2ϩ influx via depletion of intracellular Ca 2ϩ stores, we used thapsigargin (TG), the inhibitor of sarcoplasmic and endoplasmic reticulum Ca 2ϩ -AT-Pases (49). Cells were treated with 2 M TG in the absence of extracellular Ca 2ϩ for 10 min, and the extracellular solution was subsequently changed to the 2 mM Ca 2ϩ -containing solution. [Ca 2ϩ ] i increases induced by the addition of extracellular Ca 2ϩ were quite similar in TRP7␣-, TRP3-, and vector-transfected cells, indicating that CCE is not potentiated by expression of TRP7 or TRP3 (Fig. 5, A-C). The results suggest that TRP7␣ and TRP3 are unresponsive to TG treatment. However, it is possible that TRP7 and TRP3 are capable of producing CCE, but activation triggers generated by TG determine the maximum limit of CCE induced in HEK cells. Furthermore, robust endogenous CCE may have masked TRP3 and TRP7 activities in the experiments. To examine these possibilities, we monitored the store depletion-dependent activities of TRP7␣ and TRP3 in the presence of 10 M Gd 3ϩ , which effectively blocks endogenous Ca 2ϩ influx in HEK cells (34) but does not significantly affect basal and ATP-induced Ca 2ϩ influx mediated by TRP7␣ and TRP3 (data not shown). In TRP7expressing cells, amplitude of [Ca 2ϩ ] i increase (⌬[Ca 2ϩ ] i ϭ 51 Ϯ 5 nM, n ϭ 22) induced by the addition of extracellular Ca 2ϩ after TG treatment was larger than that in control cells (⌬[Ca 2ϩ ] i ϭ 19 Ϯ 5 nM, n ϭ 40) (Fig. 5, D and F). However, taking into account that basal Ca 2ϩ influx elevated [Ca 2ϩ ] i in the TRP7-expressing cells (31 Ϯ 6 nM, n ϭ 27), TG-induced Ca 2ϩ influx in TRP7-expressing cells was not significantly different from that in control cells. Gd 3ϩ -resistant, TG-induced Ca 2ϩ influx was not potentiated by TRP3 expression (Fig. 5, E and F), which is different from the data obtained using human TRP3 (34). We also tested whether ATP is capable of activating additional Ca 2ϩ influx in the TRP7-or TRP3-expressing cells where TG-induced Ca 2ϩ influx is already activated.
When TG-induced [Ca 2ϩ ] i transients in the presence of extracellular Ca 2ϩ decayed to almost stationary levels, external application of 100 M ATP induced small [Ca 2ϩ ] i increases in TRP7␣-expressing cells but induced slight decreases in control cells and, to lesser extent, in TRP3-expressing cells (Fig. 5, A-C). By contrast, in the presence of Gd 3ϩ , [Ca 2ϩ ] i transients elicited by ATP application after TG treatment became prominent in TRP7␣and TRP3-expressing cells compared with those measured in the absence of Gd 3ϩ but not in the vectortransfected control cells (Fig. 5, D-F). Thus, the results obtained using TG suggest that TRP7 is activated in response to G q -coupled stimulation through mechanisms independent of store depletion.
Interestingly, [Ca 2ϩ ] i transients induced by TG treatment in the absence of extracellular Ca 2ϩ in TRP7␣-expressing cells were significantly smaller than those in control and TRP3expressing cells (Fig. 5). This observation suggests that suppression of ATP-induced Ca 2ϩ release from the store by TRP7 expression described above (Fig. 3, B and D) derived from the reduced Ca 2ϩ content or size of internal stores.
Hofmann et al. (36) reported that human TRP3 and TRP6 are activated by diacylglycerols in a membrane delimited fashion, whereas human TRP1, mouse TRP4, and mouse TRP5 are unresponsive to the lipid mediators. We tested the effects of membrane-permeable diacylglycerol analogues 1,2-dioctanoylsn-glycerol (DOG) and 1-oleoyl-2-acetyl-sn-glycerol (OAG) on Mn 2ϩ influx mediated by TRP7 (Fig. 6, A and B). Usage of Mn 2ϩ enables measurement of divalent cation influx even when Ca 2ϩ release from the internal stores coincides. When 100 M DOG was added to the extracellular solution containing 100 M Mn 2ϩ , Mn 2ϩ influx was activated in 15 out of 21 TRP7␣-transfected, CD8-positive cells within 15 s after DOG application (Fig. 6A). Only 1 out of these 15 DOG-responsive cells and no DOG-unresponsive cells showed additional Mn 2ϩ influx induced by subsequent 100 M ATP application. No vector-transfected, CD8-positive control cells responded to DOG up to 130 s (Fig. 6A). However, enhancement of Mn 2ϩ influx was observed after 135-170 s of DOG perfusion in 2 out of 16 control cells which showed no response to subsequent application of 100 M ATP. When 100 M OAG was applied, Mn 2ϩ influx was activated in 17 out of 20 TRP7␣-transfected cells within 20 s after OAG application (Fig. 6B). Twelve out of 17 cells that immediately responded to OAG showed second enhancement of Mn 2ϩ influx after 65-190 s of OAG perfusion. One out of 3 cells that did not respond immediately to OAG showed Mn 2ϩ influx enhanced after 90 s of OAG incubation. Subsequent application of 100 M ATP did not enhance Mn 2ϩ influx in any TRP7␣-transfected cells. No control cells responded to OAG within 1 min, but 15 out of 18 cells showed Mn 2ϩ influx 65-150 s after OAG application (Fig. 6B). No control cells responded to 100 M ATP applied subsequently. The diacylglycerol lipase inhibitor RHC80267 was also tested for its effect on TRP7␣-mediated Mn 2ϩ influx. Treatment of cells with the agent was expected to inhibit metabolization of diacylglycerols and to increase passively diacylglycerol content. Slight enhancement of Mn 2ϩ influx was induced in TRP7␣transfected cells by perfusion with 50 M RHC80267-contain- D), TRP3-(B and E), and vector-transfected (C and F) cells. A-C, the perfusion solution was changed to Ca 2ϩ -free HBS containing 0.5 mM EGTA, and 2 M thapsigargin (TG) was applied to the cells in the absence of extracellular Ca 2ϩ , which was followed by the addition of 2 mM extracel- ing solution. In contrast to the experiment above using DOG and OAG, subsequent application of 1 M ATP induced further Mn 2ϩ influx in TRP7␣-expressing cells (Fig. 6C).
Arachidonic acid, a polyunsaturated fatty acid, which is synthesized from diacylglycerol by diacylglycerol lipase and activates Drosophila TRP and TRPL directly (35), slowly accelerated Mn 2ϩ influx in TRP7␣-transfected cells, but a similar acceleration of Mn 2ϩ influx by arachidonic acid was observed also in control cells (data not shown).
We also tested effects of U-73122, the inhibitor of PLC (51), since PLC is directly responsible for the production of diacylglycerol by stimulation of G q -coupled P 2Y2 ATP receptor. As expected, IP 3 -dependent Ca 2ϩ release induced by 3 M ATP was almost completely suppressed by U-73122 (1 M). However, the level of Ca 2ϩ rises due to Ca 2ϩ influx in TRP7transfected cells was only slightly reduced by U-73122, whereas the Ca 2ϩ rises were significantly slowed by the inhibitor (Fig. 7A, filled circles). U-73343 (1 M), an analogue of U-73122 that acts as a very weak inhibitor of PLC (52), showed little effect on Ca 2ϩ release from the stores or on Ca 2ϩ influx induced by 3 M ATP (Fig. 7A, open squares).
The significance of intracellular Ca 2ϩ in activation of TRP7 was suggested through electrophysiological recordings where the effects of intrapipette Ca 2ϩ concentration on TRP7 activity were tested (see below in Fig. 9) and by the previous report on TRP3 (33). We therefore examined involvement of Ca 2ϩ -calmodulin by using its antagonist N-(4-aminobutyl)-5-chloro-2naphthalenesulfonamide (W-13) (53). TRP7-mediated Ca 2ϩ influx was dose-dependently inhibited by W-13 (Fig. 7B). Interestingly, Ca 2ϩ release was increased similarly by 30 and 100 M W-13.
Functional Characterization of TRP7, Electrophysiological Measurements-To investigate directly the functional characteristics of TRP7, ionic currents in TRP7␣-transfected cells The duration of exposure to Mn 2ϩ , one of diacylglycerol analogues and RHC80267, and ATP is indicated by the filled, shaded, and hatched bars, respectively, above the graphs. The duration of exposure to BIM I or PMA is indicated by the solid lines. The results are expressed as the mean percent fluorescence intensity of the initial value in 16 -29 cells. were examined in comparison with those in control cells at a holding potential of Ϫ60 mV using the nystatin-perforated patch recording (Fig. 8A). The nystatin-perforation method allowed us to observe the development of significant spontaneous inward currents in 17 out of 17 TRP7␣-transfected cells in the 140 mM Na ϩ -containing, nominally Ca 2ϩ -free extracellular solution, as the series resistance fell on progression of membrane perforation by nystatin (Fig. 8B, lower trace; 251 Ϯ 44 pA; range: 58 -756 pA). These inward currents were markedly inhibited when the extracellular solution was changed to 2 mM Ca 2ϩ -containing solution (data not shown) and increased by a subsequent change to EGTA-containing, Ca 2ϩ -free solution (Fig. 8A, lower trace). Stimulation by 100 M ATP gradually increased the magnitude of inward currents for about 1 min, which was then followed by a gradual decline, in all the TRP7␣transfected cells examined. The degree of current increase by ATP varied depending on the cells tested. As exemplified in Fig. 8A, the cells having relatively small spontaneous inward currents (Ͻ200 pA in the Ca 2ϩ -free solution; n ϭ 6) tended to exhibit a large increase in current amplitude (2.6 -11.2-fold) in response to ATP. In contrast, ATP appeared to less potently stimulate the cells which initially exhibited large spontaneous inward currents (Ͼ200 pA; n ϭ 4; 1.3-4.0-fold increase by 100 M ATP). Addition of 2 mM Ca 2ϩ to the external solution caused a rapid reduction of the currents enhanced by ATP and subsequently a gradual increase that continued until the termination of ATP stimulation (Fig. 8A, lower trace). In the presence of 2 mM Ca 2ϩ , further addition of 100 M Gd 3ϩ only slightly inhibited the currents induced by ATP (82 Ϯ 5% that in the presence of 2 mM Ca 2ϩ ). This is consistent with our finding in [Ca 2ϩ ] i measurement that ATP-induced Ca 2ϩ entry is not appreciably affected by the addition of 100 M Gd 3ϩ in TRP7␣- and a TRP7␣-transfected (lower trace) cell at a holding potential of Ϫ60 mV and under 2-s voltage ramps from Ϫ120 to 80 mV. The perfusion solution was first changed to the Ca 2ϩ -free solution, and 100 M ATP was applied to the cells in the absence of extracellular Ca 2ϩ . During ATP stimulation, the external solution was changed back to the 2 mM Ca 2ϩ -containing solution. Finally, ATP was washed out with the 2 mM Ca 2ϩ -containing solution. The duration of exposure to the 2 mM Ca 2ϩcontaining solution, the Ca 2ϩ -free solution, and ATP is indicated by the filled, open, and hatched bars, respectively, above each trace. B, time courses of conductance increase in the process of nystatin perforation after the formation of gigaohm seals on a vector-(upper trace) and a TRP7␣-transfected (lower trace) cell. C, current-voltage relationships of the TRP7 channel. The difference current trace 8- (5,6) was yielded by subtracting the average of the currents evoked by the voltage ramps numbered 5 and 6 in A from the current numbered 8. The difference current traces 9- (1-4), 11-(1-4), and 13-(1-4) were yielded by subtracting the average of the currents numbered 1-4 from the currents 9, 11, and 13, respectively. D, time courses of ionic currents recorded from TRP7␣-transfected cells at a holding potential of Ϫ60 mV. In the nominally Ca 2ϩ -free solution, 1 M calphostin C, a PKC inhibitor (upper trace), or 250 nM phorbol 12,13-dibutyrate, a PKC activator (lower trace), was applied to the cell. In the upper trace, 100 M OAG, a diacylglycerol analogue was further added to the external solution. The duration of exposure to calphostin C or phorbol 12,13-dibutyrate was indicated by the solid line above each trace, and the duration of exposure to OAG by the hatched bar above the upper trace.
The effects of the lipid mediators observed in Mn 2ϩ influx measurements were reproducible in the electrophysiological current recordings. Responsiveness of TRP7 currents to OAG treatment (Fig. 8D, upper trace) and ATP receptor stimulation (data not shown) was not impaired by a PKC inhibitor calphostin C (1 M) (54). In addition, a PKC activator phorbol 12,13dibutyrate (250 nM) significantly reduced spontaneous TRP7 inward currents (Fig. 8D, lower trace).
In order to determine the cationic selectivity of spontaneous and ATP-enhanced inward currents, we next calculated their reversal potentials (E rev ) from the current-voltage relationships measured in the Ca 2ϩ -free, Cs ϩ -containing solution, the Ca 2ϩ -free, Na ϩ -containing solution, and the solution containing 10 mM Ca 2ϩ or Ba 2ϩ . Relative permeability ratios were calculated as follows using Equations 1 and 2 derived from the Goldman-Hodgkin-Katz equations for the biionic conditions. where ⌬E rev is a change in the reversal potential, (E rev (Na ϩ or X 2ϩ ) Ϫ E rev (Cs ϩ )), and X 2ϩ is either Ca 2ϩ or Ba 2ϩ ; F is the Faraday constant; R is the gas constant, and T is the absolute temperature. The reversal potentials of spontaneous currents were 1.2 Ϯ 0.4 (n ϭ 9), 1.9 Ϯ 0.5 (n ϭ 10), Ϫ23.8 Ϯ 2.0 (n ϭ 7), and Ϫ11.1 Ϯ 0.7 (n ϭ 10) mV, and those of ATP-enhanced currents were 0.67 Ϯ 0.2 (n ϭ 4), 4.1 Ϯ 0.8 (n ϭ 2), Ϫ2.3 Ϯ 0.7 (n ϭ 4), and Ϫ5.0 Ϯ 0.6 mV (n ϭ 3), in the Cs ϩ -, Na ϩ -, 10 mM Ca 2ϩ -, and 10 mM Ba 2ϩ -containing solutions, respectively. The calculated relative permeabilities (P Cs :P Na :P Ca :P Ba ) were 1:1.0: 1.9:3.5 for the spontaneous current and 1:1.1:5.9:5.0 for the ATP-enhanced current. Previous reports have identified intracellular Ca 2 as an important positive regulator inactivation of TRP3 (33) and TRP5 (25). To investigate the dependence of TRP7 channel activity on [Ca 2ϩ ] i , pipette solutions with various free Ca 2ϩ concentration were used in the whole-cell mode of patch clamp recording. After development of spontaneous inward currents in nystatinperforated patch recordings, patched membranes were broken in to establish the whole-cell configuration. After break-in of membranes, as the pipette solution containing 10 mM BAPTA (calculated free Ca 2ϩ concentration was Ͻ1 nM (55)) was intracellularly perfused, constitutively activated inward currents in TRP7-expressing cells at a holding potential of Ϫ60 mV were suppressed time-dependently (Fig. 9, A, upper trace, and B). Similar but weaker suppression of TRP7 currents was observed when the pipette solution containing 10 mM BAPTA and 3 mM Ca 2ϩ (54 nM calculated free Ca 2ϩ concentration) was used (Fig.  9B). In contrast, intracellular perfusion of the pipette solution containing 10 mM BAPTA and 5 mM Ca 2ϩ (125 nM calculated free Ca 2ϩ concentration) potentiated TRP7-mediated inward currents (Fig. 9A, lower trace and 9B). The current-voltage relationships of these potentiated currents were similar to those of ATP-enhanced TRP7 currents in Fig. 8 (data not shown). We also used Ca 2ϩ ionophore ionomycin to increase [Ca 2ϩ ] i . However, application of 100 M ionomycin in the presence of 2 mM extracellular Ca 2ϩ did not augment but rather slightly suppressed inward currents in TRP7-expressing cells (data not shown).
Responsiveness of TRP7 currents to ATP receptor stimulation was also dependent on [Ca 2ϩ ] i . As shown in Fig. 9, C and D, extracellular application of 100 M ATP did not significantly enhanced inward currents when the pipette solution containing 10 mM BAPTA was used, whereas ATP at the same concentration dramatically enhanced inward currents when the pipette solutions contained 10 mM BAPTA and 5 mM Ca 2ϩ . Enhancement of inward currents by ATP was not as strong when the pipette solution contained 10 mM BAPTA and 3 mM Ca 2ϩ (Fig. 9D). DISCUSSION In the present investigation, we have cloned a novel seventh member of mammalian TRP homologue, designated as TRP7. TRP7 is structurally the closest to TRP3 among the TRP homologues and constitutes a subfamily together with TRP3 and TRP6 (Fig. 1C). Comparison of functional properties of TRP7 with those of TRP3 and other homologues reveals distinctive functional features of TRP7. TRP7 is activated in response to receptor stimulation at ATP concentrations 1 order of magnitude lower than those required to elicit a detectable activation of TRP3 (Fig. 3, C and E) and TRP5 (25) and Ca 2ϩ release from the stores as well (Fig. 3D). This, together with the observation that TRP7 is still activable by ATP receptor stimulation after TG treatment (Fig. 5), suggests that enhancement of Ca 2ϩ influx activity of TRP7 via receptor stimulation occurs independently of the depletion of Ca 2ϩ stores. On the other hand, efficacy of diacylglycerol derivatives in inducing cation influx activity of TRP7, as well as those of TRP3 and TRP6, suggests that diacylglycerol is the activator of these TRP channels generated by stimulation of G q protein-coupled receptors (36). Activation of TRP7 by diacylglycerol does not involve activation of PKC, which rather suppresses TRP7 activity, since the inhib-itor of PKC potentiated and, on the other hand, the PKC activator suppressed activation of TRP7 by OAG (Fig. 6, D and  E). The action of diacylglycerol may be in a membrane-delimited manner as indicated for TRP3 and TRP6 (36). However, the inhibitor of PLC, U-73122, at concentrations sufficient to abolish IP 3 -induced Ca 2ϩ release from stores, albeit the rate reduced, only partially decreased Ca 2ϩ rises generated by Ca 2ϩ influx through TRP7 channels (Fig. 7A), in contrast to TRP3 (34). It is therefore possible that G q proteins interact also more directly with TRP7 channels for activation (32). Alternatively, diacylglycerol produced through low residual PLC␤ activity, the catalytic rate of which would not be sufficiently high to increase the cytosolic IP 3 concentrations to cause a detectable Ca 2ϩ release, may slowly activate TRP7 channels that may be closely located to PLC␤, in a membrane-delimited fashion. Consistent with this idea, activation of TRP7 requires only weak ATP receptor stimulation by ATP at concentrations lower than those necessary for Ca 2ϩ release from stores, as shown in Fig.  3, D and E. Furthermore, our experiments indicate that intracellular Ca 2ϩ plays a vital role in activation of TRP7 by stimulation of ATP receptors (Fig. 9C), as observed in activation of TRP5 (25). Since enhancement of TRP7 current by OAG is still present but is significantly diminished, when [Ca 2ϩ ] i was chelated by 10 mM BAPTA (74 Ϯ 26 pA, n ϭ 3), intracellular Ca 2ϩ may increase sensitivity of TRP7 to diacylglycerol. Suppression of Ca 2ϩ influx through TRP7 by the Ca 2ϩ -calmodulin antagonist W-13 (Fig. 7B) further suggests involvement of Ca 2ϩcalmodulin pathways in the TRP7 activation.
Our results also demonstrate that constitutive Ca 2ϩ and Mn 2ϩ influx by TRP7 is more pronounced than that by TRP3, which was reported to be constitutively active when recombinantly expressed (33,34). The mechanism that elicits the constitutively activated current still remains unclear. However, since reduced Ca 2ϩ content of internal stores is suggested from the decreased IP 3 -dependent Ca 2ϩ release (Fig. 3, B and D) and the reduced size of stores estimated by using TG (Fig. 5), the basal constitutive activity observed in TRP7-expressing cells may derive from partial store depletion. In support of this speculation, the amplitude of spontaneous cation current was approximately doubled by TG treatment in TRP7-expressing cells, where TG could thoroughly extrude Ca 2ϩ from already partially depleted stores. 2 However, it seems unlikely that endogenous store-operated channels are responsible for basal activity in unstimulated TRP7-expressing cells. The obtained results indicate that the basal Ca 2ϩ influx was not affected by 10 M Gd 3ϩ that effectively blocks endogenous TG-induced Ca 2ϩ influx (Fig. 5F), and the basal currents show only slight Ca 2ϩ selectivity in contrast to high Ca 2ϩ selectivity of Ca 2ϩ release-activated current (56,57).
The TRP7 channel has unique pore properties as compared with other TRP homologues. TRP7 is slightly selective to divalent cation over monovalent cations; the divalent cation selectivity of TRP7 is lower than those of TRP5 (25) and TRP4 (23) but higher than those of TRP3 (58) and TRP1 (21). Addition of Ca 2ϩ to the external solution rapidly suppresses inward currents, whereas it instantaneously increases TRP5 currents (25). Gradual increase by external Ca 2ϩ of cation currents observed for TRP7 appears more dramatically in TRP3- (33) and TRP5-expressing cells (25), whereas slow decreases by extracellular Ca 2ϩ of IP 3 -induced currents were observed in TRP1-expressing cells (21). It is thus conceivable that these differences of TRP7 in multiple functional parameters from other TRP homologues are essential to exert their specific physiological functions.
TRP7 channels give rise to constitutively activated and receptor stimulation-induced cation currents in HEK cells. In native systems, time-independent, spontaneous currents such as background currents have been reported (59 -63). The constitutive current induced by TRP7 shares a number of functional characteristics with native background currents, including time independence (59 -63), susceptibility to suppression by external Ca 2ϩ and Mg 2ϩ (60,62,63), selectivity among monovalent cations (60,62,63), linearity of current-voltage relationship (60,62,63), and amiloride sensitivity (60). Indeed, preparations used for the measurement of background currents are from the heart (59,60,62), lung (63), and brain (61) that are abundant in TRP7 RNA (Fig. 2). TRP7 currents induced by ATP receptor stimulation in HEK cells are considerably similar to the noradrenaline-evoked cation current in smooth muscle cells of rabbit portal vein (64 -68) and rabbit ear artery, the 2 R. Inoue, T. Okada, and Y. Mori, unpublished results. latter of which also displays constitutively active currents (69), in responsiveness to diacylglycerol (enhancement; Ref. 67) and external Ca 2ϩ (suppression and enhancement; Refs. 66 and 68), and in the selectivity for divalent over monovalent cations (64,65), and the shape of current-voltage relationships (64 -66, 69). Physiological roles of spontaneously active or background and receptor stimulation-induced non-selective cation currents mediated by TRP7 may be essential for the pace-making activity in the sino-atrial nodal cells (59,60), the spontaneous secretion in pace-making endocrine cells (61), the regulation of resting membrane potential (63), and the frequency and pattern modulation of action potentials (65) in smooth muscle cells, as suggested in the previous reports using native preparations. It is interesting to examine alteration of background and receptor stimulation-induced cation currents in preparations from mice whose TRP7 is genetically deleted in comparison with wildtype currents in normal mice.
During preparation of this manuscript, cloning of a novel protein designated as TRPC7 was reported (70). TRPC7 is structurally related to the recently cloned protein termed melastatin (the identity/similarity is 30:52%) that is implicated in suppression of tumorigenesis and shows homology to TRP proteins (14 -21:35-42%), whereas TRPC7 displays weaker homology to the members of the TRP family including TRP7 presented here (the identity/similarity is 16 -22:34 -41%) than to melastatin (71). It is interesting to examine whether melastatin and TRPC7 form Ca 2ϩ -permeable cation channels as TRP1-7.