Regulation of TRPV1 by a novel renally expressed rat TRPV1 splice variant
http://www.100md.com
《美国生理学杂志》
1Division of Nephrology and Hypertension, Department of Medicine, Oregon Health and Science University and the Portland Veterans Affairs Medical Center, Portland, Oregon
2Department of Medicine and Pharmacology and Toxicology, Michigan State University, East Lansing, Michigan
ABSTRACT
The capsaicin receptor and transient receptor potential channel TRPV1 senses heat, protons, and vanilloid agonists in peripheral sensory ganglia. Abundant data have suggested the presence of potentially novel splice variants in the kidney. We report a novel rat TRPV1 splice variant, TRPV1VAR, cloned from kidney papilla. TRPV1VAR cDNA was identified in multiple kidney tissues. Its sequence was fully compatible with potential splice donor and acceptor sites in the rat TRPV1 gene. TRPV1VAR is predicted to encode a truncated form of TRPV1 consisting of the NH2-terminal 248 residues of TRPV1 (all within the NH2-terminal intracellular domain) followed by five nonconsensus amino acids (Arg-Glu-Ala-Met-Trp) and a stop codon. The variant utilizes the same consensus Kozak sequence as canonical TRPV1. A band of the appropriate molecular mass was identified in rat kidney papillary (but not medullary) lysates immunoblotted with an antibody directed against the NH2 terminus of TRPV1, whereas an antibody recognizing the TRPV1 COOH terminus failed to detect it. Upon heterologous expression in HEK 293 cells, TRPV1VAR potentiated the ability of cotransfected TRPV1 to confer calcium influx in response to resiniferatoxin. TRPV1VAR did not influence expression or cell surface localization of cotransfected TRPV1. TRPV1VAR protein product associated with the NH2 terminus of canonical TRPV1. Interestingly, when expressed in the COS-7 epithelial cell line, TRPV1VAR functioned in a dominant-negative acting capacity, partially blocking TRPV1-dependent resiniferatoxin responsiveness. We conclude that TRPV1VAR is one of perhaps several TRPV1 splice variants expressed in rat kidney and that it may serve to modulate TRPV1 responsiveness in some tissues.
transient receptor potential channel
TRANSIENT RECEPTOR POTENTIAL CHANNEL (TRPV1) is primarily expressed in dorsal root ganglia and peripheral sensory nerve endings (consistent with its role in nociception), and, to a much lesser extent, in the central nervous system (4, 14, 20). Increasingly, TRPV1 has been implicated in the functional properties of epithelia. Birder et al. (2) showed that not only is TRPV1 expressed in the afferent nerves closely apposed to bladder epithelial cells, it is also expressed in these urothelial cells themselves. This group further demonstrated that mice deleted for the TRPV1 gene exhibited impaired afferent sensation from the bladder (3). Scant but provocative data exist with respect to the kidney. Multiple studies using Northern blot analysis and RNase protection have documented abundant TRPV1 expression in the kidney (4, 6, 19). More importantly, these studies have strongly suggested the presence of novel TRPV1 splice variants in this complex tissue.
Two splice variants of rat TRPV1 have been well characterized. The clone AF158248 [GenBank] , also known as 5'sv (23), lacks the first 0.5 kb of canonical rat TRPV1 cDNA sequence (NM_031982 [GenBank] ; see Ref. 4). The conceptual translation of this cDNA was predicted to encode a variant lacking much of the large NH2-terminal intracellular domain of TRPV1. A second putative splice variant of TRPV1 was cloned as the stretch-inhibitable nonselective channel (SIC; see Ref. 24). It was also predicted to encode a protein lacking much of the NH2 terminus of TRPV1; however, the COOH terminus of this clone appears to be the product of another TRP channel gene entirely, raising speculation about its veracity (28, 32).
We report here a novel splice variant of TRPV1, TRPV1VAR, identified in the renal papilla. This tissue is associated with a markedly increased concentration of urea and NaCl by virtue of the renal concentrating mechanism. This TRPV1 variant is predicted to encode a truncated NH2-terminal intracellular domain and is devoid of the membrane-spanning domains comprising the canonical TRPV1 channel's ion pore. Coexpression of this variant with TRPV1 influenced the vanilloid responsiveness of the latter without influencing expression of the channel or trafficking of the channel to the cell surface. In addition, the variant physically associates with canonical TRPV1, as determined via coimmunoprecipitation studies.
EXPERIMENTAL PROCEDURES
Identification of TRPV1VAR. Rat kidney mRNA was subjected to RT-PCR using primers designed to amplify canonical (wild-type) rat TRPV1 (NM_039182). A subset of the products generated with primer pair rVR1–59-79–5' (ggccacagaggatctggaaag; designed to hybridize with nucleotides 59 through 79 of NM_039182) and rVR1–2661-2641–3' (caaccctgctggttccctaag; designed to hybridize with nucleotides 2661 through 2641 of NM_039182) coded for a previously undescribed splice variant of TRPV1, which we call here TRPV1VAR. The variant was identified using multiple sources of rat kidney RNA [rat kidney mRNA (Invitrogen), rat kidney total RNA (Invitrogen), and rat kidney total RNA and oligo(dT)-purified mRNA prepared in our laboratory]. Other primer sets spanning the same region of TRPV1 similarly gave rise to the variant (data not shown). Sequencing of the variant established that it included all of NM_039182 (canonical rat TRPV1) from nucleotide 58 through nucleotide 2624 with the following exceptions: 1) the novel variant included a 101-bp cassette (comprising nucleotides 764 through 864 of TRPV1VAR: GGTAGGGAGGCCATGTGGTAGACATGAGGGAGCTAGAGGCCCAATCATCCAGGGACTAGCCTCATTTGTGGGGCTCCACTGGGGATCCTGTGTTGGCTGCA) between nucleotides 821 and 822 of NM_039182; 2) the novel variant was absent 11 bp (GTCAGGCCGAG) spanning nucleotides 2414 through 2424 of NM_039182 (at position 2456/2457 of TRPV1VAR); 3) T-to-A substitution at nucleotide 145 of NM_039182 (nucleotide 87 of TRPV1VAR); 4) A-to-C substitution at nucleotide 1348 of NM_039182 (nucleotide 1391 of TRPV1VAR); and 5) A-to-G substitution at nucleotide 1791 of NM_039182 (nucleotide 1834 of TRPV1VAR). The locations of these single-nucleotide mismatches are shown in Fig. 1B. The 101-bp cassette in TRPV1VAR but absent from canonical TRPV1 was shared with a single cDNA in GenBank [AF158248, a previously reported rat TRPV1 splice variant (23)]. Fidelity of the clone to predicted intron/exon boundaries of the rat TRPV1 genomic sequence (rat chromosome 10 genomic contig NW_042663) was tested with Spidey (National Center for Biotechnology Information; http://www.ncbi.nlm.nih.gov/spidey). Of note, at the time of this writing, NW_042663 is no longer accessible electronically although the same information is contained in NW_047336 (TRPV1 gene nucleotides 1108861 to 1134144) and in rat chromosome 10 sequence NC_005109 [GenBank] (TRPV1 gene nucleotides 60123279–60148562). For RT-PCR-based generation of additional clones only in the region of the 3' 11-bp deletion (see Fig. 1B), the following primer pair was used: rVR1–2374-5' (5'-actgtgagggcgtcaagc-3'; matching nucleotides 2374–2391 of NM_031982 [GenBank] and nucleotides 2417–2434 of TRPV1VAR) and rVR1–2568-3' (5'-tgaaaacctcagcatcctctg-3'; matching nucleotides 2568–2548 of NM_031982 [GenBank] and nucleotides 2600–2580 of TRPV1VAR). Clones harboring this deletion were identified in kidney cortex, medulla, and papilla. Full-length TRPV1VAR was initially subcloned into pcDNA3.1/V5-His-TOPO with its native stop intact such that no epitope tag was expressed. For other studies, only the open-reading frame (ORF) encoding the conceptual translation of TRPV1VAR was PCR-subcloned into this vector, absent the native stop such that a carboxy-terminal V5 epitope tag was expressed. RNA was treated with the TURBO DNA-free system (Ambion) in accordance with the manufacturer's directions, under the most stringent conditions possible ("rigorous DNase treatment," in the manufacturer's parlance).
Confirmation of the novel splice variant using "mismatched" primer pairs. Further RT-PCR-based confirmation of TRPV1VAR was accomplished by amplification of additional clones from RT product of total RNA using upstream primers designed to hybridize with a portion of TRPV1VAR common to only AF158248 [GenBank] , and "downstream" primers common to only NM-031982, such that only TRPV1VAR cDNA could generate product. Primer pairs consisted of P3–812-5' (5'-ccagggactagcctcatttg-3') and P3–1346-3' (5'-ttaggggtctcactgctgct-3'), and P3–744-5' (5'-ggaggcctggcttctacttt-3') and P3–1234-3' (5'-cggtgaacttcctggatagg-3'); template consisted of PCR product from reverse-transcribed RNA harvested from dorsal root ganglion and from kidney cortex, medulla, and papilla (see RESULTS).
Transient transfection. Transient transfection of HEK cells was performed using Lipofectamine PLUS (Life Technologies) according to the manufacturer's directions. TRPV1, consisting of only the amino-terminal intracellular portion of rat TRPV1 (absent the membrane-spanning domains), in conjunction with a V5 epitope tag, was used in some experiments where indicated. TRPV1 was amplified from rat TRPV1 using "top" and "bottom" primers 5'-aggatggaacaacgggctagcttag-3' and 5'-gcgcttgacaaatctgtcccac-3', respectively, and subcloned into pcDNA3.1/V5-His-TOPO. For rat TRPV1 devoid of the V5 epitope tag, rTRPV1-V5 in pcDNA3.1-V5/His-TOPO was subjected to site-directed mutagenesis (QuikChange; Stratagene) with the following primer pairs: 5'-ggacagatttgtcaagcgcTagggcaattctgcag-3' and 5'-ctgcagaattgccctAgcgcttgacaaatctctgtcc-3', where uppercase letters represent the mutant base pair; the resultant construct expressed only the full-length native rat TRPV1 protein. TRPV1VAR-FLAG, containing a carboxy-terminal FLAG epitope tag, was generated by annealing oligonucleotides encoding restriction site-tailed FLAG epitope followed by a stop codon (5'-ctagacgccgactacaaagacgatgacgacaagtga-3'; and 5'-ccggtcacttgtcgtcatcgtctttgtagtcggcgt-3') andligating the resultant double-stranded oligonucleotide into Xba I/Age I-digested TRPV1VAR-V5/His-TOPO.
Intracellular calcium assay. Intracellular calcium was determined as previously described (31). Briefly, 48 h after transient transfection, HEK 293 cells were loaded with fura 2-AM, incubated to effect ester cleavage, washed, and suspended in HEPES-buffered saline. An aliquot of cells (50 μl) was added to a stirred cuvette (prewarmed to 37°C and containing 2 ml HEPES-buffered saline with or without calcium) in the light path of a Hitachi F-2500 fluorescence spectrophotometer. Emission at 510 nm was monitored at 1 Hz in response to sequential excitation at 340 and 380 nm, in the presence or absence of resiniferatoxin applied at 0.1 nM final concentration (unless otherwise specified). Although not shown in Figs. 1–7, in all experiments, a stable baseline tracing was achieved for 50 s before addition of agonist. Data are reported as fura 2 ratio (340/380), either as a function of time (i.e., see Figs. 4, A, C, and D, and 7A) or at a single time point [i.e., time (t) = 200 s for HEK cells and t = 125 s for COS-7 cells]. Of note, the difference in magnitude of agonist-inducible calcium entry between cells transfected with TRPV1 + vector vs. cells transfected with TRPV1 + TRPV1VAR, which are highlighted in this fashion at a single time point, was evident at essentially all time points >50 s. A single time point was chosen purely for statistical comparison.
Immunoblotting and cell surface biotinylation. Whole cell lysates were prepared from kidneys harvested from anesthetized rats. All procedures were approved by the Institutional Animal Care and Use Committee of the Portland Veterans Affairs Medical Center. Harvested kidneys were hemisected sagitally and then dissected into cortex, medulla, and papilla. Crude lysate (40 μg) was resolved via SDS-PAGE after being boiled in 1x Laemmli buffer (final concentration). Immunoblotting was performed as previously described (31) using antibodies recognizing either the NH2 or COOH termini of TRPV1. For cultured cells, lysates were prepared with Complete Lysis Buffer (50 mM Tris, pH 7.5, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl, 1 mM orthovanadate, 1 mM AEBSF, 10 μg/ml leupeptin, and 5 μg/ml aprotinin), and 20 μg lysate were resolved via SDS-PAGE. For immunoblotting, data depicted in Figs. 1–7 are representative of three separate experiments. For immunoprecipitation experiments, transfected cells were lysed in Complete Lysis Buffer and immunoprecipitated with protein A/G-linked Sepharose beads and anti-TRPV1, anti-V5, or anti-FLAG antibody. Immunoprecipitates were washed extensively with Complete Lysis Buffer, boiled in 1x Laemmli buffer (final concentration), and resolved via SDS-PAGE. For cell-surface biotinylation and avidin affinity precipitation, confluent cells were washed three times with ice-cold PBS, biotinylated with 0.5 mg/ml NHS-LC-Biotin (Pierce) in PBS for 30 min on ice, washed three times with ice-cold PBS, quenched with glycine (100 mM in PBS), and then washed three times with ice-cold PBS. Cells were then lysed with Complete Lysis Buffer and incubated on ice for 30 min. Precipitation was performed with ImmunoPure streptavidin beads (Pierce) at 4°C for 2–4 h. Beads were pelleted, washed five times with PBS, and denatured in 1x SDS Sample Loading Buffer at 95°C for 5 min before resolution on SDS-PAGE.
Image processing and statistical analysis. For quantitation of autoradiograms, exposed films were scanned (Canon LiDE80) and data reduced using ImageJ (http://rsb.info.nih.gov/ij/; National Institutes of Health) and Excel (Microsoft). For all depicted scans of enhanced chemilumenescence exposures of immunoblots, contrast was improved by decreasing the maximum input level from 255 to 175 (Adobe PhotoShop CS) to mimic the true appearance of the exposed film. In Fig. 7C, empty intervening lanes (between lanes 2 and 3) were digitally removed; all depicted lanes are from the same autoradiogram. All experiments were performed a minimum of three times. Data are expressed as means ± SE (Excel; Microsoft). Statistical significance was ascribed using the Student t-test [for correlated samples using raw data, or for independent samples using normalized data (VassarStats; http://faculty.vassar.edu/lowry/VassarStats.html)].
RESULTS
Identification of TRPV1VAR. Using primers designed to amplify full-length rat TRPV1, we identified a novel splice variant of rat TRPV1 that we initially called TRPV1-P3 and, later, TRPV1VAR. This variant was detected repeatedly from multiple preparations and sources of rat kidney RNA using several different primer pairs (see EXPERIMENTAL PROCEDURES).
After fully sequencing the variant clone, we compared it with the rat genomic sequence to determine whether it represented a rational splice product of the rat TRPV1 gene. Using the public-domain mRNA analysis tool Spidey (www.ncbi.nlm.nih.gov/spidey), we aligned TRPV1VAR with rat chromosome 10 genomic contig (NW_042663) and found it to be fully compatible with existing putative splice donor and acceptor sites (Fig. 1A). We therefore concluded that it was unlikely to represent a cloning artifact. The exon-intron structure of canonical rat TRPV1, NM_031982 [GenBank] (4), is shown for comparison, as is the exon-intron structure of a previously identified TRPV1 splice variant, AF158248 [GenBank] (23). The first three exons of the newly identified TRPV1VAR were identical to exons 2–4 of the canonical mRNA sequence; however, exon 4 of TRPV1VAR was comprised of exons 5 and 6 (plus the intervening 101 bp of intronic sequence) of the canonical TRPV1 (Fig. 1A). The remainder of the exon-intron structure of the variant mirrored that of canonical TRPV1 with the exception of exon 13. Exon 13 of TRPV1VAR (corresponding to exon 15 of the canonical TRPV1; Fig. 1A) was absent 11 bp of canonical TRPV1 mRNA sequence (gtcaggccgag) at the 3' end; this difference is not well shown in the gross map in Fig. 1A but is highlighted in Fig. 1B.
The TRPV1VAR cDNA was then aligned with canonical rat TRPV1 cDNA (NM_031982 [GenBank] ) and splice variant AF158248 [GenBank] (Ref. 23 and Fig. 1B). The TRPV1VAR cDNA diverged from canonical rat TRPV1 (NM_031982 [GenBank] ) in the following respects: 1) it included a noncanonical 101-bp cassette toward the 5' end (Fig. 1B); and 2) it was deleted for an 11-bp stretch at the 3' end. [In addition, there were 3 insignificant point mutations (see EXPERIMENTAL PROCEDURES).] The 101-bp cassette present in TRPV1VAR but absent from canonical TRPV1 had previously been reported in AF158248 [GenBank] (23). Our newly described variant, TRPV1VAR, also closely resembles AF158248 [GenBank] with the following exceptions: 1) most importantly, the present clone includes an additional 623 bp of upstream cDNA sequence matching canonical TRPV1, which changes the reading frame (Fig. 1B); 2) TRPV1VAR includes a 180-bp cassette toward the middle of the cDNA that is also present in NM_031982 [GenBank] (where it represents exon 7) but absent from AF158248 [GenBank] ; and 3) TRPV1VAR lacks an 11-bp stretch at the 3' end that is present in AF158248 [GenBank] (and NM_031982 [GenBank] ), although this does not influence the reading frame (see below). Inclusion of the 0.6 kb of upstream cDNA sequence preserves the original reading frame of canonical TRPV1, but results in premature termination (see below). The TRPV1VAR cDNA is diagrammed in an expanded view in Fig. 1B. Regions of the cDNA matching either canonical rat TRPV1 or splice variant AF158248 [GenBank] , or matching both clones, are readily apparent. Examination of the nucleotide sequence of TRPV1VAR revealed a predicted ORF extending from nucleotides 23 through 781 and encoding 253 amino acids, followed by a stop codon. The predicted translational start site bore an ideal Kozak consensus sequence (AGGATGG) and was identical to the start site utilized by canonical rat TRPV1 (NM_031982 [GenBank] ). The ORF encompassed most of exon 1, all of exons 2 and 3, and part of exon 4 (Fig. 1B).
Rat TRPV1 cDNA encodes a protein of 838 amino acids with six membrane-spanning domains. It features three ankyrin repeat domains within a large 400-amino acid intracellular amino terminus (Fig. 1C). Because of a reading frame shift introduced by inclusion of the 101-bp cassette between exons 5 and 6 of canonical TRPV1, the newly identified TRPV1VAR is predicted to encode a truncated protein of only 253 amino acids (Fig. 1C). The first 248 are identical to canonical TRPV1; the variant diverges thereafter, terminating prematurely after amino acid 253 following the unique sequence Arg-Glu-Ala-Met-Trp. Consequently, the variant lacks two of the three ankyrin domains present in the wild-type. In marked contrast, the splice variant AF158248 [GenBank] uses a nonconsensus start site and is predicted to code for only the COOH-terminal portion of TRPV1. Therefore, although their cDNAs bear many similarities, the conceptual translations of the splice variants TRPV1VAR and AF158248 [GenBank] are predicted to match nonoverlapping regions of the canonical TRPV1 protein (Fig. 1C). For comparison, the predicted protein structure of clone AB015231 [GenBank] is also shown. The veracity of this clone, also known as SIC (24), remains inconclusive (32). Although the NH2 terminus of its conceptual translation matches TRPV1, the COOH terminus is contributed by a gene coding for another TRP channel entirely; the cDNA as originally reported cannot be generated from rat genomic DNA via conventional splicing (28, 32). Therefore, AB015231 [GenBank] is either the product of rare trans-splicing (i.e., intermolecular splicing), as has been observed with several mitochondrial transcripts in plants (29) and with selected tRNAs in a hyperthermophilic archael parasite (18), or it may potentially represent a cloning artifact as some have contended (32).
Confirmation of TRPV1VAR. We used two complementary PCR-based strategies to confirm expression of TRPV1VAR. The only unique sequence at the nucleotide level in TRPV1VAR was the absence of 11 bp at the extreme 3' end (Fig. 1B). (Although other aspects of the sequence were unique, including its apparent coding for an ORF consistent with a truncated amino-terminal TRPV1 isoform, the remainder of the sequence at the nucleotide level was present in either NM_031982 [GenBank] or AF158248 [GenBank] , or in both; Fig. 1B.) We designed PCR primers to bracket the 11-bp deletion and amplified clones from dorsal root ganglion cDNA, or from cDNA prepared from a variety of renal tissues, including cortex, medulla, and papilla. The resultant PCR product was subcloned, and 10–20 clones were screened/tissue. Of these, 25% corresponded to the –11 bp variant, and this clone could be identified in each of the four tissues examined. It appeared that the variant was more prevalent in cDNA prepared from kidney papilla (data not shown).
We sought a second strategy to confirm expression of the TRPV1VAR, taking advantage of its features shared with wild-type TRPV1 and with the previously described splice variant. An upstream PCR primer was designed to hybridize to a region of the TRPV1VAR cDNA that was shared only with clone AF158248 [GenBank] , and the downstream primer was designed to match a portion of the variant that was shared only with the canonical TRPV1 (Fig. 1B; see EXPERIMENTAL PROCEDURES). Therefore, PCR product could only arise if template contained both of these elements; neither NM_031982 [GenBank] nor AF158248 [GenBank] cDNA could give rise to PCR product. Abundant product of the appropriate size (530 bp) was generated from cDNA from all four tissues (Fig. 2). A second primer pair targeting the same regions was similarly effective (see EXPERIMENTAL PROCEDURES). In addition, when the product was excised from the gel, subcloned, and sequenced, it was found to be identical to the novel TRPV1VAR (data not shown).
Identification of a possible TRPV1VAR protein. An effort was made to identify the novel variant at the protein level. Because it includes only five unique amino acids at the COOH terminus (Fig. 1C), this was felt to be insufficient for generating a TRPV1VAR-specific antibody. However, the variant was predicted to encode an NH2-terminal truncation of TRPV1, and this was sought via immunoblotting. Whole cell lysates were prepared from kidney cortex, medulla, and papilla and subjected to immunoblotting with an antibody directed against the NH2 terminus of the canonical TRPV1 sequence. This antibody recognized TRPV1 (110 kDa) expression in medulla and papilla (Fig. 3). In papillary tissue only, a band migrating at 30 kDa was also evident; this is consistent with the predicted molecular mass for TRPV1VAR. The blot was reprobed with an antibody directed against the COOH terminus of TRPV1. Canonical TRPV1 was still immunodetectable, whereas the 30-kDa band was no longer evident (Fig. 3). Importantly, these data do not establish that the 30-kDa species is TRPV1VAR; they are, however, consistent with renal papillary expression of a truncated NH2-terminal splice variant of TRPV1. In addition, this putative truncated TRPV1 protein was not detected in all tissues where TRPV1VAR cDNA was detected by RT-PCR.
Functional properties of TRPV1VAR in HEK 293 cells. We initially speculated that the TRPV1 splice variant, predicted to lack membrane-spanning domains upon translation (Fig. 1C), might serve as a dominant-negative-acting inhibitor of TRPV1 function, consistent with other TRP (8, 21) and non-TRP (7, 9, 13, 16, 17, 22, 30, 33) model systems of channel subunit function. HEK cells were transiently transfected with canonical TRPV1, in conjunction with variant TRPV1 or with empty vector alone. After 48 h, transfected cells were then exposed to the vanilloid compound, and TRPV1 agonist, resiniferatoxin, and the effect on agonist-dependent calcium entry was examined. At saturating levels of resiniferatoxin (i.e., >1 nM), there was no clear distinction between the TRPV1 response in the presence or absence of TRPV1VAR (data not shown). When lesser concentrations of resiniferatoxin were applied, eliciting a submaximal calcium response, a clear potentiative effect of TRPV1VAR coexpression on TRPV1 signaling was evident. Specifically, at 0.1 nM resiniferatoxin, the presence of TRPV1VAR increased both the initial rate and the magnitude of the calcium transient (Fig. 4A). When a single time point is selected for analysis (i.e., fura 2 ratio at 200 s of resiniferatoxin treatment), the agonist-inducible increment in intracellular calcium was greater in the cells transfected with TRPV1 + TRPV1VAR than in those transfected with TRPV1 + vector (Fig. 4B). In some but not all experiments with HEK cells, the presence of TRPV1VAR decreased the threshold for a resiniferatoxin response; in the absence of TRPV1VAR, there was no response to 10 pM resiniferatoxin; however, in the presence of TRPV1VAR, a response could be elicited in approximately half of our experiments (data not shown). Importantly, vehicle treatment (DMSO) elicited no response under either condition, and agonist at all concentrations was without effect in untransfected cells and in cells transfected with either vector alone or with TRPV1VAR alone.
To confirm that the calcium transients we were observing were genuinely a consequence of TRPV1-dependent calcium entry, additional control experiments were performed. When HEK cells transfected with TRPV1 + TRPV1VAR were treated with resiniferatoxin in the absence of extracellular calcium, an increase in intracellular calcium was not observed (Fig. 4C). In addition, when the dual-transfected cells were treated with resiniferatoxin in the presence of the TRPV1 antagonist capsazepine, the effect was similarly abolished (Fig. 4D). In cells transfected with TRPV1 alone (in the absence of TRPV1VAR), calcium-free medium and capsazepine also prevented an increase in intracellular calcium in response to resiniferatoxin (data not shown). These data confirmed that the calcium transients observed in the present model were dependent on calcium entry and TRPV1 function.
The effect of cotransfection of TRPV1VAR on expression of wild-type TRPV1 was examined as a possible explanation for the potentiating effect of the variant. Immunoblot analysis of cell lysates with anti-TRPV1 antibody indicated no effect on expression of the transfected TRPV1 (Fig. 4E). In addition, cell-surface biotinylation experiments were performed to determine whether expression of TRPV1VAR influenced trafficking of TRPV1 to the cell membrane. Cotransfection of TRPV1VAR exerted no effect on cell surface expression of TRPV1, based on anti-TRPV1 immunoblotting of avidin-agarose affinity precipitates from cells subjected to surface biotinylation (Fig. 4E). These data suggested that the effect of TRPV1VAR vis-a-vis TRPV1 was potentially mediated via direct interaction.
Interaction of TRPV1VAR with TRPV1. The ability of TRPV1VAR to interact with TRPV1 was tested. HEK cells were transfected with TRPV1, in conjunction with either vector alone or with V5-epitope-tagged TRPV1VAR (TRPV1VAR-V5). TRPV1 expressed at the cell surface was then subjected to biotinylation; this maneuver could not label TRPV1VAR, which lacks a membrane-spanning domain and cannot be expressed at the cell surface. Avidin-agarose affinity precipitates were resolved via SDS-PAGE and then immunoblotted with anti-TRPV1 (recognizing the NH2 terminus). In cells cotransfected with TRPV1VAR-V5, coprecipitating TRPV1VAR-V5 was faintly immunodetectable via anti-TRPV1 immunoblotting (Fig. 5, top); this band was absent in the absence of cotransfected TRPV1VAR-V5, indicating specificity. To confirm the identity of this band, the membrane was reprobed with anti-V5 (Fig. 5, bottom). Here the band representing TRPV1VAR is more clearly seen, since the anti-V5 antibody is of much higher affinity. TRPV1VAR-V5 coprecipitated with biotinylated TRPV1. These data support but do not confirm a physical interaction between TRPV1 and TRPV1VAR. Although unlikely, it remains possible that TRPV1VAR associated with a different cell membrane-associated protein and was thereby "pulled down" by the avidin-agarose beads.
Next we tested for the ability of the large NH2-terminal region of canonical TRPV1 to interact with TRPV1VAR. This strategy was adopted to avoid potential difficulties frequently encountered in coimmunoprecipitation experiments using highly hydrophobic membrane-associated proteins such as TRPV1. A truncation mutant of TRPV1 (TRPV1) was engineered such that it lacked 400 amino acids of the TRPV1 COOH terminus; a stop codon was inserted immediately before the first membrane-spanning domain. When TRPV1 was cotransfected with FLAG- and epitope-tagged TRPV1VAR (TRPV1VAR-FLAG), both were detectable in whole cell lysates via immunoblotting with an antibody directed against the amino terminus of TRPV1 (Fig. 6). (In whole cell lysates, TRPV1VAR-FLAG comigrated with a nonspecific band, but expression of the tagged variant was clearly more robust than this background.) When lysates were subjected to immunoprecipitation with anti-FLAG antibody to "pull down" TRPV1VAR-FLAG, coprecipitated TRPV1 was clearly detectable via immunoblotting with the anti-TRPV1 antibody (Fig. 6). We speculate that the series of small bands migrating more rapidly than TRPV1VAR-FLAG potentially represent degradation products of this epitope-tagged variant. These data established that TRPV1VAR could interact with TRPV1 via the NH2-terminal cytoplasmic domain of the latter. These data can not, however, fully exclude the requirement for an additional intervening protein or protein complex.
Functional properties of TRPV1VAR in COS-7 cells. A second heterologous expression model was used to test the effect of TRPV1VAR vis-a-vis TRPV1 function. COS-7 cells were transfected with either TRPV1 + vector or with TRPV1 + TRPV1VAR. In this epithelial cell model, resiniferatoxin-inducible, TRPV1-dependent calcium entry was partially blocked by cotransfection with TRPV1VAR (Fig. 7A). This effect was reproducible and statistically significant (Fig. 7B). In untransfected COS-7 cells, or in cells transfected with TRPV1VAR (alone, or in conjunction with empty vector), the effect of resiniferatoxin was negligible (data not shown). We speculated that this effect of TRPV1VAR (the opposite of that observed in HEK 293 cells) might be a consequence of TRPV1VAR-dependent decrease in TRPV1 cell surface expression. Cell surface biotinylation experiments in this COS-7 model indicated that coexpression of TRPV1VAR did not decrease TRPV1 surface abundance (Fig. 7C). Specifically, surface expression in the presence of TRPV1VAR was 140 ± 30% of that observed in the absence of the variant (P = 0.15). Taken together, these data indicate that TRPV1VAR may act as either a potentiator or inhibitor of TRPV1 function, depending on the cellular context.
DISCUSSION
We believe that TRPV1VAR represents a bone fide expressed splice variant of rat TRPV1 for the following reasons: 1) its cDNA could be predictably and reproducibly identified from multiple isolates and multiple sources of RNA prepared from renal and other tissues; 2) its nucleotide sequence is fully compatible with mRNA splice donor and acceptor sites in the rat TRPV1 gene; and 3) partial cDNAs could be independently amplified using a 5' (upstream) primer specific for a region of the variant that is not shared with canonical NM_031982 [GenBank] , paired with 3' (downstream) primer specific for a region of the variant not shared with known splice variant AF158248 [GenBank] . In addition, although not conclusive, a protein of the appropriate molecular mass is immunodetectable upon anti-NH2-terminal but not anti-COOH-terminal anti-TRPV1 immunoblotting of lysates prepared from the renal tissue (papilla), yielding the variant clone with the greatest frequency. It is possible, however, that this protein represents a different splice variant of TRPV1 or of a related TRP channel.
At least two splice variants of rat TRPV1 have been described. Organization of the rat TRPV1 gene is well established; canonical TRPV1 (NM_031982 [GenBank] ) is encoded by 16 exons spanning 25 kb on rat chromosome 10 (32). A TRPV1 5' splice variant (dubbed 5'sv; AF158248 [GenBank] ) was detected by RT-PCR analysis in dorsal root ganglion and brain, but not in other tissues (23). Conceptual translation of this variant encodes a TRPV1 devoid of its intracellular NH2 terminus and consisting almost entirely of transmembrane domain (Fig. 1C). When heterologously expressed, the 5'sv variant lacked vanilloid-inducible calcium channel activity (23). We do not believe that TRPV1VAR is a full-length version of clone AF158248 [GenBank] because the latter uniquely lacks exon 7 of the canonical TRPV1 (Fig. 1A) and because of their dissimilar 5' ends (Fig. 1B). A second TRPV1 splice variant, cloned as the stretch-inactivated channel, or SIC (AB015231 [GenBank] ; see Ref. 24), was similarly identified in dorsal root ganglion and the central nervous system; mRNA was also detected in the kidney (14, 23). The fidelity of this clone has recently been questioned (28, 32) because it cannot be generated via conventional splicing from the known TRPV1 genomic sequence. Although it is conceivable that splicing in trans could occur, as has rarely been reported for other gene products (e.g., Refs. 18 and 29), the 3' half of the coding sequence appears to be the product of another TRP channel gene, TRPV4. In addition, the presence of an unidentified rat TRPV1 variant was inferred in a subset of taste receptor cells when an 0.3-kb fragment of TRPV1 could be PCR amplified, yet functional properties of the TRPV1-like activity in these cells were distinct from those of classical TRPV1 (12).
Although rat TRPV1 is perhaps the best studied, alternative splicing has recently been identified in other species as well. Wang et al. (26) characterized murine TRPV1 cDNAs, identifying TRPV1 and TRPV1 variants. TRPV1 was widely expressed and formed functional channels with agonist responsiveness resembling rat TRPV1. TRPV1, in contrast, which was predicted to lack 10 amino acids immediately upstream of the region bearing the six membrane-spanning domains, was nonfunctional when expressed in isolation; TRPV1, however, could function in a dominant-negative-acting capacity when coexpressed with TRPV1 (26). This variant, being nearly full length, bears little relation to rat TRPV1VAR described herein. Lu and colleagues (11) very recently identified a splice variant of human TRPV1, dubbed TRPV1b, that is closely related to murine TRPV1. The human b variant lacks the same 10 amino acids that are absent from the murine variant; however, it as also absent an additional 50 amino acids immediately upstream of these 10 (11). Human TRPV1b, unlike human TRPV1, fails to respond to elevated temperature; it is, however, activated by protons and the vanilloid capasaicin (11).
We initially identified TRPV1VAR through the absence of 11 bp in its extreme 3' end encoded by the terminal portion of exon 15 in canonical rat TRPV1 (NM_031982 [GenBank] ). This feature is absent from any previously reported TRPV1 cDNA and is entirely consistent with putative splice sites in the rat TRPV1 gene; however, because it is far downstream of the ORF in TRPV1VAR, its functional significance remains unclear.
Earlier evidence supported the presence of unique splice variants of TRPV1 in the kidney. Using Northern analysis, Cortright et al. (6) examined mRNA harvested from a panel of human tissues and noted abundant TRPV1 expression only in the kidney. Interestingly, whereas other tissues gave a single distinct band by Northern analysis, hybridization of electrophoresed kidney RNA with a radiolabeled TRPV1 probe yielded a "smear" characteristic of multiple mRNA species. In similar fashion, the original reports of the cloning of TRPV1 suggested the presence of novel renally expressed transcripts, based on Northern blot analysis (4). RNase protection assays from several groups were also suggestive of unidentified kidney-specific splice variants. Sanchez et al. (19) attempted to quantify the relative abundance of canonical TRPV1 mRNA, as well as each of the two putative splice variants [5'sv (AF158248 [GenBank] ) and SIC (AB015231 [GenBank] )], in various tissues. Abundant expression of at least three distinct TRPV1-related transcripts was detected in kidney mRNA; these species were not observed in other tissues and were not consistent with expression of either 5'sv or SIC. These data strongly supported the presence of additional TRPV1 splice variants.
In addition to TRPV1, splice variants have been observed for other members of the TRP channel family. At least some have documented functional effects, such as dominantly inhibiting the function of the canonical channel. For example, TRPC4 has both and splice variants and is a dominant-negative-acting modulator of (21). TRPM2 is activated by ADP-ribose; a splice variant predicted to lack the extreme COOH terminus of the canonical channel fails to respond to this agonist (27), whereas a second splice variant predicted to lack four of six membrane-spanning domains functions as a dominant negative (34). Splice variants of TRPM3 involving the pore region exhibit differences in cation selectivity (15).
The molecular mechanism through which TRPV1VAR exerts its effect remains unclear. TRP channels are predicted to assemble as aggregates of four subunits, and heteromeric channels have been identified (reviewed in Ref. 5). TRPV1VAR protein appears to physically interact with canonical TRPV1; it is conceivable that TRPV1VAR perturbs normal TRPV1 channel architecture, although it would seem unlikely that it could replace an entire subunit because of its lack of membrane-spanning domains. TRPV1VAR may also serve as a "decoy" to titrate a TRPV1-associated protein or signaling intermediate that inhibits (in HEK 293 cells) or activates (in COS-7 cells) the native channel.
a physiological perspective, TRPV1VAR potentiates the vanilloid response of canonical TRPV1 in embryonal kidney cells and inhibits it in an epithelial cell line. These effects are not achieved via alterations in either TRPV1 whole cell expression or TRPV1 trafficking to the cell membrane. TRPV1VAR interacts with the large intracellular amino terminus of canonical TRPV1, so this effect may be a direct one (Fig. 6). We conclude that the protein product of TRPV1VAR, and perhaps other related TRPV1 transcripts, may serve to modulate TRPV1 function in some tissues in vivo. For example, with respect to kidney physiology, TRPV1 has been implicated in the renal regulatory response to dietary salt loading. Multiple highly clinically relevant models of systemic hypotension appear to be mediated via release of the endogenous cannabinoid anandamide (e.g., see Refs. 1 and 25). Although much of this effect may be conferred via resultant activation of cannabinoid receptors (1), evidence implicates TRPV1, with which anandamide also interacts (35). Specifically, antagonists of TRPV1 promote vasodilation and TRPV1 agonists prevent it in vitro; this effect may itself be mediated through local release of the vasodilatory peptide calcitonin gene-related peptide (35). A corresponding picture was noted in vivo: the TRPV1 agonist capsaicin decreased blood pressure in rats fed a normal-sodium diet, whereas capsazepine, a pharmacological antagonist of TRPV1 function, potentiated the hypertensive effect of salt loading (10). The presence of TRPV1 splice variants in the kidney exhibiting unique functional properties may afford an additional locus of regulation of this phenomenon.
GRANTS
These studies were supported by the National Institutes of Health, the Department of Veterans Affairs, and the American Heart Association.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
Batkai S, Jarai Z, Wagner JA, Goparaju SK, Varga K, Liu J, Wang L, Mirshahi F, Khanolkar AD, Makriyannis A, Urbaschek R, Garcia N Jr, Sanyal AJ, and Kunos G. Endocannabinoids acting at vascular CB1 receptors mediate the vasodilated state in advanced liver cirrhosis. Nat Med 7: 827–832, 2001.
Birder LA, Kanai AJ, de Groat WC, Kiss S, Nealen ML, Burke NE, Dineley KE, Watkins S, Reynolds IJ, and Caterina MJ. Vanilloid receptor expression suggests a sensory role for urinary bladder epithelial cells. Proc Natl Acad Sci USA 98: 13396–13401, 2001.
Birder LA, Nakamura Y, Kiss S, Nealen ML, Barrick S, Kanai AJ, Wang E, Ruiz G, De Groat WC, Apodaca G, Watkins S, and Caterina MJ. Altered urinary bladder function in mice lacking the vanilloid receptor TRPV1. Nat Neurosci 5: 856–860, 2002.
Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, and Julius D. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389: 816–824, 1997.
Clapham DE. TRP channels as cellular sensors. Nature 426: 517–524, 2003.
Cortright DN, Crandall M, Sanchez JF, Zou T, Krause JE, and White G. The tissue distribution and functional characterization of human VR1. Biochem Biophys Res Commun 281: 1183–1189, 2001.
Jiang D, Xiao B, Li X, and Chen SR. Smooth muscle tissues express a major dominant negative splice variant of the type 3 Ca2+ release channel (ryanodine receptor). J Biol Chem 278: 4763–4769, 2003.
Kuzhikandathil EV, Wang H, Szabo T, Morozova N, Blumberg PM, and Oxford GS. Functional analysis of capsaicin receptor (vanilloid receptor subtype 1) multimerization and agonist responsiveness using a dominant negative mutation. J Neurosci 21: 8697–8706, 2001.
Li H, Chen Q, Moss AJ, Robinson J, Goytia V, Perry JC, Vincent GM, Priori SG, Lehmann MH, Denfield SW, Duff D, Kaine S, Shimizu W, Schwartz PJ, Wang Q, and Towbin JA. New mutations in the KVLQT1 potassium channel that cause long-QT syndrome. Circulation 97: 1264–1269, 1998.
Li J and Wang DH. Function and regulation of the vanilloid receptor in rats fed a high salt diet. J Hypertens 21: 1525–1530, 2003.
Lu G, Henderson D, Liu L, Reinhart PH, and Simon SA. TRPV1b: a functional human vanilloid receptor splice variant. Mol Pharmacol 67: 1119–1127, 2005.
Lyall V, Heck GL, Vinnikova AK, Ghosh S, Phan TH, Alam RI, Russell OF, Malik SA, Bigbee JW, and DeSimone JA. The mammalian amiloride-insensitive non-specific salt taste receptor is a vanilloid receptor-1 variant. J Physiol 558: 147–159, 2004.
Maljevic S, Lerche C, Seebohm G, Alekov AK, Busch AE, and Lerche H. C-terminal interaction of KCNQ2 and KCNQ3 K+ channels. J Physiol 548: 353–360, 2003.
Mezey E, Toth ZE, Cortright DN, Arzubi MK, Krause JE, Elde R, Guo A, Blumberg PM, and Szallasi A. Distribution of mRNA for vanilloid receptor subtype 1 (VR1), and VR1-like immunoreactivity, in the central nervous system of the rat and human. Proc Natl Acad Sci USA 97: 3655–3660, 2000.
Oberwinkler J, Lis A, Giehl KM, Flockerzi V, and Philipp SE. Alternative splicing switches the divalent cation selectivity of TRPM3 channels. J Biol Chem 280: 22540–22548, 2005.
Pereon Y, Demolombe S, Baro I, Drouin E, Charpentier F, and Escande D. Differential expression of KvLQT1 isoforms across the human ventricular wall. Am J Physiol Heart Circ Physiol 278: H1908–H1915, 2000.
Raghib A, Bertaso F, Davies A, Page KM, Meir A, Bogdanov Y, and Dolphin AC. Dominant-negative synthesis suppression of voltage-gated calcium channel Cav2.2 induced by truncated constructs. J Neurosci 21: 8495–8504, 2001.
Randau L, Munch R, Hohn MJ, Jahn D, and Soll D. Nanoarchaeum equitans creates functional tRNAs from separate genes for their 5'- and 3'-halves. Nature 433: 537–541, 2005.
Sanchez JF, Krause JE, and Cortright DN. The distribution and regulation of vanilloid receptor VR1 and VR1 5' splice variant RNA expression in rat. Neuroscience 107: 373–381, 2001.
Sasamura T, Sasaki M, Tohda C, and Kuraishi Y. Existence of capsaicin-sensitive glutamatergic terminals in rat hypothalamus. Neuroreport 9: 2045–2048, 1998.
Schaefer M, Plant TD, Stresow N, Albrecht N, and Schultz G. Functional differences between TRPC4 splice variants. J Biol Chem 277: 3752–3759, 2002.
Schulze-Bahr E, Neu A, Friederich P, Kaupp UB, Breithardt G, Pongs O, and Isbrandt D. Pacemaker channel dysfunction in a patient with sinus node disease. J Clin Invest 111: 1537–1545, 2003.
Schumacher MA, Moff I, Sudanagunta SP, and Levine JD. Molecular cloning of an N-terminal splice variant of the capsaicin receptor. Loss of N-terminal domain suggests functional divergence among capsaicin receptor subtypes. J Biol Chem 275: 2756–2762, 2000.
Suzuki M, Sato J, Kutsuwada K, Ooki G, and Imai M. Cloning of a stretch-inhibitable nonselective cation channel. J Biol Chem 274: 6330–6335, 1999.
Wagner JA, Varga K, Ellis EF, Rzigalinski BA, Martin BR, and Kunos G. Activation of peripheral CB1 cannabinoid receptors in haemorrhagic shock. Nature 390: 518–521, 1997.
Wang C, Hu HZ, Colton CK, Wood JD, and Zhu MX. An alternative splicing product of the murine trpv1 gene dominant negatively modulates the activity of TRPV1 channels. J Biol Chem 279: 37423–37430, 2004.
Wehage E, Eisfeld J, Heiner I, Jungling E, Zitt C, and Luckhoff A. Activation of the cation channel long transient receptor potential channel 2 (LTRPC2) by hydrogen peroxide. A splice variant reveals a mode of activation independent of ADP-ribose. J Biol Chem 277: 23150–23156, 2002.
Wissenbach U, Bodding M, Freichel M, and Flockerzi V. Trp12, a novel Trp related protein from kidney. FEBS Lett 485: 127–134, 2000.
Wissinger B, Schuster W, and Brennicke A. Trans splicing in Oenothera mitochondria: nad1 mRNAs are edited in exon and trans-splicing group II intron sequences. Cell 65: 473–482, 1991.
Wu FF, Ryan A, Devaney J, Warnstedt M, Korade-Mirnics Z, Poser B, Escriva MJ, Pegoraro E, Yee AS, Felice KJ, Giuliani MJ, Mayer RF, Mongini T, Palmucci L, Marino M, Rudel R, Hoffman EP, and Fahlke C. Novel CLCN1 mutations with unique clinical and electrophysiological consequences. Brain 125: 2392–2407, 2002.
Xu H, Zhao H, Tian W, Yoshida K, Roullet JB, and Cohen DM. Regulation of a transient receptor potential (TRP) channel by tyrosine phosphorylation. SRC family kinase-dependent tyrosine phosphorylation of TRPV4 on TYR-253 mediates its response to hypotonic stress. J Biol Chem 278: 11520–11527, 2003.
Xue Q, Yu Y, Trilk SL, Jong BE, and Schumacher MA. The genomic organization of the gene encoding the vanilloid receptor: evidence for multiple splice variants. Genomics 76: 14–20, 2001.
Zarei MM, Zhu N, Alioua A, Eghbali M, Stefani E, and Toro L. A novel MaxiK splice variant exhibits dominant-negative properties for surface expression. J Biol Chem 276: 16232–16239, 2001.
Zhang W, Chu X, Tong Q, Cheung JY, Conrad K, Masker K, and Miller BA. A novel TRPM2 isoform inhibits calcium influx and susceptibility to cell death. J Biol Chem 278: 16222–16229, 2003.
Zygmunt PM, Petersson J, Andersson DA, Chuang H, Sorgard M, Di Marzo V, Julius D, and Hogestatt ED. Vanilloid receptors on sensory nerves mediate the vasodilator action of anandamide. Nature 400: 452–457, 1999.(Wei Tian, Yi Fu, Donna H. Wang, and Davi)
2Department of Medicine and Pharmacology and Toxicology, Michigan State University, East Lansing, Michigan
ABSTRACT
The capsaicin receptor and transient receptor potential channel TRPV1 senses heat, protons, and vanilloid agonists in peripheral sensory ganglia. Abundant data have suggested the presence of potentially novel splice variants in the kidney. We report a novel rat TRPV1 splice variant, TRPV1VAR, cloned from kidney papilla. TRPV1VAR cDNA was identified in multiple kidney tissues. Its sequence was fully compatible with potential splice donor and acceptor sites in the rat TRPV1 gene. TRPV1VAR is predicted to encode a truncated form of TRPV1 consisting of the NH2-terminal 248 residues of TRPV1 (all within the NH2-terminal intracellular domain) followed by five nonconsensus amino acids (Arg-Glu-Ala-Met-Trp) and a stop codon. The variant utilizes the same consensus Kozak sequence as canonical TRPV1. A band of the appropriate molecular mass was identified in rat kidney papillary (but not medullary) lysates immunoblotted with an antibody directed against the NH2 terminus of TRPV1, whereas an antibody recognizing the TRPV1 COOH terminus failed to detect it. Upon heterologous expression in HEK 293 cells, TRPV1VAR potentiated the ability of cotransfected TRPV1 to confer calcium influx in response to resiniferatoxin. TRPV1VAR did not influence expression or cell surface localization of cotransfected TRPV1. TRPV1VAR protein product associated with the NH2 terminus of canonical TRPV1. Interestingly, when expressed in the COS-7 epithelial cell line, TRPV1VAR functioned in a dominant-negative acting capacity, partially blocking TRPV1-dependent resiniferatoxin responsiveness. We conclude that TRPV1VAR is one of perhaps several TRPV1 splice variants expressed in rat kidney and that it may serve to modulate TRPV1 responsiveness in some tissues.
transient receptor potential channel
TRANSIENT RECEPTOR POTENTIAL CHANNEL (TRPV1) is primarily expressed in dorsal root ganglia and peripheral sensory nerve endings (consistent with its role in nociception), and, to a much lesser extent, in the central nervous system (4, 14, 20). Increasingly, TRPV1 has been implicated in the functional properties of epithelia. Birder et al. (2) showed that not only is TRPV1 expressed in the afferent nerves closely apposed to bladder epithelial cells, it is also expressed in these urothelial cells themselves. This group further demonstrated that mice deleted for the TRPV1 gene exhibited impaired afferent sensation from the bladder (3). Scant but provocative data exist with respect to the kidney. Multiple studies using Northern blot analysis and RNase protection have documented abundant TRPV1 expression in the kidney (4, 6, 19). More importantly, these studies have strongly suggested the presence of novel TRPV1 splice variants in this complex tissue.
Two splice variants of rat TRPV1 have been well characterized. The clone AF158248 [GenBank] , also known as 5'sv (23), lacks the first 0.5 kb of canonical rat TRPV1 cDNA sequence (NM_031982 [GenBank] ; see Ref. 4). The conceptual translation of this cDNA was predicted to encode a variant lacking much of the large NH2-terminal intracellular domain of TRPV1. A second putative splice variant of TRPV1 was cloned as the stretch-inhibitable nonselective channel (SIC; see Ref. 24). It was also predicted to encode a protein lacking much of the NH2 terminus of TRPV1; however, the COOH terminus of this clone appears to be the product of another TRP channel gene entirely, raising speculation about its veracity (28, 32).
We report here a novel splice variant of TRPV1, TRPV1VAR, identified in the renal papilla. This tissue is associated with a markedly increased concentration of urea and NaCl by virtue of the renal concentrating mechanism. This TRPV1 variant is predicted to encode a truncated NH2-terminal intracellular domain and is devoid of the membrane-spanning domains comprising the canonical TRPV1 channel's ion pore. Coexpression of this variant with TRPV1 influenced the vanilloid responsiveness of the latter without influencing expression of the channel or trafficking of the channel to the cell surface. In addition, the variant physically associates with canonical TRPV1, as determined via coimmunoprecipitation studies.
EXPERIMENTAL PROCEDURES
Identification of TRPV1VAR. Rat kidney mRNA was subjected to RT-PCR using primers designed to amplify canonical (wild-type) rat TRPV1 (NM_039182). A subset of the products generated with primer pair rVR1–59-79–5' (ggccacagaggatctggaaag; designed to hybridize with nucleotides 59 through 79 of NM_039182) and rVR1–2661-2641–3' (caaccctgctggttccctaag; designed to hybridize with nucleotides 2661 through 2641 of NM_039182) coded for a previously undescribed splice variant of TRPV1, which we call here TRPV1VAR. The variant was identified using multiple sources of rat kidney RNA [rat kidney mRNA (Invitrogen), rat kidney total RNA (Invitrogen), and rat kidney total RNA and oligo(dT)-purified mRNA prepared in our laboratory]. Other primer sets spanning the same region of TRPV1 similarly gave rise to the variant (data not shown). Sequencing of the variant established that it included all of NM_039182 (canonical rat TRPV1) from nucleotide 58 through nucleotide 2624 with the following exceptions: 1) the novel variant included a 101-bp cassette (comprising nucleotides 764 through 864 of TRPV1VAR: GGTAGGGAGGCCATGTGGTAGACATGAGGGAGCTAGAGGCCCAATCATCCAGGGACTAGCCTCATTTGTGGGGCTCCACTGGGGATCCTGTGTTGGCTGCA) between nucleotides 821 and 822 of NM_039182; 2) the novel variant was absent 11 bp (GTCAGGCCGAG) spanning nucleotides 2414 through 2424 of NM_039182 (at position 2456/2457 of TRPV1VAR); 3) T-to-A substitution at nucleotide 145 of NM_039182 (nucleotide 87 of TRPV1VAR); 4) A-to-C substitution at nucleotide 1348 of NM_039182 (nucleotide 1391 of TRPV1VAR); and 5) A-to-G substitution at nucleotide 1791 of NM_039182 (nucleotide 1834 of TRPV1VAR). The locations of these single-nucleotide mismatches are shown in Fig. 1B. The 101-bp cassette in TRPV1VAR but absent from canonical TRPV1 was shared with a single cDNA in GenBank [AF158248, a previously reported rat TRPV1 splice variant (23)]. Fidelity of the clone to predicted intron/exon boundaries of the rat TRPV1 genomic sequence (rat chromosome 10 genomic contig NW_042663) was tested with Spidey (National Center for Biotechnology Information; http://www.ncbi.nlm.nih.gov/spidey). Of note, at the time of this writing, NW_042663 is no longer accessible electronically although the same information is contained in NW_047336 (TRPV1 gene nucleotides 1108861 to 1134144) and in rat chromosome 10 sequence NC_005109 [GenBank] (TRPV1 gene nucleotides 60123279–60148562). For RT-PCR-based generation of additional clones only in the region of the 3' 11-bp deletion (see Fig. 1B), the following primer pair was used: rVR1–2374-5' (5'-actgtgagggcgtcaagc-3'; matching nucleotides 2374–2391 of NM_031982 [GenBank] and nucleotides 2417–2434 of TRPV1VAR) and rVR1–2568-3' (5'-tgaaaacctcagcatcctctg-3'; matching nucleotides 2568–2548 of NM_031982 [GenBank] and nucleotides 2600–2580 of TRPV1VAR). Clones harboring this deletion were identified in kidney cortex, medulla, and papilla. Full-length TRPV1VAR was initially subcloned into pcDNA3.1/V5-His-TOPO with its native stop intact such that no epitope tag was expressed. For other studies, only the open-reading frame (ORF) encoding the conceptual translation of TRPV1VAR was PCR-subcloned into this vector, absent the native stop such that a carboxy-terminal V5 epitope tag was expressed. RNA was treated with the TURBO DNA-free system (Ambion) in accordance with the manufacturer's directions, under the most stringent conditions possible ("rigorous DNase treatment," in the manufacturer's parlance).
Confirmation of the novel splice variant using "mismatched" primer pairs. Further RT-PCR-based confirmation of TRPV1VAR was accomplished by amplification of additional clones from RT product of total RNA using upstream primers designed to hybridize with a portion of TRPV1VAR common to only AF158248 [GenBank] , and "downstream" primers common to only NM-031982, such that only TRPV1VAR cDNA could generate product. Primer pairs consisted of P3–812-5' (5'-ccagggactagcctcatttg-3') and P3–1346-3' (5'-ttaggggtctcactgctgct-3'), and P3–744-5' (5'-ggaggcctggcttctacttt-3') and P3–1234-3' (5'-cggtgaacttcctggatagg-3'); template consisted of PCR product from reverse-transcribed RNA harvested from dorsal root ganglion and from kidney cortex, medulla, and papilla (see RESULTS).
Transient transfection. Transient transfection of HEK cells was performed using Lipofectamine PLUS (Life Technologies) according to the manufacturer's directions. TRPV1, consisting of only the amino-terminal intracellular portion of rat TRPV1 (absent the membrane-spanning domains), in conjunction with a V5 epitope tag, was used in some experiments where indicated. TRPV1 was amplified from rat TRPV1 using "top" and "bottom" primers 5'-aggatggaacaacgggctagcttag-3' and 5'-gcgcttgacaaatctgtcccac-3', respectively, and subcloned into pcDNA3.1/V5-His-TOPO. For rat TRPV1 devoid of the V5 epitope tag, rTRPV1-V5 in pcDNA3.1-V5/His-TOPO was subjected to site-directed mutagenesis (QuikChange; Stratagene) with the following primer pairs: 5'-ggacagatttgtcaagcgcTagggcaattctgcag-3' and 5'-ctgcagaattgccctAgcgcttgacaaatctctgtcc-3', where uppercase letters represent the mutant base pair; the resultant construct expressed only the full-length native rat TRPV1 protein. TRPV1VAR-FLAG, containing a carboxy-terminal FLAG epitope tag, was generated by annealing oligonucleotides encoding restriction site-tailed FLAG epitope followed by a stop codon (5'-ctagacgccgactacaaagacgatgacgacaagtga-3'; and 5'-ccggtcacttgtcgtcatcgtctttgtagtcggcgt-3') andligating the resultant double-stranded oligonucleotide into Xba I/Age I-digested TRPV1VAR-V5/His-TOPO.
Intracellular calcium assay. Intracellular calcium was determined as previously described (31). Briefly, 48 h after transient transfection, HEK 293 cells were loaded with fura 2-AM, incubated to effect ester cleavage, washed, and suspended in HEPES-buffered saline. An aliquot of cells (50 μl) was added to a stirred cuvette (prewarmed to 37°C and containing 2 ml HEPES-buffered saline with or without calcium) in the light path of a Hitachi F-2500 fluorescence spectrophotometer. Emission at 510 nm was monitored at 1 Hz in response to sequential excitation at 340 and 380 nm, in the presence or absence of resiniferatoxin applied at 0.1 nM final concentration (unless otherwise specified). Although not shown in Figs. 1–7, in all experiments, a stable baseline tracing was achieved for 50 s before addition of agonist. Data are reported as fura 2 ratio (340/380), either as a function of time (i.e., see Figs. 4, A, C, and D, and 7A) or at a single time point [i.e., time (t) = 200 s for HEK cells and t = 125 s for COS-7 cells]. Of note, the difference in magnitude of agonist-inducible calcium entry between cells transfected with TRPV1 + vector vs. cells transfected with TRPV1 + TRPV1VAR, which are highlighted in this fashion at a single time point, was evident at essentially all time points >50 s. A single time point was chosen purely for statistical comparison.
Immunoblotting and cell surface biotinylation. Whole cell lysates were prepared from kidneys harvested from anesthetized rats. All procedures were approved by the Institutional Animal Care and Use Committee of the Portland Veterans Affairs Medical Center. Harvested kidneys were hemisected sagitally and then dissected into cortex, medulla, and papilla. Crude lysate (40 μg) was resolved via SDS-PAGE after being boiled in 1x Laemmli buffer (final concentration). Immunoblotting was performed as previously described (31) using antibodies recognizing either the NH2 or COOH termini of TRPV1. For cultured cells, lysates were prepared with Complete Lysis Buffer (50 mM Tris, pH 7.5, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 150 mM NaCl, 1 mM orthovanadate, 1 mM AEBSF, 10 μg/ml leupeptin, and 5 μg/ml aprotinin), and 20 μg lysate were resolved via SDS-PAGE. For immunoblotting, data depicted in Figs. 1–7 are representative of three separate experiments. For immunoprecipitation experiments, transfected cells were lysed in Complete Lysis Buffer and immunoprecipitated with protein A/G-linked Sepharose beads and anti-TRPV1, anti-V5, or anti-FLAG antibody. Immunoprecipitates were washed extensively with Complete Lysis Buffer, boiled in 1x Laemmli buffer (final concentration), and resolved via SDS-PAGE. For cell-surface biotinylation and avidin affinity precipitation, confluent cells were washed three times with ice-cold PBS, biotinylated with 0.5 mg/ml NHS-LC-Biotin (Pierce) in PBS for 30 min on ice, washed three times with ice-cold PBS, quenched with glycine (100 mM in PBS), and then washed three times with ice-cold PBS. Cells were then lysed with Complete Lysis Buffer and incubated on ice for 30 min. Precipitation was performed with ImmunoPure streptavidin beads (Pierce) at 4°C for 2–4 h. Beads were pelleted, washed five times with PBS, and denatured in 1x SDS Sample Loading Buffer at 95°C for 5 min before resolution on SDS-PAGE.
Image processing and statistical analysis. For quantitation of autoradiograms, exposed films were scanned (Canon LiDE80) and data reduced using ImageJ (http://rsb.info.nih.gov/ij/; National Institutes of Health) and Excel (Microsoft). For all depicted scans of enhanced chemilumenescence exposures of immunoblots, contrast was improved by decreasing the maximum input level from 255 to 175 (Adobe PhotoShop CS) to mimic the true appearance of the exposed film. In Fig. 7C, empty intervening lanes (between lanes 2 and 3) were digitally removed; all depicted lanes are from the same autoradiogram. All experiments were performed a minimum of three times. Data are expressed as means ± SE (Excel; Microsoft). Statistical significance was ascribed using the Student t-test [for correlated samples using raw data, or for independent samples using normalized data (VassarStats; http://faculty.vassar.edu/lowry/VassarStats.html)].
RESULTS
Identification of TRPV1VAR. Using primers designed to amplify full-length rat TRPV1, we identified a novel splice variant of rat TRPV1 that we initially called TRPV1-P3 and, later, TRPV1VAR. This variant was detected repeatedly from multiple preparations and sources of rat kidney RNA using several different primer pairs (see EXPERIMENTAL PROCEDURES).
After fully sequencing the variant clone, we compared it with the rat genomic sequence to determine whether it represented a rational splice product of the rat TRPV1 gene. Using the public-domain mRNA analysis tool Spidey (www.ncbi.nlm.nih.gov/spidey), we aligned TRPV1VAR with rat chromosome 10 genomic contig (NW_042663) and found it to be fully compatible with existing putative splice donor and acceptor sites (Fig. 1A). We therefore concluded that it was unlikely to represent a cloning artifact. The exon-intron structure of canonical rat TRPV1, NM_031982 [GenBank] (4), is shown for comparison, as is the exon-intron structure of a previously identified TRPV1 splice variant, AF158248 [GenBank] (23). The first three exons of the newly identified TRPV1VAR were identical to exons 2–4 of the canonical mRNA sequence; however, exon 4 of TRPV1VAR was comprised of exons 5 and 6 (plus the intervening 101 bp of intronic sequence) of the canonical TRPV1 (Fig. 1A). The remainder of the exon-intron structure of the variant mirrored that of canonical TRPV1 with the exception of exon 13. Exon 13 of TRPV1VAR (corresponding to exon 15 of the canonical TRPV1; Fig. 1A) was absent 11 bp of canonical TRPV1 mRNA sequence (gtcaggccgag) at the 3' end; this difference is not well shown in the gross map in Fig. 1A but is highlighted in Fig. 1B.
The TRPV1VAR cDNA was then aligned with canonical rat TRPV1 cDNA (NM_031982 [GenBank] ) and splice variant AF158248 [GenBank] (Ref. 23 and Fig. 1B). The TRPV1VAR cDNA diverged from canonical rat TRPV1 (NM_031982 [GenBank] ) in the following respects: 1) it included a noncanonical 101-bp cassette toward the 5' end (Fig. 1B); and 2) it was deleted for an 11-bp stretch at the 3' end. [In addition, there were 3 insignificant point mutations (see EXPERIMENTAL PROCEDURES).] The 101-bp cassette present in TRPV1VAR but absent from canonical TRPV1 had previously been reported in AF158248 [GenBank] (23). Our newly described variant, TRPV1VAR, also closely resembles AF158248 [GenBank] with the following exceptions: 1) most importantly, the present clone includes an additional 623 bp of upstream cDNA sequence matching canonical TRPV1, which changes the reading frame (Fig. 1B); 2) TRPV1VAR includes a 180-bp cassette toward the middle of the cDNA that is also present in NM_031982 [GenBank] (where it represents exon 7) but absent from AF158248 [GenBank] ; and 3) TRPV1VAR lacks an 11-bp stretch at the 3' end that is present in AF158248 [GenBank] (and NM_031982 [GenBank] ), although this does not influence the reading frame (see below). Inclusion of the 0.6 kb of upstream cDNA sequence preserves the original reading frame of canonical TRPV1, but results in premature termination (see below). The TRPV1VAR cDNA is diagrammed in an expanded view in Fig. 1B. Regions of the cDNA matching either canonical rat TRPV1 or splice variant AF158248 [GenBank] , or matching both clones, are readily apparent. Examination of the nucleotide sequence of TRPV1VAR revealed a predicted ORF extending from nucleotides 23 through 781 and encoding 253 amino acids, followed by a stop codon. The predicted translational start site bore an ideal Kozak consensus sequence (AGGATGG) and was identical to the start site utilized by canonical rat TRPV1 (NM_031982 [GenBank] ). The ORF encompassed most of exon 1, all of exons 2 and 3, and part of exon 4 (Fig. 1B).
Rat TRPV1 cDNA encodes a protein of 838 amino acids with six membrane-spanning domains. It features three ankyrin repeat domains within a large 400-amino acid intracellular amino terminus (Fig. 1C). Because of a reading frame shift introduced by inclusion of the 101-bp cassette between exons 5 and 6 of canonical TRPV1, the newly identified TRPV1VAR is predicted to encode a truncated protein of only 253 amino acids (Fig. 1C). The first 248 are identical to canonical TRPV1; the variant diverges thereafter, terminating prematurely after amino acid 253 following the unique sequence Arg-Glu-Ala-Met-Trp. Consequently, the variant lacks two of the three ankyrin domains present in the wild-type. In marked contrast, the splice variant AF158248 [GenBank] uses a nonconsensus start site and is predicted to code for only the COOH-terminal portion of TRPV1. Therefore, although their cDNAs bear many similarities, the conceptual translations of the splice variants TRPV1VAR and AF158248 [GenBank] are predicted to match nonoverlapping regions of the canonical TRPV1 protein (Fig. 1C). For comparison, the predicted protein structure of clone AB015231 [GenBank] is also shown. The veracity of this clone, also known as SIC (24), remains inconclusive (32). Although the NH2 terminus of its conceptual translation matches TRPV1, the COOH terminus is contributed by a gene coding for another TRP channel entirely; the cDNA as originally reported cannot be generated from rat genomic DNA via conventional splicing (28, 32). Therefore, AB015231 [GenBank] is either the product of rare trans-splicing (i.e., intermolecular splicing), as has been observed with several mitochondrial transcripts in plants (29) and with selected tRNAs in a hyperthermophilic archael parasite (18), or it may potentially represent a cloning artifact as some have contended (32).
Confirmation of TRPV1VAR. We used two complementary PCR-based strategies to confirm expression of TRPV1VAR. The only unique sequence at the nucleotide level in TRPV1VAR was the absence of 11 bp at the extreme 3' end (Fig. 1B). (Although other aspects of the sequence were unique, including its apparent coding for an ORF consistent with a truncated amino-terminal TRPV1 isoform, the remainder of the sequence at the nucleotide level was present in either NM_031982 [GenBank] or AF158248 [GenBank] , or in both; Fig. 1B.) We designed PCR primers to bracket the 11-bp deletion and amplified clones from dorsal root ganglion cDNA, or from cDNA prepared from a variety of renal tissues, including cortex, medulla, and papilla. The resultant PCR product was subcloned, and 10–20 clones were screened/tissue. Of these, 25% corresponded to the –11 bp variant, and this clone could be identified in each of the four tissues examined. It appeared that the variant was more prevalent in cDNA prepared from kidney papilla (data not shown).
We sought a second strategy to confirm expression of the TRPV1VAR, taking advantage of its features shared with wild-type TRPV1 and with the previously described splice variant. An upstream PCR primer was designed to hybridize to a region of the TRPV1VAR cDNA that was shared only with clone AF158248 [GenBank] , and the downstream primer was designed to match a portion of the variant that was shared only with the canonical TRPV1 (Fig. 1B; see EXPERIMENTAL PROCEDURES). Therefore, PCR product could only arise if template contained both of these elements; neither NM_031982 [GenBank] nor AF158248 [GenBank] cDNA could give rise to PCR product. Abundant product of the appropriate size (530 bp) was generated from cDNA from all four tissues (Fig. 2). A second primer pair targeting the same regions was similarly effective (see EXPERIMENTAL PROCEDURES). In addition, when the product was excised from the gel, subcloned, and sequenced, it was found to be identical to the novel TRPV1VAR (data not shown).
Identification of a possible TRPV1VAR protein. An effort was made to identify the novel variant at the protein level. Because it includes only five unique amino acids at the COOH terminus (Fig. 1C), this was felt to be insufficient for generating a TRPV1VAR-specific antibody. However, the variant was predicted to encode an NH2-terminal truncation of TRPV1, and this was sought via immunoblotting. Whole cell lysates were prepared from kidney cortex, medulla, and papilla and subjected to immunoblotting with an antibody directed against the NH2 terminus of the canonical TRPV1 sequence. This antibody recognized TRPV1 (110 kDa) expression in medulla and papilla (Fig. 3). In papillary tissue only, a band migrating at 30 kDa was also evident; this is consistent with the predicted molecular mass for TRPV1VAR. The blot was reprobed with an antibody directed against the COOH terminus of TRPV1. Canonical TRPV1 was still immunodetectable, whereas the 30-kDa band was no longer evident (Fig. 3). Importantly, these data do not establish that the 30-kDa species is TRPV1VAR; they are, however, consistent with renal papillary expression of a truncated NH2-terminal splice variant of TRPV1. In addition, this putative truncated TRPV1 protein was not detected in all tissues where TRPV1VAR cDNA was detected by RT-PCR.
Functional properties of TRPV1VAR in HEK 293 cells. We initially speculated that the TRPV1 splice variant, predicted to lack membrane-spanning domains upon translation (Fig. 1C), might serve as a dominant-negative-acting inhibitor of TRPV1 function, consistent with other TRP (8, 21) and non-TRP (7, 9, 13, 16, 17, 22, 30, 33) model systems of channel subunit function. HEK cells were transiently transfected with canonical TRPV1, in conjunction with variant TRPV1 or with empty vector alone. After 48 h, transfected cells were then exposed to the vanilloid compound, and TRPV1 agonist, resiniferatoxin, and the effect on agonist-dependent calcium entry was examined. At saturating levels of resiniferatoxin (i.e., >1 nM), there was no clear distinction between the TRPV1 response in the presence or absence of TRPV1VAR (data not shown). When lesser concentrations of resiniferatoxin were applied, eliciting a submaximal calcium response, a clear potentiative effect of TRPV1VAR coexpression on TRPV1 signaling was evident. Specifically, at 0.1 nM resiniferatoxin, the presence of TRPV1VAR increased both the initial rate and the magnitude of the calcium transient (Fig. 4A). When a single time point is selected for analysis (i.e., fura 2 ratio at 200 s of resiniferatoxin treatment), the agonist-inducible increment in intracellular calcium was greater in the cells transfected with TRPV1 + TRPV1VAR than in those transfected with TRPV1 + vector (Fig. 4B). In some but not all experiments with HEK cells, the presence of TRPV1VAR decreased the threshold for a resiniferatoxin response; in the absence of TRPV1VAR, there was no response to 10 pM resiniferatoxin; however, in the presence of TRPV1VAR, a response could be elicited in approximately half of our experiments (data not shown). Importantly, vehicle treatment (DMSO) elicited no response under either condition, and agonist at all concentrations was without effect in untransfected cells and in cells transfected with either vector alone or with TRPV1VAR alone.
To confirm that the calcium transients we were observing were genuinely a consequence of TRPV1-dependent calcium entry, additional control experiments were performed. When HEK cells transfected with TRPV1 + TRPV1VAR were treated with resiniferatoxin in the absence of extracellular calcium, an increase in intracellular calcium was not observed (Fig. 4C). In addition, when the dual-transfected cells were treated with resiniferatoxin in the presence of the TRPV1 antagonist capsazepine, the effect was similarly abolished (Fig. 4D). In cells transfected with TRPV1 alone (in the absence of TRPV1VAR), calcium-free medium and capsazepine also prevented an increase in intracellular calcium in response to resiniferatoxin (data not shown). These data confirmed that the calcium transients observed in the present model were dependent on calcium entry and TRPV1 function.
The effect of cotransfection of TRPV1VAR on expression of wild-type TRPV1 was examined as a possible explanation for the potentiating effect of the variant. Immunoblot analysis of cell lysates with anti-TRPV1 antibody indicated no effect on expression of the transfected TRPV1 (Fig. 4E). In addition, cell-surface biotinylation experiments were performed to determine whether expression of TRPV1VAR influenced trafficking of TRPV1 to the cell membrane. Cotransfection of TRPV1VAR exerted no effect on cell surface expression of TRPV1, based on anti-TRPV1 immunoblotting of avidin-agarose affinity precipitates from cells subjected to surface biotinylation (Fig. 4E). These data suggested that the effect of TRPV1VAR vis-a-vis TRPV1 was potentially mediated via direct interaction.
Interaction of TRPV1VAR with TRPV1. The ability of TRPV1VAR to interact with TRPV1 was tested. HEK cells were transfected with TRPV1, in conjunction with either vector alone or with V5-epitope-tagged TRPV1VAR (TRPV1VAR-V5). TRPV1 expressed at the cell surface was then subjected to biotinylation; this maneuver could not label TRPV1VAR, which lacks a membrane-spanning domain and cannot be expressed at the cell surface. Avidin-agarose affinity precipitates were resolved via SDS-PAGE and then immunoblotted with anti-TRPV1 (recognizing the NH2 terminus). In cells cotransfected with TRPV1VAR-V5, coprecipitating TRPV1VAR-V5 was faintly immunodetectable via anti-TRPV1 immunoblotting (Fig. 5, top); this band was absent in the absence of cotransfected TRPV1VAR-V5, indicating specificity. To confirm the identity of this band, the membrane was reprobed with anti-V5 (Fig. 5, bottom). Here the band representing TRPV1VAR is more clearly seen, since the anti-V5 antibody is of much higher affinity. TRPV1VAR-V5 coprecipitated with biotinylated TRPV1. These data support but do not confirm a physical interaction between TRPV1 and TRPV1VAR. Although unlikely, it remains possible that TRPV1VAR associated with a different cell membrane-associated protein and was thereby "pulled down" by the avidin-agarose beads.
Next we tested for the ability of the large NH2-terminal region of canonical TRPV1 to interact with TRPV1VAR. This strategy was adopted to avoid potential difficulties frequently encountered in coimmunoprecipitation experiments using highly hydrophobic membrane-associated proteins such as TRPV1. A truncation mutant of TRPV1 (TRPV1) was engineered such that it lacked 400 amino acids of the TRPV1 COOH terminus; a stop codon was inserted immediately before the first membrane-spanning domain. When TRPV1 was cotransfected with FLAG- and epitope-tagged TRPV1VAR (TRPV1VAR-FLAG), both were detectable in whole cell lysates via immunoblotting with an antibody directed against the amino terminus of TRPV1 (Fig. 6). (In whole cell lysates, TRPV1VAR-FLAG comigrated with a nonspecific band, but expression of the tagged variant was clearly more robust than this background.) When lysates were subjected to immunoprecipitation with anti-FLAG antibody to "pull down" TRPV1VAR-FLAG, coprecipitated TRPV1 was clearly detectable via immunoblotting with the anti-TRPV1 antibody (Fig. 6). We speculate that the series of small bands migrating more rapidly than TRPV1VAR-FLAG potentially represent degradation products of this epitope-tagged variant. These data established that TRPV1VAR could interact with TRPV1 via the NH2-terminal cytoplasmic domain of the latter. These data can not, however, fully exclude the requirement for an additional intervening protein or protein complex.
Functional properties of TRPV1VAR in COS-7 cells. A second heterologous expression model was used to test the effect of TRPV1VAR vis-a-vis TRPV1 function. COS-7 cells were transfected with either TRPV1 + vector or with TRPV1 + TRPV1VAR. In this epithelial cell model, resiniferatoxin-inducible, TRPV1-dependent calcium entry was partially blocked by cotransfection with TRPV1VAR (Fig. 7A). This effect was reproducible and statistically significant (Fig. 7B). In untransfected COS-7 cells, or in cells transfected with TRPV1VAR (alone, or in conjunction with empty vector), the effect of resiniferatoxin was negligible (data not shown). We speculated that this effect of TRPV1VAR (the opposite of that observed in HEK 293 cells) might be a consequence of TRPV1VAR-dependent decrease in TRPV1 cell surface expression. Cell surface biotinylation experiments in this COS-7 model indicated that coexpression of TRPV1VAR did not decrease TRPV1 surface abundance (Fig. 7C). Specifically, surface expression in the presence of TRPV1VAR was 140 ± 30% of that observed in the absence of the variant (P = 0.15). Taken together, these data indicate that TRPV1VAR may act as either a potentiator or inhibitor of TRPV1 function, depending on the cellular context.
DISCUSSION
We believe that TRPV1VAR represents a bone fide expressed splice variant of rat TRPV1 for the following reasons: 1) its cDNA could be predictably and reproducibly identified from multiple isolates and multiple sources of RNA prepared from renal and other tissues; 2) its nucleotide sequence is fully compatible with mRNA splice donor and acceptor sites in the rat TRPV1 gene; and 3) partial cDNAs could be independently amplified using a 5' (upstream) primer specific for a region of the variant that is not shared with canonical NM_031982 [GenBank] , paired with 3' (downstream) primer specific for a region of the variant not shared with known splice variant AF158248 [GenBank] . In addition, although not conclusive, a protein of the appropriate molecular mass is immunodetectable upon anti-NH2-terminal but not anti-COOH-terminal anti-TRPV1 immunoblotting of lysates prepared from the renal tissue (papilla), yielding the variant clone with the greatest frequency. It is possible, however, that this protein represents a different splice variant of TRPV1 or of a related TRP channel.
At least two splice variants of rat TRPV1 have been described. Organization of the rat TRPV1 gene is well established; canonical TRPV1 (NM_031982 [GenBank] ) is encoded by 16 exons spanning 25 kb on rat chromosome 10 (32). A TRPV1 5' splice variant (dubbed 5'sv; AF158248 [GenBank] ) was detected by RT-PCR analysis in dorsal root ganglion and brain, but not in other tissues (23). Conceptual translation of this variant encodes a TRPV1 devoid of its intracellular NH2 terminus and consisting almost entirely of transmembrane domain (Fig. 1C). When heterologously expressed, the 5'sv variant lacked vanilloid-inducible calcium channel activity (23). We do not believe that TRPV1VAR is a full-length version of clone AF158248 [GenBank] because the latter uniquely lacks exon 7 of the canonical TRPV1 (Fig. 1A) and because of their dissimilar 5' ends (Fig. 1B). A second TRPV1 splice variant, cloned as the stretch-inactivated channel, or SIC (AB015231 [GenBank] ; see Ref. 24), was similarly identified in dorsal root ganglion and the central nervous system; mRNA was also detected in the kidney (14, 23). The fidelity of this clone has recently been questioned (28, 32) because it cannot be generated via conventional splicing from the known TRPV1 genomic sequence. Although it is conceivable that splicing in trans could occur, as has rarely been reported for other gene products (e.g., Refs. 18 and 29), the 3' half of the coding sequence appears to be the product of another TRP channel gene, TRPV4. In addition, the presence of an unidentified rat TRPV1 variant was inferred in a subset of taste receptor cells when an 0.3-kb fragment of TRPV1 could be PCR amplified, yet functional properties of the TRPV1-like activity in these cells were distinct from those of classical TRPV1 (12).
Although rat TRPV1 is perhaps the best studied, alternative splicing has recently been identified in other species as well. Wang et al. (26) characterized murine TRPV1 cDNAs, identifying TRPV1 and TRPV1 variants. TRPV1 was widely expressed and formed functional channels with agonist responsiveness resembling rat TRPV1. TRPV1, in contrast, which was predicted to lack 10 amino acids immediately upstream of the region bearing the six membrane-spanning domains, was nonfunctional when expressed in isolation; TRPV1, however, could function in a dominant-negative-acting capacity when coexpressed with TRPV1 (26). This variant, being nearly full length, bears little relation to rat TRPV1VAR described herein. Lu and colleagues (11) very recently identified a splice variant of human TRPV1, dubbed TRPV1b, that is closely related to murine TRPV1. The human b variant lacks the same 10 amino acids that are absent from the murine variant; however, it as also absent an additional 50 amino acids immediately upstream of these 10 (11). Human TRPV1b, unlike human TRPV1, fails to respond to elevated temperature; it is, however, activated by protons and the vanilloid capasaicin (11).
We initially identified TRPV1VAR through the absence of 11 bp in its extreme 3' end encoded by the terminal portion of exon 15 in canonical rat TRPV1 (NM_031982 [GenBank] ). This feature is absent from any previously reported TRPV1 cDNA and is entirely consistent with putative splice sites in the rat TRPV1 gene; however, because it is far downstream of the ORF in TRPV1VAR, its functional significance remains unclear.
Earlier evidence supported the presence of unique splice variants of TRPV1 in the kidney. Using Northern analysis, Cortright et al. (6) examined mRNA harvested from a panel of human tissues and noted abundant TRPV1 expression only in the kidney. Interestingly, whereas other tissues gave a single distinct band by Northern analysis, hybridization of electrophoresed kidney RNA with a radiolabeled TRPV1 probe yielded a "smear" characteristic of multiple mRNA species. In similar fashion, the original reports of the cloning of TRPV1 suggested the presence of novel renally expressed transcripts, based on Northern blot analysis (4). RNase protection assays from several groups were also suggestive of unidentified kidney-specific splice variants. Sanchez et al. (19) attempted to quantify the relative abundance of canonical TRPV1 mRNA, as well as each of the two putative splice variants [5'sv (AF158248 [GenBank] ) and SIC (AB015231 [GenBank] )], in various tissues. Abundant expression of at least three distinct TRPV1-related transcripts was detected in kidney mRNA; these species were not observed in other tissues and were not consistent with expression of either 5'sv or SIC. These data strongly supported the presence of additional TRPV1 splice variants.
In addition to TRPV1, splice variants have been observed for other members of the TRP channel family. At least some have documented functional effects, such as dominantly inhibiting the function of the canonical channel. For example, TRPC4 has both and splice variants and is a dominant-negative-acting modulator of (21). TRPM2 is activated by ADP-ribose; a splice variant predicted to lack the extreme COOH terminus of the canonical channel fails to respond to this agonist (27), whereas a second splice variant predicted to lack four of six membrane-spanning domains functions as a dominant negative (34). Splice variants of TRPM3 involving the pore region exhibit differences in cation selectivity (15).
The molecular mechanism through which TRPV1VAR exerts its effect remains unclear. TRP channels are predicted to assemble as aggregates of four subunits, and heteromeric channels have been identified (reviewed in Ref. 5). TRPV1VAR protein appears to physically interact with canonical TRPV1; it is conceivable that TRPV1VAR perturbs normal TRPV1 channel architecture, although it would seem unlikely that it could replace an entire subunit because of its lack of membrane-spanning domains. TRPV1VAR may also serve as a "decoy" to titrate a TRPV1-associated protein or signaling intermediate that inhibits (in HEK 293 cells) or activates (in COS-7 cells) the native channel.
a physiological perspective, TRPV1VAR potentiates the vanilloid response of canonical TRPV1 in embryonal kidney cells and inhibits it in an epithelial cell line. These effects are not achieved via alterations in either TRPV1 whole cell expression or TRPV1 trafficking to the cell membrane. TRPV1VAR interacts with the large intracellular amino terminus of canonical TRPV1, so this effect may be a direct one (Fig. 6). We conclude that the protein product of TRPV1VAR, and perhaps other related TRPV1 transcripts, may serve to modulate TRPV1 function in some tissues in vivo. For example, with respect to kidney physiology, TRPV1 has been implicated in the renal regulatory response to dietary salt loading. Multiple highly clinically relevant models of systemic hypotension appear to be mediated via release of the endogenous cannabinoid anandamide (e.g., see Refs. 1 and 25). Although much of this effect may be conferred via resultant activation of cannabinoid receptors (1), evidence implicates TRPV1, with which anandamide also interacts (35). Specifically, antagonists of TRPV1 promote vasodilation and TRPV1 agonists prevent it in vitro; this effect may itself be mediated through local release of the vasodilatory peptide calcitonin gene-related peptide (35). A corresponding picture was noted in vivo: the TRPV1 agonist capsaicin decreased blood pressure in rats fed a normal-sodium diet, whereas capsazepine, a pharmacological antagonist of TRPV1 function, potentiated the hypertensive effect of salt loading (10). The presence of TRPV1 splice variants in the kidney exhibiting unique functional properties may afford an additional locus of regulation of this phenomenon.
GRANTS
These studies were supported by the National Institutes of Health, the Department of Veterans Affairs, and the American Heart Association.
FOOTNOTES
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
REFERENCES
Batkai S, Jarai Z, Wagner JA, Goparaju SK, Varga K, Liu J, Wang L, Mirshahi F, Khanolkar AD, Makriyannis A, Urbaschek R, Garcia N Jr, Sanyal AJ, and Kunos G. Endocannabinoids acting at vascular CB1 receptors mediate the vasodilated state in advanced liver cirrhosis. Nat Med 7: 827–832, 2001.
Birder LA, Kanai AJ, de Groat WC, Kiss S, Nealen ML, Burke NE, Dineley KE, Watkins S, Reynolds IJ, and Caterina MJ. Vanilloid receptor expression suggests a sensory role for urinary bladder epithelial cells. Proc Natl Acad Sci USA 98: 13396–13401, 2001.
Birder LA, Nakamura Y, Kiss S, Nealen ML, Barrick S, Kanai AJ, Wang E, Ruiz G, De Groat WC, Apodaca G, Watkins S, and Caterina MJ. Altered urinary bladder function in mice lacking the vanilloid receptor TRPV1. Nat Neurosci 5: 856–860, 2002.
Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, and Julius D. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389: 816–824, 1997.
Clapham DE. TRP channels as cellular sensors. Nature 426: 517–524, 2003.
Cortright DN, Crandall M, Sanchez JF, Zou T, Krause JE, and White G. The tissue distribution and functional characterization of human VR1. Biochem Biophys Res Commun 281: 1183–1189, 2001.
Jiang D, Xiao B, Li X, and Chen SR. Smooth muscle tissues express a major dominant negative splice variant of the type 3 Ca2+ release channel (ryanodine receptor). J Biol Chem 278: 4763–4769, 2003.
Kuzhikandathil EV, Wang H, Szabo T, Morozova N, Blumberg PM, and Oxford GS. Functional analysis of capsaicin receptor (vanilloid receptor subtype 1) multimerization and agonist responsiveness using a dominant negative mutation. J Neurosci 21: 8697–8706, 2001.
Li H, Chen Q, Moss AJ, Robinson J, Goytia V, Perry JC, Vincent GM, Priori SG, Lehmann MH, Denfield SW, Duff D, Kaine S, Shimizu W, Schwartz PJ, Wang Q, and Towbin JA. New mutations in the KVLQT1 potassium channel that cause long-QT syndrome. Circulation 97: 1264–1269, 1998.
Li J and Wang DH. Function and regulation of the vanilloid receptor in rats fed a high salt diet. J Hypertens 21: 1525–1530, 2003.
Lu G, Henderson D, Liu L, Reinhart PH, and Simon SA. TRPV1b: a functional human vanilloid receptor splice variant. Mol Pharmacol 67: 1119–1127, 2005.
Lyall V, Heck GL, Vinnikova AK, Ghosh S, Phan TH, Alam RI, Russell OF, Malik SA, Bigbee JW, and DeSimone JA. The mammalian amiloride-insensitive non-specific salt taste receptor is a vanilloid receptor-1 variant. J Physiol 558: 147–159, 2004.
Maljevic S, Lerche C, Seebohm G, Alekov AK, Busch AE, and Lerche H. C-terminal interaction of KCNQ2 and KCNQ3 K+ channels. J Physiol 548: 353–360, 2003.
Mezey E, Toth ZE, Cortright DN, Arzubi MK, Krause JE, Elde R, Guo A, Blumberg PM, and Szallasi A. Distribution of mRNA for vanilloid receptor subtype 1 (VR1), and VR1-like immunoreactivity, in the central nervous system of the rat and human. Proc Natl Acad Sci USA 97: 3655–3660, 2000.
Oberwinkler J, Lis A, Giehl KM, Flockerzi V, and Philipp SE. Alternative splicing switches the divalent cation selectivity of TRPM3 channels. J Biol Chem 280: 22540–22548, 2005.
Pereon Y, Demolombe S, Baro I, Drouin E, Charpentier F, and Escande D. Differential expression of KvLQT1 isoforms across the human ventricular wall. Am J Physiol Heart Circ Physiol 278: H1908–H1915, 2000.
Raghib A, Bertaso F, Davies A, Page KM, Meir A, Bogdanov Y, and Dolphin AC. Dominant-negative synthesis suppression of voltage-gated calcium channel Cav2.2 induced by truncated constructs. J Neurosci 21: 8495–8504, 2001.
Randau L, Munch R, Hohn MJ, Jahn D, and Soll D. Nanoarchaeum equitans creates functional tRNAs from separate genes for their 5'- and 3'-halves. Nature 433: 537–541, 2005.
Sanchez JF, Krause JE, and Cortright DN. The distribution and regulation of vanilloid receptor VR1 and VR1 5' splice variant RNA expression in rat. Neuroscience 107: 373–381, 2001.
Sasamura T, Sasaki M, Tohda C, and Kuraishi Y. Existence of capsaicin-sensitive glutamatergic terminals in rat hypothalamus. Neuroreport 9: 2045–2048, 1998.
Schaefer M, Plant TD, Stresow N, Albrecht N, and Schultz G. Functional differences between TRPC4 splice variants. J Biol Chem 277: 3752–3759, 2002.
Schulze-Bahr E, Neu A, Friederich P, Kaupp UB, Breithardt G, Pongs O, and Isbrandt D. Pacemaker channel dysfunction in a patient with sinus node disease. J Clin Invest 111: 1537–1545, 2003.
Schumacher MA, Moff I, Sudanagunta SP, and Levine JD. Molecular cloning of an N-terminal splice variant of the capsaicin receptor. Loss of N-terminal domain suggests functional divergence among capsaicin receptor subtypes. J Biol Chem 275: 2756–2762, 2000.
Suzuki M, Sato J, Kutsuwada K, Ooki G, and Imai M. Cloning of a stretch-inhibitable nonselective cation channel. J Biol Chem 274: 6330–6335, 1999.
Wagner JA, Varga K, Ellis EF, Rzigalinski BA, Martin BR, and Kunos G. Activation of peripheral CB1 cannabinoid receptors in haemorrhagic shock. Nature 390: 518–521, 1997.
Wang C, Hu HZ, Colton CK, Wood JD, and Zhu MX. An alternative splicing product of the murine trpv1 gene dominant negatively modulates the activity of TRPV1 channels. J Biol Chem 279: 37423–37430, 2004.
Wehage E, Eisfeld J, Heiner I, Jungling E, Zitt C, and Luckhoff A. Activation of the cation channel long transient receptor potential channel 2 (LTRPC2) by hydrogen peroxide. A splice variant reveals a mode of activation independent of ADP-ribose. J Biol Chem 277: 23150–23156, 2002.
Wissenbach U, Bodding M, Freichel M, and Flockerzi V. Trp12, a novel Trp related protein from kidney. FEBS Lett 485: 127–134, 2000.
Wissinger B, Schuster W, and Brennicke A. Trans splicing in Oenothera mitochondria: nad1 mRNAs are edited in exon and trans-splicing group II intron sequences. Cell 65: 473–482, 1991.
Wu FF, Ryan A, Devaney J, Warnstedt M, Korade-Mirnics Z, Poser B, Escriva MJ, Pegoraro E, Yee AS, Felice KJ, Giuliani MJ, Mayer RF, Mongini T, Palmucci L, Marino M, Rudel R, Hoffman EP, and Fahlke C. Novel CLCN1 mutations with unique clinical and electrophysiological consequences. Brain 125: 2392–2407, 2002.
Xu H, Zhao H, Tian W, Yoshida K, Roullet JB, and Cohen DM. Regulation of a transient receptor potential (TRP) channel by tyrosine phosphorylation. SRC family kinase-dependent tyrosine phosphorylation of TRPV4 on TYR-253 mediates its response to hypotonic stress. J Biol Chem 278: 11520–11527, 2003.
Xue Q, Yu Y, Trilk SL, Jong BE, and Schumacher MA. The genomic organization of the gene encoding the vanilloid receptor: evidence for multiple splice variants. Genomics 76: 14–20, 2001.
Zarei MM, Zhu N, Alioua A, Eghbali M, Stefani E, and Toro L. A novel MaxiK splice variant exhibits dominant-negative properties for surface expression. J Biol Chem 276: 16232–16239, 2001.
Zhang W, Chu X, Tong Q, Cheung JY, Conrad K, Masker K, and Miller BA. A novel TRPM2 isoform inhibits calcium influx and susceptibility to cell death. J Biol Chem 278: 16222–16229, 2003.
Zygmunt PM, Petersson J, Andersson DA, Chuang H, Sorgard M, Di Marzo V, Julius D, and Hogestatt ED. Vanilloid receptors on sensory nerves mediate the vasodilator action of anandamide. Nature 400: 452–457, 1999.(Wei Tian, Yi Fu, Donna H. Wang, and Davi)