Cloning and Expression of a Novel Human Brain Na Channel*

We have cloned a novel cDNA from human brain which encodes a non-voltage-dependent Na (cid:49) channel (BNC1). BNC1 has some sequence similarity (24–28%) with a new channel family that includes subunits of the mammalian epithelial Na (cid:49) channel, the Caenorhabditis elegans degenerins, and the Helix aspersa FMRF-amide- gated Na (cid:49) channel. Like other family members it is inhibited by amiloride. However, its predicted structure differs from other family members, its discrimination between Na (cid:49) and Li (cid:49) is different, and in contrast to other mammalian family members, coexpression with other cloned subunits of the family does not increase current. BNC1 has a unique pattern of expression with transcripts detected only in adult human brain and in spinal cord. Thus, BNC1 is the first cloned member of a new subfamily of mammalian Na (cid:49) channels. The function of BNC1 as a non-voltage-gated Na (cid:49) channel in human brain suggests it may play a novel role in neurotransmission. Recent studies have identified a new family of Na (cid:49) channels whose characteristic features include Na (cid:49) selectivity, inhibition by amiloride, and a conserved primary structure (1–11). Family members contain 500 to 800 residues. Sequence analysis and studies of topology suggest that the amino and carboxyl termini are intracellular, that there are two hydrophobic regions that traverse the membrane (M1 and M2), 1 and that between M1 and M2 there lies a large cysteine-rich extracellular the pMT3 vector for expression in oocytes (18). Oligonucleotides were prepared on an automated Applied Biosystems oligonucleotide synthesizer. Relationship of proteins in the phylogenetic tree was de- rived using the Pileup alignment program from Genetics Computer Group (GCG). The diagram was generated using the Distances program (GCG) with Kimura substitution, followed by the Growtree program with the UPGMA option. Northern Blot Analysis— Northern blots contained 2 (cid:109) g of poly(A) (cid:49) RNA isolated from specific adult human tissues or from sections of the Phe-Leu-Phe-Gln-Pro-Gln-Arg-Phe-NH Ala-Gly-Glu-Gly-Leu-Ser-Ser-Pro-Phe- Trp-Ser-Leu-Ala-Ala-Pro-Gln-Arg-Phe-NH 2 ; RACE, rapid amplification of cDNA ends; EST, expressed sequence tag; PCR, polymerase chain reaction; bp, base pair(s); kb, kilobase(s).

The best characterized members of this family are the amiloride-sensitive epithelial Na ϩ channels (ENaC) that control Na ϩ and fluid absorption in the kidney, colon, and lung. ENaC channels are constructed from at least three homologous subunits (␣-, ␤-, and ␥ENaC) (4 -9). These channels may also be involved in detection of salty taste (15). A closely related subunit, ␦NaCh, is expressed in pancreas, testis, ovary, and brain. ␦NaCh generates Na ϩ channels when coexpressed with ␤and ␥ENaC (10), suggesting that it may be part of the ENaC subfamily of channels. Several family members have also been discovered in C. elegans, including MEC-4, MEC-10, and DEG-1, which when mutated produce a touch-insensitive phenotype (1)(2)(3). Although the function of these gene products has not been established, several observations suggest that they form ion channels: their sequences are similar to the ENaC subunits; genetic evidence suggests that three gene products are required for function (3); M2 of ␣ENaC can substitute for M2 of MEC-4 (16); and specific mutations cause ballooning cellular degeneration (1, 2) similar to that found with overexpression of active ENaC subunits (17). Based on this ability to produce cell degeneration, family members in C. elegans are called "degenerins." The most recent addition to this family is a Phe-Met-Arg-Phe-NH 2 (FMRF-amide)-stimulated Na ϩ channel (FaNaCh) cloned from Helix (11).
Here we report the cloning and expression of a novel member of the family which is expressed in human brain. We name it BNC1 for Brain Na ϩ Channel, and 1 with the expectation that additional subunits will be discovered in the future.

EXPERIMENTAL PROCEDURES
Cloning-A complete BNC1 cDNA was obtained by extending an expressed sequence tag (GenBank TM accession number Z45660) in the 5Ј direction using rapid amplification of cDNA ends (RACE) technique according to the protocol provided with the Marathon cDNA Amplification Kit from Clontech. Human brain cDNA which had been tagged with an adapter primer at the 5Ј end (Clontech) was used as template in 5Ј RACE reactions. In brief, the tagged cDNA was used in a PCR reaction with a sense primer corresponding to the tag sequence and a gene-specific antisense primer corresponding to nucleotides 256 -282 of the EST sequence. The 3Ј end of the gene-specific primer spanned the 3Ј end of the cDNA and contained a sequence complementary to the stop codon. RACE PCR reactions were done using reagents in the Advantage cDNA PCR core kit (Clontech) which contains a combination of Klentaq-1 and Deep Vent DNA polymerases and TaqStart antibody. Thermal cycling was done in a Perkin Elmer DNA Thermal Cycler using a program of one cycle at 94°C for 1 min; 5 cycles of 94°C for 30 s and 72°C for 4 min; 5 cycles of 94°C for 30 s and 70°C for 4 min; then 20 -25 cycles of 94°C for 20 s and 68°C for 4 min. PCR products were purified on an agarose gel using ␤-agarase from New England Biolabs, cloned into the pCR II vector (Invitrogen), and sequenced. DNA sequencing was done on an Applied Biosystems automated Sequencer using fluorescent dye-labeled terminators. An 1809-bp fragment was obtained from the 5Ј RACE reaction which contained 270 nucleotides of upstream untranslated sequence and a 1539-bp open reading frame extending to the 3Ј stop codon. This fragment was digested in its entirety out of the pCR II vector as a NotI/KpnI fragment and ligated into the compatible sites of the pMT3 vector for expression in oocytes (18). Oligonucleotides were prepared on an automated Applied Biosystems oligonucleotide synthesizer. Relationship of proteins in the phylogenetic tree was derived using the Pileup alignment program from Genetics Computer Group (GCG). The diagram was generated using the Distances program (GCG) with Kimura substitution, followed by the Growtree program with the UPGMA option.
Northern Blot Analysis-Northern blots contained 2 g of poly(A) ϩ RNA isolated from specific adult human tissues or from sections of the * This work was supported in part by the Howard Hughes Medical Institute (HHMI). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EMBL Data Bank with accession number(s) U50352.
‡ Supported by NHLBI Training Grant HL-07344 from the National Institutes of Health.
§ Investigator of the HHMI. To whom correspondence should be addressed: Howard Hughes Medical Institute, University of Iowa College of Medicine, 500 EMRB, Iowa City, IA 52242. Tel.: 319-335-7619; Fax: 319-335-7623. 1 The abbreviations used are: M1 and M2, first and second membrane-spanning sequences, respectively; ENaC, epithelial Na ϩ channel with ␣, ␤, and ␥ subunits; ␦NaCh, ␦ subunit of a Na ϩ channel; FMRFamide, Phe-Met-Arg-Phe-NH 2 ; FaNaCh, FMRF-amide-gated Na ϩ channel; BNC1, brain Na ϩ channel; F-8-F-amide, Phe-Leu-Phe-Gln-Pro-Gln-Arg-Phe-NH 2 ; A-18-F-amide, Ala-Gly-Glu-Gly-Leu-Ser-Ser-Pro-Phe-Trp-Ser-Leu-Ala-Ala-Pro-Gln-Arg-Phe-NH 2 ; RACE, rapid amplification of cDNA ends; EST, expressed sequence tag; PCR, polymerase chain reaction; bp, base pair(s); kb, kilobase(s). brain (Clontech). Probes were prepared by random prime labeling (Pharmacia Biotech Inc.). PCR primers specific for the 5Ј and 3Ј ends of the protein coding sequence of the BNC1 cDNA were used in a PCR reaction to generate a fragment containing the entire coding sequence of BNC1. This fragment was cloned into the pCR II vector and used to probe the multiple tissue blots. An EcoRI/SphI 460-bp fragment was isolated from the 5Ј end of the coding region clone and used as a 5Ј end specific probe. A 299-bp PCR product corresponding to the 3Ј end of the coding region of BNC1 was cloned for use as a 3Ј end specific probe. Filters were hybridized overnight at 42°C in a buffer containing 50% formamide, 5 ϫ SSPE, 2% SDS, 10 ϫ Denhardt's solution, and 100 g/ml salmon sperm DNA. Filters were washed with 0.1 ϫ SSC, 0.1% SDS at 55°C and exposed to Kodak X-Omat AR film for 4 days at Ϫ70°C.
To determine ionic selectivity, NaCl was replaced with LiCl or KCl. Current-voltage relationships were determined by stepping from a holding potential of Ϫ60 mV to potentials between Ϫ100 and ϩ40 mV for 1 s. Amiloride-sensitive current was obtained by subtracting current during exposure to a maximal concentration of amiloride (100 M) from current prior to amiloride addition. Phe-Met-Arg-Phe-NH 2 (FMRF-amide), Phe-Leu-Phe-Gln-Pro-Gln-Arg-Phe-NH 2 (F-8-F-amide), and Ala-Gly-Glu-Gly-Leu-Ser-Ser-Pro-Phe-Trp-Ser-Leu-Ala-Ala-Pro-Gln-Arg-Phe-NH 2 (A-18-F-amide) were obtained from Sigma and were added to the bathing solution at 1-30 M.

RESULTS AND DISCUSSION
Cloning and Sequence Analysis-To identify new mammalian Na ϩ channels, we used the BLAST sequence alignment program (19) to search the data base of expressed sequence tags (EST) with the amino acid sequences of hENaC and the degenerins. We found a 299-nucleotide sequence (GenBank TM number Z45660) obtained from human brain cDNA. The EST was capable of encoding a 94-amino acid open reading frame. Using methods described above, we identified an 1809-bp cDNA containing a 1539-bp open reading frame with stops in all three reading frames upstream of the putative start methionine. Fig. 1A shows the deduced amino acid sequence of BNC1. Fig. 1B shows that BNC1 has a predicted structure with some features similar to that of other cloned amiloride-sensitive Na ϩ channels and the degenerins. Of particular interest are the two hydrophobic transmembrane segments and the extracellular cysteine-rich domains. There is also an area with limited sequence conservation between two cysteine-rich domains (cross-hatched area in Fig. 1B). However, there are also significant differences between BNC1 and other cloned members of the family (Fig. 1B). In the amino-terminal half of the extracellular domain, BNC1 seems more similar to FaNaCh because it lacks sequences found in degenerins and ENaC. Yet, in the carboxyl-terminal half of the extracellular domain, BNC1 is more similar in length to ENaC and the degenerins than to FaNaCh. BNC1 has a relatively short carboxyl-terminal intracellular tail. It lacks the conserved proline-rich sequences of ENaC that may be involved in protein-protein interactions (20). It also lacks the PPPXYXXL motif which determines the amount of cell surface protein and which is deleted from ␤hENaC or ␥hENaC in patients with Liddle's syndrome (17). BNC1 has consensus N-linked glycosylation sequences in the extracellular domain (Fig. 1A). The aminoterminal intracellular sequence contains one consensus sequence for protein kinase C phosphorylation.
Although absolute homology is relatively low, BNC1 shares slightly greater overall amino acid sequence identity with Fa-NaCh than with other members of the family. BNC1 is 28.4% identical with FaNaCh, 24.2-26.6% identical with ␣-, ␤-, and ␥ENaC and ␦NaCh, and 24.4 -25.4% identical with the degenerins. Despite the species difference, phylogenetic analysis placed BNC1 closest to FaNaCh, rather than to other mammalian members of the family (Fig. 1C).
Northern Blot Analysis-We used Northern blot analysis to examine the transcription pattern of BNC1. Fig. 2A shows that transcripts were detected in human brain and spinal cord, but not in a number of other tissues. The two BNC1 transcripts were expressed to some extent in every region of the adult human brain that was analyzed (Fig. 2B). The greatest relative abundance appeared to be in cerebellum, cerebral cortex, medulla, amygdala, and subthalamic nucleus.
The expression pattern of BNC1 is unique; expression primarily in the central nervous system contrasts with previously identified mammalian members of the family. Although transcripts of ␣and ␥ENaC and ␦NaCh have been detected in brain, they are much more prevalent in other tissues. ␣and ␥ENaC are most abundant in epithelia of kidney, colon, and lung (4 -9), and ␦NaCh is most abundant in testis, ovary, and pancreas (10). Expression of nonmammalian members of the family has been reported in excitable tissue. Transcription of FaNaCh occurs in muscle and nervous tissue of Helix (11), and the degenerins are expressed in the peripheral and central nervous system of C. elegans (1)(2)(3).
When the entire coding region of BNC1 was used as a probe, we detected two transcripts, 2.7 and 3.7 kb in length (Fig. 2, A  and B). In general, relative hybridization to the two transcripts was similar in most brain regions, although there was a greater relative abundance of the large transcript in the cerebellum, medulla, spinal cord, corpus collosum, hypothalamus, substan- Cloning and Expression of BNC1 7880 tia nigra, and thalamus. To investigate the relationship between the two transcripts, we prepared probes from the 5Ј and 3Ј regions of BNC1 cDNA (corresponding to the amino and carboxyl termini of the predicted protein) and hybridized them to a Northern blot containing human brain poly(A) ϩ RNA (Fig.  3). Whereas the 3Ј probe hybridized with both transcripts, the 5Ј probe hybridized with the 2.7-kb transcript only. These data indicate that the cDNA reported here is produced by the smaller transcript. There are at least two possible explanations for the presence of two transcripts. First, alternative splicing at the amino terminus might generate two transcripts from a single gene. Second, there may be two genes with very similar sequences corresponding to the 3Ј end of BNC1. Further investigation is necessary to distinguish between these alternatives. In either case, the data suggest the possibility of structural and thus functional complexity with multimeric channel proteins.
Expression of BNC1 in Xenopus Oocytes-Because of its homology with ENaC and FaNaCh Na ϩ channels, we tested the hypothesis that BNC1 is a Na ϩ channel. Expression of BNC1 in Xenopus oocytes generated a small inward current (holding potential Ϫ60 mV) that was reversibly inhibited by amiloride (14.0 Ϯ 2.7 nA, n ϭ 12, Fig. 4A). There was no amiloridesensitive current in control (H 2 O-injected) oocytes (Fig. 4B). The BNC1 current was highly selective for Na ϩ relative to K ϩ ; the reversal potential was 35 Ϯ 6 mV (n ϭ 6) in NaCl bathing solution, and the amiloride-sensitive current was abolished by replacing Na ϩ with K ϩ in the bathing solution (Fig. 4B). The Na ϩ current was inhibited by amiloride with half-maximal inhibition at 147 Ϯ 23 nM (Fig. 4C).
When we replaced Na ϩ with Li ϩ , we measured equal currents through BNC1 channels (Fig. 4D). This differs from ␣␤␥ENaC and ␣ENaC which are 2-fold more conductive to Li ϩ than to Na ϩ (Fig. 4D), and from FaNaCh and ␦NaCh which are more conductive to Na ϩ than to Li ϩ (9 -11). It was previously shown that Ser 589 in ␣rENaC was important for Na ϩ /Li ϩ selectivity; mutation to Ile increased Na ϩ conductance relative to Li ϩ (21). The analogous residue in BNC1 (and ␦NaCh) is alanine (Ala 442 ), suggesting that this residue might help determine relative Na ϩ /Li ϩ conductivity in BNC1 as well as in other family members. Na ϩ current generated by BNC1 was not significantly increased by coexpression with combinations of ␣-, ␤-, and/or ␥hENaC subunits (Fig. 4E). In contrast, coexpression of the three ENaC subunits significantly increased current compared with expression of only two subunits. This suggests that BNC1 functions as a novel member of the Na ϩ channel family. The data with BNC1 also contrast with ␦NaCh in which coexpression with ␤and ␥hENaC markedly increased current.
We considered the possibility that BNC1 current might be stimulated by an agonist, much as FaNaCh requires activation by the Helix aspersa neuropeptide FMRF-amide. However, BNC1 was not activated by FMRF-amide or the related mammalian peptides F-8-F-amide or A-18-F-amide. Although this suggests that BNC1 is not the mammalian homologue of the Helix FaNaCh, it does not exclude the possibility that BNC1 could be a receptor for another neurotransmitter. CONCLUSION BNC1 is a novel member of the ENaC/degenerin family. However, it has several significant differences from other FIG. 3. Northern blot analysis of human brain RNA using 5 and 3 specific BNC1 probes. 5 g of adult human brain poly(A) ϩ RNA were run on a 1.2% agarose-formaldehyde gel, transferred to a nitrocellulose filter, and hybridized with labeled probes prepared from either the 5Ј or 3Ј ends of the BNC1 cDNA as shown at bottom.

FIG. 2.
Northern blot analysis of BNC1 expression in adult human tissues (A) and in specific regions of adult human brain (B). Each lane contains approximately 2 g of poly(A) ϩ RNA; the amount of RNA in each lane was adjusted to observe identical levels of ␤-actin expression. Filters were hybridized with a probe corresponding to the coding sequence of BNC1 as described under "Experimental Procedures." Blots were exposed to film for 4 days; a 7-day exposure of B showed that both transcripts were evident to some extent in every lane.

Cloning and Expression of BNC1 7881
cloned members of the family: it has a different predicted structure; it does not discriminate between Na ϩ and Li ϩ as current carriers; expression was detected only in the central nervous system; and BNC1 current is not augmented when it is coexpressed with subunits of ENaC. These considerations suggest that BNC1 may be the first cloned member of a new subfamily of mammalian Na ϩ channels. Although expression of BNC1 generated a Na ϩ current, the magnitude was small. There are at least two explanations. It is possible that other, as yet unidentified, subunits might be required to produce a fully functional channel complex. The situation may be analogous to ␣ENaC which generates small currents when expressed alone, but produces large currents when coexpressed with ␤ and ␥ENaC (5,7). It is also possible that BNC1 might require activation by an agonist or neurotransmitter, just as FaNaCh is stimulated by FMRF-amide (11). This latter possibility seems particularly attractive because of its expression in brain. Moreover, ligand-regulated activity rather than constitutive activity would be more consistent with neuronal expression, because constitutive nonvoltage-dependent Na ϩ channel activity could depolarize the cell, thereby disrupting signal transduction, or it could cause cell toxicity. It is interesting to speculate that the large cysteine-rich extracellular domain of BNC1 might have a receptor function. Certainly the large size and presence of multiple cysteine residues is reminiscent of other receptor proteins. Identification of other brain Na ϩ channel subunits and/or receptor ligands should help us better understand the function of BNC1 and the role of non-voltage-gated Na ϩ channels in the central nervous system.