We have investigated the structure, function, and expression of the rat eye sodium/calcium+potassium exchanger NCKX1. The sequence of independent rat NCKX1 clones and the analysis of rat eye mRNA by RT-PCR revealed a region of alternative splicing that comprised four exons and encoded a stretch of 113 amino acids near the beginning of the large cytosolic loop. In comparison with other NCKX1 molecules and the rat NCKX2 protein, rat NCKX1 was highly conserved within the hydrophobic regions but was quite divergent in the two large hydrophilic loops. The only exception was the region of the cytosolic loop encoded by the second alternatively spliced exon, which was ∼60% identical. Similar to bovine, but different from human, rat NCKX1 possessed an acidic motif that was repeated 14 times in the cytoplasmic loop. Analysis of NCKX1 expression in different rat tissues by Northern blot revealed a very high level of expression of a 7-kb transcript in the eye but also lower levels of transcripts of various lengths in other tissues. The recombinant rat NCKX1 protein was tagged in the extracellular loop with the FLAG epitope and expressed in HEK-293 cells. Surface delivery and potassium-dependent sodium/calcium exchange activity were observed for each spliced variant.
- molecular cloning
- tissue distribution
- Northern analysis
- HEK-293 cells
the sodium/calcium+potassium exchanger of retinal rod photoreceptors is a critical component of the visual transduction cascade, controlling the calcium concentration of outer segments during light and darkness. The function of the exchanger has been very well characterized in situ (21, 24) and has been clearly shown to differ from the sodium/calcium exchangers of other tissues. The ubiquitously expressed sodium/calcium exchanger, exemplified by the protein expressed in the heart, transports three sodium ions in exchange for one calcium ion. The retinal rod exchanger, by contrast, transports four sodium ions in exchange for one calcium and one potassium ion (3,23). It is believed that the increase in coupling stoichiometry allows the rod exchanger to extrude calcium, even in the dark when the membrane is depolarized and the sodium gradient is reduced.
The cDNA encoding the retinal rod sodium/calcium+ potassium exchanger, gene 1 (NCKX1, SLC24A1), was cloned from cow in 1992 (19), the deduced protein sharing an overall predicted structural topology but remarkably little sequence identity with the cardiac sodium/calcium exchanger NCX1 (1, 16). Subsequently, NCKX1 was cloned from bison, human, and dolphin (4, 27). Interestingly, for species orthologs, these molecules display relatively low sequence identity with one another, especially in their large hydrophilic loops.
Until recently, it was believed that expression of NCKX1 was restricted to photoreceptors and that the particular properties of this protein had evolved to suit the special requirements for calcium homeostasis in this environment. However, physiological experiments have suggested expression of a potassium-dependent sodium/calcium exchanger in brain (5) and hemopoietic cells (8). Molecular cloning experiments have recently demonstrated the presence of NCKX1 in a human megakaryocytic cell line (9) and the expression in rat brain of a novel gene product encoding what is clearly the second member of a family of potassium-dependent sodium/calcium exchangers, NCKX2 (SLC24A2) (26). Moreover, database searches using conserved elements of sequence reveal further paralogs in various species, including nematode, plant, and yeast (17, 25). It seems likely, therefore, that the NCKX gene family has an ancient origin, possibly even predating the NCX family of sodium/calcium exchangers.
Inasmuch as NCKX1 orthologs from different species share rather low sequence identity and no paralogs within the NCKX family have been determined in the same species, we chose to clone rat eye NCKX1. In this report we demonstrate that the rat NCKX1 gene gives rise to four different alternatively spliced species, all of which are present in rat eye. We have expressed clones corresponding to each of these spliced species in HEK-293 cells and monitored their function with fura 2 imaging.
All molecular procedures were performed essentially according to standard protocols (2, 20) or the directions of reagent manufacturers, unless indicated otherwise. Chemicals were all analytic-grade quality or higher and were obtained from Fisher, BDH, or Sigma Chemical, unless indicated otherwise. Nucleic acid and protein amino acid sequence analysis was performed with the MacVector software package (Oxford Molecular Group) and by connection to the National Center for Biotechnology Information at the National Institutes of Health (http://www.ncbi.nlm.nih.gov).
Total RNA isolated from rat eye by guanidine isothiocyanate extraction and CsCl centrifugation was reverse transcribed with Superscript II (Life Technologies) with oligo(dT) as primer. For the full-length rat eye NCKX1, degenerate primers were made on the basis of the published bovine and human cDNA sequences that overlapped start and stop codons and contained EcoR I restriction sites at their 5′-ends: CGCGAATTCCACCATGGGGAAAYTGATCAGGATGGG (Sp1) and CCGGAATTCTCAGACAGATACAGGRCAGGATAT (Sp2).
Amplification was performed using the High Fidelity system (Roche Molecular Biochemicals), essentially according to the manufacturer's instructions. For analysis of the alternatively spliced region of the cytoplasmic loop, primers were based on the rat cDNA sequence flanking the sites of splicing: AGCCAAAGTCATGGCTCTAGGAGACCTCAGC (As1) and ATGTCTCCTCTGTGTTCAGCCTCATCCACATC (As2).
Amplification was carried out with Taq polymerase. In all cases, amplified products were subcloned in pBluescript II SK (−), sequenced using the AmpliTaq-FS system (Perkin-Elmer), and analyzed at the University of Calgary DNA Services facility. Full-length NCKX1 clones were fully sequenced in both directions.
Individually, the nine NCKX1 clones that were sequenced contained a number of nucleotide substitutions, presumably as a consequence of PCR incorporation errors, that resulted in amino acid replacements as follows (numbering corresponds to that shown in Fig.1): clone A, 4 nucleotide substitutions and 2 amino acid replacements: T690 replaced by A and A716T (the remaining nucleotide substitutions being silent at the amino acid level); clone B, 1 nucleotide change resulting in P399S and 1 nucleotide deletion resulting in M1169C and truncation of the subsequent 12 amino acids due to the creation of a stop codon; clone C, 3 nucleotide substitutions producing K119E and E842D; clone D, 7 nucleotide substitutions resulting in R12G, D453A, G735E, E822G, K989R, and N1092S; clone 4, 4 nucleotide substitutions producing D699E, Q930P, and L1069P; clone 6, 5 nucleotide substitutions resulting in M217T, N400D, K837T, and E1052G; clone 7, 2 nucleotide substitutions producing G856E and N1149S; clone 8, four nucleotide substitutions resulting in S23G; clone 9, 3 nucleotide substitutions producing A82V and T933A.
Northern blot analysis.
Ten micrograms of total RNA, isolated from each rat tissue by guanidine isothiocyanate extraction and CsCl centrifugation, were separated on 0.8% agaorse-formaldehyde gels and transferred to a nylon membrane by capillary diffusion overnight, and the RNA was fixed in place by ultraviolet cross-linking. Equal loading of lanes was determined by ultraviolet spectrophotometric quantitation of RNA samples and confirmed by ethidium bromide staining of rRNA bands. The blots were hybridized with digoxigenin-UTP-labeled riboprobes, essentially according to the directions of the manufacturer (Roche Molecular Biochemicals) and washed at high stringency at 65°C. Three different probes were used: probe 1 corresponded to aHind III fragment of rat NCKX1 encompassing the region encoding amino acids 1–455; probe 2 corresponded to a PCR fragment of rat NCKX1 covering the region coding for amino acids 1035–1152 (based on the full-length clone; Fig. 1); probe 3 was aHind III fragment derived from rat NCKX2 (26) and corresponded to the region encoding amino acids 27–268.
Expression of rat eye NCKX1 in HEK-293 cells.
Cloned NCKX1 molecules were opened by digestion with Spe I, which cuts within the large NH2-terminal hydrophilic extracellular domain, and a double-stranded oligonucleotide encoding the amino acid sequence of the FLAG epitope was inserted: CTAGTGACTACAAGGACGACGATGACAAGACTAG (including the Spe I overhanging ends; this fragment adds the amino acids SDYKDDDDKT, inserted between Thr-321 and Ser-322 of rat NCKX1, as illustrated in Fig. 1).
These epitope-tagged constructs were confirmed by sequencing and then subcloned into the mammalian expression vector pMT2 (Genetics Institute, Boston, MA). Qiagen-purified cDNA was used for expression in HEK-293 cells by use of standard calcium phosphate-N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid transfection procedures. Two days after transfection, protein expression was determined by immunofluorescence, immunoblotting, or fura 2 imaging, essentially as previously described (26).
The structure of the rat eye NCKX1 molecule was investigated using RT-PCR with primers based on the sequence of the regions flanking the start and stop codons from bovine and human NCKX1 cDNAs. A prominent broad band of ∼3.6 kb was observed that was absent from samples incubated without RT. Visual inspection suggested that the amplified band comprised several species. The band was excised, digested with EcoR I, and subcloned into pBluescript II SK (−). A total of 13 independently isolated clones, from 3 different RT-PCR reactions, were purified and subjected to restriction enzyme digestion mapping and nucleotide sequencing at their ends. This analysis revealed that four clones had deletion artifacts or were poorly expressed; the remaining nine appeared to be very similar. These nine clones were sequenced in their entirety and, aside from a region encoding 113 amino acids in the central hydrophilic cytoplasmic loop that was partially or entirely deleted in some cases, all sequences were essentially identical and clearly encoded the rat NCKX1 protein (Fig. 1, see Fig. 3). Each individual rat NCKX1 clone differed from the consensus sequence by two to seven nucleotides, resulting in one to six amino acid replacements. All these changes were at different positions and, we presume, arose from PCR incorporation errors. Figure 1illustrates the consensus-deduced amino acid sequence of our longest rat eye NCKX1 clones, with various features highlighted; Fig. 3compares the rat NCKX1 sequence with those of rat NCKX2 and human and bovine NCKX1.
The phenomenon of sequence “dropouts” in various clones suggested a cassette-exon type of alternative splicing process at work. To confirm and extend these results, we made a new primer set that flanked the region of apparent alternative splicing and repeated the RT-PCR amplification of rat eye RNA. The results of this experiment are shown in Fig. 2, together with a schematic diagram illustrating the location of the region of splicing within the NCKX1 protein. Four ethidium bromide-stained, amplified bands were clearly evident. Each band was excised, subcloned, and sequenced and determined to be identical to at least one of the NCKX1 clones. Indeed, the nine different clones fell into four categories, each category corresponding to one of the PCR bands, as indicated in Fig. 2. In all cases, the deleted regions did not influence the reading frame of the remaining coding region. Compared with the sequence and structure of the human NCKX1 gene (28), it is clear that the regions of dropout correspond to inclusion or exclusion of four distinct exons that we have denoted alternatively spliced exons A–D. These correspond to exons 3–6 of the human NCKX1 gene. In support of cassette-exon alternative splicing, independent clones of the human NCKX1 gene have been described that contain (9) or lack (27) exon 3.
Figure 3 shows an alignment of rat NCKX1 with orthologs from human and cow and with the rat paralog NCKX2. Compared with rat brain NCKX2, the rat eye NCKX1 protein is ∼55% identical in the regions of overlap. The identity, however, is restricted almost entirely to the two clusters of hydrophobic putative transmembrane spans, where it is >80%. Located within these hydrophobic domains are the regions of apparent sequence duplication, the α-subunit repeats (Fig. 2), which are thought to define the ion transport binding pocket (15, 25). Outside the hydrophobic areas, the identity drops to <15%. A notable exception, however, is the region encoded by exon B of the alternatively spliced region, which shares 58% identity with NCKX2. Interestingly, this segment of NCKX2 is also a distinct exon within a region of alternative splicing identified in mouse (unpublished observations).
The rat, human, and bovine NCKX1 protein orthologs share ∼60% identity. Again, this is highest within the hydrophobic regions but also includes the segments encoded by the alternatively spliced exons. The large extracellular NH2-terminal loop is ∼35% identical between the three animal species. The two potential glycosylation sites in rat NCKX1 are conserved with human and cow, although the latter has one and four additional sites, respectively.
The cytoplasmic loop of NCKX1 downstream from the region of alternative splicing shares the lowest identity between animal species. This region is highly acidic and contains an imperfectly repeated motif. As illustrated in Figs. 1 and 4, the rat NCKX1 possesses 14 repeats with a consensus of ETEAEGKEVEHEG. The bovine NCKX1 contains nine copies of a similar but slightly longer sequence (consensus, DEDEGEIQAGEGGEVEG), and the buffalo NCKX1 has a motif virtually identical to that of the cow but is present in only seven copies, whereas human NCKX1 has essentially no discernible repeat structure in this region (27).
Although it has been reported that expression of the NCKX1 protein is restricted to the eye (19), recent data demonstrate that the protein is also expressed in human platelets and a human megakaryocytic cell line (9). Nevertheless, to our knowledge, a comprehensive analysis of expression across a broad selection of tissues has not been reported previously. Therefore, we used Northern blot analysis to examine the expression of NCKX1 in different rat tissues (Fig.5). Long exposures of these blots are presented purposefully to reveal bands of lower intensity, and two different probes have been used. Probe 1 (Fig. 5 A) corresponded largely to the unique NH2-terminal extracellular portion of rat NCKX1; probe 2 (Fig. 5 B) encoded the COOH-terminal hydrophobic portion conserved between NCKX paralogs. In addition, a blot probed with a fragment from the 5′-end of the rat NCKX2 clone (encompassing unique and partially conserved sequences) is shown for comparison (Fig. 5 C).
Probe 1 recognized a transcript of ∼7 kb found exclusively in RNA from the eye and clearly did not cross-react with NCKX2 transcripts (11-kb band in lanes from brain and adrenal tissue; Fig. 5 C). The eye lane in Fig. 5 A and the brain and eye lanes in Fig.5 C have been exposed for one-tenth the time of all other lanes.Probe 1 detected bands at ∼2 kb in esophagus, lung, spleen, thymus, and adrenal tissues. These species were not recognized by the other probes and, in any event, would be too short to encode a full-length NCKX2 protein. Probe 1 also recognized faint bands at ≥11 kb in eye, lung, and thymus, as well as faint, smeared bands at ∼4 kb in most tissues that were somewhat obscured, and their migration somewhat distorted, by weak cross-hybridization to the 28S rRNA band.
Probe 2 clearly recognized NCKX1 in eye (7 kb) and NCKX2 in different brain regions (11 kb). Strong bands were also seen at 3.5 kb in stomach and at 2.5 kb in adrenal tissue that were not recognized by the other two probes. Probe 2 also recognized a band at 4 kb in many tissues that was particularly evident in diaphragm, heart, and aorta. Fainter bands were also seen at 3.5 and 2 kb, especially in brain and diaphragm. In addition, probe 2 recognized transcripts of ≥11 kb in eye, diaphragm, lung, and adrenal tissue. The 11-kb band observed in adrenal tissue, but not the other tissues, was recognized by the NCKX2 probe and so probably represents cross-reaction with NCKX2 transcripts expressed in this tissue.
Protein expression from the alternatively spliced NCKX1 clones was analyzed in transfected HEK-293 cells. To detect expressed protein, we first introduced the FLAG epitope into the external domain of the exchanger at the position illustrated in Figs. 1 and 2. Clones in the pMT2 vector were subsequently introduced into HEK-293 cells by calcium phosphate transfection. As illustrated in Fig.6 A, for six of the nine clones, protein bands ranging from 170 to 200 kDa were observed with anti-FLAG antibody in extracts from transfected, but not from control, cells. Although the relative shifts in protein size were in accordance with expectations from the extent of deletion in the different clones (Fig.2), the predicted polypeptide size ranged from only 119 to 131 kDa. The difference between predicted and actual size is likely to be caused by a combination of protein glycosylation and the acidic nature of the NCKX1 polypeptide (predicted pI of 4.37) (19).
Surface delivery of the expressed NCKX1 molecules was assessed in HEK-293 cells by immunofluorescence. Transfected cells were fixed but not permeabilized, since the FLAG epitope was expected to be exposed extracellularly, and then incubated with the M2 anti-FLAG antibody followed by a rhodamine-conjugated secondary antibody. Figure6 B shows images at two different focal planes through a selected cell from two different NCKX1 construct transfections and from a control transfection. Intense, punctate staining was observed on the surface of the transfected cells that was absent from the control. This pattern of surface staining is consistent with that previously observed for rat NCKX2 (26). Similar patterns of surface staining were observed for all six alternative constructs of Fig. 6 A (data not shown). Although there was considerable variability from experiment to experiment and from construct to construct, examination of fields of cells at low power indicated transfection efficiency of 10–30%. In general, there was a direct correlation between percent transfection and the level of protein expression observed by immunoblot for the different constructs. Protein expression tended to be better for the longer alternatively spliced molecules (all exons included, clones 9 and C; exon C only excluded, clones 8 and A) than for the shorter molecules.
Calcium transport function of the alternatively spliced rat eye NCKX1 molecules was assessed by digital imaging of fura 2-loaded HEK-293 cells that had been transfected with different constructs. The same six clones that were tested for protein expression were also tested for calcium transport function, and all showed evidence of potassium-dependent sodium/calcium exchange activity to varying extents. Figure 7 illustrates averaged data for control, NCKX2, and two different NCKX1 constructs obtained from the indicated number of responding cells collected over two to three different experiments for each construct. Loaded cells were initially perfused for 5 min with a sodium-containing potassium-free buffered solution. A switch to a potassium-free lithium-containing solution elicited a slow rise in fura 2 fluorescence for the transfected, but not for the control, cells. Subsequent addition of 5 mM potassium to the perfusate caused a marked increase in fura 2 fluorescence, consistent with sodium efflux-driven calcium uptake that is potassium dependent. A perfusion switch back to a sodium-containing solution resulted in a rapid decline in fura 2 fluorescence, as calcium was removed from the cytoplasm when entry through the NCKX1 molecules ceased. These data demonstrate that all alternatively spliced NCKX1 molecules, from the longest (clone 9) to the shortest (clone 6), are able to function in calcium transport.
We have isolated full-length cDNA clones encoding the rat eye NCKX1 molecule. As anticipated from previously published sequences (4, 27), rat NCKX1 shares a very high degree of identity with its species orthologs among the hydrophobic regions of the molecule, which are thought to associate within the membrane to form the ion binding and translocation sites. Additional blocks of similarity exist at the protein NH2 terminus, throughout the large extracellular loop between M0 and M1, and at either end of the cytosolic loop (Fig. 3). The rat paralogs, NCKX1 and NCKX2, also share a high degree of identity within the hydrophobic regions. The character and potassium dependence of calcium transport that we observed for NCKX1 and NCKX2 molecules (Fig. 7) suggest that the high degree of structural similarity underlies a striking functional similarity.
A notable feature of the bovine NCKX1 sequence is an acidic repeat sequence in the cytosolic domain that is largely absent from the human exchanger. Interestingly, the rat NCKX1 also possesses an acidic repeat structure at this site, although neither the sequence, length, nor number of repeated units is conserved (Fig. 4). The β-subunit of the bovine rod cyclic nucleotide-gated channel also contains a glutamic acid-rich, acidic motif reminiscent of the NCKX1 repeat (11). Although the functional role of this repetitive sequence is not known for the channel or the exchanger, such an acidic region might be involved in interactions with other proteins (11) or possibly in calcium regulation of transport (14, 22). It remains curious, however, that this striking repeat is not better conserved among different mammalian species, considering that photoreceptor architecture, the visual transduction process, and the role of calcium in that process are thought to be highly conserved among essentially all vertebrates.
A unique finding among our rat NCKX1 clones was an apparent region of alternative splicing near the beginning of the large central cytoplasmic domain (Figs. 1 and 2). Alternative splicing of this region in human NCKX1 had been implicated by the isolation of individual clones that include or exclude exon 3 (alternatively spliced exon A) (9, 27, 28) and has been recently documented at the mRNA level in human and cow. Curiously, the exon 3-deleted alternatively spliced species that is most common in humans is not visible in rat (unpublished observations) (Fig. 2). We have also observed that NCKX2 is subject to alternative splicing over a similar region (26; unpublished observations). Indeed, exon B of the NCKX1 alternatively spliced region is conserved with a similar exon in NCKX2, the only extensive similarity between these two proteins outside the membrane regions. This region contains a number of Ser and Thr residues, at least one of which (Ser-625; Fig. 1) falls in a weak consensus for protein kinase A or calmodulin-dependent kinase phosphorylation. Possibly this conserved exon plays a role in regulating transport activity in the NCKX family of exchangers.
Alternative splicing of the NCX1 gene has been extensively documented (10, 12, 18). Recent evidence suggests that, at least in NCX1, alternative splicing leads to significant functional differences between isoforms (6, 7, 13). The different NCX1 splice variants are expressed in a tissue-specific manner, however, although those of NCKX1 are, presumably, restricted to the outer segment of the photoreceptors in the eye. It is, thus, unclear why more than one NCKX1 isoform should be expressed. Perhaps expression of different splice variants is developmentally controlled, or the variants have distinct cellular or subcellular distributions. Resolution of this issue will require analysis of NCKX1 transcripts by RT-PCR, examination of their location by in situ hybridization, and examination of different alternatively spliced protein products by immunocytochemistry.
Recent observations from one of our laboratories have demonstrated that a region from the bovine NCKX1 molecule that corresponds almost exactly to the full extent of the alternatively spliced region reported here (exons A–D) appears to be responsible for generating a functionally silent phenotype (4). When this region was present in bovine or dolphin NCKX1 molecules, the constructs encoded proteins that did not express transport function in a heterologous expression system, although protein was expressed. Deletion of this region from either clone resulted in an exchanger capable of calcium transport. Although the amino acid sequence of this region is highly conserved between bovine and rat NCKX1, the functional silencing phenotype is not. Thus we have observed that the longest rat NCKX1 splice variant, containing the entire exon A–D region, as well as the shortest variant, lacking all four exons, displayed calcium transport function. Indeed, the longest rat NCKX1 protein isoform produced the most robust protein expression and calcium transport rates. Although it seems likely that the functionally silent phenotype is a consequence of heterologous expression (in bovine photoreceptors, where the exchanger is clearly functional, transcripts encoding the longest protein are in the majority; unpublished observations), it seems likely that this observation may provide clues to understanding targeting or regulation of the NCKX1 protein.
The tissue distribution of rat NCKX1 expression was assessed by Northern blot analysis (Fig. 5). When analyzed using a probe from a unique sequence, it was clear that expression of NCKX1 was much higher in the eye than in any other tissue. A recent report demonstrated that NCKX1 was expressed in human platelets and a megakaryocytic cell line (9). Although we have not analyzed the corresponding rat tissues or cells for NCKX1 transcripts, one presumes that the spleen or thymus would likely contain cells of similar lineage. In those tissues the NCKX1-specific probe recognized transcripts of ∼4 kb weakly (also present in many other tissues) and even fainter ones at ≥11 kb.
Interestingly, analysis with a probe corresponding to a region of the protein conserved between NCKX paralogs uncovered a number of bands that could not be accounted for by cross-reaction to NCKX1 or NCKX2 transcripts expressed in those tissues. The most evident of these were a 3.5-kb band in stomach, a 2.5-kb band in adrenal tissue, a 4 kb-band in skeletal muscle, and an 11-kb band in lung tissue. The genes encoding these transcripts represent good candidates for further members of the NCKX family.
In conclusion, we have demonstrated that the rat eye sodium/calcium+potassium exchanger is expressed as a family of alternatively spliced transcripts that encode proteins that differ in a region of the central cytoplasmic loop. All these protein isoforms are capable of catalyzing calcium transport when expressed in HEK-293 cells.
We thank Randal Kaufman (Genetics Institute, Boston, MA) for the gift of vector pMT2.
Address for reprint requests and other correspondence: J. Lytton, University of Calgary Health Science Centre, 3330 Hospital Dr., Calgary, AB, Canada T2N 4N1 (E-mail:).
This work was supported by Medical Research Council Group Operating Grant MGC-13719 (to J. Lytton and P. P. M. Schnetkamp) and by a graduate studentship (to J. E. Tucker) and a summer studentship (to S. Poon) from the Alberta Heritage Foundation for Medical Research. J. Lytton is an Established Investigator of the American Heart Association and a Medical Scholar of the Alberta Heritage Foundation for Medical Research. P. P. M. Schnetkamp is a Medical Scientist of the Alberta Heritage Foundation for Medical Research.
The consensus nucleotide sequence for rat NCKX1 has been deposited in the GenBank database (accession no. AF176688).
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. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society