Identification of a novel cysteine string protein variant and expression of cysteine string proteins in non-neuronal cells.

Cysteine string proteins (Csps) are synaptic vesicle proteins thought to be involved in calcium-dependent neurotransmitter release at nerve endings. Here, we report the cloning of two Csp variants, termed Csp1 and Csp2, from bovine adrenal medullary chromaffin cells. The bovine Csp1 appears to be the homologue of rat brain Csp, sharing 95% identity at the amino acid level. The nucleotide sequence of csp2 is identical with that of csp1 except for a 72-base insert which introduces a stop codon into the coding sequence, which would be predicted to result in a truncated protein 3.3 kDa smaller than Csp1. Furthermore, polymerase chain reaction analysis detected homologues of Csp1 and Csp2 in rat kidney, liver, pancreas, spleen, lung, and adrenal gland. Expression of Csps in non-neuronal tissues was confirmed by Northern blotting and by immunoblotting with anti-Csp1 antiserum which also demonstrated expression of both full-length and truncated Csps in spleen. The widespread tissue distribution is inconsistent with a role of Csps as specific regulators of presynaptic calcium channels as previously proposed. We suggest that Csps may have a more general role in membrane traffic in non-neuronal as well as neuronal cells.

gous to DnaJ proteins which interact with the bacterial homologue of the chaperone protein Hsp70 (5). Drosophila Csps exist in at least two forms generated by alternative RNA splicing (4), and a single related Csp was found in Torpedo (6).
An independent study found that Torpedo csp antisense cRNA virtually abolished expression of an N-type calcium channel in Xenopus oocytes injected with Torpedo mRNA (6). Torpedo Csp was, therefore, proposed to be an essential subunit or modulator of presynaptic calcium channels. However, Csps were subsequently found to copurify with synaptic vesicles and not with presynaptic membranes (7), prompting the proposal that Csps on the synaptic vesicle membrane may interact with presynaptic calcium channels converting them from an inactive to an active state. Torpedo Csp has been shown to be post-translationally modified by the palmitoylation of 11 or 12 of the 13 cysteine residues, and this fatty acylation is thought to tether Csps to the synaptic vesicle membrane (8).
Genetic analysis found that deletion of the entire Drosophila csp locus or the promoter sequence and first exon conveyed a temperature-sensitive lethal phenotype which was subsequently characterized as causing a defect in presynaptic neurotransmission (9,10). The fact that the mutant phenotype was more pronounced at 30°C than at 22°C suggests that Csps may stabilize components of the neurotransmitter release machinery. Interestingly, deletion of dnaJ in Escherichia coli causes a temperature-sensitive phenotype suggestive of chaperone activity (11).
A Csp has recently been cloned from rat brain, suggesting that these proteins may also play an important role in mammalian presynaptic neurotransmission (12). It has been suggested that Csps function in membrane fusion (13), but their exact role in the nerve terminal is still not known. In this paper we report the cloning of two csp coding sequences from bovine adrenal medullary chromaffin cells. The first encodes a protein identical in size with the rat brain form, whereas the other, which appears to be a splice variant of the first, encodes a truncated protein approximately 3.3 kDa smaller in size. Subsequent PCR, Northern, and immunoblotting analysis revealed that Csps are not brain-specific as previously reported (4), but are in fact found in a range of non-neuronal tissues, suggesting a more general function for these proteins.

EXPERIMENTAL PROCEDURES
Materials-Restriction enzymes, Taq polymerase, T4 DNA ligase, reverse transcription system, and plasmid purification kits were all obtained from Promega. RNeasy RNA isolation kit, pQE-30 plasmid DNA, Ni 2ϩ -NTA agarose, and E. coli M15[pREP4] cells were obtained from Qiagen (Surrey, UK). DNA primers for PCR reactions were obtained from Cruachem (Glasgow, UK). Multiple tissue poly(A) ϩ RNA blot was obtained from Clontech. 32 P-Labeled dUTP was obtained from Amersham (Buckinghamshire, UK) All other chemicals were of an analytical grade from Sigma Ltd. (Poole, UK).
PCR Amplification of csp Coding Sequences-RNA was extracted from freshly isolated adrenal chromaffin cells purified by differential * This work was supported in part by a grant from the Wellcome Trust. 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 1 The abbreviations used are: Csp, cysteine string protein; PCR, polymerase chain reaction; SSPE, saline-sodium phosphate-EDTA buffer; Ni 2ϩ -NTA, nickel-nitrilotriacetic acid.
plating (14), PC12 cells or the appropriate rat tissues using a RNeasy total RNA isolation kit, following the suppliers protocol (Qiagen). cDNA was synthesized with a Promega reverse transcription system, following the suppliers protocol. A HYBAID Omn-E programmable dryblock was used for PCR reactions. The 5Ј primer was designed to contain a 22-base nucleotide homology to the template, whereas the 3Ј primer had a 24-base nucleotide homology. The primers had BamHI and HindIII restriction sites incorporated to facilitate cloning. The sense and antisense primers used were (5Ј-dAAGGATCCATGGCTGACCAGAG-GCAGCG) and (3Ј-dCTCAAGCTTTTAGTTGAACCCGTCGGTGTG), respectively, and the BamHI and HindIII restriction sites are underlined. PCR consisted of an initial denaturation cycle at 95°C for 5 min, followed by 30 cycles consisting of annealing at 52°C for 1 min, elongation at 72°C for 2 min, and denaturation at 95°C for 1 min. A further cycle at 72°C for 10 min finished the amplification process. PCR products were examined by electrophoresis on 1.5% agarose gels.
Sequencing of csp Coding Regions-The PCR products were gelpurified using a Promega PCR kit, digested with BamHI and HindIII and ligated to pQE-30 (Qiagen) and the ligation mixture used to transform competent E. coli M15[pREP4] cells (Qiagen). pQE-30 plasmid DNA was prepared from transformed M15 cells using Wizard plasmid preps (Promega), and the cloned DNA was sequenced using an automated sequencer, with sequencing of both strands being carried out.
Expression and Purification of Recombinant Csp1-Isopropyl-1-thio-␤-D-galactopyranoside (1-2 mM) was added to a 1-liter culture of M15 cells containing the cloned csp1 coding sequence and incubated at 37°C for 5 h with shaking at 250 rpm, inducing expression of the recombinant protein. The induced cells were pelleted by centrifugation at 4,000 ϫ g for 20 min and resuspended in 20 ml of breaking buffer (5 mM ATP, 100 mM HEPES, 5 mM MgCl 2 , 2 mM ␤-mercaptoethanol, 500 mM KCl, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, and 1 M pepstatin A, pH 7.0). The cells were stored at Ϫ20°C overnight and thawed the following morning. 1 mg/ml lysozyme was added, and the cells were left on ice for 30 min. Following ultrasonication, 2 g/ml DNase was added, and the cells were left on ice for 15 min. Centrifugation at 100,000 ϫ g for 1 h removed cell debris. The recovered protein fraction was loaded onto a Ni 2ϩ -NTA agarose column and washed with 50 mM imidazole buffer (50 mM imidazole, 20 mM HEPES, 200 mM KCl, 2 mM ␤-mercaptoethanol, 0.5 mM ATP, 10% v/v glycerol, pH 7.0) to remove unbound protein. His 6 -tagged Csp1 was eluted by application of a gradient of 50 -500 mM imidazole, and peak fractions containing the recombinant protein were identified by SDS-polyacrylamide gel electrophoresis. The procedure yielded ϳ16 mg of essentially pure recombinant Csp1 at a concentration of 1.65 mg/ml. Chromatography was performed at 4°C using a Pharmacia FPLC system.
Northern Blotting-A multiple tissue poly(A ϩ ) RNA blot (Clontech) was used for Northern blotting. Prehybridization and hybridization were performed at 65°C in 5 ϫ SSPE containing 50% deionized formamide, 5 ϫ Denhardt's, 0.5% SDS, and 0.2 mg/ml salmon sperm in diethyl pyrocarbonate-treated water. A complementary RNA probe to Csp1 containing 32 P-labeled dUTP was generated from linearized pBluescript SK [Csp1] (following subcloning of the Csp1 coding region from pQE-30) using T7 RNA polymerase, purified on a NucTrap column (Stratagene, Cambridge, UK), and used at 2 ϫ 10 6 cpm/ml. Following hybridization, the membrane was washed twice in 0.1% SDS, 2 ϫ SSPE at room temperature for 10 min and subsequently in 0.1% SDS, 0.1 ϫ SSPE at 65°C for 20 min. A PhosphorImager was used to detect binding of the 32 P-labeled probe.
Preparation of Antiserum-A rabbit was injected with 150 g of purified recombinant Csp1 in a 1:1 mixture with Freund's complete adjuvant at multiple sites subcutaneously. 4 weeks later, 150 g of Csp1 was injected into the same rabbit at multiple sites subcutaneously in a 1:1 mixture with Freund's incomplete adjuvant, and this was repeated 2 weeks later. After an additional 7 days, blood was taken from the rabbit and the recovered antiserum was stored at Ϫ20°C.
Western Blotting-Tissue homogenates were prepared from Wistar rats by homogenization in 5 mM Tris-HCl (pH 8.0), and SDS-dissociation buffer was added to give a final protein concentration of 1 mg/ml. 2 mM dithiothreitol was added, and the homogenates were left at room temperature for 1 h before boiling for 5 min. The boiled samples were separated on a 12.5% polyacrylamide gel. Transfer of proteins to nitrocellulose was achieved by transverse electrophoresis at 90 mA overnight. The nitrocellulose was then incubated with antiserum raised in rabbits against recombinant Csp1 (1:1000 dilution). Antibody binding was detected using enhanced chemiluminescence (ECL, Amersham International).

RESULTS
Since Csps are highly conserved, a strategy was adopted to allow the PCR amplification of the coding sequence of bovine chromaffin Csp using oligonucleotide primers based on the rat brain csp sequence but incorporating restriction sites to allow direct cloning into vectors for expression of His 6 -tagged proteins. In initial experiments, mRNA was isolated from purified bovine adrenal medullary chromaffin cells, the rat clonal PC12 pheochromocytoma cell line, and rat brain. Reverse transcription of the mRNA was used to generate complementary cDNA which was subsequently amplified by PCR using oligonucleotide primers designed to the 5Ј and 3Ј regions of the rat csp coding sequence. The result of the PCR amplifications is shown as Fig. 1. Only one size of DNA fragment was amplified from rat brain, and this is in agreement with a previous study (12), suggesting the presence of only a single csp species in rat brain. However, a slightly larger DNA molecule was amplified from chromaffin and PC12 cell cDNA in addition to a fragment apparently identical in size with that from rat brain.
The DNA fragments obtained from PCR amplification of chromaffin cell cDNA were purified, ligated to BamHI/HindIIIdigested pQE-30, and transformed into E. coli M15[pREP4] cells. Of 9 clones that were isolated, 3 contained a smaller PCR product. From sequencing, 2 encoded the likely bovine homologue of the rat brain csp (Fig. 2). This species which we term bovine Csp1 showed 88% nucleotide and 95% amino acid sequence identity with the rat brain Csp. Six clones encoded a likely splice variant, which we term Csp2, since it was identical with the first coding sequence apart from a 72-base insert (GGAGGGCACTGACCTGTGCGGGAGTGTTTGTGGTGGCA- FIG. 1. Amplification of rat brain, chromaffin cell, and PC12 cell csp DNA. The DNA species were amplified by PCR from cDNA using primers designed to the 5Ј and 3Ј ends of the rat brain csp coding sequence with restriction sites incorporated. Amplification from chromaffin cell and PC12 cell cDNA generated two DNA species (indicated by arrows), the smaller of these being identical in size with that generated from rat brain cDNA.
GCGGGACGGTTGAGGTGTGAACGTGGACGCTGGA after nucleotide 492). One clone containing a smaller PCR product possessed this insert but also had an 88-base deletion at position 234 which would result in a reading frameshift after amino acid 78. This last clone was not characterized further. All sequences shown (Fig. 2) are consensuses of multiple sequence runs of all clones to eliminate problems due to PCR amplification errors. In addition, all nucleotide differences between bovine Csp1 and rat Csp1 were also present in Csp2. The additional insert present in csp2 introduces two stop codons into the amino acid sequence at positions 168 and 181, and thus the Csp1 and Csp2 proteins have expected molecular masses of 22,221 Da (Csp1) and 18,904 (Csp2).
In order to confirm that the cloned sequences encoded His 6tagged proteins of the predicted molecular mass, expression of the cloned Csps was induced by addition of isopropyl-1-thio-␤-D-galactopyranoside to transformed bacterial cultures. The expression of a new protein after induction was clearly visible, the induced proteins were subsequently partially purified using Ni 2ϩ -NTA agarose resin, and the bound proteins were analyzed (data not shown). The induced proteins binding to the Ni 2ϩ -NTA agarose have apparent molecular masses of 27 and 22 kDa, respectively. Although the apparent molecular sizes of the proteins are higher than the expected values, this has also been found for other identified Csps (12). The data on the expressed proteins confirm that the cloned cDNAs encode proteins of the expected size and that plasmids containing the coding sequence with the 72-base insert encode a truncated protein as predicted.
Csps have previously been suggested to be specifically localized to synapses (4). The data presented so far already show that they are also expressed by the non-neuronal chromaffin and PC12 cells, and this was investigated further by isolating mRNA from several different rat tissues and using the primers designed to the rat csp coding sequence to amplify the cDNA by PCR. The result of this is shown in Fig. 3a. All tissues examined resulted in the amplification of two cDNA bands identical in size with DNA-encoding Csp1 and Csp2, but with differing ratios of the two products from the various tissues. The presence of mRNA-encoding Csps was confirmed by Northern blotting. RNA blots loaded with poly(A) ϩ -enriched RNA from various rat tissues were hybridized at high stringency with an RNA probe complementary to the Csp1 coding region, and mRNAs were detected in heart, spleen, lung, liver, muscle, kidney, and testis in addition to brain (Fig. 3b). In each tissue, an mRNA species of around 5 kilobases was detected which was more abundant in brain and brain also contained a larger mRNA.
To directly confirm expression of Csp in non-neuronal tissues, an antiserum was raised using purified expressed Csp1 as the antigen. The antiserum recognized both Csp1 and Csp2 recombinant proteins, and, in immunoblotting on rat brain homogenates, recognized a polypeptide of around 36 kDa as expected for the post-translationally modified Csp1 (Fig. 4). The antiserum also specifically recognized a larger polypeptide of around 70 kDa, the amount of which varied between experiments. It is well established that Csps in Torpedo and rat brain form dimers detectable by immunoblotting (12,16), and it is likely that the antiserum recognizes only Csp1 and its dimer in brain homogenates. The antiserum also recognized an abundant 36-kDa polypeptide and the dimer in chromaffin cells. Since chromaffin cells are derived from neuronal precursor cells, we examined the possible non-neuronal expression of Csp in kidney and spleen, tissues that would not be expected to contain any neuronal-like cells. Immunoblotting showed less abundant but detectable expression of Csp1 in both kidney and spleen. Polypeptides corresponding to Csp dimers were also seen in these tissues. Immunoreactivity was also detected in liver and pancreas, but this was at the limit of detectability. An immunoreactive polypeptide, corresponding in size to that expected for post-translationally modified Csp2, was detected in spleen (Fig. 4). DISCUSSION We have cloned and characterized two Csp variants from bovine adrenal chromaffin cells. PCR amplification of cDNA from the clonal PC12 cell line also generated two DNA species identical in size with csp1 and csp2 confirming that one cell type can express two distinct Csps. DNA sequencing demonstrated that the two plasmid inserts encode variant Csps,  FIG. 3. PCR and Northern blot analysis of csp1 and csp2 distribution in rat tissues. a, amplification of csp DNA from the various named tissues was achieved using PCR with primers designed to the 5Ј and 3Ј ends of the coding sequence of rat brain csp. csp DNA was amplified from all tissues examined but with apparent differential expression of the two csps in the different tissues. b, RNA blots containing poly(A) ϩ -enriched RNA were hybridized with a probe encompassing the coding region of Csp1 and the hybridization signal was detected using a PhosphorImager.
which are highly homologous at the nucleotide and amino acid level to the previously identified Csps. Protein expression confirmed that proteins of the correct molecular mass were synthesized and most notably that csp2 encodes a truncated protein, as predicted by DNA sequence analysis. The truncated Csp2 has intact "J" and cysteine-rich domains and is thus a Csp, but lacks the extreme C terminus possessed by the other members of the Csp family. Future studies of the functions of Csp1 and Csp2 should indicate the role of the C-terminal domain.
Previous studies have only detected one Csp protein isoform in both rat brain and Torpedo electric lobe (12,15). However, in both cases, detection was by Western blotting using antibodies specific for the C-terminal region of Torpedo Csp. These antibodies would clearly not bind to the truncated Csp2 which lacks the C terminus. The present study failed to identify by PCR a larger DNA species (similar to that encoding Csp2) in rat brain, and it may well be that brain expresses only the full-length Csp1.
Using PCR we have detected expression of the two csp mRNAs in rat kidney, liver, spleen, pancreas, lung, and adrenal gland (all tissues examined), and Northern blotting detected mRNAs in brain, heart, spleen, lung, liver, muscle, kidney, and testis, implying that Csps are widespread. This finding is in contrast to previous work which has claimed that Drosophila Csps are synapse-specific (4), and that csp mRNA is not present in electric organ, liver, or muscle of Torpedo (15, 16). Analysis of protein expression by immunoblotting with antiserum raised against recombinant Csp1 confirmed expression in chromaffin cells and other non-neuronal tissues. Csp2 was readily detected only in spleen suggesting that this variant protein is expressed at low levels. In immunoblotting, the polypeptide corresponding to Csp2 was detected in spleen but not in kidney despite similar amounts of the two PCR products being detected in both these and other tissues. It should be noted, however, that the PCR method used was not quantitative and the relative amounts of the two PCR products would not reflect the relative amounts of the mRNAs for the two Csps in these tissues. The ratio of the Csp dimer to monomer detected by immunoblotting varied considerably between differ-ent tissues. The significance of these differences is not clear since the ratio of dimer to monomer varied widely between experiments even with the recombinant Csp1. We were unable to find conditions that prevented dimer formation. While this paper was under review, expression of Csp1 was demonstrated in adrenal medulla (17) consistent with our findings. Interestingly, a number of proteins involved in exocytosis previously thought to be brain-specific have since been found in nonneuronal cell types, most recently the Ca 2ϩ -sensing protein synaptotagmin (18), setting a precedent for the present finding.
The implications of the apparently ubiquitous tissue distribution of csp are manyfold. It has been suggested that Csps may regulate voltage-dependent Ca 2ϩ channels (6) and mediate membrane fusion (13). Our finding that Csps are not restricted to synapses makes their prime function unlikely to be the regulation of Ca 2ϩ channels as many of the tissues examined do not have such channels. Alternatively, Csps may be universal membrane traffic proteins such as NSF and SNAPS, which function in both constitutive and regulated secretion. It is possible that Csps could mediate membrane fusion as suggested previously (13), but our findings do not restrict the proteins to the synapse. If Csps do function in membrane fusion at the synapse during neurotransmission (13), then it seems possible that they also play a fundamental and more general role in membrane traffic within the cell.