Purification and Characterization of the Human SR 31747A-binding Protein

SR 31747A, defined as a sigma ligand, is a novel immunosuppressive agent that blocks proliferation of human and mouse lymphocytes. Using a radiolabeled chemical probe, we here purified a target of SR 31747A and called it SR 31747A-binding protein (SR-BP). Purified SR-BP retained its binding properties and migrated on SDS-polyacrylamide gel as a M r 28,000 protein. Cloning of the cDNA encoding human SR-BP shows an open reading frame for a 223-amino acid protein, which is homologous to the recently cloned sigma 1 receptor. Interestingly, the deduced amino acid sequence was found to be related to fungal C8-C7 sterol isomerase, encoded by the ERG2 gene. The ERG2 gene product has been identified recently as the molecular target of SR 31747A that mediates antiproliferative effects of the drug in yeast. Northern blot analysis of SR-BP gene expression revealed a single transcript of 2 kilobases which was widely expressed among organs, with the highest abundance in liver and the lowest abundance in brain. Subcellular localization analysis in various cells, using a specific monoclonal antibody raised against SR-BP, demonstrated that this protein was associated with the nuclear envelope. When studying the binding of SR 31747A on membranes from yeast expressing SR-BP, we found a pharmacological profile of sigma 1 receptors; binding was displaced by (+)-pentazocine, haloperidol, and (+)-SKF 10,047, with (+)-SKF 10,047 being a more potent competitor than (−)-SKF 10,047. Scatchard plot analysis revealed K d values of 7.1 nm and 0.15 nm for (+)-pentazocine and SR 31747A, respectively, indicating an affinity of SR-BP 50-fold higher for SR 31747A than for pentazocine. Additionally, we showed that pentazocine, a competitive inhibitor of SR 31747A binding, also prevents the immunosuppressive effect of SR 31747A. Taken together, these findings strongly suggest that SR-BP represents the molecular target for SR 31747A in mammalian tissues, which could be critical for T cell proliferation.

Sigma receptors, initially identified by Martin et al. (1) as a subclass of opiate receptors, were characterized by Su (2,3) in guinea pig brain. They were subsequently found to be distributed in a variety of distinct regions of the central nervous system (4), as well as in endocrine-related structures (5), gastrointestinal tract (6), liver (7), and kidney (8). Sigma receptors were further identified in human peripheral blood mononuclear cells (9 -12), and their presence was confirmed in rat spleen. The biological significance of sigma receptors in the immune system is poorly understood. However, Liu et al. (12), using a panel of sigma ligands, demonstrated a high correlation between drug binding potency at sigma sites and inhibition of splenocyte proliferation, suggesting that sigma receptors may function as immunoregulators.
SR 31747A was first described by Paul et al. (13) as a ligand able to compete with a very high potency with all known sigma ligands and to inhibit mitogen-induced mouse and human T cell proliferation with a potency similar to that of cyclosporin A (14). However, unlike cyclosporin A, SR 31747A does not alter early cellular events, but rather interferes at steps further along the pathway leading to proliferation, most probably during the S phase (14).
Pharmacological studies with SR 31747A cannot discriminate among sigma 1, sigma 2, sigma 3, and the novel sigma/ opioid binding sites since multiple types of sigma receptors can coexist on cells (19). Moreover, the cross-reactivity of many drugs that bind with high affinity to other receptors has impeded efforts to define precisely the sites for SR 31747A binding.
Recently, we examined the effect of this drug in the yeast * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) U79528.
Saccharomyces cerevisiae. We showed that SR 31747A arrested cell proliferation by inhibiting an enzyme of the sterol biosynthetic pathway, namely C8-C7 sterol isomerase (23).
In the present work, we undertook a biochemical characterization of a molecular entity that specifically binds SR 31747A using [ 3 H]SR 31747A as the specific probe and the human T leukemia cell line Ichikawa as the biological source. We purified the protein carrying the SR 31747A binding site and cloned the corresponding cDNA, which was shown to be homologous to the yeast ERG2 gene. Expression of SR-BP cDNA in yeast S. cerevisiae led to high affinity binding sites for SR 31747A indistinguishable from mammalian sites. We also present the tissue distribution and subcellular localization and discuss the structural features of this receptor and its possible functions.
To immunopurify SR-BP, active peak from the DEAE-Sepharose was loaded onto a 2-ml column of anti-SR-BP Sepharose obtained by coupling purified monoclonal anti-SR-BP antibody (see below) to CNBractivated Sepharose (ratio: 0.5 mg of proteins/ml of Sepharose). Bound proteins were eluted with 0.1 M glycine pH 3.0, 0.5 M NaCl, 0.02% digitonin. The fractions containing binding activity were neutralized to pH 7.0 with 1 M Tris-HCl and concentrated for further analysis.

Anti-SR-BP Monoclonal Antibody Obtention
Positive fractions from Q Hyper D chromatography were incubated with 2 nM [ 3 H]SR 31747A for 2 h at 20°C and chromatographed on Superdex 200 HR as described above. Fractions containing the maximum radioactivity were pooled and used as tritiated SR-BP ([ 3 H]SR-BP) for antibody screening.
Three mice were immunized by subcutaneous injection of 10 g of proteins of enriched SR-BP preparation from Superdex 200 HR chromatography in Freund's complete adjuvant. On days 30 and 60, mice were boosted with 10 g of proteins in incomplete Freund's adjuvant. Sera were collected on days 38 and 68, diluted to 100 l in PBS buffer containing 0.1% BSA and 0.1% Tween 20, and then incubated for 16 h at 4°C with 100 l of solution containing 5000 dpm of [ 3 H]SR-BP. Separation of free and bound antibodies was obtained by addition of 1 ml of 20% polyethylene glycol 4000, followed by a centrifugation at 4000 ϫ g. The pellet containing bound antibodies was dissolved in 1 ml of PBS mixed with 10 ml of scintillation liquid and counted for radioactivity.
For the production of monoclonal antibodies, 3 days before fusion, mice were boosted intravenously with 10 g of proteins. Splenocytes were collected and fused with P3 ϫ 63Ag8.653 mouse myeloma cells. Screening for anti-SR-BP antibodies was performed using [ 3 H]SR-BP. Antibodies were produced in mice and purified from ascitic fluid by affinity chromatography on protein A-Sepharose.

Electrophoresis and Immunoblot
SDS-PAGE was performed according to Laemmli (27). Proteins were transferred electrophoretically to nitrocellulose membrane. The membrane was incubated for 1 h at 20°C in 5% nonfat dry milk in 0.1% Tween 20, 150 mM NaCl, 20 mM Tris-HCl, pH 7.4, then for 1 h at 20°C with monoclonal antibody at 0.5 g/ml in the same buffer. Peroxidaseconjugated goat anti-mouse antibodies were added, and protein detection was performed with ECL kit (Amersham) as described in the manufacturer's protocol.

Partial Amino Acid Sequencing of SR-BP
Immunopurified SR-BP was subjected to SDS-PAGE. After gel staining with Coomassie Blue G-250, the 28-kDa band was cut out and subjected to trypsin hydrolysis, which generated six distinct peptides. Edman's degradation was performed using an Applied Biosystems model 477A pulse-liquid sequenator connected on line to an RP-HPLC unit (model 120A, Applied Biosystems) (28).

Molecular Cloning of Human SR-BP cDNA
Two degenerate oligonucleotides were designed based on the amino acid sequence of the longer peptide (these were derived from GARGT-NTTYTAYCCIGGIGARAC (sense) and from CATVCANGTRTTIGGIC-CCCAYTC (antisense)) ( Fig. 3). PCR was performed using 100 ng of first strand cDNA prepared from Daudi mRNA. The 80-bp PCR product was further subcloned into PGEM-T vector (Promega, Madison, WI) and sequenced. Two matched antisense primers (AS1: 5Ј-GC-CCCACTCCACAGCTGTTGCCTC-3Ј; AS2: 5Ј-ACCAGGCCCGTGTAC-TACCGTCTC-3Ј) complementary to the middle of the sequence were used to amplify the 5Ј end of cDNA with the RACE system (Life Technologies, Inc.). A PCR product of about 320 bp was subcloned into PGEM-T and sequenced. This DNA fragment was used as a probe to screen a Daudi UniZAP XR cDNA library. The longest isolated clone was fully sequenced, and sequence analyses were performed with Blast (29). Amino acid sequence alignment was produced by the clustal W program.

Expression of SR-BP cDNA in S. cerevisiae
The open reading frame of SR-BP was cloned by PCR amplification using two oligonucleotides, A and B. The oligonucleotide sequence of A (forward, BamHI, SalI) was as follows: ATCAGGATCCGTCGACAA-CATGCAATGGGCTGTTGGTAGACGGTGGGCGTGGGCC) and corresponded to the first 12 codons of SR-BP with some modifications at the DNA level to improve mRNA translation. The oligonucleotide sequence of B (reverse, BamHI, XbaI) corresponded to positions ϩ662 to ϩ684 (relative to the first base of the initiation codon). The amplified DNA fragment was cloned between the SalI and BamHI sites of the pEMR1023 polylinker (23) to yield pEMR1499, a multicopy E. coli-S. cerevisiae shuttle vector containing URA3 as the selective marker. The SR-BP open reading frame was expressed under the control of the PGK promoter.

Northern Blot Analysis
The expression of SR-BP mRNA in human tissues was examined by Northern blot analysis using membranes purchased from CLONTECH (Palo Alto, CA). Hybridization was carried out in Church buffer (30) with a 32 P-labeled 300-bp probe specific for SR-BP cDNA. Autoradiography was performed overnight with an intensifying screen.

Analysis of Subcellular Localization of SR-BP by Confocal Microscopy
Cells were fixed overnight with 1% formaldehyde at 4°C, washed once in PBS, and permeabilized for 10 min with a solution of 0.1% saponin in PBS containing 1% BSA. Biotinylated conjugated anti-SR-BP mAb, and anti-mitochondria M117 or anti-nuclear membrane AE-5 mAbs (Leinco Technologies Inc., St. Louis, MO) were simultaneously incubated with cells in the 0.1% saponin, 1% BSA solution for 30 min at room temperature. After two washes, cells were simultaneously incubated for 30 min with fluorescein isothiocyanate-conjugated goat anti-mouse IgG1 Abs (Southern Biotechnology Inc., Birmingham, AL) to stain M117 or AE-5 mAbs, and with Cy5-conjugated streptavidin (from Jackson Immunoresearch, West Grove, PA) to stain biotinylated conjugated anti-SR-BP mAb. Cell pellets were resuspended in 20 l of glycerol containing the anti-bleaching reagent DABCO at 50 mg/ml. A laser-scanning confocal microscope (LSM410, Zeiss, Oberkochen, Germany) equipped with a planapo oil (ϫ 63) immersion lens (numerical aperture ϭ 1.4) was used to analyze the three-dimensional distribution of SR-BP as described previously (31).

In Vitro Lymphocyte Proliferation
Mouse splenocytes were cultured in 96-well flat-bottomed plates in quadruplicate at 4 ϫ 10 5 cells/well with 1 g/ml staphylococcal enterotoxin B. Cells were cultured in complete medium for 4 days in the presence of various concentrations of SR 31747A or (ϩ)-pentazocine, then pulsed with 1 Ci/well [ 3 H]thymidine (Amersham, les Ulis, France), and harvested 4 h later on glass fiber papers using a Skatron harvester system (Pharmacia). Incorporated radioactivity was measured by using a Betaplate liquid scintillation spectrometer (Pharmacia). Table I. From the solubilized membrane preparation, the binding activity was eluted as a single peak with DEAE anion exchange chromatography. A straightforward enrichment of SR-BP was obtained using Mono Q gel followed by Q Hyper D gel. The apparent size of SR-BP determined by gel filtration gave a value of approximately 100 kDa (Fig. 1), whereas electrophoresis revealed several bands between 80 and 100 kDa and a band close to 30 kDa (Fig. 2, lane 5). At this stage, the enriched SRBP preparation was used to immunize animals to generate specific monoclonal antibodies against SR-BP for further purification.

Purification of Human SR-BP-Results of SR-BP purification from human T leukemia Ichikawa cells are summarized in
We developed a screening test for antibody detection, taking advantage of the fact that the binding of [ 3 H]SR 31747A to partially purified SR-BP was not dissociated upon gel filtration chromatography. As shown in Fig. 1, Q Hyper D fraction incubated with [ 3 H]SR 31747A gave a single radioactive peak associated with eluted proteins. We used this property to prepare a [ 3 H]SR 31747A/SR-BP complex (tritiated antigen) for radioimmunoassay to detect antibody specific to SR-BP. The hybridoma 30G10 producing the antibody with the highest affinity was selected; antibodies were produced in mice, purified, and linked to CNBr-Sepharose. The subsequent immunopurification step drastically improved the purity of SR-BP (Table I) since only a single band of 28 kDa could be detected by silver staining and immunobloting (Fig. 2). In the purified preparation, the SR 31747A binding activity was enriched 914-fold over the crude digitonin extract. This preparation was used for partial amino acid sequencing.
SR-BP cDNA Cloning-To isolate the human SR-BP-encoding cDNA, we used the trypsic peptides obtained from the purified human SR-BP protein. Based on amino acid sequence of the longest peptide, reverse transcriptase-PCR and then 5Ј RACE experiments led us to clone a 320-bp fragment.
This fragment was then used to screen an oligo(dT)-primed Daudi cDNA library in UniZAP phage. The cDNA and deduced amino acid sequence of the largest clone are shown in Comparisons to the GenBank data base revealed that human SR-BP displayed 93% amino acid sequence identity with the predicted amino acid sequence of a guinea pig protein called sigma 1 receptor (32) and was further found identical to the recently published human sigma 1 receptor (33). In addition, a significant homology (29.9% amino acid sequence identity) was found with the product of the gene ERG2, a fungal gene encoding a sterol isomerase (34). It is noteworthy that we have already shown that ERG2P is the molecular target of SR 31747A in yeast (23). The three proteins shared the same hydropathy plot, including a highly hydrophobic domain located at the N terminus, with a second stretch of hydrophobic amino acids located in the middle of these proteins. Multiple alignments showed that this domain is highly conserved among human SR-BP, guinea pig sigma 1 receptor, and the ERG2 gene product from yeast S. cerevisiae as well as from other fungi (Magnaportae grisea, Ustilago maydis, Neurospora crassa) (Fig. 4). SR-BP Expression-We next investigated SR-BP functions by first transfecting SR-BP cDNA in a different host cell. We used a ERG2 gene disruptant strain of S. cerevisiae (EMY47), as host to express SR-BP. Untransformed EMY47 cells did not detectably display any SR-BP binding sites. In contrast, SR-BP-expressing cells exhibited SR 31747A binding sites at a high level. Flow cytometry analysis showed a marked labeling in SR-BP-expressing cells (data not shown). These results confirmed that the cDNA sequence encoded the SR-BP protein.
Analysis of the SR-BP transcript expression by Northern blotting identified a major band of 2 kilobases (Fig. 5A). The human tissues that contained the highest amount of human SR-BP mRNA were liver along with colon, prostate, placenta, small intestine, heart, and pancreas. SR-BP mRNA is also present at a slightly lower level in spleen, lung, thymus, ovary, peripheral blood leukocytes, and brain. These observations were confirmed in the human Master Blot (Fig. 5B).
Subcellular distribution of SR-BP was examined in the human promonocytic cell line THP1 by immunological analysis using confocal microscopy. Cells were simultaneously labeled with anti-SR-BP mAb and mAbs specific for different subcellular organelles. We clearly observed SR-BP localized in association with the nuclear envelope (Fig. 6). This localization was also observed on a variety of different cell types including freshly isolated lymphocytes and monocytes. Another localization was found in the cytosol distinct from mitochondria (Fig. 6) and might be the endoplasmic reticulum.
Pharmacological Characterization of SR-BP-We studied the binding properties of human SR-BP expressed in yeast and compared it to the S. cerevisiae ERG2 gene product (Erg2p). Binding assays were performed in transformed cells of the ERG2 disruptant strain EMY47, which produce either Erg2p (control) or SR-BP. SR 31747A binding assays were also carried out on purified SR-BP. A variety of competing compounds were tested, including drugs with high affinity for sigma receptors as well as compounds known to bind yeast sterol isomerase (Table  II). Among the various molecules tested, the rank order of drug potency at SR 31747A sites was: tridemorph ϭ fenpropimorph Ͼ SR 31747A Ͼ haloperidol Ͼ (ϩ)-pentazocine Ͼ ifenprodil. Purified soluble SR-BP showed an essentially similar affinity profile for various tested compounds when compared with membraneous SR-BP. Only minor decreases in affinity were observed for haloperidol and ifenprodil, instead of major decreases for (ϩ)-pentazocine, which could be related to a difference in protein conformation.
Comparable ranking of tested ligands was observed with  4. Comparative analysis of human SR-BP amino acid sequence. The alignment was done using the clustal W alignment program. Identical amino acid residues conserved in three or more sequences are in bold letters. The proteins listed include the SR-BP, the guinea pig sigma 1 receptor (GP-sigma 1R) (32), and the sterol isomerase from the yeast (S. cerevisiae) (34), the rice blast fungus M. grisea (39), the maize smut pathogen U. maydis (39), and N. crassa (GenBank data base accession number Z22775).

FIG. 5. Northern blot analysis of human SR-BP expression.
A, commercial blots (CLONTECH) containing 2 g of poly(A) ϩ mRNAs from human tissues shown above each lane were hybridized with human SR-BP cDNA as described under "Materials and Methods." Blots were stripped and rehybridized with actin probe to assess levels of RNA between lanes (data not shown). B, a human master blot (CLONTECH) containing 50 different human tissues immobilized in separate dots was hybridized with human SR-BP cDNA as described under "Materials and Methods." The diagram on the left shows the nature and the position of poly(A) ϩ RNAs and controls.
with the above findings, [ 3 H]pentazocine, which bound to membraneous SR-BP, did not bind to Erg2p (data not shown).
Scatchard plot analysis (Fig. 7A)  We have shown previously (14) that SR 31747A is active at inhibiting lymphocyte proliferation while pentazocine is inactive. Since pentazocine does not block T lymphocyte proliferation while it prevents [ 3 H]SR 31747A binding, if SR 31747A is acting through SR-BP to block T lymphocyte mitogenesis, pentazocine should thereby prevent SR 31747A inhibition. We hence analyzed the effect of SR 31747A on cell growth in the absence or presence of pentazocine. As shown in Fig. 8A, cellular proliferation was inhibited by SR 31747A and restored in the presence of increasing pentazocine concentrations (Fig.  8B). Although it is not known whether pentazocine could completely reverse the SR 31747A effect due to the high pentazocine concentration-induced toxicity, this observation strongly suggests that the SR 31747A antiproliferative effects could be mediated by SR-BP. DISCUSSION SR 31747A is an immunomodulating agent eliciting high affinity for sigma receptors. It demonstrates immunosuppressive properties both in vitro and in vivo (14). This compound, which blocks the proliferation of lymphocytes at nanomolar concentrations, displays a spectrum of activity distinct from current immunosuppressive agents (35). However, these stud-FIG. 6. Nuclear envelope localization of SR-BP in THP1 cells by confocal microscopy. THP1 cells were simultaneously labeled with anti-SR-BP and anti-nuclear envelope mAbs or with anti-SR-BP and anti-mitochondria (anti-mt) mAbs as described under "Materials and Methods." The red (Cy5) left side corresponds to the labeling of SR-BP, the green (fluorescein isothiocyanate) central part to the labeling of subcellular organelles, and the right side represents the merged images. A single optical section representative of each cell type is shown in this image. Ab, antibody; mb, membrane.

TABLE II Inhibition of [ 3 H]SR 31747A or [ 3 H](ϩ)-pentazocine binding to soluble SR-BP, SR-BP, and Erg2p expressed in yeast by different drugs
Binding experiments were performed as described under "Materials and Methods." The values shown are the mean Ϯ S.E. of triplicate determinations. Data are from one representative experiment out of two or three. Binding parameters were obtained by nonlinear curve fitting of the inhibition data to the general dose-response equation (41 ies suffered from a lack of information on the binding protein involved to relate the observed effects to a biochemical process (36). In an attempt to decipher this, we have purified, cloned, sequenced, and expressed the SR-BP. We also studied the tissue distribution, subcellular localization, and pharmacological properties of this protein.
Molecular Characterization of SR-BP-Partial purification of the digitonin-solubilized SR-BP from cells of the human T leukemic Ichikawa cell line was achieved by three successive ion-exchange chromatographic steps followed by a gel filtration. The specific activity obtained was 510 pmol/mg of protein, representing a 160-fold increase over the specific activity of the crude extract. This preparation was used to generate a monoclonal antibody anti-SR-BP, which in turn enabled the subsequent purification of SR-BP. Immunopurified SR-BP was used for partial amino acid sequence analyses, after which the corresponding cDNA was cloned. The nucleotide sequence of the SR-BP cDNA encodes a 223-amino acid protein. Furthermore, the characterization of the heterologous expression of the SR-BP cDNA in yeast demonstrates that the cloned cDNA encodes a protein that has a high affinity [ 3 H]SR 31747Abinding domain and is specifically recognized by anti SR-BP-antibodies.
In the course of this study, we cloned the guinea pig and human sigma 1 receptor cDNA and found a 93% and 100% amino acid sequence identity, respectively, with SR-BP (32,33). We discuss below successively (i) the relationship of SR-BP and sigma 1 receptor, and (ii) the relationship of SR-BP and Erg2p.
Comparison of SR-BP and Sigma 1 Receptor-The sigma 1 receptor was recently purified and characterized from guinea pig liver by Hanner et al. (32) (see, for review, Moebius et al. (37)). The high degree of homology between the two proteins suggests that it could be the guinea pig counterpart of human SR-BP. We show here that the SR 31747A binding profile fits with the pharmacological definition of sigma 1 receptor: 1) SR-BP binds the sigma ligands, pentazocine and haloperidol with high affinity; 2) only one stereoisomer of pentazocine efficiently competed for [ 3 H]SR 31747A-labeled sites, which is a trait characteristic of the sigma 1 receptor previously characterized in rat spleen (10,11,12). Our results obtained with From saturation studies with SR 31747A and (ϩ)-pentazocine binding on SR-BP, we concluded that both ligands bound to a single population of sites. As SR 31747A was competitively displaced by (ϩ)-pentazocine and reciprocally, it is very likely that SR 31747A and (ϩ)-pentazocine bind the same site. Nevertheless, B max values (6195 and 3860 fmol/mg proteins for SR 31747A and (ϩ)-pentazocine, respectively) were slightly different and the SR 31747A affinity was 50-fold higher than that of (ϩ)-pentazocine. We observed an irreversible (or very slowly dissociable) binding of [ 3 H]SR 31747A at SR-BP.
Comparison between SR-BP and Erg2p-SR-BP shares a significant sequence identity with Erg2p (29.9%). We have shown that this latter protein is the target that mediates the anti-proliferation effects of SR 31747A in yeast (23). Results presented here as well as those of Hanner et al. (32) show that yeast Erg2p and mammalian sigma 1 receptor or SR-BP, whether of guinea-pig or human source, are related to one another not only structurally but also pharmacologically. In fact, only pentazocine clearly distinguishes both proteins, whereas SR 31747A, haloperidol, ifenprodil, and drugs of the N-substituted morpholine derivative family, like tridemorph and fenpropimorph, all bind to both proteins with high affinity. Tridemorph and fenpropimorph are known to inhibit C8-C7 sterol isomerase, as well as other enzymes of the sterol biosynthetic pathway, presumably by mimicking the high energy carbocationic reaction intermediate (38). These drugs can therefore be considered as belonging to a non-sterol class of sterol analogs that bind to the reactive sterol-binding pocket of sterol biosynthesis enzymes. Sequence alignment of the C8-C7 sterol isomerase from S. cerevisiae (34), the rice blast fungus M. grisea (39), and the maize smut pathogen U. maydis (39) show a strikingly high percentage of similarity within the central hydrophobic region. For this reason, this central domain has been suggested to contain the sterol isomerase catalytic site. Interestingly, SR-BP, which displays a hydropathy profile similar to that of fungal sterol isomerase, also presents this highly conserved central domain. This observation raises two hypotheses: first, the presumably catalytic domain of Erg2p, and the equivalent central hydrophobic domain of SR-BP might contain the sigma ligand binding site. This could be illustrated by the property of fenpropimorph to competitively inhibit C8-C7 sterol isomerase and other sterol biosynthesis enzymes (23,38). The second hypothesis is that SR-BP may exhibit a C8-C7 sterol isomerase activity. This hypothesis is supported by the high affinity of the sterol isomerase competitive inhibitors tridemorph and fenpropimorph to bind to SR-BP. However, it is not supported by expression studies in yeast, since no complementation of any of the Erg2p defect, i.e. C8-sterol accumulation and ergosterol prototrophy, was observed by expressing either the guinea-pig sigma 1 receptor (32) or the human SR-BP in an ERG2 gene disruptant (this study). No ergosterol could be detected in pEMR1499-transformed EMY47 cells (data not shown). We cannot exclude that recombinant SR-BP produced in yeast is synthesized in an active form but localized in subcellular compartments where sterol biosynthesis does not take place, as has been suggested in the case of recombinant human squalene synthase (40). Alternatively, the enzymatically active form of SR-BP might require specific post-translation modification events that do not occur in yeast cells. Finally, one should not exclude that SR-BP function(s) might not include the sterol isomerization reaction as performed by Erg2p in yeast. It is worth noting that we have already cloned a mammalian enzyme by complementation of the Erg2p defect in yeast (38). This microsomal protein displays all the biochemical characteristics expected for mammalian sterol isomerase, including sensitivity to drugs that specifically inhibit the activity of this enzyme in mammals (38). Surprisingly, this enzyme is absolutely not related phylogenetically to Erg2p or SR-BP; it does not share any striking structural similarity with any members of the Erg2p family including SR-BP. However, mammalian sterol isomerase does contain a sigma ligand binding site (32,38). This observation supports the hypothesis that the sterol-binding pocket present in sterol isomerase, whether belonging to the Erg2p family or not, indeed constitutes the sigma ligand binding site. It also suggests that other sterol metabolism enzymes might present sigma ligand binding sites as well.
Distribution and Subcellular Localization of SR-BP-The tissular distribution of human SR-BP mRNA revealed a major transcript of 2 kilobases, which was found to be ubiquitous, although its abundance varied among tissues. The transcript was most abundant in the liver, and lower in brain. This latter observation clearly contrasts with the high expression level of sigma 1 receptor mRNA in brain guinea pig reported by Hanner et al. (32). This suggests a major difference in tissue expression of sigma 1 receptor or SR-BP in guinea pig and human. Further investigations using specific anti-SR-BP antibodies are required to document this in particular to interpret the neurological activities of sigma ligands in rodent and human.
In cells expressing SR-BP, the subcellular localization using specific monoclonal antibodies showed that SR-BP is always associated with the nuclear envelope, strongly suggesting that SR-BP is an ubiquitous component of the nuclear membrane. We also observed that, when the nuclear envelope disassembles at the onset of mitosis and reassembles at the end of mitosis, the SR-BP follows the same dynamics (data not shown). Our finding of a subcellular localization of SR-BP in nuclear membrane was substantiated by the presence of a hydrophobic domain near the N terminus. The cytosolic localization, which was also observed, was different from one cell line to another and remains to be further explored. Within the nuclear membrane, the lack of recognition of SR-BP by mAbs applied to intact isolated nuclei indicates that the epitope recognized by the mAb is not exposed to the cytoplasm. However, further investigations are required to determine whether SR-BP belongs to the inner or outer nuclear membrane.
Biological Effect of SR 31747A and SR-BP-We have shown previously that SR 31747A blocks T lymphocyte proliferation. Several findings from this study support the possibility that SR 31747A effects are mediated by SR-BP.
1) We showed the presence of SR-BP in immune tissue at the mRNA level. The expression of the protein in purified T cells as well as in macrophages was also demonstrated by flow cytometry using an anti-SR-BP antibody (data not shown). These results are in agreement with recent studies from Wolfe et al. (22), who showed, using in vitro receptor autoradiography in frozen sections of rat spleen, that sigma 1 receptors were present throughout the spleen and particularly concentrated in the T cell zones.
2) We demonstrated that the effect of SR 31747A on T cell proliferation was blocked by the competitive ligand (ϩ)-pentazocine. This finding is consistent with a sigma 1 receptormediated event. Although the clear definition of agonist and antagonist at the sigma 1 receptor is difficult, these results suggest that SR 31747A and pentazocine display opposite biological effects.
We have already established that SR 31747A affects a late event in the activation process of T lymphocytes which occurs during the S/G 2 /M phases of the cell cycle (14). This suggests that the function of SR-BP could be essential for T lymphocyte mitogenesis. Furthermore, and taking into account (i) the nuclear membrane localization of SR-BP, (ii) the dramatic changes in nuclear membrane during mitosis, (iii) the homology of SR-BP with sterol isomerase, and (iv) the known effects of cholesterol on the activity of membrane proteins, one may speculate that some of the effects of SR 31747A on T cell mitogenesis could be mediated by changes in membrane lipids.
Even though our study does not definitively prove that SR-BP is the site through which SR 31747A modulates immune functions, we cannot rule out that SR 31747A could be effective through other sites not yet identified. An accurate structureactivity relationship study should help to elucidate such an important question.
In conclusion, we have identified and characterized a high affinity binding protein for SR 31747 and shown that this protein is the human sigma 1 receptor. We also demonstrated that both binding and activity of SR 31747 could be reversed by (ϩ)-pentazocine. This important property strongly suggests that the sigma 1 receptor displays biological functions and provides new tools for further mechanistic investigations.