Human Molecular Genetics, 2000, Vol. 9, No. 20 3001-3009
© 2000 Oxford University Press
Alternative splicing at the MEFV locus involved in familial Mediterranean fever regulates translocation of the marenostrin/pyrin protein to the nucleus
Institut National de la Santé et de la Recherche Médicale (INSERM) U468, Hôpital Henri-Mondor, 51 avenue du Maréchal de-Lattre-de-Tassigny, 94010 Créteil, France, 1Laboratoire ‘Complexes macromoléculaires en cellules vivantes’ and 2Laboratoire ‘Transport nucléocytoplasmique’, Département de biologie supramoléculaire et cellulaire, Institut Jacques-Monod, CNRS UMR 7592, Université Paris VI, 75005 Paris, France
Received 31 July 2000; Revised and Accepted 16 October 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Mutations in MEFV, a gene encoding a protein (marenostrin/pyrin) of unknown function, are associated with familial Mediterranean fever, a genetic condition characterized by febrile episodes of serosal inflammation. Based on its primary structure, this 781 residue protein is thought to function as a nuclear effector molecule. However, recent transient expression studies indicated a perinuclear cytoplasmic localization. Here, we describe the isolation and expression of a novel human MEFV isoform, MEFV-d2, generated by in-frame alternative splicing of exon 2. This transcript, expressed in leukocytes, predicts a 570 residue protein designated marenostrin-d2. To investigate differences in subcellular localization between the full-length protein (marenostrin-fl) and marenostrin-d2, while providing against the overexpression of transiently expressed proteins, we have generated CHO cell lines stably expressing these two isoforms fused to the green fluorescent protein. The localization pattern of marenostrin-d2 differs dramatically from that of marenostrin-fl. Marenostrin-fl is homogeneously distributed over the entire cytoplasm, whereas marenostrin-d2 concentrates into the nucleus. To map the critical domain(s) specifying these differences, deletion mutants have been generated. Deletion of the putative nuclear localization signals (NLS) does not alter the nuclear localization of marenostrin-d2 whereas, despite the lack of discernible NLS in the domain encoded by the exon 1–exon 3 splice junction, deletion of this domain indeed disrupts this localization. These data, which challenge the current domain organization model of marenostrin, strongly suggest that MEFV encodes a nuclear protein and raises the possibility that MEFV alternative splicing may control functions of wild-type and mutant marenostrin proteins by regulating their translocation to the nucleus.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
MEFV was isolated by positional cloning methods as the gene involved in familial Mediterranean fever (FMF) (1,2). This autosomal recessive condition (MIM 249100) that primarily affects populations of Mediterranean extraction—the frequency of carriers reaching 1 in 5 among individuals of Armenian or Sephardic Jewish extraction—is characterized by recurrent episodes of fever and serosal inflammation manifested by sterile peritonitis, arthritis and/or pleurisy (3). The severity of the disease is due to the risk of occurrence of renal amyloidosis, which, in the absence of daily and life-long administration of colchicine, leads to terminal renal failure (4).
The cloning of the MEFV gene has opened up new ways to manage the disease—especially by providing the first objective test of diagnostic value (5)—and to decipher both the pathophysiology of this disorder and the biological properties of the protein encoded by MEFV. The MEFV gene, which spans 15 kb of the 16p13.3 region, is composed of 10 coding exons. To date, a single MEFV transcript has been identified. This 3.7 kb mRNA is expressed in polynuclear leukocytes and predicts a protein of 781 residues named pyrin (1) or marenostrin (2). Although the function of this protein remains unknown, several lines of evidence—including those deduced from the disease phenotype of FMF patients, the tissue-specific expression of MEFV, as well as results of recent in vitro studies (6)—strongly suggest that marenostrin is implicated in the regulation of inflammatory processes. However, the mechanisms by which this protein may regulate such processes are poorly understood. Analysis of the marenostrin primary structure provided indirect evidence suggesting that the protein functions as a nuclear effector molecule (1,2). This analysis indeed revealed the presence of two overlapping potential nuclear localization signals (NLSs) (amino acids 419–422 and 420–437) (7), as well as a bZIP basic domain (amino acids 266–280) (8), a B-box-type zinc finger motif (amino acids 375–407) (9) and a B30.2 domain (amino acids 577–757) (10), encoded by exons 3 and 4, 2, 3 and 8–10, respectively. However, in contrast to this prediction, recent in vitro transient expression studies of a fusion protein between marenostrin and the green fluorescent protein (GFP) indicated that the protein encoded by MEFV is exclusively located in the cytoplasm (11,12).
As a first step in approaching the biological properties of marenostrin, we investigated MEFV gene expression in human leukocytes and identified a novel MEFV transcript. To provide against any possible intracellular protein transport disruption due to marked overexpression of recombinant proteins in transient expression assays, we established several Chinese hamster ovary (CHO) cell lines stably expressing the two different isoforms of the marenostrin protein. The documented differences in cell localization of these isoforms prompted us to construct marenostrin deletion mutants that led (i) to the identification of a protein domain specifying these subcellular compartmentalization patterns; and (ii) to the revision of the current domain organization model of marenostrin.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Identification of an exon 2-deleted MEFV splice variant expressed in human peripheral blood leukocytes
In an attempt to clone the full-length human MEFV cDNA for the purpose of protein characterization studies, MEFV-specific primers (P1 and P2) were designed to amplify the entire 2346 bp coding sequence of MEFV from human peripheral blood leukocytes. Somewhat unexpectedly, this experiment, which was performed under standard PCR conditions, did not result in the amplification of a 2.4 kb product, but in a smaller PCR product (1.7 kb) (Fig. 1A, left), whereas similar experiments performed with mouse leukocyte RNAs as templates gave rise to Mefv PCR products of expected sizes and sequences (data not shown). Cloning and sequencing of the human 1.7 kb molecular species demonstrated that it corresponded to a new MEFV transcript. The relationship between this isoform and the previously described MEFV mRNA species is depicted in Figure 1B. The nucleotide sequence of the shorter transcript was identical to that of MEFV, except that it lacked the entire exon 2, an observation that is in keeping with an alternative splice event leading to the full-length or an exon 2-deleted MEFV transcript (referred to as MEFV-fl or MEFV-d2 transcript, respectively). This hypothesis was confirmed in two ways. Firstly, the MEFV-fl transcript was indeed amplified from the same cDNA template, with the use of the same forward primer and an exon 2-specific reverse primer (Fig. 1A). Secondly, we looked for PCR conditions that would allow the co-amplification of the two MEFV transcripts in the same experiment. Assuming that the particularly high GC content (i.e. 67%) of this 633 bp exon would have prevented the amplification of MEFV-fl transcripts, we performed PCR amplifications in the presence of 10% DMSO. Under these experimental conditions, MEFV-fl, which was amplified from different leukocyte RNA samples, was easily detectable after ethidium bromide staining (Fig. 1C), whereas the MEFV-d2 isoform was visualized after hybridization with MEFV-specific probes (Fig. 1D).
|
Following similar experimental conditions, the expression of MEFV-d2 transcripts was also detected in polymorphonuclear cells, as well as in the population of peripheral blood leukocytes enriched for mononuclear cells (Fig. 2); however, the MEFV-d2:MEFV-fl transcript ratio was found to be
3-fold higher in mononuclear cells than in polymorphonuclear leukocytes. In vitro stimulation of these two leukocyte subpopulations with interferon (IFN)-
for 2 h resulted in increased levels of both MEFV-fl and MEFV-d2 transcripts (Fig. 2); however, in both leukocyte subpopulations, this treatment did not significantly alter the MEFV-d2:MEFV-fl transcript ratio.
|
Stable expression and subcellular localization of the marenostrin isoforms encoded by MEFV-fl and MEFV-d2 in CHO cells
To characterize the subcellular localization of the two marenostrin splice variants (i.e. the full-length marenostrin isoform and the isoform lacking the domain encoded by exon 2, referred to as marenostrin-fl and marenostrin-d2, respectively), we first fused in a eukaryotic expression vector the corresponding MEFV cDNAs to the 5' end of the gene encoding GFP. The resulting plasmids carrying the chimeric MEFV-fl–GFP and MEFV-d2–GFP inserts were used to generate CHO cell lines stably expressing GFP-coupled marenostrin-fl or marenostrin-d2 isoforms (Figs 3 and 4).
|
|
We found that marenostrin-fl–GFP localized exclusively to the cytoplasm, as revealed by a homogeneously distributed labelling over the entire cytoplasm (Fig. 5A and B). A similar pattern of subcellular distribution of marenostrin-fl was observed in nine other independent CHO cell lines stably expressing this protein (data not shown). The presence of two overlapping potential NLSs in this isoform leaves open the possibility that its cytoplasmic localization results from rapid nuclear export of the protein by a nuclear export sequence. To test this hypothesis, we treated CHO cells stably expressing the GFP-coupled marenostrin-fl isoform with leptomycin B, a drug that inhibits CRM1-mediated nuclear export (13,14). This treatment did not modify the subcellular distribution of marenostrin-fl which remained cytoplasmic (data not shown), thereby demonstrating that this localization was not the result of protein shuttling between nucleus and cytoplasm controlled by a CRM1-dependent mechanism.
|
In contrast, the study of CHO cell lines stably expressing the marenostrin-d2–GFP isoform revealed that this protein was localized in the cell nucleus, although a small amount was localized in the cytoplasm (Fig. 5C and D). Inside each nucleus, non-fluorescent round bodies resembling nucleoli were easily visualized. A similar pattern of subcellular distribution of marenostrin-d2 was observed in nine other independent CHO cell lines stably expressing this protein (data not shown).
To further investigate and confirm the subcellular distribution of the marenostrin–GFP fusion proteins, two of the cell lines stably expressing marenostrin-fl–GFP and marenostrin-d2–GFP were subjected to fractionation and the distribution of these two isoforms was subsequently determined by immunoblotting with the use of a monoclonal anti-GFP antibody. Western blot analyses confirmed the exclusive cytoplasmic loalization of marenostrin-fl, whereas marenostrin-d2 was readily detectable in the nuclear fraction. A small amount of the recombinant protein was detected in the cytosolic fraction (Fig. 5E), in agreement with the small amount of this isoform that was observed in the cytosol in the fluorescence studies (Fig. 5C). This experiment revealed the presence of recombinant proteins of 115 and 90 kDa in CHO cells stably expressing marenostrin-fl–GFP and marenostrin-d2–GFP, respectively. These proteins of expected sizes (considering that the molecular weight of GFP, marenostrin-fl and marenostrin-d2 is 27, 86 and 62 kDa, respectively) were not seen in the non-transfected CHO cells (Fig. 5E).
The two putative NLSs of marenostrin are not necessary for nuclear localization
The domain encoded by residues 419–437 of marenostrin-fl (HKKKIQKQLEHLKKLRKSG) matches two overlapping NLSs (Fig. 3). Although these putative NLSs are present in both marenostrin-fl and marenostrin-d2, the possibility cannot be excluded that the nuclear localization of marenostrin-d2 depends on the integrity of these sequences and that marenostrin-fl is localized in the cytoplasm by virtue of either an inhibitor protein whose binding masks these NLSs or an exon 2-specific protein folding leading to a similar masking of these sequences.
To test whether the nuclear localization of marenostrin-d2 was dependent on these putative NLSs, we deleted these sequences and transiently expressed in CHO cells the mutant protein fused to the same GFP tag, designated marenostrin-d2-dNLS–GFP (Fig. 3). Somewhat unexpectedly, the mutant protein was still detected in the nucleus (Fig. 6A and B), implying that the NLS motifs of marenostrin-d2 are not required for nuclear localization of the splice variant. This result also suggests that the nuclear localization of this isoform is controlled by marenostrin-d2-specific structural determinants that are absent from marenostrin-fl.
|
Nuclear localization of marenostrin-d2 requires the integrity of the domain encoded by the MEFV exon 1–exon 3 junction
Having shown that the deletion of the putative NLSs of marenostrin-d2 does not affect its nuclear localization, we investigated the possibility that the alternative splice event leading to this isoform results in the generation of a nuclear import sequence encoded by the exon 1–exon 3 junction of MEFV-d2. Sequence examination of this splice junction did not reveal any domain resembling a classical NLS (7). Therefore, to test further the possible existence of a nuclear import sequence at this junction, a deletion mutant that lacked 15 acids of the domain encoded by the exon 1–exon 3 junction was constructed (Fig. 3). Study of CHO cells expressing this mutant fused to GFP (designated marenostrin-d2-dj–GFP) demonstrated that this sequence was essential for nuclear localization, as the fusion protein was detected in the cytoplasm (Fig. 6C and D). Overall, these results suggest that the domain encoded by the exon 1–exon 3 junction of MEFV-d2 acts to target marenostrin-d2 to the nucleus.
To determine whether this 15 amino acid junction domain (referred to as marenostrin-j) contains a nuclear import sequence that can functionally substitute for a classical NLS, it was fused to GFP (Fig. 3) and the cellular distribution of the marenostrin-j–GFP fusion protein was determined by direct fluorescence in CHO cells. This distribution was found to be very similar to that of native GFP, with a diffuse expression pattern of both proteins (data not shown). Given the well known potency of short NLS sequences fused to GFP to target the GFP into the nucleus (15,16), this result indicates that the 15 amino acid domain encoded by the exon 1–exon 3 junction of MEFV-d2 is unable to function per se as a nuclear import signal.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
MEFV is a gene involved in the pathogenesis of FMF, a recessively inherited disorder characterized by recurrent episodes of fever and serosal inflammation. To date, nothing is known about the function of its product, but based on its primary sequence, there is widespread speculation that marenostrin may be a nuclear protein. This assumption has recently been challenged by transient expression studies that showed a cytoplasmic localization of this protein (11,12). The data presented in this study revise this concept to include the existence of an alternative splice event leading to a novel marenostrin isoform that is targeted to the nucleus.
We have isolated MEFV-d2, the first splice variant of MEFV, generated by an in-frame splice removal of exon 2, a large 633 bp sequence that accounts for 27% of the coding sequence. Like MEFV-fl, MEFV-d2 is expressed in human leukocytes. More precisely, we observed MEFV-d2 expression in polymorphonuclear cells, as well as in the subpopulation of peripheral blood leukocytes enriched for mononuclear cells, an expression pattern that is similar to that of MEFV-fl (6, and this study). We also showed that IFN- upregulated both MEFV-fl levels in polymorphonuclear and mononuclear subpopulations, in keeping with a previous report (6), and MEFV-d2 levels in the same cells. Taken together, these observations support the hypothesis that the two MEFV isoforms may act as IFN -mediated regulators of the inflammatory response.
Given the absence of available anti-marenostrin antibodies, we elected to analyse the subcellular localization of these two marenostrin isoforms with expression vectors that produce GFP-tagged proteins. To avoid possible artefacts in subcellular localization due to the overexpression of recombinant proteins in transient expression assays, we constructed CHO cell lines stably expressing the marenostrin-fl–GFP and marenostrin-d2–GFP chimeric proteins. Consistent with a recent report, we found that marenostrin-fl localized exclusively to the cytoplasm. However, in the previous studies in which marenostrin-fl was transiently expressed at high levels either in COS-7 cells (12; S. Papin and S. Amselem, unpublished data) or in COS-1 cells (11), labelling was concentrated in a perinuclear localization, which is different from the diffuse cytoplasmic staining observed in our CHO cell lines stably expressing the same protein. We believe that these different labelling patterns reflect the different expression levels of the recombinant protein, as also attested to by a similar intense perinuclear labelling observed in CHO cells transiently expressing marenostrin-fl (data not shown).
The subcellular localization of marenostrin-d2 differs dramatically from that of the full-length protein, the spliced variant being mainly located in the nucleus. This latter result, which was confirmed by western blot analysis of cytoplasmic and nuclear extracts of a CHO cell line stably expressing marenostrin-d2, is in agreement with the subcellular distribution pattern of a nuclear effector molecule. Again, when expressed at high levels in CHO cells by means of transient expression assays, marenostrin-d2–GFP also accumulated in the perinuclear region, whereas, in the same transient transfection experiments, cells that expressed low levels of the protein showed a nuclear labelling similar to that observed in stable transfectants (data not shown); this observation, therefore, further supports the hypothesis that exaggerated overexpression of recombinant proteins in transiently transfected cells may result in disruption of physiological protein transport pathways (17). At this point, it is also important to note that, given the molecular weight of the marenostrin-d2–GFP chimeric protein (90 kDa), the hypothesis of a diffusion of the protein into the nucleus can be ruled out (18).
In an attempt to decipher the mechanism(s) by which the two isoforms are differently compartmentalized and to identify the structural determinants involved in this phenomenon, several experiments were subsequently performed. The following conclusions can be drawn. Firstly, as the two overlapping NLS motifs are present on both marenostrin-fl and marenostrin-d2, we tested and excluded the hypothesis that the cytoplasmic localization of marenostrin-fl results from a rapid nuclear export which may be ensured by the CRM1-mediated nuclear export system; indeed, the use of leptomycin B, a specific inhibitor of the CRM1 receptor (13,14,19–22), did not affect the cytoplasmic localization of marenostrin-fl. Secondly, surprisingly, these two potential NLS motifs do not appear to be critical for nuclear localization of marenostrin-d2, as a mutant protein carrying a deletion encompassing these motifs did not localize to the cytoplasm, but still concentrated into the nucleus. This unusual observation is reminiscent of the data obtained with Cdc18, a human protein that also contains a canonical NLS motif that turned out to be unnecessary for its nuclear localization (23). Thirdly, we also tested and confirmed the hypothesis that the domain encoded by the exon 1–exon 3 junction plays a critical role in targeting marenostrin-d2 to the nucleus; a deletion of 15 residues spanning this junction is indeed sufficient to disrupt the nuclear localization of this isoform, a result which also argues against the nuclear localization of marenostrin-d2 by simple diffusion.
Overall, these data raise the larger question of the mechanisms by which marenostrin-d2 mainly localizes to the nucleus, in contrast to the cytoplasmic localization of marenostrin-fl. One possibility is that a nuclear localization signal is created by the alternative splice event which leads to the juxtaposition of the domains encoded by exons 1 and 3. Such a mechanism has already been documented in a few other proteins, such as the CaM kinase (24), the protein 4.1 (25) and NF2 (26). However, sequence examination of the domain encoded by the MEFV exon 1–exon 3 splice junction (EELHRAAIQGRPPDT) showed that there was neither an SV40 T antigen NLS consensus (27) nor a nucleoplasmin bipartite NLS consensus (28), two motifs that are recognized by a soluble heterodimeric carrier that consists of two proteins, importin-
and -ß (7). In addition, this analysis did not reveal any extensive homologies with the M9 motif, a sequence recognized by transportin (29), thereby raising the possibility that nuclear import of marenostrin-d2 occurs via a still unidentified receptor-mediated pathway or via its association with another protein that contains an NLS. In this regard, it is of note that the 15 amino acid domain encoded by the exon 1–exon 3 junction was revealed to be necessary but not sufficient for targeting marenostrin-d2 to the nucleus, since we found it unable to functionally substitute for an NLS. This observation therefore leaves open the possibility that this domain may be required for the activity of an NLS-like motif located elsewhere in marenostrin-d2. The identification of proteins that interact with the exon 2-deleted but not the full-length marenostrin protein should provide additional insight into how nuclear import of marenostrin-d2 is accomplished.
The identification of marenostrin-d2 leads to a revision of the current domain organization model of marenostrin. Strikingly, nuclear targeting of this novel marenostrin isoform does not depend on the putative NLSs that had been identified in this protein, but on the domain created by the splice junction which does not contain any sequence resembling such signals. At first glance, it is also somewhat surprising that the domain encoded by the MEFV exon 2 sequence, which is absent from the nuclear marenostrin isoform, contains a 15 amino acid sequence (residues 266–280) which predicts a bZIP transcription factor basic domain (8), as previously identified (1) by a PROSITE search (30). However, several lines of evidence rather indicate that this sequence does not actually belong to a bZIP domain: search for protein motifs using other research tools like Pfam (31), SMART (32) or InterPro (http://www.ebi.ac.uk/interpro/ ) did not identify any similarity of this marenostrin sequence with a bZIP basic domain; in addition, this putative bZIP basic domain (KTAANLDSATEPRAR) contains only three basic residues. Furthermore, sequence examination of the marenostrin region immediately following these 15 residues did not reveal any leucine-rich domain, which is part of the characteristic bipartite structural motif of bZIP domains (8). Importantly, like marenostrin-fl, marenostrin-d2 indeed contains a B30.2 domain (residues 577–757) which is present in several nuclear proteins (10,33,34) and which represents the main target for FMF-associated mutations (5,35). In addition, marenostrin-d2 shares with marenostrin-fl a B-box zinc finger, a motif also found in a family of nuclear proteins (9). Taken together, it is therefore attractive to speculate that, depending on the subcellular localization of the marenostrin isoforms, the FMF-associated MEFV mutations might lead to different biological consequences.
The Mefv gene is also present in the genome of mice and rats; however, RT–PCR analyses of mouse leukocyte transcripts did not reveal the presence of an exon 2-deleted Mefv isoform. This observation is actually in keeping with the evolutionary divergence already documented between the human and murine species at this particular locus (36). It is indeed striking that, although the overall genomic organization of human, mouse and rat genes is quite similar, a substantial divergence does exist among these species in several important gene domains (36). Firstly, the sequence homology between the human and murine MEFV genes is particularly low in exon 2 which consists of 633 bp in humans, compared with 846 and 852 bp in mice and rats, respectively. At the protein level, the domain encoded by this exon is only 42% identical between humans and mice and 26% identical between humans and rats. Eight of the 15 amino acid residues encoded by the human MEFV exon 1–exon 3 junction are not conserved in the mouse predicted protein. In addition, Mefv also diverges from MEFV in the exon–intron structure of the 3' part of the gene, the predicted murine proteins lacking the B30.2 domain (36). Although the function of this motif is still unknown, it is noteworthy that, as mentioned above, the large majority of the human mutations associated with the FMF phenotype are clustered in this domain, these mutations accounting for up to 88% of the mutated MEFV alleles in different populations (5,35). Although the possibility cannot be excluded that the Mefv-d2 isoform is expressed at very low levels in mouse peripheral blood or in tissues that have not yet been explored, all these structural differences strongly support the hypothesis that particular functions of the MEFV gene may have been altered through evolution (36). Such differences between human and murine genes have already been widely documented (37–40) and include species-specific alternative splice events of functional significance (41,42). The study of murine marenostrin proteins should shed light on the functional consequences of these species-specific structural differences.
Overall, the isolation of a novel marenostrin isoform which is targeted to the nucleus opens up a new field of research to decipher the role that marenostrin plays in this cell compartment. In this regard, it is tempting to hypothesize that this alternative splice event may be affected under physiological or pathological conditions, thereby regulating amounts of the encoded isoforms in the cytoplasm and the nucleus.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
RT–PCR analyses
Poly(A)+ RNA was prepared from human peripheral blood samples, using the mRNA isolation kit for blood/bone marrow (Roche, Meylan, France). Total RNA was also prepared from peripheral blood polymorphonuclear leukocytes and mononuclear cells isolated from whole blood by standard density centrifugation with Ficoll-Paque (Amersham Pharmacia Biotech, Orsay, France), after erythrocytes were lysed in hypotonic solution. RT–PCR analyses were performed using the Titan One Tube RT–PCR system (Roche), according to the manufacturer’s instructions. Two sets of human MEFV-specific oligonucleotides were used. The first one, consisting of P1 (5'-CGCGTCGACGCGCTCGAGCCTCTCCTGCTCAGCACC-3') located in exon 1 and P2 (5'-CCGGAATTCCGGTGGGCATTCAGTCAGGCCCCTGACC-3') located in exon 10, amplifies two MEFV cDNA isoforms of 2.4 and 1.7 kb, designated MEFV-fl (full-length) and MEFV-d2 (exon 2-deleted), respectively. The second set of primers, consisting of P1 and P3 (5'-TTCCTCGACAGCCCCCTCCCGGCCT-3') located in exon 2, was used to amplify the full-length isoform only. P1 and P2 were designed to contain SalI and EcoRI restriction sites, respectively. Cycle parameters were as follows: 1 min of denaturation (94°C), 30 s of annealing (60°C) and 2 min of elongation (72°C) for 35 cycles, with a 7 min final elongation (72°C). To co-amplify the two marenostrin isoforms, the PCR amplifications were performed in the presence of 10% DMSO. The resulting products were blotted onto a Hybond nylon membrane and hybridized with two labelled oligonucleotides (Alk Phos; Amersham Pharmacia Biotech) derived from sequences of exons 4 and 5. In addition, two sets of mouse Mefv-specific primers were used to amplify Mefv transcripts expressed in mouse peripheral blood leukocytes. The first, consisting of P1m (5'-CTGGCCCGCACCATGGCCAAGACC-3') located in exon 1 and P2m (5'-TTGTCTCCCTGTAGCCTGTGCTCC-3') located in exon 4, yields a PCR product of 1.6 kb, whereas the second, consisting of P1m and P3m (5'-GGAACTTTGCACACAGGTATCACC-3') located in exon 3, yields a PCR product of 1.2 kb.
The expression of MEFV-fl and MEFV-d2 was also studied in peripheral blood polymorphonuclear leukocytes and mononuclear cells before and after a treatment for 2 h in vitro with interferon (IFN)- (1000 U/ml). The two MEFV isoforms were co-amplified in the presence of 10% DMSO and the resulting products were subjected to Southern blotting as described above. The MEFV-d2:MEFV-fl transcript ratio was assessed by use of a Storm 840 Phosphorimager (Molecular Dynamics, Sunnyvale, CA). In each RNA sample, ß-actin transcripts were also amplified using the following primers: 5'-CCAAGGCCAACCGCGAGAAGATGAC-3' and 5'-AGGGTACATGGTGGTGCCGCCAGAC-3'.
Plasmid construction and site-directed mutagenesis
The cDNAs coding for MEFV-fl and MEFV-d2 were digested with SalI and EcoRI and inserted into the corresponding sites of the simian virus (SV40) promoter-based expression vector pECE (43). The resulting expression vectors were designated pMEFV-fl and pMEFV-d2, respectively. The MEFV stop codon was deleted by site-directed mutagenesis on both vectors, using the QuickChange site-directed mutagenesis system (Stratagene, Amsterdam, The Netherlands) and two complementary oligonucleotides (P4, 5'-GGGCCTGACTCAATCCCCACCGGAA-3', and P5, 5'-TTCCGGTGGGGATTGAGTCAGGCCC-3'). The cDNA encoding the GFP was amplified by PCR from the pEGFP-N3 vector (Clontech, Palo Alto, CA), using GFP-specific primers: the sense primer (P6, 5'CCGGAATTCTGGGCGGGGGCGTGAGCAAGGGCGAGGAG-3') was designed to delete the initiation methionine of the GFP cDNA and to introduce an EcoRI site followed by three glycine residues to increase the flexibility of the chimeric protein structure (44); the antisense primer (P7, 5'-CCGGAATTCTTAGCCCCCGCCCTTGTACAGCTCGTCCAT-3') was designed to create an EcoRI site at the 3' end of the GFP cDNA. The resulting PCR product (740 bp) was cloned into the EcoRI site of pMEFV-fl and pMEFV-d2, thereby generating two constructs named pMEFV-fl–GFP and pMEFV-d2–GFP, respectively. The two overlapping NLSs (residues 419–437) were deleted by means of site-directed mutagenesis, using primers P8 (5'-GAGGTCGCCCTGGAAGAGGAGCAGCGATCC-3') and P9 (5'-GGATCGCTGCTCCTCTTCCAGGGCGACCTC-3') and the pMEFV-d2–GFP vector as a template, generating a plasmid named pMEFV-d2-dNLS–GFP. Similarly, residues 84–92 and 304–309 encoded by the exon 1–exon 3 junction were deleted by means of site-directed mutagenesis performed on pMEFV-d2–GFP with primers P10 (5'-CGCCTGCTGGCCGCTGCGAGTCCC-3') and P11 (5'-GGGACTCGCAGCGGCCAGCAGGCG-3'), generating a plasmid named pMEFV-d2-dj–GFP. Finally, the nucleotide sequence of the exon 1–exon 3 splice junction of MEFV encoding residues 84–92 and 304–309 of marenostrin-fl were fused to the GFP cDNA sequence by means of site-directed mutagenesis performed on the expression vector pEGFP-N3 with primers P12 (5'-ATCGCCACCATGGAGGAGCTCCACAGGGCAGCCATTCAGGG-AAGGCCACCAGACACGGTGAGCAAGGGC-3') and P13 (5'-GCCCTTGCTCACCGTGTCTGGTGGCCTTCCCTGAATGGCTGCCCTGTGGAGCTCCTCCATGGTGGCGAT-3'), generating a plasmid named pMEFV-j–GFP.
Construction of CHO cell lines stably expressing MEFV-fl–GFP, MEFV-d2–GFP and GFP and transient expression of MEFV-d2-dNLS–GFP, MEFV-d2-dj–GFP, MEFV-j–GFP and GFP in CHO cells
CHO cells were obtained from the American Type Culture Collection (Manassas, VA) and grown in Iscove medium (Life Technologies, Cergy-Pontoise, France) containing 10% fetal calf serum (FCS) (Life Technologies). To generate cell lines expressing the different constructs, CHO cells were cotransfected with a neomycin resistance plasmid and pMEFV-fl–GFP, pMEFV-d2–GFP or pEGFP-N3. Transfections were performed at 60% confluence by the Lipofectamine-Plus method (Life Technologies) in OptiMEM medium and selection in 500 µg/ml G418 was carried out for 3 weeks before subcloning survivors. Stably transfected and non-transfected CHO cells were maintained at 37°C in Iscove supplemented with 10% FCS with or without G418 and processed for RNA, protein and fluorescence analyses. In addition, transient transfection of pMEFV-d2-dNLS–GFP, pMEFV-d2-dj–GFP, pMEFV-j–GFP and pEGFP-N3 was performed in CHO cells using Lipofectamine-Plus (Life Technologies) according to the manufacturer’s standard protocol.
RNA isolation and northern blot analysis
Total cellular RNA was isolated in parallel from non-transfected and marenostrin- or GFP-expressing CHO cells and transferred by vacuum blotting onto a Hybond nylon membrane (Amersham Pharmacia Biotech). Filters were hybridized with the human MEFV-d2 cDNA probe or with GFP cDNA probes labelled with [32P]dCTP by random oligonucleotide priming (Roche). The relative abundance of RNA in each lane was assessed by comparing ethidium bromide staining intensity of the ribosomal bands. For further confirmation, the blots were hybridized with a probe for the endogenous housekeeping ß-actin gene.
Fluorescence analysis and microscopy
Transfectants expressing MEFV-fl–GFP, MEFV-d2–GFP, MEFV-d2-dNLS–GFP, MEFV-d2-dj–GFP, MEFV-j or GFP were plated onto glass coverslips and fixed in 4% (v/v) formaldehyde, phosphate-buffered saline (PBS). They were subsequently permeabilized in 0.1% Triton, PBS and DNA was stained with 4',6'-diamidino-2-phenylindole. Cells were then examined with a Leica DMR fluorescence microscope. In addition, CHO cells stably expressing MEFV-fl–GFP were treated with various concentrations of leptomycin B (i.e. 20, 100 and 200 nM) for at least 6 h. The fusion protein was then directly visualized in live cells with an inverted Olympus IX50 fluorescence microscope.
Preparation of whole cell, cytoplasmic and nuclear extracts and western blot analysis
To prepare whole cell extracts, 107 cells were lysed in 500 µl of lysis buffer (20 mM HEPES pH 7, 150 mM NaCl, 1 mM EDTA, 1% Izepal) containing protease inhibitors (Protease inhibitor cocktail tablets; Roche). Nuclear and cytoplasmic extracts were prepared as previously described (45). Protein extracts (10–50 µg) were analysed on 10% SDS–polyacrylamide gel and transferred onto a nitrocellulose membrane which was blocked in 5% non-fat dry milk and incubated with a monoclonal anti-GFP antibody (Roche). The secondary antibody was coupled to peroxidase allowing revelation with the ECL kit (Amersham Pharmacia Biotech).
|
ACKNOWLEDGEMENTS |
|---|
We wish to thank J. Coppey and N. Raich for helpful discussions. This work was supported by grants from the Association Française contre les Myopathies (AFM). S.P. is the recipient of a fellowship from the Ministère de la Recherche et de la Technologie.
|
FOOTNOTES |
|---|
+ To whom correspondence should be addressed. Tel: +33 1 49 81 28 73; Fax: +33 1 49 81 28 42; Email address: amselem@im3.inserm.fr
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 The International FMF Consortium (1997) Ancient missense mutations in a new member of the RoRet gene family are likely to cause familial Mediterranean fever. Cell, 90, 797–807.[ISI][Medline]
2 The French FMF Consortium (1997) A candidate gene for familial Mediterranean fever. Nature Genet., 17, 25–31.[ISI][Medline]
3 Sohar, E., Gafni, J., Pras, M. and Heller, H. (1967) Familial Mediterranean fever. A survey of 470 cases and review of the literature. Am. J. Med., 43, 227–253.[ISI][Medline]
4 Zemer, D., Pras, M., Sohar, E., Modan, M., Cabili, S. and Gafni, J. (1986) Colchicine in the prevention and treatment of the amyloidosis of familial Mediterranean fever. N. Engl. J. Med., 314, 1001–1005.[Abstract]
5 Cazeneuve, C., Sarkisian, T., Pêcheux, C., Dervichian, M., Nédelec, B., Reinert, P., Ayvazyan, A., Kouyoumdjian, J.-C., Ajrapetyan, H., Delpech, M. et al. (1999) MEFV gene analysis in Armenian patients with familial Mediterranean fever: diagnostic value and unfavorable renal prognosis of the M694V homozygous genotype—genetic and therapeutic implications. Am. J. Hum. Genet., 65, 88–97.[ISI][Medline]
6 Centola, M., Wood, G., Frucht, D.M., Galon, J., Aringer, M., Farrell, C., Kingma, D.W., Horwitz, M.E., Mansfield, E., Holland, S.M. et al. (2000) The gene for familial Mediterranean fever, MEFV, is expressed in early leukocyte development and is regulated in response to inflammatory mediators. Blood, 95, 3223–3231.
7 Nigg, E.A. (1997) Nucleocytoplasmic transport: signals, mechanisms and regulation. Nature, 386, 779–787.[Medline]
8 Vinson, C.R., Sigler, P.B. and McKnight, S.L. (1989) Scissors-grip model for DNA recognition by a family of leucine zipper proteins. Science, 246, 911–916.
9 Reddy, B.A., Etkin, L.D. and Freemont, P.S. (1992) A novel zinc finger coiled-coil domain in a family of nuclear proteins. Trends Biochem. Sci., 17, 344–345.[ISI][Medline]
10 Henry, J., Ribouchon, M.-T., Offer, C. and Pontarotti, P. (1997) B30.2-like domain proteins : a growing family. Biochem. Biophys. Res. Commun., 235, 162–165.[ISI][Medline]
11 Tidow, N., Chen, X., Muller, C., Kawano, S., Gombart, A.F., Fischel-Ghodsian, N. and Koeffler, H.P. (2000) Hematopoietic-specific expression of MEFV, the gene mutated in familial Mediterranean fever and subcellular localization of its corresponding protein, pyrin. Blood, 95, 1451–1455.
12 Chen, X., Bykhovskaya, Y., Tidow, N., Hamon, M., Bercovitz, Z., Spirina, O. and Fischel-Ghodsian, N. (2000) The familial mediterranean fever protein interacts and colocalizes with a putative Golgi transporter. Proc. Soc. Exp. Biol. Med., 224, 32–40.
13 Ossareh-Nazari, B., Bachelerie, F. and Dargemont, C. (1997) Evidence for a role of CRM1 in signal-mediated nuclear protein export. Science, 278, 141–144.
14 Fornerod, M., Ohno, M., Yoshida, M. and Mattaj, I.W. (1997) CRM1 is an export receptor for leucine-rich nuclear export signals. Cell, 90, 1051–1060.[ISI][Medline]
15 Shields, J.M. and Yang, V.W. (1997) Two potent nuclear localization signals in the gut-enriched Kruppel-like factor define a subfamily of closely related Kruppel proteins. J. Biol. Chem., 272, 18504–18507.
16 Wu, J., Zhou, L., Tonissen, K., Tee, R. and Artzt, K. (1999) The quaking I-5 protein (QKI-5) has a novel nuclear localization signal and shuttles between the nucleus and the cytoplasm. J. Biol. Chem., 274, 29202–29210.
17 Kremmidiotis, G., Lensink, I.L., Bilton, R.L., Woollatt, E., Chataway, T.K., Sutherland, G.R. and Callen, D.F. (1999) The Batten disease gene product (CLN3p) is a Golgi integral membrane protein. Hum. Mol. Genet., 8, 523–531.
18 Ohno, M., Fornerod, M. and Mattaj, I.W. (1998) Nucleocytoplasmic transport: the last 200 nanometers. Cell, 92, 327–336.[ISI][Medline]
19 Stade, K., Ford, C.S., Guthrie, C. and Weis, K. (1997) Exportin 1 (Crm1p) is an essential nuclear export factor. Cell, 90, 1041–1050.[ISI][Medline]
20 Fukuda, M., Asano, S., Nakamura, T., Adachi, M., Yoshida, M., Yanagida, M. and Nishida, E. (1997) CRM1 is responsible for intracellular transport mediated by the nuclear export signal. Nature, 390, 308–311.[Medline]
21 Bardoni, B., Sittler, A., Shen, Y. and Mandel, J.L. (1997) Analysis of domains affecting intracellular localization of the FMRP protein. Neurobiol. Dis., 4, 329–336.[ISI][Medline]
22 Henderson, B.R. and Eleftheriou, A. (2000) A comparison of the activity, sequence specificity and CRM1-dependence of different nuclear export signals. Exp. Cell Res., 256, 213–224.[ISI][Medline]
23 Saha, P., Chen, J., Thome, K.C., Lawlis, S.J., Hou, Z.H., Hendricks, M., Parvin, J.D. and Dutta, A. (1998) Human CDC6/Cdc18 associates with Orc1 and cyclin-cdk and is selectively eliminated from the nucleus at the onset of S phase. Mol. Cell. Biol., 18, 2758–2767.
24 Srinivasan, M., Edman, C.F. and Schulman, H. (1994) Alternative splicing introduces a nuclear localization signal that targets multifunctional CaM kinase to the nucleus. J. Cell Biol., 126, 839–852.
25 Luque, C.M., Lallena, M.J., Alonso, M.A. and Correas, I. (1998) An alternative domain determines nuclear localization in multifunctional protein 4.1. J. Biol. Chem., 273, 11643–11649.
26 Schmucker, B., Tang, Y. and Kressel, M. (1999) Novel alternatively spliced isoforms of the neurofibromatosis type 2 tumor suppressor are targeted to the nucleus and cytoplasmic granules. Hum. Mol. Genet., 8, 1561–1570.
27 Stochaj, U., Bossie, M.A., van Zee, K., Whalen, A.M. and Silver, P.A. (1993) Analysis of conserved binding proteins for nuclear localization sequences. J. Cell Sci., 104, 89–95.[Abstract]
28 Robbins, J., Dilworth, S.M., Laskey, R.A. and Dingwall, C. (1991) Two interdependent basic domains in nucleoplasmin nuclear targeting sequence: identification of a class of bipartite nuclear targeting sequence. Cell, 64, 615–623.[ISI][Medline]
29 Pollard, V.W., Michael, W.M., Nakielny, S., Siomi, M.C., Wang, F. and Dreyfuss, G. (1996) A novel receptor-mediated nuclear protein import pathway. Cell, 86, 985–994.[ISI][Medline]
30 Bairoch, A., Bucher, P. and Hofmann, K. (1997) The PROSITE database, its status in 1997. Nucleic Acids Res., 25, 217–221.
31 Bateman, A., Birney, E., Durbin, R., Eddy, S.R., Howe, K.L. and Sonnhammer, E.L. (2000) The Pfam protein families database. Nucleic Acids Res., 28, 263–266.
32 Schultz, J., Copley, R.R., Doerks, T., Ponting, C.P. and Bork, P. (2000) SMART: a web-based tool for the study of genetically mobile domains. Nucleic Acids Res., 28, 231–234.
33 Vernet, C., Boretto, J., Mattei, M.G., Takahashi, M., Jack, L.J., Mather, I.H., Rouquier, S. and Pontarotti, P. (1993) Evolutionary study of multigenic families mapping close to the human MHC class I region. J. Mol. Evol., 37, 600–612.[ISI][Medline]
34 Bellini, M., Lacroix, J.C. and Gall, J.G. (1993) A putative zinc-binding protein on lampbrush chromosome loops. EMBO J., 12, 107–114.[ISI][Medline]
35 Yalcinkaya, F., Cakar, N., Misirlioglu, M., Tumer, N., Akar, N., Tekin, M., Tastan, H., Kocak, H., Ozkaya, N. and Elhan, A.H. (2000) Genotype–phenotype correlation in a large group of Turkish patients with familial mediterranean fever: evidence for mutation-independent amyloidosis. Rheumatology (Oxford), 39, 67–72.
36 Chae, J.J., Centola, M., Aksentijevich, I., Dutra, A., Tran, M., Wood, G., Nagaraju, K., Kingma, D.W., Liu, P.P. and Kastner, D.L. (2000) Isolation, genomic organization and expression analysis of the mouse and rat homologs of MEFV, the gene for familial mediterranean fever. Mamm. Genome, 11, 428–435.[ISI][Medline]
37 Park, C., Lecomte, M.J. and Schindler, C. (1999) Murine Stat2 is uncharacteristically divergent. Nucleic Acids Res., 27, 4191–4199.
38 Pantel, J., Machinis, K., Sobrier, M.L., Duquesnoy, P., Goossens, M. and Amselem, S. (2000) Species-specific alternative splice mimicry at the growth hormone receptor locus revealed by the lineage of retroelements during primate evolution. J. Biol. Chem., 275, 18664–18669.
39 Hsu, M.H., Ember, J.A., Wang, M., Prossnitz, E.R., Hugli, T.E. and Ye, R.D. (1997) Cloning and functional characterization of the mouse C3a anaphylatoxin receptor gene. Immunogenetics, 47, 64–72.[ISI][Medline]
40 Gemel, J., Gorry, M., Ehrlich, G.D. and MacArthur, C.A. (1996) Structure and sequence of human FGF8. Genomics, 35, 253–257.[ISI][Medline]
41 Dastot, F., Sobrier, M.L., Duquesnoy, P., Duriez, B., Goossens, M. and Amselem, S. (1996) Alternatively spliced forms in the cytoplasmic domain of the human growth hormone (GH) receptor regulate its ability to generate a soluble GH-binding protein. Proc. Natl Acad. Sci. USA, 93, 10723–10728.
42 Dastot, F., Duquesnoy, P., Sobrier, M.L., Goossens, M. and Amselem, S. (1998) Evolutionary divergence of the truncated growth hormone receptor isoform in its ability to generate a soluble growth hormone binding protein. Mol. Cell. Endocrinol., 137, 79–84. [Erratum (1998) Mol. Cell. Endocrinol., 141, 195.][ISI][Medline]
43 Ellis, L., Clauser, E., Morgan, D.O., Edery, M., Roth, R.A. and Rutter, W.J. (1986) Replacement of insulin receptor tyrosine residues 1162 and 1163 compromises insulin-stimulated kinase activity and uptake of 2-deoxyglucose. Cell, 45, 721–732.[ISI][Medline]
44 Moriyoshi, K., Richards, L.J., Akazawa, C., O’Leary, D.D. and Nakanishi, S. (1996) Labeling neural cells using adenoviral gene transfer of membrane- targeted GFP. Neuron, 16, 255–260.[ISI][Medline]
45 Krishnamoorthy, R.R., Crawford, M.J., Chaturvedi, M.M., Jain, S.K., Aggarwal, B.B., Al-Ubaidi, M.R. and Agarwal, N. (1999) Photo-oxidative stress down-modulates the activity of NF-
ß via involvement of caspase-1, leading to apoptosis of photoreceptor cells. J. Biol. Chem., 274, 3734–3743.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F. Bedoya, L. L. Sandler, and J. A. Harton Pyrin-Only Protein 2 Modulates NF-{kappa}B and Disrupts ASC:CLR Interactions J. Immunol., March 15, 2007; 178(6): 3837 - 3845. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Simon and J. W. M. van der Meer Pathogenesis of familial periodic fever syndromes or hereditary autoinflammatory syndromes Am J Physiol Regulatory Integrative Comp Physiol, January 1, 2007; 292(1): R86 - R98. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C Cazeneuve, S Papin, I Jeru, P Duquesnoy, and S Amselem Subcellular localisation of marenostrin/pyrin isoforms carrying the most common mutations involved in familial Mediterranean fever in the presence or absence of its binding partner ASC J. Med. Genet., March 1, 2004; 41(3): e24 - 24. [Full Text] [PDF]
|
|
|
|
 |
|
 J. Milledge, P. J. Shaw, A. Mansour, S. Williamson, B. Bennetts, T. Roscioli, J. Curtin, and J. Christodoulou Allogeneic bone marrow transplantation: cure for familial Mediterranean fever Blood, July 18, 2002; 100(3): 774 - 777. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Richards, P. Schaner, A. Diaz, J. Stuckey, E. Shelden, A. Wadhwa, and D. L. Gumucio Interaction between Pyrin and the Apoptotic Speck Protein (ASC) Modulates ASC-induced Apoptosis J. Biol. Chem., October 12, 2001; 276(42): 39320 - 39329. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Mansfield, J. J. Chae, H. D. Komarow, T. M. Brotz, D. M. Frucht, I. Aksentijevich, and D. L. Kastner The familial Mediterranean fever protein, pyrin, associates with microtubules and colocalizes with actin filaments Blood, August 1, 2001; 98(3): 851 - 859. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||