- Email: [email protected]
Structure, Genetic Localization, and Identification of the Cardiac and Skeletal Muscle Transcripts of the Human Integrin α7 Gene (ITGA7)
Biochemical and Biophysical Research Communications 260, 357–364 (1999) Article ID bbrc.1999.0916, available online at http://www.idealibrary.com on
Structure, Genetic Localization, and Identification of the Cardiac and Skeletal Muscle Transcripts of the Human Integrin a7 Gene (ITGA7) Nicolas Vignier,* Behzad Moghadaszadeh,* Franc¸oise Gary,† Jacques Beckmann,† Ulrike Mayer,‡ and Pascale Guicheney* ,1 *INSERM U153, Groupe Hospitalier Pitie´-Salpeˆtrie`re, 75651 Paris Cedex 13, France; †Ge´ne´thon-CNRS URA 1922, Evry, France; and ‡Max-Planck Institute for Biochemistry, D-82152 Martinsried, Germany
Received April 27, 1999
We have determined the structure and the exon size pattern of the human integrin a7 subunit gene (ITGA7), which has been shown to be affected in a form of congenital myopathy. The gene is composed of at least 27 exons spanning a region of about 22.5 kb. The sequence of all exon/intron boundaries was determined and conforms to the GT/AG splicing consensus. We investigated the different splicing forms previously described in human and rodents. The major cytoplasmic variants a7A and a7B, which are developmentally regulated and tissue specific, were identified in human tissues, as well as the extracellular isoforms X1 and X2. The recently described D variant was detected in adult tissues by RT-PCR but not the C variant. We localized ITGA7 on chromosome 12q13 by high-resolution radiation hybrid mapping between D12S312 and D12S90 and identified a new CA-repeat microsatellite in intron 1. © 1999 Academic Press
Integrins are a large family of heterodimeric cell surface receptors which mediate cell– extracellular matrix and cell-cell interactions (2– 4). These transmembrane glycoproteins are composed of an a and a b subunit which are non covalently associated (5). In mammals, 16 a and 8 b subunits have been identified so far which form more than 20 distinct receptors with diverse but often overlapping ligand-binding functions (6, 7). This association is apparently required for the transport of the dimer to the cell surface, the interaction with cytoskeletal proteins, and the formation of the ligand-binding site. In addition, alternative splic1 To whom correspondence should be addressed. Fax: (33 1) 42 16 57 00. E-mail: [email protected]. All genomic sequences have been deposited on EMBL database under Accession Nos. AJ228836 to AJ228862. The (CA) 13 sequence has been deposited on EMBL database under Accession No. AJ006533.
ing occurring in the extracellular domain and/or the cytoplasmic domain enlarge the specificity of action of these proteins. The extracellular matrix surrounding skeletal muscle is composed of two organized networks of laminins and collagens, linked by various proteins such as fibulin, nidogen or perlecan (8). The link of extracellular matrix to muscle fibers, which is essential for the normal development and function of the skeletal muscle, is mediated through at least two types of transmembrane protein complexes, the a7b1 integrin and the dystrophin-associated glycoprotein complex. The integrin a7 chain, originally discovered on skeletal myoblasts (9 –11), is complexed with the b1 subunit to form the principal integrin on skeletal myoblasts during the early stage of development and on adult muscle fibers (12). It mediates cell adhesion to laminin-1 (9, 10), as well as laminin-2 and -4 (13–15). Besides, laminins bind also to a dystroglycan, an extracellular protein, part of the membrane glycoproteins complex referred as to the dystrophin-associated glycoprotein, to which the cytoskeletal component dystrophin is tightly associated (16, 17). Comparison of the integrin a7 expression pattern with that of different laminin isoforms and dystroglycan revealed a coordinate temporal expression of dystroglycan, a7 integrin, and laminin-2 and -4 in the forming skeletal muscle (18). Identification of mutations in the genes encoding these proteins such as dystrophin, sarcoglycans and laminin a2 chain resulting in muscular dystrophies has underlined their role in the maintenance of the myofibers cytoarchitecture as well as for their anchorage and viability (17, 19). The importance of the integrin a7b1 for the functional integrity of muscle has recently been highlighted. Altered expression of the a7b1 integrin was reported in human and murine muscular dystrophies (14, 15). In addition, a null allele of mouse integrin a7
357
0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
gene demonstrated that integrin a7b1 is not essential for myogenesis but revealed typical symptoms of a progressive muscular dystrophy accompanied by an impairment of function of the myotendinous junction (20). More recently, three mutations in human integrin a7 gene, ITGA7, have been identified in patients affected by a form of congenital myopathy (1). These mutations induced premature stop codon and an integrin a7 deficiency in muscle fibers. The human integrin a genes are quite dispersed in the genome, but most of them are located in proximity to each other, indicating their evolutionary relationship. ITGA7 is localized on chromosome 12q13 (21), close to a5 and b7 subunit genes (22–24), but the protein encoded by ITGA7 presents a higher degree of homology with a3 and a6 subunits (25). Different isoforms of a7/b1 integrin appear to have distinct functions in skeletal muscle. Alternative forms of both the a7 (12, 25–28) and b1 chains (29 –32) are expressed in a developmentally regulated fashion during myogenesis, and different isoforms localize at specific sites on myofibers (33). Detailed characterization of the ITGA7 gene is essential for the study of its regulation, analysis of mutations, and its implication in the pathogenesis of congenital myopathy. In this work, we determined the entire sequence of human ITGA7 gene, and showed that the integrin a7 subunit is organized into 27 exons spanning a region of about 22.5 kb. We refined the ITGA7 localization to perform accurate linkage analysis and determined primers to amplify each exon and its intron boundaries. We investigated the different splicing forms previously described in human or rodents. EXPERIMENTAL PROCEDURES Human integrin a7 cDNA (ITGA7) sequence. Based on the partial 39 end cDNA sequence HSA7BINT (Accession No. X74295) (12) and the high homology between human and mouse cDNA sequences, cross species oligonucleotides were designed using OLIGO 4.0 program, to amplify human ITGA7 cDNA fragments by RT-PCR from human skeletal muscle. The 59 end of human cDNA was obtained using rapid amplification of cDNA ends procedure (59 RACE system for rapid amplification of cDNA ends, Version 2.0, GibcoBRL). Both strands of all fragments were sequenced. Genomic organization of ITGA7. The intronic sequences of ITGA7 were also determined by PCR. Oligonucleotides were designed on cDNA sequence to amplify exon to exon fragments on genomic DNA. According to the size of introns, we used either a standard PCR protocol for introns smaller than 1 kb (Taq polymerase Eurobio) or a long distance PCR protocol for introns of 1 to 5.5 kb (Advantage Genomic Polymerase Mix, Clontech). Both strands of the amplified fragments were sequenced. Introns sizes were determined on agarose gel or by sequencing. Based on these new sequences, oligonucleotides were designed to amplify each exon and its exon/intron boundaries. Amplifications from genomic DNA were performed using the “touch down” method, with decreasing of the annealing temperature by 0.5°C after each cycle and with 30 cycles at the lowest temperature. At each cycle denaturation was performed at 94°C for 30 s, extension at 72°C for 1
min, and final extension at 72°C for 2.5 min. The annealing temperature range was 65 to 50°C depending on the primer pair T m. Each exon was amplified from at least five different genomic DNA samples, the two strands sequenced and analyzed using GCG sequence analysis package. Sequencing. Sequencing was performed by the dideoxynucleotide chain termination method with fluorescent dideoxynucleotides using a 377 Applied Biosystem DNA sequencer (Perkin/Elmer/ABI). Identification of a microsatellite. A (CA) 13 repeat in intron 1 was found by amplification of a genomic DNA fragment using a forward primer in exon 1, and a reverse primer in exon 2. This fragment was cloned in a pGEM-T plasmid vector and transformed in JM109 cells using TA cloning protocol (pGEM-T and pGEM-T Easy Vector Systems, Promega). One clone was isolated, purified using a plasmid purification system (Qiagen) and sequenced with T7 and Sp6 primers. The two following oligonucleotides were designed to amplify by PCR, a 252-bp fragment containing the (CA) 13 repeat: 59 AGAGAAGGGGCAAATGTCAGTT 39 and 59 CAAGGAGGGGGATAGGGAAG 39. PCR conditions were 1 min at 94°C, 1 min at 55°C, and 1.5 min at 72°C for 30 cycles. Seventy-four unrelated Caucasian individuals were genotyped as previously described (34) and the allele frequencies were determined. Localization of ITGA7. The Genebridge 4 radiation hybrid panel (Research Genetics, Huntsville, AL) and the primers designed to amplify exon 2 were used to refine the localization of ITGA7. The PCR products were analyzed by 1.5% agarose gel electrophoresis. LOD score calculations were processed by the RHMAPPER program on the Whitehead Institute/MIT center server (http://www-genome. wi.mit.edu/cgi-bin/contig/rhmapper.pl) with a required LOD score for linkage equal to 15 (35). Localization of the Ge´ne´thon markers D12S312, D12S1707, and D12S90 was also reassessed. Identification of the human ITGA7 variants. The different variants previously described in rodents were investigated in human cardiac and skeletal muscle tissues. Myoblasts and myotubes were obtained from a skeletal muscle biopsy by the explant technique (36). The medium was changed from 20% fetal calf serum to 5% horse serum to facilitate differentiation into myotubes as previously described (37). Total RNAs were prepared using RNA Plus kit (Bioprobe Systems) and cDNA using First-Strand cDNA Synthesis kit (Pharmacia Biotech) with random hexadeoxynucleotides. To amplify specifically the N-terminal X1 and X2 variants, primers were chosen in exons 5 (5F) and 7 (7R), and in exons 6 (6F) and 7 (7R), respectively. The X1 primer pair (5F: 59 GCCAGGGTGGAGCTCTG 39, 7R: 59 CAGCCAGTGAGTAGCCAAAG 39) generated a fragment of 313 bp, and the X2 primer pair (6F: 59 TGATAGCTCAGACCCCGACCA 39, 7R: 59 CAGCCAGTGAGTAGCCAAAG 39) amplified a fragment of 284 bp. For the C-terminal variants, a first PCR was performed with primers located in exons 25 (25F: 59 AGCACTGCTGGTGCTGCTCC 39) and 27 (27R1: 59 AGACGAAACCACGAAACCACTA 39), which allowed the amplification of three fragments of 834, 721, and 81 bp, corresponding to the A, B, and C variants, respectively. Then specific nested PCRs were used for a better investigation of these three transcripts. A second set of primer was designed to analyze separately A, B variants and A, B, C variants. The A/B primer pair (25F: 59 AGCACTGCTGGTGCTGCTCC 39 and 27R2: 59 TGGTACTGGGGCACGGTGGC 39) produced two fragments, one of 81 bp in size corresponding to the B variant and the other of 194 bp in size corresponding to the A variant (Fig. 2C). The A/B/C primer pair (25F: 59 AGCACTGCTGGTGCTGCTCC 39 and 27R3: 59 ACCACGAAACCACTAGACTGAT 39) should generate three fragments of 827 bp, 714 bp and 74 bp corresponding to A, B and C variants, respectively, but the C variant was never identified (data not shown). The D variant, recently described in human tissues by Leung et al. corresponds to partial deletion of exon 15 and the deletion of exon 16 (28). Primers in exons 14 (14F: 59 CTTCGGGCCATTGTAGTGAC 39) and 17 (17R: 59 TTGGACAGGCAGAGTGGCTT 39) generated two
358
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
fragments, the normal transcript of 464 bp and a deleted fragment of 239 bp. cDNA amplifications were performed using the “touch down” method with annealing temperatures decreasing from 65 to 60°C. All fragments were isolated on electrophoresis gel (Nusieve 2%), reamplified and both strands were sequenced.
RESULTS Determination of ITGA7 cDNA sequence. The cDNA sequence was established by a combination of RT-PCR using cross species oligonucleotides and rapid amplification of cDNA ends procedure from human skeletal muscle. Six overlapping fragments were sequenced and continued to generate ITGA7 cDNA sequence in accordance with the recently published human cDNA sequence (Accession No. AF032108) (28). The deduced a7 protein sequence was highly homologous to human a6 and a3 proteins with seven conserved repeat domains, three divalent cation-binding regions and 18 cysteine residues conserved in the extracellular domain (Fig. 1). The deduced human a7 protein sequence (X2/A isoform) is 59% similar with the sequence of the corresponding human a6 homologue, and is 48% similar with the sequence of the corresponding human a3 homologue. Determination of ITGA7 genomic sequence. Exonto-exon PCR amplification on genomic DNA allowed to detect 27 exons of 62 to 256 bp (Table 1). The gene is about 22500 bp in size. The sizes of the introns vary from 82 to 4800 bp, most of them are less than 1 kb, the longest is intron 1 (Table 1). Based on the intronic sequences, primer pairs were designed to amplify each exon and flanking intronic sequences, and PCR conditions were established for specific amplification. The intron– exon boundaries of ITGA7, determined by comparing the genomic sequence to the cDNA sequence, fitted with the characteristic consensus sequences of 59 and 39 intronic splice sites, which are GTRAGY and YNYAG, respectively (38). The latter was found in all the acceptor splice sites. Donor splice sites were mostly GTRAGY, however introns 5, 6, 8, 13, 17, 26, contained GTAGGG, GTTTGT, GTGTGA, GTGGAC, GTTGGC, and GTAACT, respectively. The sequences of exon/intron boundaries are reported in Table 1. The extracellular domain was encoded by exon 1 to exon 24. Exon 25 encoded the transmembrane domain, whereas exon 26 and exon 27 encoded the cytoplasmic domain (Fig. 1). Identification of an intragenic (CA)n repeat polymorphism. A dinucleotide (CA)n repeat was identified in intron 1. The determination of the sequences flanking this microsatellite allowed us to determine specific forward and reverse primers, to amplify a 252-bp fragment containing the (CA) 13 repeat. By genotyping 74 unrelated Caucasian individuals, 3 alleles with the following sizes and frequencies were detected: allele 1,
252 bp, 0.35; allele 2, 250 bp, 0.09; and allele 3, 248 bp, 0.56. The heterozygosity rate was 0.54. The genotype of the control individual (1347– 02) used for the Ge´ne´thon human linkage map (39) was 252/250. Refined localization of ITGA7. Localization of ITGA7 which has been previously mapped on chromosome 12q13 by FISH (21) was refined by using GB4 radiation hybrid panel. We mapped ITGA7 at 5 cRay of WI-6672. Analysis of Ge´ne´thon markers localized nearby showed that ITGA7 was localized between D12S312 and D12S90 in a 2 cM interval and closely linked to D12S1707. ITGA7 transcripts. To determine if the developmentally regulated isoforms described in rodents are also expressed in humans, we studied the various transcripts by RT-PCR in cardiac and skeletal muscles, and in cell cultures (Fig. 2). In myoblasts, X1, X2 and B variants were abundantly expressed while the A variant was present at low level. In myotubes X1, X2, A and B transcripts were detected, but the transcription rate of the X1 variant was lower than in myoblasts, and the A transcript was upregulated. In fetal and adult skeletal muscles, X1, X2, A and B variants were expressed. In heart, a comparable pattern was observed, except for the A variant which was not detectable. The D variant was not detected in myoblasts and myotubes but was found at low level in fetal and adult skeletal muscle, and in heart. Nevertheless, in these three tissues the normal undeleted transcript was clearly the most abundant form. We were not able to amplify the C variant in our samples by RT-PCR (data not shown). We also amplified three additional transcripts corresponding to non-classical splicing mechanisms of exons 5 and 6, which are usually alternatively spliced (Figs. 2B and 2C). The first, called X3, resulted from the association of exons 5 and 6, and was present at low level in all the tissues and cultured cells. The second, called X4, corresponding to exon 5 and exon 6 deleted of its 28 first bases was only found in fetal skeletal muscle. The third one, X5, corresponded to exon 5, intron 5 and exon 6. This variant was detected in fetal and adult skeletal muscle, and cardiac muscle. Only a limited number of copies of these transcripts was detected, the major transcript is clearly the X2 form in most of the tissues. Translation of the X4 and X5 transcripts lead to premature stop codons. DISCUSSION The present work describes the cDNA sequence and the complete genomic organization of human a7 integrin gene, ITGA7. The cDNA sequence that we established was identical to the human sequence recently reported by Leung et al. (28), with in addition the sequence of exon 26, itself previously reported as a human C-terminal 113-bp insertion (12). We identified
359
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Alignment of the amino acid sequences of human a7, a6 and a3 integrin subunits. The deduced amino acid sequences of human a7 (variant X1/A), a6 (Accession No. P23229) and a3 (Accession No. P26006) integrin subunits have been aligned. Amino acid number 1 corresponds to the first amino acid in the mature a integrins. Position minus 33 of a7 corresponds to the initiating methionine and the beginning of the signal peptide. The positions of ITGA7 introns are indicated by vertical arrows. Identical amino acids are indicated by an asterisk (*), related amino acids are indicated by a . or : depending of the degree of their physicochemical properties. Gaps indicated by hyphens have been introduced to maximize the alignments. Conserved cysteines (C) in the extracellular domain of a7 are indicated in bold. The seven conserved repeat domains are overlined and numbered, the putative metal-binding domains are shaded and the predicted transmembrane domain boxed, as well as the GFF(H/K)R sequence, common to a integrins.
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
The Exon–Intron Boundaries in ITGA7 Exon
Position on cDNA
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
1 207 335 415 671 803 923 1131 1325 1414 1542 1638 1700 1870 2020 2136 2329 2410 2490 2565 2668 2845 2977 3091 3190 3316 3429
Splice acceptor site
Exon size (bp)
Splice donor site
GGATCGTCCCATGGCCGGGG.. 206 ..CCCAGAGCTGgtgagtcacc catcttgcagGCTGCTGGTG.. 128 ..GACCAGGGAGgtgtggccct ctgcctctagCTGATATGCA.. 80 ..CAAGATTGTTgtgagtattg cctcccccagACCTGTGCAC.. 256 ..AATTGGAAGGgtgagtcact gtccttgcagGCACGGCCAG.. 132 ..AGCTACTTTGgtagggacct ccccccacagGGTTGCTTTT.. 120 ..AGCTACTTAGgtttgtaagc cttcccccagGCTTCTCTAT.. 208 ..ACAGTGATGGgtgagtgggt ctctccatagCTGGCCAGAC.. 194 ..GGCTTTCCAGgtgtgacggg tgccctgcagATATTGCAGT.. 89 ..ACCTTCACAGgtgaggggag ggttctctagGTGCTGGAGG.. 128 ..TGCTCTTCAGgtgagcccct ccaatcccagGGCCAGACCC.. 96 ..ACTCGGTCTGgtgaggtggg gcccccgcagTGTGGACCTA.. 62 ..CCTACTGTGGgtgagtgcgg catgccacagCCCTGGACTA.. 170 ..CCAGCTCCAGgtggacactg ctcttcctagGAAAATGTCA.. 150 ..GCGGGCAGAGgtgagcatgg ctctccccagATCCACTTCC.. 116 ..CTCTGCCCATgtgagggggg gactctccagGGATGTGGAT.. 193 ..GGACCCTGCGgtgaggacct cccattccagGAGAAGCCAC.. 81 ..AGGTGCCCAGgttggcacat cttggcctagGTCACCTTCT.. 80 ..TGTTGGCCACgtaagccagg cactcaccagGATCAGTGAG.. 75 ..CCATTGCAGGgtgagcctgg tccacgccagAATGGCCATT.. 103 ..TGAGGTCACGgtaagtgtca tttcccacagGTTTCCAACC.. 177 ..CCTCCACCTGgtgaggctta ccgtgtccagGATGTGGACA.. 132 ..CATCACCCTGgtgagggcag ttatccctagGACTGCGCCC.. 114 ..CTTTCTGGAGgtgaggatac gttcccttagGAGTACTCAG.. 99 ..CTCCACAGTGgtgagctgca tgctccccagATCCCAGTGA.. 126 ..CCTGTGGAAGgtgaggcttg gacatcctagTGTGGCTTCT.. 113 ..GGACTGTGGGgtaactgttg gtccttatagATGGGATTCTTCAAACGGGCGAAGCACCCCGAGGCCACCGTGCCCCAGTACCATG CGGTGAAGATTCCTCGGGAAGACCGACAGCAGTTCAAGGAGGAGAAGACGGGCACCATCCTGAGG AACAACTGGGGCAGCCCCCGGCGGGAGGGCCCGGATGCACACCCCATCCTGGCTGCTGACGGGCA TCCCGAGCTGGGCCCCGATGGGCATCCAGGGCCAGGCACCGCCTAGGTTCCCATGTCCCAGCCTG CGCTGTGGCTGCCCTCCATCCCTTCCCCAGAGATGGCTCCTTGGGATGAAGAGGGTAGAGTGGGC TGCTGGTGTCACATCAAGATTTGGCAGGATCGGCTTCCTCAGGGGCACAGACCTCTCCCACCCAC AAGAACTCCTCCCACCCAACTTCCCCTTAGAGTGCTGTGAGATGAGAGTGGGTAAATCAGGGACA GGGCCATGGGGTAGGGTGAGAAGGGCAGGGGTGTCCTGATGCAAAGGTGGGGAGAAGGATCCTAA TCCCTTCCTCTCCCATTCACCCTGTGTAACAGGACCCCAAGGACCTGCCTCCCCGGAAGTGCCTT AACCTAGAGGGTCGGGGAGGAGGTTGTGTCACTGACTCAAGGCTGCTCCTTCTCTAGTTTCCCCT CTCATCTGACCTTAGTTTGCTGCCATCAGTCTAGTGGTTTCGTGGTTTCGTCTATTTATTAAAAA ATATTTGAGAACAAAAAAAAAAAA
Intron size (bp) '4800* 99 '2500* 405 272 953 134 356 115 148 173 147 '700* 647 551 82 92 112 132 700 165 '2000* 540 105 '2000* 913
Note. Exon and intron sequences are in uppercase and lowercase letters, respectively. Splice site consensus nucleotides conforming to the ag/gt rule, transcription, and the stop codons corresponding to the A (58 bp after the beginning of exon 27) and B (228 bp after the beginning of exon 27) isoforms are in boldface. Nucleotide cDNA sequence numbering starts at the transcription codon. The underlined sequence corresponds to the sequence spliced in the rat C isoform. Intron sizes were determined by PCR (*) or sequencing. All exon sequences have been deposited on EMBL Database under Accession Nos. AJ228836 to AJ228862.
only two non silent polymorphisms in the coding sequence. The deduced amino acid sequence of ITGA7 showed high homology with human a3 and a6 integrin subunits (40, 41) as it has been previously reported for the mouse a7 integrin (25). This gene is about 22 500 bp and contains as in mouse 27 exons (20) (Mayer, personal communication). Several major isoforms, which may modify ligand affinity and specificity, cytoplasmic interactions and signaling, have been described. Their variations in the extracellular and the cytoplasmic domains correspond to the splicing of exons 5, 6, and 26. These exons present 100% homology with mouse (Mayer, personal communication). In addition, the 39 splice sites of hu-
man exons 5, 6, and 26 which are GTAGG, GTTTG, GTAAC, respectively, are mild consensus sites as in mouse (Mayer, personal communication), supporting the prediction of the same major isoforms in the two species. The integrin human a7 integrin isoforms are composed of an extracellular domain of 1042 aa for isoform X1, and of 1038 aa for isoform X2, a 23-amino-acid transmembrane domain and a short cytoplasmic domain of 57 amino acids for isoform A, or of 76 amino acids for isoform B. The two extracellular isoforms, designed X1 and X2, derive from the mutually exclusive splicing of exons 5 or 6 (11, 25). These two isoforms were present in equal
361
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. The human a7 integrin gene, ITGA, and its various splice variants. (A) Genomic organization of ITGA7 and the various splice variants. The rectangles denote exons, numbered from 1 to 27, and the thin lines represent either intronic sequences, or regions that are spliced in/out in the variants. Alternatively spliced exons. (B) Schematic representation of the divergent parts of the 8 transcriptional variants detected by RT-PCR in human myoblasts, myotubes, and skeletal or cardiac muscles. (C) Detection of splice variants by specific RT-PCR amplification of total RNAs extracted from various human tissues and separation by 2% agarose gel electrophoresis. The PCR products were stained with ethidium bromide. The M r marker is a 100-bp ladder. On the left: identification of the variants according to their sizes (the primers are given under Experimental Procedures). Each band has been cloned and sequenced.
amounts in mouse C2C12 myoblasts and adult heart, while the X2 variant was found nearly exclusively expressed in adult skeletal muscle (25). In human tissues the two variants were found in human cardiac, and fetal and adult skeletal muscles. Nevertheless, X1 variant was more abundant in myoblasts and fetal skeletal muscle and was downregulated during development. Two major cytoplasmic isoforms, designed A and B, have been described in rat, mouse and human (12, 26) and specific anti-integrin antibodies have been developed (12, 15, 18, 33). The A isoform, which corresponds to a normal splicing of exon 26 (38 amino acids) and formation of a stop codon after the 19 first amino acids encoded by exon 27, is muscle specific (12, 26). The B isoform is the longest isoform with a different C-terminal amino acid sequence. This is due to the splicing of exon 26 and to the formation of a stop codon after 76 amino acids in exon 27. This isoform is widely expressed not only in skeletal, cardiac and vascular smooth muscles but also in a number of other tissues, such as liver, spleen or brain (18, 26). We observed also that the isoform A is muscle specific, upregulated during muscle development and not expressed in human heart.
A third cytoplasmic variant, designed a7C, was isolated from a rat muscle cDNA library and detected in low amounts in poly(A) 1 RNA, prepared from rat myotubes (12) but we were not able to detect the corresponding variant in total RNA from human skeletal muscle even after two nested PCRs. If the same splicing as in rat is considered, no in frame stop codon can be formed due to a 8-bp insertion in the human compared to rat 39 part of exon 27 and in addition the potential splice donor site which is CGGGCG in human does not agree with consensus splice donor sequence. This variant was also neither found in mouse (Mayer, personal communication). Another isoform, designed as a7D, presenting a 225-bp in-frame deletion was recently reported by Leung et al. (28). This new isoform appears to be a partial deletion of exon 15 associated with the entire deletion of exon 16. Indeed, a 39 splice site consensus, GTCAG is present at the position 85 in exon 15. The use of this splice site should induce the deletion of the last 32 nucleotides of exon 15 and the 193 nucleotides of exon 16. We did detect this transcript in human cardiac and skeletal muscle but only at the adult stage of development.
362
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
We detected three additional minor transcripts, X3 to X5, in this study. The translation of the X4 and X5 transcripts would result in truncated proteins of 297 and 362 residues, respectively. These variants may be due to immature transcripts present at low level in nucleus or cytoplasm and amplified by PCR, without any physiological function. The X3 variant which corresponds to an in-frame association of exons 5 and 6 has been described in the human a6 subunit, and associated with b subunits forms functional laminin receptors (42). The identification of a new intragenic microsatellite marker in intron 1 of the gene and the refinement of its localization between two highly informative markers, D12S312 and D12S90, provides useful tools for linkage analysis. Intronic primers to amplify each exon and its intron boundaries should facilitate the identification of new ITGA7 mutations in congenital myopathy. So far the three first mutations identified in ITGA7 were mutations leading to premature stop codons (1) in patients with a complete absence of a7b1 integrin in muscle biopsies. Mutations in other integrin subunits cause human inherited diseases. A homozygous splicing mutation in the human a6 integrin gene, ITGA6, was identified in a patient with junctional epidermolysis bullosa with congenital duodenal atresia (43). Missense mutations occurring in functional sites may also cause diseases, as it has been shown for b3 integrin subunit in Glanzmann thrombasthenia, a recessive autosomal bleeding disorder characterized by abnormal platelet aggregation due to a qualitative or quantitative defect of the glycoprotein IIb–IIIa complex (integrin aIIbb3) (44). Identification of missense mutations in ITGA7 will help to define the ligand binding sites and the functional sites of this protein, which plays an important role in human skeletal muscle development and integrity. ACKNOWLEDGMENTS We thank Marc Fiszman for helpful discussion, Martine Verdie`reSahuque´ for her help in preparing skeletal muscle cells, and Delphine Muselet for technical assistance. This work was supported by the Institut National de la Sante´ et de la Recherche Me´dicale and the Association Franc¸aise contre les Myopathies.
REFERENCES 1. Hayashi, Y. K., Chou, F.-L., Engvall, E., Ogawa, M., Mastuda, C., Hirabayashi, S., Yokochi, K., Ziober, B. L., Kramer, R. H., Kaufman, S. J., Ozawa, E., Goto, Y., Nonaka, I., Tsukahara, T., Wang, J. Z., Hoffman, E. P., and Arahata, K. (1998) Nature Genet. 19, 94 –97. 2. Hynes, R. O., and Lander, A. D. (1992) Cell 68, 303–322. 3. Gullberg, D., Velling, T., Lohikangas, L., and Tiger, C. F. (1998) Front Biosci. 15, D1039 –D1050. 4. Ekblom, P., and Timpl, R. (1996) Curr. Opin. Cell Biol. 8, 599 – 601.
5. Ruoslahti, E. (1988) Annu. Rev. Biochem. 57, 375– 413. 6. Camper, L., Hellman, U., and Lundgren-Akerlund, E. (1998) J. Biol. Chem. 273, 20383–20389. 7. Horwitz, A. F. (1997) Sci. Am., 46 –53. 8. Timpl, R., and Brown, J. C. (1996) Bioessays 18, 123–132. 9. Kramer, R. H., Vu, M. P., Cheng, Y. F., Ramos, D. M., Timpl, R., and Waleh, N. (1991) Cell Regul. 2, 805– 817. 10. von der Mark, H., Du¨rr, J., Sonnenberg, A., von der Mark, K., Deutzmann, R., and Goodman, S. L. (1991) J. Biol. Chem. 266, 23593–23601. 11. Song, W. K., Wang, W., Foster, R. F., Bielser, D. A., and Kaufman, S. J. (1992) J. Cell Biol. 117, 643– 657. 12. Song, W. K., Wang, W., Sato, H., Bielser, D. A., and Kaufman, S. J. (1993) J. Cell. Sci. 106, 1139 –1152. 13. Yao, C. C., Ziober, B. L., Squillace, R. M., and Kramer, R. H. (1996) J. Biol. Chem. 271, 25598 –25603. 14. Hodges, B. L., Hayashi, Y. K., Nonaka, I., Wang, W., Arahata, K., and Kaufman, S. J. (1997) J. Cell Sci. 110, 2873–2881. 15. Vachon, P. H., Xu, H., Liu, L., Loechel, F., Hayashi, Y., Arahata, K., Reed, J., Wewer, U. M., and Engvall, E. (1997) J. Clin. Invest. 100, 1870 –1881. 16. Ervasti, J. M., and Campbell, K. P. (1993) J. Cell Biol. 122, 809 – 823. 17. Campbell, K. (1995) Cell 80, 675– 679. 18. Velling, T., Collo, G., Sorokin, L., Durbeej, M., Zhang, H., and Gullberg, D. (1996) Dev. Dyn. 207, 355–371. 19. Helbling-Leclerc, A., Zhang, X., Topaloglu, H., Cruaud, C., Tesson, F., Weissenbach, J., Tome´, F. M. S., Schwartz, K., Fardeau, M., Tryggvason, K., and Guicheney, P. (1995) Nat. Genet. 11, 216 –218. 20. Mayer, U., Saher, G., Fassler, R., Bornemann, A., Echtermeyer, F., von der Mark, H., Miosge, N., Poschl, E., and von der Mark, K. (1997) Nat. Genet. 17, 318 –323. 21. Wang, W., Wu, W., Desai, T., Ward, D. C., and Kaufman, S. J. (1995) Genomics 26, 568 –570. 22. Sosnoski, D. M., Emanuel, B. S., Hawkins, A. L., van Tuinen, P., Ledbetter, D. H., Nussbaum, R. L., Kaos, F. T., Schwartz, E., Phillips, D., Bennett, J. S., et al. (1988) J. Clin. Invest. 81, 1993–1998. 23. Baker, E. (1992) Mamm. Genome 2, 272–273. 24. Krissansen, G. W., Yuan, Q., Jenkins, D., Jiang, W., Rooke, L., Spurr, N. K., Eccles, M., Leung, E., and Watson, J. D. (1992) Immunogenetics 35, 58 – 61. 25. Ziober, B. L., Vu, M. P., Waleh, N., Crawford, J., Lin, C. S., and Kramer, R. H. (1993) J. Biol. Chem. 268, 26773–26783. 26. Collo, G., Starr, L., and Quaranta, V. (1993) J. Biol. Chem. 268, 19019 –19024. 27. Hodges, B. L., and Kaufman, S. J. (1996) Basic Appl. Myologie 6, 437– 446. 28. Leung, E., Pheng Lim, S., Berg, R., Yang, Y., Ni, J., Wang, S. X., and Krissansen, G. W. (1998) Biochem. Biophys. Res. Commun. 243, 317–325. 29. Balzac, F., Belkin, A. M., Koteliansky, V. E., Balabanov, Y. V., Altruda, F., Silengo, L., and Tarone, G. (1994) J. Cell Biol. 121, 171–178. 30. Zhidkova, B. L., Belkin, A. M., and Mane, R. (1995) Biochem. Biophys. Res. Commun. 214, 279 –285. 31. van der Flier, A., Kuikman, I., Baudoin, C., van der Neut, R., and Sonnenberg, A. (1995) FEBS Lett. 369, 340 –344. 32. Belkin, A. M., Zhidkova, N. I., Balzac, F., Altruda, F., Tomatis, D., Maier, A., Tarone, G., Koteliansky, V. E., and Burridge, K. (1996) J. Cell Biol. 132, 211–226.
363
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
33. Martin, P. T., Kaufman, S. J., Kramer, R. H., and Sanes, J. R. (1996) Dev. Biol. 174, 125–139. 34. Vignal, A., Gyapay, G., Hazan, J., Nguyen, S., Dupraz, C., Cheron, N., Becuwe, N., Tranchant, M., and Weissenbach, J. (1993) in Non-radioactive Multiplex Procedure for Genotyping of Microsatellite Markers (Series Ed.), Vol. 211–221, Academic Press, New York. 35. Slonim, D., Kruglyak, L., Stein, L., and Lander, E. (1997) J. Comput. Biol. 4, 487–504. 36. Askanas, V., and Engel, W. K. (1975) Neurology 25, 58 – 67. 37. Verdie`re-Sahuque´, M., Akaaboune, M., Lachkar, S., Festoff, B. W., Jandrot-Perrus, M., Garcia, L., Barlovatz-Meimon, G., and Hantai, D. (1996) Exp. Cell Res. 222, 70 –76. 38. Shapiro, M. B., and Senapathy, P. (1987) Nucleic Acids Res. 15, 7155–7174.
39. Dib, C., Faure´, S., Fizames, C., Samson, D., Drouot, N., Vignal, A., Millasseau, P., Seboun, E., Marc, S., Lathrop, M., Gyapay, G., Morissette, J., and Weissenbach, J. (1996) Nature 380, 152– 154. 40. Takada, Y., Murphy, E., Pil, P., Chen, C., Ginsberg, M. H., and Hemler, M. E. (1991) J. Cell Biol. 115, 257–266. 41. Hogervorst, F., Kuikman, I., van Kessel, A. G., and Sonnenberg, A. (1991) Eur. J. Biochem. 199, 425– 433. 42. Delwel, G. O., Kuikman, I., and Sonnenberg, A. (1995) Cell Adhes. Commun. 3, 143–161. 43. Pulkkinen, L., Kurtz, K., Xu, Y., Bruckner-Tuderman, L., and Uitto, J. (1997) Lab. Invest. 76, 823– 833. 44. Bajt, M. L., Ginsberg, M. H., Frelinger AL, d., Berndt, M. C., and Loftus, J. C. (1992) J. Biol. Chem. 267, 3789 –3794.
364
Structure, Genetic Localization, and Identification of the Cardiac and Skeletal Muscle Transcripts of the Human Integrin a7 Gene (ITGA7) Nicolas Vignier,* Behzad Moghadaszadeh,* Franc¸oise Gary,† Jacques Beckmann,† Ulrike Mayer,‡ and Pascale Guicheney* ,1 *INSERM U153, Groupe Hospitalier Pitie´-Salpeˆtrie`re, 75651 Paris Cedex 13, France; †Ge´ne´thon-CNRS URA 1922, Evry, France; and ‡Max-Planck Institute for Biochemistry, D-82152 Martinsried, Germany
Received April 27, 1999
We have determined the structure and the exon size pattern of the human integrin a7 subunit gene (ITGA7), which has been shown to be affected in a form of congenital myopathy. The gene is composed of at least 27 exons spanning a region of about 22.5 kb. The sequence of all exon/intron boundaries was determined and conforms to the GT/AG splicing consensus. We investigated the different splicing forms previously described in human and rodents. The major cytoplasmic variants a7A and a7B, which are developmentally regulated and tissue specific, were identified in human tissues, as well as the extracellular isoforms X1 and X2. The recently described D variant was detected in adult tissues by RT-PCR but not the C variant. We localized ITGA7 on chromosome 12q13 by high-resolution radiation hybrid mapping between D12S312 and D12S90 and identified a new CA-repeat microsatellite in intron 1. © 1999 Academic Press
Integrins are a large family of heterodimeric cell surface receptors which mediate cell– extracellular matrix and cell-cell interactions (2– 4). These transmembrane glycoproteins are composed of an a and a b subunit which are non covalently associated (5). In mammals, 16 a and 8 b subunits have been identified so far which form more than 20 distinct receptors with diverse but often overlapping ligand-binding functions (6, 7). This association is apparently required for the transport of the dimer to the cell surface, the interaction with cytoskeletal proteins, and the formation of the ligand-binding site. In addition, alternative splic1 To whom correspondence should be addressed. Fax: (33 1) 42 16 57 00. E-mail: [email protected]. All genomic sequences have been deposited on EMBL database under Accession Nos. AJ228836 to AJ228862. The (CA) 13 sequence has been deposited on EMBL database under Accession No. AJ006533.
ing occurring in the extracellular domain and/or the cytoplasmic domain enlarge the specificity of action of these proteins. The extracellular matrix surrounding skeletal muscle is composed of two organized networks of laminins and collagens, linked by various proteins such as fibulin, nidogen or perlecan (8). The link of extracellular matrix to muscle fibers, which is essential for the normal development and function of the skeletal muscle, is mediated through at least two types of transmembrane protein complexes, the a7b1 integrin and the dystrophin-associated glycoprotein complex. The integrin a7 chain, originally discovered on skeletal myoblasts (9 –11), is complexed with the b1 subunit to form the principal integrin on skeletal myoblasts during the early stage of development and on adult muscle fibers (12). It mediates cell adhesion to laminin-1 (9, 10), as well as laminin-2 and -4 (13–15). Besides, laminins bind also to a dystroglycan, an extracellular protein, part of the membrane glycoproteins complex referred as to the dystrophin-associated glycoprotein, to which the cytoskeletal component dystrophin is tightly associated (16, 17). Comparison of the integrin a7 expression pattern with that of different laminin isoforms and dystroglycan revealed a coordinate temporal expression of dystroglycan, a7 integrin, and laminin-2 and -4 in the forming skeletal muscle (18). Identification of mutations in the genes encoding these proteins such as dystrophin, sarcoglycans and laminin a2 chain resulting in muscular dystrophies has underlined their role in the maintenance of the myofibers cytoarchitecture as well as for their anchorage and viability (17, 19). The importance of the integrin a7b1 for the functional integrity of muscle has recently been highlighted. Altered expression of the a7b1 integrin was reported in human and murine muscular dystrophies (14, 15). In addition, a null allele of mouse integrin a7
357
0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
gene demonstrated that integrin a7b1 is not essential for myogenesis but revealed typical symptoms of a progressive muscular dystrophy accompanied by an impairment of function of the myotendinous junction (20). More recently, three mutations in human integrin a7 gene, ITGA7, have been identified in patients affected by a form of congenital myopathy (1). These mutations induced premature stop codon and an integrin a7 deficiency in muscle fibers. The human integrin a genes are quite dispersed in the genome, but most of them are located in proximity to each other, indicating their evolutionary relationship. ITGA7 is localized on chromosome 12q13 (21), close to a5 and b7 subunit genes (22–24), but the protein encoded by ITGA7 presents a higher degree of homology with a3 and a6 subunits (25). Different isoforms of a7/b1 integrin appear to have distinct functions in skeletal muscle. Alternative forms of both the a7 (12, 25–28) and b1 chains (29 –32) are expressed in a developmentally regulated fashion during myogenesis, and different isoforms localize at specific sites on myofibers (33). Detailed characterization of the ITGA7 gene is essential for the study of its regulation, analysis of mutations, and its implication in the pathogenesis of congenital myopathy. In this work, we determined the entire sequence of human ITGA7 gene, and showed that the integrin a7 subunit is organized into 27 exons spanning a region of about 22.5 kb. We refined the ITGA7 localization to perform accurate linkage analysis and determined primers to amplify each exon and its intron boundaries. We investigated the different splicing forms previously described in human or rodents. EXPERIMENTAL PROCEDURES Human integrin a7 cDNA (ITGA7) sequence. Based on the partial 39 end cDNA sequence HSA7BINT (Accession No. X74295) (12) and the high homology between human and mouse cDNA sequences, cross species oligonucleotides were designed using OLIGO 4.0 program, to amplify human ITGA7 cDNA fragments by RT-PCR from human skeletal muscle. The 59 end of human cDNA was obtained using rapid amplification of cDNA ends procedure (59 RACE system for rapid amplification of cDNA ends, Version 2.0, GibcoBRL). Both strands of all fragments were sequenced. Genomic organization of ITGA7. The intronic sequences of ITGA7 were also determined by PCR. Oligonucleotides were designed on cDNA sequence to amplify exon to exon fragments on genomic DNA. According to the size of introns, we used either a standard PCR protocol for introns smaller than 1 kb (Taq polymerase Eurobio) or a long distance PCR protocol for introns of 1 to 5.5 kb (Advantage Genomic Polymerase Mix, Clontech). Both strands of the amplified fragments were sequenced. Introns sizes were determined on agarose gel or by sequencing. Based on these new sequences, oligonucleotides were designed to amplify each exon and its exon/intron boundaries. Amplifications from genomic DNA were performed using the “touch down” method, with decreasing of the annealing temperature by 0.5°C after each cycle and with 30 cycles at the lowest temperature. At each cycle denaturation was performed at 94°C for 30 s, extension at 72°C for 1
min, and final extension at 72°C for 2.5 min. The annealing temperature range was 65 to 50°C depending on the primer pair T m. Each exon was amplified from at least five different genomic DNA samples, the two strands sequenced and analyzed using GCG sequence analysis package. Sequencing. Sequencing was performed by the dideoxynucleotide chain termination method with fluorescent dideoxynucleotides using a 377 Applied Biosystem DNA sequencer (Perkin/Elmer/ABI). Identification of a microsatellite. A (CA) 13 repeat in intron 1 was found by amplification of a genomic DNA fragment using a forward primer in exon 1, and a reverse primer in exon 2. This fragment was cloned in a pGEM-T plasmid vector and transformed in JM109 cells using TA cloning protocol (pGEM-T and pGEM-T Easy Vector Systems, Promega). One clone was isolated, purified using a plasmid purification system (Qiagen) and sequenced with T7 and Sp6 primers. The two following oligonucleotides were designed to amplify by PCR, a 252-bp fragment containing the (CA) 13 repeat: 59 AGAGAAGGGGCAAATGTCAGTT 39 and 59 CAAGGAGGGGGATAGGGAAG 39. PCR conditions were 1 min at 94°C, 1 min at 55°C, and 1.5 min at 72°C for 30 cycles. Seventy-four unrelated Caucasian individuals were genotyped as previously described (34) and the allele frequencies were determined. Localization of ITGA7. The Genebridge 4 radiation hybrid panel (Research Genetics, Huntsville, AL) and the primers designed to amplify exon 2 were used to refine the localization of ITGA7. The PCR products were analyzed by 1.5% agarose gel electrophoresis. LOD score calculations were processed by the RHMAPPER program on the Whitehead Institute/MIT center server (http://www-genome. wi.mit.edu/cgi-bin/contig/rhmapper.pl) with a required LOD score for linkage equal to 15 (35). Localization of the Ge´ne´thon markers D12S312, D12S1707, and D12S90 was also reassessed. Identification of the human ITGA7 variants. The different variants previously described in rodents were investigated in human cardiac and skeletal muscle tissues. Myoblasts and myotubes were obtained from a skeletal muscle biopsy by the explant technique (36). The medium was changed from 20% fetal calf serum to 5% horse serum to facilitate differentiation into myotubes as previously described (37). Total RNAs were prepared using RNA Plus kit (Bioprobe Systems) and cDNA using First-Strand cDNA Synthesis kit (Pharmacia Biotech) with random hexadeoxynucleotides. To amplify specifically the N-terminal X1 and X2 variants, primers were chosen in exons 5 (5F) and 7 (7R), and in exons 6 (6F) and 7 (7R), respectively. The X1 primer pair (5F: 59 GCCAGGGTGGAGCTCTG 39, 7R: 59 CAGCCAGTGAGTAGCCAAAG 39) generated a fragment of 313 bp, and the X2 primer pair (6F: 59 TGATAGCTCAGACCCCGACCA 39, 7R: 59 CAGCCAGTGAGTAGCCAAAG 39) amplified a fragment of 284 bp. For the C-terminal variants, a first PCR was performed with primers located in exons 25 (25F: 59 AGCACTGCTGGTGCTGCTCC 39) and 27 (27R1: 59 AGACGAAACCACGAAACCACTA 39), which allowed the amplification of three fragments of 834, 721, and 81 bp, corresponding to the A, B, and C variants, respectively. Then specific nested PCRs were used for a better investigation of these three transcripts. A second set of primer was designed to analyze separately A, B variants and A, B, C variants. The A/B primer pair (25F: 59 AGCACTGCTGGTGCTGCTCC 39 and 27R2: 59 TGGTACTGGGGCACGGTGGC 39) produced two fragments, one of 81 bp in size corresponding to the B variant and the other of 194 bp in size corresponding to the A variant (Fig. 2C). The A/B/C primer pair (25F: 59 AGCACTGCTGGTGCTGCTCC 39 and 27R3: 59 ACCACGAAACCACTAGACTGAT 39) should generate three fragments of 827 bp, 714 bp and 74 bp corresponding to A, B and C variants, respectively, but the C variant was never identified (data not shown). The D variant, recently described in human tissues by Leung et al. corresponds to partial deletion of exon 15 and the deletion of exon 16 (28). Primers in exons 14 (14F: 59 CTTCGGGCCATTGTAGTGAC 39) and 17 (17R: 59 TTGGACAGGCAGAGTGGCTT 39) generated two
358
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
fragments, the normal transcript of 464 bp and a deleted fragment of 239 bp. cDNA amplifications were performed using the “touch down” method with annealing temperatures decreasing from 65 to 60°C. All fragments were isolated on electrophoresis gel (Nusieve 2%), reamplified and both strands were sequenced.
RESULTS Determination of ITGA7 cDNA sequence. The cDNA sequence was established by a combination of RT-PCR using cross species oligonucleotides and rapid amplification of cDNA ends procedure from human skeletal muscle. Six overlapping fragments were sequenced and continued to generate ITGA7 cDNA sequence in accordance with the recently published human cDNA sequence (Accession No. AF032108) (28). The deduced a7 protein sequence was highly homologous to human a6 and a3 proteins with seven conserved repeat domains, three divalent cation-binding regions and 18 cysteine residues conserved in the extracellular domain (Fig. 1). The deduced human a7 protein sequence (X2/A isoform) is 59% similar with the sequence of the corresponding human a6 homologue, and is 48% similar with the sequence of the corresponding human a3 homologue. Determination of ITGA7 genomic sequence. Exonto-exon PCR amplification on genomic DNA allowed to detect 27 exons of 62 to 256 bp (Table 1). The gene is about 22500 bp in size. The sizes of the introns vary from 82 to 4800 bp, most of them are less than 1 kb, the longest is intron 1 (Table 1). Based on the intronic sequences, primer pairs were designed to amplify each exon and flanking intronic sequences, and PCR conditions were established for specific amplification. The intron– exon boundaries of ITGA7, determined by comparing the genomic sequence to the cDNA sequence, fitted with the characteristic consensus sequences of 59 and 39 intronic splice sites, which are GTRAGY and YNYAG, respectively (38). The latter was found in all the acceptor splice sites. Donor splice sites were mostly GTRAGY, however introns 5, 6, 8, 13, 17, 26, contained GTAGGG, GTTTGT, GTGTGA, GTGGAC, GTTGGC, and GTAACT, respectively. The sequences of exon/intron boundaries are reported in Table 1. The extracellular domain was encoded by exon 1 to exon 24. Exon 25 encoded the transmembrane domain, whereas exon 26 and exon 27 encoded the cytoplasmic domain (Fig. 1). Identification of an intragenic (CA)n repeat polymorphism. A dinucleotide (CA)n repeat was identified in intron 1. The determination of the sequences flanking this microsatellite allowed us to determine specific forward and reverse primers, to amplify a 252-bp fragment containing the (CA) 13 repeat. By genotyping 74 unrelated Caucasian individuals, 3 alleles with the following sizes and frequencies were detected: allele 1,
252 bp, 0.35; allele 2, 250 bp, 0.09; and allele 3, 248 bp, 0.56. The heterozygosity rate was 0.54. The genotype of the control individual (1347– 02) used for the Ge´ne´thon human linkage map (39) was 252/250. Refined localization of ITGA7. Localization of ITGA7 which has been previously mapped on chromosome 12q13 by FISH (21) was refined by using GB4 radiation hybrid panel. We mapped ITGA7 at 5 cRay of WI-6672. Analysis of Ge´ne´thon markers localized nearby showed that ITGA7 was localized between D12S312 and D12S90 in a 2 cM interval and closely linked to D12S1707. ITGA7 transcripts. To determine if the developmentally regulated isoforms described in rodents are also expressed in humans, we studied the various transcripts by RT-PCR in cardiac and skeletal muscles, and in cell cultures (Fig. 2). In myoblasts, X1, X2 and B variants were abundantly expressed while the A variant was present at low level. In myotubes X1, X2, A and B transcripts were detected, but the transcription rate of the X1 variant was lower than in myoblasts, and the A transcript was upregulated. In fetal and adult skeletal muscles, X1, X2, A and B variants were expressed. In heart, a comparable pattern was observed, except for the A variant which was not detectable. The D variant was not detected in myoblasts and myotubes but was found at low level in fetal and adult skeletal muscle, and in heart. Nevertheless, in these three tissues the normal undeleted transcript was clearly the most abundant form. We were not able to amplify the C variant in our samples by RT-PCR (data not shown). We also amplified three additional transcripts corresponding to non-classical splicing mechanisms of exons 5 and 6, which are usually alternatively spliced (Figs. 2B and 2C). The first, called X3, resulted from the association of exons 5 and 6, and was present at low level in all the tissues and cultured cells. The second, called X4, corresponding to exon 5 and exon 6 deleted of its 28 first bases was only found in fetal skeletal muscle. The third one, X5, corresponded to exon 5, intron 5 and exon 6. This variant was detected in fetal and adult skeletal muscle, and cardiac muscle. Only a limited number of copies of these transcripts was detected, the major transcript is clearly the X2 form in most of the tissues. Translation of the X4 and X5 transcripts lead to premature stop codons. DISCUSSION The present work describes the cDNA sequence and the complete genomic organization of human a7 integrin gene, ITGA7. The cDNA sequence that we established was identical to the human sequence recently reported by Leung et al. (28), with in addition the sequence of exon 26, itself previously reported as a human C-terminal 113-bp insertion (12). We identified
359
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Alignment of the amino acid sequences of human a7, a6 and a3 integrin subunits. The deduced amino acid sequences of human a7 (variant X1/A), a6 (Accession No. P23229) and a3 (Accession No. P26006) integrin subunits have been aligned. Amino acid number 1 corresponds to the first amino acid in the mature a integrins. Position minus 33 of a7 corresponds to the initiating methionine and the beginning of the signal peptide. The positions of ITGA7 introns are indicated by vertical arrows. Identical amino acids are indicated by an asterisk (*), related amino acids are indicated by a . or : depending of the degree of their physicochemical properties. Gaps indicated by hyphens have been introduced to maximize the alignments. Conserved cysteines (C) in the extracellular domain of a7 are indicated in bold. The seven conserved repeat domains are overlined and numbered, the putative metal-binding domains are shaded and the predicted transmembrane domain boxed, as well as the GFF(H/K)R sequence, common to a integrins.
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
The Exon–Intron Boundaries in ITGA7 Exon
Position on cDNA
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
1 207 335 415 671 803 923 1131 1325 1414 1542 1638 1700 1870 2020 2136 2329 2410 2490 2565 2668 2845 2977 3091 3190 3316 3429
Splice acceptor site
Exon size (bp)
Splice donor site
GGATCGTCCCATGGCCGGGG.. 206 ..CCCAGAGCTGgtgagtcacc catcttgcagGCTGCTGGTG.. 128 ..GACCAGGGAGgtgtggccct ctgcctctagCTGATATGCA.. 80 ..CAAGATTGTTgtgagtattg cctcccccagACCTGTGCAC.. 256 ..AATTGGAAGGgtgagtcact gtccttgcagGCACGGCCAG.. 132 ..AGCTACTTTGgtagggacct ccccccacagGGTTGCTTTT.. 120 ..AGCTACTTAGgtttgtaagc cttcccccagGCTTCTCTAT.. 208 ..ACAGTGATGGgtgagtgggt ctctccatagCTGGCCAGAC.. 194 ..GGCTTTCCAGgtgtgacggg tgccctgcagATATTGCAGT.. 89 ..ACCTTCACAGgtgaggggag ggttctctagGTGCTGGAGG.. 128 ..TGCTCTTCAGgtgagcccct ccaatcccagGGCCAGACCC.. 96 ..ACTCGGTCTGgtgaggtggg gcccccgcagTGTGGACCTA.. 62 ..CCTACTGTGGgtgagtgcgg catgccacagCCCTGGACTA.. 170 ..CCAGCTCCAGgtggacactg ctcttcctagGAAAATGTCA.. 150 ..GCGGGCAGAGgtgagcatgg ctctccccagATCCACTTCC.. 116 ..CTCTGCCCATgtgagggggg gactctccagGGATGTGGAT.. 193 ..GGACCCTGCGgtgaggacct cccattccagGAGAAGCCAC.. 81 ..AGGTGCCCAGgttggcacat cttggcctagGTCACCTTCT.. 80 ..TGTTGGCCACgtaagccagg cactcaccagGATCAGTGAG.. 75 ..CCATTGCAGGgtgagcctgg tccacgccagAATGGCCATT.. 103 ..TGAGGTCACGgtaagtgtca tttcccacagGTTTCCAACC.. 177 ..CCTCCACCTGgtgaggctta ccgtgtccagGATGTGGACA.. 132 ..CATCACCCTGgtgagggcag ttatccctagGACTGCGCCC.. 114 ..CTTTCTGGAGgtgaggatac gttcccttagGAGTACTCAG.. 99 ..CTCCACAGTGgtgagctgca tgctccccagATCCCAGTGA.. 126 ..CCTGTGGAAGgtgaggcttg gacatcctagTGTGGCTTCT.. 113 ..GGACTGTGGGgtaactgttg gtccttatagATGGGATTCTTCAAACGGGCGAAGCACCCCGAGGCCACCGTGCCCCAGTACCATG CGGTGAAGATTCCTCGGGAAGACCGACAGCAGTTCAAGGAGGAGAAGACGGGCACCATCCTGAGG AACAACTGGGGCAGCCCCCGGCGGGAGGGCCCGGATGCACACCCCATCCTGGCTGCTGACGGGCA TCCCGAGCTGGGCCCCGATGGGCATCCAGGGCCAGGCACCGCCTAGGTTCCCATGTCCCAGCCTG CGCTGTGGCTGCCCTCCATCCCTTCCCCAGAGATGGCTCCTTGGGATGAAGAGGGTAGAGTGGGC TGCTGGTGTCACATCAAGATTTGGCAGGATCGGCTTCCTCAGGGGCACAGACCTCTCCCACCCAC AAGAACTCCTCCCACCCAACTTCCCCTTAGAGTGCTGTGAGATGAGAGTGGGTAAATCAGGGACA GGGCCATGGGGTAGGGTGAGAAGGGCAGGGGTGTCCTGATGCAAAGGTGGGGAGAAGGATCCTAA TCCCTTCCTCTCCCATTCACCCTGTGTAACAGGACCCCAAGGACCTGCCTCCCCGGAAGTGCCTT AACCTAGAGGGTCGGGGAGGAGGTTGTGTCACTGACTCAAGGCTGCTCCTTCTCTAGTTTCCCCT CTCATCTGACCTTAGTTTGCTGCCATCAGTCTAGTGGTTTCGTGGTTTCGTCTATTTATTAAAAA ATATTTGAGAACAAAAAAAAAAAA
Intron size (bp) '4800* 99 '2500* 405 272 953 134 356 115 148 173 147 '700* 647 551 82 92 112 132 700 165 '2000* 540 105 '2000* 913
Note. Exon and intron sequences are in uppercase and lowercase letters, respectively. Splice site consensus nucleotides conforming to the ag/gt rule, transcription, and the stop codons corresponding to the A (58 bp after the beginning of exon 27) and B (228 bp after the beginning of exon 27) isoforms are in boldface. Nucleotide cDNA sequence numbering starts at the transcription codon. The underlined sequence corresponds to the sequence spliced in the rat C isoform. Intron sizes were determined by PCR (*) or sequencing. All exon sequences have been deposited on EMBL Database under Accession Nos. AJ228836 to AJ228862.
only two non silent polymorphisms in the coding sequence. The deduced amino acid sequence of ITGA7 showed high homology with human a3 and a6 integrin subunits (40, 41) as it has been previously reported for the mouse a7 integrin (25). This gene is about 22 500 bp and contains as in mouse 27 exons (20) (Mayer, personal communication). Several major isoforms, which may modify ligand affinity and specificity, cytoplasmic interactions and signaling, have been described. Their variations in the extracellular and the cytoplasmic domains correspond to the splicing of exons 5, 6, and 26. These exons present 100% homology with mouse (Mayer, personal communication). In addition, the 39 splice sites of hu-
man exons 5, 6, and 26 which are GTAGG, GTTTG, GTAAC, respectively, are mild consensus sites as in mouse (Mayer, personal communication), supporting the prediction of the same major isoforms in the two species. The integrin human a7 integrin isoforms are composed of an extracellular domain of 1042 aa for isoform X1, and of 1038 aa for isoform X2, a 23-amino-acid transmembrane domain and a short cytoplasmic domain of 57 amino acids for isoform A, or of 76 amino acids for isoform B. The two extracellular isoforms, designed X1 and X2, derive from the mutually exclusive splicing of exons 5 or 6 (11, 25). These two isoforms were present in equal
361
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. The human a7 integrin gene, ITGA, and its various splice variants. (A) Genomic organization of ITGA7 and the various splice variants. The rectangles denote exons, numbered from 1 to 27, and the thin lines represent either intronic sequences, or regions that are spliced in/out in the variants. Alternatively spliced exons. (B) Schematic representation of the divergent parts of the 8 transcriptional variants detected by RT-PCR in human myoblasts, myotubes, and skeletal or cardiac muscles. (C) Detection of splice variants by specific RT-PCR amplification of total RNAs extracted from various human tissues and separation by 2% agarose gel electrophoresis. The PCR products were stained with ethidium bromide. The M r marker is a 100-bp ladder. On the left: identification of the variants according to their sizes (the primers are given under Experimental Procedures). Each band has been cloned and sequenced.
amounts in mouse C2C12 myoblasts and adult heart, while the X2 variant was found nearly exclusively expressed in adult skeletal muscle (25). In human tissues the two variants were found in human cardiac, and fetal and adult skeletal muscles. Nevertheless, X1 variant was more abundant in myoblasts and fetal skeletal muscle and was downregulated during development. Two major cytoplasmic isoforms, designed A and B, have been described in rat, mouse and human (12, 26) and specific anti-integrin antibodies have been developed (12, 15, 18, 33). The A isoform, which corresponds to a normal splicing of exon 26 (38 amino acids) and formation of a stop codon after the 19 first amino acids encoded by exon 27, is muscle specific (12, 26). The B isoform is the longest isoform with a different C-terminal amino acid sequence. This is due to the splicing of exon 26 and to the formation of a stop codon after 76 amino acids in exon 27. This isoform is widely expressed not only in skeletal, cardiac and vascular smooth muscles but also in a number of other tissues, such as liver, spleen or brain (18, 26). We observed also that the isoform A is muscle specific, upregulated during muscle development and not expressed in human heart.
A third cytoplasmic variant, designed a7C, was isolated from a rat muscle cDNA library and detected in low amounts in poly(A) 1 RNA, prepared from rat myotubes (12) but we were not able to detect the corresponding variant in total RNA from human skeletal muscle even after two nested PCRs. If the same splicing as in rat is considered, no in frame stop codon can be formed due to a 8-bp insertion in the human compared to rat 39 part of exon 27 and in addition the potential splice donor site which is CGGGCG in human does not agree with consensus splice donor sequence. This variant was also neither found in mouse (Mayer, personal communication). Another isoform, designed as a7D, presenting a 225-bp in-frame deletion was recently reported by Leung et al. (28). This new isoform appears to be a partial deletion of exon 15 associated with the entire deletion of exon 16. Indeed, a 39 splice site consensus, GTCAG is present at the position 85 in exon 15. The use of this splice site should induce the deletion of the last 32 nucleotides of exon 15 and the 193 nucleotides of exon 16. We did detect this transcript in human cardiac and skeletal muscle but only at the adult stage of development.
362
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
We detected three additional minor transcripts, X3 to X5, in this study. The translation of the X4 and X5 transcripts would result in truncated proteins of 297 and 362 residues, respectively. These variants may be due to immature transcripts present at low level in nucleus or cytoplasm and amplified by PCR, without any physiological function. The X3 variant which corresponds to an in-frame association of exons 5 and 6 has been described in the human a6 subunit, and associated with b subunits forms functional laminin receptors (42). The identification of a new intragenic microsatellite marker in intron 1 of the gene and the refinement of its localization between two highly informative markers, D12S312 and D12S90, provides useful tools for linkage analysis. Intronic primers to amplify each exon and its intron boundaries should facilitate the identification of new ITGA7 mutations in congenital myopathy. So far the three first mutations identified in ITGA7 were mutations leading to premature stop codons (1) in patients with a complete absence of a7b1 integrin in muscle biopsies. Mutations in other integrin subunits cause human inherited diseases. A homozygous splicing mutation in the human a6 integrin gene, ITGA6, was identified in a patient with junctional epidermolysis bullosa with congenital duodenal atresia (43). Missense mutations occurring in functional sites may also cause diseases, as it has been shown for b3 integrin subunit in Glanzmann thrombasthenia, a recessive autosomal bleeding disorder characterized by abnormal platelet aggregation due to a qualitative or quantitative defect of the glycoprotein IIb–IIIa complex (integrin aIIbb3) (44). Identification of missense mutations in ITGA7 will help to define the ligand binding sites and the functional sites of this protein, which plays an important role in human skeletal muscle development and integrity. ACKNOWLEDGMENTS We thank Marc Fiszman for helpful discussion, Martine Verdie`reSahuque´ for her help in preparing skeletal muscle cells, and Delphine Muselet for technical assistance. This work was supported by the Institut National de la Sante´ et de la Recherche Me´dicale and the Association Franc¸aise contre les Myopathies.
REFERENCES 1. Hayashi, Y. K., Chou, F.-L., Engvall, E., Ogawa, M., Mastuda, C., Hirabayashi, S., Yokochi, K., Ziober, B. L., Kramer, R. H., Kaufman, S. J., Ozawa, E., Goto, Y., Nonaka, I., Tsukahara, T., Wang, J. Z., Hoffman, E. P., and Arahata, K. (1998) Nature Genet. 19, 94 –97. 2. Hynes, R. O., and Lander, A. D. (1992) Cell 68, 303–322. 3. Gullberg, D., Velling, T., Lohikangas, L., and Tiger, C. F. (1998) Front Biosci. 15, D1039 –D1050. 4. Ekblom, P., and Timpl, R. (1996) Curr. Opin. Cell Biol. 8, 599 – 601.
5. Ruoslahti, E. (1988) Annu. Rev. Biochem. 57, 375– 413. 6. Camper, L., Hellman, U., and Lundgren-Akerlund, E. (1998) J. Biol. Chem. 273, 20383–20389. 7. Horwitz, A. F. (1997) Sci. Am., 46 –53. 8. Timpl, R., and Brown, J. C. (1996) Bioessays 18, 123–132. 9. Kramer, R. H., Vu, M. P., Cheng, Y. F., Ramos, D. M., Timpl, R., and Waleh, N. (1991) Cell Regul. 2, 805– 817. 10. von der Mark, H., Du¨rr, J., Sonnenberg, A., von der Mark, K., Deutzmann, R., and Goodman, S. L. (1991) J. Biol. Chem. 266, 23593–23601. 11. Song, W. K., Wang, W., Foster, R. F., Bielser, D. A., and Kaufman, S. J. (1992) J. Cell Biol. 117, 643– 657. 12. Song, W. K., Wang, W., Sato, H., Bielser, D. A., and Kaufman, S. J. (1993) J. Cell. Sci. 106, 1139 –1152. 13. Yao, C. C., Ziober, B. L., Squillace, R. M., and Kramer, R. H. (1996) J. Biol. Chem. 271, 25598 –25603. 14. Hodges, B. L., Hayashi, Y. K., Nonaka, I., Wang, W., Arahata, K., and Kaufman, S. J. (1997) J. Cell Sci. 110, 2873–2881. 15. Vachon, P. H., Xu, H., Liu, L., Loechel, F., Hayashi, Y., Arahata, K., Reed, J., Wewer, U. M., and Engvall, E. (1997) J. Clin. Invest. 100, 1870 –1881. 16. Ervasti, J. M., and Campbell, K. P. (1993) J. Cell Biol. 122, 809 – 823. 17. Campbell, K. (1995) Cell 80, 675– 679. 18. Velling, T., Collo, G., Sorokin, L., Durbeej, M., Zhang, H., and Gullberg, D. (1996) Dev. Dyn. 207, 355–371. 19. Helbling-Leclerc, A., Zhang, X., Topaloglu, H., Cruaud, C., Tesson, F., Weissenbach, J., Tome´, F. M. S., Schwartz, K., Fardeau, M., Tryggvason, K., and Guicheney, P. (1995) Nat. Genet. 11, 216 –218. 20. Mayer, U., Saher, G., Fassler, R., Bornemann, A., Echtermeyer, F., von der Mark, H., Miosge, N., Poschl, E., and von der Mark, K. (1997) Nat. Genet. 17, 318 –323. 21. Wang, W., Wu, W., Desai, T., Ward, D. C., and Kaufman, S. J. (1995) Genomics 26, 568 –570. 22. Sosnoski, D. M., Emanuel, B. S., Hawkins, A. L., van Tuinen, P., Ledbetter, D. H., Nussbaum, R. L., Kaos, F. T., Schwartz, E., Phillips, D., Bennett, J. S., et al. (1988) J. Clin. Invest. 81, 1993–1998. 23. Baker, E. (1992) Mamm. Genome 2, 272–273. 24. Krissansen, G. W., Yuan, Q., Jenkins, D., Jiang, W., Rooke, L., Spurr, N. K., Eccles, M., Leung, E., and Watson, J. D. (1992) Immunogenetics 35, 58 – 61. 25. Ziober, B. L., Vu, M. P., Waleh, N., Crawford, J., Lin, C. S., and Kramer, R. H. (1993) J. Biol. Chem. 268, 26773–26783. 26. Collo, G., Starr, L., and Quaranta, V. (1993) J. Biol. Chem. 268, 19019 –19024. 27. Hodges, B. L., and Kaufman, S. J. (1996) Basic Appl. Myologie 6, 437– 446. 28. Leung, E., Pheng Lim, S., Berg, R., Yang, Y., Ni, J., Wang, S. X., and Krissansen, G. W. (1998) Biochem. Biophys. Res. Commun. 243, 317–325. 29. Balzac, F., Belkin, A. M., Koteliansky, V. E., Balabanov, Y. V., Altruda, F., Silengo, L., and Tarone, G. (1994) J. Cell Biol. 121, 171–178. 30. Zhidkova, B. L., Belkin, A. M., and Mane, R. (1995) Biochem. Biophys. Res. Commun. 214, 279 –285. 31. van der Flier, A., Kuikman, I., Baudoin, C., van der Neut, R., and Sonnenberg, A. (1995) FEBS Lett. 369, 340 –344. 32. Belkin, A. M., Zhidkova, N. I., Balzac, F., Altruda, F., Tomatis, D., Maier, A., Tarone, G., Koteliansky, V. E., and Burridge, K. (1996) J. Cell Biol. 132, 211–226.
363
Vol. 260, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
33. Martin, P. T., Kaufman, S. J., Kramer, R. H., and Sanes, J. R. (1996) Dev. Biol. 174, 125–139. 34. Vignal, A., Gyapay, G., Hazan, J., Nguyen, S., Dupraz, C., Cheron, N., Becuwe, N., Tranchant, M., and Weissenbach, J. (1993) in Non-radioactive Multiplex Procedure for Genotyping of Microsatellite Markers (Series Ed.), Vol. 211–221, Academic Press, New York. 35. Slonim, D., Kruglyak, L., Stein, L., and Lander, E. (1997) J. Comput. Biol. 4, 487–504. 36. Askanas, V., and Engel, W. K. (1975) Neurology 25, 58 – 67. 37. Verdie`re-Sahuque´, M., Akaaboune, M., Lachkar, S., Festoff, B. W., Jandrot-Perrus, M., Garcia, L., Barlovatz-Meimon, G., and Hantai, D. (1996) Exp. Cell Res. 222, 70 –76. 38. Shapiro, M. B., and Senapathy, P. (1987) Nucleic Acids Res. 15, 7155–7174.
39. Dib, C., Faure´, S., Fizames, C., Samson, D., Drouot, N., Vignal, A., Millasseau, P., Seboun, E., Marc, S., Lathrop, M., Gyapay, G., Morissette, J., and Weissenbach, J. (1996) Nature 380, 152– 154. 40. Takada, Y., Murphy, E., Pil, P., Chen, C., Ginsberg, M. H., and Hemler, M. E. (1991) J. Cell Biol. 115, 257–266. 41. Hogervorst, F., Kuikman, I., van Kessel, A. G., and Sonnenberg, A. (1991) Eur. J. Biochem. 199, 425– 433. 42. Delwel, G. O., Kuikman, I., and Sonnenberg, A. (1995) Cell Adhes. Commun. 3, 143–161. 43. Pulkkinen, L., Kurtz, K., Xu, Y., Bruckner-Tuderman, L., and Uitto, J. (1997) Lab. Invest. 76, 823– 833. 44. Bajt, M. L., Ginsberg, M. H., Frelinger AL, d., Berndt, M. C., and Loftus, J. C. (1992) J. Biol. Chem. 267, 3789 –3794.
364