THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 267, No. 18, Issue of June 25, pp. 12845-12850,1992 Printed in U.S.A.

Chicken Vinculin and meta-Vinculin Are Derived from a Single Gene by Alternative Splicing of a 207-Base Pair Exon Unique to meta-Vinculin”

(Received for publication, February 4, 1992)

Barry J. ByrneS, Yolanda J. Kaczorowski, Michelle D. Coutue, and Susan W. Craig7 From the Departments of Biological Chemistry and $Pediatrics, The Johns Hopkins School of Medicine, Baltimore, Maryland 21205

meta-Vinculin and vinculin are closely related pro- teins that are cytoplasmic components of microfila- ment-associated cell junctions. This report describes the structural relationship between these two proteins and the genetic basis for tissue-specific expression of meta-vinculin. Analysis of genomic DNA coding for amino acids 676-1066 of vinculin revealed 9 exons spanning an 11.7-kilobase pair region of the genome. In the 4 kilobase pairs of intervening sequence that separates vinculin exons E896-E915 and E916-E984, there is an open reading frame that predicts a sequence homologous to the 68-amino acid peptide specific to porcine meta-vinculin (Gimona, M., Small, J. V., Moeremans, M., Van Damme, J., Puype, M., and Van- dekerckhove, J. (1988) EMBO J. 7, 2329-2334). Analysis of the corresponding cDNA established that chicken meta-vinculin contains a 69-amino acid inser- tion between residues 915 and 916 of vinculin and that there are no other amino acid sequence differences between chicken vinculin and meta-vinculin. Muscle- specific expression of meta-vinculin occurs by alter- native splicing of a transcript produced from a single gene because: all 20 genomic isolates that contain the 3’ vinculin exons, also contain the meta-vinculin-spe- cific exon; Southern blots performed at high stringency with exon-specific probes indicate the presence of a single gene; and the 3”untranslated sequences of vin- culin and meta-vinculin cDNAs are identical.

Vinculin (1) and metu-vinculin (2, 3) are proteins that localize in the cytoplasmic plaque of microfilament-associated cel1:cell and cel1:substrate junctions (4,5). In non-muscle cells, these load-bearing structures include cell-to-extracellular ma- trix junctions, such as focal adherens, and cell-to-cell junc-

* This work was supported by the Harriet Lane Fellowship, Na- tional Institute of Child Health and Human Development Grant 5P30-HD27799-02 (to B. J. B.), Grant GM 41605 from the National Institutes of Health, Grant 0973 from the American Heart Associa- tion, and by grants from the W. W. Smith Foundation and the Muscular Dystrophy Association (to S. W. C.). The costs of publica- tion of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertise- rnent” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide seguence(s) reported in this paper has been submitted

M87837. to the GenBankTM/EMBL Data Bank with accession number(s)

I Present address: Dept. of Genetics, Washington University School of Medicine, 4566 Scott Ave., St. Louis, MO 63110.

11 To whom correspondence should be addressed Dept. of Biological Chemistry, The Johns Hopkins School of Medicine, Baltimore, MD 21205. Fax: 410-55-5759.

tions of the zonulae adherens type; in muscle cells, they are represented by costameres, myotendinous junctions, and fas- ciae adherens.

Although the function of vinculin is unknown, the protein is essential for development of Cuenorhubditis eleguns past the 2-fold stage and for movement of the body wall muscles (6). Recent observations of rat cardiocytes cultured on a silastic rubber substrate provide evidence that vinculin plays a role in assembly or function of transmembrane, force- transducing structures. Cardiocytes assemble myofibrils and beat rhythmically for several days before they acquire the ability to contract the silastic membrane. Development of this latter capacity is correlated with prominent localization of vinculin at each pleat in the silastic substrate (7). The mor- phology of these contact points, their relationship to the underlying myofibrils, and the presence of vinculin closely resembles the costamere structure of muscle in situ (8-10).

Vinculin is present in many cell types, including muscle where it is constitutively expressed. In contrast, rnetu-vinculin has been found only in muscle cells (3, 12), where it is expressed in a developmentally regulated manner (13) and in platelets (11). The probable role of vinculin in transduction of force across cellular membranes suggests that metu-vincu- lin may play a specialized role, perhaps even an essential role, in transcellular force transduction in muscle. For this reason we have initiated studies to describe structural and functional relationships between vinculin and metu-vinculin (14).

Currently it is known that metu-vinculin is antigenically similar to vinculin (2, 3) but larger on SDS’ gels. Vinculin has an electrophoretic mobility consistent with its calculated molecular mass from cDNA of 117 kDa (15, 16), whereas metu-vinculin has an apparent molecular mass 17 kDa greater (3, 17). On isoelectric focusing gels, porcine and avian metu- vinculin are more acidic than vinculin. In addition, chicken gizzard metu-vinculin has different solubility properties than vinculin, and an 8-10-fold higher specific activity of endoge- nous Ser/Thr phosphorylation (17).’ The subcellular localiza- tion of vinculin and rnetu-vinculin, evaluated by microinjec- tion of fluorescently labeled proteins (13) and by immunoflu- orescence localization with antibody specific to human metu- vinculin (18), is coincident at 0.2-pm resolution. Vinculin (19, 21) and metu-vinculin (17, 21) bind to talin, and vinculin binds to acidic phospholipids (22, 23) as does metu-~inculin.~

Definition of the structural differences between vinculin and metu-vinculin will provide a basis for assessing their functional properties. Early studies addressing the relation-

’ The abbreviations used are: SDS, sodium dodecyl sulfate; kb,

* C. DeBerry, unpublished data. R. Johnson, unpublished data.

kilobase pair(s); bp, base pair(s); PCR, polymerase chain reaction.

12845

12846 Sequence and Genetic Origin of Chicken meta- Vinculin

ship between the two proteins showed that they have distin- guishable although nearly identical peptide maps (3, 17), are synthesized from separate mRNAs (3), and do not have a precursor-product relationship (17). Recently, Gimona et al. (24) isolated and sequenced a 68-residue meta-vinculin-spe- cific peptide from porcine meta-vinculin. Based on sequence homology between residues at the termini of the porcine meta- vinculin-specific peptide with the sequence of chicken vincu- lin, the meta-vinculin peptide was positioned within the se- quence of chicken vinculin between residues 912 and 913. Because the meta-vinculin peptide accounts for only 7.6 kDa of the apparent mass difference between vinculin and meta- vinculin, the question remained whether the meta-vinculin peptide represented the full extent of the difference between the two proteins.

The present study was undertaken to define the nature and location of the primary sequence differences between chicken vinculin and meta-vinculin and to establish the genetic basis for tissue-specific expression of meta-vinculin.

EXPERIMENTAL PROCEDURES

Reagent~-[a-~*P]dCTP, [-y-32P]ATP, and [a-32P]CTP were ob- tained from Du Pont-New England Nuclear. Restriction endonucle- ases were supplied by New England Biolabs and Bethesda Research Laboratories. Reverse transcriptase was obtained from Life Science Biologicals (Tampa, Florida). RNAse inhibitor was from Bohringer Mannheim. Thermus aquaticus (Taq) polymerase was obtained from Perkin-Elmer Cetus. Bluescript vectors were purchased from Stra- tagene (San Diego, CA); pCRlOOO was from InVitrogen, Inc.

SDS-Polyacrylamide Gel Electrophoresis and Western Blotting- Vinculin and meta-vinculin were separated by SDS-gel electropho- resis (25) and transferred by electrophoresis (26) to a nitrocellulose membrane. Proteins were detected with an affinity-purified poly- clonal antibody to vinculin or an affinity-purified polyclonal antibody to P2, a synthetic peptide with the sequence (K)EFSLPSDIEDDYEP, which corresponds to a portion of the cDNA-derived chicken meta- vinculin-specific sequence.

Isolation and Analysis of Vinculin Genomic Clones-A chicken genomic library (gift of D. Engel, Department of Biochemistry, North- western University, Evanston, IL) was screened by hybridization with cDNA probes derived from pCE45V and pCE68V (27). Library screening, isolation of phage DNA, digestion of DNA, gel electropho- resis of DNA or RNA, transfer of gels to nitrocellulose, and hybridi- zation were performed as described (28). Hybridization washes were done at high stringency as follows: 6 X SSC for two washes at room temperature, 2 X SSC for two washes at 37 "C and 0.1 X SSC for two washes at 65 'C. DNA fragments used as probes were radiolabeled by the random prime method (29). SalI and EcoRI subfragments were subcloned into the Bluescript vector, and deletions were generated as described (Bluescript Exo/Mung Bean Manual, Stratagene Cloning Systems, San Diego, CA). Double-stranded DNA sequencing was done with Sequenase (U. S. Biochemical Corp.) by the method of Chen (30). DNA and amino acid sequences were analyzed using the Uni- versity of Wisconsin GCG analysis programs. Plasmid DNA was purified by CsCl gradient centrifugation or by miniprep lysis (28). A combination of gene-specific primers and nested deletion sets were used to obtain 11.6 kb of sequence data of the total 11.7 kb expected on the basis of subfragment electrophoretic mobility.

cDNA Cloning of meta- Vinculin-A chicken gizzard cDNA library in Agt-11 was obtained from Clonetech. Plaques were screened with a 150-bp EcoRI-SmaI fragment of the meta-vinculin-specific region isolated from genomic clone VGen-2 and one positive isolate, h MV Q, was identified and plaque-purified. This clone is a partial cDNA which begins with an EcoRI site within the meta-vinculin exon followed by sequence coding for vinculin amino acids 915-1066, as well as 1.4 kb of the 3'-untranslated region. The 5' portion of the meta-vinculin cDNA not represented in A MV Q was isolated by amplification of reverse transcribed mRNA from 18-day embryonic gizzard (31). Oligo(dT) was used to prepare cDNA which was ampli- fied in a PCR reaction with vinculin primers V12 (bp 2498-2522) and V14 (bp 2910-2886) which flank the meta-vinculin exon. Two prod- ucts were obtained which differed by approximately 200 base pairs. Sequence analysis of the larger product confirmed this to be a cDNA which contained 207 base pairs of additional sequence coding for the

additional 69 amino acids found in meta-vinculin. A meta-vinculin- specific primer (MV6, mv-bp 141-124) was used to generate a second pool of cDNA which served as a template for amplification of the remaining meta-vinculin cDNA. A second meta-vinculin-specific primer (MV9, mv-bp 72-55) (Fig. 3) was used in combination with a vinculin primer V54 (bp 25-390) to amplify a 3-kb segment of the meta-vinculin cDNA. To obtain the final segment of the message, oligonucleotide V50 (bp -50 to -32 in the 5'-untranslated region (16)) was used in combination with V34 (bp 454-471) to generate the 5' most fragment in a PCR reaction with the MV6-primed reverse transcriptase reaction products serving as template. These PCR prod- ucts were sequenced following subcloning into pCRlOOO (InVitrogen, Inc.) or were sequenced directly on an Applied Biosystems Inc. 373A DNA sequencing system.

RESULTS AND DISCUSSION

Identification and Characterization of Genomic Clones for Vinculin-A chicken erythrocyte genomic library in EMBL 3 was screened with a pool of vinculin cDNA clones represent- ing bp 175 - 2645 of the 6.2-kb vinculin message (15, 27). Twenty-seven isolates were plaque-purified and characterized by hybridization with individual vinculin cDNA subfrag- ments. Twenty of these clones hybridized to pCEV 3'8-3 (bp 2649 -3227), which codes for vinculin amino acids 877-1066 (Fig. 1). These clones could contain the putative exon for a meta-vinculin-specific peptide analogous to the porcine meta- vinculin peptide that is known from peptide mapping studies to be present in this region (24). Therefore, the isolate with the largest genomic insert, Vgen-2, was chosen for analysis of exonlintron structure by restriction mapping and DNA se- quencing.

Determination of the ZntronlExon Structure of Genomic DNA Coding for the Carboxyl-terminal 370 Amino Acids of Vinculin-Vgen-2, shown schematically in Fig. 2, is comprised of an 11.7-kb genomic SalI fragment and contains five con- secutive EcoRI subfragments of 2.8, 0.4, 2.4, 0.6, and 5.5 kb. The 5' end of the gene is joined to the left arm of EMBL 3. From nucleotide sequencing, we determined that amino acids 676-1066, found both in chicken embryo (15) and chicken fibroblast (16, 32) vinculin, are encoded by 9 exons ranging in size from 33 to 336 bp (Table I). The 5.5-kb EcoRI subfrag- ment of Vgen-2 also encodes 1.4 kb of uninterrupted sequence which matches the 3'-untranslated region of the mRNA from vinculin (32) and meta-vinculin (this report).

Genomic Clone No. 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

mi:; Rv t t t t t t

t t t t t t t

I f t t t t t t t t t t t t t t t t t t t t

1 5 t t t t t t t t t t t t t t t t t t t t

3'

FIG. 1. Hybridization characteristics of vinculin-positive genomic isolates. Two million plaques from a chicken erythrocyte genomic library were screened with a pool of vinculin cDNA probes, corresponding to bp 175-2645 of the 6.2-kb vinculin message. Twenty- seven positive isolates were plaque-purified and hybridized with the probes indicated at the left of the schematic. Homology with a probe is indicated by +. Two distinct groups of clones are evident; those with homology to the sequences 5' to the proline region, and those with homology to the vinculin tail domain which is in the carboxyl- terminal one-third of the molecule. All isolates in the second group were positive for both vinculin and meta-vinculin coding region probes.

Sequence and Genetic Origin of Chicken meta- Vinculin 12847 sol I R1 R I R1 R I s o l I

L I n I1 I

8 0 11.7 kb

5‘ I+ 3 3’ ~*t;idmwaJl;

w: 676 1135

FIG. 2. Comparison of the physical maps of genomic clone Vgen-2 and metu-vinculin mRNA. The 11.7-kb genomic DNA Sal1 has five EcoRI subfragments of 2.8, 0.6, 2.4,0.6, and 5.5 kb. The X left arm of EMBL 3 is joined to the 2.8-kb SaZI-EcoRI fragment. Positions of subfragments used for vinculin and meta-vinculin exon- specific probes are indicated by shaded boxes beneath the correspond- ing exon. The 150-bp EcoRI-SmaI fragment from the 3’ portion of the 0.6-kb EcoRI fragment is the meta-vinculin specific probe; it includes 19 bp of intron sequence at its 3‘ end. A vinculin exon probe was derived from the E985-E1018 sequence. The 6.2-kb meta-vinculin mRNA is shown with the 5’- and 3”untranslated regions as thin lines and the coding region as a shaded box. The clone Vgen-2 encodes vinculin amino acids 676-1066,69 residues of meta-vinculin-specific amino acid sequence and 1.4 kb of the 3”untranslated region.

TABLE I Location and size of exons in the chicken vinculin and meta-uinculin

gene Designation of the exon positions in Vgen-2 is based on sequence

comparison to vinculin and meta-vinculin cDNA. Each exon is iden- tified by the number of the first and last amino acid that it encodes. The contribution of each exon to structural domains is shown in the last column.

Exon position No’ Of amino Corresponding acids domain

676-710 35 Head 711-812 112 Head 813-842 30 Head 843-853 11 Head/proline 854-895 42 Proline 896-915 20 Proline

mvl-mv69 69 meta-Vinculin 916-984 69 Tail 985-1018 34 Tail

1019-1066 48 Tail

Identification and Characterization of the Exon Encoding Chicken meta- Vinculin-specific Sequence-We observed that amino acids 915 and 916 are encoded in separate exons, E896- E915 and E916-E984, which are separated by 4 kb of inter- vening sequence (Fig. 2; Table I). This exon structure would permit generation of meta-vinculin by alternative splicing if the intervening 4 kb contained a meta-vinculin specific exon(s). Analysis of the open reading frames in this region revealed homology between the translated product of one open reading frame (Fig. 3) and the 68-amino acid sequence specific to porcine meta-vinculin (24). Direct evidence that this open reading frame encodes sequences present in chicken meta- vinculin mRNA was obtained by Northern blot analysis of poly(A+) mRNA from 8-day chick embryo and adult gizzard using a riboprobe generated from the meta-vinculin-specific exon (Fig. 4). This probe detects a 6.2-kb mRNA in the embryo and gizzard. Duplicate lanes probed with a segment of vinculin cDNA in common with meta-vinculin cDNA also detects a 6.2-kb message. At the protein level, antibody raised to a portion of the meta-vinculin-specific peptide sequence, (K)EFSLPSDIEDDYEP, recognizes only a 134-kDa meta- vinculin isoform on Western blots (Fig. 5). This result shows that there is not a 116-kDa isoform that contains the meta- vinculin-specific sequence, but lacks some other exon.

Translated Gemmlc ORF IMCTGTCCCCTGTTTCTCCACCCTTGCTGCCACCTGCMAG L S P V S P P L L P P A K B

N m N L L S L P R G L I C AAGTGGTCTAGTMGGTACTCCTGAGTCTCCCCCGGGWCTCTGT

K Y S S K

GTGCATTGGGCA~GGMCGCTGACACATTTGCATCACTCTT~~ V H Y A F G T L T H L H H S

FIG. 3. DNA sequence of a genomic open reading frame from Vgen-2 and comparison of the translated protein se- quence to the porcine metu-vinculin peptide. Genomic DNA sequence is presented in the first line. The translated product of the chicken genomic DNA sequence is shown in the second line with the boxed region representing the corresponding segment of the translated cDNA sequence. The last line shows the peptide sequence of porcine meta-vinculin determined by Gimona et al. (24). Vertical lines repre- sent identity, and dots represent conservative amino acid substitu- tions. The genomic DNA open reading frame from 1 to 327 bp and cDNA sequence has been submitted to GenBank with the accession number M87837.

9.5 -

7.5 -

4.4 - 2.4 -

Vlnclllln Meta-vhculh Robe Robe

FIG. 4. Northern blot of chicken embryo and gizzard RNA. Poly(A+) RNA was isolated from an 8-day embryo and adult gizzard and fractionated on a 0.8% agarose-formaldehyde gel. Duplicate lanes were run and blotted onto nitrocellulose. Riboprobes to the meta- vinculin exon (EcoRI-Sa& 150 bp) or vinculin E985-El018 were used to detect the 6.2-kb mRNA in both embryo and gizzard.

To determine the portion of the chicken meta-vinculin open reading frame that is expressed in mRNA, a cDNA for meta- vinculin was isolated (Fig. 6; see “Experimental Procedures”) and sequenced. Of the 107 amino acids encoded by the ge- nomic open reading frame (Fig. 3), 69 are present in meta- vinculin cDNA (Fig. 3, boxed region of the amino acid se- quence). Of particular interest is a set of repeated peptides (boxed segments, Fig. 7) which occur at the carboxyl-terminal end of the vinculin exon preceding the meta-vinculin insertion and at the end of the meta-vinculin-specific sequence. These repeats are an indication that a duplication event has occurred with the vinculin exon E896-E915 in higher eukaryotes, since these sequences are not represented in the deb-1 gene of C. elegans, which is homologous to vinculin (33).

Comparison of the Avian and Porcine meta- Vinculin-specific Sequence-The homology between porcine and avian meta- vinculin-specific peptides is greatest in the carboxyl-terminal portion where 37 of the last 40 amino acids are identical. In the amino-terminal one-third, however, identity drops to 44%)

12848 Sequence and Genetic Origin of Chicken meta- Vinculin

I j

S V MV V MV

200 -

116 - 97 -

66 -

45 -

Anti-vincuiin Anti-MV peptide FIG. 5. Western blot of vinculin and meta-vinculin with

anti-vinculin or anti-meta-vinculin-specific peptide. Purified vinculin and meta-vinculin were separated on a 7% polyacrylamide, SDS gel and transferred to nitrocellulose in duplicate lanes. Proteins were detected with either affinity-purified polyclonal anti-vinculin antibody or affinity purified polyclonal antibody raised to a portion of the meta-vinculin-specific peptide sequence, (K)EFSLPSD- IEDDYEP. The meta-vinculin-specific antibody detects only the 134- kDa protein.

5.- r;;h I 1 3 ’

MV - €I - V12iV14 m V54lMW I

V50iV34 -

1.0 kb

FIG. 6. Isolation of meta-vinculin cDNA. The complete meta- vinculin cDNA is shown in the first line. Clone X MV-Q was isolated from a Xgt -11 library and is truncated at an EcoRI site within the meta-vinculin-specific exon. PCR amplification of reverse-tran- scribed mRNA was used to obtain the remaining clones. The V12/ V14 product was obtained from mRNA reverse transcribed with oligo(dT). Products V54/MV9 and V50/V34 were generated from mRNA reverse transcribed with a meta-vinculin-specific primer, as detailed under “Experimental Procedures.” Together, these four over- lapping clones provide the complete coding region sequence of the meta-vinculin cDNA.

.,.e I!‘

* ~ . ~ Q P V T V I N ~ A A ~ A G V D I D E E D D A D V B Y

E F S g P S D I E D D Y E P E L L L M P T N C i . , 7

P V ~ I L A ~ ~ s ~ H ~ R ~ ~ [ K I S S ~ ~ I * * * K ~ ?

FIG. 7. Amino acid sequence of the derived chicken meta- vinculin cDNA sequence at the insertion site in vinculin. Vinculin protein sequence is indicated by the small letters up to amino acid 915 and resuming at amino acid 916. At the insertion of the meta-vinculin exon, the sequence unique to meta-vinculin is shown in large letters. The first line of 23 amino acids is a segment which is divergent from the porcine sequence. The second line is the portion which is unique to meta-vinculin and nearly identical to the porcine sequence. The third line of sequence contains a series of peptides (boxed) which are also found in the carboxyl-terminal portion of the adjacent vinculin exon.

while the overall similarity is 72%. Three regions of the meta- vinculin-specific exon are discernible based on their degree of conservation and relationship to vinculin. The amino-termi- nal one-third is divergent between the porcine and avian peptides and unique to meta-vinculin. The carboxyl-terminal one-third is completely conserved between the two species and is similar to the carboxyl-terminal end of the preceding vinculin exon by a set of repeated sequences (Fig. 6). The middle third of the meta-vinculin segment is highly conserved and unique to meta-vinculin and is likely therefore to encode a specialized function for meta-vinculin.

The chicken meta-vinculin-specific sequence is character- ized by a calculated isoelectric point of 5.37. On isoelectric focusing gels, meta-vinculin from chicken gizzard has one prominent isoform of PI 5.9 and a minor isoform of PI 6.2. Both of these are more acidic than the gizzard vinculin isoforms which range from PI 6.3 to 6.5 (34) (Fig. 8).

Determination of the Complete meta-Vinculin Primary Structure from cDNA-The apparent molecular mass differ- ence between vinculin and meta-vinculin on SDS gels is 17 kDa. The 69-amino acid sequence specific to chicken meta- vinculin accounts for only 7.6 kDa of this mass difference. To determine whether there are additional sequence differences, a full-length cDNA for meta-vinculin was obtained by a PCR strategy (Fig. 7 and “Experimental Procedures”), subcloned, and sequenced with oligonucleotide primers. No other se- quence differences between vinculin and meta-vinculin were found. Consequently, the 17-kDa shift in apparent molecular mass must in part reflect a conformational difference in meta- vinculin as compared with vinculin. In rotary shadowed im- ages both vinculin and meta-vinculin show a globular head domain of 8 nm and an extended tail domain (35). The meta- vinculin tail region measures 21.5 uersus 19.4 nm for vinculin (35); this difference may contribute to the slower mobility of meta-vinculin on SDS gels.

The existing data do not rule out the possibility that there are additional primary sequence isoforms of vinculin and meta-vinculin. The recent characterization of a cDNA having an in-frame deletion coding for amino acids 167-207, the putative talin-binding region, suggests that there may be

23 -

9.4 -

6.6 -

4.4 -

2.3 - 2.0 -

0.6 - I

Mculin Probe MetavlncuAn Probe

FIG. 8. Southern blot of chicken genomic DNA. High molec- ular weight cellular DNA (8 mg) from a stage 18 chick embryo was digested with Sal1 and redigested with the enzymes listed above each lane and run on a 0.8% agarose gel. The gel was transferred to nitrocellulose and hybridized with the indicated probes which contain coding region nucleotides within an exon. The vinculin probe spans nucleotides 2755-2910 ( VZNC probe in Fig. 2) and the meta-vinculin probe spans meta-vinculin nucleotides 75-207 plus 19 bp of adjacent intervening sequence (MV probe in Fig. 2). Markers (M) are the Hind111 digest of X DNA.

Sequence and Genetic Origin of Chicken meta- Vinculin 12849

additional sequence isoforms (36). Also, the structural basis of the multiple isoelectric forms of vinculin and meta-vinculin (5, 34,37, 38) remains to be established.

Evidence That a Single Gene Codes for Both Vinculin and meta-Vinculin-The second goal of this study was to deter- mine whether vinculin and meta-vinculin are encoded in separate genes or are alternatively spliced products of a single gene. Clear information on this point is required prior to studies of the mechanism of muscle-specific expression of meta-vinculin. Earlier analyses of Southern blots probed with segments of vinculin cDNA produced variable results, some- times identifying multiple bands (27, 32).

We have obtained three independent sets of data to address the genetic origin of meta-vinculin. The first evidence to suggest vinculin and meta-vinculin are encoded by a single gene is provided by the hybridization characteristics of 27 vinculin-positive isolates from the EMBL 3 library. All 20 of the isolates that contain sequence homologous to the tail domain of vinculin, detected by probe pCEV3 ‘83 (15), are also recognized by the meta-vinculin-specific probe (Fig. 1). These genetic data make it improbable that separate genes code for vinculin and meta-vinculin.

Second, unambiguous Southern blots were obtained using probes derived from sequence-defined exons lacking the re- striction enzyme sites used to digest the genomic DNA (Fig. 8). Chicken genomic DNA was digested with Sal1 followed by a second digestion with either BamHI, PstI, PuuII, or EcoRI. A duplicate set of lanes was run on the same gel and then probed with either vinculin exon E916-E984, or the meta- vinculin-specific probe, an EcoRI-SmaI fragment from Vgen- 2 (Fig. 2). Single bands were detected in all four digests, indicating that a single gene is recognized by both probes. In the BamHI, PstI, and PuuII digests, both probes recognized the same size fragments, the sizes of which match those from corresponding digests of genomic clone Vgen-2 (data not shown). Results of the EcoRI digest demonstrate that the vinculin and meta-vinculin probes reside on separate frag- ments as predicted from the map in Fig. 2. The E916-E984 probe identifies a 5.5-kb EcoRI-Sal1 fragment, and the meta- vinculin probe identifies the smaller 0.6-kb EcoRI-EcoRI frag- ment. A restriction fragment length polymorphism in the 5.5- kb EcoRI-Sal1 fragment produced a doublet of 5.5 and 4 kb in several of 10 additional chicken genomic DNAs analyzed (data not shown). Heterozygotes at this locus may also have

NH 2 COOH

actin paxiilin vincuiin

Globular Head Domain Baric Tail Domain

FIG. 9. Schematic of meta-vinculin domain structure. The meta-vinculin protein is shown schematically with relevant structural features indicated along the 1135 amino acids. The globular head domain is separated from the basic tail domain (15) by the proline- rich region and the meta-vinculin-specific segment (black rectangle). The meta-vinculin head domain contains several structural elements and putative functional domains based on sequence identity to the vinculin head segment. A segment consisting of three repeats of a 110 amino acids is located between amino acid 259 and 589 (16). A talin- binding domain has been proposed to reside between amino acids 167-207 (36) and is indicated by a circle. The region implicated in the binding of vinculin to actin is located in the carboxyl-terminal portion of the head domain between amino acids 587 and 851 and is indicated by a line below the area (39). Other possible functional domains, characterized in vinculin, are located in the tail segment and include vinculin binding or self-association (35) and paxillin binding (40). Intact vinculin has been found to bind to the 27-kDa thermolysin fragment of a-actinin (41).

generated multiple bands in earlier experiments. This data strongly supports the presence of a single gene for both vinculin and meta-vinculin, since the coding segments physi- cally map to the same DNA segment.

Last, the 1.4 kb of the 3”untranslated sequence of meta- vinculin cDNA is identical to the 3”untranslated sequence of vinculin cDNA (data not shown), and this sequence is encoded as a single segment in Vgen-2. In general, the 3“untranslated portions of mRNA are not highly conserved between related genes, suggesting that vinculin and meta-vinculin derive from the same gene.

Together, these three independent sets of data establish that the tissue-specific expression of meta-vinculin occurs by alternative splicing of the mRNA product of a single gene that encodes the information for both vinculin and meta- vinculin; this gene is designated vincl. Based on the sequence identity of vinculin and meta-vinculin, we have summarized the structural and putative functional domains of meta-vin- culin in Fig. 9. A key challenge for future experiments is to determine the role(s) that vinculin and meta-vinculin have in assembly or function of the transmembrane, force-transduc- ing structures in muscle cells and other cell types.

Acknowledgments-We acknowledge the expert technical assist- ance of Nancy Larson-Kelly on a portion of this study. Special thanks are given to Candy DeBerry and Robert Johnson for help with editing and proof reading the manuscript.

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.

REFERENCES

Geiger, B. (1979) Cell 18, 193-205 Siliciano, J. D., and Craig, S. W. (1982) Nature 300,533-535 Feramisco, J. R., Smart, J. E., Burridge, K., Helfman, D. M., and

Geiger, B., Tokuyasu, K. T., Dutton, A. H., and Singer, S. J.

Geiger, B. (1982) J. Mol. Biol. 159, 685-701 Barstead, R. J. (1991) J. Cell Biol. 114, 715-724 Danowski, B. A., ImanakaYoshida, J. M., Sanger, J. M., and

Sanger, J . W. (1991) J . Cell Biol. 115, 368a Pardo, J. V., Siliciano, J. D., and Craig, S. W. (1983) Proc. Natl.

Acad. Sci. U. S. A. 80, 1008-1012 Pardo, J. V., Siliciano, J. D., and Craig, S. W. (1983) J. Cell Biol.

97,1081-1088 Craig, S. W., and Pardo, J. V. (1983) Cell Motil. 3,449-462 Turner, C. E., and Burridge, K. (1989) Eur. J. Cell Biol. 49,202-

Gimona, M., Furst, D. O., and Small, J. V. (1987) J. Muscle Res.

Saga, S., Hamaguchi, M., Hoshino, M., and Kojima, K. (1985)

Byrne, B. J., Kaczorowski, Y. J., and Craig, S. W. (1991) J. Cell

Coutu, M. D., and Craig, S. W. (1988) Proc. Natl. Acad. Sci.

Price, G. J., Jones, P., Davison, M. D., Patel, B., Bendori, R.,

461 Geiger, B., and Critchley, D. R. (1989) Biochem. J. 259, 453-

Siliciano, J . D., and Craig, S. W. (1987) J. Cell Biol. 104 , 473- 482

Belkin, A. M., Ornatsky, 0. I., Glukhova, M. A., and Koteliansky, V. E. (1988) J . Cell Biol. 107 , 545-553

Otto, J. J. (1983) J. Cell Biol. 9 7 , 1283-1287 Wilkins, J. A., Chen, K. Y., and Lin, S. (1983) Biochem. Biophys.

Burridge, K., and Mangeat, P. (1984) Nature 308, 744-746 Ito, S., Werth, D. K., Richert, N. D., and Pastan, I. (1983) J. Biol.

Chem. 258, 14626-14631 Niggli, V., Dimitrov, D. P., Brunner, J., and Burger, M. M. (1986)

J. Biol. Chem. 261, 6912-6918 Gimona, M., Small, J. V., Moeremans, M., Van Damme, J.,

Puype, M., and Vandekerckhove, J . (1988) EMBO J. 7, 2329- 2334

Thomas, G. P. (1982) J. Bid. Chem. 257, 11024-11031

(1980) Proc. Natl. Acad. Sci. U. S. A. 77,4127-4131

206

Cell Motil. 8, 329-341

Exp. Cell Res. 156 , 45-56

Biol. 11 1, 38a

U. S. A. 85,8535-8539

Res. Commun. 116,1026-1032

Laemmli, U. K. (1970) Nature 227, 680-685

12850 Sequence and Genetic Origin of Chicken meta- Vinculin 26. Towbin, H., Staehelin, T., and Gordon, J . (1979) Proc. Natl. Acad.

Sci. U. S. A. 76 , 4350-4354 27. Coutu, M. D., Simon, D. J., Brown, A. E., and Craig, S. W. (1987)

Biochem. Biophys. Res. Commun. 147,637-643 28. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular

Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

29. Feinberg, A. P., and Vogelstein, B. (1983) Anal. Biochem. 132,

30. Chen, E. Y., and Seeburg, P. H. (1985) D N A ( N Y ) 4 , 165-170 31. Frohman, M. A., Dush, M. K., and Martin, G. R. (1988) Proc.

32. Price, G. J., Jones, P., Davison, M. D., Patel, B., Eperon, I. C.,

33. Barstead, R. J., and Waterston, R. H. (1989) J. Biol. Chem. 2 6 4 ,

6-13

Natl. Acad. Sci. U. S. A. 85,8998-9002

and Critchley, D. R. (1987) Biochem. J. 245,595-603

10177-10185

34. Endo, T., and Masaki, T. (1982) Bwchem. Biophys. Res. Commun.

35. Molony, L., and Burridge, K. (1985) J. Cell Biochem. 2 9 , 3136 36. Jones, P., Jackson, P., Price, G. J., Patel, B., Ohanion, V., b a r ,

A. L., and Critchley, D. R. (1989) J. Cell Bwl. 109,29172927 37. Belkin, A. M., Ornatsky, 0. I., Kabakov, A. E., Glukhova, M. A.,

and Koteliansky, V. E. (1988) J. Biol. Chem. 263,6631-6635 38. Glukhova, M. A., Kabakov, A. E., Frid, M. G., Ornatsky, 0. I.,

Belkin, A. M., Mukhin, D. N., Orekhov, A. N., Koteliansky, V. E.. and Smirnov. V. N. (1988) Proc. Natl. Acad. Sci. U. S. A.

107, 1467-1474

. , 86,9542-9546 '

39. Westmeyer, A., Ruhnau, K., Wegner, A., and Jockusch, B. M.

40. Turner. C. E.. Glennev. J. R.. Jr.. and Burridge. K. (1990) J. Cell (1990) EMBO J. 9,2071-2078

Biol. '1 11,1059-1068

491

, , - . .

41. Pavalko, F. M., and Burridge, K. (1991) J. Cell Biol. 114 , 481-