Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Human fibronectin: cell specific alternative mRNA splicing generates polypeptide chains dfffering in the number of internal repeats

Human fibronectin: cell specific alternative mRNA splicing generates polypeptide chains dfffering... Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 volume 12 Number 14 1984 Nucleic Acids Research Hainan fibrODectin: cefl specific alternative mRNA spiking generates porypeptide chains differing in the number of internal repeats Alberto R.Komblihtt*, Karen Vibe-Pedersen and Fransciso E.BaraDe Sir William Dunn School of Pathology, University of Oxford, South Parks Road, Oxford OX1 3RE, UK Received 8 May 1984; Revised and Accepted 27 June 1984 ABSTRACT The nucleotide sequence of five independent cDNA clones, which cover 4843 nucleotides from the poly(A) addition site of human fibronectin (FN) mRNA was determined. The deduced amino acid sequence (1383 residues) covers the COOH- terminal 60% of human FN, spanning the C-terminus, fibrin-, heparin- and cell-binding domains, and shows the exact location of the only two free sulphydryl groups present in each subunit chain. We have recently reported two different FN mRNA species; one of them containing an additional 270 nucleotide insert (ED) that encodes exactly one of the homology type III repeats of the protein. The two mRNAs arise by alternative splicing of a common precursor. SI nuclease mapping of cDNA/RNA hybrids shows that the expression of the two mRNAs is cell specific. Liver only produces the mRNA without the ED, whereas hepatoma cells, breast tumor cells and normal flbroblasts produce both forms of mRNA. Another area of alternative splicing generating three different FN mRNAs in rat liver has been reported by Schwarzbauer e_t al (16). We here provide evidence for the existence in human cells of a fourth mRNA species different from the three described in rat liver. INTRODUCTION Fibronectin (FN) is a high molecular weight glycoprotein involved in various contact processes such as attachment of cells onto substrata, cell migration, mainten- ance of normal cell morphology, cell differentiation, opsonization and wound healing. All these biological activities imply interaction of FN with cells and with extracellular materials. Several binding activities have been located in different domains of the FN molecule. Indeed, FN has affinities for collagen, heparin, fibrin, cell surfaces, bacteria, DNA and itself (for reviews, see 1 - 3). FN is probably one of the most versatile proteins known so far, both functionally and structurally. It exists in at least two similar but not identical forms. Plasma FN (previously known as cold insoluble globulin) is present at high concentration in plasma (300(ig/ml) as a soluble heterodimer of polypeptide chains of mol. wt. — 220,000. Cellular FN appears either as an insoluble multimer, deposited in the extracellular matrix of various cell types (4), or as a secreted soluble molecule. Cellular FN attracted a great deal of interest because its expression is greatly affected by oncogenic transformation (5-7). Plasma and cellular FNs differ in solubility, electrophoretic behaviour, certain biological activities (8) and © IRL Press Limited, Oxford, England. 5853 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research immunoqenicity (9). Nevertheless, partial primary structure data have revealed highly conserved amino acid sequences both between the two FN forms and among FNs from different species: bovine plasma (10), bovine cellular (11), human plasma (12,13), human cellular (11,14,15), rat plasma (16). All these data tend to confirm that the basic FN polypeptide is structured by three different types of internal repeats (homology types I, II and III) as originally shown in bovine plasma FN (17,10). One of the interesting questions concerning the repetitive structure of FN is, whether variations of a basic FN polypeptide "theme" could account for the differences between cellular and plasma fibronectins and also between the polypeptide chains of both forms. Accordingly, we have recently reported the existence of two FN mRNA species in a human cell line. One of them contains an additional 270 nucleotide insert (ED) that encodes exactly one of the homology type III repeats of the protein. Most interestingly, the ED insert seems to be absent in plasma FN (14). The two mRNA species arise by alternative splicing of a common precursor (Vibe-Pedersen e_t al, submitted). Schwarzbauer et al (16) have simultaneously reported three different FN mRNAs rising by alternative splicing in rat liver, which differ in another area downstream of the ED. The difference sequence does not belong to any of the known internal homologies and it is inserted between the last two type III homology repeats, near the COOH terminus. All these findings, together with the accumulated evidence for the existence of a single FN gene (11,16) draw a complex picture of multiple FN mRNAs arising from the same gene by differential splicing in at least two distinctive regions of the primary transcript. We report here the sequence of five cDNA clones covering 4.9 Kb from the polyA tail of human FN mRNA (estimated size 7.9 Kb), which provides the amino acid sequence of two thirds of human FN. We show the existence in human cells of a fourth mRNA species, different from any of the three reported in rat liver, varying in the same area of differential splicing reported by Schwarzbauer et al (16). Furthermore, and most interestingly, in order to assess the biological relevance of multiple FN mRNAs, we have examined the expression of some of the different FN mRNA species in a variety of human cell types. EXPERIMENTAL PROCEDURES RNA preparation Human cell lines Hs578T (18) and Hep 3B (19) were cultured in Dulbecco's modified Eagle's medium containing 10% foetal calf serum. Total RNA was extracted from confluent cell monolayers by the guanidine-HCl method (20). Between 2 and 4 mg of total RNA were extracted from 4 x 108 cells. Total RNA from RJK 735 fibroblasts was kindly provided by Dr M Goldman. 5854 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research Isolation of fibronectin cDNA clones All the cDNA clones depicted in Figure 1 were obtained using Hs578T cell RNA as template. Isolation of clone pFHl by oligonucleotide probing was described in Kornblihtt et al (11). Isolation of clones pFH23, pFH37, pFHll l and pFH154 by "mRNA walking" (oligonucleotide priming) was described in Kornblihtt et al (14). Sequence determination Inserts from clones were excised from the vector DNA by digestion with appropriate restriction enzymes, separated in agarose gel electrophoresis, and recovered by electroelution (21). Most of the sequencing was performed by the chemical degradation procedure of Maxam and Gilbert (22). The upstream half of clone pFHl insert was sequenced by the chain terminator method. For that purpose, a 1 Kb EcoRl fragment of pFHl was isolated, digested either with Alul or Haelll and ligated to a Smal digested M13mp9 vector (23), previously treated with calf intestinal phosphatase to prevent its circularization. The ligation mixtures were used to transform competent E.coli JM101 and recombinants were selected as clear plaques by insertional inactivation of the f£ -galactosidase gene (24). Single stranded DNA was prepared by standard procedures (25) and the inserts were sequenced by the method of Sanger et a\ (26) using a "universal" 17-nucleotide long primer (27). 51 nuclease mapping SI nuclease mapping was performed as described by Berk and Sharp (28). Probes were labelled at their 5au96I sites by "filling in" with the Klenow fragment of DNA polymerase I and [ct -' 2P ] dGTP. Probes isolated by polyacrylamide gel electro- phoreais were strand separated according to Maxam and Gilbert (22). 10 ng of total RNA were mixed with the appropriate probe in the presence of 20 (ig of poly(A) and co-precipitated with ethanol. PelleU were resuspended in 12.5 nl of 80% formamide, 10 mM PIPES (pH 6.5), 1 mM EDTA, 0.4 M NaCl, heated at 73°C for 10 min under paraffin oil and hybridized overnight at 52°C (probe I) or 46°C (probe II). The hybrids were diluted with 150 \il of SI buffer (250 mM NaCl, 30 mM NaOAc pH 4.4 1 mM ZnSC^,), digested with 3000 u of SI nuclease (Boehringer) for 60 min at 30°C, mixed with 50 |il of SI stop (15 mM EDTA, 600 ng/ml yeast tRNA), ethanol precipitated and analyzed on 6% polyacrylamide 7 M urea gels. RESULTS AND DISCUSSION DNA sequence analysis Figure 1 shows restriction enzyme maps and strategy of sequencing of five independent cDNA clones, which contiguoualy cover 4843 nucleotides from the poly(A) addition site of human FN mRNA. This represents more than 60% of the estimated length of human FN mRNA (7.9 Kb). As previously reported (14), clones pFH37 and 5855 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research pFHl o SxiMI & Hin f l • Dd« I pFH23 pFH111 PFH37 PFH154 Figure 1 Restriction enzyme map and strategy for determining the nucleotide sequences of the inserts of five FN cDNA clones. The 3' end of the map is the 3' end of FN mRNA, and the map covers approximately two-thirds of the FN mRNA molecule. Wavy lines represent vector sequences. Clone inserts are aligned to show overlapping areas. Thick and thin lines represent the coding and non-coding region respectively. ^ ^ indicates contiguity. ED, extra domain. •-» and •-» indicate extent and direction of each sequence analysis performed according to Maxam and Gilbert (22) and Sanger et al (26) respectively. pFH154 lack a 270 bases long internal segment coding for one of the homology type III repeats of FN, called ED for extra domain, while clones pFH23 and pFHll l contain the ED segment. Figure 2 shows the complete nucleotide sequence and its deduced amino acid sequence of the clones depicted in Figure 1. It contains the complete 3' non-coding region (694 nucleotides) and 4149 nucleotides of coding region, accounting for 1383 amino acids of the C-terminal part of FN. A normal polyadenylation signal AATAAA (29) is present at position 4846. From nucleotide 1593 and downstream, the sequence In Figure 2 is homologous to the one reported for rat liver FN mRNA (16), 5856 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research J ; ora TO»CTT T , O A ^ 70 90 00 10O 110 1» TIUTCAATOAAACTIMTTC T AC TO H X 130 14© 190 34O 330 MO T T I W I T U T TOCAOCCTCaOAOCTCT ATTCCACCT T At CTOAOACCACCATCQ TQATCACATOtWCOl *TTQQTT T T AAOCTQQOTI 37 0 390 40 0 410 420 430 43 0 4K0 33 0 300 37O 3*0 10*0 1100 1110 1120 U3O lit© 1170 11*0 11M 1200 ACTTTTCTOAT ATT AC TOCCAACTC T TT TAC TO TOCAC TOOA T TQCTCC TCOAOCCACCATC ACTQOCT ACAOOATCCOCCATCATOXOAOCACrny«nOOCUW«XTCaAa**OATC 1210 1220 1Z30 1240 1230 12S0 1270 12*0 12*0 13OO 1310 13X0 Trr*ftn^ 1330 1340 1330 1380 1370 13*0 1390 14O0 1410 1420 143O 1440 C *ACAO TT TC TM TO TTCCQAOOOACCT DOAA QT TQTT OC TOC OACCCCCACCAOCC T AC TOATCAOCTOOOATllCriXCTQCTQTCACAOTOAtlATATTACAaQATCACTTAaiQAQAAA 1430 1460 1470 14M 14»0 13O0 1910 1320 1330 134O 133O 13S0 CAOaAOCWWWTAOCCCTOTCCAnQAOTTCACTOTOCXTtroOACICAAOTCTACAnCTAa^TCAaailtfCTTAAAC^^ 1370 13S0 1390 l«00 1110 1020 1*30 1*40 1B30 1H0 1B7O l»0 T AQTOTCAAaTOQC TOCCT T 1W0 1700 1710 172O 1730 1740 1730 17*0 1770 1790 17 IB10 1920 4O 1130 11S0 l«70 1B90 ISM l«O0 ltlO lt20 ITJO 1540 1*30 lt 1*70 1BSO lBSO 1000 1010 Z020 203O 2O4O TCOATTCCATCAMAT Ti 2O30 2Ofl0 ZCT^ 20*0 20*0 210O 2110 1120 2130 2140 2150 21C0 Z17O 21tO 21*0 23O0 2310 2320 2330 234O 233© 23*0 2370 2390 2400 TQAAAQAAArCAAO.I 101.ILL IIWaMCTCATCCQTttOTTOTATCAaOACTTATI 5857 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research JITC*CC*T T«OCT(XM(IAMXAMACTIMfMC0ATCACTtXlC TTCC R3 0 X34O 2330 13*0 2370 0*0 I3»0 ZSOO 2*10 2*20 2*30 2*40 VOAUPANOOT P [ d ft T I H PDUtlYT I TQL9P0TDY I IYLY T A«.I I IIIA HIT m i I M A<if r AAT^ nrr iflA f TTTA^TrrAflAflflArrATTA<>nrrww>TBTrwflrtflflrTf>rArTfiTrfif ftiMii t n'H tut i wnrf u i nw iflf ft^rift i r Trtrr T" T*^ * 2*30 2**0 2*70 2**0 TOO m>0 2710 2720 2730 2740 2730 17*4 LH9HASIIPVV 1 DAlT A IDAPlWLRFLATTPMELLViM9 P reTTQA*TnACJM»TOCTC3OAaCTgXCraTgCTCATCgACC^KXj«TQCCAT^ r770 27*0 27*0 Z*O0 Jflfl HBO 2S30 2*40 2*30 236O 2170 28*0 TWO 2C00 W1C 2*20 2t30 2B4© ZB3C 2B0O 217O ZOCO 2*»0 3OO0 3020 3030 7040 3030 30C0 307O 30*0 30*0 3100 3110 3120 iC TOOAAATOOT A TTCAOCTTCCTOOCACT TCTQQTCAOCAACCCA 3130 3140 3130 3110 3170 31*O 31B0 32O0 3210 322O 3230 3240 3230 32*0 3270 32B0 3200 3300 3310 3320 3330 3340 333O 3360 3370 33*0 33*0 3400 3410 342O 3430 3440 3430 34*0 3470 34*0 ATLTOLTROATY N I IVEALKDaQRHKVRECVVTVaHlVM K Q rOCCtfTCTO«r\aOCTTCiWXAfl*tt]TOCC*CCTACAACATC*Tft«TOQ^ WO 3300 3310 3320 333O 3340 3330 33S0 337O 3310 33*0 3*00 rtrftr«TrTrrrftTTftTorniTTmwy>TqftOTnftnr>ftrn^ 3110 3*20 3«3O 3040 3630 3S*0 3070 3**0 3**O 3 TOO 3710 3720 FQeaHPftCDIflRMCHOMDUNYKIQEKHDRBOINOgnni C rrrrOMAdTOOTCA rrrCAOA TQni*TTCATC T AOATOOTQCC A TOACAATQaTOT 0A«rrACAi*aATTlH»ClAaAAaTOaOACCaTCAOOOAaA*UWT0OCCAOATl»ATOAOCT 373C 3740 373O 37SO 377O 37*0 37*0 3S0O 3S1O 31120 3*30 3*40 CLaMaxaiFKCDfHEATCYDOQKTYHUOCQuaKIYLQA I rTTftTBAorirwirnTnTTftniftTfWTTmivw^ofrftTftrcftfTrrftowwu'f WToncftrwftnnrviTi'iTrTrnnTfTrTft 3*3 0 39*0 3S7O 3**O 3**O 3*OO 3*10 3*20 3*30 3*40 3*3O 3*90 CS C TCFQQQRQUftCD H CrtffPQQEPSPGQTTQQSYMGYSS R TTTHCTCCTaCACATOCTTTOOAOaCCAQCtmOdC I OttCOCTQT(lAC*ACTqCCOCJ«ACCTaDQ(MTQA<»CCCJMTCCCOA*OOCACTACTCaCCAO 3170 3S*0 3**0 40OO 4010 4020 4O3O 4040 4030 40S0 4O70 40*0 YHQRTMTHVHC P ICCFRPLDVOADRIDIR I OATACC*TaMAaA*CAAJtfACTAaTgTTA*TTacCCA*TTnj«lT0CTTC^TgC?TTTAOAT0TAC*QOCTOtt 4C*0 4100 4:10 4120 4i±O 4140 4190 41*0 4170 41*0 41*0 4200 * /^* ffl^ TT*TTTfTrrriminTctriTrTftiVri^fTniMrmnTorr* wmmrr r 4Z10 4220 4230 424O 4Z30 42*0 4270 42*0 42*0 43O0 4310 4320 C ACAOCTTCTCCAAOCATCACCC TOanAOrTTC CTOAOaOT T TTCTCA T ArfMTllAOQaCTOCi^ 433O 4340 4330 43O0 4370 438O 43*0 4400 4410 4420 4430 4440 ••"'"'"TTnTTTTTTfflnTrnnTrtmriviTTrnTnTnrnm^ ^ 4430 44SO 4470 44*0 44SO 43O0 4310 4320 433O 4340 4330 43*0 TA*OTOTCTtmCCCaCA*TACTirrACnA«CA*qaaaATCTTflTTACTQTMTAT^ 437O 43*0 43*0 4«00 «10 4*20 4*30 4*40 4*30 4*S0 4*70 4**0 TTiVTTTTTTftTTTnTTWTTTTTCtTflin'nTT^TKTnCtliVllVwVlrtTTqTftTTfWV^ 4*»O 470© 4710 4720 473O 4740 4730 47*0 4770 47*0 47*0 4*O0 4110 4*20 4*30 4940 «&" 4CS0 4*70 4*90 4*90 Figure 2 Nucleotide sequence and predicted amino acid sequence (top row) resulting from the sequencing of the five overlapping FN cDNA clones depicted in Figure 1. Part of this sequence (from positions 1728 to 2538, and from 3883 to 4892) has been reported before (11,14). The first 23 nucleotides of the 5' end of clone pFH154 insert (In brackets) have not been translated because we have evidence (not shown) that it is an inverted sequence due to cloning artefacts. Arrows indicate the ends of the 270 5858 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research bases long ED segment (position 1999 and 2268) and of the 267 bases long IIICS segment (positions 3082 and 3348). Nucleotide 2525 is a T in clones pFH23 and pFHll l but an A in pFH 37 and pFH154. The following sequences were underlined: polyadenylation signal (positions 4846 to 4851), region complementary to the synthetic oligonucleotide used as a probe in the isolation of the bovine cDNA clone which allowed the isolation of the first human FN cDNA clone, pFHl (see 11) (positions 4122 to 4135) and region complementary to the synthetic oligonucleotide U3ed as primer in "mRNA walking" (see Experimental Procedures) (positions 2808 to 2823). A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, He; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gin; R, Arg; S, Ser; T, Thr; V, Val; W, Trp and Y, Tyr. except within the two areas of RNA variability due to alternative splicing, ED (positions 1999 to 2268) and IIICS (positions 3082 to 3348) (see below). The comparable areas within the coding region showed 88% homology at the nucleotide level and 92% homology at the amino acid level. Surprisingly, rat cDNA contains an extra codon (GCT) coding for alanine [position 2260 in Figure 3 of Schwarzbauer et al, (16)] which is not present in the human or the bovine (11) cDNAs. However, this amino acid is located in an area of high interspecies variation whose primary structure is believed not to be critical for protein function. Comparison between rat and human 3' non- coding regions has already been analyzed by Schwarzbauer et al (16) and it will not be discussed here. From nucleotides 1409 to 1718, the sequence shown in Figure 2 is identical to the sequence of a 3hort cDNA clone coding for the cell attachment domain of human FN, pHFN-1, isolated by Oldberg et al (15). These authors used a human fibrosarcoma tumour as source of mRNA, while we used an epithelial cell line derived from a human breast carcinosarcoma (18). Both tissues presumably make cellular FN, and comparison of a stretch of 300 bases does not show any difference between their FN mRNAs. Amino acid sequence and internal homoloqies Figure 3 shows the amino acid sequence deduced from the nucleotide sequence of the human cDNA clones depicted in Figure 1. The 1383 amino acids are aligned to show internal homologies. The sequence spans the C-terminus, fibrin-(Fib-2), heparin- (Hep-2) and cell-binding domains of human FN. The sequence also shows the exact localization of the only two sulfhydryl groups known to exist per subunit chain of FN. One of them (Figure 3, line 4) is the same as the one identified in bovine plasma FN by Vibe-Pedersen et al (30). The second one (line 14) has been predicted by SH titration experiments and it has been mapped in a 31 kilodalton tryptic fragment, near the Fib-2 domain of human plasma FN (31). Furthermore, a peptide of bovine plasma FN containing a second free SH group has been sequenced. The bovine sequence: IISCHPVGIDEEPLGF fits the human sequence in line 14 (Dr T Petersen, personal communication). Two types of internal homologies are represented in Figure 3: three "fingers" or 5859 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 CD o_ (n HoMOtogy to V) tc -^ 3" 0 T - - • lQlQl!l[9<<ITD[IPLQrlTMTITg > - I * tj A LJrft^Q Q I Figure 3 Amino acid sequence deduced from nucleotide sequence shown in Figure 2, The 1383 amino acids are aligned to show internal homologles. Identical residues in the three "fingers" (Iine9 15, 16, and 17) and in the 12i repeats of homoiogy type III are boxed. Gaps (-) are inserted to obtain maximal homoiogy. Arrowheads indicate amino acids conserved in all the type III repeats. Black dots are on top of free sulfhydryl groups. ED, extra domain; IIICS, homoiogy type III connecting segment. Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research repeats of homology type I (lines 15, 16 and 17) and 12£ repeats of homology type III (lines 1 to 12 and 14). The last type III homology (line 14) is separated from the rest of the type III repeats by a stretch of 89 amino acids (line 13), whose sequence is not homologous to any of the three kind3 of homologies. We called this segment IIICS, for homology type III Connecting Siegment. Our results show for the first time that most of FN is formed by type III sequences. The degree of homology within the type III repeats is quite remarkable, except for repeat in line 3 in which only 11 positions are conserved, compared to an average of 30 conserved positions per repeat. Two amino acids, tryptophan and tyrosine, have a conserved position in all the repeats (arrows in Figure 3), as is the case in bovine (10) and rat (16) FNs. Furthermore, the conserved sequences seem to be distributed in two "peak" areas around both tryptophan and tyrosine, separated by a "valley" of low homology. The N-terminal portion of the amino acid sequence in Figure 3 does not overlap with any other published FN sequence. However, it is possible to predict that between the N-terminus of the sequence reported here and the C-terminus of the collagen-binding fragment [partially sequenced in bovine by Skorstengaard et al, (34)], there is room for as many as 2f repeats of homology type III (see Figure 5A). One of these predicted repeats should be contained in the DNA binding domain, since the N-terminal sequence of a fragment with that activity (12) is homologous to type III sequences and it is not present in Figure 3. Primary structure of fibronectin Over the last decade, structural studies on the FN molecule have been performed mainly by the use of proteolytic enzymes in peptide mapping. Only recently, partial but extensive data on primary structure have been obtained by direct sequencing of bovine plasma fibronectin (10). This information has been invaluable for the identification of cloned fragments of DNA containing gene (32) or cDNA (11,14,15,16) sequences for fibronectins. The use of recombinant DNA techniques by us and by other laboratories, has proved to be successful in accelerating the completion of FN primary structure and also in producing new evidence on the existence of multiple FN mRNAs (see below). The amino acid sequence shown in Figure 3 represents the longest stretch of FN sequence reported so far, covering 60% of the molecule. The large size of FN not only imposes difficulties to the "peptide sequencers", but also to the "DNA cloners", who pursue the "full length" cDNA. In fact, current protocols rarely yield cDNA clones as long as 8.0 Kb (the size of FN mRNA) and "mRNA walking" strategies (33) have to be used. The sequence in Figure 3, together with the amino acid sequence of the 29 kilodalton N-terminal domain of human plasma FN (13) account for about 80% of the human FN molecule. If we also consider the protein sequences of the collagen-binding area of bovine FN (34), we can draw the model of a typical FN 5861 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research polypeptide, regarding the three types of internal homologies (Figure 5A). it contains, from N- to C-terminus, 6 repeats of type I, 2 repeats of type n, 3 repeats of type I, 13 to 15 repeats of type III, 3 repeats of type I homologies and a COOH-terminal non homologous stretch where the disulfide bridges that hold FN monomers together are located. Figure 5A alao shows the distribution of the most relevant binding activities along the molecule, as well as the location of the only two free sulfhydryl groups present per FN monomer. Two different FN mRNAs with cell-specificity In their expression Two different FN mRNA species have been identified in the human cell line H8578T (14). One species (mRNAI) contains an additional 270 nucleotide insert (ED), while the other (mRNAII) lacks it. In "Northern" blots of total RNA from Hs578T cells and normal fibroblasts, a specific probe containing only ED sequences hybridizes to bands of the same size as mature FN mRNA ( ~ 7.9 Kb) (results not shown). This indicates that mRNAI is not an unspliced long precursor of mRNAII. Furthermore, the polypsptide coded by the ED is a homology type III repeat in itself (Figure 3, line 9), and it is located between the cell- and heparin-binding domains. We have put forward the idea that the presence of the ED in an FN polypeptide may reflect one of the differences between cellular and plasma FNs. This is based on the fact that while the cell line Hs578T (used for the isolation of the clones in Figure 1) produced both mRNAs, liver tissue contained none or undetectable levels of mRNAI, the ED-bearing species. Since liver has been reported to be the source of plasma FN (35), our results indicated that the presence of the ED could be a particular feature of cellular FN. This conclusion may also be supported by a series of results obtained by other laboratories and summarized as follows: (1) The ED is not present in two independent FN cDNA clones from rat liver (A rlf3 and X rlf6), isolated by Schwarzbauer et al (16). (2) The ED has not been seen at the polypeptide level in bovine plasma FN [segment 6 in Petersen et al, (10)1 (3) Hayashi and Yamada (36) showed that, among other differences, heparin-binding fragments are 10,000 daltons larger in cellular than in plasma chicken FNs. The size (10,000 daltons = 90 amino acids) and location of this polypeptide are coincident with those of the ED. (4) A difference region between human cellular and plasma FNs has been seen in the same area as we see the extra domain (Drs S Akiyama and K Yamada, personal communication). To test this idea further, we have investigated the expression of both mRNAs in different cell types, by SI nuclease mapping of cDNA/RNA hybrids. Two probes were prepared: probe I was the 446 nucleotide 3' end-labelled anti-sense strand 5au961-HaeII of clone pFH23 insert and probe II was the 176 nucleotide 3' end-labelled anti-sense strand Sau96I-HaeII of clone pFH37 insert. Total RNA from the following human cells was analyzed: epithelial cell line Hs578T, derived from a breast carcinoaarcoma; cell 5862 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research EHF - LE EHF-LE • • Sau9fc : liae II ..„ PROBE I 3 ' =>— s 446 PROBE II 3 Figure 4 SI nuclease analysis of RNA/cDNA hybrids. Probe I was the Sau96I-HaeII fragment of pFH23 insert and probe II was the 5au96I-HaeII fragment of pFH37 insert. Probes were hybridized to human RNAs from the following sources: E, epithelial tumour cell line Hs578T; H, hepatoma cell line Hep 3B; F, RJK 735 fibroblasts and L, liver. - control without RNA in the hybridization mixture. Probe fragments, SI nuclease-resistant DNA products and mRNA species are shown in the diagram below the Figure. Correspondence between diagrams and gels are indicated by letters. Arrows indicate the expected position of bands A and D in liver RNA. 5863 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research line Hep 3B, derived from a hepatocellular carcinoma; RJK 735 normal fibroblasts and total liver. When probe I was hybridized to RNA from Hs578T, Hep 3B or RJK 735 cells, two SI nuclease-reslatant DNA fragments were detected (Figure 4, products A and B). Product A (446 nucleotides) represents the fully protected probe and indicates the existence of the mRNA species containing the ED (mRNAI). Product B (119 nucleotides) exactly covers the distance between the Sau961 site and the base in which mRNAs I and II diverge. When probe II was used, again two SI nuclease-resistant products were observed (Figure 4, products C and D). Band C (176 nucleotides) is the fully protected probe, this time indicating the existence of mRNA molecules (mRNAII) lacking the EDj while band D, co-migrating with band B, again represents the distance between the labelled end of the probe and the point of divergence between the two types of RNAs. When similar SI analysis was carried out with human liver RNA, only bands B and C were detected (Figure 4, L lanes). Bands A and D cannot be distinguished above the background, Indicating the existence only of the mRNA lacking the ED. These experiments showed that, while liver tissue only produces mRNAII, both forms of FN mRNA are produced by cultured cells of various origins. Most interestingly, one of these cell lines (Hep 3B) is of liver origin and it has been shown to synthesize and secrete many of the major human plasma proteins (19). However, we show here that, with respect to FN, Hep 3B cells process their RNA in a different way to that of the parental tissue. This still raises the guestion whether the synthesis of FN mRNA having the ED is a phenomenon restricted to tumour cell lines. The presence of the ED in FN mRNA from normal fibroblasts (Figure 4, lane F) seems to rule out this possibility. If, as discussed above, the ED is a difference sequence characteristic of cellular FN, then we could argue that hepatoma cells, like other cells in culture, are making the cellular form of FN, while liver cells are mainly making plasma FN. These conclusions are limited, because we are comparing RNA from an organ with RNA from cultured cells. However, the observation of cell specificity in the expression of the ED will stand independently of whether it correlates with a change in the environment of the cells (hepatocytes in liver —» hepatocytes in culture) or with a change in the differentiation state (normal liver —» hepatoma cells). Two regions of alternative splicing generate multiple FN mRNAs The two FN mRNAs differing by the ED segment arise by alternative splicing of a common precursor, and the ED segment is represented by one exon in the human genome (Vibe-Pedersen et al, submitted). But surprisingly, the ED is not the only difference sequence found in FN mRNAs. Schwarzbauer et al (16) found another region of variability, 0.8 Kb downstream of the ED, in rat liver FN mRNA. They 5864 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research COOH Coll.gtn DNA Hcparln Fibrin IIICS ED pFHJ l II I II I pFHtti • 0 CM PFHJ7 II L III <T ^ J IIII I WHIM III ArtiO CM Figure 5 A. Model of a typical fibronectin polypeptide with the localization of the internal rfomologies (types I, II and III). Unbroken lines and boxes represent sequences known in human FNs: towards the NH2-terminus, reported by Garcia-Pardo et al (13); towards the COOH-terminus, reported in this paper. Broken lines and boxes represent sequences unknown in human, but known in bovine FN (34). B. Schematic representation of the different FN polypeptides that could arise by the translation of the RNAs generated in the ED region (left) and in the IIICS region (right). The name of the cDNA clones representing the corresponding encoding mRNA species is writte n down to the right of each polypeptide. Cell and Hep name the cell- and heparin-binding domains. /^\ indicates contiguity. reported three FN mRNAs arising from alternative splicing. The three RNAs differ in a region not belonging to any of the three internal homologies, which corresponds to the above defined IIICS. Comparison of the human IIICS nucleotide sequence, reported for the first time in this paper, (Figure 2, positions 3082 to 3348) with the rat liver nucleotide sequences in the same area reveals that the human sequence (present in clone pFHl) constitutes a fourth mRNA species. In fact, the human IIICS encodes an 89 amino acids long polypeptide which is highly homologous with the firs t 89 of the 120 amino acids encoded by the rat IIICS in clone X rlf 2 of Schwarzbauer et al (16). Preliminary evidence shows the existence of more than one mRNA in the human IIICS region (not shown). Figure 5A shows the localization of the two regions of variability 5865 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research along the FN molecule. Figure 5B shows the different kinds of peptides that could arise from the translation of the different human and rat RNAs generated in the ED region (left) and in the IIICS region (right). It seems important to point out that all these differences have been seen at the RNA level and, except for the evidence of polypeptides lacking the ED in bovine plasma FN [segment 6 in Petersen et al, (10)], none of the other alternatives have yet been demonstrated at the protein level. Genomic blot experiments are consistent with the existence of a single FN gene in the human haploid genome (11 and unpublished results). A similar kind of analysis in rat supports this conclusion (16). If there is only one gene, and if we assume that rat and human FN genes are directly comparable, the two regions of alternative splicing could in theory originate eight different species of FN mRNA from a single primary transcript. We wil l attempt to discuss, blending evidence and speculation, if there are any biological activities associated with the difference sequences. With respect to the IIICS area, Schwarzbauer et al (16) have speculated that, among other possibilities, this region might represent a site for interaction with fibrin or for fibronectin-fibronectin binding. Ehrismann et al (37) have mapped the self- association site in a region of horse plasma FN corresponding to the IIICS area and have proposed that the difference in size of the two chains of plasma FN Is due to the absence of a self-association site in one of them. We have already suggested that the ED could be present in cellular but not in plasma FN polypeptides. If so, the presence of the ED may explain particular features of cellular FN, such as its lower solubility, the ability to aggregate and form covalently bound multimers, its higher activity in haemagglutination and restoration of normal morphology in transformed cells (8), and its greater affinity for hyaluronlc acid (38). It is not clear whether these abilities reside in a differential sequence of cellular FN or whether they are consequences of its aggregated state. It may be that the differential biological activity resides in the ED itself, in which case, one should look at stretches of non conserved amino acids, compared to other type III homology repeats. Since the ED polypeptide does not contain any cysteine, It cannot be responsible for multimer formation. Alternatively, the ED can be increasing the distance between binding sites or active centres, interfering with or creating biologi- cal activities. Whatever role is played by the ED polypeptide, our results show that the expression of its gene counterpart is cell-specific. Nothing is known about cell- specificity in the IIICS area and since, in our experiments, the ED and IIICS areas are covered by different clones, we cannot assess whether there Is a "functional linkage" in their expression. For example, one could speculate that the expression of the ED always correlates with the expression of one or some of the possible segments in the IIICS area. 5866 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research We are just starting to understand this complex system. The analysis of the actual number of alternatively spliced regions and their biological significance requires further extensive studies on primary structure and expression of fibronectin. ACKNOWLEDGEMENTS We thank Dr K Umezawa for helpful discussions, and Drs M Goldman and R. Corte3e for providing the RJK 735 RNA and Hep 3B cells, respectively. K.V-P is a recipient of a Danish National Research Council Fellowship. *To whom reprint requests should be sent REFERENCES 1. Hynes, R.O. and Yamada, K.M. (1982) J.Cell Biol. 95, 369-377. 2. Yamada, K.M. (1983) Ann.Rev.Biochem. 52, 761-799. 3. Vartio, T. and Vaheri, A. (1983) TIBS, , 442-444. 4. Hynes, R.O. and Destree, A. (1977) Proc.Natl.Acad.Sci.USA 74, 2855-2859. 5. Yamada, K.M. and Olden, KJ. (1978) Nature 275, 179-184. 6. Fagan, J.B., Sobel, M.E., Yamada, K.M., de Crombrugghe, B. and Pastan, I. (1981) J.Biol.Chem. 256, 520-525. 7. Smith, H.S., Riggs, J.R. and Mosesson, M.W. (1979) Cancer Res. 39, 4138-4144. 8. Yamada, K.M. and Kennedy, D.W. (1979) J.Cell Biol. 80, 492-498. 9. Atherton, B.T. and Hynes, R.O. (1981) Cell 25, 133-141. 10. Peteraen, T.E., ThiSgersen, H.C., Skorstengaard, K., Vibe-Pedersen, K., Sahl, P., Sottrup-Jensen, L. and Magnusson, S. (1983) Proc.Natl.Acad.Sci.USA 80, 137-141. 11. Kornblihtt, A.R., Vibe-Pedersen, K. and Baralle, F.E. (1983) Proc.Natl.Acad.Sci.- USA 80, 3218-3222. 12. Pande, H. and Shively, J.E. (1982) Arch.Biochem.Biophys. 213, 258-265. 13. Garcia-Pardo, A., Pearlstein, E. and Frangione, B. (1983) J.Biol.Chem. 258, 12670-12674. 14. Kornblihtt, A.R., Vibe-Pedersen, K. and Baralle, F.E. (1984) EMBO Jnl. 3, 221- 15. Oldberg, A., Linney, E. and Ruoslahti, E. (1983) J.Biol.Chem. 258, 10193-10196. 16. Schwarzbauer, J.E., Tamkun, J.W., Lemischka, I.R. and Hynes, R.O. (1983) Cell 35, 421-431. 17. Skorstengaard, K., Thtfgersen, H.C., Vibe-Peder3en, K., Peterson, T.E. and Magnusson, S. (1982) Eur.J.Blochem. 128, 605-623. 18. Hackett, A.J., Smith, H.S., Springer, E.L., Owens, R.B., Nelson-Rees, W.A., Riggs, J.L. and Gardner, M.B. (1977) J.Natl.Cancer Inst. 58, 1795-1800. 19. Knowles, B., Howe, C.C. and Aden, D.P. (1980) Science 209, 497-499. 20. Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J. (1979) Biochemistry .18, 5294-5299. 21. Girwitz, S.C., Bacchetti, S., Rainbow, A.J. and Graham, F.L. (1980) Anal.Bio- chem. 106, 492-496. 22. Maxam, A.M. and Gilbert, W. (1980) Methods Enzymol. 65, 499-580. 23. Messing, J. and Vieira, J. (1982) Gene 19, 269-276. 24. Messing, J. , Crea, R. and Seeburg, P.H7T1981 ) Nucl.Acids Res. 9, 309-321. 25. Winter, G. and Fielda, S. (1980) Nucl.Acids Res. 8, 1965-1974. 26. Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc.Natl.Acad.Sci.USA 74, 5463-5467. 5867 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research 27. Duckworth, M.L., Gait, M.J., Goelet, P., Hong, G.F., Singh, M. and Titmas, R.C. (1981) Nucl.Acida Res. 9, 1691-1706. 28. Berk, A.J. and Sharp, P.A. (1977) Cell 12, 721-732. 29. Proudfoot, N.J. and Brownlee, G.G. (1976) Nature 263, 211-214. 30. Vibe-Pedersen, K., Sahl, P., Skorstengaard, K. and Petersen, T.E. (1982) FEBS Lett. 142, 27-30. 31. Smith, D.G., Mosher, D.F., Johnson, R.B. and Furcht, L.T. (1982) J.Biol.Chem. 257, 5831-5838. 32. Hirano, H., Yamada, Y., Sullivan, M., De Crombrugghe, B., Pastan, I. and Yamada, KM. (1983) Proc.Natl.Acad.Sci.USA 80, 46-50. 33. Lawn, R.M., Adelman, J., Bock, S.C., Franke, A.E., Houck, CM. , Najarian, R.C, Seeburg, P.H. and Wion, K.L. (1981) Nucl.Aclds Res. 9, 34. Skorstengaard, K., Thrfgersen, H.C. and Petersen, T.E. (1984) Eur.J.Biochem., in press. 35. Tamkun, J.W. and Hynes, R.O. (1983) J.Biol.Chem. 258, 4641-4647. 36. Hayashi, M. and Yamada, K.M. (1981) J.Biol.Chem.~236, 11291-11300. 37. Ehrismann, R., Roth, D.E. Epponberger, H.M. and Turner, D.C. (1982) J.Biol.- Chem. 257, 7381-7387. 38. Laterra, J. and Culp, L.A. (1982) J.Biol.Chem. 257, 719-726. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nucleic Acids Research Oxford University Press

Human fibronectin: cell specific alternative mRNA splicing generates polypeptide chains dfffering in the number of internal repeats

Download PDF
16 pages

Loading next page...
 
/lp/oxford-university-press/human-fibronectin-cell-specific-alternative-mrna-splicing-generates-cWVfG8NdXg

References (0)

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Oxford University Press
Copyright
© IRL Press Limited
ISSN
0305-1048
eISSN
1362-4962
DOI
10.1093/nar/12.14.5853
Publisher site
See Article on Publisher Site

Abstract

Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 volume 12 Number 14 1984 Nucleic Acids Research Hainan fibrODectin: cefl specific alternative mRNA spiking generates porypeptide chains differing in the number of internal repeats Alberto R.Komblihtt*, Karen Vibe-Pedersen and Fransciso E.BaraDe Sir William Dunn School of Pathology, University of Oxford, South Parks Road, Oxford OX1 3RE, UK Received 8 May 1984; Revised and Accepted 27 June 1984 ABSTRACT The nucleotide sequence of five independent cDNA clones, which cover 4843 nucleotides from the poly(A) addition site of human fibronectin (FN) mRNA was determined. The deduced amino acid sequence (1383 residues) covers the COOH- terminal 60% of human FN, spanning the C-terminus, fibrin-, heparin- and cell-binding domains, and shows the exact location of the only two free sulphydryl groups present in each subunit chain. We have recently reported two different FN mRNA species; one of them containing an additional 270 nucleotide insert (ED) that encodes exactly one of the homology type III repeats of the protein. The two mRNAs arise by alternative splicing of a common precursor. SI nuclease mapping of cDNA/RNA hybrids shows that the expression of the two mRNAs is cell specific. Liver only produces the mRNA without the ED, whereas hepatoma cells, breast tumor cells and normal flbroblasts produce both forms of mRNA. Another area of alternative splicing generating three different FN mRNAs in rat liver has been reported by Schwarzbauer e_t al (16). We here provide evidence for the existence in human cells of a fourth mRNA species different from the three described in rat liver. INTRODUCTION Fibronectin (FN) is a high molecular weight glycoprotein involved in various contact processes such as attachment of cells onto substrata, cell migration, mainten- ance of normal cell morphology, cell differentiation, opsonization and wound healing. All these biological activities imply interaction of FN with cells and with extracellular materials. Several binding activities have been located in different domains of the FN molecule. Indeed, FN has affinities for collagen, heparin, fibrin, cell surfaces, bacteria, DNA and itself (for reviews, see 1 - 3). FN is probably one of the most versatile proteins known so far, both functionally and structurally. It exists in at least two similar but not identical forms. Plasma FN (previously known as cold insoluble globulin) is present at high concentration in plasma (300(ig/ml) as a soluble heterodimer of polypeptide chains of mol. wt. — 220,000. Cellular FN appears either as an insoluble multimer, deposited in the extracellular matrix of various cell types (4), or as a secreted soluble molecule. Cellular FN attracted a great deal of interest because its expression is greatly affected by oncogenic transformation (5-7). Plasma and cellular FNs differ in solubility, electrophoretic behaviour, certain biological activities (8) and © IRL Press Limited, Oxford, England. 5853 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research immunoqenicity (9). Nevertheless, partial primary structure data have revealed highly conserved amino acid sequences both between the two FN forms and among FNs from different species: bovine plasma (10), bovine cellular (11), human plasma (12,13), human cellular (11,14,15), rat plasma (16). All these data tend to confirm that the basic FN polypeptide is structured by three different types of internal repeats (homology types I, II and III) as originally shown in bovine plasma FN (17,10). One of the interesting questions concerning the repetitive structure of FN is, whether variations of a basic FN polypeptide "theme" could account for the differences between cellular and plasma fibronectins and also between the polypeptide chains of both forms. Accordingly, we have recently reported the existence of two FN mRNA species in a human cell line. One of them contains an additional 270 nucleotide insert (ED) that encodes exactly one of the homology type III repeats of the protein. Most interestingly, the ED insert seems to be absent in plasma FN (14). The two mRNA species arise by alternative splicing of a common precursor (Vibe-Pedersen e_t al, submitted). Schwarzbauer et al (16) have simultaneously reported three different FN mRNAs rising by alternative splicing in rat liver, which differ in another area downstream of the ED. The difference sequence does not belong to any of the known internal homologies and it is inserted between the last two type III homology repeats, near the COOH terminus. All these findings, together with the accumulated evidence for the existence of a single FN gene (11,16) draw a complex picture of multiple FN mRNAs arising from the same gene by differential splicing in at least two distinctive regions of the primary transcript. We report here the sequence of five cDNA clones covering 4.9 Kb from the polyA tail of human FN mRNA (estimated size 7.9 Kb), which provides the amino acid sequence of two thirds of human FN. We show the existence in human cells of a fourth mRNA species, different from any of the three reported in rat liver, varying in the same area of differential splicing reported by Schwarzbauer et al (16). Furthermore, and most interestingly, in order to assess the biological relevance of multiple FN mRNAs, we have examined the expression of some of the different FN mRNA species in a variety of human cell types. EXPERIMENTAL PROCEDURES RNA preparation Human cell lines Hs578T (18) and Hep 3B (19) were cultured in Dulbecco's modified Eagle's medium containing 10% foetal calf serum. Total RNA was extracted from confluent cell monolayers by the guanidine-HCl method (20). Between 2 and 4 mg of total RNA were extracted from 4 x 108 cells. Total RNA from RJK 735 fibroblasts was kindly provided by Dr M Goldman. 5854 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research Isolation of fibronectin cDNA clones All the cDNA clones depicted in Figure 1 were obtained using Hs578T cell RNA as template. Isolation of clone pFHl by oligonucleotide probing was described in Kornblihtt et al (11). Isolation of clones pFH23, pFH37, pFHll l and pFH154 by "mRNA walking" (oligonucleotide priming) was described in Kornblihtt et al (14). Sequence determination Inserts from clones were excised from the vector DNA by digestion with appropriate restriction enzymes, separated in agarose gel electrophoresis, and recovered by electroelution (21). Most of the sequencing was performed by the chemical degradation procedure of Maxam and Gilbert (22). The upstream half of clone pFHl insert was sequenced by the chain terminator method. For that purpose, a 1 Kb EcoRl fragment of pFHl was isolated, digested either with Alul or Haelll and ligated to a Smal digested M13mp9 vector (23), previously treated with calf intestinal phosphatase to prevent its circularization. The ligation mixtures were used to transform competent E.coli JM101 and recombinants were selected as clear plaques by insertional inactivation of the f£ -galactosidase gene (24). Single stranded DNA was prepared by standard procedures (25) and the inserts were sequenced by the method of Sanger et a\ (26) using a "universal" 17-nucleotide long primer (27). 51 nuclease mapping SI nuclease mapping was performed as described by Berk and Sharp (28). Probes were labelled at their 5au96I sites by "filling in" with the Klenow fragment of DNA polymerase I and [ct -' 2P ] dGTP. Probes isolated by polyacrylamide gel electro- phoreais were strand separated according to Maxam and Gilbert (22). 10 ng of total RNA were mixed with the appropriate probe in the presence of 20 (ig of poly(A) and co-precipitated with ethanol. PelleU were resuspended in 12.5 nl of 80% formamide, 10 mM PIPES (pH 6.5), 1 mM EDTA, 0.4 M NaCl, heated at 73°C for 10 min under paraffin oil and hybridized overnight at 52°C (probe I) or 46°C (probe II). The hybrids were diluted with 150 \il of SI buffer (250 mM NaCl, 30 mM NaOAc pH 4.4 1 mM ZnSC^,), digested with 3000 u of SI nuclease (Boehringer) for 60 min at 30°C, mixed with 50 |il of SI stop (15 mM EDTA, 600 ng/ml yeast tRNA), ethanol precipitated and analyzed on 6% polyacrylamide 7 M urea gels. RESULTS AND DISCUSSION DNA sequence analysis Figure 1 shows restriction enzyme maps and strategy of sequencing of five independent cDNA clones, which contiguoualy cover 4843 nucleotides from the poly(A) addition site of human FN mRNA. This represents more than 60% of the estimated length of human FN mRNA (7.9 Kb). As previously reported (14), clones pFH37 and 5855 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research pFHl o SxiMI & Hin f l • Dd« I pFH23 pFH111 PFH37 PFH154 Figure 1 Restriction enzyme map and strategy for determining the nucleotide sequences of the inserts of five FN cDNA clones. The 3' end of the map is the 3' end of FN mRNA, and the map covers approximately two-thirds of the FN mRNA molecule. Wavy lines represent vector sequences. Clone inserts are aligned to show overlapping areas. Thick and thin lines represent the coding and non-coding region respectively. ^ ^ indicates contiguity. ED, extra domain. •-» and •-» indicate extent and direction of each sequence analysis performed according to Maxam and Gilbert (22) and Sanger et al (26) respectively. pFH154 lack a 270 bases long internal segment coding for one of the homology type III repeats of FN, called ED for extra domain, while clones pFH23 and pFHll l contain the ED segment. Figure 2 shows the complete nucleotide sequence and its deduced amino acid sequence of the clones depicted in Figure 1. It contains the complete 3' non-coding region (694 nucleotides) and 4149 nucleotides of coding region, accounting for 1383 amino acids of the C-terminal part of FN. A normal polyadenylation signal AATAAA (29) is present at position 4846. From nucleotide 1593 and downstream, the sequence In Figure 2 is homologous to the one reported for rat liver FN mRNA (16), 5856 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research J ; ora TO»CTT T , O A ^ 70 90 00 10O 110 1» TIUTCAATOAAACTIMTTC T AC TO H X 130 14© 190 34O 330 MO T T I W I T U T TOCAOCCTCaOAOCTCT ATTCCACCT T At CTOAOACCACCATCQ TQATCACATOtWCOl *TTQQTT T T AAOCTQQOTI 37 0 390 40 0 410 420 430 43 0 4K0 33 0 300 37O 3*0 10*0 1100 1110 1120 U3O lit© 1170 11*0 11M 1200 ACTTTTCTOAT ATT AC TOCCAACTC T TT TAC TO TOCAC TOOA T TQCTCC TCOAOCCACCATC ACTQOCT ACAOOATCCOCCATCATOXOAOCACrny«nOOCUW«XTCaAa**OATC 1210 1220 1Z30 1240 1230 12S0 1270 12*0 12*0 13OO 1310 13X0 Trr*ftn^ 1330 1340 1330 1380 1370 13*0 1390 14O0 1410 1420 143O 1440 C *ACAO TT TC TM TO TTCCQAOOOACCT DOAA QT TQTT OC TOC OACCCCCACCAOCC T AC TOATCAOCTOOOATllCriXCTQCTQTCACAOTOAtlATATTACAaQATCACTTAaiQAQAAA 1430 1460 1470 14M 14»0 13O0 1910 1320 1330 134O 133O 13S0 CAOaAOCWWWTAOCCCTOTCCAnQAOTTCACTOTOCXTtroOACICAAOTCTACAnCTAa^TCAaailtfCTTAAAC^^ 1370 13S0 1390 l«00 1110 1020 1*30 1*40 1B30 1H0 1B7O l»0 T AQTOTCAAaTOQC TOCCT T 1W0 1700 1710 172O 1730 1740 1730 17*0 1770 1790 17 IB10 1920 4O 1130 11S0 l«70 1B90 ISM l«O0 ltlO lt20 ITJO 1540 1*30 lt 1*70 1BSO lBSO 1000 1010 Z020 203O 2O4O TCOATTCCATCAMAT Ti 2O30 2Ofl0 ZCT^ 20*0 20*0 210O 2110 1120 2130 2140 2150 21C0 Z17O 21tO 21*0 23O0 2310 2320 2330 234O 233© 23*0 2370 2390 2400 TQAAAQAAArCAAO.I 101.ILL IIWaMCTCATCCQTttOTTOTATCAaOACTTATI 5857 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research JITC*CC*T T«OCT(XM(IAMXAMACTIMfMC0ATCACTtXlC TTCC R3 0 X34O 2330 13*0 2370 0*0 I3»0 ZSOO 2*10 2*20 2*30 2*40 VOAUPANOOT P [ d ft T I H PDUtlYT I TQL9P0TDY I IYLY T A«.I I IIIA HIT m i I M A<if r AAT^ nrr iflA f TTTA^TrrAflAflflArrATTA<>nrrww>TBTrwflrtflflrTf>rArTfiTrfif ftiMii t n'H tut i wnrf u i nw iflf ft^rift i r Trtrr T" T*^ * 2*30 2**0 2*70 2**0 TOO m>0 2710 2720 2730 2740 2730 17*4 LH9HASIIPVV 1 DAlT A IDAPlWLRFLATTPMELLViM9 P reTTQA*TnACJM»TOCTC3OAaCTgXCraTgCTCATCgACC^KXj«TQCCAT^ r770 27*0 27*0 Z*O0 Jflfl HBO 2S30 2*40 2*30 236O 2170 28*0 TWO 2C00 W1C 2*20 2t30 2B4© ZB3C 2B0O 217O ZOCO 2*»0 3OO0 3020 3030 7040 3030 30C0 307O 30*0 30*0 3100 3110 3120 iC TOOAAATOOT A TTCAOCTTCCTOOCACT TCTQQTCAOCAACCCA 3130 3140 3130 3110 3170 31*O 31B0 32O0 3210 322O 3230 3240 3230 32*0 3270 32B0 3200 3300 3310 3320 3330 3340 333O 3360 3370 33*0 33*0 3400 3410 342O 3430 3440 3430 34*0 3470 34*0 ATLTOLTROATY N I IVEALKDaQRHKVRECVVTVaHlVM K Q rOCCtfTCTO«r\aOCTTCiWXAfl*tt]TOCC*CCTACAACATC*Tft«TOQ^ WO 3300 3310 3320 333O 3340 3330 33S0 337O 3310 33*0 3*00 rtrftr«TrTrrrftTTftTorniTTmwy>TqftOTnftnr>ftrn^ 3110 3*20 3«3O 3040 3630 3S*0 3070 3**0 3**O 3 TOO 3710 3720 FQeaHPftCDIflRMCHOMDUNYKIQEKHDRBOINOgnni C rrrrOMAdTOOTCA rrrCAOA TQni*TTCATC T AOATOOTQCC A TOACAATQaTOT 0A«rrACAi*aATTlH»ClAaAAaTOaOACCaTCAOOOAaA*UWT0OCCAOATl»ATOAOCT 373C 3740 373O 37SO 377O 37*0 37*0 3S0O 3S1O 31120 3*30 3*40 CLaMaxaiFKCDfHEATCYDOQKTYHUOCQuaKIYLQA I rTTftTBAorirwirnTnTTftniftTfWTTmivw^ofrftTftrcftfTrrftowwu'f WToncftrwftnnrviTi'iTrTrnnTfTrTft 3*3 0 39*0 3S7O 3**O 3**O 3*OO 3*10 3*20 3*30 3*40 3*3O 3*90 CS C TCFQQQRQUftCD H CrtffPQQEPSPGQTTQQSYMGYSS R TTTHCTCCTaCACATOCTTTOOAOaCCAQCtmOdC I OttCOCTQT(lAC*ACTqCCOCJ«ACCTaDQ(MTQA<»CCCJMTCCCOA*OOCACTACTCaCCAO 3170 3S*0 3**0 40OO 4010 4020 4O3O 4040 4030 40S0 4O70 40*0 YHQRTMTHVHC P ICCFRPLDVOADRIDIR I OATACC*TaMAaA*CAAJtfACTAaTgTTA*TTacCCA*TTnj«lT0CTTC^TgC?TTTAOAT0TAC*QOCTOtt 4C*0 4100 4:10 4120 4i±O 4140 4190 41*0 4170 41*0 41*0 4200 * /^* ffl^ TT*TTTfTrrriminTctriTrTftiVri^fTniMrmnTorr* wmmrr r 4Z10 4220 4230 424O 4Z30 42*0 4270 42*0 42*0 43O0 4310 4320 C ACAOCTTCTCCAAOCATCACCC TOanAOrTTC CTOAOaOT T TTCTCA T ArfMTllAOQaCTOCi^ 433O 4340 4330 43O0 4370 438O 43*0 4400 4410 4420 4430 4440 ••"'"'"TTnTTTTTTfflnTrnnTrtmriviTTrnTnTnrnm^ ^ 4430 44SO 4470 44*0 44SO 43O0 4310 4320 433O 4340 4330 43*0 TA*OTOTCTtmCCCaCA*TACTirrACnA«CA*qaaaATCTTflTTACTQTMTAT^ 437O 43*0 43*0 4«00 «10 4*20 4*30 4*40 4*30 4*S0 4*70 4**0 TTiVTTTTTTftTTTnTTWTTTTTCtTflin'nTT^TKTnCtliVllVwVlrtTTqTftTTfWV^ 4*»O 470© 4710 4720 473O 4740 4730 47*0 4770 47*0 47*0 4*O0 4110 4*20 4*30 4940 «&" 4CS0 4*70 4*90 4*90 Figure 2 Nucleotide sequence and predicted amino acid sequence (top row) resulting from the sequencing of the five overlapping FN cDNA clones depicted in Figure 1. Part of this sequence (from positions 1728 to 2538, and from 3883 to 4892) has been reported before (11,14). The first 23 nucleotides of the 5' end of clone pFH154 insert (In brackets) have not been translated because we have evidence (not shown) that it is an inverted sequence due to cloning artefacts. Arrows indicate the ends of the 270 5858 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research bases long ED segment (position 1999 and 2268) and of the 267 bases long IIICS segment (positions 3082 and 3348). Nucleotide 2525 is a T in clones pFH23 and pFHll l but an A in pFH 37 and pFH154. The following sequences were underlined: polyadenylation signal (positions 4846 to 4851), region complementary to the synthetic oligonucleotide used as a probe in the isolation of the bovine cDNA clone which allowed the isolation of the first human FN cDNA clone, pFHl (see 11) (positions 4122 to 4135) and region complementary to the synthetic oligonucleotide U3ed as primer in "mRNA walking" (see Experimental Procedures) (positions 2808 to 2823). A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, He; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gin; R, Arg; S, Ser; T, Thr; V, Val; W, Trp and Y, Tyr. except within the two areas of RNA variability due to alternative splicing, ED (positions 1999 to 2268) and IIICS (positions 3082 to 3348) (see below). The comparable areas within the coding region showed 88% homology at the nucleotide level and 92% homology at the amino acid level. Surprisingly, rat cDNA contains an extra codon (GCT) coding for alanine [position 2260 in Figure 3 of Schwarzbauer et al, (16)] which is not present in the human or the bovine (11) cDNAs. However, this amino acid is located in an area of high interspecies variation whose primary structure is believed not to be critical for protein function. Comparison between rat and human 3' non- coding regions has already been analyzed by Schwarzbauer et al (16) and it will not be discussed here. From nucleotides 1409 to 1718, the sequence shown in Figure 2 is identical to the sequence of a 3hort cDNA clone coding for the cell attachment domain of human FN, pHFN-1, isolated by Oldberg et al (15). These authors used a human fibrosarcoma tumour as source of mRNA, while we used an epithelial cell line derived from a human breast carcinosarcoma (18). Both tissues presumably make cellular FN, and comparison of a stretch of 300 bases does not show any difference between their FN mRNAs. Amino acid sequence and internal homoloqies Figure 3 shows the amino acid sequence deduced from the nucleotide sequence of the human cDNA clones depicted in Figure 1. The 1383 amino acids are aligned to show internal homologies. The sequence spans the C-terminus, fibrin-(Fib-2), heparin- (Hep-2) and cell-binding domains of human FN. The sequence also shows the exact localization of the only two sulfhydryl groups known to exist per subunit chain of FN. One of them (Figure 3, line 4) is the same as the one identified in bovine plasma FN by Vibe-Pedersen et al (30). The second one (line 14) has been predicted by SH titration experiments and it has been mapped in a 31 kilodalton tryptic fragment, near the Fib-2 domain of human plasma FN (31). Furthermore, a peptide of bovine plasma FN containing a second free SH group has been sequenced. The bovine sequence: IISCHPVGIDEEPLGF fits the human sequence in line 14 (Dr T Petersen, personal communication). Two types of internal homologies are represented in Figure 3: three "fingers" or 5859 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 CD o_ (n HoMOtogy to V) tc -^ 3" 0 T - - • lQlQl!l[9<<ITD[IPLQrlTMTITg > - I * tj A LJrft^Q Q I Figure 3 Amino acid sequence deduced from nucleotide sequence shown in Figure 2, The 1383 amino acids are aligned to show internal homologles. Identical residues in the three "fingers" (Iine9 15, 16, and 17) and in the 12i repeats of homoiogy type III are boxed. Gaps (-) are inserted to obtain maximal homoiogy. Arrowheads indicate amino acids conserved in all the type III repeats. Black dots are on top of free sulfhydryl groups. ED, extra domain; IIICS, homoiogy type III connecting segment. Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research repeats of homology type I (lines 15, 16 and 17) and 12£ repeats of homology type III (lines 1 to 12 and 14). The last type III homology (line 14) is separated from the rest of the type III repeats by a stretch of 89 amino acids (line 13), whose sequence is not homologous to any of the three kind3 of homologies. We called this segment IIICS, for homology type III Connecting Siegment. Our results show for the first time that most of FN is formed by type III sequences. The degree of homology within the type III repeats is quite remarkable, except for repeat in line 3 in which only 11 positions are conserved, compared to an average of 30 conserved positions per repeat. Two amino acids, tryptophan and tyrosine, have a conserved position in all the repeats (arrows in Figure 3), as is the case in bovine (10) and rat (16) FNs. Furthermore, the conserved sequences seem to be distributed in two "peak" areas around both tryptophan and tyrosine, separated by a "valley" of low homology. The N-terminal portion of the amino acid sequence in Figure 3 does not overlap with any other published FN sequence. However, it is possible to predict that between the N-terminus of the sequence reported here and the C-terminus of the collagen-binding fragment [partially sequenced in bovine by Skorstengaard et al, (34)], there is room for as many as 2f repeats of homology type III (see Figure 5A). One of these predicted repeats should be contained in the DNA binding domain, since the N-terminal sequence of a fragment with that activity (12) is homologous to type III sequences and it is not present in Figure 3. Primary structure of fibronectin Over the last decade, structural studies on the FN molecule have been performed mainly by the use of proteolytic enzymes in peptide mapping. Only recently, partial but extensive data on primary structure have been obtained by direct sequencing of bovine plasma fibronectin (10). This information has been invaluable for the identification of cloned fragments of DNA containing gene (32) or cDNA (11,14,15,16) sequences for fibronectins. The use of recombinant DNA techniques by us and by other laboratories, has proved to be successful in accelerating the completion of FN primary structure and also in producing new evidence on the existence of multiple FN mRNAs (see below). The amino acid sequence shown in Figure 3 represents the longest stretch of FN sequence reported so far, covering 60% of the molecule. The large size of FN not only imposes difficulties to the "peptide sequencers", but also to the "DNA cloners", who pursue the "full length" cDNA. In fact, current protocols rarely yield cDNA clones as long as 8.0 Kb (the size of FN mRNA) and "mRNA walking" strategies (33) have to be used. The sequence in Figure 3, together with the amino acid sequence of the 29 kilodalton N-terminal domain of human plasma FN (13) account for about 80% of the human FN molecule. If we also consider the protein sequences of the collagen-binding area of bovine FN (34), we can draw the model of a typical FN 5861 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research polypeptide, regarding the three types of internal homologies (Figure 5A). it contains, from N- to C-terminus, 6 repeats of type I, 2 repeats of type n, 3 repeats of type I, 13 to 15 repeats of type III, 3 repeats of type I homologies and a COOH-terminal non homologous stretch where the disulfide bridges that hold FN monomers together are located. Figure 5A alao shows the distribution of the most relevant binding activities along the molecule, as well as the location of the only two free sulfhydryl groups present per FN monomer. Two different FN mRNAs with cell-specificity In their expression Two different FN mRNA species have been identified in the human cell line H8578T (14). One species (mRNAI) contains an additional 270 nucleotide insert (ED), while the other (mRNAII) lacks it. In "Northern" blots of total RNA from Hs578T cells and normal fibroblasts, a specific probe containing only ED sequences hybridizes to bands of the same size as mature FN mRNA ( ~ 7.9 Kb) (results not shown). This indicates that mRNAI is not an unspliced long precursor of mRNAII. Furthermore, the polypsptide coded by the ED is a homology type III repeat in itself (Figure 3, line 9), and it is located between the cell- and heparin-binding domains. We have put forward the idea that the presence of the ED in an FN polypeptide may reflect one of the differences between cellular and plasma FNs. This is based on the fact that while the cell line Hs578T (used for the isolation of the clones in Figure 1) produced both mRNAs, liver tissue contained none or undetectable levels of mRNAI, the ED-bearing species. Since liver has been reported to be the source of plasma FN (35), our results indicated that the presence of the ED could be a particular feature of cellular FN. This conclusion may also be supported by a series of results obtained by other laboratories and summarized as follows: (1) The ED is not present in two independent FN cDNA clones from rat liver (A rlf3 and X rlf6), isolated by Schwarzbauer et al (16). (2) The ED has not been seen at the polypeptide level in bovine plasma FN [segment 6 in Petersen et al, (10)1 (3) Hayashi and Yamada (36) showed that, among other differences, heparin-binding fragments are 10,000 daltons larger in cellular than in plasma chicken FNs. The size (10,000 daltons = 90 amino acids) and location of this polypeptide are coincident with those of the ED. (4) A difference region between human cellular and plasma FNs has been seen in the same area as we see the extra domain (Drs S Akiyama and K Yamada, personal communication). To test this idea further, we have investigated the expression of both mRNAs in different cell types, by SI nuclease mapping of cDNA/RNA hybrids. Two probes were prepared: probe I was the 446 nucleotide 3' end-labelled anti-sense strand 5au961-HaeII of clone pFH23 insert and probe II was the 176 nucleotide 3' end-labelled anti-sense strand Sau96I-HaeII of clone pFH37 insert. Total RNA from the following human cells was analyzed: epithelial cell line Hs578T, derived from a breast carcinoaarcoma; cell 5862 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research EHF - LE EHF-LE • • Sau9fc : liae II ..„ PROBE I 3 ' =>— s 446 PROBE II 3 Figure 4 SI nuclease analysis of RNA/cDNA hybrids. Probe I was the Sau96I-HaeII fragment of pFH23 insert and probe II was the 5au96I-HaeII fragment of pFH37 insert. Probes were hybridized to human RNAs from the following sources: E, epithelial tumour cell line Hs578T; H, hepatoma cell line Hep 3B; F, RJK 735 fibroblasts and L, liver. - control without RNA in the hybridization mixture. Probe fragments, SI nuclease-resistant DNA products and mRNA species are shown in the diagram below the Figure. Correspondence between diagrams and gels are indicated by letters. Arrows indicate the expected position of bands A and D in liver RNA. 5863 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research line Hep 3B, derived from a hepatocellular carcinoma; RJK 735 normal fibroblasts and total liver. When probe I was hybridized to RNA from Hs578T, Hep 3B or RJK 735 cells, two SI nuclease-reslatant DNA fragments were detected (Figure 4, products A and B). Product A (446 nucleotides) represents the fully protected probe and indicates the existence of the mRNA species containing the ED (mRNAI). Product B (119 nucleotides) exactly covers the distance between the Sau961 site and the base in which mRNAs I and II diverge. When probe II was used, again two SI nuclease-resistant products were observed (Figure 4, products C and D). Band C (176 nucleotides) is the fully protected probe, this time indicating the existence of mRNA molecules (mRNAII) lacking the EDj while band D, co-migrating with band B, again represents the distance between the labelled end of the probe and the point of divergence between the two types of RNAs. When similar SI analysis was carried out with human liver RNA, only bands B and C were detected (Figure 4, L lanes). Bands A and D cannot be distinguished above the background, Indicating the existence only of the mRNA lacking the ED. These experiments showed that, while liver tissue only produces mRNAII, both forms of FN mRNA are produced by cultured cells of various origins. Most interestingly, one of these cell lines (Hep 3B) is of liver origin and it has been shown to synthesize and secrete many of the major human plasma proteins (19). However, we show here that, with respect to FN, Hep 3B cells process their RNA in a different way to that of the parental tissue. This still raises the guestion whether the synthesis of FN mRNA having the ED is a phenomenon restricted to tumour cell lines. The presence of the ED in FN mRNA from normal fibroblasts (Figure 4, lane F) seems to rule out this possibility. If, as discussed above, the ED is a difference sequence characteristic of cellular FN, then we could argue that hepatoma cells, like other cells in culture, are making the cellular form of FN, while liver cells are mainly making plasma FN. These conclusions are limited, because we are comparing RNA from an organ with RNA from cultured cells. However, the observation of cell specificity in the expression of the ED will stand independently of whether it correlates with a change in the environment of the cells (hepatocytes in liver —» hepatocytes in culture) or with a change in the differentiation state (normal liver —» hepatoma cells). Two regions of alternative splicing generate multiple FN mRNAs The two FN mRNAs differing by the ED segment arise by alternative splicing of a common precursor, and the ED segment is represented by one exon in the human genome (Vibe-Pedersen et al, submitted). But surprisingly, the ED is not the only difference sequence found in FN mRNAs. Schwarzbauer et al (16) found another region of variability, 0.8 Kb downstream of the ED, in rat liver FN mRNA. They 5864 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research COOH Coll.gtn DNA Hcparln Fibrin IIICS ED pFHJ l II I II I pFHtti • 0 CM PFHJ7 II L III <T ^ J IIII I WHIM III ArtiO CM Figure 5 A. Model of a typical fibronectin polypeptide with the localization of the internal rfomologies (types I, II and III). Unbroken lines and boxes represent sequences known in human FNs: towards the NH2-terminus, reported by Garcia-Pardo et al (13); towards the COOH-terminus, reported in this paper. Broken lines and boxes represent sequences unknown in human, but known in bovine FN (34). B. Schematic representation of the different FN polypeptides that could arise by the translation of the RNAs generated in the ED region (left) and in the IIICS region (right). The name of the cDNA clones representing the corresponding encoding mRNA species is writte n down to the right of each polypeptide. Cell and Hep name the cell- and heparin-binding domains. /^\ indicates contiguity. reported three FN mRNAs arising from alternative splicing. The three RNAs differ in a region not belonging to any of the three internal homologies, which corresponds to the above defined IIICS. Comparison of the human IIICS nucleotide sequence, reported for the first time in this paper, (Figure 2, positions 3082 to 3348) with the rat liver nucleotide sequences in the same area reveals that the human sequence (present in clone pFHl) constitutes a fourth mRNA species. In fact, the human IIICS encodes an 89 amino acids long polypeptide which is highly homologous with the firs t 89 of the 120 amino acids encoded by the rat IIICS in clone X rlf 2 of Schwarzbauer et al (16). Preliminary evidence shows the existence of more than one mRNA in the human IIICS region (not shown). Figure 5A shows the localization of the two regions of variability 5865 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research along the FN molecule. Figure 5B shows the different kinds of peptides that could arise from the translation of the different human and rat RNAs generated in the ED region (left) and in the IIICS region (right). It seems important to point out that all these differences have been seen at the RNA level and, except for the evidence of polypeptides lacking the ED in bovine plasma FN [segment 6 in Petersen et al, (10)], none of the other alternatives have yet been demonstrated at the protein level. Genomic blot experiments are consistent with the existence of a single FN gene in the human haploid genome (11 and unpublished results). A similar kind of analysis in rat supports this conclusion (16). If there is only one gene, and if we assume that rat and human FN genes are directly comparable, the two regions of alternative splicing could in theory originate eight different species of FN mRNA from a single primary transcript. We wil l attempt to discuss, blending evidence and speculation, if there are any biological activities associated with the difference sequences. With respect to the IIICS area, Schwarzbauer et al (16) have speculated that, among other possibilities, this region might represent a site for interaction with fibrin or for fibronectin-fibronectin binding. Ehrismann et al (37) have mapped the self- association site in a region of horse plasma FN corresponding to the IIICS area and have proposed that the difference in size of the two chains of plasma FN Is due to the absence of a self-association site in one of them. We have already suggested that the ED could be present in cellular but not in plasma FN polypeptides. If so, the presence of the ED may explain particular features of cellular FN, such as its lower solubility, the ability to aggregate and form covalently bound multimers, its higher activity in haemagglutination and restoration of normal morphology in transformed cells (8), and its greater affinity for hyaluronlc acid (38). It is not clear whether these abilities reside in a differential sequence of cellular FN or whether they are consequences of its aggregated state. It may be that the differential biological activity resides in the ED itself, in which case, one should look at stretches of non conserved amino acids, compared to other type III homology repeats. Since the ED polypeptide does not contain any cysteine, It cannot be responsible for multimer formation. Alternatively, the ED can be increasing the distance between binding sites or active centres, interfering with or creating biologi- cal activities. Whatever role is played by the ED polypeptide, our results show that the expression of its gene counterpart is cell-specific. Nothing is known about cell- specificity in the IIICS area and since, in our experiments, the ED and IIICS areas are covered by different clones, we cannot assess whether there Is a "functional linkage" in their expression. For example, one could speculate that the expression of the ED always correlates with the expression of one or some of the possible segments in the IIICS area. 5866 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research We are just starting to understand this complex system. The analysis of the actual number of alternatively spliced regions and their biological significance requires further extensive studies on primary structure and expression of fibronectin. ACKNOWLEDGEMENTS We thank Dr K Umezawa for helpful discussions, and Drs M Goldman and R. Corte3e for providing the RJK 735 RNA and Hep 3B cells, respectively. K.V-P is a recipient of a Danish National Research Council Fellowship. *To whom reprint requests should be sent REFERENCES 1. Hynes, R.O. and Yamada, K.M. (1982) J.Cell Biol. 95, 369-377. 2. Yamada, K.M. (1983) Ann.Rev.Biochem. 52, 761-799. 3. Vartio, T. and Vaheri, A. (1983) TIBS, , 442-444. 4. Hynes, R.O. and Destree, A. (1977) Proc.Natl.Acad.Sci.USA 74, 2855-2859. 5. Yamada, K.M. and Olden, KJ. (1978) Nature 275, 179-184. 6. Fagan, J.B., Sobel, M.E., Yamada, K.M., de Crombrugghe, B. and Pastan, I. (1981) J.Biol.Chem. 256, 520-525. 7. Smith, H.S., Riggs, J.R. and Mosesson, M.W. (1979) Cancer Res. 39, 4138-4144. 8. Yamada, K.M. and Kennedy, D.W. (1979) J.Cell Biol. 80, 492-498. 9. Atherton, B.T. and Hynes, R.O. (1981) Cell 25, 133-141. 10. Peteraen, T.E., ThiSgersen, H.C., Skorstengaard, K., Vibe-Pedersen, K., Sahl, P., Sottrup-Jensen, L. and Magnusson, S. (1983) Proc.Natl.Acad.Sci.USA 80, 137-141. 11. Kornblihtt, A.R., Vibe-Pedersen, K. and Baralle, F.E. (1983) Proc.Natl.Acad.Sci.- USA 80, 3218-3222. 12. Pande, H. and Shively, J.E. (1982) Arch.Biochem.Biophys. 213, 258-265. 13. Garcia-Pardo, A., Pearlstein, E. and Frangione, B. (1983) J.Biol.Chem. 258, 12670-12674. 14. Kornblihtt, A.R., Vibe-Pedersen, K. and Baralle, F.E. (1984) EMBO Jnl. 3, 221- 15. Oldberg, A., Linney, E. and Ruoslahti, E. (1983) J.Biol.Chem. 258, 10193-10196. 16. Schwarzbauer, J.E., Tamkun, J.W., Lemischka, I.R. and Hynes, R.O. (1983) Cell 35, 421-431. 17. Skorstengaard, K., Thtfgersen, H.C., Vibe-Peder3en, K., Peterson, T.E. and Magnusson, S. (1982) Eur.J.Blochem. 128, 605-623. 18. Hackett, A.J., Smith, H.S., Springer, E.L., Owens, R.B., Nelson-Rees, W.A., Riggs, J.L. and Gardner, M.B. (1977) J.Natl.Cancer Inst. 58, 1795-1800. 19. Knowles, B., Howe, C.C. and Aden, D.P. (1980) Science 209, 497-499. 20. Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J. (1979) Biochemistry .18, 5294-5299. 21. Girwitz, S.C., Bacchetti, S., Rainbow, A.J. and Graham, F.L. (1980) Anal.Bio- chem. 106, 492-496. 22. Maxam, A.M. and Gilbert, W. (1980) Methods Enzymol. 65, 499-580. 23. Messing, J. and Vieira, J. (1982) Gene 19, 269-276. 24. Messing, J. , Crea, R. and Seeburg, P.H7T1981 ) Nucl.Acids Res. 9, 309-321. 25. Winter, G. and Fielda, S. (1980) Nucl.Acids Res. 8, 1965-1974. 26. Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc.Natl.Acad.Sci.USA 74, 5463-5467. 5867 Downloaded from https://academic.oup.com/nar/article/12/14/5853/1204054 by DeepDyve user on 20 August 2020 Nucleic Acids Research 27. Duckworth, M.L., Gait, M.J., Goelet, P., Hong, G.F., Singh, M. and Titmas, R.C. (1981) Nucl.Acida Res. 9, 1691-1706. 28. Berk, A.J. and Sharp, P.A. (1977) Cell 12, 721-732. 29. Proudfoot, N.J. and Brownlee, G.G. (1976) Nature 263, 211-214. 30. Vibe-Pedersen, K., Sahl, P., Skorstengaard, K. and Petersen, T.E. (1982) FEBS Lett. 142, 27-30. 31. Smith, D.G., Mosher, D.F., Johnson, R.B. and Furcht, L.T. (1982) J.Biol.Chem. 257, 5831-5838. 32. Hirano, H., Yamada, Y., Sullivan, M., De Crombrugghe, B., Pastan, I. and Yamada, KM. (1983) Proc.Natl.Acad.Sci.USA 80, 46-50. 33. Lawn, R.M., Adelman, J., Bock, S.C., Franke, A.E., Houck, CM. , Najarian, R.C, Seeburg, P.H. and Wion, K.L. (1981) Nucl.Aclds Res. 9, 34. Skorstengaard, K., Thrfgersen, H.C. and Petersen, T.E. (1984) Eur.J.Biochem., in press. 35. Tamkun, J.W. and Hynes, R.O. (1983) J.Biol.Chem. 258, 4641-4647. 36. Hayashi, M. and Yamada, K.M. (1981) J.Biol.Chem.~236, 11291-11300. 37. Ehrismann, R., Roth, D.E. Epponberger, H.M. and Turner, D.C. (1982) J.Biol.- Chem. 257, 7381-7387. 38. Laterra, J. and Culp, L.A. (1982) J.Biol.Chem. 257, 719-726.

Journal

Nucleic Acids Research – Oxford University Press

Published: Jul 25, 1984

There are no references for this article.