Cell type specific trans-acting factors are involved in alternative splicing of human fibronectin pre-mRNA.

ED-A and ED-B are facultative type III homologies of fibronectin, encoded by alternatively spliced exons, described in man and in rat. A hybrid alpha-globin-fibronectin minigene containing the ED-B region from the human gene has been transfected in human cell lines derived from various tissues, in order to study the processing of the generated precursor RNA in the different cell environments. In most tested lines the pre-RNA is alternatively spliced and produces two mature RNAs, with and without the ED-B exon, in different ratios that closely resemble the corresponding endogenous fibronectin RNAs. In a hepatoma cell line, Hep 3B, only one RNA is produced, in which the ED-B exon is absent; the same pattern of splicing is observed in liver. The data show that all the information required to produce accurate and regulated alternative splicing of the ED-B exon is contained in the fragment used and cell specific factors are necessary for the pre-RNA to be differentially spliced in the various cell lines. In contrast, expression in Hep 3B of a similar gene containing the ED-A area failed to reproduce the liver specific splicing pattern. Therefore regulation of ED-A processing is likely to involve different mechanisms to those responsible for control of ED-B splicing.     

Sequence analysis and in vivo expression show that alternative splicing of ED-B and ED-A regions of the human fibronectin gene are independent events.

The structure of two alternatively spliced regions. ED-A and ED-B, of human fibronectin gene, was determined, in order to show whether any similarity was present between the two. Although some interesting features are present in each, no obvious common structure or sequence homology was found. Functional analysis of the alternative splicing events was carried out by transient expression in Hela cells. A hybrid gene was constructed by inserting the ED-B region into the third exon of the human alpha 1-globin gene. The transfected hybrid gene is expressed and produces, in Hela cells, two alternatively spliced RNAs, showing a pattern very similar to that observed for the endogenous fibronectin gene in fibroblasts. Cotransfection of this gene with a similar gene containing the ED-A region, shows that no interference is present between the two alternative splicing processes.     

The type III-9 repeat of human fibronectin is encoded by a single exon which is not alternatively spliced.

Type III homologies of human fibronectin are generally encoded by two exons, with the exception of the ED-A and ED-B repeats which are encoded by a single exon undergoing alternative splicing. We report that also the type III-9 homology is encoded by a single exon. Further more, RT-PCR analysis, performed on mRNA purified from fetal and adult tissues and from normal and tumor-derived cell types, showed that the III-9 region is not undergoing alternative splicing in all samples tested.     

Patterns of alternative splicing of fibronectin pre-mRNA in human adult and fetal tissues

Alternative splicing of fibronectin pre-mRNA at two distinct regions, termed ED-A and IIICS, was investigated with human adult and fetal tissues by the nuclease S1 protection assay. A clear tissue specificity was observed in the splicing pattern at the ED-A region. More ED-A+ than ED-A- mRNAs were identified in lung, whereas ED-A- mRNAs were predominantly expressed in liver. Endometrium contained nearly equal amounts of ED-A+ and ED-A- mRNAs. The splicing pattern at the ED-A region was also different between adult and fetal liver but not between adult and fetal lung. Tissue type specific splicing was also observed at the IIICS region. Although the mRNA species containing the complete IIICS sequence comprised 40-65% of the total fibronectin mRNAs irrespective of tissue types, expression of the mRNA species lacking a part or all of the IIICS sequence was more pronounced in adult liver than in other tissues including fetal liver. These results strongly suggest that the alternative splicing of fibronectin pre-mRNA in vivo is regulated in a tissue type specific manner at both the ED-A and IIICS regions and that it is developmentally regulated in liver but not in lung. On the basis of these and other observations reported previously, a possibility that a part of the fibronectins synthesized and secreted by hepatocytes is deposited in the tissue matrix is discussed.     

Multiple sites of alternative splicing of the rat fibronectin gene transcript

We describe analyses of the structure and expression of the rat fibronectin gene with particular attention to the 40-kb stretch from the center of the gene which encodes 17 type-III repeating units. Each repeat is precisely separated from its neighbors by introns and most are encoded by pairs of exons. Three repeats are encoded precisely by single exons and two of these (EIIIA and EIIIB) are alternatively spliced in a cell type-specific fashion. A third site of alternative splicing (EIIIB) reported here is similar in expression to the previously described EIIIA segment. Both are excluded from mRNA in liver cells and are, therefore, absent from plasma fibronectin. These two alternative splices, plus a third one (V) reported previously, can occur in all possible combinations giving 12 fibronectin mRNAs from a single gene. These splicing variations account for most but not all of the known fibronectin subunit variants. We report investigations designed to detect other regions of alternative splicing. We also show that the pattern of alternative splicing is somewhat altered on oncogenic transformation.     

Deregulation of alternative splicing of fibronectin pre-mRNA in malignant human liver tumors

Alternative splicing of fibronectin pre-mRNA at the ED-A region has been shown to be regulated in a tissue- and developmental stage-specific manner. We investigated the splicing pattern at this region in malignant and nonmalignant human liver tissues and found that the relative population of the fibronectin mRNA containing the ED-A sequence is markedly increased in malignant liver tumors. Nontumorous liver tissues including those with chronic hepatitis and cirrhosis did not show any significant change in the alternative splicing at the ED-A region. It was also found that the increased expression of the ED-A-containing fibronectin mRNA closely correlates with the occurrence of portal vein tumor thrombus and intrahepatic metastasis, which are two characteristic features of invasive liver tumors. These results indicate that tissue-specific alternative splicing of fibronectin mRNA is modified in human liver cancer and raise a possibility that the putative molecular machinery governing alternative RNA splicing of not only fibronectin but also other cellular proteins is deregulated in malignant human tumors.     

A family of fibronectin mRNAs in human normal and transformed cells

Previously, two fibronectin mRNAs, generated by alternative splicing of the extra domain (ED) and type III connecting segments (IIICS) sequences, have been described in a human transformed cell line and in human liver, respectively. We now report on a family of fibronectin mRNAs identified by Northern blotting analysis in two normal human fibroblast strains (HEL 299 and Flow 7000) and five transformed cell lines (8387 and HT-1080, fibrosarcomas; G-361, melanoma; JEG-3, choriocarcinoma; and RD, rhabdomyosarcoma). Seven different fibronectin mRNA forms with electrophoretic mobilities ranging between 8.6 and 7.7 kb were identified. Each cell line contains three (HEL 299, Flow 7000 and 8387) or two (HT-1080, G-361, JEG-3 and RD) fibronectin mRNAs species with characteristic size. In all cell lines we detected one fibronectin mRNA form which lacks the ED sequence (ED- fibronectin mRNA) and one or two fibronectin mRNAs containing it (ED+ fibronectin mRNA). These data show that the presence of ED+ and ED- fibronectin mRNAs is a general feature of all cells tested. Moreover, the fibronectin mRNA pattern is characteristic of the cell type analyzed, suggesting the occurrence of specifically programmed splicing mechanisms in each cell line.     

Expression of fibronectin variants in vascular and visceral smooth muscle cells in development

Monoclonal antibodies recognizing extra domain A (ED-A) and extra domain B (ED-B) fibronectin (FN) sequences were used to characterize FN variants expressed in human vascular smooth muscle cells (SMC) during fetal and postnatal development and to compare spectrum of FN variants produced by vascular and visceral SMC. In 8- to 12-week-old fetuses both ED-A-containing FN (A-FN) and ED-B-containing FN (B-FN) were found in all smooth muscles studied--aorta, esophagus, stomach, and jejunum. By 20-25 weeks of gestation relative amounts of both A-FN and B-FN were reduced significantly in the aortic media (fivefold for A-FN and twofold for B-FN), while in visceral SMC only B-FN content was decreased. All the adult visceral smooth muscles examined contained A-FN rather than B-FN. Therefore, the cells from adult aortic media appear to be the only SMC so far known to produce FN that contains neither ED-A nor ED-B. Moreover, the data obtained show that, unlike other cells, medial SMC are embedded in vivo in the extracellular matrix that contains FN lacking both ED-A and ED-B. SMC from the minor intimal thickenings in the human child aorta as well as those from the atherosclerotic plaques produce A-FN rather than B-FN. We conclude that (1) vascular SMC change the spectrum of produced FN variants at least twice--during prenatal development between 12 and 20 weeks of gestation, and during the postnatal period, when they are recruited into the intimal cell population; (2) the production of FN variants in visceral SMC is also developmentally regulated; (3) all visceral SMC unlike the cells from adult aortic media produce A-FN; (4) the presence of ED-A and ED-B sequences in the FN molecule is not necessary for the extracellular matrix assembly in vivo.     

Human placental fibronectin: demonstration of structural differences between the A and B chains in the extra domain-A region

Fibronectins are a class of cell adhesion proteins produced from a single gene. Soluble plasma fibronectin plays a role in wound healing and the insoluble cellular fibronectin form anchors cells to the substrata. The proteins possess multiple macromolecular binding domains including collagen, fibrin, and heparin. Alternative RNA splicing in at least three regions (ED-A, ED-B, and III CS) is responsible for this fibronectin polymorphism. We have been studying this polymorphism at the protein level in placental fibronectin, a poorly soluble form of cellular fibronectin. Cathepsin D-digested placental fibronectin applied to a heparin-agarose column and eluted with a NaCl stepwise gradient (0.1, 0.3, 0.5 M) gave two polypeptides (80-100 and 65 kDa) in the 0.3 M NaCl peak. Immunoblots with monoclonal antibodies IST-2 (specific for the carboxy-terminal heparin-binding domain) and IST-9 (specific for the ED-A portion of fibronectin) suggest that both peptides contain the carboxy-terminal heparin-binding (Hep-2) domain, but that only the larger fragment possesses the ED-A segment. The two peptides were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis, electrotransferred to Polybrene-coated polyvinyl difluoride membranes, and characterized by microsequence analysis. This analysis confirmed that both fragments start with the same amino acid sequence, 17 amino acids before the start of ED-A. These results demonstrate that placental fibronectin is a heterodimer, structurally distinct from plasma fibronectin due to the presence of a unique domain modification that is not seen in the plasma form.     

Detection and measurement of alternative splicing using splicing-sensitive microarrays

Splicing and alternative splicing are major processes in the interpretation and expression of genetic information for metazoan organisms. The study of splicing is moving from focused attention on the regulatory mechanisms of a selected set of paradigmatic alternative splicing events to questions of global integration of splicing regulation with genome and cell function. For this reason, parallel methods for detecting and measuring alternative splicing are necessary. We have adapted the splicing-sensitive oligonucleotide microarrays used to estimate splicing efficiency in yeast to the study of alternative splicing in vertebrate cells and tissues. We use gene models incorporating knowledge about splicing to design oligonucleotides specific for discriminating alternatively spliced mRNAs from each other. Here we present the main strategies for design, application, and analysis of spotted oligonucleotide arrays for detection and measurement of alternative splicing. We demonstrate these strategies using a two-intron yeast gene that has been altered to produce different amounts of alternatively spliced RNAs, as well as by profiling alternative splicing in NCI 60 cancer cell lines.     

Cis requirements for alternative splicing of the cardiac troponin T pre-mRNA.

The cardiac troponin T (cTNT) pre-mRNA splices 17 exons contiguously but alternatively splices (includes or excludes) the fifth exon. Because both alternative splice products are processed from the same pre-mRNA species, the cTNT pre-mRNA must contain cis-acting sequences which specify exon 5 as an alternative exon. A cTNT minigene (SM-1) transfected into cultured cells produces mRNAs both including and excluding exon 5. The junctions of exons 4-5-6 and 4-6 in the cTNT minigene mRNAs are identical to those of endogenous cTNT mRNAs and no other exons are alternatively spliced. Thus, the SM-1 pre-mRNA is correctly alternatively spliced in transfected cells. To circumscribe the pre-mRNA regions which are required for the alternative nature of exon 5, we have constructed a systematic series of deletion mutants of SM-1. Transfection of this series demonstrates that a 1200 nt pre-mRNA region containing exons 4, 5, and 6 is sufficient to direct alternative splicing of exon 5. Within this region are two relatively large inverted repeats which potentially sequester the alternative exon via intramolecular base-pairing. Such sequestration of an alternative exon is consistent with models which propose pre-mRNA conformation as being determinative for alternative splicing of some pre-mRNAs. However, deletion mutants which remove the majority of each of the inverted repeats retain the ability to alternatively splice exon 5 demonstrating that neither is required for cTNT alternative splice site selection. Taken together, deletion analysis has limited cis elements required for alternative splicing to three small regions of the pre-mRNA containing exons 4, 5, and 6. In addition, the cTNT minigene pre-mRNA expresses both alternative splice products in a wide variety of cultured non-muscle cells as well as in cultured striated muscle cells, although expression of the cTNT pre-mRNA is normally restricted to striated muscle. This indicates that cis elements involved in defining the cTNT exon 5 as an alternative exon do not require muscle-specific factors in trans to function.     

Mitochondrial damage modulates alternative splicing in neuronal cells: implications for neurodegeneration

Mitochondrial damage is linked to many neurodegenerative conditions, such as Parkinson's disease, Alzheimer's disease and amyotrophic lateral sclerosis. These diseases are associated with changes in the splicing pattern of individual mRNAs. Here, we tested the hypothesis that mitochondrial damage modulates alternative splicing, not only of a few mRNAs, but in a general manner. We incubated cultured human neuroblastoma cells with the chemical agent paraquat (a neurotoxin that interferes with mitochondrial function, causing energy deficit and oxidative stress) and analysed the splicing pattern of 13 genes by RT-PCR. For all mRNAs that are alternatively spliced, we observed a dose- and time-dependent increase of the smaller isoforms. In contrast, splicing of all constitutive splicing exons that we monitored did not change. Using other drugs, we show that the modulation of alternative splicing correlates with ATP depletion, not with oxidative stress. Such drastic changes in alternative splicing are not observed in cell lines of non-neuronal origin, suggesting a selective susceptibility of neuronal cells to modulation of splicing. As a significant percentage of all mammalian mRNAs undergo alternative splicing, we predict that mitochondrial failure will unbalance a vast number of isoform equilibriums, which would give an important contribution to neurodegeneration.