Alternative splicing of mouse transcription factors affects their DNA-binding domain architecture and is tissue specific
© Taneri et al.; licensee BioMed Central Ltd. 2004
Received: 28 May 2004
Accepted: 18 August 2004
Published: 30 September 2004
Analyzing proteins in the context of all available genome and transcript sequence data has the potential to reveal functional properties not accessible through protein sequence analysis alone. To analyze the impact of alternative splicing on transcription factor (TF) protein structure, we constructed a comprehensive database of splice variants in the mouse transcriptome, called MouSDB3 containing 461 TF loci.
Our analysis revealed that 62% of these loci in MouSDB3 have variant exons, compared to 29% of all loci. These variant TF loci contain a total of 324 alternative exons, of which 23% are in-frame. When excluded, 80% of in-frame alternative exons alter the domain architecture of the protein as computed by SMART (simple modular architecture research tool). Sixty-eight % of these exons directly affect the coding regions of domains important for TF function. Seventy-five % of the domains affected are DNA-binding domains. Tissue distribution analyses of variant mouse TFs reveal that they have more alternatively spliced forms in 14 of the 18 tissues analyzed when compared to all the loci in MouSDB3. Further, TF isoforms are homogenous within a given single tissue and are heterogeneous across different tissues, indicating their tissue specificity.
Our study provides quantitative evidence that alternative splicing preferentially adds or deletes domains important to the DNA-binding function of the TFs. Analyses described here reveal the presence of tissue-specific alternative splicing throughout the mouse transcriptome. Our findings provide significant biological insights into control of transcription and regulation of tissue-specific gene expression by alternative splicing via creation of tissue-specific TF isoforms.
Alternative splicing is a widespread mechanism involved in regulation of gene expression, which enables production of many structurally and functionally different forms of proteins from a single gene, adding to the complexity of the genomes [1–3]. Different mRNA transcripts of a gene can be expressed in different tissues or developmental stages or physiological conditions [4, 5].
An expanding body of expressed sequence data from the human and mouse genomes indicates that alternative splicing is an important mechanism in creating protein diversity, and adds to functional complexity encoded in eukaryotic genomes. Earlier studies indicate that at least 50% of the genes in the human genome are alternatively spliced . Examples include the vast majority of immune system and nervous system genes .
Comprehensive analysis of alternative splicing is essential to understand fully the proteomes of organisms . Several reports have indicated that variant splice forms result in proteins with different functions. These can range from minimal changes in function to absolutely opposite functions. For example, the cAMP-response element modulator has three different isoforms with entirely different DNA-binding domains, which are all transcription activators. On the other hand, isoforms of the human transcription factor AML1 function both as positive and as negative regulators of transcription . However, for the majority of genes, the functional significance of alternative splicing is still not known .
Transcription is a critical process that specifies the mRNAs and the proteins expressed within a cell. Expression of a given gene is dependent on the interactions of different transcription factors and their cofactors with the regulatory regions of that gene. These transcription factors are in turn regulated by processes that include interaction with other proteins and signaling cascades .
Alternative splicing is a mechanism that regulates transcription factor (TF) activity by generating a variety of protein isoforms from a single gene. As noted by Lopez, alternative splicing can affect TF structure in two primary ways : alterations can be in the DNA-binding domains affecting their affinity or specificity; or alterations can modulate interactions of transcription factors with their cofactors. Such changes have been observed experimentally to alter specificity or binding strength or to switch between activator and repressor isoforms of the same TF . TF isoforms can have stage-specific and tissue-specific expression patterns throughout the development of an organism . Little is known about the tissue specificity of alternative splicing .
In this paper, we use an integrated approach to analyze DNA and protein sequence data jointly to determine the potential effect of alternative splicing on protein structure and function. We perform a detailed analysis of tissue-specific distribution of alternatively spliced mouse TFs to gain biologically meaningful insights into regulation of gene expression by alternative splicing.
For our joint DNA-protein analysis described here, we developed MouSDB3 , which identifies, classifies, computes, stores and answers queries about splice variants within the mouse genome. As described in Materials and methods, MouSDB3 uses the mouse genome and expressed sequences in GenBank  and dbEST  to compute splice variants of mouse transcripts organized by genomic loci. This section provides definitions of terms used in MouSDB3 and in the joint DNA-protein analysis method described here. A 'transcript' is a sequence transcribed from the genomic DNA sequence. MouSDB3 is restricted to transcripts with at least one splice junction. A 'locus' is a genomic region that includes a set of overlapping transcripts mapped to the genome such that a transcript appears in only one locus and all transcripts whose genome coordinates overlap by at least one nucleotide are included in the locus. Within a locus, a 'cassette exon' is completely included in some transcripts and completely excluded in others. A 'length variant exon' has alternative 5' or 3' splice sites, or both, in different transcripts. An exon can be both length variant and cassette. A 'variant exon' is either cassette or length variant or both. We consider an exon whose number of nucleotides is a multiple of three and which starts at the first base of a codon to be an 'in-frame exon'. Such exons do not introduce an amino-acid substitution or a stop codon when skipped, unless they are terminal exons within the coding sequence. A 'genomic exon' is an uninterrupted series of nucleotides, each of which is mapped to a transcript. By this definition the genomic exon for a length variant exon reflects the outermost splice sites. A 'cluster' is the set of transcripts that map to a locus. A 'variant cluster' contains one or more variant exons. An 'invariant cluster' has no variant exons.
MouSDB3 cluster analysis
Cluster analyses of transcription factors and entire MouSDB3
Total number of clusters
Number of invariant clusters
Number of variant clusters
Cassette exon analysis
The 287 variant TF loci contain 324 cassette exons of which 23% (76 exons) are in-frame. Only 11% of cassette exons are expected to be multiples of three and in codon position 1 randomly. The twofold difference between expected and observed numbers indicates a bias towards in-frame cassette exons. The exons which are a multiple of three and in codon position 2 and 3 comprise 10% and 7%, respectively. When deleted, these exons introduce an amino-acid substitution to the sequence. As exons which are a multiple of three starting at codon position 1 are enriched and do not introduce an amino-acid substitution when deleted, our study focuses on these exons only.
Assessing domain architecture alterations
SMART [15, 16] and Pfam [17, 18] entries for the altered domains revealed that 75% of the domains affected by alternative splicing with known functions are DNA-binding domains. The names of all altered domains and links to their annotated biological functions are provided on our web page . There we provide the 53 in-frame cassette exons (shown in Figure 2), which alter the domain architecture of their transcripts when skipped. Links to MouSDB3 clusters containing these transcripts and links to their GenBank entries are provided. In addition, we provide the names of the domains altered by these 53 exons as active links to their SMART and Pfam annotations. All sequences for long transcripts, altered transcripts and in-frame cassette exons are provided as links to fasta files on the same web page. Our domain-alteration results correlate with recent findings of Resch et al. , who show that alternative splicing preferentially removes certain domains more frequently.
Part two of our analysis assessed the tissue distribution of alternatively spliced transcription factors. We chose 18 tissues from the existing libraries in MouSDB3 on the basis of the fact that they contain both variant and invariant transcripts annotated as TFs. There are a total of 1,413 library names in MouSDB3 imported from expressed sequence records in GenBank and dbEST. Of these, 328 are ambiguous in that they list several different tissues or cell types for a single library, such as 'mixture of brain and testis' or no tissues at all, such as 'embryo or carcinoma'. For the work described here we did not include tissue information from such ambiguous libraries. There are a total of 95 libraries in MouSDB3 for which there are TF transcripts. In addition, to account for library ambiguities within these 95 libraries, we pooled different parts of a tissue into one library. For example, the term 'brain' corresponds to all parts of the brain found in MouSDB3, including cerebellum, thalamus, hippocampus and 16 other libraries. When analyzing the tissue distribution of all genes, only the libraries that contain TF transcripts have been used.
Figures 5 and 6 show that the majority of TF isoforms and the isoforms of all alternatively spliced genes differ across tissues: within a given single tissue there generally is only one isoform. These data indicate the presence of tissue-specific alternative splicing throughout the mouse transcriptome. In addition, our findings indicate expression of different TF isoforms in different tissues. This implies contribution of alternative splicing to regulation of gene expression in a tissue-specific manner by controlling activation or repression of different sets of genes in different tissues via variant TF isoforms. These data have significant implications in further understanding the regulation of tissue-specific gene expression and control of transcription.
Through integrated analyses of DNA and protein sequences for TF genes, we show that alternative splicing of TFs are more prevalent in the entire mouse transcriptome and in specific tissues when compared to alternatively spliced forms of all the genes. In 78% of the tissues analyzed, higher proportions of TFs exhibit alternative splicing compared to all the genes in the mouse transcriptome. This result, along with the finding that 62% of TF loci are variant, indicates the widespread impact of alternative splicing on regulation of TF function.
We also show that alternative splicing changes TF structure by adding or deleting domains. This study reveals that 80% of alternatively spliced TFs have different domain architectures due to introduction of an in-frame cassette exon by alternative splicing. Of the altered domains, 75% have a role in DNA binding. These findings provide quantitative evidence for the role of alternative splicing in controlling the presence of domains in the proteins. They also suggest that alternative splicing might regulate TF activity by changing the architecture of the DNA-binding domains of these proteins.
Our analyses revealed that within a single tissue there generally is only one TF isoform, and that across tissues, isoforms differ. This finding indicates tissue specificity of alternatively spliced TFs and suggests that TFs might regulate gene expression in a tissue-specific manner by having different isoforms in different tissues. These findings further indicate the role of alternative splicing in regulation of tissue-specific gene expression. Activation and repression of different sets of genes within different tissues can be regulated through variant TF isoforms created by alternative splicing. These findings will significantly aid further understanding of control of transcription and tissue-specific gene expression. In addition, our study shows that all variant loci in the mouse transcriptome display isoform homogeneity within single tissues and heterogeneity across tissues. This finding demonstrates the presence of tissue-specific alternative splicing across the mouse transcriptome and greatly expands the knowledge on the tissue specificity of alternatively spliced genes.
Overall, our study provides quantitative evidence for the effect of alternative splicing on protein structure and sheds light on how alternative splicing might regulate transcription factor function in a tissue-specific manner. This, in turn, reveals the contribution of alternative splicing to regulation of gene expression via tissue-specific TF isoforms. The work described here implies that future high-throughput screens of gene expression analyses should be sensitive to multiple alternatively spliced forms of TFs. Because gene-expression arrays are intended to measure transcription, the next generation of arrays should contain probes specific to all known isoforms of genes represented on the arrays. Given that alternatively spliced exons are highly conserved across species [21, 22], it would be of further interest to extend this study to other organisms. Strong sequence homology between mouse, human and rat exons suggests that a comparative analysis of human, mouse and rat TF variations will be a natural extension of the studies described here.
Materials and methods
Development of the alternative splicing database MouSDB3
For this analysis, we constructed a database of alternatively spliced mouse transcripts called MouSDB3 , using the methods described in . Briefly, full-length transcript nucleotide sequences were obtained by an Entrez query on 5 August 2003 from GenBank  with molecule selected as mRNA and limits used to exclude expressed sequence tags (ESTs), sequence-tagged sites (STSs), genome sequence survey (GSS), third-party annotation (TPA), working draft and patents. EST sequences were downloaded on 31 July 2003 from dbEST  by extracting only Mus musculus entries. All expressed sequences were mapped to a region of the University of California Santa Cruz (UCSC) February 2003 version mm3 of the mouse genome assembly using BLAT . BLAT tools gfServer and gfClient were installed from jksrc444 dated 15 July 2002 . This was followed by a careful alignment by SIM4  version 3/3/2002 to establish splice sites of exons. A post-processing analysis computed genomic exons and determined types of variation for each exon, transcript and locus.
Cassette exon analysis
We identified in-frame cassette exons and extracted from MouSDB3 nucleotide and amino-acid sequences for transcripts containing these exons. The selected amino-acid sequences were then analyzed with SMART [29, 30] to compute protein-domain architecture for each transcript within a cluster.
Tissue distribution of alternatively spliced TFs
From MouSDB3, we then extracted library information for the transcripts within clusters and their annotations. We used these data to compute the tissue distribution of variant transcripts as reported in Results. All scripts and README files used to carry out this data-gathering process are available upon request from the Laboratory of Computational Genomics of The Rockefeller University.
We acknowledge support from Mathers Foundation and Hirschl Foundation. This work has been partially funded by NSF grant DBI9984882 and NIH grant GM62529 to T.G. We thank Joseph A. Sorge for suggestions regarding the tissue-distribution analyses and members of Laboratory of Computational Genomics for their support. Corresponding author T.G. can be reached at firstname.lastname@example.org as well as at email@example.com.
- Caceres JF, Kornblihtt AR: Alternative splicing: multiple control mechanisms and involvement in human disease. Trends Genet. 2002, 18: 186-193. 10.1016/S0168-9525(01)02626-9.PubMedView ArticleGoogle Scholar
- Brett D, Popisil H, Valcarel J, Reich J, Bork P: Alternative splicing and genome complexity. Nat Genet. 2002, 30: 29-30. 10.1038/ng803.PubMedView ArticleGoogle Scholar
- Black DL: Protein diversity from alternative splicing: a challenge for bioinformatics and post-genome biology. Cell. 2000, 103: 367-370. 10.1016/S0092-8674(00)00128-8.PubMedView ArticleGoogle Scholar
- Graveley BR: Alternative splicing: increasing diversity in the proteomic world. Trends Genet. 2001, 17: 100-107. 10.1016/S0168-9525(00)02176-4.PubMedView ArticleGoogle Scholar
- O'Donovan KJ, Darnell RB: Neuronal signaling through alternative splicing: some exons CaRRE. Sci STKE. 2001, PE2-Google Scholar
- Modrek B, Lee C: A genomic view of alternative splicing. Nat Genet. 2002, 30: 13-19. 10.1038/ng0102-13.PubMedView ArticleGoogle Scholar
- Modrek B, Resch A, Grasso C, Lee C: Genome-wide detection of alternative splicing in expressed sequences of human genes. Nucleic Acids Res. 2001, 29: 2850-2859. 10.1093/nar/29.13.2850.PubMedPubMed CentralView ArticleGoogle Scholar
- Cline MS, Shigeta R, Wheeler RL, Siani-Rose MA, Kulp D, Loraine AE: The effects of alternative splicing on transmembrane proteins in the mouse genome. Pac Symp Biocomput. 2004, 17-28.Google Scholar
- Lopez AJ: Developmental role of transcription factor isoforms generated by alternative splicing. Dev Biol. 1995, 172: 396-411. 10.1006/dbio.1995.8050.PubMedView ArticleGoogle Scholar
- Foulkes NS, Sassone-Corsi P: More is better activators and repressors from the same gene. Cell. 1992, 68: 411-414. 10.1016/0092-8674(92)90178-F.PubMedView ArticleGoogle Scholar
- Xu Q, Modrek B, Lee C: Genome-wide detection of tissue-specific alternative splicing in the human transcriptome. Nucleic Acids Res. 2002, 30: 3754-3766. 10.1093/nar/gkf492.PubMedPubMed CentralView ArticleGoogle Scholar
- MouSDB3. [http://genomes.rockefeller.edu/autodb/sdb.php?db=MouSDB3]
- Bilofsky HS, Burks C, Fickett JW, Goad WB, Lweitter FI, Rindone WP, Swindell CD, Tung CS: The GenBank genetic sequence databank. Nucleic Acids Res. 1986, 14: 1-4.PubMedPubMed CentralView ArticleGoogle Scholar
- Boguski MS, Lowe TM, Tolstoshev CM: dbEST - database for 'expressed sequence tags'. Nat Genet. 1993, 4: 332-333. 10.1038/ng0893-332.PubMedView ArticleGoogle Scholar
- Schultz J, Milpetz F, Bork P, Ponting CP: SMART, a simple modular architecture research tool: Identification of signaling domains. Proc Natl Acad Sci USA. 1998, 95: 5857-5864. 10.1073/pnas.95.11.5857.PubMedPubMed CentralView ArticleGoogle Scholar
- Letunic I, Copley RR, Schmidt S, Ciccarelli FD, Doerks T, Schultz J, Ponting CP, Bork P: SMART 4.0: towards genomic data integration. Nucleic Acids Res. 2004, 32 Database issue: D142-D144. 10.1093/nar/gkh088.View ArticleGoogle Scholar
- Bateman A, Coin L, Durbin R, Finn RD, Hollich V, Griffiths-Jones S, Khanna A, Marshall M, Moxon S, Sonnhammer EL, Studholme DJ, et al: The Pfam protein families database. Nucleic Acids Res. 2004, 32 Database issue: D138-D141. 10.1093/nar/gkh121.View ArticleGoogle Scholar
- Pfam. [http://www.sanger.ac.uk/Software/Pfam/index.shtml]
- Supplementary Web Page. [http://genomes.rockefeller.edu/~bahar/TF.html]
- Resch A, Xing Y, Modrek B, Gorlick M, Riley R, Lee C: Assessing the impact of alternative splicing on domain interactions in the human proteome. J Proteome Res. 2004, 3: 76-83. 10.1021/pr034064v.PubMedView ArticleGoogle Scholar
- Sugnet CW, Kent WJ, Ares M, Haussler D: Transcriptome and genome conservation of alternative splicing events in humans and mice. Pac Symp Biocomput. 2004, 66-77.Google Scholar
- Thanaraj TA, Clark F, Muilu J: Conservation of human alternative splice events in mouse. Nucleic Acids Res. 2003, 31: 2544-2552. 10.1093/nar/gkg355.PubMedPubMed CentralView ArticleGoogle Scholar
- Zavolan M, van Nimwegen E, Gaasterland T: Splice variation in mouse full-length cDNAs identified by mapping to the mouse genome. Genome Res. 2002, 12: 1377-1385. 10.1101/gr.191702.PubMedPubMed CentralView ArticleGoogle Scholar
- Entrez nucleotide. [http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=Nucleotide]
- FTP directory/repository/dbEST. [ftp://ftp.ncbi.nih.gov/repository/dbEST]
- Kent WJ: BLAT - the BLAST-like alignment tool. Genome Res. 2002, 12: 656-664. 10.1101/gr.229202. Article published online before March 2002.PubMedPubMed CentralView ArticleGoogle Scholar
- Jim Kent's web page. [http://www.cse.ucsc.edu/~kent]
- Florea L, Hartzell G, Zhang Z, Rubin GM, Miller W: A computer program for aligning a cDNA sequence with a genomic DNA sequence. Genome Res. 1998, 8: 967-974.PubMedPubMed CentralGoogle Scholar
- Letunic I, Goodstadt L, Dickens NJ, Doerks T, Schultz J, Mott R, Ciccarelli F, Copley RR, Ponting CP, Bork P: Recent improvements to the SMART domain-based sequence annotation resource. Nucleic Acids Res. 2002, 30: 242-244. 10.1093/nar/30.1.242.PubMedPubMed CentralView ArticleGoogle Scholar
- SMART - simple modular architecture research tool. [http://smart.embl-heidelberg.de]
- Cluster scl24819. [http://genomes.rockefeller.edu/autodb/cluster_map.php?cluster_id=scl24819&db=MouSDB3]
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.