SEARCH · Search PubMed

Results for “long noncoding RNA”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,477 records · Page 82Linked to original sources

PubMed2003

Molecular correlates of genes exhibiting RNAi phenotypes in Caenorhabditis elegans.

Understanding genome-wide links between genotype and phenotype has generally been difficult due to both the complexity of phenotypes, and until recently, inaccessibility to large numbers of genes that might underlie a trait. To address this issue, we establish the association between particular RNAi phenotypes in Caenorhabditis elegans and sequence characteristics of the corresponding proteins and DNA. We find that genes showing RNAi phenotypes are long and highly expressed with little noncoding DNA and high rates of synonymous site substitution (KS). In addition, genes conferring RNAi phenotypes have significantly lower rates of nonsynonymous site substitution (KA). Collectively, these sequence features explain nearly 20% of the difference between the sets of loci that display or lack a RNAi-mediated effect, and reflect aspects both of the RNAi mechanism and the biological function of the genes. For example, the particularly low rate of evolution of genes in the sterility RNAi phenotype class suggests a role of C. elegans life history in shaping these patterns of sequence and expression characteristics on phenotypes. This approach also allows prediction of a set of heretofore-uncharacterized loci for which we expect future RNAi studies to reveal phenotypic effects (i.e., false negatives in present screens).

Animals↗

PubMed1990

On the flexible interaction of yeast factor tau with the bipartite promoter of tRNA genes.

Yeast transcription factor tau (analogous to vertebrate TFIIIC) interacts specifically with the internal split promoter of tRNA genes. Binding to the two promoter elements (A block and B block) occurs within 30 seconds even when they are separated by a long intervening sequence. Dimethylsulfate protection analysis of contact points between tau and the noncoding strand of a series of internally deleted tRNA3(Leu) genes shows that the specificity of the interaction is not affected by changes in the distance or in the relative helical orientation of the promoter elements. This result is consistent with the results of previous footprinting experiments (Baker, R.E., Camier, S., Sentenac, A. and Hall, B.D., 1987, Proc. Natl. Acad. Sci. USA, 84, 8768-8772). To test if any physical constraint is imposed on the DNA molecule upon tau binding, we analyzed the effect of introducing random single-strand breaks in the noncoding strand of the tRNA gene. Whereas some nicks located in the A block were found to prevent tau binding, no single-strand break in the B block region or in the DNA between the A and B blocks were observed to inhibit or facilitate the binding of tau. We therefore propose that the great flexibility of the tau-tDNA interaction is mostly due to the tau protein itself.

DNA, Fungal↗

PubMed1990

Molecular analysis of structural protein genes of the Yamagata-1 strain of defective subacute sclerosing panencephalitis virus. II. Nucleotide sequence of a cDNA corresponding to the P plus M dicistronic mRNA.

The nucleotide sequence of a cloned cDNA corresponding to the P + M dicistronic mRNA of a subacute sclerosing panencephalitis (SSPE) virus was determined and compared with data of measles virus (MV). The dicistronic mRNA of the SSPE virus consisted of the 3' proximal 626 nucleotides of P mRNA, intercistronic trinucleotides, a full length of M mRNA, and 75 poly A nucleotides. The part encoding the P protein had a high homology to MV, except at the noncoding region. The terminating consensus sequence of the P gene and the intercistronic trinucleotides of the SSPE virus were CTAC(A)6 and CCT; in MV they are TTAT(A)6 and CTT, respectively. In the M gene, the starting consensus sequence was exactly the same as MV, but at the 5' proximal end, one third of this gene was different: The first ATG codon of the MV M gene signaling opening of the reading frame was changed to ACG in the SSPE virus and one long open reading frame started from the third ATG codon. The stop codon (TAG) of the MV M gene was also changed to CAG in the SSPE virus. Thus, the deduced SSPE-virus M protein lacked 50 amino acids at the amino terminal and had 15 extra amino acids at the carboxyl end when compared with the MV M protein.

Amino Acid Sequence↗

PubMed1989

Structure and transcription of the glycoprotein gene of attenuated HEP-Flury strain of rabies virus.

The mRNA-encoding G protein of the attenuated HEP-Flury strain of rabies virus was sequenced by the cDNA cloning technique. The G-mRNA was composed of 2059 nucleotides, with the coding region located from the 28th to 1602nd nucleotide, and was capable of encoding a polypeptide of 524 amino acids. Although the coding region was highly homologous (90% or more) to that of ERA and PV strains, the 3' noncoding region of the HEP virus G-mRNA was longer than that reported for other strains by some 400 nucleotides. The extra sequence was homologous to the long G-L intergenic sequence of the PV viral genome. The HEP virus genome lacked the postulated polyadenylating signal (TG-AAAAAAAA) that should have been found just before the "long G-L intergenic region," which indicates that the long G-L intergenic region of the HEP virus is integrated into the preceding G gene, and is transcribed only as a portion of the G-mRNA molecule. In the ERA virus-infected cells, however, two species of G-mRNA (1.9 and 2.3 kb long) were produced. The longer G-mRNA also contained the sequence complementary to the long G-L intergenic region and the shorter one did not. These findings suggest that two different poly(A)-tailing signals (one is present just before and another at the end of the long G-L intergenic region) work toward terminating the transcription of the ERA virus G gene and that the longer G-mRNA is produced as a readthrough product.

Amino Acid Sequence↗

PubMed2006

Transcription of bxd noncoding RNAs promoted by trithorax represses Ubx in cis by transcriptional interference.

Much of the genome is transcribed into long noncoding RNAs (ncRNAs). Previous data suggested that bithoraxoid (bxd) ncRNAs of the Drosophila bithorax complex (BX-C) prevent silencing of Ultrabithorax (Ubx) and recruit activating proteins of the trithorax group (trxG) to their maintenance elements (MEs). We found that, surprisingly, Ubx and several bxd ncRNAs are expressed in nonoverlapping patterns in both embryos and imaginal discs, suggesting that transcription of these ncRNAs is associated with repression, not activation, of Ubx. Our data rule out siRNA or miRNA-based mechanisms for repression by bxd ncRNAs. Rather, ncRNA transcription itself, acting in cis, represses Ubx. The Trithorax complex TAC1 binds the Ubx coding region in nuclei expressing Ubx, and the bxd region in nuclei not expressing Ubx. We propose that TAC1 promotes the mosaic pattern of Ubx expression by facilitating transcriptional elongation of bxd ncRNAs, which represses Ubx transcription.

Animals↗

PubMed1994

Complete nucleotide sequence of a strain of coxsackie B4 virus of human origin that induces diabetes in mice and its comparison with nondiabetogenic coxsackie B4 JBV strain.

The E2 strain of coxsackie B4 virus (CB4), which is of human origin, can induce a diabetes-like syndrome in mice. The cDNA of the genome of the E2 strain was cloned and sequenced. The E2 viral genome was found to comprise 7,396 bases, which appear to encode a polyprotein of 2,183 amino acids with an overall similarity of 94.91% to nondiabetogenic CB4 prototype JBV strain. The E2 genome is organized like other enteroviruses. It has a 5' noncoding region of 744 nucleotides, a single long open translational reading frame starting at nucleotide 745 and extending to nucleotide 7293, a 3' noncoding region of 100 nucleotides, and a poly (A) tract. Genomic sequence comparison of the E2 and JBV strains showed 1,369 nucleotide substitutions in the genome of the E2 strain, most of which are single and silent. There were 111 resultant amino acid changes arising from some of these substitutions, including 82 amino acid changes in the noncapsid proteins, and 29 amino acid changes in the capsid proteins VP1, VP2, VP3, and VP4, which showed 11, 13, 4, and 1 substitution(s), respectively. Noncapsid protein P2-C showed eight amino acid substitutions. On the basis of the sequence comparison of E2 and JBV strains of CB4, we suggest that some of the amino acid changes in the capsid and noncapsid proteins of the E2 strain may be involved in the determination of its diabetogenicity.

Amino Acid Sequence↗

PubMed1996

Genomic organization of the human beta-catenin gene (CTNNB1).

The cytoplasmic beta-catenin protein is implicated in signal transduction and associates with both the cell-cell adhesion protein E-cadherin and the tumor suppressor gene product APC. We determined the primary structure of the human beta-catenin gene (CTNNB1) by analysis of cDNA and genomic clones. The size of the complete gene was determined to be 23.2 kb. Restriction mapping and partial sequence analysis revealed 16 exons. All splice donor and acceptor sites were conformable to the GT/AG rule. The exon size ranged from 61 to 790 bp. Half of the introns were smaller than 550 bp, with the smallest being 84 bp and the longest being 6700 bp. The intron-exon boundaries did not coincide either with conserved sites in the 12 armadillo repeat sequences of beta-catenin or with intron-exon boundaries in the armadillo gene of Drosophila. A major site for transcription initiation was identified as an A residue 214 nucleotides upstream of the ATG initiation codon. The resulting transcript is 3362 nucleotides long. Compared to the previously published mRNA sequence, additional residues were identified, 16 at the 5' end and 766 at the 3' end of the mRNA. An alternative splice acceptor site within exon 16 reduced the 3' UTR sequence by 159 bp. Polymerase chain reaction on cDNA from 14 human cell lines demonstrated the general occurrence of both splice variants. The 5'-flanking region is highly GC-rich and lacks a CCAAT box, but contains a TATA box and potential binding sites for several transcription factors, such as NF kappa B, SP1, AP2, and EGR1. Both a 437-bp fragment and a 6-kb fragment, containing about 4.7 kb of the 5'-flanking region in addition to the noncoding exon 1 and 1 kb of intron 1, showed clear promoter activity when these fragments were linked to a secreted alkaline phosphatase reporter gene and transfected into a mouse epithelial cell line.

Alternative Splicing↗

PubMed1999

Molecular cloning and characterization of the human CLOCK gene: expression in the suprachiasmatic nuclei.

The Clock gene is an essential regulator of circadian rhythms. It encodes a member of the basic helix-loop-helix/PER-ARNT-SIM family of transcription factors known to play a central role in the control of diverse cellular events. Previously we described the functional identification and molecular isolation of the Clock gene in the mouse, its interaction with the BMAL1 protein, and the role of this complex as a transcriptional activator in the circadian pacemaker. Here, we report the cloning, exon organization, chromosomal location, and mRNA expression of the human CLOCK gene. The coding sequence of human CLOCK extends for 2538 bp and is 89% identical to its mouse ortholog; its deduced amino acid sequence is 846 residues long and is 96% identical to mouse CLOCK. Radiation hybrid mapping localized human CLOCK to the long arm of human chromosome 4 (4q12). Direct sequencing of a genomic CLOCK clone indicated that the coding sequence of human CLOCK extends over 20 exons and that its intron/exon organization is identical to that of the mouse ortholog. Northern blot analysis indicated widespread expression of two major transcripts of 8 and 10 kb, and in situ hybridization of human brain tissue revealed elevated expression of CLOCK mRNA in the suprachiasmatic nuclei, the locus of circadian control in mammals, and in the cerebellum. Comparison of cDNA clones revealed two single nucleotide polymorphisms in noncoding sequence flanking the CLOCK open reading frame. The central role of Clock in the organization of circadian rhythms suggests that it will be a useful candidate gene for genetic analyses of disorders associated with dysfunction of the circadian system.

Alleles↗

PubMed1998

Sequence analysis of Mus dunni endogenous virus reveals a hybrid VL30/gibbon ape leukemia virus-like structure and a distinct envelope.

Mus dunni endogenous virus (MDEV) can be activated from M. dunni cells by exposing the cells to hydrocortisone or 5-iodo-2'-deoxyuridine. Interference analysis has revealed that MDEV uses a receptor for cell entry that is different from those used by other murine retroviruses. The entire genome has now been sequenced, revealing a long terminal repeat (LTR)-gag-pol-env-LTR structure typical of simple retroviruses of the murine leukemia virus genus, with no additional open reading frames between env and the 3' LTR. The LTRs and other noncoding regions of MDEV are most closely related to those of VL30 elements, while the majority of the coding sequences are most closely related to those of gibbon ape leukemia virus. MDEV represents the first example of a naturally occurring, replication-competent virus with sequences closely related to VL30 elements. The U3 region of MDEV contains six nearly perfect 80-bp repeats and the beginning of a seventh, and the region expected to contain the packaging sequence contains approximately four imperfect 33-bp repeats. The receptor specificity domains of the envelope are unique among retroviruses and show no apparent similarity to regions of known proteins.

Animals↗

PubMed1988

Fusion glycoprotein (F) of rinderpest virus: entire nucleotide sequence of the F mRNA, and several features of the F protein.

The full-length cDNA corresponding to the mRNA for the fusion protein of rinderpest virus (RV) was cloned and its complete nucleotide sequence was determined. The mRNA for the F protein was composed of 2359 nucleotides and contained a single large open reading frame which was capable of encoding 566 amino acids with a molecular weight (MW) of 58,929. The RV-F mRNA had a long noncoding region at the 5' end (586 bases) which was C-rich like the measles virus (MV)-F mRNA but they did not appear to be homologous with each other. Their secondary structure with long G-C stems suggested that they are easily folded. The coding region of RV-F mRNA was significantly homologous with that of MV-F; 74% of the nucleotides and 79.0% [corrected] of the amino acids were identical. The predicted RV-F protein had a basic amino acid region (104-108) which may be cleaved by protease to yield an activated form of F1,2. Three regions (1-19, 109-133, 418-513) were highly hydrophobic, and the N-terminal hydrophobic region of F1 or the positions of cysteines were significantly conserved compared with those of the other paramyxovirus F proteins. Three potential sites for glycosylation existed only in the F2 protein. Several features of the predicted RV-F protein were confirmed in polyacrylamide gel electrophoresis.

Amino Acid Sequence↗

PubMed2003

Regulatory roles of conserved intergenic domains in vertebrate Dlx bigene clusters.

Dlx homeobox genes of vertebrates are generally arranged as three bigene clusters on distinct chromosomes. The Dlx1/Dlx2, Dlx5/Dlx6, and Dlx3/Dlx7 clusters likely originate from duplications of an ancestral Dlx gene pair. Overlaps in expression are often observed between genes from the different clusters. To determine if the overlaps are a result of the conservation of enhancer sequences between paralogous clusters, we compared the Dlx1/2 and the Dlx5/Dlx6 intergenic regions from human, mouse, zebrafish, and from two pufferfish, Spheroides nephelus and Takifugu rubripes. Conservation between all five vertebrates is limited to four sequences, two in Dlx1/Dlx2 and two in Dlx5/Dlx6. These noncoding sequences are >75% identical over a few hundred base pairs, even in distant vertebrates. However, when compared to each other, the four intergenic sequences show a much more limited similarity. Each intergenic sequence acts as an enhancer when tested in transgenic animals. Three of them are active in the forebrain with overlapping patterns despite their limited sequence similarity. The lack of sequence similarity between paralogous intergenic regions and the high degree of sequence conservation of orthologous enhancers suggest a rapid divergence of Dlx intergenic regions early in chordate/vertebrate evolution followed by fixation of cis-acting regulatory elements.

Animals↗

PubMed1991

Bluetongue virus evolution: sequence analyses of the genomic S1 segments and major core protein VP7.

The S1 segments, encoding the group-specific antigen, VP7, from the five United States prototype BTV serotypes were cloned as full-length entities. The nucleotide and deduced amino acid sequences of segment S1 of BTV-2 were determined and compared with BTV-10, -11, -13, and -17, completing the sequencing of this cognate gene segment from all five US BTV serotypes. Each segment is 1156 bp long and contains an open reading frame encoding the 349-amino acid VP7 protein. Most (greater than 94%) of the amino acids of VP7 among the serotypes are conserved, including the location (position 255) of a single lysine residue. Secondary structure analyses of VP7 predict a putative eight-stranded beta-barrel between amino acid positions 150 and 250, a structure similar to that observed in ssRNA viruses. The S1 genes are flanked by conserved 5' and 3' noncoding regions. Stem-loop structures are predicted at the 3' end of each gene (nucleotide positions 1058-1097). The S1 segments of BTV-2, -10, -11, and -17 have greater than 93% of the nucleotides conserved, while less than 80% of their bases are identical with BTV-13. Analyses of nucleotide mismatches in each codon position of the VP7 open reading frame, transition frequencies, and evolutionary distances show that of the five, BTV-13 is the most distantly related and that BTV-10 and -17 are the most closely related serotypes. Evolutionary distance calculations of segment L2 from BTV-10, -11, and -17 concur with these observations. Comparison of this relationship with hybridization data of segment M3, which codes for VP5, suggests that BTV-17 has evolved by a combination of genetic drift and genomic reassortment. The data also indicate that the five US BTV serotypes are derived from two distinct gene pools. Evolution distances were used to estimate an evolution rate of 2.2 x 10(-3) nucleotide substitution/site/year for BTV segment S1. This rate is similar to the genes of retroviruses and implies an absence of RNA polymerase proofreading activity for dsRNA viruses.

Amino Acid Sequence↗

PubMed2004

Genomic structure and differential expression of two tandem-arranged GSTZ genes in rice.

Glutathione S-transferases (GSTs) are scavenging enzymes that detoxify cellular xenobiotics and toxins by catalyzing the conjugation of these substrates with a tripeptide glutathione. GSTs are classified depending on gene organization and sequence similarity. The sequence analysis of genomic DNA for zeta class GST (GSTZ) locus in rice indicated that two homologous GSTZ genes lay in a tandem orientation with a short (0.4 kb) intergenic spacer. The upstream OsGSTZ1 and downstream OsGSTZ2 spanned 3.5 and 3.2 kb with nine coding exons, respectively. The transcript of OsGSTZ1 had a long 3' untranslated region (3' UTR) that was mostly encoded by a 10th noncoding exon, whereas OsGSTZ2 mRNA contained a long 5' UTR. Northern blot analysis showed that OsGSTZ1/2 messages were strongly expressed in leaf blades, while transcripts from roots were low level. Because OsGSTZ1/2 messages in leaf tissues were strongly induced only by water treatment, it was difficult to assay for the induction of OsGSTZ1/2 transcripts by various stress treatments. Thus, using rice culture cells, we analyzed the respective responses of OsGSTZ1 and OsGSTZ2 genes against various treatments by real-time reverse transcriptase-polymerase chain reaction (RT-PCR). The results showed that OsGSTZ1 was expressed at a level ca. 1000-fold higher than OsGSTZ2 in suspension cells without stress treatment. OsGSTZ1 was expressed constitutively under various stress conditions. In contrast, the expression of OsGSTZ2 gene was strongly enhanced to 30-fold by treatment with jasmonic acid. These observations suggested that the expression of OsGSTZ1 and OsGSTZ2 genes are differentially regulated in the culture cell of rice.

Amino Acid Sequence↗

PubMedDate not supplied

Complete sequence of rabbit kidney aminopeptidase N and mRNA localization in rabbit kidney by in situ hybridization.

We have isolated and sequenced a full-length cDNA for rabbit kidney aminopeptidase N (APN). The 3-kilobase cDNA contains 12 nucleotides of the 5' noncoding region, a 2898 nucleotide long open reading frame, and 113 nucleotides of the 3' untranslated region. The open reading frame encodes a type II membrane protein of 966 amino acid residues composed of a 10 residue NH2-terminal cytosolic domain, a 23 residue transmembrane domain, and a large 933 residue ectodomain that contains the active site. Rabbit APN has eight potential N-glycosylation sites and seven cysteine residues, one of which is located in the transmembrane domain. Computer analysis showed that the enzymes from human, rat, and rabbit were highly conserved, except for the stalk region immediately downstream from the transmembrane domain. Using in situ hybridization techniques we showed that in rabbit kidney, APN mRNA is present in proximal tubules but not in glomeruli, which corresponds to the localization of the protein observed by immunohistochemistry. Taken together, our results strongly suggest that the expression of APN in kidney is modulated at mRNA levels and not at translational and (or) posttranslational levels.

Amino Acid Sequence↗

PubMed1984

Yeast LEU1. Repression of mRNA levels by leucine and relationship of 5'-noncoding region to that of LEU2.

Yeast LEU1 encodes the second enzyme in leucine biosynthesis. A 3.5-kilobase pair (kb) yeast genomic DNA fragment which complements a leu1 auxotroph was isolated by yeast transformation. After recloning into an integrating vector, a subfragment (of the 3.5-kb fragment) directs a URA3 marker to integrate at the LEU1 locus. About 1.9 kb was sequenced from the 5'-end of the 3.5-kb insert, and a long open reading frame and potential ATG start codon were located. S1 nuclease mapping showed a major start for LEU1 transcripts at 79 nucleotides upstream of the ATG codon. Northern blots with a LEU1-specific probe showed the size of the LEU1 transcript (about 2.9 kb) is consistent with the size of the enzyme and steady state levels of the transcript are sharply reduced in cells grown in the presence of an elevated leucine concentration. The latter observation correlates with the repression by leucine of LEU1 gene product levels. Other work has shown that the level of the LEU2 gene product is also repressed by leucine. Sequence comparisons between LEU1 and LEU2 show that the LEU2 5'-sequences which are cognate to leucine are not found in LEU1; and three blocks of nucleotide sequence homology between LEU1 and LEU2 occur in the 330 nucleotides upstream of the respective start codons.

Amino Acid Sequence↗

PubMed1984

Complete nucleotide sequence of the mRNA coding for the N protein of vesicular stomatitis virus (New Jersey serotype).

The nucleotide sequence of the mRNA encoding the nucleocapsid protein of the New Jersey serotype (Ogden strain) of vesicular stomatitis virus (VSV) was determined from two overlapping cDNA clones spanning almost entirely the coding region of the mRNA. The 5'-terminal noncoding sequence present in the mRNA but not in the cDNA clones was determined from a primer extended to the 5' terminus of the mRNA. The mRNA is 1329 nucleotides long (excluding polyadenylic acid) and encodes a protein of 422 amino acids. The nucleotide sequence was compared with the previously determined nucleotide sequence of the nucleocapsid protein of the Indiana serotype. An overall identity of 67.7% was found between the two serotypes. The only place where insertions and/or deletions have occurred during the evolution of the two viral genes from their presumed common ancestor is in the untranslated region. The nonidentical nucleotides are distributed throughout the length of the mRNA although not in an entirely random manner. The predicted amino acid sequence demonstrates that both proteins are initiated from the initiator codon located at the same distance from the 5' end (nucleotides 14 to 16) and contain the same number of amino acids. An overall identity of more than 80% of the amino acid sequence was observed between the two proteins when conservative replacements of amino acids were considered.

Amino Acid Sequence↗

PubMed1992

Nuclear-encoded chloroplast ribosomal protein L27 of Nicotiana tabacum: cDNA sequence and analysis of mRNA and genes.

A tobacco (Nicotiana tabacum cv. Petite Havana) leaf cDNA library was constructed in the expression vector lambda gt11. Immunological and nucleic acid hybridization screening yielded several cDNAs encoding an M(r) 19,641 precursor to an M(r) 14,420 mature protein which is homologous to Escherichia coli ribosomal protein L27. One cDNA (L27-1; 882 nucleotides long) contains 104 bp of 5'-noncoding sequence, 51 codons for a transit peptide, 128 codons for the predicted mature L27 polypeptide, and 241 bp of 3'-noncoding sequence, including the poly(A)29 tail. A beta-galactosidase-L27 fusion protein was bound to nitrocellulose filters, expressed, and used as an affinity matrix to purify monospecific antibody to L27 protein from an antiserum of rabbits immunized with 50S chloroplast ribosomal proteins. Using this monospecific antibody, protein L27 was identified among HPLC-purified tobacco chloroplast ribosome 50S subunit proteins. The predicted amino terminus of the mature L27 protein was confirmed by partial sequencing of the HPLC-purified L27 protein. The mature L27 protein has 66%, 61%, 56%, and 48% amino acid sequence identity with the L27-type ribosomal proteins of Bacillus subtilis, E. coli, Bacillus stearo-thermophilus, and yeast mitochondria (MRP7), respectively, in the homologous overlapping regions. The transit peptide of tobacco chloroplast ribosomal protein L27 has 41% amino acid sequence similarity with the MRP7 mitochondrial targeting sequence. Tobacco chloroplast L27 protein also has a 40 amino acid long carboxyl-terminal extension (compared to its bacterial counterparts) which is similar to the corresponding portion of yeast MRP7.(ABSTRACT TRUNCATED AT 250 WORDS)

Amino Acid Sequence↗

PubMed2001

Promoter choice influences alternative splicing and determines the balance of isoforms expressed from the mouse bcl-X gene.

Differential splicing from the bcl-X gene generates several isoforms with opposite effects on the apoptotic response. To explore the mechanism controlling the balance between the various isoforms, we have characterized the 5' region of the mouse bcl-X gene. We identified three new promoters in addition to the two previously described (Grillot, D. A., M., G.-G., Ekhterae, D., Duan, L., Inohara, N., Ohta, S., Seldin, M. F., and Núñez, G. (1997) J. Immunol. 158, 4750-4757). These five promoters (P1-P5) would give rise to at least five mRNAs with different 5'-untranslated region, all sharing the same translation initiation site. Except for the product of the most proximal promoter (P1), the other mRNAs are generated by alternative splicing of noncoding exons to a common acceptor site located in the first translated exon. Reverse transcriptase-polymerase chain reaction, primer extension, and RNase protection assays demonstrate a tissue-specific pattern of promoter usage. P1 and P2 are active in all tissues analyzed, whereas the other three promoter show tissue-specific activities. P3 is active in spleen, liver, and kidney, P4 is active in uterus and spleen, and P5 is active in spleen, liver, brain, and thymus. We present evidence suggesting that promoter selection influences the outcome of the splice process. Transcripts from P1 generate mainly the mRNA for the long isoform Bcl-X(L), whereas transcripts from P2 generate mRNAs for the isoforms Bcl-X(L), Bcl-X(S), and Bcl-X(gamma) and transcripts from P3 yield mainly mRNAs for the isoform Bcl-X(gamma). Our results suggest a key role of promoter choice in determining alternative splicing and, thus, the balance of Bcl-X isoforms.

5' Untranslated Regions↗