Use Of Alternative Polyadenylation Signals Research Articles

The Gene of Chicken Axonin-1

We have isolated and characterised the gene encoding the chicken axonal cell adhesion molecule axonin-1. This gene comprises 23 exons distributed over approximately 40 kb. Each of the six immunoglobulin-like domains and the four fibronectin-type-III-like domains of axonin-1 is encoded by two exons. The introns between two domains are exclusively phase I. Their exonintron borders correspond to the domain borders of the protein, suggesting that the gene of axonin-1 had been generated by exon shuffling. Three transcripts with a length of 4.3 kb, 5 kb, and 8 kb are found, and we provide evidence that they result from alternative use of polyadenylation signals. In situ hybridization revealed co-localisation of these transcripts in time and space in the developing chicken retina. Several identical transcription initiation sites were found in retina, brain, and cerebellum by RNase protection assay and anchored polymerase chain reaction. By transfection of HeLa cells, rat PC-12 phaeochromocytoma cells, and chicken embryonic fibroblasts with serially truncated segments of the 5′-flanking region linked to a luciferase reporter gene, we have found that the sequence from −91 to +56 relative to the transcription initiation site is sufficient to promote efficient gene expression. Tissue-specific expression of the axonin-1 gene seems to be regulated in part by sequences more than 1 kb upstream of the transcription initiation site. As revealed by computer analysis, the sequence immediately upstream of exon 1 contains an AP-2 binding site, a tumor phorbol-ester-responsive element, and a homeodomain protein binding site, but no canonical TATA box. A second AP-2 binding site and a homeodomain protein binding site are located within exon 1.

Structural Characterization and Expression Studies ofDbyand Its Homologs in the Mouse

In spite of recent evidence showing the importance of DBY (DEAD-box RNA helicase Y) in spermatogenesis in human, the biologic role of its homolog Dby (also known as Ddx3y) in the mouse is less clear. The present study aims at characterizing the molecular structure of Dby and comparing its expression with its X- and autosome-linked homologs in embryonic gonads and developing germ cells in mice. Molecular cloning by rapid amplification of 3'-cDNA ends showed that the Dby gene in the mouse gives rise to 2 transcripts that differ only in the length of the 3'-untranslated region as a consequence of the use of alternative polyadenylation signals. Measurement by quantitative real-time polymerase chain reaction showed that both transcripts were ubiquitously expressed and were present in male germ cells and Sertoli cells. They were more abundant in type A spermatogonia compared with pachytene spermatocytes and round spermatids. Expression of Dby in the embryonic gonad increased from day 10.5 and reached a peak at day 17.5. The expression level of Dby decreased after birth and remained low in adult male gonads. Although the level of expression of Dby was much lower than its X chromosome homolog, Ddx3 (also known as Ddx3x) in all samples examined, the pattern of expression of the 2 genes was comparable. In contrast, their autosomal homolog, D1Pas1(also known as PL10), was predominantly expressed in pachytene spermatocytes and round spermatids. This result is in accord with meiotic sex chromosome inactivation in that Dby and Ddx are replaced in pachytene spermatocytes by their autosomal retroposon. These observations indicate that unlike DBY in humans, the role of Dby in spermatogenesis is less obvious in the mouse and its biologic activity may be replaced by that of Ddx3 and D1Pas1.

Expression of the human TPT1 gene coding for translationally controlled tumor protein (TCTP) is regulated by CREB transcription factors

Re-evaluation of genomic and cDNA data revealed that the human TPT1 gene coding for the translationally controlled tumor protein (TCTP) consists of at least 4211 base pairs. It is transcribed into two transcripts of about 0.8 and 1.2 kb, which contain the same coding region and 5′-UTR, but differ in the length of 3′-UTRs by the use of alternative polyadenylation signals. 459 bp promoter sequences were analyzed by theoretical evaluation, reporter-gene assays, gelshift and footprinting experiments to search for transcription factor binding sites. The promoter contains two highly conserved CRE sites between − 50 and − 89 in close vicinity to a TATA-box at − 30. Supershift assays identified CREB I and Fra II of the CREB/ATF1/AP1 family as factors interacting with the CRE/AP1 site. A 3–5-fold stimulation of TCTP synthesis by forskolin and phorbolester in T24 cells and promoter–reporter experiments using CRE-deletion constructs suggested a transcriptional control by cAMP signaling via phosphorylation dependent activation of CRE/CREB interaction.

Genomic organization and promoter analysis of the mouse ADP-ribosylarginine hydrolase gene

Mono-ADP-ribosylation is a reversible modification of proteins with NAD:arginine ADP-ribosyltransferases and ADP-ribosylarginine hydrolases (ADPRH) catalyzing the opposing arms of an ADP-ribosylation cycle. The ADPRH cDNA had been cloned from human, rat, and mouse tissues and high levels of mRNA were found in brain, spleen, and testis. To begin to understand the molecular mechanisms that regulate ADPRH gene expression, we cloned the full-length cDNA, determined the genomic structure of mouse ADPRH, and investigated promoter function. Northern analysis using different regions of the ADPRH cDNA as probes identified two mRNAs of 1.7 and 3.0 kb, which resulted from the use of alternative polyadenylation signals, CATAAC and ATTAAA, beginning at positions 1501 and 2885, respectively, of the nucleotide sequence (A of ATG = 1). The ADPRH gene, represented in two overlapping genomic clones, spans ∼9 kilobases with four exons and three introns. The 5′-flanking region contains the features of a housekeeping gene; it has neither a TATA nor a CAAT box, but is, instead, highly GC-rich with multiple transcription initiation sites. Promoter analysis, assessed using transient transfection of PC12, NB41A3, NIH/3T3, and Hepa 1–6 cells with truncated constructs, revealed potent stimulatory (−119 to −89) and inhibitory (−161 to −119) elements, which were utilized similarly in the different cell lines. Further mutational analysis of the promoter and electrophoretic mobility-shift assays identified a positive GC-box element (−107 to −95); Sp1 and Sp3, which bound to this motif, were also detected by supershift assays. In co-transfection experiments using Drosophila SL2 cells that lack endogenous Sp1, Sp1 trans-activated the ADPRH promoter in a manner dependent on the presence of an Sp1-binding motif. The promoter activity pattern and involvement of Sp transcription factors are consistent with prior observations of widespread hydrolase expression in mammalian tissues.

Aspartyl β-Hydroxylase (Asph) and an Evolutionarily Conserved Isoform of Asph Missing the Catalytic Domain Share Exons with Junctin

The mouse aspartyl beta-hydroxylase gene (Asph, BAH) has been cloned and characterized. The mouse BAH gene spans 200 kilobase pairs of genomic DNA and contains 24 exons. Of three major BAH-related transcripts, the two largest (6,629 and 4,419 base pairs) encode full-length protein and differ only in the use of alternative polyadenylation signals. The smallest BAH-related transcript (2,789 base pairs) uses an alternative 3' terminal exon, resulting in a protein lacking a catalytic domain. Evolutionary conservation of this noncatalytic isoform of BAH (humbug) is demonstrated in mouse, man, and Drosophila. Monoclonal antibody reagents were generated, epitope-mapped, and used to definitively correlate RNA bands on Northern blots with protein species on Western blots. The gene for mouse junctin, a calsequestrin-binding protein, was cloned and characterized and shown to be encoded from the same locus. When expressed in heart tissue, BAH/humbug preferably use the first exon and often the fourth exon of junctin while preserving the reading frame. Thus, three individual genes share common exons and open reading frames and use separate promoters to achieve differential expression, splicing, and function in a variety of tissues. This unusual form of exon sharing suggests that the functions of junctin, BAH, and humbug may be linked.

A new structure for the murine Xist gene and its relationship to chromosome choice/counting during X-chromosome inactivation.

In this report, we present structural data for the murine Xist gene. The data presented in this paper demonstrate that the murine Xist transcript is at least 17.4 kb, not 14.3 kb as previously reported. The new structure of the murine Xist gene described herein has seven exons, not six. Exon VII encodes an additional 3.1 kb of information at the 3' end. Exon VII contains seven possible sites for polyadenylation; four of these sites are located in the newly discovered 3' end. Consequently, it is possible that several distinct transcripts may be produced through differential polyadenylation of a primary transcript. Alternative use of polyadenylation signals could result in size changes for exon VII. Two major species of Xist are detectable by Northern analysis, consistent with differential polyadenylation. In this paper, we propose a model for the role of the Xist 3' end in the process of X-chromosome counting and choice during embryonic development.

A variant transcript encoding an isoform of the human protein tyrosine kinase EPHB2 is generated by alternative splicing and alternative use of polyadenylation signals.

We previously isolated and characterized cDNA clones of DRT (EPHB2), encoding a receptor protein-tyrosine kinase of the EPH family. Northern blot analysis showed that EPHB2 transcripts are expressed in three sizes of approximately 4, 5, and 11 kb, suggesting that these transcripts are generated by alternative splicing and/or alternative use of polyadenylation sites. To explore this possibility, we isolated additional EPHB2 cDNA clones, including clone 5K-1, by re-screening the human fetal brain cDNA library. Nucleotide sequence analysis of clone 5K-1 revealed that it represents a variant transcript of EPHB2 (EPHB2v). Relative to the EPHB2 cDNA sequence previously reported, clone 5K-1 has two coding region deletions of 3 and 93 nucleotides. Nucleotide sequence analyses of EPHB2 genomic DNA fragments corresponding to these deletions suggest that the EPHB2v transcript is generated by alternative splicing. The 3' end of clone 5K-1 contains a polyadenosine stretch preceded by a potential polyadenylation signal, which is not used to generate the EPHB2 transcript. Taken together, these data indicate that EPHB2v is generated by both alternative splicing and alternative use of polyadenylation sites. The EPHB2v protein lacks one arginine residue that resides immediately following the EPHB2 transmembrane domain. In contrast, as a result of the frame shift caused by the 93 nucleotide deletion, the C-terminus of the EPHB2v protein is longer by 70 amino acids than that of EPHB2. We also show that the human neuroblastoma cell line SY5Y and NTera-2N neurons express high levels of EPHB2 and lower levels of EPHB2v. These structural variations found between the EPHB2 and EPHB2v proteins may reflect functional heterogeneity of EPHB2.

ES2, a gene deleted in DiGeorge syndrome, encodes a nuclear protein and is expressed during early mouse development, where it shares an expression domain with a Goosecoid-like gene.

ES2 is a gene deleted in DiGeorge syndrome (DGS) and velocardiofacial syndrome (VCFS) which has homologs in species as distant as Caenorhabditis elegans and Drosophila . The function of ES2 is unknown, and the predicted protein sequence does not contain motifs which suggest a particular role in the developmental defects present in DGS and VCFS. Here we show that the mouse homolog, Es2 , is transcribed in two forms resulting from the use of alternative polyadenylation signals. Structural analysis programs predict that the Es2 -encoded peptide has a coiled-coil domain, and transfection experiments with an Es2 -green fluorescent protein (GFP) fusion construct show that the peptide is recruited into the nucleus. Es2 is highly expressed during mouse embryogenesis from E7 onwards. In situ hybridization with an RNA probe revealed that the gene is widely expressed; however, relatively higher expression was detected in the nervous system, with a particularly high area of expression in a sub-region of the pons. The Es2 expression domain in the pons is shared with a Goosecoid-like gene ( Gscl) which is located upstream of Es2 , and raises the possibility that the two genes share regulatory elements and/or interact in this region of the developing brain. This finding suggests that different genes in the deleted region may be functionally related and might explain the occurrence of the characteristic phenotype in patients with non-overlapping genetic lesions.

Genomic Structure and Chromosomal Mapping of the Nuclear Orphan Receptor RORγ (RORC) Gene

The nuclear orphan receptor subfamily ROR/RZR is part of the steroid and thyroid hormone/retinoid receptor superfamily and consists of three different genes, α, β, and γ. In this study, we determined the genomic structure of mouse RORγ and the chromosomal localization of both mouse RORγ and human RORγ (HGMW-approved symbol RORC). The genomic structure of the mouse RORγ gene was derived from the analysis of P1 vector clones containing large genomic fragments encoding RORγ. These results revealed that the mRORγ gene has a complex structure consisting of 11 exons separated by 10 introns spanning more than 21 kb of genomic DNA. The DNA-binding domain is contained in two exons, 3 and 4, each encoding one zinc-finger. The splice site between exon 3 and exon 4 is identical to that found in RAR and TR3 receptors. RORγ is expressed as two mRNAs, 2.3 and 3.0 kb in size, that are derived by the use of alternative polyadenylation signals. We show by fluorescencein situhybridization that the mouse RORγ gene is located on chromosome 3, in a region that corresponds to band 3F2.1–2.2. The human RORγ was mapped to chromosome region 1q21. The results demonstrate that the RORγ genes are located in chromosomal regions that are syntenic between mouse and human.

Presence of multiple functional polyadenylation signals in the 3′-untranslated region of human corticotropin receptor cDNA

We present 2.59 kb of the 3′-non-coding region of the ACTH receptor cDNA that contains seven potential polyadenylation signals. Among these signals, five are functional as detected by 3′-RACE and are consistent with the transcripts of 1.8, 3.4 and 4 kb visualized on Northern blots. We propose that the most likely molecular mechanism for the multiple ACTH-R mRNA transcripts is the alternative use of polyadenylation signals.

Alternative transcriptional initiation and alternative use of polyadenylation signals in the αB-crystallin gene expressed in different chicken tissues

Overexpression of αB-crystallin is associated with numerous neurodegenerative diseases and abnormal cell growth patterns. To study the mechanisms involved in the control of the transcriptional activity of the gene we have characterized its expression in different chicken tissues. The sequence of the αB-crystallin cDNA isolated from chicken testis and 6-day-old chick embryo is highly homologous to the duck αB-crystallin cDNA and differs from the previously reported chicken lens αB-crystallin cDNA in the 5′ untranslated region (5′- UTR) and in one amino acid of the coding sequence. Four forms of the αB-crystallin cDNA detected in chicken testes arise from the use of alternative transcription initiation sites and alternative polyadenylation signals. The two principal hybridizing bands found in lens and embryonic tissues possess a short 5′- UTR and differ in the length of the 3′- UTR. Forms with longer 5′- UTR are present in testis, muscle, and heart. The use of different start sites and polyadenylation signals could modulate transcriptional activity and the stability of the messages. The expression of the αB-crystallin gene decreases from day 6 to day 8 of chick embryogenesis, in parallel with the expression of the polyubiquitin gene UbII.

Molecular Sites of Regulation of Expression of the Rat Cationic Amino Acid Transporter Gene

Cat-1 is a protein with a dual function, a high affinity, low capacity cationic amino acid transporter of the y+ system and the receptor for the ecotropic retrovirus. We have suggested that Cat-1 is required in the regenerating liver for the transport of cationic amino acids and polyamines in the late G1 phase, a process that is essential for liver cells to enter mitosis. In our earlier studies we had shown that the cat-1 gene is silent in the quiescent liver but is induced in response to hormones, insulin, and glucocorticoids, and partial hepatectomy. Here we demonstrate that cat-1 is a classic delayed early growth response gene in the regenerating liver, since induction of its expression is sensitive to cycloheximide, indicating that protein synthesis is required. The peak of accumulation of the cat-1 mRNA (9-fold) by 3 h was not associated with increased transcriptional activity of the cat-1 gene in the regenerating liver, indicating post-transcriptional regulation of expression of this gene. Induction of the cat-1 gene results in the accumulation of two mRNA species (7.9 and 3.4 kilobase pairs (kb)). Both mRNAs hybridize with the previously described rat cat-1/2.9-kb cDNA clone. However, the 3' end of a longer rat cat-1 cDNA (rat cat-1/6.5-kb) hybridizes only to the 7.9-kb mRNA transcript. Sequence analysis of this clone indicated that the two mRNA species result from the use of alternative polyadenylation signals. The 6. 5-kb clone contains a number of AT-rich mRNA destabilizing sequences which is reflected in the half-life of the cat-1 mRNAs (90 min for 7. 9-kb mRNA and 250 min for 3.4-kb mRNA). Treatment of rats with cycloheximide superinduces the level of the 7.9-kb cat-1 mRNA in the kidney, spleen, and brain, but not in the liver, suggesting that cell type-specific labile factors are involved in its regulation. We conclude that the need for protein synthesis for induction of the cat-1 mRNA, the short lived nature of the mRNAs, and the multiple sites for regulation of gene expression indicate a tight control of expression of the cat-1 gene within the regenerating liver and suggest that y+ cationic amino acid transport in liver cells is regulated at the molecular level.

The 3'-nontranslated region of rat renal glutaminase mRNA contains a pH-responsive stability element.

Rat kidney expresses two forms of glutaminase (GA) mRNA which probably result from the use of alternative polyadenylation signals. The two mRNAs are increased coordinately in response to metabolic acidosis via a mechanism that apparently does not involve transcriptional or translational regulation. A 956-bp fragment that contains the 3'-nontranslated sequence of the smaller GA cDNA was cloned into an expression vector (p beta G) that encodes a chimeric beta-globin growth hormone mRNA. Both the parent and the derived construct (p beta G-GA) were transfected into LLC-PK1-F+ cells. Stable transfectants express sixfold lower levels of beta G-GA mRNA than that of the parent beta G mRNA. However, only the beta G-GA mRNA is increased 2.5-fold by growth in acidic medium (pH 6.9, 10 mM HCO3-). The apparent half-life of the beta G mRNA (> 24 h) is unaffected by the pH of the growth media. In contrast, the apparent half-life of the beta G-GA mRNA is increased from 4.5 h to approximately 24 h when cells are transferred to acidic medium for 8 h. The observed pH response is not reproduced when the beta G-GA construct is stably transfected into COS-7 cells or when a beta-globin-phosphoenolpyruvate carboxykinase chimeric gene is expressed in LLC-PK1-F+ cells. Thus the 3'-nontranslated region of the GA mRNA contains a pH-responsive stability element.

Characterization of Multiple mRNAs That Encode Mammalian Translation Initiation Factor 5 (eIF-5)

Eukaryotic translation initiation factor 5 (eIF-5) interacts with the 40 S initiation complex (40S.mRNA.MettRNAf.eIF-2.GTP) to promote the hydrolysis of bound GTP with the concomitant joining of the 60 S ribosomal subunit to the 40 S initiation complex to form a functional 80 S initiation complex. In this paper, the multiple mRNAs that encode mammalian eIF-5 have been characterized. In rat tissues, three major eIF-5 mRNAs of 3.5, 2.8, and 2.2 kilobases in length are detected. All major eIF-5 mRNAs are initiated from a single transcription initiation site, contain identical 5'-untranslated and coding regions, but differ from one another only in the length of their 3'-untranslated regions. The different lengths of the 3'-untranslated region of eIF-5 mRNAs are generated by the use of alternative polyadenylation signals. Additionally, we demonstrate tissue-specific variations in eIF-5 mRNA expression as well as preference for polyadenylation sites. These results should lead to increased understanding of the regulation of eIF-5 gene expression.

The S locus receptor kinase gene encodes a soluble glycoprotein corresponding to the SRK extracellular domain in Brassica oleracea

Self-incompatibility in Brassica is controlled by the S locus which contains at least two genes. SLG encodes a secreted S locus glycoprotein whilst SRK encodes a putative S locus receptor kinase which consists of three domains: an extracellular domain sharing extensive sequence identity with SLG, transmembrane region, and a cytoplasmic domain exhibiting a serine/threonine protein kinase activity. Here, the existence of truncated forms of the SRK protein corresponding to the extracellular domain of the putative receptor is reported. These proteins were detected by an antibody which recognizes the N-terminus of SRK3 and, in an F2 progeny segregating for the S3 haplotype, were only expressed in plants possessing the S3 haplotype. The truncated SRK proteins were expressed specifically in stigmas but, unlike the membrane-spanning SRK3 protein, were soluble and occurred as four different glycoforms sharing the same amino acid backbone as shown by deglycosylation experiments. Several SRK3 transcripts that may code for these truncated SRK3 proteins have been identified by RACE PCR, stigma cDNA library screening and RNA blot analysis. These transcripts are apparently generated by a combination of alternative splicing and the use of alternative polyadenylation signals. The existence of truncated forms of the S locus receptor kinase highlights some similarities between plant and animal receptor kinases. In animals, soluble extracellular domains of receptors have been described and, in some cases, have been shown to play a role in the modulation of signal transduction. By analogy, the soluble, truncated SRK proteins may play a similar role in the self-incompatibility response.

Expression of cell-surface proteoglycan mRNA by human articular chondrocytes

The expression of six cell-surface proteoglycans (syndecan, fibroglycan, amphiglycan, glypican, betaglycan and CD44) was studied at the mRNA level. Analysis was performed by Northern blotting using total RNA preparations from freshly isolated articular chondrocytes obtained from both juveniles and adults. Similar results were obtained for both age groups. By far the most abundant message was that for amphiglycan, CD44 message was next in relative abundance, and the messages for fibroglycan, glypican and betaglycan were all expressed at low levels. Syndecan message could not be detected by this technique. This pattern of expression was different to that observed in cultured skin fibroblasts, where the messages for amphiglycan, CD44, fibroglycan and glypican were all expressed at a similar level. In contrast with the fibroblasts, where the amphiglycan message exhibits no size polymorphism, the chondrocyte amphiglycan message is present in three polymorphic forms, due to the use of alternative polyadenylation signals. When the newly isolated chondrocytes are maintained in monolayer culture for several passages, the amphiglycan message heterogeneity reverts to that characteristic of the fibroblasts. Thus human articular chondrocytes are characterized by both their high level of amphiglycan message expression and their use of alternative polyadenylation signals.

The gene of chicken axonin-1. Complete structure and analysis of the promoter.

We have isolated and characterised the gene encoding the chicken axonal cell adhesion molecule axonin-1. This gene comprises 23 exons distributed over approximately 40 kb. Each of the six immunoglobulin-like domains and the four fibronectin-type-III-like domains of axonin-1 is encoded by two exons. The introns between two domains are exclusively phase I. Their exon/intron borders correspond to the domain borders of the protein, suggesting that the gene of axonin-1 had been generated by exon shuffling. Three transcripts with a length of 4.3 kb, 5 kb, and 8 kb are found, and we provide evidence that they result from alternative use of polyadenylation signals. In situ hybridization revealed co-localisation of these transcripts in time and space in the developing chicken retina. Several identical transcription initiation sites were found in retina, brain, and cerebellum by RNase protection assay and anchored polymerase chain reaction. By transfection of HeLa cells, rat PC-12 phaeochromocytoma cells, and chicken embryonic fibroblasts with serially truncated segments of the 5'-flanking region linked to a luciferase reporter gene, we have found that the sequence from -91 to +56 relative to the transcription initiation site is sufficient to promote efficient gene expression. Tissue-specific expression of the axonin-1 gene seems to be regulated in part by sequences more than 1 kb upstream of the transcription initiation site. As revealed by computer analysis, the sequence immediately upstream of exon 1 contains an AP-2 binding site, a tumor phorbol-ester-responsive element, and a homeodomain protein binding site, but no canonical TATA box. A second AP-2 binding site and a homeodomain protein binding site are located within exon 1.

Identification and characterization of a second mouse Nramp gene

The Nramp gene was isolated as a candidate for the host resistance locus Bcg/Ity/Lsh, which controls natural resistance of mice to several types of infections. We have isolated by cross-hybridization cDNA clones corresponding to a second mouse Nramp gene, which we designate Nramp2. Nucleotide and predicted amino acid sequence analyses of full-length cDNA clones for Nramp2 indicate that this novel Nramp protein is closely homologous to the previously described Nramp and that the two genes form part of a small gene family. The two Nramp proteins encode integral membrane proteins that share 63% identical residues and an overall homology of 78%. They share very similar secondary structure, including identical hydropathy profiles and predicted membrane organization, with a minimum of 10 and most probably 12 transmembrane domains, a cluster of predicted N-linked glycosylation sites, and a consensus transport motif. Analysis of the distribution of Nramp2 mRNA transcripts in normal mouse tissues by Northern blotting revealed that the Nramp2 gene produces several mRNAs, including prominent 3.3- and 2.3-kb species generated by the use of alternative polyadenylation signals. In contrast to the previously described macrophage-specific Nramp gene, Nramp2 mRNAs were found to be expressed at low levels in all tissues tested. Using a polymorphic (GT) 26 dinucleotide repeat identified in the 3′ untranslated region of the mRNA, we have mapped the Nramp2 gene to the distal part of mouse chromosome 15 between markers D15Mit41 and D15Mit15, with the gene order and intergene distance (in cM): centromere—56.1— D15Mit41—(1 ± 1)— Nramp2— (5 ± 2)— D15Mit15.

Analysis of human and rodent beta 3-adrenergic receptor messenger ribonucleic acids.

Recent evidence indicates that the human and rodent beta 3-receptor genes differ with respect to their intron/exon organization and putative regulatory elements in the proximal promotor regions. The present work examined the molecular basis for heterogeneity among human and rodent beta 3-receptor messenger RNAs (mRNAs) and mRNA transcription start site selection with gene-derived human and rodent probes. This analysis indicates that the mRNA size heterogeneity of the human beta 3-receptor mRNA seen in Northern blots results from the use of alternative polyadenylation signals. In contrast, mRNA heterogeneity of the rat beta 3-receptor mRNA is due to the use of widely separated promoters. Analysis of mRNA transcription sites within the proximal promoters of the human and rat genes indicates that multiple homologous start sites are used. However, the pattern of transcription initiation appears to be both species and tissue specific. In rat adipose tissues where abundant expression of the beta 3 gene occurs, most transcripts began at -161 relative to translation initiation ATG, whereas minor transcription began around -123 and -109. Interestingly, virtually all transcripts were found to begin at the -123 and -109 sites in the rat gastric fundus, which expresses only low levels of beta 3-receptor mRNA. In human brown fat and neuroepithelioma cells, the vast majority (> 80%) of beta 3 transcripts began at multiple sites around -130. Exposure of murine 3T3-F442A adipocytes to isoproterenol, 8-bromo-cAMP, phorbol 12-myristate 13-acetate, or dexamethasone dramatically reduced murine beta 3-receptor mRNA levels. In contrast to mouse adipocytes, 8-bromo-cAMP increased human beta 3-receptor mRNA levels in SK-N-MC neuroepithelioma cells, whereas isoproterenol, phorbol 12-myristate 13-acetate, and dexamethasone were without effect. These data identify important differences in the structure and regulation of the human and rodent beta 3-receptor genes and suggest that the data obtained in rodent models may not be directly applicable to the regulation of the human gene.

Alternative splicing in the fragile X gene FMR1.

The FMR1 gene, associated with fragile X syndrome, has recently been cloned and the sequence of partial cDNA clones is known. We have determined additional cDNA sequences both at the 5' and 3' end. We have characterized the expressed gene by means of RT-PCR in various tissues and have found that alternative splicing takes place in the FMR1 gene, which does not seem to be tissue specific. When the different alternative splicing events are combined, 12 distinct mRNA products could result from FMR1 expression in each tested tissue. In all these transcripts the open reading frame is maintained until the same stop codon. At the 3' end alternative use of polyadenylation signals is found. The alternative splicing allows functional diversity of the FMR-1 gene. Whether all the possible proteins will be synthesized and whether they will be functionally active has to be determined.