Putative Polyadenylation Signal Research Articles

Human liver arylacetamide deacetylase. Molecular cloning of a novel esterase involved in the metabolic activation of arylamine carcinogens with high sequence similarity to hormone-sensitive lipase.

Microsomal arylacetamide deacetylase (DAC) competes against the activity of cytosolic arylamine N-acetyltransferase, which catalyzes one of the initial biotransformation pathways for arylamine and heterocyclic amine carcinogens in many species and tissues. Activity determination and immunoblot analysis of DAC in human target tissues for arylamine carcinogens revealed that in extrahepatic tissues, additional enzymes are responsible for any deacetylation activity, whereas a single enzyme predominantly catalyzes this hydrolytic reaction in liver. We isolated and characterized a full-length cDNA from a human liver lambda gt11 library. This clone encodes an open reading frame of 400 amino acids with a deduced molecular mass of 45.7 kDa and contains two putative glycosylation sites. The 3'-untranslated region contains two putative polyadenylation signals. The cDNA was confirmed to be that for DAC in tryptic peptides from the purified human liver protein. Highest sequence similarity of DAC was found in a series of prokaryotic esterases encompassing the putative active site. Two extended regions of significant sequence homology with hormone-sensitive lipase and with lipase 2 from Moraxella TA144 were identified, whereas similarity to carboxyl esterases was restricted to the region encompassing the putative active site, indicating that DAC should be classified as esterase. This cDNA provides an important tool to study deacetylation and its effects on the metabolic activation of arylamine and heterocyclic amine carcinogens.

Utilization of multiple polyadenylation signals in the human RHOA protooncogene

Little is known regarding the regulation of expression of the RHOA protooncogene, a member of the family of genes encoding Ras-related GTP-binding proteins. We have previously reported that the 3′ untranslated region ( UTR) of RHOA was contained within a genomic sequence which flanked the 5′ end of the human glutathione peroxidase 1-encoding gene [J.A. Moscow et al. J. Biol. Chem. 267 (1992) 5949–5958]. Our previous studies revealed the presence of multiple (1.8 and 1.5kb) RHOA mRNA species in breast cancer cell lines and of three putative polyadenylation signals in the RHOA 3′ UTR. In this report, we have isolated several RHOA cDNAs from a multidrug-resistant MCF-7 human breast cancer cell line. Sequence analyses of these RHOA cDNA clones indicate that multiple polyadenylation signals are used to terminate RHOA transcripts. RNase-protection analysis demonstrated that all three polyadenylation signals are utilized in breast cancer cell lines and RNA stability studies demonstrated that RHOA RNA species with different 3′ ends have equivalent stability. Since little is known about the RNA expression of RHOA in human tumors, and since both activated and non-activated RHOA genes possess transformation potential, we analyzed RHOA mRNA in lung and colon tumors by Northern blot and RNase-protection analyses. In all eight lung tumors examined, RHOA RNA levels were decreased relative to the level in normal surrounding tissue, whereas RHOA expression was decreased in only two of six colon tumors. We also found that lovastatin-induced cell cycle arrest resulted in increased RHOA RNA expression in breast cancer cell lines. RNase-protection analysis of RHOA RNA from these tumor and cell line specimens demonstrated that the relative abundance of RNA transcripts utilizing these three polyadenylation signals did not vary with total RHOA mRNA levels.

Structure of the mouse growth/differentiation factor 9 gene

Using a mouse GDF-9 cDNA as a probe, over 20 kb of sequence encompassing the GDF-9 gene was isolated from a mouse 129SvEv genomic library. Sequence analysis of the exons, exon-intron boundaries, and 5′- and 3′-non-translated regions was used to establish the structure of the mouse GDF-9 gene. The GDF-9 gene is encoded by two exons separated by a 2.9 kb intron. Multiple putative transcription start sites are detected between 31 and 57 bp upstream of the start site of translation and a putative polyadenylation signal lies 342 bp 3′ of the end of translation. Knowledge of the GDF-9 gene structure will enable us to further understand the role of GDF-9 in ovarian physiology and development.

Binding of nuclear factors to a satellite DNA of retroviral origin with marked differences in copy number among species of the rodent Ctenomys.

The major satellite DNA of the subterranean rodent Ctenomys, named RPCS, contains several consensus sequences characteristic of the U3 region of retroviral long terminal repeats (LTRs), such as a polypurine tract, CCAAT boxes, binding sites for the CCAAT/enhancer-binding protein (C/EBP), a TATA box and putative polyadenylation signals. RPCS presents an enormous variation in abundance between species of the same genus: while C. australis or C. talarum have approximately 3 x 10(6) copies per genome, C. opimus has none. A sequence (RPCS-I) with identity to the SV40-enhancer core element, present in all the repeating units of the satellite is specifically protected in DNase I footprintings. Competitions of band-shift assays with different transcription factor binding sites indicate that binding to RPCS-I is specific and involves CCAAT proteins related to NF-1, but not to C/EBP. By the use of quantitative protein/DNA binding assays we determined that, despite of their conspicuous difference in RPCS copy number, C. talarum and C. opimus have equivalent amounts and identical quality of RPCS-binding proteins. These results are consistent with the observation, by in situ hybridization, that RPCS is clustered in heterochromatic regions, where it might have restricted accessibility to transcription factors in vivo. This is the first report of the binding of transcription factors to a satellite DNA of retroviral origin.

The Drosophila forked gene encodes two major RNAs, which, in gypsy or springer insertion mutants, are partially or completely truncated within the 5′-LTR of the inserted retrotransposon

Mutations in the forked (f) gene of Drosophila cause deformation of bristles and hairs. Our molecular analysis showed the f gene to span more than 30 kb, and to encode two major RNAs, 6.0 and 2.5 kb long, both of which are prematurely terminated in gypsy and springer insertion mutants. These truncated RNAs were polyadenylated using putative polyadenylation signals within the 5'-LTR of the inserted retrotransposon. No evidence was found for effects of the retrotransposon insertions on the promoters for transcription of the 6.0 and 2.5 kb RNAs. In f1 and fx, a single gypsy element was found to be inserted at identical sites in the second intron of region encoding the 2.5 kb f RNA and both truncated and wild-type sized RNAs were detected. Recessive mutations at suppressor of forked (su(f)) increased the fraction of wild-type sized RNAs considerably, suggesting that the wild-type su(f) product either stimulates premature termination at the gypsy LTR or inhibits normal splicing. In f36a, a springer element inserted in the third exon of the region encoding the 2.5 kb f RNA completely suppressed the formation of apparently wild-type transcripts.

Isolation and characterization of the gene encoding the rat α 1B adrenergic receptor

Using a rat α 1B adrenergic receptor(AR)-encoding cDNA probe, we isolated two genomic clones from a rat liver genomic DNA library. Southern blot analysis and nucleotide sequencing indicate that the rat α 1B AR gene has two exons and a single large intron of at least 16 kb. Analysis of the sequence of the 5'-flanking region suggests that this gene has the features of a housekeeping gene: it has neither a TATA box nor a CAAT box, but has multiple transcription start points ( tsp), multiple Sp1-binding sites, and a high G + C content. The 5'-flanking region also contains consensus sequences for AP1-and AP2-binding sites, and putative sites for cyclic AMP ( CRE), glucocorticoid (GRE), and thyroid responses ( TRE). The 3'-flanking region contains a putative polyadenylation signal (ATTAAA) 492 bp downstream from the stop codon. A comparison of the rat α 1B AR sequence with that of the recently cloned human α 1B AR gene showed that the overall structure of this gene is highly conserved, with some notable differences in the 5'-flanking region.

Expression of ferrochelatase mRNA in erythroid and non-erythroid cells

Ferrochelatase, which catalyses the last step in haem biosynthesis, i.e. the insertion of Fe(II) into protophorphyrin IX, is present in all cells, but is particularly abundant in erythroid cells during haemoglobinization. Using mouse ferrochelatase cDNA as a probe two ferrochelatase transcripts, having lengths of 2.9 kb and 2.2 kb, were found in extracts of mouse liver, kidney, brain, muscle and spleen, the 2.9 kb transcript being more abundant in the non-erythroid tissues and the 2.2 kb transcript more predominant in spleen. In mouse erythroleukemia cells the 2.9 kb ferrochelatase transcript is also more abundant; however, following induction of erythroid differentiation by dimethyl sulphoxide there is a preferential increase in the 2.2 kb transcript, which eventually predominates. With mouse reticulocytes, the purest immature erythroid cell population available, over 90% of the total ferrochelatase mRNA is present as the 2.2 kb transcript. Since there is probably only one mouse ferrochelatase gene, the occurrence of two ferrochelatase transcripts could arise from the use of two putative polyadenylation signals in the 3' region of ferrochelatase DNA. This possibility was explored by using a 389 bp DNA fragment produced by PCR with synthetic oligoprimers having sequence similarity with a region between the polyadenylation sites. This fragment hybridized only to the 2.9 kb ferrochelatase transcript, indicating that the two transcripts differ at their 3' ends and suggesting that the 2.2 kb transcript results from the utilization of the upstream polyadenylation signal. The preferential utilization of the upstream polyadenylation signal may be an erythroid-specific characteristic of ferrochelatase gene expression.

Complementary DNA Sequence of an Adenylate Translocator from Arabidopsis thaliana

The adenylate translocator (also called ADP/ATP translocator or carrier) has been intensively studied as a protein of the mitochondrial inner membrane (5, 9). In mitochondria, the ADP/ATP translocator mediates the exchange of mitochondrial ATP, generated by oxidative phosphorylation, with cytosolic ADP in a 1:1 exchange. This exchange process links the energy metabolism of the two different compartments. Multiple nuclear genes have been found to code for the translocator in different organisms: two in maize (10), three in yeast (6), and at least three in humans (4). In plants, it has been shown that an adenylate translocator is also present in the inner membranes of amyloplasts (1, 7) and is able to mediate the uptake of ADP-glucose (8). To study the expression of the gene(s) and the localization of the translocator in different plant organelles, we screened a XgtlO cDNA library from Arabidopsis thaliana (L. Heynh, Columbia strain) using the pANT1 cDNA from maize (10) as a probe. Preliminary screening yielded a partial cDNA of 800 bp encoding for the COOH terminus of a protein homologous to the maize translocator as determined by the analysis of the nucleotide sequence. A second screening, using the 800bp partial clone, yielded a putative full-length cDNA of 1393 bp. This cDNA has been called AANT1 (Table I). The nucleotide sequence of this cDNA shows the occurence of an open reading frame of 1137 nucleotides and a 230nucleotide long noncoding 3' end followed by a polyadenylation stretch of 23 nucleotides (Fig. 1). In the 3'-untranslated region, a putative polyadenylation signal 5'-AATAAA-3' is present, 18 nucleotides upstream from the beginning of the polyadenylation sequence. The open reading frame codes a 379-amino acid polypeptide, as deduced from the nucleotide sequence. This translation product shows 87.0, 75.8, 68.4, and 53.2% identities with the maize (MANT1), yeast (AAC2), Neurospora, and human ADP/ATP translocators, respectively (2, 3, 6, 10). The NH2-terminal amino acid regions differ significantly. Indeed, this region is highly heterologous among the different organisms in terms of length and amino acid sequence but also within one species, Saccharomyces cerevisiae, where the three adenylate translocators so far known do not share the same NH2-terminal region (6). According to the Arabidopsis and maize sequences, it seems that the NH2-terminal region of the adenylate translocator is much longer in the higher plant protein. The comparisons between the mol wts as estimated by immunobloting and the ones deduced from genomic or cDNA sequences did not show significant differences for most of the organisms previously cited. However, in the case of the cDNA that we isolated, the mol wt of the deduced protein is 41,295, compared to the apparent mol wt of 31,000 as determined from immunoblot analysis of partially purified proteins of Arabidopsis leaves (data not shown). To account for this discrepancy, one might propose that translation proceeds from another initiation site located further downstream from the first ATG codon. Another possibility may be that a larger precursor protein is made, which undergoes posttranslational processing to yield a smaller mature protein. In such a case, the plant system appears to be different from other reported systems. Further experiments are currently in progress to confirm the length of the full mRNA and to distinguish between these two alternatives.

Genomic organization and expression of the human alpha 1B-adrenergic receptor.

alpha 1-Adrenergic receptors (ARs) are members of the guanine nucleotide-binding protein-coupled receptor superfamily. The genes for all ARs described thus far are intronless. We report here the cloning and the nucleotide sequence of the gene for the human alpha 1B-AR. It consists of two exons and a single large intron of at least 20 kilobases which interrupts the coding region at the end of the putative sixth transmembrane domain. The deduced amino acid sequence of the encoded receptor has a high degree of homology to the cloned hamster, rat, and dog alpha 1B-ARs. To characterize the encoded protein, a fusion gene constructed by splicing together exon 1 and exon 2 was expressed transiently in COS-1 cells. The transfected gene fusion product resulted in the production of an alpha 1B-AR with ligand binding characteristics indistinguishable from those of the expressed hamster alpha 1B cDNA. Evidence that the human alpha 1B-AR gene we have isolated is indeed transcribed is the finding of similar sized (2.8-kilobase) transcripts in human heart and other tissues by Northern blot analysis when either exon 1 or exon 2 is used as a probe. Moreover, using primers designed to span the exon 1/exon 2 boundary, a polymerase chain reaction product generated from single-stranded DNA prepared from human heart mRNA had the exact size and nucleotide sequence predicted for a transcript in which exon 1 is spliced to exon 2. The 5'-flanking region (924 base pairs (bp)) of exon 1 contains neither a TATA box nor a CAAT box but is high in GC content (70%) and contains several Sp1 binding sites (GC boxes), consistent with promoters described for housekeeping genes. The 5'-untranslated region also contains a putative cyclic AMP response element. Primer extension studies and RNase protection assays suggested that there are several potential transcription start sites in most tissues with a predominant site located 173 bp upstream from the translation start site. The 3'-flanking region contains a putative polyadenylation signal (ATTAAA) 492 bp downstream from the stop codon. The genomic organization of the human alpha 1B-AR with a single large intron interrupting its coding region differs from those of other ARs as well as muscarinic and 5-hydroxy-tryptamine receptors, which are intronless. The location of the intron in the human alpha 1B-AR gene is also unique among those members of the G-protein-coupled receptor family that do possess introns.(ABSTRACT TRUNCATED AT 400 WORDS)

Characterization by cDNA cloning of the mRNA of a new growth factor from bovine seminal plasma: Acidic seminal fluid protein

A cDNA expression library in λgt11 prepared from cDNA derived of seminal vesicle tissue was screened by means of monospecific rabbit anti-aSFP IgG. The sequence of clone pTF21, containing an insert of 668 bp comprised an open reading frame from position 7 to 411 terminated by two stop codons. From this sequence a protein of 134 amino acid residues can be deduced. The mature aSFP was preceded by a signal peptide of 20 amino acids length. The protein sequence contains no signal for N-glycosylation. The molecular weight calculated from the amino acid sequence is 12922 Da. The start codon ATG is part of the sequence AAG ATG A which fulfills the criteria of an initiation consensus sequence. The coding region was followed by 257bp of the complete 3′-untranslated region (3′UTR). A putative polyadenylation signal AATAAT, although not of the standard type, is observed at position 650. According to Northern analysis, aSFP mRNA is expressed in seminal vesicle tissue, ampulla and weakly in tissue of epididymis, but not in testis or other bovine tissue. aSFP is specified by a single copy gene. Attempts to detect homologies to known protein sequences were not successful.

Dissection of a pollen-specific promoter from maize by transient transformation assays

We have previously reported the isolation and characterization of a gene (Zm13) from Zea mays which shows a pollen-specific pattern of expression. Stably transformed tobacco plants containing a reporter gene linked to portions of the Zm13 5' flanking region show correct temporal and spatial expression of the gene. Here we present a more detailed analysis of the 5' regions responsible for expression in pollen by utilizing a transient expression system. Constructs containing the beta-glucuronidase (GUS) gene under the control of various sized fragments of the Zm 13 5' flanking region were introduced into Tradescantia and Zea mays pollen via high-velocity microprojectile bombardment, and monitored both visually and with a fluorescence assay. The results suggest that sequences necessary for expression in pollen are present in a region from -100 to -54, while other sequences which amplify that expression reside between -260 and -100. The replacement of the normal terminator with a portion of the Zm13 3' region containing the putative polyadenylation signal and site also increased GUS expression. While the -260 to -100 region contains sequences similar to other protein-binding domains reported for plants, the -100 to -54 region appears to contain no significant homology to other known promoter fragments which direct pollen-specific expression. The microprojectile bombardment of Tradescantia pollen appears to be a good test system for assaying maize and possibly other monocot promoter constructs for pollen expression.

Normal Chlamydomonas nuclear gene structure on linkage group xix

The unusual Chlamydomonas linkage group XIX-called the uni linkage group for the uni mutants that lack one of the paired flagellae of wild-type cells--has been reported to be physically located exclusively at the basal bodies. To learn whether the structure of genes on this linkage group differs from the structure of nuclear genes in this organism, we determined the primary structure of a gene that maps to linkage group XIX. This analysis reveals the presence of nine intervening sequences; the nucleotides at exon/intron boundaries conform with nuclear gene intron junction sequences. Also typical for C. reinhardtii nuclear genes are the position and sequence of the putative polyadenylation signal. These findings suggest that transcripts from linkage group XIX are likely to be processed in the nucleus. The open reading frame, which displays weak but easily detected Chlamydomonas codon bias, potentially encodes a protein similar to a membrane anchor for cytoskeletal proteins. The observation that expression of this gene is regulated during interphase and in gametes is not consistent with the hypothesis that linkage group XIX may be expressed only during mitotic and meiotic processes.

Amber codon suppression: the in vivo and in vitro analysis of two C-hordein genes from barley

A 1420 bp genomic fragment (lambda-hor1-17) encompassing a Hor-1 gene encoding a C-hordein polypeptide is presented. The deduced amino acid sequence is 261 residues long. It comprises a 20 amino acid signal peptide, unique NH2- and COOH-terminal regions and a coding region comprised of pentapeptide (PQQPY) and octapeptide (PQQPFPQQ) repeat motifs. The 431 bp of 5' non-coding region contains a 'TATA box' at -105, a 'CACA box' (-181 to -201) and a -300 prolamin element. In the 3' non-coding region there are two putative polyadenylation signals located 88 and 142 bp downstream of the stop codon. The structure of lambda-hor1-17 is compared with that of another gene (lambda-hor1-14) encoding a C-hordein polypeptide, which contains an amber codon interrupting the ORF. A functional assay in which the 5' non-coding regions of the two genes were fused to the beta-glucuronidase (GUS) gene demonstrated that both genes were transcriptionally active and that circa 430 bp of the C-hordein promoters were sufficient to drive the expression of the GUS gene in developing barley endosperms. It also demonstrated that both promoters had transcriptional efficiencies comparable with that of the 35S CaMV promoter. The in vitro translation of the coding region of lambda-hor1-14 in the wheat germ system showed that the premature stop codon could be partially suppressed. The suppression was also demonstrated in a transient expression assay in vivo using isolated barley endosperms.

Cloning and expression of a new human polypeptide which regulates protein phosphorylation in Escherichia coli

A 1,820 bp full-length clone encoding for a new human protein was isolated from a lambda gt11 placental cDNA library using anti-human hexokinase antibodies. The cDNA complete sequence includes a 12 bp 5' non-coding region, a single open reading frame encoding a protein of 55 KDa (HP-10) and a 177 bp non-coding with two putative polyadenylation signals upstream of 3' poly(A)tail. The deduced amino acid sequence reveals a sequence of 492 amino acids that contains a stretch of 7 glutamic acid from position 169 and one potential glycosylation site at position 274. Although antibodies against hexokinase recognize the fusion protein and antibodies against the fusion protein recognize hexokinase, HP-10 is not human hexokinase, by a number of criteria including the alignment of determined amino acid sequences. In searching for a possible functional role of HP-10 its cDNA was inserted into a procaryotic vector which allows the expression of the non-fused protein. Bacteria expressing the HP-10 encoded protein were isolated and found to have a dramatic increase in endogenous phosphorylated proteins. Since HP-10 does not have a protein kinase activity per se it should be considered a new regulatory phosphorylation protein which is active in E. coli.

Human Spermidine Synthase Gene: Structure and Chromosomal Localization

The human spermidine synthase (EC 2.5.1.16) gene was isolated from a genomic library constructed with DNA obtained from a human immunoglobulin G (IgG) myeloma cell line. Subsequent sequence analyses revealed that the gene comprised of 5,818 nucleotides from the cap site to the last A of the putative polyadenylation signal with 8 exons and 7 intervening sequences. The 5'-flanking region of the gene was extremely GC rich, lacking any TATA box but containing CCAAT consensus sequences. No perfect consensus sequence for the cAMP-responsive element for the AP-1 binding site was found, yet the gene contained seven AP-2 binding site consensus sequences. The putative polyadenylation signal was an unusual AATACA instead of AATAAA. Polymerase chain reaction analysis with DNA obtained from human x hamster somatic cell hybrids indicated that human spermidine synthase genomic sequences segregate with human chromosome 1. Transfection of the genomic clone into Chinese hamster ovary cells displaying a low endogenous spermidine synthase activity revealed that the gene was transiently expressed and hence in all likelihood represents a functional gene.

Influence of the poly(A) tail and putative polyadenylation signal on the infectivity of white clover mosaic potexvirus

A series of mutations has been constructed in a cDNA clone of white clover mosaic virus (WCIMV) which decreases the poly(A) tail length of run-off transcripts from 74 to 27, 10 or zero 3′-terminal (A) residues. Although transcripts with short poly(A) tails were less infectious than wild-type RNA, complete removal of the (A) tail did not abolish infectivity. Addition of nonviral nucleotides to the 3′ terminus of transcripts with no 3′-terminal (A) residues eliminated infectivity. Heterogenous-length poly(A) tails, indistinguishable from the wild type, were synthesized de novo in plants inoculated with transcripts with no 3′-terminal (A) residues, demonstrating the presence of a poly(A) polymerase activity in WCIMV-infected tissue. Mutation of a putative polyadenylation motif found in the 3′ noncoding region of WCIMV decreased the efficiency of polyadenylation of the progeny of transcripts with 10 3′-terminal (A) residues. The same mutation in transcripts with no 3′-terminal (A) residues abolished infectivity.

Variations of nucleotide sequences and transcription of the SH gene among mumps virus strains

We have compared nucleotide sequences of the SH genes as well as their flanking regions of six mumps virus strains and found a high amino acid diversity (up to 23%) of the putative SH proteins among these strains. It was found, in addition, that one of these strains (Enders strain) contained a point mutation in putative polyadenylation signal for the F gene mRNA (TTTAGAAAAAAA to TTTAGAAGAAAA ). Northern blot analysis showed that the Enders strain was also unique in that neither monocistronic SH nor bicistronic SH-HN mRNA could be detected in the cells infected with this particular strain.

Multiple polyadenylation sites are active in the α 1-tubulin gene from Zea mays

Two cDNA clones (MR2 and MR29) encoding the same α-tubulin isotype (α1) have been identified and characterized from maize roots. The sequence of these two cDNA clones is identical to a previously described cDNA clone (MR19) except in the location of their polyadenylation site. The sequence of the cDNA and its homologous genomic clone (MG 19 14 ) shows two putative polyadenylation signals which could direct the variable 3′ processing of the observed transcripts. Endonuclease Sl protection analysis in this 3' flanking region confirms the presence in the α1-tubulin gene from Zea mays of these two main functional polyadenylation sites and possibly other related ones. The relative accumulation of RNAs bearing the two main polyadenylation sites has been tested by using a RNA-slot analysis of several tissues of the plant. It appears that a higher proportion of shorter mRNA species is found in actively dividing tissues.

Molecular cloning and sequence analysis of cDNA encoding human ferrochelatase

The cDNA encoding human ferrochelatase [EC 4.99.1.1] was isolated from a human placenta cDNA library in bacteriophage lambda gt11 by screening with a radiolabeled fragment of mouse ferrochelatase cDNA. The cDNA had an open reading frame of 1269 base pairs (bp) encoding a protein of 423 amino acid residues (Mr. 47,833) with alternative putative polyadenylation signals in the 3' non-coding regions and poly (A) tails. Amino acid sequencing showed that the mature protein consists of 369 amino acid residues (Mr. 42,158) with a putative leader sequence of 54 amino acid residues. The human enzyme showed an 88% identity to mouse enzyme and 46% to yeast enzyme. Northern blot analysis showed two mRNAs of about 2500 and 1600 bp for ferrochelatase in K562 and HepG2 cells. As full-length cDNA for human ferrochelatase is now available, molecular lesions related to erythropoietic protoporphyria can be characterized.

Atlantic Salmon (Salmo salar) Serum Albumin: cDNA Sequence, Evolution, and Tissue Expression

Atlantic salmon serum albumin is one of the most abundant proteins in salmon liver, representing 1.6% of all clones in a cDNA library made from salmon liver RNA. The DNA from a number of clones was sequenced to reveal an open reading frame of 1,827 bases encoding a 608-amino-acid protein. The sequenced 5' untranslated region is 69 bases long and the 3' untranslated region contains two putative polyadenylation signals and poly(A) tail. Sequence analysis of different clones indicates the presence of a second cDNA for salmon serum albumin. Multiple alignments of salmon serum albumin deduced amino acid sequence with Xenopus laevis, rat, bovine, and human serum albumins shows significant conservation of cysteine residues. The triple domain structure of serum albumin proteins is maintained. Unlike mammalian systems where serum albumin expression appears to be specific to liver only, salmon serum albumin is expressed in muscle also.