Role of the conserved AAUAAA sequence: four AAUAAA point mutants prevent messenger RNA 3' end formation

Analysis of the primary sequence and microtubule-binding region of the Drosophila 205K MAP

We have sequenced cDNA clones encoding the Drosophila 205K microtubule-associated protein (MAP), a protein that may be the species specific homologue of mammalian MAP4. The peptide sequence deduced from the longest open-reading frame reveals a hydrophilic protein, which has basic and acidic regions that are similar in organization to mammalian MAP2. Using truncated forms of the 205K MAP, a 232-amino acid region could be defined that is necessary for microtubule binding. The amino acid sequence of this region shares no similarity with the binding motif of MAP2 or tau. We also analyzed several embryonic cDNA clones, which show the existence of differentially spliced mRNAs. Finally, we identified several potential protein kinase target sequences. One of these is distal to the microtubule-binding site and fits the phosphorylation consensus sequence of proteins phosphorylated by the mitosis specific protein kinase cdc2. Our data suggest that the 205K MAP uses a microtubule-binding motif unlike that found in other MAPs, and also raise the possibility that the activities of the 205K MAP may be regulated by alternative splicing and phosphorylation.

Structure of the gene encoding the M1 protein of sonchus yellow net virus

The gene encoding the M1 protein of sonchus yellow net virus (SYNV), a plant rhabdovirus, has been sequenced and identified by Western blot analysis of SYNV proteins using antibodies directed against a fusion protein derived from a portion of the sequenced gene. The M1 gene is positioned between nucleotides 4039 and 5109 relative to the 3' end of the viral RNA and is the fourth gene from the 3' end of the genome. The 1071-nucleotide (nt) M1 gene lies between a putative nonstructural gene of unknown function and the gene encoding the glycoprotein and is bordered on either side by the same GG intergenic dinucleotide that separates other genes in the SYNV genome. The M1 mRNA (scRNA 6) consists of a 71-nt untranslated region at the 5' terminus followed by an 858-nt open reading frame (ORF) capable of encoding a protein with a calculated molecular weight of 31,779. The amino acid sequence deduced from this ORF is not highly homologous to those of other rhabdovirus matrix proteins, but has some localized regions of similarity. The UGA codon that terminates the M1 ORF is followed by a 3' untranslated region of 142 nt. The viral RNA (minus-sense) sequence corresponding to the extreme 3' end of the mRNA contains a 9-nt tract (3'-AUUGUUUUU-5') that is identical to the sequences at the termini of other SYNV genes.

Characterization of three distinct size classes of human 3-hydroxy-3-methylglutaryl coenzyme A reductase mRNA: expression of the transcripts in hepatic and nonhepatic cells

3-Hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase mRNA is expressed in two highly differentiated human hepatoma cell lines, HepG2 and Hep3B, at exceptionally high levels relative to human fetal liver and fibroblasts. Blotting experiments revealed that the mRNA consists of three major size classes of approximately 4.7, 4.5, and 4.2 kb that responded coordinately to agents that alter HMG-CoA reductase activity. In view of the markedly elevated levels of reductase mRNA in the hepatoma cell lines, we compared the pattern of transcriptional initiation in these cells with those in normal liver and fibroblasts. These analyses revealed a complex pattern of initiation sites, all of which were suppressed by oxysterols, extending over approximately 300 nucleotides. However, all of the major sites detected in the hepatomas could also be found in human liver and fibroblasts. Heterogeneity of transcriptional initiation does not account for the three major size classes of mRNA detected by RNA blotting. RNase H mapping demonstrates that these are produced by use of three polyadenylation sites. To determine the extent to which these sites have been conserved between the human gene and the previously characterized Chinese hamster gene, we cloned and sequenced the 3' untranslated region of the longest form of the human mRNA. These studies revealed that, despite a high overall degree of sequence conservation, the spectrum of polyadenylation sites used differs qualitatively between the two species. Features of the mRNA sequence that may contribute to these differences are described.

Point mutations in AAUAAA and the poly (A) addition site: effects on the accuracy and efficiency of cleavage and polyadenylation in vitro

Three sequences in the vicinity of poly (A) addition sites are conserved among vertebrate mRNAs. We analyze the effects of single base changes in each position of AAUAAA and in the nucleotide to which poly (A) is added on 3' end formation in vitro. All 18 possible single base changes of the AAUAAA sequence greatly reduce addition of poly (A) to RNAs that end at the poly (A) addition site, and prevent cleavage of RNAs that extend beyond. The magnitude of reduction varies greatly with the position changed and the base introduced. For any given mutation, cleavage and polyadenylation are reduced to similar extents, strongly suggesting that the same factor interacts with AAUAAA in both reactions. Mutations at and near the conserved adenosine to which poly (A) is added disturb the accuracy, but not the efficiency, of 3' end formation. For example, point mutations at the conserved adenosine shift the 3' end of the most abundant 5' half-molecule downstream by a single nucleotide. The mechanism by which these mutations might exert their effects on the precision of 3' end formation are discussed.

Molecular cloning and characterization of rat liver catechol-O-methyltransferase

The coding sequence of rat liver catechol-O-methyl-transferase (COMT; EC 2.1.1.6) was determined from rat cDNA and genomic libraries were screened with DNA probes and specific antiserum. The open reading frame consisted of 663 nucleotides coding for a 221-amino acid (aa) polypeptide with a deduced Mr of 24,747. No obvious hydrophobic signal sequence, membrane-spanning domains, or potential N-glycosylation sites were found in this sequence. The identity of the clone and the accuracy of the sequence was verified by direct aa sequencing of the tryptic peptides derived from the purified rat liver enzyme. Primer extension analysis showed that the transcription start point of the rat liver COMT mRNA was 450 bp upstream from the translation start codon. A putative polyadenylation signal (ATTAAA) was found in the 3'-noncoding region. The predicted size of the COMT transcript was 1.8-2.0 kb, which could be confirmed from Northern hybridization analyses of the isolated rat liver mRNA. One polypeptide of 25 kDa, could be immunoprecipitated with anti-COMT antibody from in vitro translation of rat liver mRNA. Employing the DNA blot analysis only one COMT-encoding gene was found in the rat genome.

A presumed DNA helicase encoded by ERCC-3 is involved in the human repair disorders xeroderma pigmentosum and Cockayne's syndrome

The human gene ERCC-3 specifically corrects the defect in an early step of the DNA excision repair pathway of UV-sensitive rodent mutants of complementation group 3. The predicted 782 amino acid ERCC-3 protein harbors putative nucleotide, chromatin, and helix-turn-helix DNA binding domains and seven consecutive motifs conserved between two superfamilies of DNA and RNA helicases, strongly suggesting that it is a DNA repair helicase. ERCC-3-deficient rodent mutants phenotypically resemble the human repair syndrome xeroderma pigmentosum (XP). ERCC-3 specifically corrects the excision defect in one of the eight XP complementation groups, XP-B. The sole XP-B patient presents an exceptional conjunction of two rare repair disorders: XP and Cockayne's syndrome. This patient's DNA contains a C----A transversion in the splice acceptor sequence of the last intron of the only ERCC-3 allele that is detectably expressed, leading to a 4 bp insertion in the mRNA and an inactivating frameshift in the C-terminus of the protein. Because XP is associated with predisposition to skin cancer, ERCC-3 can be considered a tumor-preventing gene.

Evidence for an alternate splicing in the thyroperoxidase messenger from patients with Graves' disease

An initial lambda gt 11 cDNA library constructed from a human Graves' patient thyroid was screened with an immunopurified rabbit anti-human thyroperoxidase (hTPO) polyclonal antibody. A 869 bp clone was obtained. It presents a 130 bp deletion as compared to the published sequence and a 77 bp insertion in the 3' non-coding region. Screening of a pUC cDNA library from another Graves' patient thyroid exhibited the same 130 bp deletion in two other cDNA clones. PCR analysis of mRNA transcripts confirmed the presence of the two messengers in two other Graves' thyroid tissues. In all the cases, this new spliced mRNA species represents between 40% and 50% of the total hTPO mRNAs. With respect to the structure of the hTPO gene, the present deletion suggests an alternate splicing of exon 16. The juxtaposition of exon 17 to exon 15 encoding the transmembrane domain leads to a shift in the reading frame. By the use of a different stop codon, the spliced mRNA generates a modified 56 - COOH terminal aminoacids (aa) sequence.

Production of shorter mRNA for protein phosphatase 2A beta by alternative poly(A) addition

Two types of cDNA clones of protein phosphatase 2A beta, PP-2A beta L and PP-2A beta S, were isolated from a rat testis cDNA library. PP-2A beta L was 1.8 kb in length and corresponded to the 2.0 kb mRNA that is the major species expressed in somatic tissues. PP-2A beta S was 1.3 kb in length and corresponded to the 1.4 kb mRNA that is the major species expressed in germ cells of the testis. Nucleotide sequencing of these two clones revealed that their open reading frames were identical. The only difference between these two cDNAs was at the 3' non-coding region due to alternative poly(A) addition: the 1.8 kb cDNA had a consensus poly(A) addition signal, AATAAA, but the 1.3 kb cDNA had an unusual poly(A) addition signal, probably with the sequence AATATA. These data suggest that in germ cells there may be some mechanism that utilizes the specific poly(A) addition signal.

RNA processing in vitro produces mature 3' ends of a variety of Saccharomyces cerevisiae mRNAs

Ammonium sulfate fractionation of a Saccharomyces cerevisiae whole-cell extract yielded a preparation which carried out correct and efficient endonucleolytic cleavage and polyadenylation of yeast precursor mRNA substrates corresponding to a variety of yeast genes. These included CYC1 (iso-1-cytochrome c), HIS4 (histidine biosynthesis), GAL7 (galactose-1-phosphate uridyltransferase), H2B2 (histone H2B2), PRT2 (a protein of unknown function), and CBP1 (cytochrome b mRNA processing). The reaction processed these pre-mRNAs with varying efficiencies, with cleavage and polyadenylation exceeding 70% in some cases. In each case, the poly(A) tail corresponded to the addition of approximately 60 adenosine residues, which agrees with the usual length of poly(A) tails formed in vivo. Addition of cordycepin triphosphate or substitution of CTP for ATP in these reactions inhibited polyadenylation but not endonucleolytic cleavage and resulted in accumulation of the cleaved RNA product. Although this system readily generated yeast mRNA 3' ends, no processing occurred on a human alpha-globin pre-mRNA containing the highly conserved AAUAAA polyadenylation signal of higher eucaryotes. This sequence and adjacent signals used in mammalian systems are thus not sufficient to direct mRNA 3' end formation in yeast. Despite the lack of a highly conserved nucleotide sequence signal, the same purified fraction processed the 3' ends of a variety of unrelated yeast pre-mRNAs, suggesting that endonuclease cleavage and polyadenylation may produce the mature 3' ends of all mRNAs in S. cerevisiae.

A genomic clone encoding a novel proliferation-dependent histone H2A.1 mRNA enriched in the poly(A)+ fraction

Replication-dependent histone mRNAs are prime examples of nonpolyadenylated mRNAs. We isolated and characterized cDNAs and a genomic clone for a replication-dependent histone H2A.1 mRNA which segregated into the poly(A)+ fraction during mRNA isolation through an oligo(dT)-cellulose column. However, the results of sequencing of the genomic clone suggested that the mRNA did not contain a poly(A) tail. Instead, the genomic sequence revealed a nonterminal oligo(A) tract directly upstream from the typical 3'-terminal hairpin loop of replication-dependent histone mRNAs. The nonterminal oligo(A) tract consisted of 14 adenylate residues interrupted by one guanylate residue (A4GA10). We concluded that this short oligo(A) stretch mediated binding of the mRNA to oligo(dT) even after stringent washes with 0.1 M NaCl, indicating that rather short oligo(A) sequences can ensure binding to oligo(dT)-cellulose. The cDNA and genomic clones contained an AAATAAG sequence at the end of the coding region. It has been suggested that this sequence contains a polyadenylation signal in some yeast and mouse transcripts, but it does not function as a polyadenylation signal in the histone transcript described in this paper.

A genomic clone encoding a novel proliferation-dependent histone H2A.1 mRNA enriched in the poly(A)+ fraction

Replication-dependent histone mRNAs are prime examples of nonpolyadenylated mRNAs. We isolated and characterized cDNAs and a genomic clone for a replication-dependent histone H2A.1 mRNA which segregated into the poly(A)+ fraction during mRNA isolation through an oligo(dT)-cellulose column. However, the results of sequencing of the genomic clone suggested that the mRNA did not contain a poly(A) tail. Instead, the genomic sequence revealed a nonterminal oligo(A) tract directly upstream from the typical 3'-terminal hairpin loop of replication-dependent histone mRNAs. The nonterminal oligo(A) tract consisted of 14 adenylate residues interrupted by one guanylate residue (A4GA10). We concluded that this short oligo(A) stretch mediated binding of the mRNA to oligo(dT) even after stringent washes with 0.1 M NaCl, indicating that rather short oligo(A) sequences can ensure binding to oligo(dT)-cellulose. The cDNA and genomic clones contained an AAATAAG sequence at the end of the coding region. It has been suggested that this sequence contains a polyadenylation signal in some yeast and mouse transcripts, but it does not function as a polyadenylation signal in the histone transcript described in this paper.

Isolation and expression of the gene for a major surface protein of Giardia lamblia

To study the interactions between the parasitic protozoan Giardia lamblia and its environment, we have cloned the gene that encodes the two major surface-labeled trophozoite protein species. Sequence analysis of this gene reveals a single open reading frame specifying a hydrophilic, cysteine-rich (11.8%) protein of 72.5-kDa molecular mass with an amino-terminal signal peptide and a postulated hydrophobic membrane-spanning anchor region near the carboxyl terminus. Most of the cysteine residues (58 of 84) are in the motif Cys-Xaa-Xaa-Cys, which is dispersed 29 times throughout the sequence. Antibodies against the recombinant protein react with the entire surface of live trophozoites, including flagella and adhesive disc. These antibodies inhibit trophozoite attachment, prevent growth, and immunoprecipitate the major approximately 66- and 85-kDa proteins from surface-labeled live trophozoites. The recombinant Escherichia coli also expresses polypeptides of approximately 66- and 85-kDa molecular mass, which are not fusion proteins. This suggests that the processing and/or conformational changes that lead to production of these two peptide species in E. coli reflect those that occur in Giardia. The abundance of cysteine residues suggests that the native proteins on the parasite surface may contain numerous disulfide bonds, which would promote resistance to intestinal fluid proteases and to the detergent activity of bile salts and would help to explain the survival of Giardia in the human small intestine.

RNA processing in vitro produces mature 3' ends of a variety of Saccharomyces cerevisiae mRNAs

Ammonium sulfate fractionation of a Saccharomyces cerevisiae whole-cell extract yielded a preparation which carried out correct and efficient endonucleolytic cleavage and polyadenylation of yeast precursor mRNA substrates corresponding to a variety of yeast genes. These included CYC1 (iso-1-cytochrome c), HIS4 (histidine biosynthesis), GAL7 (galactose-1-phosphate uridyltransferase), H2B2 (histone H2B2), PRT2 (a protein of unknown function), and CBP1 (cytochrome b mRNA processing). The reaction processed these pre-mRNAs with varying efficiencies, with cleavage and polyadenylation exceeding 70% in some cases. In each case, the poly(A) tail corresponded to the addition of approximately 60 adenosine residues, which agrees with the usual length of poly(A) tails formed in vivo. Addition of cordycepin triphosphate or substitution of CTP for ATP in these reactions inhibited polyadenylation but not endonucleolytic cleavage and resulted in accumulation of the cleaved RNA product. Although this system readily generated yeast mRNA 3' ends, no processing occurred on a human alpha-globin pre-mRNA containing the highly conserved AAUAAA polyadenylation signal of higher eucaryotes. This sequence and adjacent signals used in mammalian systems are thus not sufficient to direct mRNA 3' end formation in yeast. Despite the lack of a highly conserved nucleotide sequence signal, the same purified fraction processed the 3' ends of a variety of unrelated yeast pre-mRNAs, suggesting that endonuclease cleavage and polyadenylation may produce the mature 3' ends of all mRNAs in S. cerevisiae.

Rat RI beta isoform of type I regulatory subunit of cyclic adenosine monophosphate-dependent protein kinase: cDNA sequence analysis, mRNA tissue specificity, and rat/mouse difference in expression in testis

A rat complementary DNA (cDNA) for the RI beta isoform of type I cyclic adenosine monophosphate (cAMP)-dependent protein kinase regulatory subunit was cloned and sequenced and was found to contain the entire protein coding and 3'-untranslated regions, with a single (ATTAAA) poly-adenylation site. The largest open reading frame was preceded by a short out-of-phase open reading frame, which is not seen in the corresponding mouse RI beta cDNA due to a single base substitution. The rat RI beta cDNA clone was 2,374 bases long and detected a rat mRNA of approximately 2.8 kilobases. Rat RI beta mRNA was abundant in adult rat brain and testis but was undetectable in other rat tissues. The rat RI beta cDNA also detected RI beta mRNA in mouse brain, but not mouse testis, from 10-week-old BALB/c or 10- and 6-week-old Swiss Webster mice. Thus, despite a 96% nucleotide identity in the coding region of RI beta in rat vs. mouse, there are at least two differences in these closely related species. First, there is a short open reading frame, which precedes the coding region in the rat but not the mouse. Second, unlike the mouse testis, the rat testis contains abundant levels of RI beta mRNA.

Partial amino acid sequence of the major intrinsic protein (MIP) of the chicken lens deduced from the nucleotide sequence of a cDNA clone

A cDNA clone of the major intrinsic protein (MIP) of the chicken lens was isolated. This clone covers the C-terminal half of the coding region and 3'-untranslated region including a polyadenylation signal. Comparison with the bovine MIP cDNA sequence revealed that: (1) the amphilphilic transmembrane helix in bovine MIP is highly hydrophobic in chicken MIP, and is thus unlikely to offer a hydrophilic lining of the transmembrane pore, and (2) the possible calmodulin binding site is conserved especially at amino acid residues which are postulated to be important in its binding with calmodulin. Northern blotting revealed the presence of transcripts of different lengths, two of which correspond closely to the transcripts of bovine MIP.

Unusual features of integrated cDNAs generated by infection with genome-free retroviruses

We previously demonstrated that when nonretroviral RNAs are encapsidated in retroviral particles they can be reverse transcribed into cDNAs, which are then integrated into the cellular genome. This transfer of genetic information via retroviral infection has been designated retrofection. Further analyses of three genes transferred in this manner (retrogenes) revealed that each was present in a single copy at a different site in the recipient quail cell genome and included a transcriptional promoter encoded by the encapsidated neo RNA. A unique feature of the retrogenes was a common 16-nucleotide sequence at or near a recombination border, which was not present in either recombination partner. The existence of this sequence suggests a common mechanism of retrogene formation and/or integration mediated by retrofection.

Mediation of meiotic and early mitotic chromosome segregation in Drosophila by a protein related to kinesin

Contrary to the traditional view that microtubules pull chromosomes polewards during the anaphase stage of meiotic and mitotic cell divisions, new evidence suggests that the chromosome movements are driven by a motor located at the kinetochore. The process of chromosome segregation involves proper arrangement of kinetochores for spindle attachment, followed by spindle attachment and chromosome movement. Mechanisms in Drosophila for chromosome segregation in meiosis differ in males and females, implying the action of different gene products in the two sexes. A product encoded at the claret locus in Drosophila is required for normal chromosome segregation in meiosis in females and in early mitotic divisions of the embryo. Here we show that the predicted amino-acid sequence of this product is related to the heavy chain of kinesin. The conserved region corresponds to the kinesin motor domain and includes the ATP-binding site and a region that can bind microtubules. A second region contains a leucine repeat motif which may mediate protein-subunit interactions necessary for attachment of chromosomes to the spindle. The mutant phenotype of chromosome nondisjunction and loss, and its similarity to the kinesin ATP-binding domain, suggest that the product encoded at claret not only stabilizes chromosome attachments to the spindle, but may also be a motor that drives chromosome segregation in female meiosis.

Unusual features of integrated cDNAs generated by infection with genome-free retroviruses

We previously demonstrated that when nonretroviral RNAs are encapsidated in retroviral particles they can be reverse transcribed into cDNAs, which are then integrated into the cellular genome. This transfer of genetic information via retroviral infection has been designated retrofection. Further analyses of three genes transferred in this manner (retrogenes) revealed that each was present in a single copy at a different site in the recipient quail cell genome and included a transcriptional promoter encoded by the encapsidated neo RNA. A unique feature of the retrogenes was a common 16-nucleotide sequence at or near a recombination border, which was not present in either recombination partner. The existence of this sequence suggests a common mechanism of retrogene formation and/or integration mediated by retrofection.

Sequences 5' to the polyadenylation signal mediate differential poly(A) site use in hepatitis B viruses

Most genetic elements that employ reverse transcription generate a terminally redundant genomic RNA that serves as the template for this reaction. Because the identical polyadenylation signal is present in each terminally redundant segment, synthesis of this RNA requires that this signal be ignored on the first pass of the transcription machinery, then recognized and used on the second pass. We have studied the mechanism of this differential poly(A) site use in one family of retroid elements, the hepatitis B viruses (hepadnaviruses). Our results indicate that two features are involved: the presence of a variant poly(A) signal (TATAAA) and the participation of multiple sequences 5' to this signal that act to increase the efficiency of its use. Deletion of these upstream elements abolishes proper poly(A) site use, despite the presence of the poly(A) signal and downstream GT- and T-rich motifs known to be required for polyadenylation. Sequences from the corresponding regions of retroviral genomes can restore proper processing to these hepadnaviral deletion mutants. Thus, functionally analogous upstream elements exist in other classes of retroid elements, including those employing the canonical AATAAA hexanucleotide signal.

A multicomponent complex is required for the AAUAAA-dependent cross-linking of a 64-kilodalton protein to polyadenylation substrates

A 64-kilodalton (kDa) polypeptide that is cross-linked by UV light specifically to polyadenylation substrate RNAs containing a functional AAUAAA element has been identified previously. Fractionated HeLa nuclear components that can be combined to regenerate efficient and accurate polyadenylation in vitro have now been screened for the presence of the 64-kDa protein. None of the individual components contained an activity which could generate the 64-kDa species upon UV cross-linking in the presence of substrate RNA. It was necessary to mix two components, cleavage stimulation factor and specificity factor, to reconstitute 64-kDa protein-RNA cross-linking. The addition of cleavage factors to this mixture very efficiently reconstituted the AAUAAA-specific 64-kDa protein-RNA interaction. The 64-kDa protein, therefore, is present in highly purified, reconstituted polyadenylation reactions. However, it is necessary to form a multicomponent complex to efficiently cross-link the protein to a substrate RNA.

A multicomponent complex is required for the AAUAAA-dependent cross-linking of a 64-kilodalton protein to polyadenylation substrates

A 64-kilodalton (kDa) polypeptide that is cross-linked by UV light specifically to polyadenylation substrate RNAs containing a functional AAUAAA element has been identified previously. Fractionated HeLa nuclear components that can be combined to regenerate efficient and accurate polyadenylation in vitro have now been screened for the presence of the 64-kDa protein. None of the individual components contained an activity which could generate the 64-kDa species upon UV cross-linking in the presence of substrate RNA. It was necessary to mix two components, cleavage stimulation factor and specificity factor, to reconstitute 64-kDa protein-RNA cross-linking. The addition of cleavage factors to this mixture very efficiently reconstituted the AAUAAA-specific 64-kDa protein-RNA interaction. The 64-kDa protein, therefore, is present in highly purified, reconstituted polyadenylation reactions. However, it is necessary to form a multicomponent complex to efficiently cross-link the protein to a substrate RNA.

The enzyme that adds poly(A) to mRNAs is a classical poly(A) polymerase

Virtually all mRNAs in eucaryotes end in a poly(A) tail. This tail is added posttranscriptionally. In this report, we demonstrate that the enzyme that catalyzes this modification is identical with an activity first identified 30 years ago, the function of which was previously unknown. This enzyme, poly(A) polymerase, lacks any intrinsic specificity for its mRNA substrate but gains specificity by interacting with distinct molecules: a poly(A) polymerase from calf thymus, when combined with specificity factor(s) from cultured human cells, specifically and efficiently polyadenylates only appropriate mRNA substrates. Our results thus demonstrate that this polymerase is responsible for the addition of poly(A) to mRNAs and that its interaction with specificity factors is conserved.

The mouse surfeit locus contains a cluster of six genes associated with four CpG-rich islands in 32 kilobases of genomic DNA

The clustered arrangement (no two adjacent genes are separated by more than 73 base pairs [bp] and two genes overlap by 133 bp at their 3' ends) of the four genes (Surf-1 to -4) identified so far in the mouse surfeit locus (T. Williams, J. Yon, C. Huxley, and M. Fried, Proc. Natl. Acad. Sci. USA 85:3527-3530, 1988) is the tightest gene clustering found in any mammalian genome to date and strongly suggests the possibility of cis-interaction and/or coregulation of gene expression. Thus, we are analyzing the surfeit genes in detail and are defining the extent of the cluster. Here we present the sequence of the entire Surf-4 gene and define the 3' and 5' extents of its mRNAs. The Surf-4 gene has heterogeneous transcriptional start sites, and its 5' end lies in a CpG-rich island. The gene specifies three mRNAs, with the two most abundant mRNAs differing in the locations of their 3' polyadenylation sites. Only the most abundant Surf-4 mRNA would overlap the 3' end of the Surf-2 gene by 133 bp. Two new genes (Surf-5 and Surf-6) have been identified in the surfeit gene cluster by Northern (RNA) blot analysis. The 5' end of Surf-6 lies within the CpG-rich island about 8 kilobases (kb) from the CpG-rich island containing the 5' end of Surf-3, and Surf-5 lies between Surf-3 and Surf-6. Thus, the cluster contains a unique arrangement of four CpG-rich islands within 32 kb associated with the 5' ends of the six surfeit genes. The neighboring CpG-rich islands have been located 500 and 100 kb distant on either side of the surfeit cluster, indicating that the end of the cluster of islands has been reached.

Sequences upstream of AAUAAA influence poly(A) site selection in a complex transcription unit

The adenovirus major late transcription unit (MLTU) encodes five colinear mRNA families, L1 through L5, each distinguished by a unique poly(A) site. Site selection is regulated during the course of infection, predominating early at the L1 site and late at the L2 through L5 sites. Two general mechanisms can be invoked to explain predominant usage of the L1 site early in infection. MLTU site selection may proceed in a first-come, first-serve manner whereby the L1 site is used most frequently because it is closest to the promoter. Alternatively, specific sequences flanking the L1 site may control predominant L1 site usage in a position-independent manner. To distinguish between these mechanisms, we constructed deletions in the L1 flanking sequences and inserted the mutated sites into either simple transcription units or mini-MLTUs encoding two poly(A) sites. The pattern of site selection for each construct was then quantitated by S1 nuclease analysis after transfection into 293 cells. The results indicated that L1 sequences upstream of AAUAAA define a novel selector element that can cause predominant L1 site usage at either position of a tandem transcription unit. The element did not significantly affect the stability or nucleocytoplasmic transport of L1 transcripts and was not required for efficient 3'-end processing in simple transcription units. Predominant L1 site usage required physical linkage of the processing signals and was independent of the major late promoter.

A putative transmembrane protein with histidine-rich charge clusters encoded in the H-2K/tw5 region of mice

The H-2 complex of mice contains many genes in addition to the gene families involved in immune reactions. Some of them are believed to function in mouse development, as suggested by the findings that several embryonic lethal mutations map within or near the H-2 complex. We have analyzed the H-2K/tw5 region in an attempt to study non-H-2 genes encoded in this region. Overlapping cosmid clones spanning about 170 kilobase pairs of DNA, including the H-2K/tw5 region of the mouse, have been screened for genes expressed in embryonic carcinoma cells. A transcript of 2.8 kilobase pairs (K. Abe. J.-F. Wei, F.-S. Wei, Y.-C. Hsu, H. Uehara, K. Artzt, and D. Bennett, EMBO J. 7:3441-3449, 1988) encoded by the KE 4 gene flanking H-2K distally was identified. The transcript was abundantly expressed in embryonic carcinoma cells but was present at low levels in other tissues in adults. A cDNA for this transcript was isolated from the F9 embryonic carcinoma cell line and sequenced. It potentially encodes a protein of 436 amino acids with several interesting features. First, it contains two regions made of well-conserved repeats unusually rich in histidine residues. In the repeats, histidine alternates with other amino acids, notably glycine or serine. Second, the two histidine-rich regions are separated by three putative membrane-spanning domains. Third, the N-terminal part of the sequence shows characteristics of a signal peptide. The results indicate that the protein coded by the gene may be a transmembrane protein with histidine-rich charge clusters. A similar sequence motif found in other known genes allows speculation on the possible functional of this gene.

The mouse surfeit locus contains a cluster of six genes associated with four CpG-rich islands in 32 kilobases of genomic DNA

The clustered arrangement (no two adjacent genes are separated by more than 73 base pairs [bp] and two genes overlap by 133 bp at their 3' ends) of the four genes (Surf-1 to -4) identified so far in the mouse surfeit locus (T. Williams, J. Yon, C. Huxley, and M. Fried, Proc. Natl. Acad. Sci. USA 85:3527-3530, 1988) is the tightest gene clustering found in any mammalian genome to date and strongly suggests the possibility of cis-interaction and/or coregulation of gene expression. Thus, we are analyzing the surfeit genes in detail and are defining the extent of the cluster. Here we present the sequence of the entire Surf-4 gene and define the 3' and 5' extents of its mRNAs. The Surf-4 gene has heterogeneous transcriptional start sites, and its 5' end lies in a CpG-rich island. The gene specifies three mRNAs, with the two most abundant mRNAs differing in the locations of their 3' polyadenylation sites. Only the most abundant Surf-4 mRNA would overlap the 3' end of the Surf-2 gene by 133 bp. Two new genes (Surf-5 and Surf-6) have been identified in the surfeit gene cluster by Northern (RNA) blot analysis. The 5' end of Surf-6 lies within the CpG-rich island about 8 kilobases (kb) from the CpG-rich island containing the 5' end of Surf-3, and Surf-5 lies between Surf-3 and Surf-6. Thus, the cluster contains a unique arrangement of four CpG-rich islands within 32 kb associated with the 5' ends of the six surfeit genes. The neighboring CpG-rich islands have been located 500 and 100 kb distant on either side of the surfeit cluster, indicating that the end of the cluster of islands has been reached.

Single gene encodes glycophospholipid-anchored and asymmetric acetylcholinesterase forms: alternative coding exons contain inverted repeat sequences

Polymorphic forms of acetylcholinesterase are tethered extracellularly either as dimers membrane-anchored by a glycophospholipid or as catalytic subunits disulfidelinked to a collagen tail that associates with the basal lamina. Genomic clones of acetylcholinesterase from T. californica revealed that individual enzyme forms are encoded within a single gene that yields multiple mRNAs. Each enzyme form is encoded in three exons: the first two exons, bases -22 to 1502 and 1503 to 1669, encode sequence common to both forms, while alternative third exons encode a hydrophobic C-terminal region, to which a glycophospholipid is added upon processing, and a nonprocessed C-terminus, yielding a catalytic subunit that disulfide-links with a collagen-like structural unit. The 3' untranslated region of each alternative exon contains tandem repeat sequences that are inverted with respect to the other exon. This may either dictate alternative exon usage by formation of cis stem-loops or affect the abundance of translatable mRNA by trans-hybridization between the alternative spliced mRNA species.

The enzyme that adds poly(A) to mRNAs is a classical poly(A) polymerase

Virtually all mRNAs in eucaryotes end in a poly(A) tail. This tail is added posttranscriptionally. In this report, we demonstrate that the enzyme that catalyzes this modification is identical with an activity first identified 30 years ago, the function of which was previously unknown. This enzyme, poly(A) polymerase, lacks any intrinsic specificity for its mRNA substrate but gains specificity by interacting with distinct molecules: a poly(A) polymerase from calf thymus, when combined with specificity factor(s) from cultured human cells, specifically and efficiently polyadenylates only appropriate mRNA substrates. Our results thus demonstrate that this polymerase is responsible for the addition of poly(A) to mRNAs and that its interaction with specificity factors is conserved.

Genetic transfer of a functional human interferon alpha receptor into mouse cells: cloning and expression of its cDNA

A cDNA coding for the human interferon alpha receptor has been cloned using a gene transfer approach. This consists of transferring human DNA to mouse cells and selecting for cells sensitive to human interferon alpha. The transfected cells expressed the human interferon alpha receptor, and a 5 kb human DNA was isolated from a secondary transfectant. This DNA defects an mRNA present in human cells and was used to clone a 2.7 kb cDNA from a library constructed from human Daudi cells. The sequence of the cDNA is presented. It codes for a glycoprotein of 557 amino acids with an N-terminal hydrophobic region and a single transmembrane-spanning segment. Mouse cells expressing the cDNA become sensitive to the antiviral activity of and express binding sites for human interferon alpha, demonstrating that the cloned cDNA encodes a functional human interferon alpha receptor.

Expression of REX-1, a gene containing zinc finger motifs, is rapidly reduced by retinoic acid in F9 teratocarcinoma cells

In the presence of retinoic acid (RA), cultured F9 murine teratocarcinoma stem cells differentiate into nontumorigenic cells resembling the extraembryonic endoderm of the early mouse embryo. By differential hybridization screening of an F9 cell cDNA library, we isolated a 1,745-nucleotide cDNA for a gene, REX-1 (for reduced expression), whose steady-state mRNA level began to decline in F9 cells in monolayer culture within 12 h after the addition of RA. By 48 to 96 h after RA treatment of F9 cells in monolayer culture, the REX-1 steady-state mRNA level was more than sevenfold lower than the level in undifferentiated F9 stem cells. The REX-1 mRNA decrease did not result from the reduction in cell growth rate associated with the differentiation process, since the REX-1 mRNA level did not decline in F9 cells that were partially growth arrested after 48 h of isoleucine deprivation. The RA-associated REX-1 mRNA decrease resulted primarily from a reduction in the transcription rate of the REX-1 gene in the presence of RA. In contrast to results in F9 cells, we have been unable thus far to detect REX-1 mRNA in day 7.5 to 12.5 mouse embryo RNA samples or in the P19 teratocarcinoma stem cell line. The putative REX-1 protein identified by DNA sequence analysis contains four repeats of the zinc finger nucleic acid-binding motif and a potential acidic activator domain, suggesting that REX-1 encodes a regulatory protein. The REX-1 gene is not identical to the previously reported murine genes that encode zinc finger-containing proteins.

Murine beta 1,4-galactosyltransferase: both the amounts and structure of the mRNA are regulated during spermatogenesis

Previously we have shown that the gene encoding murine beta 1,4-galactosyltransferase (beta 1,4-GT; UDPgalactose:N-acetyl-D-glucosaminyl-glycopeptide 4-beta-D-galactosyltransferase, EC 2.4.1.38) is unusual in that it specifies two sets of mRNAs of about 3.9 and 4.1 kilobases (kb). Translation of the 3.9- and 4.1-kb mRNAs results in the predicted synthesis of two related membrane-bound forms of the protein of 386 amino acids (short form) and 399 amino acids (long form), respectively. In this study we have examined the expression of beta 1,4-GT during murine spermatogenesis. Spermatogonia contain a 4.1-kb transcript that is comparable in size to the beta 1,4-GT mRNA identified in somatic cells. During differentiation from spermatogonia (2n) to pachytene spermatocytes (4n), the amount of beta 1,4-GT mRNA is reduced to barely detectable levels. Continued differentiation to round spermatids (n) is coincident with a renewed production of beta 1,4-GT mRNA to levels comparable with those detected in spermatogonia. However, the characteristic 4.1-kb mRNA detected in spermatogonia is replaced by two truncated transcripts of 2.9 and 3.1 kb. By S1 nuclease analysis, the 2.9- and 3.1-kb transcripts were shown to encode the same open reading frame as the 4.1-kb transcript found in somatic cells. The shorter round spermatid transcripts arise as a consequence of the use of alternative poly(A) signals. Lastly, we show that, in direct contrast to all somatic tissues and cell lines examined to date, male germ cells synthesize only the long form of the beta 1,4-GT polypeptide.

A putative transmembrane protein with histidine-rich charge clusters encoded in the H-2K/tw5 region of mice

The H-2 complex of mice contains many genes in addition to the gene families involved in immune reactions. Some of them are believed to function in mouse development, as suggested by the findings that several embryonic lethal mutations map within or near the H-2 complex. We have analyzed the H-2K/tw5 region in an attempt to study non-H-2 genes encoded in this region. Overlapping cosmid clones spanning about 170 kilobase pairs of DNA, including the H-2K/tw5 region of the mouse, have been screened for genes expressed in embryonic carcinoma cells. A transcript of 2.8 kilobase pairs (K. Abe. J.-F. Wei, F.-S. Wei, Y.-C. Hsu, H. Uehara, K. Artzt, and D. Bennett, EMBO J. 7:3441-3449, 1988) encoded by the KE 4 gene flanking H-2K distally was identified. The transcript was abundantly expressed in embryonic carcinoma cells but was present at low levels in other tissues in adults. A cDNA for this transcript was isolated from the F9 embryonic carcinoma cell line and sequenced. It potentially encodes a protein of 436 amino acids with several interesting features. First, it contains two regions made of well-conserved repeats unusually rich in histidine residues. In the repeats, histidine alternates with other amino acids, notably glycine or serine. Second, the two histidine-rich regions are separated by three putative membrane-spanning domains. Third, the N-terminal part of the sequence shows characteristics of a signal peptide. The results indicate that the protein coded by the gene may be a transmembrane protein with histidine-rich charge clusters. A similar sequence motif found in other known genes allows speculation on the possible functional of this gene.

Cloning and analysis of the tomato nitrate reductase-encoding gene: protein domain structure and amino acid homologies in higher plants

We have cloned and sequenced the nitrate reductase (NR)-encoding gene (nia) from tomato. When compared to the two Nicotiana tabacum nia structural genes, this 5-kb tomato gene shows a highly conserved structure, the coding sequence being interspersed with three introns at the same positions. Nucleotide sequences of the 5' promoter regions are not homologous, except for a 250-bp fragment. This small region might be involved in the similar regulation of the nia expression in tomato and tobacco plant species. The tomato gene codes for a 911 amino acid (aa) polypeptide chain. This sequence was aligned with and compared to other higher plant NR sequences. This alignment clearly identifies the three catalytic domains of NR, namely, a molybdopterin cofactor-binding domain, a heme domain and a FAD/NADH domain. On the other hand, it suggests that the less conserved 80-aa N-terminal region, containing a striking acidic aa cluster, is an additional domain bearing regulatory or structural function.

The nucleotide sequence of a mouse renin-encoding gene, Ren-1d, and its upstream region

The renin-encoding genes have been cloned from high (Ren-1d, Ren-2d)- and low (Ren-1c)-renin-producing strains of mice (DBA/2J and C57BL/10). Each of the genes is approx. 9.6 kb in length and consists of nine exons and eight introns. The entire nucleotide sequence of the Ren-1d gene has been determined and the 5'-flanking regions of the three genes, Ren-1c, Ren-1d and Ren-2d, have been compared. The significance of several potential regulatory signals found in the DNA is discussed.

Expression of REX-1, a gene containing zinc finger motifs, is rapidly reduced by retinoic acid in F9 teratocarcinoma cells

In the presence of retinoic acid (RA), cultured F9 murine teratocarcinoma stem cells differentiate into nontumorigenic cells resembling the extraembryonic endoderm of the early mouse embryo. By differential hybridization screening of an F9 cell cDNA library, we isolated a 1,745-nucleotide cDNA for a gene, REX-1 (for reduced expression), whose steady-state mRNA level began to decline in F9 cells in monolayer culture within 12 h after the addition of RA. By 48 to 96 h after RA treatment of F9 cells in monolayer culture, the REX-1 steady-state mRNA level was more than sevenfold lower than the level in undifferentiated F9 stem cells. The REX-1 mRNA decrease did not result from the reduction in cell growth rate associated with the differentiation process, since the REX-1 mRNA level did not decline in F9 cells that were partially growth arrested after 48 h of isoleucine deprivation. The RA-associated REX-1 mRNA decrease resulted primarily from a reduction in the transcription rate of the REX-1 gene in the presence of RA. In contrast to results in F9 cells, we have been unable thus far to detect REX-1 mRNA in day 7.5 to 12.5 mouse embryo RNA samples or in the P19 teratocarcinoma stem cell line. The putative REX-1 protein identified by DNA sequence analysis contains four repeats of the zinc finger nucleic acid-binding motif and a potential acidic activator domain, suggesting that REX-1 encodes a regulatory protein. The REX-1 gene is not identical to the previously reported murine genes that encode zinc finger-containing proteins.

Partial deduced sequence of the 110-kD-calmodulin complex of the avian intestinal microvillus shows that this mechanoenzyme is a member of the myosin I family

The actin bundle within each microvillus of the intestinal brush border is laterally tethered to the membrane by bridges composed of the protein complex, 110-kD-calmodulin. Previous studies have shown that avian 110-kD-calmodulin shares many properties with myosins including mechanochemical activity. In the present study, a cDNA molecule encoding 1,000 amino acids of the 110-kD protein has been sequenced, providing direct evidence that this protein is a vertebrate homologue of the tail-less, single-headed myosin I first described in amoeboid cells. The primary structure of the 110-kD protein (or brush border myosin I heavy chain) consists of two domains, an amino-terminal "head" domain and a 35-kD carboxy-terminal "tail" domain. The head domain is homologous to the S1 domain of other known myosins, with highest homology observed between that of Acanthamoeba myosin IB and the S1 domain of the protein encoded by bovine myosin I heavy chain gene (MIHC; Hoshimaru, M., and S. Nakanishi. 1987. J. Biol. Chem. 262:14625-14632). The carboxy-terminal domain shows no significant homology with any other known myosins except that of the bovine MIHC. This demonstrates that the bovine MIHC gene most probably encodes the heavy chain of bovine brush border myosin I (BBMI). A bacterially expressed fusion protein encoded by the brush border 110-kD cDNA binds calmodulin. Proteolytic removal of the carboxy-terminal domain of the fusion protein results in loss of calmodulin binding activity, a result consistent with previous studies on the domain structure of the 110-kD protein. No hydrophobic sequence is present in the molecule indicating that chicken BBMI heavy chain is probably not an integral membrane protein. Northern blot analysis of various chicken tissue indicates that BBMI heavy chain is preferentially expressed in the intestine.

Sequences upstream of AAUAAA influence poly(A) site selection in a complex transcription unit

The adenovirus major late transcription unit (MLTU) encodes five colinear mRNA families, L1 through L5, each distinguished by a unique poly(A) site. Site selection is regulated during the course of infection, predominating early at the L1 site and late at the L2 through L5 sites. Two general mechanisms can be invoked to explain predominant usage of the L1 site early in infection. MLTU site selection may proceed in a first-come, first-serve manner whereby the L1 site is used most frequently because it is closest to the promoter. Alternatively, specific sequences flanking the L1 site may control predominant L1 site usage in a position-independent manner. To distinguish between these mechanisms, we constructed deletions in the L1 flanking sequences and inserted the mutated sites into either simple transcription units or mini-MLTUs encoding two poly(A) sites. The pattern of site selection for each construct was then quantitated by S1 nuclease analysis after transfection into 293 cells. The results indicated that L1 sequences upstream of AAUAAA define a novel selector element that can cause predominant L1 site usage at either position of a tandem transcription unit. The element did not significantly affect the stability or nucleocytoplasmic transport of L1 transcripts and was not required for efficient 3'-end processing in simple transcription units. Predominant L1 site usage required physical linkage of the processing signals and was independent of the major late promoter.

A protein with several possible membrane-spanning domains encoded by the Drosophila segment polarity gene patched

The patterning of cells in insect segments requires the exchange of information between cells, which in Drosophila depends on the activity of members of the segment-polarity class of genes. Here we report the molecular characterization of one such gene, patched. We find that patched encodes a large protein with several possible membrane-spanning domains and is expressed in a complex pattern during embryogenesis.

Multiple polyadenylation sites in a large 3'-most exon of the rat insulin-like growth factor II gene

The rat insulin-like growth factor II (rIGFII) gene produces, in addition to three major mRNA species 3.6 kilobases (kb), 4.6 kb and 3.8 kb in length which represent transcripts from three independent leader-exons, multiple smaller-sized products that distribute broadly in the 1-3 kb region on Northern blots. Structural constituents of these RNAs were analyzed by hybridization with region-specific probes prepared from the entire rIGFII genome. Most of these shorter RNAs contained both 5'-untranslated and coding regions, but only parts of the 3'-untranslated region. At least nine protected sites were mapped within a single 3'-most exon E6 by S1 nuclease analysis. Some but not all of these sites were associated with the upstream polyadenylation signal, AATAAA, or its variants. Since none of the shorter subspecies contained intronic sequences, aberration in splicing is not involved in their generation. Thus, the main parts of submature materials are a collection of discrete species of RNAs, most, if not all, of which are produced by alternative polyadenylation site selection.

Alterations in the pre-mRNA topology of the bovine growth hormone polyadenylation region decrease poly(A) site efficiency

RNase mapping experiments show that the bovine growth hormone (bGH) poly(A) region forms an extensive hairpin loop. Mutants were prepared to change poly(A) region pre-mRNA structure and cleavage site efficiency without altering necessary sequences. An inverted repeat which includes the poly(A) cleavage site was created by insertion of a linker upstream of the poly(A) region to compete with any wild-type secondary structure. RNA mapping analyses show alterations in the nuclease accessibility of this mutant at the natural site of cleavage. This mutant shows a 75% drop in relative reporter gene expression at the steady-state protein and RNA levels. When the linker is inserted as a direct repeat, expression is equivalent to wildtype levels. To show that transcription was not terminated by the inverted repeat, the SV40 late poly(A) region was inserted downstream. These mutants show restored expression and processing at the downstream site. Our experiments reveal that the conformation of the poly(A) site pre-mRNA is important in mediating efficient cleavage-polyadenylation.

The poly(A) binding protein is required for poly(A) shortening and 60S ribosomal subunit-dependent translation initiation

Depletion of the essential poly(A) binding protein (PAB) in S. cerevisiae by promoter inactivation or by the utilization of a temperature-sensitive mutation (pab1-F364L) results in the inhibition of translation initiation and poly(A) tail shortening. Reversion analysis of pab1-F364L yielded seven independent, extragenic cold-sensitive mutations (spb1-spb7) that also suppress a PAB1 deletion. These mutations allow translation initiation without significantly changing poly(A) tail lengths in the absence of PAB, and they affect the amount of 60S ribosomal subunit. Consistent with this, SPB2 encodes the ribosomal protein L46. These data suggest that the 60S subunit mediates the PAB requirement of translation initiation, thereby ensuring that only intact poly(A)+ mRNA will be translated efficiently in vivo.

Sequence and polyadenylation site determination of the murine immunoglobulin gamma 2a membrane 3' untranslated region

We have determined the nucleotide sequence of the murine immunoglobulin gamma 2a membrane 3' untranslated region (1413 nucleotides) and approximately 679 nucleotides of downstream sequence. Two AATAAA hexanucleotide sequences are present in the 2092 nucleotide interval. The first one functions as the major polyA signal, directing cleavage and polyadenylation at a site 20 nucleotides downstream. Within 41 nucleotides downstream of the major membrane polyA signal are two sequences with 75% homology to the consensus sequence, (C/T)GTGTT(C/T)(C/T), identified by McLauchlan et al. [Nucl. Acids Res. 13, 1347-1365 (1985)]. An 80% homology match to the Berget consensus sequence, CA(C/T)TG, begins five nucleotides 3' of the major polyA site (used 20 times more than the second, downstream polyA site) [Berget Nature 309, 179-182 (1984)]. The second AATAAA, located 73 nucleotides 3' of the first, directs cleavage and polyadenylation 18 nucleotides downstream at a minor polyA site. One match with 75% homology to the McLauchlan consensus sequence begins 17 nucleotides 3' of the second (weaker) polyA site. No matches to the Berget consensus sequence are located near this second, weaker polyA site.