Search

The transcriptional elongation factor TFIIS causes stimulation of RNA polymerase II elongation and readthrough of some of the elongation blocks. We present cloning and sequence characterization of the human TFIIS gene and a pseudogene. The intron-less organization of both of these genes indicates that previously identified cDNAs which suggested the presence of an intron were the products of cloning artifacts. The gene is organized in an uninterrupted ORF which codes for 301 amino acids, whereas the pseudogene lacks an ORF able to code for a full-length protein. The potential promoter for the gene has two putative GC-box-type consensus sequences, two CCAAT-box consensus sequences, and is bounded by a human Alu sequence. Two potential transcriptional termination signal sequences downstream from the consensus polyadenylation signal are proposed.

Gastrin is expressed in the gastric antrum and in fetal pancreatic islets but not in adult islets. We have now identified the hepatobiliary tract as another, previously unknown, potential site of gastrin gene expression. Two human gastrin simian virus 40 large tumor antigen (SV40 T antigen) fusion genes containing 1.5 kb of 5' flanking sequence and 10.5 kb that included 5.5 kb upstream, 1.5 kb downstream, and the entire transcribed region were used to generate transgenic mice. Analysis of several transgenic lines, derived from both fusion genes, revealed development of transmissible hepatobiliary tract tumors and pancreatic islet cell tumors. Analysis of each of the tumor cells demonstrates expression of SV40 T antigen but no expression of gastrin. Of the two fusion genes, only the 10.5-kb sequence induces hyperplasia of gastrin-producing cells in the antrum. Analysis of these cells demonstrates expression of SV40 T antigen and gastrin, suggesting that the 10.5-kb sequence is sufficient for gastrin cell hyperplasia in the antrum. These data raise the possibility that gastrin is transiently expressed in the hepatobiliary tract.

The cDNA for the human elongation factor, TFIIS, has been cloned and expressed in E. coil with the T7 expression system. This 280-amino acid TFIIS protein is shorter by 21 residues than that of the mouse. The missing 21 residues are located in the amino-terminal region, which is not thought to be required for transcriptional stimulation. Apart from this gap, human and mouse proteins reveal 96% overall identity and 98.5% sequence similarity If conservative substitutions are taken into account. The bacterially expressed human protein and the purified calf thymus proteins are indistinguishable in their ability to stimulate transcript elongation by purified RNA polymerase II. Estimation of the native molecular size of the human protein in solution indicates that It exists as a dimer. Accepted version

The crystal structure of d[Cp(CH3)G] has been determined as part of a project to study the mechanism of the B----Z transition in DNA. The asymmetric unit contains two dinucleotides and the equivalent of 7.5 water molecules, partially disordered over 12 definable positions. The two symmetry-independent dinucleotides form a duplex with Watson-Crick base-pairing and a right-handed helical sense. Comparison with previously determined structures of the B and A conformation showed that this duplex is closer to B than to A but significantly different from B. It corresponds to a stretched out helix with a 4 A rise per base pair and a helical twist of 32 degrees. This structure may serve as a model for the bending of DNA in certain situations. The configuration at the methyl phosphonate is RP, and a mechanism, based on this assignment, is presented for the B----Z transition in DNA.

Transcriptional elongation involves dynamic interactions among RNA polymerase and single-stranded and double stranded nucleic acids in the ternary complex1–4. In prokaryotes its regulation pro-vides an important mechanism of genetic control1. Analogous eukaryotic mechanisms are not well understood5, but may control expression of proto-oncogenes6,7 and viruses, including the human immunodeficiency virus HIV-1 (ref. 8). The highly conserved euk-aryotic transcriptional elongation factor TFIIS9 enables RNA polymerase II (RNAPII) to read though pause or termination sites, nucleosomes and sequence-specific DNA-binding proteins10–14. Two distinct domains of human TFIIS, which bind RNAPII and nucleic acids, regulate read-through10 and possibly nascent transcript cleavage11–15. Here we describe the three-dimensional NMR16 structure of a Cys4 nucleic-acid-binding domain from human TFIIS9,10. Unlike previously characterized zinc modules17–21, which contain an α-helix, this structure consists of a three-stranded β-sheet. Analogous Cys4 structural motifs may occur in other proteins involved in DNA or RNA trans-actions22–24, including RNAPII itself25. This new structure, desig-nated the Zn ribbon, extends the repertoire of Zn-mediated peptide architectures26 and highlights the growing recognition of the β-sheet as a motif of nucleic-acid recognition27,28. Accepted version

The transcriptional factor TFIIS helps overcome elongation barriers and enhances proofreading by RNA polymerase II. These TFIIS functions may be modulated by the TFIIS zinc ribbon domain through interactions with nucleic acids in the elongation complex. Within this zinc ribbon domain, the dipeptide sequences Asp261-Glu262 and Arg276-Trp277 have been shown to be critical for its function by mutant analysis. The sequence Asp261-Glu262 has been suggested to participate in metal binding within the RNA polymerase II active site. We now show that the sequence Arg276-Trp277 interacts with nucleic acids through a combination of electrostatic and stacking interactions. The interaction of the indole side chain of the tryptophan residue with nucleic acid bases is demonstrated by a characteristic and reversible decrease in the zinc ribbon fluorescence intensity as a function of oligonucleotide concentration. These interactions are salt sensitive (maximum interaction at 200 mM and no interaction at 500 mM NaCl), suggesting that the tryptophan stacking with nucleic acid base accompanies electrostatic contacts. The oligonucleotide-zinc ribbon interactions exhibit small but significant base preferences, as shown by the dependence of Keq on base composition, with decreasing Keq in the order U > T > A > C >> G. Within the variety of homopolymeric single- and double-stranded deoxy- and ribooligonucleotides, the oligonucleotide rU12-18.dA20 exhibited a 2-6-fold binding preference relative to other oligonucleotides. This preferential binding of the zinc ribbon to sequences composed of rU.dA base pairs, which are generally associated with elongation blocks, may help in overcoming elongation barriers. Since the mRNA proofreading and enhancement of elongation involve cleavage of ribonucleotide of the mismatched pair and the weakly paired rU.dA nucleotides, but not the stably paired rC.dG nucleotides, we propose that the Arg276-Trp277 sequence in the TFIIS zinc ribbon may serve as a scanner connected to the transcript cleavage apparatus for weakly paired or mismatched nucleotides by employing indole ring stacking with the bases as a criterion of determining their subsequent removal. The striking similarity in preference for mismatched and weakly paired nucleotides for binding and for excision suggests a functional relationship between binding and cleavage reactions.

Fidelity of DNA and protein synthesis is regulated by a proofreading mechanism but function of a similar mechanism during RNA synthesis has not been demonstrated. Analysis of transcriptional fidelity and its control has been hampered by the necessity to employ complex DNA templates requiring either a promoter and initiation factors or 3′-extended templates. To circumvent this difficulty, we have created an RNA–DNA dumbbell template that can be recognized as a template-primer and extended by RNA polymerase II. By employing this system, we demonstrate that RNA polymerase II can misincorporate a nucleotide and carry out template-dependent elongation at the mispaired end. The transcripts containing misincorporated residues can be cleaved by the very slow 3′ → 5′ ribonuclease activity of the RNA polymerase II, but enhancement of this activity by the elongation factor TFIIS generates RNA with a high degree of fidelity. This enhanced preferential cleavage of misincorporated transcripts suggests an important role for TFIIS in maintaining transcriptional fidelity.

The eukaryotic transcription factor TFIIS enhances elongation and nascent transcript cleavage activities of RNA polymerase II in a stalled elongation complex. By site-directed mutagenesis, we have demonstrated that invariant residues Asp-261 and Glu-262 of the nucleic acid-binding TFIIS Zn ribbon are critical for stimulation of both elongation and RNA cleavage activities of RNA polymerase II. Substitution of either of these residues inactivates both TFIIS functions, suggesting a related role in both activities. These acidic residues may participate in phosphoryl transfer reactions by a two-metal-ion mechanism in a manner analogous to Klenow fragment. The RNA polymerase II itself may contain a Zn ribbon, in as much as the polymerase's 15-kDa subunit contains a sequence that aligns well with the TFIIS Zn ribbon sequence, including a similarly placed pair of acidic residues.

We describe an in vitro system that emulates the specific and efficient transcriptional termination associated with the human gastrin gene terminator in vivo. The system involves a dC-tailed DNA template containing the gastrin gene terminator sequence, purified RNA polymerase II, and purified elongation factor TFIIS. In this system, the basal level of termination by RNA polymerase II at the gastrin gene terminator is specifically enhanced by netropsin, an (A + T)-rich minor groove-binding peptide. This enhanced termination is maintained even with TFIIS, which normally suppresses termination at this site. In vitro termination is terminator sequence-specific. Mutant sequences that reduce or abolish termination in vivo show corresponding reductions in activity in the in vitro system. This in vitro emulation of in vivo activities of wild-type and mutant terminators strongly suggests that netropsin and a putative termination factor may share some aspects of their biochemical mechanisms. The general applicability of this system to the study of RNA polymerase II elongation and termination is suggested by the enhancement of termination seen at both the gastrin and human histone H3.3 gene terminators.

The eukaryotic transcriptional factor TFIIS enhances transcript elongation by RNA polymerase II. Here we describe two functional domains in the 280 amino acid human TFIIS protein: residues within positions 100-230 are required for binding to polymerase, and residues 230-280, which form a zinc finger, are required in conjunction with the polymerase binding region for transcriptional stimulation. Interestingly, a mutant TFIIS with only the polymerase binding domain actually inhibits transcription, whereas a mutant in which the polymerase binding and zinc finger domains are separated by an octapeptide is only weakly active. The zinc finger itself has no effect on transcription, but in contrast to the wild-type protein, it binds to oligonucleotides. These findings suggest that TFIIS may interact with RNA polymerase II such that the normally masked zinc finger can specifically contact nucleotides in the transcription elongation zone at a position juxtaposed to the polymerization site.

We located and characterized a downstream transcriptional regulatory element in the human gastrin gene by transferring the gastrin gene 3' fragment, from which the polyadenylation signal sequence was deleted, into the shuttle vector pSCAT10 at a site located immediately downstream from the chloramphenicol acetyltransferase (CAT) gene and upstream from the simian virus 40 polyadenylation region. Study of CAT RNA derived from the hybrid plasmids, indicated regulation of transcription on the gastrin gene fragment. Analysis of deletion mutants generated from the 5' region of the fragment by CAT assay and by S1 nuclease mapping of mRNAs indicated the possible involvement of an oligothymidylate-rich sequence in transcription regulation. Mapping of gastrin gene RNA 3' ends to the 5' side proximal to the oligothymidylate-rich sequence clearly demonstrated that this sequence is a transcriptional terminator element. This unique sequence, interspersed with one or two adenines, which also functions in an orientation-dependent manner, is located 192 nucleotides downstream from the gastrin gene polyadenylation site, and serves as a transcriptional termination signal.

We have isolated a human gastrin gene from a genomic library by employing a human gastrin cDNA clone as a hybridization probe. The total length of the gene is approximately 4.0 kilobase pairs, and the gene is separated into three exons and two introns. A 130-base-pair intron interrupts the coding region and a 3.0-kilobase-pair intron is located in the 5' untranslated region. Nucleotide sequence analysis showed that all of the exon-intron boundaries follow the A-G/G-T consensus sequences. A putative transcription initiation site is assigned to the adenine 60 nucleotides upstream from the exon-intron junction on the basis of S1 nuclease protection mapping. A possible "TATA" equivalent sequence T-T-A-T-A-A is located 28 base pairs upstream from the transcription initiation site. A "CAT box" sequence, C-A-T-T, is located 99 nucleotides upstream of the transcription initiation site. A poly(A)-addition signal, A-A-U-A-A-A, is located 80 base pairs downstream from the termination codon. Comparison of the nucleotide sequences of the human cDNA and the genomic clone revealed that the aspartic acid codon at position 71 of preprogastrin is interrupted by the small intron (130 base pairs). The 3' region of the large intron contains a sequence of 300 nucleotides that is flanked by 15-nucleotide direct repeats. This sequence exhibits a striking homology to the human Alu-type sequence.

Chemical methods for the synthesis of short deoxyribooligonucleotides containing methyl and phenylphosphonodiester linkages have been developed. The interaction of two such nonionic dinucleotide analogs, T(pCH3)T and T(pC6H5)T, with several enzymes has been investigated. Because of the phosphonate linkage each dinucleotide exists as a diastereomeric pair as shown by thin layer chromatography and enzymatic studies. Both isomers of each dinucleotide can be phosphorylated by T4-polynucleotide kinase in the presence of [gamma-32P]ATP. Only one of the diastereoisomers of each dinucleotide is slowly hydrolyzed by snake venom phosphodiesterase and acts as an inhibitor of the enzyme-catalyzed hydrolysis of 5'-labeled oligothymidylic acid. Both isomers of each dinucleotide analog are completely resistant to hydrolysis by spleen phosphodiesterase.

To determine which functional groups of bases within the grooves of double-helical DNA interact with the HpaI endonuclease, we have employed chemically synthesized octanucleotides containing base analogues. The 5-methyl group of thymine was probed as a contact between the HpaI endonuclease and its recognition sequence by using the oligonucleotides d(G-G-T-T-A-A-C-C), d(G-G-T-U-A-A-C-C), and d(G-G-T-U(Br)-A-A-C-C). The 2-amino group of guanine was probed as a contact for the HpaI endonuclease by using the octanucleotide d(G-I-T-T-A-A-C-C). The HpaI endonuclease cleaves octanucleotides d(G-G-T-T-A-A-C-C) and d(G-G-T-B-A-A-C-C) according to Michaelis-Menten kinetics. However, both the Km and turnover number for d(G-G-T-B-A-A-C-C) were severalfold lower than those for cleavage of d(G-G-T-T-A-A-C-C). In addition, d(G-G-T-U-A-A-C-C) was not cleaved by HpaI endonuclease, suggesting that the 5-methyl group of thymine is a contact between the HpaI endonuclease and its recognition sequence. d(G-I-T-T-A-A-C-C) was not cleaved by the HpaI endonuclease which may be due in part to the low thermal stability of the duplex. Nevertheless, our results suggest that the 2-amino group of guanine is a contact for the HpaI endonuclease. A phosphate group 5' external to the HpaI recognition sequence has been identified as a contact between the HpaI endonuclease and DNA. The HpaI endonuclease cleaved 5'-phosphorylated octanucleotide 30-fold faster than unphosphorylated octanucleotide. In addition, the Km of the d(G-G-T-T-A-A-C-C) was 8000-fold higher than the Km of the phage f1 RFI DNA, suggesting that the octanucleotide is too short to take advantage of the entire DNA binding site of the enzyme.

By employing S1 nuclease mapping of total RNA isolated from porcine cerebral cortex, cerebellum, hypothalamus, pituitary, kidney, liver, pancreas, intestine, and antral mucosa, we have investigated gastrin gene expression in these tissues. Our results show that a gastrin gene is expressed only in the antral mucosal and pituitary tissues. Based on the amount of gastrin specific probe protected from S1 nuclease digestion in the presence of a given weight of total RNA, the amount of gastrin mRNA present in pituitary is approximately 330 times lower than in antral mucosa. These findings help establish the tissue distribution of gastrin gene expression.

The synthesis of a self-complementary octanucleotide, d(G-G-T-T-A-A-C-C-), using a modified triester approach is described. The protected dinucleotides, d(Me2O)TribG(C1C6H4) ibG, d(Me2O)TrT(ClC6H4)T, d(Me2O)TrbzA(ClC6H4)bzA, and d(Me2O)TranC(ClC6H4)anC were synthesized by a one step triester procedure. After removal of the trityl group, the dinucleotides, dT(ClC6H4)T and danC (ClC6H4)anC were coupled to d(Me2O)TribG(ClC6H4)ibG and d(Me2O)TrbzA(ClC6H4)bzA, respectively to yield the respective tetranucleotides. The tetranucleotide, d(Me2O)TrbzA(ClC6H4)bzA(ClC6H4) and (ClC6H4)anC was detritylated and then coupled with d(Me2O)TribG(ClC6H4)ibG(ClC6H4)T(Cl6H4)T to give octanucleotide. The fully protected octanucleotide was deblocked by treatment with aqueous NH3 followed by acid and was characterized by nucleotide sequence analysis.

The purification and catalytic properties of the homogeneous Hpa I methylase is described. The enzyme exists as a single polypeptide chain with a molecular weight of 37,000 +/- 2,000 was shown by sedimentation equilibrium and polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. The Hpa I methylase transfers methyl groups of S-adenosylmethionine to adenine present in the recognition sequence d(G-T-T-A-A*-C), A* is the N6 methyl adenosine. An average of 2.1 methyl groups per recognition site are transferred by the Hpa I methylase.

Three oligodeoxynucleotides have been chemically synthesized and used as hybridization probes to detect gastrin-specific mRNA within an heterogeneous population of RNA molecules. Using these oligonucleotides whose nucleotide sequences were deduced from the amino acid sequence of the hormone, 0.2 fmol of gastrin mRNA can be detected/microgram of poly(A)-RNA from hog antrums. The 32P-labeled oligonucleotides hybridize specifically to RNA covalently coupled to DBM-paper (Alwine, J.C., Kemp, D.J., and Stark, G.R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5350-5354). Labeled oligonucleotides also hybridize specifically to RNA separated by gel electrophoresis in the presence of CH3HgOH and covalently coupled to DBM-paper. Using this procedure, we have found that the mRNA coding for gastrin contains about 620 nucleotides, which is in agreement with results obtained using gastrin cDNA to determine mRNA size.

A chiral derivatizing agent, 1-menthyl chloroformate, has been used as a 3′-OH blocking group to facilitate the resolution of diastereomers of methylphosphonate dinucleotides by silica gel column chromatography.

Two proteins exhibiting Hpa II methylase activity have been purified to homogeneity from Haemophilus parainfluenzae and their physical and catalytic properties have been studied. Separation of the two Hpa II methylase activities was achieved by DEAE-Sephadex A-50 chromatography. In subsequent steps, each methylase was purified separately by chromatography on Sephacryl S-200, phosphocellulose, and hydroxylapatite. The proteins have molecular weights of 38,500 +/- 1,000 (Hpa II) and 41,500 +/- 1,000 (Hpa II') as judged by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Sedimentation equilibrium analyses of the native proteins yield molecular weights of 38,800 +/- 3,000 and 42,200 +/- 3,000 for Hpa II and Hpa II', respectively, indicating that both enzymes are composed of a single subunit. Furthermore, both methylases exhibit identical specificity in the methylation of the nucleotide sequence dC-C-G-G in simian virus 40 (SV40) DNA and in a short synthetic oligonucleotide duplex. Although pH, temperature, and salt optima are the same for both enzymes, homogeneous Hpa II' methylase is more stable than Hpa II methylase. Preliminary peptide mapping indicates that the two enzymes are structurally related, suggesting the possibility that Hpa II' methylase may represent a precursor form of Hpa II methylase.

Recently, we identified a specific DNA sequence from the gastrin gene that regulates RNA polymerase II transcription termination in vivo. In the studies presented here, we examined the processing and termination activity of this sequence in vitro. When present in an in vitro synthesized RNA, this sequence (U9A2U5AU4AU4AU5) does not serve as an RNA processing signal on incubation with HeLa whole-cell extract. However, transcription of template DNA in HeLa whole-cell extract does terminate near the 5' end of this sequence. Nuclease S1 and exonuclease VII mapping of the 3' region of the in vitro synthesized RNAs confirm these results. The termination activity of the sequence A9T2A5TA4TA4TA5 is independent of the distance from the promoter and of the nature of the DNA template (linear vs. circular). The termination activity of the sequence shows a strong orientation dependence. These results strongly suggest that the termination activity of this cis-acting element is modulated by a trans-acting cellular factor. The unique structural feature of this sequence, a 10.5-base-pair inverted repeat, may determine the specificity of interaction of the trans-acting factor with the cis-acting element, resulting in accurate termination of transcription.

Many small biologically active peptides are derived from larger precursor forms which fulfil a variety of roles in the synthesis, segregation and intracellular migration of secretory products. Limited proteolysis may occur at several stages during this process, giving rise to products that are either degraded (e. g. the prepeptides) or discharged coordinately from their cells of origin during exocytosis (e. g. insulin and C-peptide). Molecular defects have recently been found to occur at cleavage sites in proinsulin as well as in other proproteins, and these point mutations may, in some instances, be responsible for familial metabolic disorders. The nature and cell specificity of the proteolytic enzymes involved in the conversion of the various precursor forms remains unresolved. Recent studies in our laboratory have led to the identification of precursors of glucagon and somatostatin in rat islets of Langerhans. Analysis of tryptic maps of these precursors has shown that a trypsin-like enzyme would be sufficient to cleave the C-terminally located somatostatin sequence from its precursor (relative molecular mass 12500), but that both trypsin-like and carboxypeptidase B-like enzymes would be necessary to cleave the internal glucagon sequence from its prohormone (relative molecular mass 18000). Molecular cloning techniques have provided valuable new approaches to analysing the structures of a variety of precursor forms, including those for insulin, gastrin, growth hormone, adrenocorticotrophic hormone and the endorphins, and in the future will undoubtedly shed more light on the structures of their chromosomal genes, the mechanisms regulating their expression, and their evolutionary origins.

RNA ligase, purified extensively from Escherichia coli infected with wild-type or DNA ligase mutants of bacteriophage T4, catalyzes the joining of 5′-phosphoryl terminated DNA to DNA and RNA acceptors. This was shown by the conversion of [5′-32P]deoxyoligomers to a form resistant to phosphatase, the increased chain length of the joined RNA-DNA copolymers, the circularity of the DNA joined product, and the transfer of the 5′-32P label of the donor DNA to the 3′-end of both RNA and DNA acceptors. The novel DNA joining activity is intrinsic to RNA ligase since it co-purifies with RNA joining activity and has the same requirements, inhibitors, and thermolability. These results suggest that RNA ligase should be a useful reagent for the synthesis of defined sequence DNA and RNA-DNA copolymers and raise the possibility of a role for RNA ligase in DNA metabolism.

The interaction of (±)-BPDE ( 1 ) with DNA at neutral pH was investigated by the application of relaxed circular pBR322 DNA, (±)-BPDE causes a rapid positive supercoiling of this DNA followed by a slower spontaneous relaxation. The results indicate that there are two clearly discernable types of chemical interactions between 1 and DNA, a rapid intercalative covalent binding and a slower strand breakage. The implications of these findings are discussed.

We have used a synthetic deoxydecanucleotide to generate an insulin-specific cDNA probe suitable for selecting transformants that contain nearly full-length cDNAs corresponding to the mRNAs coding for rat insulins I and II. Double-stranded cDNA was synthesized from x-ray-induced rat insulinoma poly(A)-RNA, inserted in pBR322 plasmid DNA by the homopolymeric tailing technique, and cloned in Escherichia coli chi 1776. Colony hybridization with oligonucleotide-primed cDNA yielded 16 positive clones of which 7 corresponded to rat insulin I mRNA and 9 to rat insulin II mRNA. Restriction endonuclease maps of representative clones of each group indicated that these contained the complete coding sequences, as was confirmed by nucleotide sequence analysis of the 5' region of the cloned DNA for rat insulin II. Nucleotide sequence analysis also established the amino acid sequence of the prepeptide of rat preproinsulin II. Comparison of the amino acid sequence of the prepeptides of rat preproinsulin I and II shows that three conservative amino acid substitutions have occurred in this region of the molecule.