Comparative analysis of human chromosomal segments bearing nonallelic dispersed tRNAimet genes

Rearrangements of human mitochondrial DNA (mtDNA): new insights into the regulation of mtDNA copy number and gene expression

Mitochondria from patients with Kearns-Sayre syndrome harboring large-scale rearrangements of human mitochondrial DNA (mtDNA; both partial deletions and a partial duplication) were introduced into human cells lacking endogenous mtDNA. Cytoplasmic hybrids containing 100% wild-type mtDNA, 100% mtDNA with partial duplications, and 100% mtDNA with partial deletions were isolated and characterized. The cell lines with 100% deleted mtDNAs exhibited a complete impairment of respiratory chain function and oxidative phosphorylation. In contrast, there were no detectable respiratory chain or protein synthesis defects in the cell lines with 100% duplicated mtDNAs. Unexpectedly, the mass of mtDNA was identical in all cell lines, despite the fact that different lines contained mtDNAs of vastly different sizes and with different numbers of replication origins, suggesting that mtDNA copy number may be regulated by tightly controlled mitochondrial dNTP pools. In addition, quantitation of mtDNA-encoded RNAs and polypeptides in these lines provided evidence that mtDNA gene copy number affects gene expression, which, in turn, is regulated at both the post-transcriptional and translational levels.

Interference between effector RNAs expressed from conventional dual-function anti-HIV retroviral vectors can be circumvented using dual-effector-cassette retroviral vectors

Coexpression of different effector molecules from a single vector (a dual-function vector) may provide enhanced efficacy. Thus far most of the reported anti-HIV dual-function vectors express different effector RNAs as a chimeric molecule. In our study involving retroviral vectors coexpressing a U5 ribozyme and either an anti-tat or anti-rev antisense RNA, chimeric vectors exhibit poor potency in several important functional aspects, including inhibition of HIV replication, protection against cytopathic effects, and suppression of target gene function. Surprisingly, such a poor efficacy of chimeric vector function was not associated with a lower level of effector RNA expression. These results indicate that expression of two effector RNAs as a chimeric molecule can lead to interference, reducing their global biological effects. More importantly, we have demonstrated that such interference can be avoided by coexpressing these effector RNAs as separate molecules through a new dual-function vector, called a dual-effector cassette (Dec) vector, developed in this study. We also define some of the design alterations that might affect the efficacy of the Dec vector and demonstrate that forward-designed Dec vectors are more efficacious than reverse-designed Dec vectors, which express a lower level of effector RNA owing to the instability of the 5' effector cassettes in the provirus. We believe that the principle of Dec vector design may also be applicable for the coexpression of other therapeutic RNA effectors in many gene therapy applications.

Construction of an approximately 700-kb transcript map around the familial Mediterranean fever locus on human chromosome 16p13.3

We used a combination of cDNA selection, exon amplification, and computational prediction from genomic sequence to isolate transcribed sequences from genomic DNA surrounding the familial Mediterranean fever (FMF) locus. Eighty-seven kb of genomic DNA around D16S3370, a marker showing a high degree of linkage disequilibrium with FMF, was sequenced to completion, and the sequence annotated. A transcript map reflecting the minimal number of genes encoded within the approximately 700 kb of genomic DNA surrounding the FMF locus was assembled. This map consists of 27 genes with discreet messages detectable on Northerns, in addition to three olfactory-receptor genes, a cluster of 18 tRNA genes, and two putative transcriptional units that have typical intron-exon splice junctions yet do not detect messages on Northerns. Four of the transcripts are identical to genes described previously, seven have been independently identified by the French FMF Consortium, and the others are novel. Six related zinc-finger genes, a cluster of tRNAs, and three olfactory receptors account for the majority of transcribed sequences isolated from a 315-kb FMF central region (between D16S468/D16S3070 and cosmid 377A12). Interspersed among them are several genes that may be important in inflammation. This transcript map not only has permitted the identification of the FMF gene (MEFV), but also has provided us an opportunity to probe the structural and functional features of this region of chromosome 16.

Identification of an autonomously initiating RNA polymerase III holoenzyme containing a novel factor that is selectively inactivated during protein synthesis inhibition

Transcription by RNA polymerase III (Pol III) requires multiple general initiation factors that, in isolated form, assemble onto the promoter in an ordered fashion. Here, it is shown that all components required for transcription of the VA1 and tRNA genes, including TFIIIB, TFIIIC, and RNA Pol III, can be coimmunopurified from a HeLa cell line that constantly expresses a FLAG epitope-tagged subunit of human RNA Pol III. This finding of an RNA Pol III "holoenzyme" suggests similarities between transcription initiation by RNA Pol II and RNA Pol III and has led to the identification of a novel general initiation factor (TDF, translation dependent factor) that is present within the holoenzyme. TDF is selectively inactivated during protein synthesis inhibition by cycloheximide and at a late stage of adenovirus infection, thus accounting for the loss of RNA Pol III-mediated transcription of the tRNA and VA RNA genes under these conditions. On the basis of these observations, possible mechanisms for the global regulation of transcription by RNA Pol III and for disassembly of RNA Pol III initiation complexes are proposed.

Use of a nonviral vector to express a chimeric tRNA-ribozyme against lymphocytic choriomeningitis virus: cytoplasmic accumulation of a catalytically competent transcript but minimal antiviral effect

RNA polymerase III promoters direct the ubiquitous, high-level, expression of small, stable RNAs such as tRNAs, and thus are attractive candidates for achieving stable expression of small therapeutic (e.g., antiviral) molecules, such as ribozymes or antisense RNAs. In this article, we describe the use of a nonviral vector containing a tRNA promoter to express an antilymphocytic choriomeningitis virus (LCMV) ribozyme (tRNA-Rib5). The chimeric tRNA-ribozyme is specifically and efficiently transcribed by pol III in cell-free extracts, and the resulting transcript has appropriate ribozyme activity. In tissue culture studies, high levels of chimeric transcripts were readily detectable and were transported to the cytoplasm, the site of LCMV replication. Despite accumulation of tRNA-Rib5 in the cytoplasm of stably transformed cell clones, antiviral effects were minimal or absent. The implications of these findings and the potential use of this vector system for in vivo studies requiring the delivery of small molecules are discussed.

RNA transport

RNA molecules synthesized in the nucleus are transported to their sites of function throughout the eukaryotic cell by specific transport pathways. This review focuses on transport of messenger RNA, small nuclear RNA, ribosomal RNA, and transfer RNA between the nucleus and the cytoplasm. The general molecular mechanisms involved in nucleocytoplasmic transport of RNA are only beginning to be understood. However, during the past few years, substantial progress has been made. A major theme that emerges from recent studies of RNA transport is that specific signals mediate the transport of each class of RNA, and these signals are provided largely by the specific proteins with which each RNA is associated.

Antisense suppression of S-RNase expression in Nicotiana using RNA polymerase II- and III-transcribed gene constructs

In the Solanaceae, self-incompatibility is controlled by a single, multi-allelic ('S') locus. One product of this locus is a ribonuclease, the S-RNase, which is expressed predominantly in mature pistils and has recently been shown to cause allele-specific pollen rejection in transgenic plants. Hybrid Nicotiana plumbaginifolia x N. alata plants were used to test the effects of antisense suppression of the SA2-RNase from N. alata using three different gene constructs: two driven by RNA polymerase II-transcribed promoters, and the third, containing a truncated soybean tRNA (met-i) gene, transcribed by RNA polymerase III. All three constructs caused suppression of S-RNase activity in the transgenic plants. Unexpectedly, the CaMV 35S promoter was more effective for antisense suppression than the tissue specific tomato ChiP promoter. Antisense suppression of S-RNase correlated with low sense SA2 transcript levels and high antisense SA2 transcript levels. Untransformed hybrids that contained the N. alata SA2 allele were incompatible with N. alata SA2 pollen, while transgenic plants with suppressed SA2 gene expression accepted the pollen. The utility of this hybrid plant system for studying some aspects of antisense gene suppression is discussed.

Structure and function of a human transcription factor TFIIIB subunit that is evolutionarily conserved and contains both TFIIB- and high-mobility-group protein 2-related domains

Transcription factor TFIIIB plays a central role in transcription initiation by RNA polymerase III on genes encoding tRNA, 5S rRNA, and other small structural RNAs. We report the purification of a human TFIIIB-derived complex containing only the TATA-binding polypeptide (TBP) and a 90-kDa subunit (TFIIIB90) and the isolation of a cDNA clone encoding the 90-kDa subunit. The N-terminal half of TFIIIB90 exhibits sequence similarity to the yeast TFIIIB70 (BRF) and the class II transcription factor TFIIB and interacts weakly with TBP. The C-terminal half of TFIIIB90 contains a high-mobility-group protein 2 (HMG2)-related domain and interacts strongly with TBP. Recombinant TFIIIB90 plus recombinant human TBP substitute for human TFIIIB in a complementation assay for transcription of 5S, tRNA, and VA1 RNA genes, and both the TFIIB-related domain and the HMG2-related domain are required for this activity. TFIIIB90 is also required for transcription of human 7SK and U6 RNA genes by RNA polymerase III, but apparently within a complex distinct from the TBP/TFIIIB90 complex.

Inhibition of human immunodeficiency virus type 1 in human T cells by a potent Rev response element decoy consisting of the 13-nucleotide minimal Rev-binding domain

Intracellular immunization is an anti-viral gene therapy strategy based on the introduction of DNA templates into cells to stably express genetic elements which inhibit viral gene expression and replication. We have recently developed an intracellular immunization strategy for human immunodeficiency virus (HIV) infection that uses RNA decoys. RNA decoys are short RNA oligonucleotides corresponding to the HIV trans activation response element (TAR) or Rev response element (RRE) sequences, which function by inhibiting the binding of the HIV regulatory proteins Tat and Rev to the authentic HIV RNA TAR and RRE regions, respectively. In this report we describe the characterization of potent RRE decoys containing the minimal 13-nucleotide primary Rev binding domain of the RRE. Using an improved tRNA cassette to express high levels of RRE transcripts in CEM cells, we found that this new generation of minimal RRE decoys were more potent inhibitors of HIV in isolated cell lines than previously described TAR or RRE decoys. CEM cells expressing RRE decoys exhibited diminished Rev function in cotransfection assays, confirming the specificity of inhibition of HIV by RRE decoys and indicating that the 13-nucleotide minimal Rev binding domain defined by using in vitro binding studies also binds Rev in vivo. Significant differences in the degree of HIV inhibition between individual CEM cell lines transduced with RRE decoy vectors which were not due to sequence alterations in the tRNA-RRE DNA template, differences in RRE decoy expression level, or endogenous variations in the resistance of CEM clonal cell lines to HIV were observed. In order to evaluate the efficacy of RRE decoys in a more realistic fashion than by comparison of individual clonal cell lines, polyclonal populations of transduced CEM cells were infected with HIV. By using a novel flow cytometric method for quantitating intracellular p24 expression, one version of the RRE decoys tested in this study was found to be capable of durably protecting polyclonal populations of CEM cells from HIV.

Influenza virus NS1 protein inhibits pre-mRNA splicing and blocks mRNA nucleocytoplasmic transport

The influenza virus RNA segment 8 encodes two proteins, NS1 and NS2, by differential splicing. The collinear transcript acts as mRNA for NS1 protein, while the spliced mRNA encodes NS2 protein. The splicing of NS1 mRNA was studied in cells transfected with a recombinant plasmid that has the cDNA of RNA segment 8 cloned under the SV40 late promoter and polyadenylation signals. As described for influenza virus-infected cells, NS1 mRNA was poorly spliced to yield NS2 mRNA. However, inactivation of the NS1 gene, but not the NS2 gene, led to a substantial increase in the splicing efficiency, as shown by the relative accumulations of NS1 and NS2 mRNAs. This effect was not specific for NS1 mRNA, since the splicing of the endogenous SV40 early transcript was altered in such a way that t-Ag mRNA was almost eliminated. These changes in the splicing pattern coincided with a strong inhibition of the mRNA nucleocytoplasmic transport. Both NS1 and NS2 mRNAs were retained in the nucleus of cells expressing NS1 protein, but no effect was observed when only NS2 protein was expressed. Furthermore, other mRNAs tested, such as T-Ag mRNA and the non-spliceable nucleoprotein transcript, were also retained in the nucleus upon expression of NS1 protein, suggesting that it induced a generalized block of mRNA export from the nucleus.

An intron-containing tRNAArg gene within a large cluster of human tRNA genes

The insert within lambda Ht363, a recombinant selected from a bank of human genomic DNA cloned in lambda Ch4A, is described. Southern blot hybridization with a mixed tRNA[32P]pCp probe revealed the presence of four tRNA genes, which were shown to represent further copies of genes previously identified as a solitary tRNAGly gene and as a three gene cluster on two different recombinants. In vitro transcription of a fragment containing the three gene cluster revealed the presence of a further pol III gene, which was shown to be that for a tRNAArgTCT. This gene contains a 15 bp intron, the presence of which presumably prevented its detection on Southern blots by tRNA hybridisation. The gene is present in the previously reported cluster and occurs in higher copy number (> 7) in other arrangements in the genome. Most of the copies of the gene have related intron sequences.

A human DNA segment encompassing leucine and methionine tRNA pseudogenes localized on chromosome 6

A human genomic clone, designated LHtlm8, that strongly hybridized to a mammalian leucine tRNA(IAG) probe, was found to encompass a pair of tRNA pseudogenes that are transcribed in a homologous cell extract. A leucine tRNA(AAG) pseudogene (TRLP1) is 2.1-kb upstream and of opposite polarity to a methionine elongator tRNA(CAU) pseudogene (TRMEP1). TRLP1 has three nucleotide variations (97% identity) from its cognate leucine tRNA(IAG), while TRMEP1 has a 78% identity with its cognate tRNA. Similar to a number of other eukaryotic tRNA pseudogenes, presumptive precursor tRNA transcripts are generated from the two pseudogenes in vitro, but possibly due to their aberrant and unstable secondary and tertiary structures, no detectable mature tRNA products are observed. The two tRNA pseudogenes are encompassed within a 9.6-kb EcoRI fragment that has been assigned to the chromosomal locus, 6pter-q13, by Southern blot hybridization of human-rodent somatic cell hybrid DNAs with probes derived from the cloned tRNA pseudogenes and flanking sequences. A 4.4-kb EcoRI fragment also harbored in clone LHtlm8 was mapped to human chromosome 11, suggesting that the two EcoRI fragments were inadvertantly ligated together during construction of the genomic library.

Sequence-specific binding of transfer RNA by glyceraldehyde-3-phosphate dehydrogenase

A transfer RNA (tRNA) binding protein present in HeLa cell nuclear extracts was purified and identified as the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Studies with mutant tRNAs indicated that GAPDH recognizes both sequence and structural features in the RNA. GAPDH discriminated between wild-type tRNA and two tRNA mutants that are defective in nuclear export, which suggests that the protein may participate in RNA export. The cofactor nicotinamide adenine dinucleotide disrupted complex formation between tRNA and GAPDH and thus may share a common binding site with the RNA. Indirect immunofluorescence experiments showed that GAPDH is present in the nucleus as well as in the cytoplasm.

The first human genes for tRNA(ArgICG), tRNA(GlyUCC), and tRNA(ThrIGU) and more tRNA(Val) pseudogenes: expression and pre-tRNA maturation in HeLa cell-free extracts

A functional tRNA(Val) gene, which codes for the major tRNA(ValIAC) isoacceptor species, and three new tRNA(Val) pseudogenes have been isolated from human genomic DNA. Two tRNA(Val) pseudogenes and a tRNA(Val) variant gene were found to be associated with tRNA genes encoding tRNA(ArgICG), tRNA(GlyUCC), and tRNA(ThrIGU), respectively, on distinct DNA fragments. All tRNA genes, including the pseudogenes, are actively transcribed in HeLa nuclear extract. Pre-tRNAs of tRNA(Val), tRNA(Arg), tRNA(Thr), and tRNA(Gly) genes are correctly processed to mature-sized tRNAs, whereas the three tRNA(Val) pseudogenes yield stable pre-tRNAs in vitro. These findings reveal that, together with the three known pseudogenes, half of the members of the human tRNA(Val) gene family are pseudogenes, all of which are active in homologous nuclear extracts in vitro and presumably also in vivo.

Isolation of a mouse genomic clone containing four tRNACys-encoding genes

A cluster of four tRNACys-encoding genes with the anticodon GCA was found on a murine genomic clone containing an 18-kb DNA insert. Three of the four genes encode the identical tRNA, whereas the fourth gene has an altered nucleotide (nt) sequence. Two of the genes within a 2-kb PvuII fragment have the same polarity and are separated by only 921 bp. These two tRNAs have a different primary sequence. The changes in the nt sequences occur within three stems of the tRNA cloverleaf structure and weaken the strength of the H-bonds within the stems. All four genes (designated i-iv) have the 3' structural element that has been proposed as the transcription termination signal [Bogenhagen and Brown, Cell 24 (1981) 261-270]. The remainder of the flanking regions of the three identical tRNAs are very similar to each other, whereas the flanking regions of the fourth tRNA are distinctly different.

Expression of chimeric tRNA-driven antisense transcripts renders NIH 3T3 cells highly resistant to Moloney murine leukemia virus replication

NIH 3T3 cells infected with Moloney murine leukemia virus (MoMLV) express high levels of virus-specific RNA. To inhibit replication of the virus, we stably introduced chimeric tRNA genes encoding antisense templates into NIH 3T3 cells via a retroviral vector. Efficient expression of hybrid tRNA-MoMLV antisense transcripts and inhibition of MoMLV replication were dependent on the use of a particular type of retroviral vector, the double-copy vector, in which the chimeric tRNA gene was inserted in the 3' long terminal repeat. MoMLV replication was inhibited up to 97% in cells expressing antisense RNA corresponding to the gag gene and less than twofold in cells expressing antisense RNA corresponding to the pol gene. RNA and protein analyses suggest that inhibition was exerted at the level of translation. These results suggest that RNA polymerase III-based antisense inhibition systems can be used to inhibit highly expressed viral genes and render cells resistant to viral replication via intracellular immunization strategies.

Nuclear tRNA(Tyr) genes are highly amplified at a single chromosomal site in the genome of Arabidopsis thaliana

We have examined the organization of tRNA(Tyr) genes in three ecotypes of Arabidopsis thaliana, a plant with an extremely small genome of 7 x 10(7) bp. Three tRNA(Tyr) gene-containing EcoRI fragments of 1.5 kb and four fragments of 0.6, 1.7, 2.5 and 3.7 kb were cloned from A. thaliana cv. Columbia (Col-O) DNA and sequenced. All EcoRI fragments except those of 0.6 and 2.5 kb comprise an identical arrangement of two tRNA(Tyr) genes flanked by a tRNA(Ser) gene. The three tRNA genes have the same polarity and are separated by 250 and 370 bp, respectively. The tRNA(Tyr) genes encode the known cytoplasmic tRNA(G psi ATyr). Both genes contain a 12 bp long intervening sequence. Densitometric evaluation of the genomic blot reveals the presence of at least 20 copies, including a few multimers, of the 1.5 kb fragment in Col-O DNA, indicating a multiple amplification of this unit. Southern blots of EcoRI-digested DNA from the other two ecotypes, cv. Landsberg (La-O) and cv. Niederzenz (Nd-O) also show 1.5 kb units as the major hybridizing bands. Several lines of evidence support the idea of a strict tandem arrangement of this 1.5 kb unit: (i) Sequence analysis of the EcoRI inserts of 2.5 and 0.6 kb reveals the loss of an EcoRI site between 1.5 kb units and the introduction of a new EcoRI site in a 1.5 kb dimer. (ii) Complete digestion of Col-O DNA with restriction enzymes which cleave only once within the 1.5 kb unit also produces predominantly 1.5 kb fragments. (iii) Partial digestion with EcoRI shows that the 1.5 kb fragments indeed arise from the regular spacing of the restriction sites.(ABSTRACT TRUNCATED AT 250 WORDS)

Human tRNAGlu genes: their copy number and organisation

The tRNAGlu gene copy number, determined by genomic blot analysis of human placental DNA, is approximately thirteen. These studies, using several probes and DNA digested with several restriction enzymes singly or in combination, show that most of these tRNAGlu genes are flanked by DNA of very similar sequence for at least 5 kb. This conclusion is supported by the close similarity of the restriction maps of two lambda Charon-4A recombinants of human genomic DNA containing two different tRNAGlu genes.

Expression of chimeric tRNA-driven antisense transcripts renders NIH 3T3 cells highly resistant to Moloney murine leukemia virus replication

NIH 3T3 cells infected with Moloney murine leukemia virus (MoMLV) express high levels of virus-specific RNA. To inhibit replication of the virus, we stably introduced chimeric tRNA genes encoding antisense templates into NIH 3T3 cells via a retroviral vector. Efficient expression of hybrid tRNA-MoMLV antisense transcripts and inhibition of MoMLV replication were dependent on the use of a particular type of retroviral vector, the double-copy vector, in which the chimeric tRNA gene was inserted in the 3' long terminal repeat. MoMLV replication was inhibited up to 97% in cells expressing antisense RNA corresponding to the gag gene and less than twofold in cells expressing antisense RNA corresponding to the pol gene. RNA and protein analyses suggest that inhibition was exerted at the level of translation. These results suggest that RNA polymerase III-based antisense inhibition systems can be used to inhibit highly expressed viral genes and render cells resistant to viral replication via intracellular immunization strategies.

Overexpression of TAR sequences renders cells resistant to human immunodeficiency virus replication

Overexpression of TAR-containing sequences (TAR decoys) was used to render cells resistant to HIV replication. A chimeric tRNA(meti)-TAR transcription unit contained in a double copy murine retroviral vector was used to express high levels of HIV-1 TAR-containing transcripts in CEM SS cells. Replication of HIV-1 was inhibited over 99% in cells expressing chimeric tRNA-TAR transcripts, but an amphotropic murine retrovirus replicated normally in these cells. Expression of TAR sequences in CEM SS cells had no adverse effects on cell viability, indicating that essential cellular factors are not being sequestered in these cells. TAR decoy RNA-mediated HIV inhibition may also be effective against natural HIV isolates in spite of their hypervariable nature, as suggested by the fact that replication of SIVmac was also inhibited in cells expressing HIV-1 TAR decoys.

Genes, variant genes and pseudogenes of the human tRNA(Val) gene family. Expression and pre-tRNA maturation in vitro

Nine different members of the human tRNA(Val) gene family have been cloned and characterized. Only four of the genes code for one of the known tRNA(Val) isoacceptors. The remaining five genes carry mutations, which in two cases even affect the normal three-dimensional tRNA structure. Each of the genes is transcribed by polymerase III in a HeLa cell nuclear extract, but their transcription efficiencies differ by up to an order of magnitude. Conserved sequences immediately flanking the structural genes that could serve as extragenic control elements were not detected. However, short sequences in the 5' flanking region of two genes show striking similarity with sequences upstream from two Drosophila melanogaster tRNA(Val) genes. Each of the human tRNA(Val) genes has multiple, i.e. two to four, transcription initiation sites. In most cases, transcription termination is caused by oligo(T) sequences downstream from the structural genes. However, the signal sequences ATCTT and CTTCTT also serve as effective polymerase III transcription terminators. The precursors derived from the four tRNA(Val) genes coding for known isoacceptors and those derived from two mutant genes are processed first at their 3' and subsequently at their 5' ends to yield mature tRNAs. The precursor derived from a third mutant gene is incompletely maturated at its 3' end, presumably as a consequence of base-pairing between 5' and 3' flanking sequences. Finally, precursors encoded by the genes that carry mutations affecting the tRNA tertiary structure are completely resistant to 5' and 3' processing.

Localization of three DNA segments encompassing tRNA genes to human chromosomes 1, 5, and 16: proposed mechanism and significance of tRNA gene dispersion

The chromosomal locations of three cloned human DNA fragments encompassing tRNA genes have been determined by Southern analysis of human-rodent somatic cell hybrid DNAs with subfragments from these cloned genes and flanking sequences used as hybridization probes. These three DNA segments have been assigned to human chromosomes 1, 5, and 16, and homologous sequences are probably located on chromosome 14 and a separate locus on chromosome 1. These studies, combined with previous results, indicate that tRNA genes and pseudogenes are dispersed on at least seven different human chromosomes and suggest that these sequences will probably be found on most, if not all, human chromosomes. Short (8-12 nucleotide) direct terminal repeats flank many of the dispersed tRNA genes. The presence of these flanking repeats, combined with the dispersion of tRNA genes throughout the human genome, suggests that many of these genes may have arisen by an RNA-mediated retroposition mechanism. The possible functional significance of this gene dispersion is considered.

A human tRNA gene cluster encoding the major and minor valine tRNAs and a lysine tRNA

A human genomic DNA clone hybridizing to mammalian valine tRNA(IAC) contained a cluster of three tRNA genes. Two valine tRNA genes with anticodons of AAC and CAC, encoding the major and minor cytoplasmic valine tRNA isoacceptors, respectively, and a lysine tRNA(CUU) gene were identified by Southern blot hybridization and DNA sequence analysis of a 7.1-kb region. At least nine Alu family members were interspersed throughout the 18.5-kb human DNA fragment, with three Alu elements in the intergenic region between the valine tRNA(AAC) gene and the lysine tRNA gene. Each of the five Alu family members in the sequenced region can be categorized into one of the four Alu subfamilies. The coding regions of all three tRNA genes contain characteristic internal split promoter sequences and typical RNA polymerase III termination signals in the 3'-flanking regions. The tRNA genes are accurately transcribed by RNA polymerase III in a HeLa cell extract, since the RNase T1 fingerprints of the mature-sized tRNA transcription products are consistent with the structural genes. The lysine tRNA(CUU) gene was transcribed only slightly more efficiently than the valine tRNA(CAC) gene in the homologous in vitro transcription system. Surprisingly, the valine tRNA(CAC) gene was transcribed about eightfold more efficiently than the valine tRNA(AAC) gene, implicating the presence of a modulatory element in the upstream region flanking the tRNA(CAC) gene.

Purification of human transcription factor IIIC and its binding to the gene for ribosomal 5S RNA

Transcription factor hTFIIIC was purified from cytoplasmic extracts of HeLa cells using four different chromatographic steps. This procedure yields a protein fraction which actively supports transcription in reconstitution assays and contains five major polypeptide chains with a molecular mass ranging from 25 to 250 kDa as estimated by SDS-PAGE and silver staining. In this fraction a polypeptide with a molecular mass of approximately 110 kDa could be identified as a specific DNA-binding component of hTFIIIC. By electrophoretic mobility shift and footprinting analyses it could be demonstrated that purified hTFIIIC binds specifically to the 5S gene. The protected region encompasses the A-Box promoter element and flanking sequences extending toward the 5'-proximal end of the gene. By addition of hTFIIIC to preformed TFIIIA/5S DNA complexes, we observe an additive effect of both factors on the footprint boundaries.

Flanking sequences are required for efficient transcription and stable complex formation for the human tRNAiMet3-coding gene

An analysis of 5' and 3' deletions of the human tRNAiMet3 gene has revealed upstream regions required for efficient transcription and stable complex formation in vitro. The 5' boundary of this essential region lies between nucleotides -39 to -18 (start point = + 1), and it has been shown that 3'-flanking sequences near the first termination site are also important for stable complex formation. The transcriptional efficiency of two non-allelic loci (TMET3 and TMET2) has been compared and TMET2 is more active. An analysis of chimeric (hybrid) genes indicates that much of the difference seen is due to 5'-flanking sequences and that there may be complex interactions between 5' and 3' sequences.

Initiator tRNAMet genes from the nematode Caenorhabditis elegans

Five different members of the initiator tRNAMet gene family have been isolated and characterized from the nematode Caenorhabditis elegans. All five show identical tRNA coding sequences, followed by a block of T residues associated with termination by RNA polymerase III. Nucleotide sequences flanking the tDNAs are completely divergent, except for two distinct members with identical flanking sequences, which may have arisen from a recent gene duplication event. Each tDNA is also flanked by middle-repetitive DNA, but the lack of cross-hybridization to each other suggests that these repetitive sequences have no common functional significance. The tRNAMeti genes do not appear to be closely linked to each other, although in vitro transcription reveals a putative tDNA adjacent to one member. Finally, there are large differences in the extent to which the five genes are transcribed by a homologous C. elegans cell-free extract, suggesting that flanking sequences have a significant effect on transcription by RNA polymerase III.

Chromosomal assignment of a glutamic acid transfer RNA (tRNAGlu) gene to 1p36

A gene for tRNAGlu has been assigned to human chromosome 1p36 by in situ hybridisation using a [3H]-labelled or biotinylated 2.4-kb (human) DNA fragment containing a tRNAGlu gene as a probe. With the biotinylated DNA probe a secondary statistically significant site of hybridisation was observed at 1q21-22 which might represent a pseudogene or related sequence. In fibroblasts from gorilla (Gorilla gorilla) using biotin labelling, a single site of hybridisation occurred at 1qter which provides further support for homology of 1q in the higher apes and human 1p.

An efficiently transcribed human tRNA(Val) gene variant produces a stable pre-tRNA: repair of the processing defect by in vitro mutations

A tRNA(CACVal) gene variant, pHtV4, was cloned from human placenta genomic DNA. This gene differs from a closely related, functional tRNA(CACVal) gene by four base exchanges: T residues in place of C25, C62, and C66 create G:U pairs, and an A instead of G65 creates an A:C mismatch in the corresponding RNA transcript. The tRNA(Val) gene variant in pHtV4 is efficiently transcribed in HeLa cell nuclear extracts; however, the resulting pre-tRNA is processing-deficient, i.e., neither its 5'- nor its 3'-flanking sequences are removed to generate mature tRNA. Reversion of all four point mutations in pHtV4 by oligonucleotide-directed mutagenesis yielded a functional tRNA(CACVal) gene within the flanks of pHtV4, the pre-tRNA of which was processed to mature tRNA. Construction of a chimeric tRNA(Val) gene and site-directed mutagenesis of the tRNA(Val) gene in pHtV4, respectively, followed by transcription and processing studies showed that each of the four mutations contributes to the processing defect of the pHtV4-derived pre-tRNA. Moreover, this revealed that G:U pairs, which are common in all tRNAs, can impair pre-tRNA processing and therefore do not occur in certain positions in eukaryotic tRNAs.

Two human genes encoding tRNA(GCCGly)

Two human DNA fragments of 16.7 and 15.5 kb have been selected from a human lambda Charon-4A library by hybridization to an unfractionated tRNA probe. Restriction mapping and Southern and Northern hybridization analyses revealed the presence of a single tRNA-hybridizing region in each of the human DNA fragments. Nucleotide sequence analysis has identified two identical members of the tRNA(GCCGly) gene family. These tRNA(GCCGly) genes encode all of the conserved and semiconserved nucleotides of the tDNA split promoter sequences. Neither gene contains introns or encodes the CCA sequence present on the 3' terminus of mature tRNA. One of these identical tRNA(GCCGly) genes was found to be expressed at a substantially greater efficiency than the other in a HeLa cell lysate in vitro transcription system. No similarity was detected in the nucleotide sequences flanking these genes other than the characteristic, 3' oligo[dT] transcription termination signals and a TCTTT sequence located 7 to 10 bp upstream. These data are consistent with the hypothesis that as yet unidentified tDNA flanking sequences may have an important role in modulating human tRNA gene expression.

A human tRNA(iMet) gene produces multiple transcripts

A third nonallelic locus of the human methionyl-tRNA multigene family (tRNA(iMet-3) was isolated. This gene, unlike two other tRNA(iMet) loci, lacks a remarkable run of T and C residues which functions as a termination of transcription signal. Instead, three tandem termination signals, each containing no more than four thymidylate residues, function as relatively inefficient termination signals. As a result, polymerase readthrough generates at least three transcripts in vitro. The efficiency of apparent termination varies significantly at these sites. All resulting transcripts appear to be processed in vitro.

Increased levels of glycine tRNA associated with collagen synthesis

Analysis of codon usage for chick Type I collagen indicates that 89% of glycine codons are GGU/C. Since collagens are one-third glycine, chick Type I collagen synthesis should require large amounts of tRNAGly with the anticodon GCC. Earlier chromatographic studies of chick tRNA had indicated that connective tissues showed altered tRNAGly isoacceptor profiles [P. J. Christner and J. Rosenbloom (1976) Arch. Biochem. Biophys. 172, 399-409; H. J. Drabkin and L. N. Lukens (1978) J. Biol. Chem. 253, 6233-6241]. We have therefore used both two-dimensional gel electrophoresis and hybridization analysis to investigate whether collagen synthesis in chick connective tissues is associated with expression of a novel tRNAGly. Liver and calvaria tRNAs produced qualitatively similar patterns when separated on 2-D gels. Northern blots of 2-D-separated tRNAs from liver and calvaria, when hybridized to genes for vertebrate tRNAGly isoacceptors with GCC or UCC anticodons, showed hybridization to the same tRNAs in both tissues. Quantitation of tRNA species by dot blot hybridization indicated an increase in levels of the tRNAGly isoacceptor with anticodon GCC. Tissues synthesizing Type I collagen had a two- to threefold increase in this tRNA while tissues synthesizing Type II collagen showed a more modest increase. We conclude that elevated tRNAGly levels associated with collagen synthesis are due to increased amounts of the same isoacceptor which is the major tRNAGly in other tissues.

A human tRNA(iMet) gene produces multiple transcripts

A third nonallelic locus of the human methionyl-tRNA multigene family (tRNA(iMet-3) was isolated. This gene, unlike two other tRNA(iMet) loci, lacks a remarkable run of T and C residues which functions as a termination of transcription signal. Instead, three tandem termination signals, each containing no more than four thymidylate residues, function as relatively inefficient termination signals. As a result, polymerase readthrough generates at least three transcripts in vitro. The efficiency of apparent termination varies significantly at these sites. All resulting transcripts appear to be processed in vitro.

Transcription complexes for various class III genes differ in parameters of formation and stability towards salt

RNA polymerase III faithfully transcribes the genes for ribosomal 5 S RNA, tRNA(1Met) or adenovirus VA RNA in vitro in the presence of required transcription factors. These genes display distinct differences in the kinetics of transcription complex formation and in their response to excess template. In contrast to tRNA and VA RNA synthesis, 5 S RNA synthesis displays a lag phase of 15 minutes before the onset of transcription and is clearly inhibited by high concentrations of template. Once formed, transcription complexes for the RNA polymerase III genes listed can be isolated by glycerol gradient centrifugation and display a remarkable stability against transient treatment with high salt concentrations. Complexes for 5 S RNA and tRNA remain functionally active up to 2.5 M-KCl. The activity of transcription complexes for VA RNA, however, is significantly diminished after treatment with high salt concentrations. This effect is shown to be due to an irreversible loss of transcription factors. RNA polymerase III is dissociated by high salt concentrations from all the transcription complexes studied but remains part of these complexes during the normal reinitiation cycle at 60 mM-KCl. An additional method for the purification of partial transcription complexes was developed that involves equilibrium centrifugation on cesium sulfate gradients. This method completely releases TFIIIB from 5 S complexes and a core complex, composed of the 5 S RNA gene, factors IIIA and IIIC, is retained. In the case of tRNA and VA RNA, core complexes are obtained that remain partly associated with TFIIIC and TFIIIB. These results indicate a qualitatively and/or quantitatively different interaction of individual factors in different polymerase III transcription complexes.

Modulation of transcriptional activity and stable complex formation by 5'-flanking regions of mouse tRNAHis genes

We determined the nucleotide sequences of three mouse tRNAHis genes and a tRNAGly gene present in two different lambda clones. One lambda clone contained two tRNAHis genes 600 base pairs (bp) apart in opposite orientations. The other clone contained a tRNAHis and a tRNAGly gene 569 bp apart in the same orientation. The coding regions of the three tRNAHis genes were identical to sequenced mammalian tRNAHis if posttranscriptional modifications are not considered. Notably, the three tRNAHis genes and a fourth gene previously sequenced by us contained within the flanking regions, various amounts of short, conserved 5' leader sequences and 3' trailer sequences directly abutting the coding regions. Otherwise the flanking regions were not homologous. Deletion mutants of one of the tRNAHis genes were constructed which contained 228, 99, 9, and 3 bp of the wild-type 5'-flanking region, respectively. Deletion of 5'-flanking sequences from positions -9 to -4 reduced transcriptional activity substantially (ca. fivefold) in a HeLa cell S-100 lysate. This effect was independent of the vector sequences in the deletion clone, implying that the region from -4 to -9 of the intact gene contains a positive modulatory element for transcription in vitro. The deletion mutant containing 3 bp of wild-type 5'-flanking sequence also had a greatly reduced ability to inhibit the transcription of a second tRNA gene in a competition assay. Thus, the normal 5'-flanking region influences the ability of the gene to form stable complexes with transcription factors. These data further indicate that a mammalian transcription extract is sensitive to 5'-flanking-region effects if a suitable tRNA gene is assayed.

Analysis of a human gene cluster coding for tRNA(GAAPhe) and tRNA(UUULys)

A 13.8-kb fragment of human DNA isolated from a human lambda Charon-4A DNA library was found to contain four human tRNA genes. Nucleotide sequence analysis of approx. 3.7 kb of this segment of human DNA identified two lysine tRNA(UUU) genes identical in coding sequence to a previously reported human lysine tRNA gene [Roy et al., Nucl. Acids Res. 10 (1982) 7313-7322]. The other two tRNA genes were phenylalanine tRNA(GAA) genes, the first to be isolated from a mammalian source. These phenylalanine tRNA(GAA) genes were identical in sequence with the exception of a G/A polymorphism at coordinate 57. None of these tRNA genes contains introns. The tRNA(UUULys) and tRNA(GAAPhe) genes are organized in alternating order and are irregularly spaced, by intergenic regions of approx. 1.0, 2.6 and 5.0 kb, and randomly oriented. There was no evidence to indicate that any of these genes arose by gene duplication, since flanking sequence homology was limited to the putative RNA polymerase III termination signals in the 3'-flanking regions. A mature tRNA-sized product was identified following the transcription of each tRNA gene in a homologous in vitro transcription system. Interestingly, different levels of transcriptional activity of the three identical lysine tRNA genes were observed, suggesting modulation of tDNA expression by extragenic sequences. In addition, a minimum of eight regions of homology to Alu-type repetitive elements were detected in this human DNA fragment, one of which was located 53 bp upstream from a tRNA(GAAPhe) gene.

Two human tyrosine tRNA genes contain introns

A human lambda-phage recombinant which contains at least four tRNA genes, has been isolated. Two DNA fragments were subcloned to give the recombinant plasmids pM6 and pM6128. Nucleotide sequence analysis showed that each plasmid contained a different tyrosine acceptor tRNA (tRNATyr) gene. Both tRNATyr genes are interrupted by 21-bp introns. These recombinant plasmids have been shown to direct the in vitro synthesis of tRNA-sized products in a HeLa cell transcription system.

Nucleotide sequence and transcription of a human glycine tRNAGCC gene and nearby pseudogene

A bacteriophage lambda clone containing a 15.4-kb human DNA fragment was isolated and found to contain a glycine tRNA gene and, 758 bp away, a pseudogene, both with an anticodon of GCC. The nucleotide (nt) sequence of a 1362-bp segment of this clone, encompassing the gene, pseudogene, and their flanking regions, was determined. The gene and pseudogene have an identical sequence of eight nt (5'-CAGCTGGA-3') in their 5'-flanking regions immediately preceding the coding regions, as well as characteristic transcription termination sites of five consecutive T nt in the 3'-flanking regions. Neither of these genes has intervening sequences. Only one of the two genes was efficiently transcribed in vitro by RNA polymerase III in a HeLa cell-free system. During the course of transcription, primary transcripts of one gene were processed to yield mature-sized products. In contrast, the level of transcription of the second gene was significantly less than that of the first, and no mature-sized products could be detected. The nt sequence of the inefficiently transcribed gene has two base substitutions compared to the sequence of the efficiently transcribed gene, and the DNA sequence predicted from the human placental tRNAGlyGCC sequence. One of these nt substitutions is a C to T transition in the TTCG sequence within the B block of the characteristic internal split promoter sequence. The precursor-product relationships of the tRNA transcripts were established by comparing the RNase T1 and RNase A fingerprints of the precursors and products.

The human tRNAVal gene family: organization, nucleotide sequences and homologous transcription of three single-copy genes

At least 13 independent tRNAVal gene loci were detected in the human genome. Three of these genes were isolated and shown to occur only once in the haploid genome. No further functional tRNA genes are located on the isolated clones. Two tRNAVal genes encode the known major and minor tRNAVal isoacceptors, the third may be a pseudogene because a corresponding tRNAVal is not yet known. Comparison of extragenic sequences did not reveal significant homologies, indicating the separation of these genes early in vertebrate evolution. An Alu-type repeat was found in two of the clones within several hundred bp distance from the tDNA. All three genes are transcriptionally active in a HeLa nuclear extract. We show here for the first time that homologous in vitro transcription of mammalian tRNA genes strongly depends on extragenic control regions: interestingly, as a consequence of different flanking regions, the transcription efficiencies vary by an order of magnitude among the genes for the major and the minor tRNAVal and thus reflect the concentrations of these tRNAs in vivo.

Modulation of transcriptional activity and stable complex formation by 5'-flanking regions of mouse tRNAHis genes

We determined the nucleotide sequences of three mouse tRNAHis genes and a tRNAGly gene present in two different lambda clones. One lambda clone contained two tRNAHis genes 600 base pairs (bp) apart in opposite orientations. The other clone contained a tRNAHis and a tRNAGly gene 569 bp apart in the same orientation. The coding regions of the three tRNAHis genes were identical to sequenced mammalian tRNAHis if posttranscriptional modifications are not considered. Notably, the three tRNAHis genes and a fourth gene previously sequenced by us contained within the flanking regions, various amounts of short, conserved 5' leader sequences and 3' trailer sequences directly abutting the coding regions. Otherwise the flanking regions were not homologous. Deletion mutants of one of the tRNAHis genes were constructed which contained 228, 99, 9, and 3 bp of the wild-type 5'-flanking region, respectively. Deletion of 5'-flanking sequences from positions -9 to -4 reduced transcriptional activity substantially (ca. fivefold) in a HeLa cell S-100 lysate. This effect was independent of the vector sequences in the deletion clone, implying that the region from -4 to -9 of the intact gene contains a positive modulatory element for transcription in vitro. The deletion mutant containing 3 bp of wild-type 5'-flanking sequence also had a greatly reduced ability to inhibit the transcription of a second tRNA gene in a competition assay. Thus, the normal 5'-flanking region influences the ability of the gene to form stable complexes with transcription factors. These data further indicate that a mammalian transcription extract is sensitive to 5'-flanking-region effects if a suitable tRNA gene is assayed.

Nucleotide sequence and transcription of a human tRNA gene cluster with four genes

A bacteriophage lambda clone containing a 20-kb human DNA segment was isolated and found to harbor a cluster of four tRNA genes. An 8.2-kb HindIII subfragment encompassing the genes was cloned into pBR322 for restriction mapping and DNA sequence analysis. The genes were found to be arranged as two tandem pairs, separated by 3 kb. A proline tRNAAGG gene is separated from a leucine tRNAAAG gene by a 724-bp intergenic region in the first pair, and a second proline tRNAAGG gene is 316 bp from a threonine tRNAUGU gene in the second pair, with the leucine tRNA gene being of opposite polarity to the other three genes. A putative Alu-like element was found to occur within a 2.0-kb DNA fragment, at least 0.7 kb from the tRNA gene cluster. The coding sequences of the two proline tRNAAGG genes are identical. The coding regions of all four tRNA genes contain consensus internal split promoter sequences and do not have intervening sequences nor the CCA trinucleotide found in mature tRNAs. The 3'-flanking regions of these four tRNA genes have normal RNA polymerase III termination sites of at least four consecutive T nucleotides. No apparent homologies occur between the 5'-flanking regions of these genes. All four tRNA genes are accurately transcribed in an in vitro HeLa cell-free system, and the RNase T1 fingerprints of the mature-sized tRNA transcripts were found to be consistent with the DNA sequences of the genes.

tRNA nuclear transport: defining the critical regions of human tRNAimet by point mutagenesis

We recently described a carrier-mediated nuclear transport system for tRNA in Xenopus laevis oocytes. A natural human tRNAimet variant with a G to T transversion in position 57 is defective in transport across the nuclear membrane. In addition, processing of the primary transcript of the variant gene is much less efficient than the wild type. We now describe the nuclear transport and processing phenotypes of 30 different point mutants generated by in vitro mutagenesis of a wild-type human tRNAimet gene. The effects of each nucleotide change on processing and transport were analyzed in X. laevis oocytes following nuclear microinjection of each mutant gene. Mutants exhibiting transport-defective behavior were further characterized by measuring transport kinetics of the purified mature tRNA. Our studies demonstrate that many mutations affect tRNAimet nuclear transport, although those with the most deleterious effects are clustered in the highly conserved D stem-loop and T stem-loop regions.

Repeat sequence families derived from mammalian tRNA genes

Short interspersed repetitive DNA sequences (SINEs) are the major component of dispersed repetitive DNA in all mammalian genomes. Most SINEs contain an intragenic RNA polymerase III promoter that initiates transcription at the 5' end of the repeated DNA sequence and which has been proposed to facilitate the transposition and amplification of these sequences by an RNA-intermediate mechanism. We have discovered several SINE families in the prosimian Galago crassicaudatus which have promoter regions similar to transfer RNA genes. To determine the relationship between Galago SINEs and mammalian tRNA genes, we have compared their sequences. Here, we demonstrate that the Galago monomer and type II SINE families are 68 and 62% homologous, respectively, with a human methionine tRNA gene. We have extended our analysis to include the rat identifier and mouse B2 families and show that their sequences are closely related to alanine and serine tRNA genes, respectively. Our observations suggest that many mammalian SINE families are amplified tRNA pseudogenes.

Conserved sequences in both coding and 5' flanking regions of mammalian opal suppressor tRNA genes

The rabbit genome encodes an opal suppressor tRNA gene. The coding region is strictly conserved between the rabbit gene and the corresponding gene in the human genome. The rabbit opal suppressor gene contains the consensus sequence in the 3' internal control region but like the human and chicken genes, the rabbit 5' internal control region contains two additional nucleotides. The 5' flanking sequences of the rabbit and the human opal suppressor genes contain extensive regions of homology. A subset of these homologies is also present 5' to the chicken opal suppressor gene. Both the rabbit and the human genomes also encode a pseudogene. That of the rabbit lacks the 3' half of the coding region. Neither pseudogene has homologous regions to the 5' flanking regions of the genes. The presence of 5' homologies flanking only the transcribed genes and not the pseudogenes suggests that these regions may be regulatory control elements specifically involved in the expression of the eukaryotic opal suppressor gene. Moreover the strict conservation of coding sequences indicates functional importance for the opal suppressor tRNA genes.