tRNA Base Research Articles

Possible depletion of a DNA repair enzyme in human lymphoma cells by subversive repair.

Mex+ human lymphoma cell lines contain O6-methylguanine-DNA methyltransferase, a DNA repair enzyme that undergoes suicide inactivation on interaction with its substrate. The cells are therefore competent to remove the alkylation lesion O6-methylguanine from their DNA. However, several repair-deficient lymphoma cell lines (Mex-) are also known. It is shown here that Mex+ cells can be converted temporarily to a Mex- phenotype by growth in nontoxic concentrations of free O6-methylguanine. The depletion of methyltransferase activity is not a result of O6-methylguanine incorporation into DNA and subsequent demethylation by the enzyme. It is proposed that O6-methylguanine is mistakenly incorporated into tRNA molecules by means of a post-transcriptional ribosyl transfer reaction. The demethylation of such bases in tRNA has been demonstrated by using bacterial and human DNA repair enzymes. The existence of such a subversive repair of a methylated base in tRNA raises the possibility of competition between DNA and RNA for cellular DNA repair enzymes. Furthermore, it is proposed that the known aberrant methylation of tRNA in certain transformed cells, together with subversive tRNA repair, could account for the Mex- phenotype.

TRNA-rRNA sequence homologies: Evidence for an ancient modular format shared by tRNAs and rRNAs

Homologies between tRNAs and rRNAs are identified in searches using various combinations of Escherichia coli, yeast, Halobacterium volcanii and bovine mitochondrial sequences. As in previously reported comparisons, the homologies are too frequent and long to be attributed to coincidence, and similar frequencies from inter- and intraspecies comparisons preclude evolutionary convergence as an explanation. In contrast to the earlier studies, patterns in the positioning of the homologies are now described. Graphing the positions of the homologies along orthogonal axes that represent numbers of bases in tRNA and rRNA shows recurring patterns in the alignments. Preferred spacings of integral multiples of 9 bases are found, suggesting a periodicity in the ancestral structure from which the tRNAs and rRNAs were derived. The periodicity also suggests persistence of a modular format in both classes of molecules that survived changes in sequence that occured during evolution. A model is proposed for the generation of the ancestral molecule and the early evolution of the coding mechanism. Elongation by self-priming and self-templating gave a hairpin with a 9 base stem. Two additional cycles gave a 70–80 base tRNA-like structure. Additional cycles yielded a tandem repeat of this unit, roughly equivalent in size to the combined rRNAs of prokaryotes. The larger RNA would contain the information and materials for generating the smaller RNAs. It is proposed that multiple recombination among such molecules gave composite structures, presumed progenitors of today's t- and rRNAs. The distribution of the conserved domains among today's species argues for the existence of the ancestral molecule prior to divergence of lines leading to the various kingdoms. Their presence in the different nucleic acids suggests the existence of a nucleic acid with multiple functions prior to partitioning of these functions among the nucleic acids that exist today. The occurrence of overlaps, overlays and consensus alignments among the homologies provides the means for identifying contiguous and neighboring conserved regions and holds promise for the reconstruction of the sequence of an ancestral molecule.

Modified bases in transfer RNA

Transfer RNA is uniquely enriched with modified bases. A large body of information has accumulated about the occurrence, nature and distribution of modified bases in tRNA. But similar investigations on the enzymes involved in this post-transcriptional modification have been hampered by the instability of the enzymes and lack of suitable substrates. The present review summarises briefly, the occurrence and methods of detection of modified bases, the enzymes involved in their formation and also certain suggestive evidence for the role of modification in cellular metabolism.

TRNA base modifications and gene regulation

tRNA base modifications and gene regulation

Mutations affecting excision of the intron from a eukaryotic dimeric tRNA precursor.

The nucleotide sequences of a Schizosaccharomyces pombe opal suppressor serine tRNA gene (sup9-e) and of 12 in vivo-generated mutant genes, which have lost the ability to suppress UGA mutations, have been determined. Analysis of the expression of these genes in Saccharomyces cerevisiae in vitro and in vivo systems has revealed defects in tRNA gene transcription and precursor tRNA processing. Single base changes in the D-loop, the intron and the extra arm affect the efficiency of splicing of the tRNA precursors while an anti-codon stem mutation may affect the accuracy of this process. Two mutations which occur in the intervening sequence of the sup9-e gene allow an alternate tRNA base pairing configuration. Transcription of the sup9-e gene and of the adjacent tRNAMet gene (located 7 bp downstream) is essentially abolished in vivo by a G----A19 mutation in the tRNASer gene, suggesting that tRNAMet may be derived solely via processing of the tRNASer-tRNAMet dimeric precursor.

Comparison of ribosomal entry and acceptor transfer ribonucleic acid binding sites on Escherichia coli 70S ribosomes. Fluorescence energy transfer measurements from Phe-tRNAPhe to the 3' end of 16S ribonucleic acid.

Distances were measured by nonradiative energy transfer from fluorescent probes specifically located on one of three points of yeast or Escherichia coli Phe-tRNAPhe enzymatically bound to the entry site or to the acceptor site of E. coli 70S ribosomes to energy-accepting probes on the 3' end of the 16S ribonucleic acid (RNA) of the 30S subunit. The Y base in the anticodon loop of yeast tRNAPhe was replaced by proflavin. Fluorescein isothiocyanate was attached to the X base (position 47) of E. coli tRNAPhe. E. coli tRNAPhe which had been photochemically cross-linked between positions 8 and 13 followed by chemical reduction to form a fluorescent probe was also used. Labeled tRNAs were aminoacylated and enzymatically bound to the ribosome in the presence of elongation factor Tu and guanosine 5'-triphosphate (acceptor-site binding) or a nonhydrolyzable analogue (entry-site binding). Nonradiative energy transfer measurements were made of the distances between fluorophores located on the Phe-tRNA and the fluorophore at the 3' end of 16S RNA. Calculations were based on comparison of the fluorescence lifetime of the energy donor, located on the Phe-tRNA, in the absence and presence of an energy acceptor on the 3' end of the 16S RNA. Under both sets of binding conditions, the distances to the 3' end of 16S RNA were found to be the following: cross-linked tRNA, greater than 69 A; Y base of tRNA, greater than 61 A. The distance between the 3' end of 16S RNA and the X base of tRNA was found to be 81 A under acceptor-site binding conditions but greater than 86 A under entry-site binding conditions.

Alterations in post-transcriptional modification of the Y base in phenylalanine tRNA from tumor cells.

Various tumor cells contain chromatographically distinct isoacceptor tRNA species. To decide whether the tumor-specific species represent an expression of a separate tRNA gene or only an undermodified form of normal tRNAPhe, nucleotide sequences of tRNAPhe isolated from neuroblastoma and normal mouse liver were determined by postlabeling techniques. The results showed identical sequences except for the changes of post-transcriptional modifications in the anticodon loop. Normal mouse liver tRNAPhe contained Cm32, Gm34, and the hypermodified YOH next to the 3' end of the anticodon. On the contrary, tRNAPhe from neuroblastoma contained C32, G34, and, instead of YOH base m1G. A small proportion of tRNAPhe species contained an undermodified YOH base. For the examination of the conditions leading to the undermodified tRNAPhe, Vero cells derived from the kidney of African green monkey in culture were used. In these cells, deprivation of methionine or lysine resulted in changes in tRNAPhe modification similar to those in tumor cells. Ehrlich ascites tumor cells were examined to determine whether the presence of altered tRNAPhe species in various tumors is also the result of starvation of some nutritional factors. Results obtained with these cells showed that tRNAPhe species lacking the Y base disappeared in tumor-bearing mice after intraperitoneal injection with a mixture of amino acids and vitamins. Thus it is concluded that tumor-specific tRNAPhe species are the products of aberrant post-transcriptional modification, not the transcripts of different, normally repressed genes.

A spectrophotometric study of tRNA-Cu(II) complexes

The spectrum of the aqueous copper(II) ion is changed in the presence of the tRNA molecule: a marked increase in absorption and a shift in its maximum from 800 to 740 nm is observed (Fig. 1). This ▪ effect is similar to the one observed on coordination of Cu(II) with ammonia and ethylene diamine [1] and is due to the stronger ligand field of the nitrogen derivatives. Therefore our data seem to indicate that the copper hydrate ion changes a water molecule with a heterocylic base of tRNA at one of its nitrogen donor sites creating coordination complex. The ΔA 710 value obtained from the differential spectrum (2) in Fig. 1 is proportional to the bound copper concentration. We observed that ΔA 710 increased immediately as Cu(II) ions were added and the process shows a saturation on for a [Cu 2+]/[PO 4 −] ratio greater than 0.5 (Fig. 2). An increase in the NaCl concentration from 0.1 to 0.7 M has some influence, however small, on the formation of ▪ ▪ the tRNA-Cu(II) complex. This influence manifest mostly at values of the [Cu 2+]/[PO 4 − ratio grater than 0.1. The formation of the copper complex with tRNA is on the contrary hindered by the presence of small quantities of Co 2+ or Mn 2+ and this fact suggests competition of the Cu 2+ ions with the other transition metals for the same sites of the tRNA molecule. A further indication of the transition metal binding to the tRNA heterocyclic bases may be obtained from the analysis of the changes in optical properties of bases which accompany the binding of Cu 2+, Mn 2+ and Co 2+ ions [2]. When Cu 2+ is added to the tRNA solution a change in the tRNA electronic spectrum is observed with the appearance of an absorption band in the 280–300 nm range (Fig. 3). A comparison of the absorption changes which follow the varying of the copper concentration at 280 nm (see insert on the top of Fig. 3) with the analogue measure previously performed at 710 nm (Fig. 2) confirmed that both absorption changes completed in the same copper concentration range, that is they are measurement of the same effect.

Molecular basis for the genetic code.

It was found by using the CPK molecular model that holes on the complexes of four nucleotides (C4N) on the tRNAs, namely complexes of the anticodon bases with the discriminator base at 4th position of 3′ end, had lock and key relations to the corresponding protein amino acids. Various general features of the universal and mitochondrial genetic codes were easily explained in terms of the C4N model. The recognition mechanism of the tRNA by the aminoacyl-tRNA-synthetase is closely correlated with the formation of the C4N on the Rossmann fold on the synthetase. The meaning of the hypermodification of the tRNA base next to the third anticodon base and other phenomena were also discussed.

Changes of post-transcriptional modification of wye base in tumor-specific tRNAPhe.

Nucleotide sequences of normal mouse liver tRNAPhe and tumor-specific tRNAPhes isolated from Ehrlich ascites tumor and neuroblastoma cells were examined by post-labeling techniques. The results showed that their sequences are identical, except for changes in post-transcriptional modifications that are located in the anticodon region. Normal mouse liver tRNAPhe contained Cm32, Gm34 and YOH37. On the other hand, tumor-specific tRNAPhes were found in one of two possible configurations: 1) Cm32, Gm34 and Y*OH37 (under-modified YOH) or 2) C32, G34 and m1G37. The ratio of the two forms of tRNAPhes differed in different tumor cells; Ehrlich ascites tumor tRNAPhe had mainly Y*OH-containing tRNAPhe whereas neuroblastoma tRNAPhe has predominantly m1G-containing tRNAPhe. It was concluded that tumor-specific tRNAPhes are products of different extents of modification, rather than of new tRNA transcription.

TRNA-guanine transglycosylase and guanine-accepting transfer RNAs of Drosophila melanogaster

tRNA-guanine transglycosylase replaces a base in tRNA with guanine and uses free guanine as the substrate. This enzyme was obtained from adult Drosphila melanogaster and characterized as follows: K m for guanine is 2.8 × 10 −7 M; pH optimum is 7.4. The enzyme activity is present in third instar larvae and adult flies but is absent in late pupae. When Drosphila tRNAs were guanylated with the comparable enzyme from rabbit reticulocytes, at least four species of tRNA could be demonstrated to have incorporated radioactive guanine. The role for this enzyme appears to be to exchange queuine into tRNA and thus replace a guanine in the anticodon of tRNAs for Asn, Asp, His, and Tyr. In the absence of queuine it will catalyze the exchange reaction described above.

A phase relationship associates tRNA structural gene sequences with nucleosome cores

DNA (760 bp) isolated from nucleosome tetramers of staphylococcal nuclease-digested chicken embryo chromatin was highly enriched for tRNA genes and subsequently cloned in E. coli X 1776. The location of genes coding for chicken embryo tRNA Lys, tRNA Phe and tRNA i Met within the cloned nucleosome tetramer DNA was determined using restriction endonucleases for which single cleavage sites could be predicted from the respective tRNA base sequence. All our tRNA genes reside nonrandomly at four locations on nucleosome tetramer DNA. The spacing between the tRNA gene locations is approximately 190 bp, similar to the DNA repeat length of chicken embryo chromatin. The four tRNA gene locations were also defined in noncloned nucleosome tetramer DNA highly enriched for tRNA genes. The majority of genes coding for tRNA Lys, tRNA Phe and tRNA i Met, respectively, are located in equal proportion 40–45, 230, 420 and 610 bp distant from the 5′ end of the tRNA-identical strand. Thus the tRNA structural gene sequences all appear to begin about 20 bp “inside” the nucleosome core. As observed with nucleosomal DNA not enriched for tRNA genes, the phase relationship between tRNA genes and nucleosome location is maintained over a distance of 4–6 subsequent nucleosomes. A cloned molecule of nucleosomal DNA containing both a tRNA Lys gene and a tRNA i Met gene in the same polarity reveals that a phase adjustment might be necessary for the nucleosomes between these two tRNA genes in chicken embryo chromatin.

Fluorescence studies of the binding of a yeast tRNA Phe derivative to Escherichia coli ribosomes

The equilibrium binding of a highly fluorescent derivative of yeast tRNA Phe to Escherichia coli 70 S ribosomes was studied fluorimetrically at 7 °C in 25 m m-magnesium. Under these conditions 70 S ribosomes bind two deacylated tRNAs stoichiometrically. An analysis of the binding data using a model in which occupancy of the weaker site requires prior occupancy of the stronger site leads to apparent association constants of (1.00 ± 0.05) × 10 9 m −1 and (3.4 ± 0.2) × 10 7 m −1. The use of an independent site model does not change these values appreciably. The observed binding constants do not depend upon the presence or absence of the messenger RNA, poly(U). However, spectroscopic evidence strongly suggests that the anticodons of both bound tRNAs are in contact with the message. This evidence further suggests that in the presence of poly(U) the environment of the hypermodified base adjacent to the anticodon is substantially different in the two sites. This may reflect a difference in the conformation of the anticodon loops or an interaction between the hypermodified base of the weak site tRNA and the anticodon loop of the strong site tRNA.

Deficiency of the peroxy-Y base in oocyte phenylalanine tRNA

Deficiency of the peroxy-Y base in oocyte phenylalanine tRNA

Incorporation of lysine into Y base of phenylalanine tRNA in Vero cells

Vero cells, a line derived from African green monkey kidney, contains a hypermodified base, called Y, adjacent to the 3' end of the anticodon of tRNAPhe. Two types of evidence are presented suggesting that lysine is involved in biosynthesis of Y base in these cells. First, when Vero cells are starved for lysine, a new, early-eluting species of tRNAPhe which lacks the fully modified Y base can be detected by reversed phase chromatography (RPC-5). After addition of lysine to the medium, this new species disappears. Second, when these cells are grown in low-lysine medium and then exposed to [3H]lysine, radioactivity from the lysine comigrates with tRNAPhe. The Y base can be selectively excised from tRNAPhe by incubation at pH 2.9, and extracted into ethyl acetate. Thin-layer chromatography of acid-excised material from these cells reveals that lysine-derived radioactivity comigrates with genuine Y base from calf liver tRNAPhe and the acid-excised tRNA no longer contains radioactivity. These results are consistent with the model that lysine is a structural precursor of Y base in tRNAPhe of Vero cells.

HPLC determination of the base composition of yeast tRNA

A relatively simple procedure was developed to quantitate normal and methylated tRNA bases by isocratic HPLC. tRNA was extracted with phenol from lyophilyzed cells, purified and precipitated with isopropanol. After perchloric acid hydrolysis, the samples were subjected to HPLC analysis. The mole % composition of normal and methylated bases was determined in yeasts grown on unusual carbon sources including hydrocarbons and light alcohols. Normal and methylated base composition of total tRNA depends on the microorganism and on the carbon source used.

Selective inhibition of uracil tRNA methylases of E. coli by ethionine

L-ethionine has been found to inhibit uracil tRNA methylating enzymes in vitro under conditions where methylation of other tRNA bases is unaffected. No selective inhibitor for uracil tRNA methylases has been identified previously. 15 mM L-ethionine or 30 mM D,L-ethionine caused about 40% inhibition of tRNA methylation catalyzed by enzyme extracts from E. coli B or E. coli M3S (mixtures of methylases for uracil, guanine, cytosine, and adenine) but did not inhibit the activity of preparations from an E. coli mutant that lacks uracil tRNA methylase. Analysis of the 14CH3 bases in methyl-deficient E. coli tRNA after its in vitro methylation with E. coli B3 enzymes in the presence or absence of ethionine showed that ethionine inhibited 14CH3 transfer to uracil in tRNA, but did not diminish significantly the 14CH3 transfer to other tRNA bases. Under similar conditions 0.6 mM S-adenosylethionine and 0.2 mM ethylthioadenosine inhibited the overall tRNA base methylating activity of E. coli B preparations about 50% but neither of these ethionine metabolites preferentially inhibited uracil methylation. Ethionine was not competitive with S-adenosyl methionine. Uracil methylation was not inhibited by alanine, valine, or ethionine sulfoxide. It is suggested that the thymine deficiency that we found earlier in tRNA from ethionine-treated E. coli B cells, resulted from base specific inhibition by the amino acid, ethionine, of uracil tRNA methylation in vivo.

Modified bases in tRNA: the structures of 5-carbamoylmethyl- and 5-carboxymethyl uridine.

The crystal structures of two nucleosides, 5-carbamoylmethyluridine (1) and 5-carboxymethyluridine (2), were determined from three-dimensional x-ray diffraction data, and refined to R = 0.036 and R = 0.047, respectively. Compound 1 is in the C3'-endo conformation with chi +5.2 degrees (anti), psiinfinity = +63.4 degrees and psialpha = +180.0 degrees (tt); 2 is in the C2'endo conformation with chi +49.4 degrees (anti), psiinfinity -60.5 degrees and psialpha +60.0 degrees (gg). For each derivative, the plane of the side chain substituent is skewed with respect to the plane of the nucleobase; for 1, the carboxamide group is on the same side of the uracil plane vis a vis the ribose ring; for 2, the carboxyl group is on the opposite side of this plane. No base pairing is observed for either structure. Incorporation of structure 1 into a 3'-stacked tRNA anticodon appears to place 08 within hydrogen bonding distance of the 02' hydroxyl of ribose 33, which may limit the ability of such a molecule of tRNA to "wobble".

Mass spectral studies of probe pyrolysis products of intact oligoribonucleotides.

Pyrolysis of underivatized homogeneous oligoribonucleotides in the source of a mass spectrometer leads to production of simple mass spectra which resemble the spectra of the bases themselves. It is proposed that pyrolysis occurs by hydrogen transfer to the base moiety followed by elimination of the neutral base. Ionization by electron impact gives rise to the observed spectra. Mixed oligoribonucleotides pyrolyze readily to give spectra of adenine and uracil, but evidence for the presence of guanine and cytosine in mixed polymers and native RNA's is difficult to obtain presumably because of their low vapor pressure. The method may be useful for detection of modified bases in tRNA and for studies of temperature effects on RNA pyrolysis.

Inhibition oftRNA methylation in vitro and in whole cells by an oncostatic S-adenosyl-homocysteme (SAIl) analogue: 5′ deoxy 5′ -S-isobutyl adenosine (SIBA)

A high increase in the amount of methylated tRNA bases was found in vivo in Rous sarcoma virus infected and transformed chick embryo fibroblasts in comparison with normal cells, tRNA methylases extracted from transformed cells showed also higher activity in vitro with a heterologous substrate. 5'-deoxy-5'-S-isobutyl adenosine, (a structural analogue of S-adenosyl-L homocysteine), which inhibits virus-induced cell transformation, inhibits also the increase of incorporation of labelled methyl groups into tRNA in infected and transformed cells. When normal cells are grown in the presence of this inhibitor, undermethylated tRNAs are obtained. The effect of the drug is different in normal, infected and transformed cells. The methylation of the different bases is inhibited in vitro and in vivo to various extent. The effect of this substance on tRNA methylation may be the cause of its inhibitory effect on cell transformation.