Initiation of in vivo protein synthesis with non-methionine amino acids

Methionine is the universal amino acid for initiation of protein synthesis in all known organisms. The amino acid is coupled to a specific initiator methionine tRNA by methionyl-tRNA synthetase. In Escherichia coli, attachment of methionine to the initiator tRNA (tRNA(fMet)) has been shown to be dependent on synthetase recognition of the methionine anticodon CAU (complementary to the initiation codon AUG), [Schulman, L. H., & Pelka, H. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 6755-6759]. We show here that alteration of the anticodon of tRNA(fMet) to GAC or GAA leads to aminoacylation of the initiator tRNA with valine or phenylalanine. In addition, tRNA(fMet) carrying these amino acids initiates in vivo protein synthesis when provided with initiation codons complementary to the modified anticodons. These results indicate that the sequence of the anticodon of tRNA(fMet) dictates the identity of the amino acid attached to the initiator tRNA in vivo and that there are no subsequent steps which prevent initiation of E. coli protein synthesis by valine and phenylalanine. The methods described here also provide a convenient in vivo assay for further examination of the role of the anticodon in tRNA amino acid acceptor identity.     

Anticodon sequence mutants of Escherichia coli initiator tRNA: effects of overproduction of aminoacyl-tRNA synthetases,methionyl-tRNA formyltransferase,and initiation factor 2 on activity in initiation

Anticodon sequence mutants of Escherichia coli initiator tRNA initiate protein synthesis with codons other than AUG and amino acids other than methionine. Because the anticodon sequence is, in many cases, important for recognition of tRNAs by aminoacyl-tRNA synthetases, the mutant tRNAs are aminoacylated in vivo with different amino acids. The activity of a mutant tRNA in initiation in vivo depends on (i) the level of expression of the tRNA, (ii) the extent of aminoacylation of the tRNA, (iii) the extent of formylation of the aminoacyl-tRNA to formylaminoacyl-tRNA (fAA-tRNA), and (iv) the affinity of the fAA-tRNA for the initiation factor IF2 and the ribosome. Previously, using E. coli overproducing aminoacyl-tRNA synthetases, methionyl-tRNA formyltransferase, or IF2, we identified the steps limiting the activity in initiation of mutant tRNAs aminoacylated with glutamine and valine. Here, we have identified the steps limiting the activity of mutant tRNAs aminoacylated with isoleucine and phenylalanine. The combined results of experiments involving a variety of initiation codons (AUG, UAG, CAG, GUC, AUC, and UUC) provide support to the hypothesis that the ribosome.fAA-tRNA complex can act as an intermediate in initiation of protein synthesis. Comparison of binding affinities of various fAA-tRNAs (fMet-, fGln-, fVal-, fIle-, and fPhe-tRNAs) to IF2 using surface plasmon resonance supports the idea that IF2 can act as a carrier of fAA-tRNA to the ribosome. Other results suggest that the C1xA72 base pair mismatch, unique to eubacterial and organellar initiator tRNAs, may also be important for the binding of fAA-tRNA to IF2.     

Functionally impaired tRNA from ethionine treated rats as detected in injected Xenopus oocytes.

Treatment of rats with ethionine was found to cause severe impairment in the aminoacylation capacity of tRNA. This effect was only observed when assayed in injected oocytes, while invitro assays of aminoacylation failed to detect differences between normal tRNA and tRNA from ethionine treated animals. The effect of ethionine on the tRNA population was not uniform and differed for various amino acid specific tRNAs. Thus liver tRNA from ethionine treated rats showed a decreased capacity for phenylalanine aminoacylation, while no change was found in the case of leucine. On the other hand, the level of histidine aminoacylation was higher for tRNA from ethionine treated animals. An even more complex response was observed with methionine aminoacylation where tRNA from ethionine treated animals showed an initially faster rate than control tRNA. With more prolonged incubation periods, the methionyl-tRNA from ethionine treated animals was deacylated at an accelerated rate while the level of normal methionyl-tRNA remained almost constant.In addition to the aminoacylation reaction, the participation of aminoacyl-tRNA in protein synthesis was severely impaired. In this case, both the injected oocyte system and the cell-free wheat germ assay revealed these differences which were manifested with various mRNA and viral RNA preparations.     

Base substitutions in the wobble position of the anticodon inhibit aminoacylation of E. coli tRNAfMet by E. coli Met-tRNA synthetase

Derivatives of E. coli tRNAfMet containing single base substitutions at the wobble position of the anticodon have been enzymatically synthesized in vitro. The procedure involves excision of the normal anticodon, CAU, by limited digestion of intact tRNAfMet with RNase A. RNA ligase is then used to join each of four trinucleotides, NAU, to the 5' half molecule and to subsequently link the 3' and modified 5' fragments to regenerate the anticodon loop. Synthesis of intact tRNAfMet containing the anticodon CAU by this procedure yields a product which is indistinguishable from native tRNAfMet with respect to its ability to be aminoacylated by E. coli methionyl-tRNA synthetase. Substitution of any other nucleotide at the wobble position of tRNAfMet drastically impairs the ability of the synthetase to recognize the tRNA. Measurement of methionine acceptance in the presence of high concentrations of pure enzyme has established that the rate of aminoacylation of the AAU, GAU and UAU anticodon derivatives of tRNAfMet is four to five orders of magnitude slower than that of the native or synthesized tRNA containing C as the wobble base. In addition, the inactive tRNA derivatives fail to inhibit aminoacylation of normal tRNAfMet, indicating that they bind poorly to the enzyme. These results support a model involving direct interaction between Met-tRNA synthetase and the C in the wobble position during aminoacylation of tRNAfMet.     

Enzymatic tRNA acylation by acid and alpha-hydroxy acid analogues of amino acids

Incorporation of unnatural amino acids with unique chemical functionalities has proven to be a valuable tool for expansion of the functional repertoire and properties of proteins as well as for structure-function analysis. Incorporation of alpha-hydroxy acids (primary amino group is substituted with hydroxyl) leads to the synthesis of proteins with peptide bonds being substituted by ester bonds. Practical application of this modification is limited by the necessity to prepare corresponding acylated tRNA by chemical synthesis. We investigated the possibility of enzymatic incorporation of alpha-hydroxy acid and acid analogues (lacking amino group) of amino acids into tRNA using aminoacyl-tRNA synthetases (aaRSs). We studied direct acylation of tRNAs by alpha-hydroxy acid and acid analogues of amino acids and corresponding chemically synthesized analogues of aminoacyl-adenylates. Using adenylate analogues we were able to enzymatically acylate tRNA with amino acid analogues which were otherwise completely inactive in direct aminoacylation reaction, thus bypassing the natural mechanisms ensuring the selectivity of tRNA aminoacylation. Our results are the first demonstration that the use of synthetic aminoacyl-adenylates as substrates in tRNA aminoacylation reaction may provide a way for incorporation of unnatural amino acids into tRNA, and consequently into proteins.     

Editing of errors in selection of amino acids for protein synthesis.

All living cells must conduct protein synthesis with a high degree of accuracy maintained in the transmission and flow of information from gene to finished protein product. One crucial "quality control" point in maintaining a high level of accuracy is the selectivity by which aminoacyl-tRNA synthetases furnish correctly activated amino acids, attached to tRNA species, as the building blocks for growing protein chains. During selection of amino acids, synthetases very often have to distinguish the cognate substrate from a homolog having just one fewer methyl group in its structure. The binding energy of a methyl group is estimated to contribute only a factor of 100 to the specificity of binding, yet synthetases distinguish such closely related amino acids with a discrimination factor of 10,000 to 100,000. Examples of this include methionine versus homocysteine, isoleucine versus valine, alanine versus glycine, and threonine versus serine. Many investigators have demonstrated in vitro the ability of certain aminoacyl-tRNA synthetases to edit, that is, correct or prevent incorrect attachment of amino acids to tRNA molecules. Several major editing pathways are now established from in vitro data. Further, at least some aminoacyl-tRNA synthetases have recently been shown to carry out the editing function in vivo. Editing has been demonstrated to occur in both Escherichia coli and Saccharomyces cerevisiae. Significant energy is expended by the cell for editing of misactivated amino acids, which can be reflected in the growth rate. Because of this, cellular levels of aminoacyl-tRNA synthetases, as well as amino acid biosynthetic pathways which yield competing substrates for protein synthesis, must be carefully regulated to prevent excessive editing. High-level expression of recombinant proteins imposes a strain on the biosynthetic capacity of the cell which frequently results in misincorporation of abnormal or wrong amino acids owing in part to limited editing by synthetases. Unbalanced amino acid pools associated with some genetic disorders in humans may also lead to errors in tRNA aminoacylation. The availability of X-ray crystallographic structures of some synthetases, combined with site-directed mutagenesis, allows insights into molecular details of the extraordinary selectivity of synthetases, including the editing function.     

CHO cell mutants for arginyl-, asparagyl-, glutaminyl-, histidyl- and methionyl-transfer RNA synthetases: identification and initial characterization

A number of conditional lethal mutants of CHO cells that are defective in protein synthesis have been characterized with respect to their biochemical lesions. A defective aminoacyl-tRNA synthetase appears to be the basis of each mutant phenotype. In each strain, the specific activity in vitro of the synthetase cognate for one of the following amino acids was substantially reduced: arginine, asparagine, glutamine, histidine or methionine. One mutant, Arg-1, gave no detectable arginyl-tRNA synthetase activity, suggesting that it contains an altered enzyme that is unstable in vitro. Most of the mutants correspondingly exhibited impaired aminoacylation in vivo under nonpermissive conditions. However, two mutants, Arg-1 and His-1, appeared to have normal levels of acylated tRNA under the nonpermissive conditions which inhibited protein synthesis to approximately 50% and 10%, respectively. The expression of each mutant's phenotype, measured by rates of protein synthesis or growth, was a function of temperature and/or the concentration of amino acid cognate for the synthetase found to be deficient in vitro. The properties of these mutants make them applicable to diverse problems related to translation in mammalian cells.     

In Vivo Aminoacylation of Transfer Ribonucleic Acid in Bacillus subtilis and Evidence for Differential Utilization of Lysine-Isoaccepting Transfer Ribonucleic Acid Species

The presence or absence of certain amino acids has different effects on the ability of Bacillus subtilis to sporulate, and the intracellular pool size of amino acids has been reported to vary during sporulation. The idea that these variations might exert a regulatory effect through aminoacylation of transfer ribonucleic acid (tRNA) was investigated by studying the levels of aminoacylation in vivo in the logarithmic or stationary phase of growth. Both the periodate oxidation method and the amino acid analyzer were used to evaluate in vivo aminoacylation. The results indicated that in general the level of aminoacylation of tRNA's remained constant through stage III of sporulation, although there were detectable variations for specific amino acid groups. Our studies also showed that periodate oxidation damaged certain tRNA's; therefore, the results obtained by such a method should be interpreted with caution. Because the damage can affect certain isoaccepting species specifically, the periodate oxidation method cannot be used to establish which isoaccepting species are acylated in vivo. We also investigated the possibility of preferential use of particular tRNA species by polyribosomes. These results demonstrated a preferential use of lysyl-tRNA's at different growth stages. Control mechanisms operating during the early stages of sporulation, therefore, do not affect the overall level of aminoacylation. However, there is an effect on the levels of aminoacylation of specific amino acids and on which isoaccepting species are utilized by the polyribosome system.     

Distinctive acceptor-end structure and other determinants of Escherichia coli tRNAPro identity.

The previously uncharacterized determinants of the specificity of tRNAPro for aminoacylation (tRNAPro identity) were defined by a computer comparison of all Escherichia coli tRNA sequences and tested by a functional analysis of amber suppressor tRNAs in vivo. We determined the amino acid specificity of tRNA by sequencing a suppressed protein and the aminoacylation efficiency of tRNA by examining the steady-state level of aminoacyl-tRNA. On substituting nucleotides derived from the acceptor end and variable pocket of tRNAPro for the corresponding nucleotides in a tRNAPhe gene, the identity of the resulting tRNA changed substantially but incompletely to that of tRNAPro. The redesigned tRNAPhe was weakly active and aminoacyl-tRNA was not detected. Ethyl methanesulfonate mutagenesis of the redesigned tRNAPhe gene produced a mutant with a wobble pair in place of a base pair in the end of the acceptor-stem helix of the transcribed tRNA. This mutant exhibited both a tRNAPro identity and substantial aminoacyl-tRNA. The results speak for the importance of a distinctive conformation in the acceptor-stem helix of tRNAPro for aminoacylation by the prolyl-tRNA synthetase. The anticodon also contributes to tRNAPro identity but is not necessary in vivo.     

Alteration of the kinetic parameters for aminoacylation of Escherichia coli formylmethionine transfer RNA by modification of an anticodon base.

Treatment of Escherichia coli formylmethionine tRNA with 2 M sodium bisulfite, pH 7.0, in 10 mM MgCl2 at 25 degrees results in formation of uridine/bisulfite adducts at U18 in the dihydrouridine loop, U37 in the anticodon, and U48 in the variable loop. Two products, corresponding to the two diastereoisomers of 5,6-dihydrouridine-6-sulfonate, are formed at each reactive site in the tRNA. Although none of the modifications cause complete loss of methionine acceptor activity, the modified tRNA is amino-acylated at a reduced rate and has a decreased affinity for E. coli methionyl-tRNA synthetase. Aminoacylation of [35S]bisulfite-labeled tRNAfMet with a limiting amount of purified enzyme followed by separation of the acylated and unacylated molecules and structural analysis has shown that the presence of a specific diastereoisomer of the uridine/bisulfite adduct in the anticodon base U37 alters the kinetic parameters for aminoacylation of tRNAfMet.     

Importance of the anticodon sequence in the aminoacylation of tRNAs by methionyl-tRNA synthetase and by valyl-tRNA synthetase in an Archaebacterium

The mode of recognition of tRNAs by aminoacyl-tRNA synthetases and translation factors is largely unknown in archaebacteria. To study this process, we have cloned the wild type initiator tRNA gene from the moderate halophilic archaebacterium Haloferax volcanii and mutants derived from it into a plasmid capable of expressing the tRNA in these cells. Analysis of tRNAs in vivo show that the initiator tRNA is aminoacylated but is not formylated in H. volcanii. This result provides direct support for the notion that protein synthesis in archaebacteria is initiated with methionine and not with formylmethionine. We have analyzed the effect of two different mutations (CAU-->CUA and CAU-->GAC) in the anticodon sequence of the initiator tRNA on its recognition by the aminoacyl-tRNA synthetases in vivo. The CAU-->CUA mutant was not aminoacylated to any significant extent in vivo, suggesting the importance of the anticodon in aminoacylation of tRNA by methionyl-tRNA synthetase. This mutant initiator tRNA can, however, be aminoacylated in vitro by the Escherichia coli glutaminyl-tRNA synthetase, suggesting that the lack of aminoacylation is due to the absence in H. volcanii of a synthetase, which recognizes the mutant tRNA. Archaebacteria lack glutaminyl-tRNA synthetase and utilize a two-step pathway involving glutamyl-tRNA synthetase and glutamine amidotransferase to generate glutaminyl-tRNA. The lack of aminoacylation of the mutant tRNA indicates that this mutant tRNA is not a substrate for the H. volcanii glutamyl-tRNA synthetase. The CAU-->GAC anticodon mutant is most likely aminoacylated with valine in vivo. Thus, the anticodon plays an important role in the recognition of tRNA by at least two of the halobacterial aminoacyl-tRNA synthetases.     

Quality control in tRNA charging -- editing of homocysteine

All living organisms conduct protein synthesis with a high degree of accuracy maintained in the transmission and flow of information from a gene to protein product. One crucial 'quality control' point in maintaining a high level of accuracy is the selectivity by which aminoacyl-tRNA synthetases furnish correctly activated amino acids, attached to tRNA species, as the building blocks for growing protein chains. When differences in binding energies of amino acids to an aminoacyl-tRNA synthetase are inadequate, editing is used as a major determinant of enzyme selectivity. Some incorrect amino acids are edited at the active site before the transfer to tRNA (pre-transfer editing), while others are edited after transfer to tRNA at a separate editing site (post-transfer editing). Access of natural non-protein amino acids, such as homocysteine, homoserine, or ornithine to the genetic code is prevented by the editing function of aminoacyl-tRNA synthetases. Disabling editing function leads to tRNA mischarging errors and incorporation of incorrect amino acids into protein, which is detrimental to cell homeostasis and inhibits growth. Continuous homocysteine editing by methionyl-tRNA synthetase, resulting in the synthesis of homocysteine thiolactone, is part of the process of tRNA aminoacylation in living organisms, from bacteria to man. Excessive homocysteine thiolactone synthesis in hyperhomocysteinemia caused by genetic or nutritional deficiencies is linked to human vascular and neurological diseases.     

Effects of p-chlorophenylalanine and alpha-methylphenylalanine on amino acid uptake and protein synthesis in mouse neuroblastoma cells.

The phenylalanine analogues p-chlorophenylalanine and alpha-methylphenylalanine were used to inhibit phenylalanine hydroxylase in animal models for phenylketonuria. The present report examines the affects of these analogues on the metabolism of neuroblastoma cells. p-Chlorophenylalanine inhibited growth and was toxic to neuroblastoma cells. Although in vivo this analogue increased cell monoribosomes by 42%, it did not significantly affect poly(U)-directed protein synthesis in vitro. P-Chlorophenylalanine did not compete with phenylalanine or tyrosine for aminoacylation of tRNA and was therefore not substituted for those amino acids in nascent polypeptides. The initial cellular uptake of various large neutral amino acids was inhibited by this analogue but did not affect the flux of amino acids already in the cell; this suggested that an alteration of the cell's amino acid pools was not responsible for the cytotoxicity of the analogues. In contrast with p-chlorophenylalanine, alpha-methylphenylalanine did not exert these direct toxic effects because the administration of alpha-methylphenylalanine in vivo did not affect brain polyribosomes and a comparable concentration of this analogue was neither growth inhibitory nor cytotoxic to neuroblastoma cells in culture. The suitability of each analogue as an inhibitor of phenylalanine hydroxylase in animal models for phenylketonuria is discussed.     

Superspecificity of aminoacyl-tRNA-synthases

The correct aminoacylation of tRNA with the proper aminoacid by aminoacyl-tRNA synthetase is one of the key reactions which determines the overall high fidelity of protein biosynthesis. The initial selection of the amino acid is achieved in the active centre of the synthetase at the activation step due to differences in the side chains binding energies of specific substrate and the competing amino acids present in cell. If, nevertheless, the activation of amino acids structurally similar to the cognate one does proceed, additional mechanisms of correction which are based on the decomposition of unstable noncognate (intermediate or final) product of the tRNA aminoacylation reaction, by synthetase are switched on. In this review the literature on the specificity of aminoacyl-tRNA synthetases at amino acid activation step is analyzed along with the proofreading mechanisms which allow the elimination of the errors, leading to so called superspecifity of aminoacyl-tRNA synthetases.     

Conversion of a methionine initiator tRNA into a tryptophan-inserting elongator tRNA in vivo.

The role of the anticodon and discriminator base in aminoacylation of tRNAs with tryptophan has been explored using a recently developed in vivo assay based on initiation of protein synthesis by mischarged mutants of the Escherichia coli initiator tRNA. Substitution of the methionine anticodon CAU with the tryptophan anticodon CCA caused tRNA(fMet) to be aminoacylated with both methionine and tryptophan in vivo, as determined by analysis of the amino acids inserted by the mutant tRNA at the translational start site of a reporter protein containing a tryptophan initiation codon. Conversion of the discriminator base of tRNA(CCA)fMet from A73 to G73, the base present in tRNA(Trp), eliminated the in vivo methionine acceptor activity of the tRNA and resulted in complete charging with tryptophan. Single base changes in the anticodon of tRNA(CCA)fMet containing G73 from CCA to UCA, GCA, CAA, and CCG (changes underlined) essentially abolished tryptophan insertion, showing that all three anticodon bases specify the tryptophan identity of the tRNA. The important role of G73 in tryptophan identity was confirmed using mutants of an opal suppressor derivative of tRNA(Trp). Substitution of G73 with A73, C73, or U73 resulted in a large loss of the ability of the tRNA to suppress an opal stop codon in a reporter protein. Base pair substitutions at the first three positions of the acceptor stem of the suppressor tRNA caused 2-12-fold reductions in the efficiency of suppression without loss of specificity for aminoacylation of the tRNA with tryptophan.(ABSTRACT TRUNCATED AT 250 WORDS)     

Control of protein synthesis in mammalian cells by aminoacylation of transfer ribonucleic acid.

The dependence of protein synthesis on the intracellular content of aminoacylated tRNA has been studied in mouse ascites tumor cells deprived for various amino acids. A remarkable reduction in net protein synthesis has been found only after a drastic decrease in aminoacylation of tRNA. The quantitative correlation of protein synthesis with the degree of aminoacylation suggests that a moderate amino acid starvation primarily influences the rate of elongation at the codon concerned. These results are in contrast to the findings previously reported for HeLa cells. Some crucial steps during the determination of intracellular aminoacyl-tRNA have been investigated. The reliability of the method employed has been discussed on a theoretical basis.     

The effects of chronic hyperphenylalaninaemia on mouse brain protein synthesis can be prevented by other amino acids.

A prolonged elevation in the concentrations of circulating phenylalanine was maintained in newborn mice by daily injections of phenylalanine and a phenylalanine hydroxylase inhibitor, alpha-methylphenylalanine. The result of this chronic hyperphenylalaninaemia was an accumulation of vacant or inactive monoribosomes that persisted for 18 h of each day. An elongation assay in vitro with brain postmitochondrial supernatants demonstrated that, in addition, there was an equally prolonged decrease in the rates of polypeptide-chain elongation by the remaining brain polyribosomes. Analyses of the free amino acid composition in the brains of hyperphenylalaninaemic mice showed a loss of several amino acids from the brain, particularly the large, neutral amino acids, which are co- or counter-transported across plasma membranes with phenylalanine. When a mixture of these amino acids (leucine, isoleucine, valine, threonine, tryptophan, tyrosine, methionine) was injected into hyperphenylalaninaemic mice, there was an immediate cessation of monoribosome accumulation in the brain and there was no inhibition of the rates of polypeptide-chain elongation. Although the concentrations of the large, neutral amino acids in the brain were partially preserved by treatment of hyperphenylalaninaemic mice with the amino acid mixture, the elevated concentrations of phenylalanine remained unaltered. The amino acid mixture had no detectable effect on brain protein synthesis in the absence of the hyperphenylalaninaemic condition.     

Biochemical research on oogenesis. Transfer RNA is fully charged in the 42-S storage particles of Xenopus laevis oocytes.

1. Transfer RNA makes up 30-40% of total RNA in previtellogenic oocytes of Xenopus laevis. The bulk of tRNA is associated with 5-S RNA and two proteins in a high-molecular-weight complex sedimenting at 42S. 2. We show here that all kinds of tRNA are present in the 42-S particles and all of them sediment coincidently. Particle tRNA is fully charged in vivo. During purification of the 42-S particles tRNA becomes partially uncharged. When purified particles are incubated in vitro with amino acids and ATP a charging reaction occurs without disruption of the nucleoprotein complex. Many aminoacyl-tRNA synthetases can be shown to co-sediment with the 42-S particles. We conclude that complete aminoacylation of tRNA within the storage particles results from the activity of particle-bound aminoacyl-tRNA synthetases.     

Response of specific transfer-ribonucleic-acid levels to amino-acid deprivation in Friend leukemia cells.

In eukaryotes, the levels of specific tRNAs are closely correlated with the demands for their cognate amino acids in protein synthesis. To account for this phenomenon, we have proposed that the extent of aminoacylation of a given tRNA species in vivo controls the relative rate of synthesis or turnover of that species. Previously, we reported that Friend leukemia cells respond to histidine deprivation by increasing their relative level of tRNAHis by as much as two-fold, with no change in the relative level of tRNALeu. In this paper, we show that deprivation of leucine or tryptophan also causes a specific increase in the relative level of tRNAs cognate to the deprived amino acid. At least in the case of tRNATrp, the increases in relative tRNA levels are preceded by extensive declines in the steady-state extent of aminoacylation of the tRNA in vitro. We also find that different isoacceptors may respond differently to amino acid deprivation. These results suggest that decreased extents of aminoacylation of a given tRNA species in vivo cause increases in the relative rate of synthesis or decreases in the relative rate of degradation of that species.     

Initial position of aminoacylation of individual Escherichia coli, yeast, and calf liver transfer RNAs.

Transfer RNAs from Escherichia coli, yeast (Sacharomyces cerevisiae), and calf liver were subjected to controlled hydrolysis with venom exonuclease to remove 3'-terminal nucleotides, and then reconstructed successively with cytosine triphosphate (CTP) and 2'- or 3'-deoxyadenosine 5'-triphosphate in the presence of yeast CTP(ATP):tRNA nucleotidyltransferase. The modified tRNAs were purified by chromatography on DBAE-cellulose or acetylated DBAE-cellulose and then utilized in tRNA aminoacylation experiments in the presence of the homologous aminoacyl-tRNA synthetase activities. The E. coli, yeast, and calf liver aminoacyl-tRNA synthetases specific for alanine, glycine, histidine, lysine, serine, and threonine, as well as the E. coli and yeast prolyl-tRNA synthetases and the yeast glutaminyl-tRNA synthetase utilized only those homologous modified tRNAs terminating in 2'-deoxyadenosine (i.e., having an available 3'-OH group). This is interpreted as evidence that these aminoacyl-tRNA synthetases normally aminoacylate their unmodified cognate tRNAs on the 3'-OH group. The aminoacyl-tRNA synthetases from all three sources specific argining, isoleucine, leucine, phenylalanine, and valine, as well as the E. coli and yeast enzymes specific for methionine and the E. coli glutamyl-tRNA synthetase, used as substrates exclusively those tRNAs terminating in 3'-deoxyadenosine. Certain aminoacyl-tRNA synthetases, including the E. coli, yeast, and calf liver asparagine and tyrosine activating enzymes, the E. coli and yeast cysteinyl-tRNA synthetases, and the aspartyl-tRNA synthetase from yeast, utilized both isomeric tRNAs as substrates, although generally not at the same rate. While the calf liver aspartyl- and cysteinyl-tRNA synthetases utilized only the corresponding modified tRNA species terminating in 2'-deoxyadenosine, the use of a more concentrated enzyme preparation might well result in aminoacylation of the isomeric species. The one tRNA for which positional specificity does seem to have changed during evolution is tryptophan, whose E. coli aminoacyl-tRNA synthetase utilized predominantly the cognate tRNA terminating in 3'-deoxyadenosine, while the corresponding yeast and calf liver enzymes were found to utilize predominantly the isomeric tRNAs terminating in 2'-deoxyadenosine. The data presented indicate that while there is considerable diversity in the initial position of aminoacylation of individual tRNA isoacceptors derived from a single source, positional specificity has generally been conserved during the evolution from a prokaryotic to mammalian organism.