Mitochondrial tRNA 3' end metabolism and human disease
http://www.100md.com
《核酸研究医学期刊》
1 York College/CUNY, 94-20 Guy R. Brewer Boulevard, Jamaica, NY 11451, USA, 2 UPR 9002 du CNRS, IBMC 15 rue René Descartes, F-67084 Strasbourg, France and 3 Max-Planck-Institute for Evolutionary Anthropology, Deutscher Platz 6, D-04103 Leipzig, Germany
* To whom correspondence should be addressed. Tel: +1 718 262 2704; Fax: +1 718 262 2652; Email: louie@york.cuny.edu
ABSTRACT
Over 150 mutations in the mitochondrial genome have been shown to be associated with human disease. Remarkably, two-thirds of them are found in tRNA genes, which constitute only one-tenth of the mitochondrial genome. A total of 22 tRNAs punctuate the genome and are produced together with 11 mRNAs and 2 rRNAs from long polycistronic primary transcripts with almost no spacers. Pre-tRNAs thus require precise endonucleolytic excision. Furthermore, the CCA triplet which forms the 3' end of all tRNAs is not encoded, but must be synthesized by the CCA-adding enzyme after 3' end cleavage. Amino acid attachment to the CCA of mature tRNA is performed by aminoacyl-tRNA synthetases, which, like the preceding processing enzymes, are nuclear-encoded and imported into mitochondria. Here, we critically review the effectiveness and reliability of evidence obtained from reactions with in vitro transcripts that pathogenesis-associated mutant mitochondrial tRNAs can lead to deficiencies in tRNA 3' end metabolism (3' end cleavage, CCA addition and aminoacylation) toward an understanding of molecular mechanisms underlying human tRNA disorders. These defects probably contribute, individually and cumulatively, to the progression of human mitochondrial diseases.
INTRODUCTION
The human mitochondrial genome encodes 13 out of more than 80 polypeptide subunits of the mitochondrial respiratory chain complexes and contains 24 additional genes required for mitochondrial protein biosynthesis (1,2). These include 2 rRNA genes and 22 tRNA genes, one for each of 18 amino acids, and 2 each for tRNALeu (which read UUR and CUN codons) and tRNASer (which read UCN and AGY codons). The mitochondrial genome undergoes a 10- to 17-fold higher rate of mutation than the nuclear genome (3). Most mitochondrial mutations lead to polymorphisms (considered to be harmless), while a growing number correlate with various disorders. Since 1988 when the first pathogenesis-associated mutations in the mitochondrial genome were characterized (4,5), over 150 disease-correlated mutations have been found, about 100 of them located in tRNA genes . The tRNA genes make up only 10% of the mitochondrial genome (Figure 1A); these genes are thus ‘hot-spots’ for mutation-based mitochondrial pathogenesis.
Figure 1. (A) Human mitochondrial gene map. H with counterclockwise arrow and L with clockwise arrow designate direction of heavy and light DNA strand transcription. Outer ring is the light transcript, and inner ring is the heavy transcript. rRNAs and mRNAs are filled in with colors. Black bars designate tRNAs which punctuate the mitochondrial genome. (B) Heteroplasmy. Mixed (heteroplasmic) populations of wild-type and mutant mitochondrial genomes are present. Filled circles indicate mutant mitochondrial genomes and open circles indicate wild-type. (C) Thresholds. The thresholds for pathology are typically between 15 and 50% of mitochondrial tRNA function, affected by the extent of heteroplasmy (B) combined with effectiveness of the mutant tRNA in the series of reactions illustrated (Figure 2), in translation, and with its stability. A lower functional level would be lethal and a higher level would be without a phenotype.
Figure 2. The tRNA end processing pathway followed by aminoacylation. (A) tRNA is transcribed as a precursor, with a 5' end leader and a 3' end trailer. (B) RNase P has endonucleolytically cleaved the tRNA at +1. (C) tRNAse Z endonucleolytically cleaves the precursor on the 3' side of the discriminator base (N; +73). (D) CCA-adding enzyme (CCAse) adds CCA to the 3' end of the tRNA (N) produced by tRNase Z cleavage. (E) tRNA is charged with the cognate amino acid by a specific aminoacyl-tRNA synthetase (aaRS). Dashed line from CCA in (D) to tRNAse Z between (B) and (C) with an X through it indicates that 3'-CCA of mature tRNA is a tRNase Z anti-determinant (39).
The tRNA-linked disorders cover a wide range of symptoms and syndromes, including myopathies, encephalopathies, cardiopathies, deafness, ophthalmoplegia, diabetes and others. The relationship between mutations and disorders is complex; a single mutation can lead to different symptoms and different mutations to the same disorder. All these mitochondrial tRNA disorders more or less severely affect the energy metabolism of the cell. Efforts to understand the molecular mechanisms underlying their genotype/phenotype relationships have been made over the last 15 years; numerous reviews summarize the clinical and biological aspects (6–13) or the molecular mechanisms affected during the tRNA life cycles (14–18). Owing to their central function in protein synthesis, the malfunction of mutant tRNAs is thought to play a key role in the corresponding mitochondrial pathologies and has therefore been extensively investigated. Various aspects were explored, emphasizing aminoacylation properties and steady-state levels of aminoacyl-tRNA in mitochondria, but also including tRNA synthesis, structure, stability and interaction with partners of the translational machinery. However, so far only a limited number of mutations and of tRNAs have been explored, and for each, only one or a subset of possibilities were investigated. Neither general rules nor common features have emerged, and further studies are thus necessary.
Studies are complicated due to intrinsic properties of mitochondria and mitochondrial tRNAs (heteroplasmy, limited access to natural material and almost no possibility for genetic manipulation). Mitochondria are present in hundreds of copies per cell, each of them carrying more than one copy of DNA. A mitochondrial mutation can thus be heteroplasmic, varying in proportion to wild-type from 0 to 100% mutated DNA in the different tissues of an affected patient (Figure 1B). The threshold for effect of a mitochondrial mutation is usually between 15 and 50% of the normal functional level (referring to a combination of the proportion of the mutant tRNA in the heteroplasmic population with the effect of the mutation on processing, aminoacylation and function in translation; Figure 1C). Above the upper threshold, no pathology is observed, and below the lower threshold, the condition would be lethal. The heteroplasmic state also complicates analysis and interpretation of mutational effects. To minimize this problem, cybrid cells homoplasmic for a specific mutation can be produced by fusion of enucleated patient cells with a 0 cell line lacking mitochondria (19). Although such cells allowed the evaluation of several properties of mutated tRNAs (e.g. steady-state level of aminoacylation, stability and post-transcriptional modification), their construction is difficult and time consuming, so that only a handful of mutations have been analyzed so far. Moreover, the variety of cell types used to generate cybrids with different nuclear backgrounds (e.g. HeLa, osteosarcoma and lymphoblasts) does not allow straightforward comparisons (14). Ex vivo analyses, based on purification of a specific tRNA from cultured cells or from biopsies, are very powerful but have quantitative limitations (20,21).
The alternative approach, which avoids heteroplasmic effects in the study of pathological tRNAs and gives access to a sufficient amount of material, uses enzymes that are either recombinantly expressed or extracted from cultured cells without mitochondrial diseases, and tRNA substrates or precursors that are produced by in vitro transcription. This approach has recently made it possible to investigate the effects of several point mutations on various aspects of tRNA metabolism (Figure 2). Here, we review the defects in mitochondrial tRNA 3' end metabolism revealed by in vitro studies, including processing (cleavage of the 3'-trailer and CCA addition) and aminoacylation. Point mutations in mitochondrial tRNAs may contribute to tRNA malfunction by affecting the efficiency of one or a combination of these reactions. Advantages and limitations of the in vitro approach within the framework of investigating molecular mechanisms of human mitochondrial tRNA disorders will be discussed below.
tRNA 3' END METABOLISM
The 22 human mitochondrial tRNAs are synthesized from polycistronic precursor transcripts where they are interspersed among 11 mRNAs and two rRNAs (2,22,23), often with no spacers (Figure 1A). Table 1 provides details on the neighborhood of each tRNA gene. In about half the cases, tRNA excision from primary transcripts leads to mature mRNAs and rRNAs with proper ends, others are either flanked by a small (1–2 nt) spacer and several even overlap by one residue with the adjacent RNA, as in the case of tRNATyr/tRNACys. Therefore, the processing of mitochondrial tRNAs requires precise endonucleolytic cleavage at both 5' and 3' ends (Figure 2). Pre-tRNA 3' ends are subsequently completed by the addition of the CCA triplet, and a subset of nucleotides becomes post-transcriptionally modified. Finally, mature and correctly folded tRNAs are esterified at the 3' end with the cognate amino acid by the corresponding aminoacyl-tRNA synthetase. They are subsequently carried to the ribosome by translation factor , allowing protein synthesis. This brief overview illustrates the importance of 3' end metabolism for both tRNA synthesis and tRNA function.
Table 1. tRNA punctuation and 3' end sequence
Mutations in tRNAs may affect the reaction efficiency of the enzymes involved in tRNA 3' end metabolism in a direct way, or indirectly via changes in structural properties of the precursor or mature tRNA. Importantly, in this context, human mitochondrial tRNAs deviate structurally from classical tRNAs (24). Most extreme is tRNASer(AGY), in which the entire D-domain is missing, and an extended anticodon stem is present. Other tRNAs possess all expected secondary structural domains and can be folded into a cloverleaf. Large variations in the size of the D-loop, and especially T-loop (which is strictly maintained at 7 nt in canonical, nuclear-encoded tRNAs), are observed, and classically conserved or semi-conserved nucleotides (e.g. GG in the D-loop, UUC converted into TC in the T-loop), however, are absent. These common deviations from canonical secondary structure lead, in turn, to unpredictable tertiary interactions, suggesting greater flexibility of mitochondrial tRNAs. Owing to the nucleotide composition bias of the two mitochondrial DNA strands (1), 14 tRNAs (transcribed from the heavy strand; Figure 1A) are A-rich (36%) and G-poor (14%), and eight tRNAs (transcribed from the light strand) are U-rich (31.6%) and C-poor (16.4%), weakening the helical domains of tRNA secondary structures (24). These structural peculiarities may influence all aspects of the life-cycle of human mitochondrial tRNAs and contribute to the effects of pathogenic mutations.
CLEAVAGE OF tRNA 3'-TRAILERS
tRNA end processing
Excision of tRNAs from primary polycistronic mitochondrial transcripts is catalyzed by two specialized enzymes, RNase P and tRNase Z (the tRNA 3' end processing endonuclease, previously known as 3'-tRNase or RNase Z). Although not obligatory, a reaction order favoring cleavage by RNase P at the 5' end of a pre-tRNA followed by tRNase Z 3' end cleavage (Figure 2A–C) is observed in many systems, including higher plants (25), human mitochondria (26) and budding yeast (27). While RNase P efficiently cleaves substrates with long 3' end trailers, tRNase Z is apparently sensitive to the length of the 5'-leader (28). Experiments discussed below have therefore been performed on simplified model substrates.
Characterization of tRNase Z
Interestingly, the single fruit fly tRNase Z gene encodes both the nuclear and mitochondrial enzymes (29), and the ratio of nuclear/mitochondrial distribution in such cases is often 20:1 (A. Hopper, personal communication). tRNase Z RNAi knockdowns cause the accumulation of both nuclear-encoded and mitochondrial pre-tRNAs with 3' end extensions, but did not produce a phenotype or reduce the amount of mature cytoplasmic tRNA. The amount of mature mitochondrial tRNAIle and tRNATyr, however, was clearly reduced (29). The concentration of tRNase Z is thus more likely to be limiting in mitochondria than in nuclei, but issues of steady-state concentration and half-life of both the processing enzyme and the tRNA product must also be considered.
Human tRNase Z has also been cloned, recombinantly expressed and partially characterized, leading to the conclusion that these enzymes belong to the ELAC family (30). Both human homologs ELAC1 and ELAC2 (the short and long forms, respectively) possess the tRNase Z activity (31). It is unclear why humans and a number of other metazoans encode multiple forms of tRNase Z. Sorting servers (e.g. Mitoprot; http://ihg.gsf.de/ihg/mitoprot.html) predict that only the long form (ELAC2) carries an N-terminal mitochondrial targeting sequence, and ELAC2 has a lower KM and a wider substrate range than ELAC1 (32), consistent with ELAC2 function as mitochondrial tRNase Z. Since the human mitochondrial form of tRNase Z has not yet been proven, however, tRNase Z activity was obtained from HeLa mitoplast extracts for the analysis of mitochondrial tRNA 3' end processing (26,33,34).
ELAC2 was first characterized as a prostate cancer susceptibility gene (35). Two of the identified mutations which are weakly linked to prostate cancer are located in or close to the phosphodiesterase II domain, the active site of tRNase Z. ELAC2 is associated with mitotic spindles (36), and the RNAi knockdown of the Caenorhabditis elegans homolog of ELAC2 interferes with germline proliferation (37), but a possible function in mitochondrial tRNA processing, in which certain mutations increase the risk of human prostate cancer, is also worth considering.
CCA, a tRNase Z anti-determinant
The CCA triplet needed at the 3' end of all mature tRNAs is not mitochondrially encoded (rightmost column in Table 1), and must thus be added to the discriminator base (the last encoded 3' end nucleotide) by CCA-adding enzyme after 3' end processing. Mature in vitro transcribed tRNA with CCA at its 3' end is a tRNase Z anti-determinant (38,39); dashed line with an X through it; Figure 2DB,C; , being neither a substrate for nor an inhibitor of tRNase Z. This ensures that mature tRNA proceeds smoothly to aminoacylation without recycling through tRNase Z. Interestingly, a partial CCA terminus consisting of –CC is sufficient to produce the anti-determinant effect.
The CCA terminus is also a tRNase Z anti-determinant in Bacillus subtilis, in which the CCA triplet is encoded in some tRNA genes, while it must be added post-transcriptionally to the transcripts of others. Knockout of B.subtilis tRNase Z shows that most of the CCA-containing pre-tRNAs are processed by a 3'-exonuclease, while those without encoded CCA are processed endonucleolytically (41). In Thermotoga maritima, CCA is transcriptionally encoded at the 3' end of tRNA genes. Surprisingly, CCA is a cleavage site determinant in this organism, guiding tRNase Z to efficiently cleave downstream of the CCA triplet instead of after the discriminator base (42).
Defective tRNA end processing as a model for pathogenesis in vivo
The evidence which led to a model linking defective tRNA processing to pathogenesis was first obtained from studies of mitochondrial RNA in cells harboring disease-associated mutations in tRNALeu(UUR) (43,44). This tRNA gene is located between the genes for 16S rRNA and the ND1 subunit (Figure 1A). There is no spacer between the 3' end of 16S rRNA and position +1 of tRNALeu(UUR), and only a 2 nt spacer is present between the discriminator base of tRNALeu(UUR) and the initiation codon of the ND1 message (Table 1). An increase in the steady-state concentration of a discrete precursor (RNA 19) consisting of the linked 16S rRNA-tRNALeu(UUR)-ND1 RNA was observed with the A3243G and A3302G mutations (43,44), which are located in the D-loop and acceptor stem of the gene product, respectively (see Figure 3).
Figure 3. Mitochondrial tRNAs and pathogenesis-associated mutations analyzed for their effects on 3' end processing, CCA addition and/or aminoacylation. (A) tRNASer(UCN), (B) tRNAGly, (C) tRNAIle, (D) tRNALeu(UUR) and (E) tRNALys. tRNAs are represented in their secondary cloverleaf structures. Mutations detected so far in the corresponding genes are indicated by arrows, and numbering is according to the mitochondrial genome (1). Black arrows and numbers refer to pathology-related mutations for which one or more of the enzymatic activities described here have been studied (see Table 2); grey numbers and arrows identify those pathology-related mutations which have not yet been characterized as to their effects on any of these reactions (MITOMAP).
Table 2. Effects of mutations in human mitochondrial tRNAs on 3' end metabolism
Reduced tRNase Z processing efficiency of pathogenesis-associated tRNA precursors
The hypothesis that pathogenesis-associated mutations could cause defective tRNA end processing, thereby contributing to disease, was further investigated in vitro using precursors with both 5' and 3' end extensions (26,33,45). We have tested the hypothesis that pathogenesis-associated mutations in tRNAs could reduce the efficiency of precursor 3' end cleavage using wild-type and pathogenesis-associated mutations in tRNASer(UCN) (26), tRNAIle (46) and tRNALeu(UUR) (34) (Figure 3 and Table 2). According to this hypothesis, poor excision of tRNA could contribute to pathology. This would only be the case, however, if the intra-mitochondrial concentration of the processing enzyme is limiting.
The clearest example of a pathogenesis-associated mutation which could cause defective tRNA 3' end processing is the T7445C substitution in tRNASer(UCN), which causes non-syndromic deafness (47,48). T7445C falls just on the 3'-side of the discriminator base of tRNASer(UCN) (Figure 3), outside of the mature tRNA, suggesting that processing reactions depending on tRNA structure or sequence would be unaffected. The mutation, however, changes 3' end processing, leading to defective 3' ends (47). The cell lines homoplasmic for T7445C display approximately one-third of the steady-state level of tRNASer(UCN) observed in homoplasmic wild-type sibling cell lines (48). Interestingly, the T7445C substitution changes the sequence following the discriminator base (underlined) from G/UCU– to G/CCU–, a tRNase Z anti-determinant . The pathogenesis-associated T7445C tRNA precursor could not be cleaved in vitro, while wild-type tRNASer(UCN) could be processed efficiently (26). Additionally, tRNASer(UCN) with an extra G at the boundary between the V-loop and the T-stem (7472insG) seems to be processed and aminoacylated less efficiently than the normal tRNA (49,50). A possible rescue function for 3'-exonucleases in mitochondria has been discussed previously (26,50). In principle, any endonucleolytic cleavage which unlinks tRNASer(UCN) from the neighboring tRNATyr located 2.5 kb downstream (Figure 1A) could be followed by 3'-exonucleolytic cleavage to the discriminator.
Among trinucleotides found at the 3' end of human mitochondrial tRNAs following the discriminator base, /CCA is not observed, nor is /CC, and /C is found only once (in tRNACys; rightmost column in Table 1). In six cases , the trinucleotide following the discriminator base is a mitochondrial translation initiation codon (/AUG, /AUA, /AUU or /AAU). Genes for tRNAPhe and tRNAThr, like tRNALys, also have the trinucleotide AAU following the discriminator base, although the sequence immediately downstream does not encode a polypeptide in these cases. In conclusion, there is no natural tRNase Z anti-determinant in the sequences of 21 tRNA genes, and /C in tRNACys would not be expected to have much effect.
The kinetics of tRNase Z processing was performed using eight substitutions in tRNAIle associated with cardiomyopathy or ophthalmoplegia (46), and three in tRNALeu(UUR) correlated with mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes (MELAS) and assorted myopathies (34) (Figure 3). Four of the mutations (A4269G, A4295G, G4309A and A4317G) reduce the tRNase Z reaction efficiency up to 10-fold, and three tRNALeu(UUR) mutations decrease the efficiency up to 3-fold relative to wild type (Table 2), all due to lower Vmax. Interestingly, in tRNALeu(UUR), the reduction in processing efficiency increases with proximity to the 3'cleavage site (WT > A3243G > A3302G C3303U). These results support the hypothesis that 3' end processing defects are among the molecular events involving mutant mitochondrial tRNAs which contribute to mitochondrial tRNA disorders. Note, however, that there can be discrepancies between results obtained with wild-type and mutant tRNAs from cybrid cells and those from in vitro transcripts when it is possible to directly compare them. For example, the tRNAIle mutation A4269G was demonstrated to affect stability without any evident effect on maturation when the tRNAs were extracted from homoplasmic cybrid cells (51).
RNase P
This review is principally concerned with reactions at the 3' end of tRNA and mitochondrial pathogenesis, but for the sake of completeness, 5' end cleavage by RNase P should also be mentioned. Human mitochondrial RNase P activity was found by some to be indistinguishable from nuclear RNase P (52,53), while others found rat and human mitochondrial and nuclear RNase P to have different characteristics (33,54). The composition of human mitochondrial RNase P has not been completely resolved because it is difficult to purify mitochondria free of cytosolic or nuclear contamination (55). The number of RNase P molecules per mitochondrion required to process the mitochondrial transcripts could be very low (53), and if the intra-mitochondrial concentration of enzymes involved in tRNA metabolism is limiting, the modest reductions in processing efficiency arising from mutations in tRNA could lead to the accumulation of precursors and reduce the amount of functional mature tRNA.
SYNTHESIS OF THE CCA TERMINUS
Cloning, expression and characterization of tRNA nucleotidyltransferase
tRNA requires the base triplet CCA at the 3'-terminus for its aminoacylation and participation in protein synthesis. In mitochondrial tRNAs, this sequence must be added post-transcriptionally as an essential step in tRNA maturation (Figure 2). CCA addition is catalyzed by CCA-adding enzymes which are conserved in all organisms of all kingdoms. In humans, the mitochondrial, nuclear and cytosolic forms of this enzyme are encoded by a single gene located on chromosome 3 (56). A mitochondrial target sequence encoded in exon 1 ensures delivery of a proportion of the CCA-adding enzyme to mitochondria. Importantly, both the tRNase Z and CCA-adding enzyme are encoded in both forms by using two or three in-frame start codons separated by a short distance, with the mitochondrial targeting sequence being present only when translation initiates at the upstream AUG. Since CCA-adding enzymes are closely related to poly(A) polymerases, these enzymes were sometimes mistakenly identified by sequence analysis (57,58). The recombinant enzymes, however, can easily be distinguished based on their biochemical activity. The human CCA-adding enzyme was therefore cloned, recombinantly expressed and characterized (56,59). While showing kinetic properties comparable to those of other CCA-adding enzymes, the human enzyme has a specific affinity for mitochondrial tRNAs which lack some of the canonical elements, such as the TC-loop. The N-terminal domain of this enzyme contains the catalytic core as well as features such as nucleotide recognition/discrimination and specificity for tRNA (60,61). Recent domain swap experiments between the CCA-adding enzyme and poly(A) polymerase indicate that C-terminal domains also contribute to the discriminatory power of CCA addition (62).
Influence of pathology-related mutations on CCA addition
The activity of CCA-adding enzyme can be influenced by changes in sequence and/or structure of substrate tRNAs which can consequently contribute to pathogenicity, as first reported in certain disease-associated mutations in mitochondrial tRNAIle and tRNAGly (63) as well as tRNALeu(UUR) (34). In both tRNAIle and tRNAGly, an A was replaced by G in the TC-loop (A4317G in tRNAIle and A10044G in tRNAGly; Figure 3). These two domains combine to form the ‘elbow’ of the L-shaped three-dimensional (3D) structure of tRNA, a recognition element for CCA-adding enzymes (64–67). Structure probing experiments on in vitro transcripts carrying these replacements revealed that they lead to an aberrantly stable TC-arm or to a weakened interaction between the TC- and D-loop, respectively (63). In the kinetic analysis, KM values were only slightly affected, indicating that the enzyme's affinity for the mutated tRNAs was not reduced. The observed decrease in kcat suggests that the chemical step of catalysis is affected, possibly due to an inappropriate orientation of the tRNA substrate in the catalytic core of the enzyme (63). Besides tRNAIle and tRNAGly, five pathogenic mutations in tRNALeu(UUR) were investigated for their effects on CCA addition (Figure 3) (34). Only the substitution closest to the 3'-terminus, C3303T (causing cardiomyopathy), showed a significant effect on CCA addition (Table 2), reducing nucleotide incorporation by a factor of 5.5. While Vmax was unaffected, KM increased, indicating that C3303T reduces the affinity of the CCA-adding enzyme for tRNA without interfering with catalysis.
These examples show that pathogenesis-associated mutations which affect the same enzymatic reaction can do so by different means, either interfering with catalysis (as in A4317G in tRNAIle and A10044G in tRNAGly) or affecting the strength of enzyme–substrate interaction . Furthermore, the same mutation affects two different enzymes differently, causing a decrease in the Vmax of tRNase Z and an increase in the KM of CCA-adding enzyme (34).
AMINOACYLATION
Once the tRNA is synthesized (i.e. the steps of 5' and 3' end cleavage and CCA addition are finished as in Figure 2AD, post-transcriptional modifications are complete and 3D folding has taken place), the amino acid can be attached by the cognate aminoacyl-tRNA synthetase (aaRS; Figure 2DE). It can then be brought to the ribosome by transient binding to translation factors, allowing transfer of the amino acid to the growing polypeptide chain, thereby fulfilling its fundamental role in protein synthesis.
Mitochondrial aminoacyl-tRNA synthetases
Eukaryotic cells contain at least two sets of aminoacyl-tRNA synthetases, one devoted to cytoplasmic protein synthesis, the other to translation of the mitochondrial genetic information. In humans, cytosolic and mitochondrial enzymes are generally encoded by distinct genes, with the exception of GlyRS and LysRS . Enzymes targeted to mitochondria are synthesized with a mitochondrial signal sequence which is cleaved after import. Interestingly, mitochondrial synthetases recognize not only mitochondrial but also cytosolic tRNAs and tRNAs from other organisms; on the other hand, mitochondrial tRNAs are not recognized by non-cognate enzymes (69), probably due to their unusual structural properties.
Specific recognition of a tRNA by its cognate aaRS is based on the presence of ‘identity elements’, usually a set of bases or base pairs in the tRNA. Such elements have been defined precisely for a large number of bacterial, archaeal and lower eukaryotic tRNAs (70,71) but remain largely unknown for human mitochondrial tRNAs. Efficient aminoacylation also depends on structural features within the tRNA and on more global conformational properties.
Mutations in mitochondrial tRNAs reduce aminoacylation efficiency
To evaluate the impact of the mutations in pathology-related tRNAs on aminoacylation, wild-type and mutant mitochondrial tRNAs have been prepared by in vitro transcription and aminoacylated by the corresponding recombinantly expressed aminoacyl-tRNA synthetases. The first aminoacylations tested in such an in vitro system were those specific for serine and isoleucine. The serine system contributed to a fundamental understanding of tRNA/aaRS recognition, but pathology-related mutations were not analyzed (72,73). In contrast, 10 pathology-related mutations within tRNAIle associated with either ophthalmoplegia or cardiomyopathy were tested in the presence of a crude mitochondrial enzymatic extract (74) or with cloned and over-expressed IleRS (75,76) (Figure 3). Some mutations strongly reduce the ability of the tRNA to be aminoacylated, while others have only a very limited effect (Table 2). Mutations correlated with ophthalmoplegia reduced the isoleucylation efficiencies (as measured by the ratio kcat/KM compared to wild-type) from between 25- and 50-fold to more than 1000-fold. More modest reductions in aminoacylation efficiencies (or no effect) were reported for mutations correlated with cardiomyopathies (75,76). Mutants with compensatory mutations that restore disrupted base pairing regain wild-type aminoacylability. The structural effects of the mutations thus provide a stronger explanation for the reductions in aminoacylation efficiency than direct interference in tRNA/synthetase recognition. The analysis of tRNAIle variants also revealed that a mutant tRNA can potentially compete for the corresponding wild-type tRNA by inhibiting its aminoacylation. This situation could mimic the in vivo case, in which wild-type and mutant tRNAs co-exist in heteroplasmic patient cells (Figure 1B).
The use of in vitro transcripts leads to difficulties for investigation of aminoacylation properties of several human mitochondrial tRNAs other than tRNAIle and tRNASer because the corresponding wild-type transcripts do not fold into the expected cloverleaf secondary structures. The cases of tRNALys and tRNALeu(UUR) are of particular interest.
Wild-type tRNALys folds into an extended hairpin (77) due to the absence of a particular post-transcriptional modification, the N1-methyl group on adenine 9 (78). This methyl group prevents base pairing of nucleotides 9 and 64 (position 8355) from producing an extended acceptor stem in native tRNALys. The role of this methyl group has been mimicked in an in vitro transcript by mutating either nucleotide 9 or 64 to prevent them from base pairing (77). The corresponding ‘pseudo-wild-type’ transcripts fold into the cloverleaf structure and become active in aminoacylation with kinetic properties close to those of cytoplasmic human tRNALys (79). By second-site mutation of such ‘pseudo-wild-type’ transcripts, it became possible to investigate the effect on aminoacylation of all pathology-related mutations described so far in the mitochondrial tRNALys gene (79). Two (G8313A and G8328A) out of the eleven pathology-related mutations lead to a very strong loss (5000- to 7000-fold) in aminoacylation efficiency (kcat/KM); the other mutations had mild (2- to 10-fold losses in efficiency) or no effects. Thus, as in the case of tRNAIle, only a subset of mutations affects aminoacylation. The mechanisms by which the G8313A and G8328A mutations affect lysylation are unknown. Both mutations are located in the anticodon branch (anticodon domain and D-domain) on the same side on a model L-shaped 3D tRNA structure where they could directly interfere with LysRS interaction (79), while mutations without effect are located on the opposite side or in the amino acid accepting branch. Alternatively, these mutations could cause large structural rearrangements of the tRNA, leading indirectly to failure of synthetase recognition.
tRNALeu(UUR) has also been prepared by in vitro transcription. The cloned and purified human mitochondrial LeuRS accepts this transcript as a substrate at a 16- to 50-fold lower efficiency than native tRNALeu(UUR) extracted from cell lines (21,80). Here, the transcript only folds partially into a tRNA structure, despite the presence of sequence elements typical of classical tRNAs required for correct folding. Indeed, the amino acid acceptor stem and the TC-arm fold normally, but a large ‘floppy’ domain is observed instead of well-defined D and anticodon domains (34,80,81). Footprinting experiments suggest that the synthetase induces a structural change within the tRNA, leading to a productive interaction. Several mutations were analyzed to determine their effects on leucylation (Figure 3 and Table 2). For example, the effects of MELAS mutation A3243G and a second mutation at the same position correlated with encephalomyopathy (A3243T) lead to different effects (21,80). While mutation A3243G displays 10-fold reduced efficiency of aminoacylation, A3243T leads to a 300-fold decrease. The T3271C substitution in the anticodon stem has only a 3-fold negative effect, which can be reversed by a compensatory mutation (82). As was found for tRNAIle and tRNALys, mutations can affect aminoacylation efficiency either through kcat or through KM. Local structural perturbations were observed between some pathogenesis-associated mutations and wild-type tRNALeu(UUR) (80,82).
The structurally unorganized domain in tRNALeu(UUR) prepared by in vitro transcription, as opposed to the well-folded native tRNA extracted from cell lines (83), can be stabilized so as to more closely resemble a canonical structure. The site-directed mutagenesis that replaces C?A mismatches in either the D-stem or the anticodon stem by G-C pairs leads to cloverleaf structures, thus improving aminoacylation efficiency (81,83). These ‘pseudo-wild-types’ represent valuable reference molecules for a systematic investigation of pathology-related mutations on tRNALeu(UUR) functions.
CHANGE IN 3' END METABOLISM OF MITOCHONDRIAL tRNAs IS AN IMPORTANT CONSEQUENCE OF MUTATION
Pathology-related mutations affect 3' end metabolism
Analysis of transcribed wild-type or engineered ‘pseudo-wild-type’ substrates (tRNAs and tRNA precursors), or mutated derivatives, allow the effects of mutations to be screened in vitro. Mutations were found to affect any combination of the activities involved with 3' end metabolism of tRNAs (Figure 2B–E): cleavage of the 3'-trailer, CCA addition and aminoacylation. Although data summarized in Table 2 are not yet complete, it is clear that for many mutations, at least one of the three activities is significantly affected, making the 3' end of tRNAs a reasonable place to look for the effect of pathogenesis-associated mutations. Only three mutations have been tested for their effect on all three functions. Among these, mutation A3243G in tRNALeu(UUR) leads to a weak effect on 3' end cleavage (2.2-fold decrease in efficiency), no effect on CCA addition and a mild effect on aminoacylation (2- to 12-fold decrease in efficiency). The other two, A4317G in tRNAIle and C3303T in tRNALeu(UUR), substantially reduce the efficiency of 3' end cleavage and CCA addition but have little or no effect on aminoacylation. Nine mutations in tRNAIle have been tested for their effect on both 3' end cleavage and aminoacylation. For some of them (4269, 4295 and 4320), cleavage of the 3' end trailer of the primary tRNA transcript is reduced but aminoacylation is not affected; for others (4274, 4284 and 4300), 3' end maturation is unaffected, but reduced efficiency of aminoacylation is observed.
Whatever the affected activity, the end result is expected to be the same, a decrease in aminoacyl-tRNA available for protein synthesis. While the effect can differ at every level, an accumulation of small effects on several aspects of tRNA metabolism could lead to a strong global negative effect, sufficient to place the level of mitochondrial protein synthesis within the zone of pathology (Figure 1C). Indeed, limiting the amount of aminoacyl-tRNA, by any means, reduces the rate of translation (84,85). Such reductions have been observed in vivo in many cases involving pathogenesis-associated mitochondrial tRNAs (49,86–89). For some mutations, the 3' end metabolism was not affected (see Table 2); in these cases, other levels are presumably responsible for pathology, as suggested by the results of in vitro and in vivo studies by numerous investigators . The defects were observed in tRNA synthesis and maturation, including impaired termination of transcription, impaired RNase P processing and absence of post-transcriptional modifications (especially of the anticodon triplet, hindering codon reading). Post-aminoacylation functions of tRNA, including interaction with EF-Tu, codon reading or polysome formation, were sometimes affected. Decreased stability is the most common feature of mutant mitochondrial tRNAs associated with pathogenesis.
Structural aspects
As discussed above, the mechanisms by which mutations in tRNA affect the enzymatic activities of tRNA metabolism remain largely unknown. Mutations may interfere directly with recognition by the concerned enzyme, or change the overall structure enough to perturb the recognition. Owing to the intrinsic structural weakness of human mitochondrial tRNAs, the effect of mutations on tRNA structure may be amplified. Most of the pathology-associated mutations are conservative nucleotide transitions (90). When found in stems, canonical base pairs are often changed to weak G-U (U-G) wobble pairs or to C?A (A?C) mismatches (see Figure 3). In many cases, the effect on stem structure would be expected to be mild, because G-U and C?A can hydrogen bond with strength comparable to the A-U pair (91). Their structural effect could be magnified, however, since they are often found in regions which are already thermodynamically weak (18). Further illustrating this point, A4309G in the T-stem of tRNAIle displays a weakened secondary structure (46), and T3271C in the anticodon stem of tRNALeu(UUR) loses its negative effect on aminoacylation when combined with a stabilizing substitution in the D-stem (82).
EFFECTIVENESS AND CAVEATES CONCERNING USE OF IN VITRO TRANSCRIPTS AS TOOLS FOR SYSTEMATIC SCREENING
The results summarized here on the potential impact of mitochondrial tRNA mutations on 3' end metabolism were obtained using in vitro transcribed substrate tRNAs or tRNA precursors. The in vitro strategy of RNA preparation is of particular interest for human mitochondrial tRNAs since it allows the preparation of a large amount of any tRNA (either wild-type or mutant). Furthermore, in vitro kinetic analysis can lead to mechanistic insight. The structural flexibility of these tRNAs leads, in some cases, to the need to create ‘pseudo-wild-type’ molecules which have a single mutation introduced to stabilize structure and allow correct folding. Thoroughly documented in at least two cases to react in a manner comparable to that of the corresponding natural tRNAs extracted from cells, we believe they are relevant substitutes which allow useful analyses of second-site (pathogenesis-associated) mutations.
The lack of post-transcriptional modifications can be a drawback since such modifications can affect the structure and function of tRNA (92). Post-transcriptional modifications in classical tRNAs play either of two major roles, stabilizing tRNA structure and/or providing a recognition element for interacting partners such as aminoacyl-tRNA synthetases or mRNA codons. Therefore, aminoacylation properties of in vitro transcripts may be different from those of fully modified tRNAs in vivo. However, mutations introduced into the ‘pseudo-wild-type’ molecules are designed to stabilize the tRNA structure, replacing the structural role of the naturally occurring modifications to the extent that is feasible and allowing for efficient aminoacylation.
The in vitro approach is also valid when testing 3' end maturation activities such as tRNase Z activity, because this enzyme processes primary transcripts which may not yet be modified. Concerning CCA addition, this reaction can take place before all modifications are introduced into the tRNA (93). CCA-adding enzyme requires a properly folded tRNA acceptor stem to synthesize the CCA triplet; this domain has been shown to be correctly folded in the case of in vitro transcripts which have been analyzed for CCA addition . Modified nucleotides may thus be unnecessary for CCA addition and by inference, for the preceding tRNase Z reaction.
Significance of small effects of mutations on enzyme activities measured in vitro
Two major questions affect the interpretation of measured kinetic values. (i) To what extent do in vitro effects reflect the in vivo situation? (ii) Do small effects on reaction efficiency contribute to the expression of disorders? The first question is difficult to answer since in vitro assays take into account neither possible competition events nor the cellular environment, which the in vivo approach does. Moreover, one approach measures kinetic parameters of an enzymatic reaction leading to a value for efficiency, while the other measures the steady-state level of reaction products (e.g. aminoacyl-tRNA). These parameters cannot be compared directly, leading to apparent discrepancies between some of the in vitro and in vivo results. A valuable approach would be to measure reaction kinetics using in vitro wild-type and mutated transcripts and corresponding native wild-type and mutated tRNAs extracted from cybrid cell lines. In the case of tRNALys, results were the same (20,79). In the case of tRNALeu(UUR), however, results were significantly different (21). The native mutated tRNALeu(UUR) (A3243G) showed a 25-fold loss in aminoacylation efficiency when compared to the corresponding wild-type tRNA, while the mutated in vitro transcript only led to a 2.5- to 3-fold loss compared to the wild-type transcript. The transcripts can thus underestimate the impact of mutations; they can nonetheless usefully serve as indicators/detectors of the effects of mutation on aminoacylation.
Concerning the magnitude of effects and their relevance to pathology, Table 2 shows between 2.1- and 12.5-fold effects of mutations on cleavage of the 3'-trailer and at most a 5.5-fold decrease in efficiency of CCA addition. On the other hand, effects on aminoacylation properties of transcripts can be much greater, ranging from a 2-fold to a several 1000-fold decrease in efficiency. While several 1000-fold reductions in aminoacylation would be expected to contribute strongly to pathogenicity, even small reductions in individual maturation steps can contribute to the onset of disease, amplified by the cumulative effects of these impairments. Along these lines, the A4269G mutation in tRNAIle displays a 2-fold decrease in affinity for EF-Tu (94).
FUTURE PROSPECTS
Pathogenesis-associated tRNAs may contribute to mitochondrial disease due to individual effects or to accumulation of a number of deficiencies in end maturation, including 3' end cleavage by tRNase Z and CCA addition by tRNA nucleotidyltransferase, and in aminoacylation. Effects of the pathogenesis-associated substitutions on these reactions, analyzed using in vitro transcripts, does not preclude other events, including 5' end processing by RNase P, post-transcriptional modifications, elongation factor binding, ribosome binding, etc.
One promising direction is the analysis of nuclear-encoded enzymes involved in mitochondrial tRNA metabolism. In the case of enzymes encoded by single genes which are targeted to the mitochondria, cytoplasm and/or nucleus, it will be important to establish the relative enzyme concentrations in the individual compartments, which will greatly influence tRNA metabolism.
Virtually all pathogenesis-associated tRNA mutations can be investigated using in vitro transcripts. It will thus be possible to quantify and classify the effect(s) of individual mutations on metabolism as it relates to pathogenicity and to accumulate fundamental information concerning structure/function relations in human mitochondrial tRNAs. Although the in vitro strategy is not universally applicable, it is a valuable approach for unraveling many aspects of tRNA-based mitochondrial pathology before undertaking more costly in vivo analyses.
ACKNOWLEDGEMENTS
We gratefully acknowledge helpful conversations with M. Sissler and her help in the preparation of Figure 3. This work was supported by NIH grants F33GM64266, S06GM08153 and T34GM08498 (L.L.), by DFG grant Mo 634/2 and MPG (M.M.) and by CNRS, AFM, EC-QLG2-CT-1999-00660 (C.F.).
REFERENCES
Anderson,S., Bankier,A.T., Barrel,B.G., de Bruijn,M.H.L., Coulson,A.R., Drouin,J., Eperon,J.C., Nierlich,D.P., Roe,B.A., Sanger,F. et al. ( (1981) ) Sequence and organization of the human mitochondrial genome. Nature, , 290, , 457–465.
Montoya,J., Ojala,D. and Attardi,G. ( (1981) ) Distinctive features of the 5'-terminal sequences of the human mitochondrial mRNAs. Nature, , 290, , 465–470.
Pesole,G., Gissi,C., DeChirico,A. and Saccone,C. ( (1999) ) Nucleotide substitution rate of mammalian mitochondrial genomes. J. Mol. Evol., , 48, , 427–434.
Holt,I.J., Harding,A.E. and Morgan-Hughes,J.A. ( (1988) ) Deletions of muscle mitochondrial DNA in patients with mitochondrial myopathies. Nature, , 331, , 717–719.
Lestienne,P. and Ponsot,G. ( (1988) ) Kearns–Sayre syndrome with muscle mitochondrial DNA deletion. Lancet, , 1, , 885.
Larsson,N.-G. and Clayton,D.A. ( (1995) ) Molecular genetic aspects of human mitochondrial disorders. Annu. Rev. Genet., , 29, , 151–178.
Schon,E.A., Bonilla,E. and DiMauro,S. ( (1997) ) Mitochondrial DNA mutations and pathogenesis. J. Bioenerg. Biomembr., , 29, , 131–149.
DiMauro,S. and Schon,E. ( (1998) ) Nuclear power and mitochondrial disease. Nature Genet., , 19, , 214–215.
Wallace,D.C. ( (1999) ) Mitochondrial diseases in man and mouse. Science, , 283, , 1482–1488.
Chinnery,P.F. and Turnbull,D.M. ( (2000) ) Mitochondrial DNA mutations in the pathogenesis of human disease. Mol. Med. Today, , 6, , 425–432.
Shapira,A. ( (2000) ) Mitochondrial disorders. Curr. Opin. Neurol., , 13, , 527–532.
Dimauro,S. and Schon,E. ( (2001) ) Mitochondrial DNA mutations in human disease. Am. J. Med. Genet., , 106, , 18–26.
Betts,J., Lightowlers,R.N. and Turnbull,D. ( (2004) ) Neuropathological aspects of mitochondrial DNA disease. Neurochem. Res., , 29, , 505–511.
Jacobs,H.T. and Holt,I.J. ( (2000) ) The np 3243 MELAS mutation: damned if you aminoacylate, damned if you don't. Hum. Mol. Genet., , 9, , 463–465.
Florentz,C. ( (2002) ) Molecular investigations on tRNAs involved in human mitochondrial disorders. Biosci. Rep., , 22, , 81–98.
Jacobs,H.T. ( (2003) ) Disorders of mitochondrial protein synthesis. Hum. Mol. Genet., , 12, , R293–R301.
Florentz,C., Sohm,B., Tryoen-Tóth,P., Pütz,J. and Sissler,M. ( (2003) ) Human mitochondrial tRNAs in health and disease. Cell. Mol. Life Sci., , 60, , 1356–1375.
Wittenhagen,L.M. and Kelley,S.O. ( (2003) ) Impact of disease-related mitochondrial mutations on tRNA structure and function. Trends Biochem. Sci., , 28, , 605–611.
King,M.P. and Attardi,G. ( (1989) ) Human cells lacking mtDNA: repopulation with exogenous mitochondria by complementation. Science, , 246, , 500–503.
Yasukawa,T., Suzuki,T., Ishii,N., Ohta,S. and Watanabe,K. ( (2001) ) Wobble modification defect in tRNA disturbs codon-anticodon interaction in a mitochondrial disease. EMBO J., , 20, , 4794–4802.
Park,H., Davidson,E. and King,M.P. ( (2003) ) The pathogenic A3243G mutation in human mitochondrial tRNALeu(UUR) decreases the efficiency of aminoacylation. Biochemistry, , 4, , 958–64.
Ojala,D., Montoya,J. and Attardi,G. ( (1981) ) tRNA punctuation model of RNA processing in human mitochondria. Nature, , 290, , 470–474.
Chomyn,A., Cleeter,M.W., Ragan,C.I., Riley,M., Doolittle,R.F. and Attardi,G. ( (1986) ) URF6, last unidentified reading frame of human mtDNA, codes for an NADH dehydrogenase subunit. Science, , 234, , 614–618.
Helm,M., Brulé,H., Friede,D., Giegé,R., Pütz,J. and Florentz,C. ( (2000) ) Search for characteristic structural features of mammalian mitochondrial tRNAs. RNA, , 6, , 1356–1379.
Kunzmann,A., Brennicke,A. and Marchfelder,A. ( (1998) ) 5' end maturation and RNA editing have to preceed tRNA 3' processing in plant mitochondria. Proc. Natl Acad. Sci. USA, , 95, , 108–113.
Levinger,L., Jacobs,O. and James,M. ( (2001) ) In vitro 3' end endonucleolytic processing defect in a human mitochondrial tRNASer(UCN) precursor with the U7445C substitution, which causes non-syndromic deafness. Nucleic Acids Res., , 29, , 4334–4340.
Kuffel,J. and Tollervey,D. ( (2003) ) 3'-processing of yeast tRNATrp precedes 5'-processing. RNA, , 9, , 202–208.
Nashimoto,M., Wesemann,D.R., Geary,S., Tamura,M. and Kaspar,R.L. ( (1999) ) Long 5' leaders inhibit removal of a 3' trailer from a precursor tRNA by mammalian tRNA 3' processing endoribonuclease. Nucleic Acids Res., , 27, , 2770–2776.
Dubrovsky,E.B., Dubrovskaya,V.A., Levinger,L., Schiffer,S. and Marchfelder,A. ( (2004) ) Drosophila RNase Z processes mitochondrial and nuclear pre-tRNA 3' ends in vivo. Nucleic Acid Res., , 32, , 255–262.
Schiffer,S., Rosch,S. and Marchfelder,A. ( (2002) ) Assigning a function to a conserved group of proteins: the tRNA 3'-processing enzymes. EMBO J., , 21, , 2769–2777.
Takaku,H., Minegawa,A., Takagi,M. and Nashimoto,M. ( (2003) ) A candidate prostate cancer susceptibility gene encodes tRNA 3' processing endoribonuclease. Nucleic Acid Res., , 31, , 2272–2278.
Takaku,H., Minegawa,A., Takagi,M. and Nashimoto,M. ( (2004) ) T N-terminal half domain of the long form of tRNase Z is required for the RNase 65 activity. Nucleic Acid Res., , 32, , 4429–4438.
Rossmanith,W., Tullo,A., Potuschak,T., Karwan,R. and Sbisa,E. ( (1995) ) Human mitochondrial tRNA processing. J. Biol. Chem., , 270, , 12885–12891.
Levinger,L., Oestereich,I., Florentz,C. and M?rl,M. ( (2004) ) A pathogenesis-associated mutation in human mitochondrial tRNALeu(UUR) leads to reduced 3'-end processing and CCA addition. J. Mol. Biol., , 337, , 535–544.
Tavtigian,S.V., Simard,J., Teng,D.H., Abtin,V., Baumgard,M., Beck,A., Camp,N.J., Carillo,A.R., Chen,Y., Dayananth,P. et al. ( (2001) ) A candidate prostate cancer susceptibility gene at chromosome 17p. Nature Genet., , 27, , 172–180.
Korver,W., Guevara,C., Chen,Y., Neutoboom,S., Bookstein,R., Tavtigian,S. and Lees,E. ( (2003) ) The product of the candidate prostate cancer susceptibility gene ELAC2 interacts with the gamma-tubulin complex. Int. J. Cancer, , 104, , 283–288.
Smith,M.M. and Levitan,D.J. ( (2004) ) The Caenorhabditis elegans homolog of the putative prostate cancer susceptibility gene ELAC2, hoe-1, plays a role in germline proliferation. Dev. Biol., , 266, , 151–160.
Nashimoto,M. ( (1997) ) Distribution of both length and 5' terminal nucleotides of mammalian pre-tRNA 3' trailers reflect properties of 3' processing endoribonuclase. Nucleic Acids Res., , 25, , 1148–1154.
Mohan,A., Whyte,S., Wang,X., Nashimoto,M. and Levinger,L. ( (1999) ) The 3' end CCA of mature tRNA is an antideterminant for eukaryotic 3'-tRNase. RNA, , 5, , 245–256.
Schiffer,S., Rosch,S. and Marchfelder,A. ( (2003) ) Recombinant RNase Z does not recognize CCA as part of the tRNA and its cleavage efficiency is influenced by acceptor stem length. Biol. Chem., , 384, , 333–342.
Pelligrini,O., Nezzar,J., Marchfelder,A., Putzer,H. and Condon,C. ( (2003) ) Endonucleolytic processing of CCA-less tRNA precursors by RNase Z in Bacillus subtilis. EMBO J., , 22, , 4534–4543.
Minegawa,A., Takaku,H., Takagi,M. and Nashimoto,M. ( (2004) ) A novel endonucleolytic mechanism to generate the CCA 3'-termini of tRNA molecules in Thermotoga maritima. J. Biol. Chem., , 279, , 15688–15697.
King,M.P., Koga,Y., Davidson,M. and Schon,E.A. ( (1992) ) Defects in mitochondrial protein synthesis and respiratory chain activity segregate with the tRNALeu(UUR) mutation associated with mitochondrial myopathy, encephalopathy, lactic acidosis and strokelike episodes. Mol. Cell. Biol., , 12, , 480–490.
Bindoff,L.A., Howell,N., Poulton,J., McCullough,D.A., Morten,K.J., Lightowlers,R.N., Turnbull,D.M. and Weber,K. ( (1993) ) Abnormal RNA processing associated with a novel tRNA mutation in mitochondrial DNA. A potential disease mechanism. J. Biol. Chem., , 268, , 19559–19564.
Rossmanith,W. and Karwan,R.M. ( (1998) ) Impairment of tRNA processing by point mutations in mitochondrial tRNALeu(UUR) associated with mitochondrial diseases. FEBS Lett., , 433, , 269–274.
Levinger,L., Giegé,R. and Florentz,C. ( (2003) ) Pathology-related substitutions in human mitochondrial tRNAIle reduce precursor 3'-end processing efficiency in vitro. Nucleic Acids Res., , 31, , 1904–1912.
Reid,F., Rovio,A.T., Holt,I.J. and Jacobs,H.T. ( (1997) ) Molecular phenotype of a human lymphoblastoid cell-line homoplasmic for the np-7445 deafness-associated mitochondrial mutation. Hum. Mol. Genet., , 6, , 443–449.
Guan,M.-X., Enriquez,J.A., Fishel-Ghodsian,N., Puranam,R.S., Lin,C.P., Maw,M.A. and Attardi,G. ( (1998) ) The deafness-associated mitochondrial DNA mutation at position 7445, which affects tRNASer(UCN) precursor processing, has long-range effects on NADH dehydrogenase subunit ND6 gene expression. Mol. Cell. Biol., , 18, , 5868–5879.
Toompuu,M., Yasukawa,T., Suzuki,T., Hakkinen,T., Spelbrink,J., Watanabe,K. and Jacobs,H. ( (2002) ) The 7472insC mitochondrial DNA mutation impairs the synthesis and extent of aminoacylation of tRNASer(UCN) but not its structure or rate of turnover. J. Biol. Chem., , 22240–22250.
Toompuu,M., Levinger,L., Nadal,A. and Jacobs,H. ( (2004) ) The 7472insG mutation impairs 5' and 3' processing of tRNASer(UCN). Biochem. Biophys. Res. Commun., , 322, , 803–813.
Yasukawa,T., Hino,N., Suzuki,T., Watanabe,K., Ueda,T. and Ohta,S. ( (2000) ) A pathogenic point mutation reduces stability of mitochondrial mutant tRNAIle. Nucleic Acids Res., , 28, , 3779–3784.
Doersen,C.J., Guerrier-Takada,C., Altman,S. and Attardi,G. ( (1985) ) Characterization of an RNase P activity from HeLa cell mitochondria. Comparison with the cytosol RNase P activity. J. Biol. Chem., , 260, , 5942–5949.
Puranam,R.S. and Attardi,G. ( (2001) ) The RNase P associated with HeLa cell mitochondria contains an essantial RNA component identical in sequence to that of the nuclear RNase P. Mol. Cell. Biol., , 21, , 548–561.
Manam,S. and Van Tuyle,G.C. ( (1987) ) Separation and characterization of 5'- and 3'-tRNA processing nucleases from rat liver mitochondria. J. Biol. Chem., , 262, , 10272–10279.
Rossmanith,W. and Potuschak,T. ( (2001) ) Difference between Mitochondrial RNase P and Nuclear RNase P. Mol. Cell. Biol., , 21, , 8236–8237.
Reichert,A., Thurlow,D. and M?rl,M. ( (2001) ) A eubacterial origin for the human tRNA nucleotidyltransferase? Biol. Chem., , 382, , 1431–1438.
Raynal,L., Krisch,H. and Carpousis,A. ( (1998) ) The Bacillus subtilis nucleotidyltransferase is a tRNA CCA-adding enzyme. J. Bacteriol., , 180, , 6276–6282.
Sohlberg,B., Huang,J. and Cohen,S. ( (2003) ) The Streptomyces coelicolor polynucleotide phosphorylase homologue, and not the putative poly(A) polymerase, can polyadenylate RNA. J. Bacteriol., , 185, , 7273–7278.
Nagaike,T., Suzuki,T., Tomari,Y., Takemoto-Hori,C., Negayama,F., Watanabe,K. and Ueda,T. ( (2001) ) Identification and characterization of mammalian mitochondrial tRNA nucleotidyltransferase. J. Biol. Chem., , 276, , 40041–40049.
Li,F., Xiong,Y., Wang,J., Cho,H., Tomita,K., Weiner,A. and Steitz,T. ( (2002) ) Crystal structures of the Bacillus stearothermophilus CCA-adding enzyme and its complexes with ATP or CTP. Cell, , 111, , 815–824.
Augustin,M., Reichert,A., Betat,H., Huber,R., Morl,M. and Steegborn,C. ( (2003) ) Crystal structure of the human CCA-adding enzyme: insights into template-independent polymerization. J. Mol. Biol., , 328, , 985–994.
Betat,H., Rammelt,C., Martin,G. and Moerl,M. ( (2004) ) Exchange of regions between bacterial poly(A) polymerase and CCA adding enzyme generates altered specificities. Mol. Cell, , 15, , 389–398.
Tomari,Y., Hino,N., Nagaike,T., Suzuki,T. and Ueda,T. ( (2003) ) Decreased CCA-addition in human mitochondrial tRNAs bearing a pathogenic A4317G or A10044G mutation. J. Biol. Chem., , 278, , 16828–16833.
Spacciapoli,P., Doviken,L., Mulero,J.J. and Thurlow,D.L. ( (1989) ) Recognition of tRNA by the enzyme ATP/CTP:tRNA nucleotidyl transferase. J. Biol. Chem., , 264, , 3799–3805.
Spacciapoli,P. and Thurlow,D.L. ( (1990) ) Purines in tRNAs required for recognition by ATP/CTP:tRNA nucleotidyltransferase from rabbit liver. J. Mol. Recognit., , 3, , 149–155.
Hegg,L.A. and Thurlow,D.L. ( (1990) ) Cytidines in tRNAs that are required intact by ATP/CTP:tRNA nucleotidyltransferases from Escherichia coli and Saccharomyces cerevisiae. Nucleic Acids Res., , 18, , 5975–5979.
Shi,P.-Y., Maizels,N. and Weiner,A.M. ( (1998) ) CCA addition by tRNA nucleotidyltransferase: polymerization without translocation? EMBO J., , 17, , 3197–3206.
Sissler,M., Pütz,J., Fasiolo,F. and Florentz,C. ( (2004) ) Chapter 28. In Ibba,M., Francklyn,C. and Cusack,S. (eds), The Aminoacyl-tRNA Synthetases. Landes Biosciences, Georgetown, TX.
Kumazawa,Y., Himeno,H., Miura,K.-I. and Watanabe,K. ( (1991) ) Unilateral aminoacylation specificity between bovine mitochondria and eubacteria. J. Biochem., , 109, , 421–427.
Giegé,R., Sissler,M. and Florentz,C. ( (1998) ) Universal rules and idiosyncratic features in tRNA identity. Nucleic Acids Res., , 26, , 5017–5035.
Giegé,R. and Frugier,M. ( (2003) ) In Lapointe,J. and Brakier-Gringas,L. (eds), Translation Mechanisms. Landes Sciences, Georgetown, TX, pp. 1–24.
Yokogawa,T., Shimada,N., Takeuchi,N., Benkowski,L., Suzuki,T., Omori,A., Ueda,T., Nishikawa,K., Spremulli,L.L. and Watanabe,K. ( (2000) ) Characterization and tRNA recognition of mammalian mitochondrial seryl-tRNA synthetase. J. Biol. Chem., , 275, , 19913–19920.
Shimada,N., Suzuki,T. and Watanabe,K. ( (2001) ) Dual mode of recognition of two isoacceptor tRNAs by mammalian mitochondrial seryl-tRNA synthetase. J. Biol. Chem., , 276, , 46770–46778.
Degoul,F., Brulé,H., Cepanec,C., Helm,M., Marsac,C., Leroux,J.-P., Giegé,R. and Florentz,C. ( (1998) ) Isoleucylation properties of native human mitochondrial tRNAIle and tRNAIle transcripts. Implications for cardiomyopathy-related point mutations (4269, 4317) in the tRNAIle gene. Hum. Mol. Gen., , 7, , 347–354.
Kelley,S., Steinberg,S. and Schimmel,P. ( (2000) ) Functional defects of pathogenic human mitochondrial tRNAs related to structural fragility. Nature Struct. Biol., , 7, , 862–865.
Kelley,S.O., Steinberg,S.V. and Schimmel,P. ( (2001) ) Fragile T-stem in disease-associated human mitochondrial tRNA sensitizes structure to local and distant mutations. J. Biol. Chem., , 276, , 10607–10611.
Helm,M., Brulé,H., Degoul,F., Cepanec,C., Leroux,J.-P., Giegé,R. and Florentz,C. ( (1998) ) The presence of modified nucleotides is required for cloverleaf folding of a human mitochondrial tRNA. Nucleic Acids Res., , 26, , 1636–1643.
Helm,M., Giegé,R. and Florentz,C. ( (1999) ) A Watson–Crick base-pair-disrupting methyl group (m1A9) is sufficient for cloverleaf folding of human mitochondrial tRNALys. Biochemistry, , 38, , 13338–13346.
Sissler,M., Helm,M., Frugier,M., Giegé,R. and Florentz,C. ( (2004) ) Aminoacylation properties of pathology-related variants of human mitochondrial tRNALys variants. RNA, , 10, , 841–853.
Sohm,B., Frugier,M., Brulé,H., Olszak,K., Przykorska,A. and Florentz,C. ( (2003) ) Towards understanding human mitochondrial leucine aminoacylation identity. J. Mol. Biol., , 328, , 995–1010.
Roy,M.D., Wittenhagen,L.M., Vozzella,B.E. and Kelley,S.O. ( (2004) ) Interdomain communication between weak structural elements within a disease-related human tRNA. Biochemistry, , 43, , 384–392.
Wittenhagen,L.M., Roy,M.D. and Kelley,S.O. ( (2003) ) The pathogenic U3271C human mitochondrial tRNALeu(UUR) mutation disrupts a fragile anticodon stem. Nucleic Acid Res., , 31, , 596–601.
Sohm,B., Sissler,M., Park,H., King,M.P. and Florentz,C. ( (2004) ) Recognition of human mitochondrial tRNALeu(UUR) by its cognate leucyl-tRNA synthetase. J. Mol. Biol., , 339, , 17–29.
Brandsma,M., Janssen,M. and Moller,W. ( (1997) ) Termination of Quiescence in Crustacea. The role of transfer RNA aminoacylation in the brine shrimp Artemia. J. Biol. Chem., , 272, , 28912–28917.
Sorensen,M.A. ( (2001) ) Charging levels of four tRNA species in Escherichia coli Rel(+) and Rel(–) strains during amino acid starvation: a simple model for the effect of ppGpp on translational accuracy. J. Mol. Biol., , 307, , 785–798.
Enriquez,J.A., Chomyn,A. and Attardi,G. ( (1995) ) MtDNA mutation in MERRF syndrome causes defective aminoacylation of tRNALys and premature translation termination. Nature Genet., , 10, , 47–55.
Hao,H. and Moraes,C.T. ( (1997) ) A disease-associated G5703A mutation in human mitochondrial DNA causes a conformational change and a marked decrease in steady-state levels of mitochondrial tRNAAsn. Mol. Cell. Biol., , 17, , 6831–6837.
B?rner,G.V., Zeviani,M., Tiranti,V., Carrara,F., Hoffmann,S., Gerbitz,K.D., Lochmuller,H., Pongratz,D., Klopstock,T., Melberg,A. et al. ( (2000) ) Decreased aminoacylation of mutant tRNAs in MELAS but not in MERRF patients. Hum. Mol. Genet., , 9, , 467–475.
Chomyn,A., Enriquez,J.A., Micol,V., Fernandez-Silva,P. and Attardi,G. ( (2000) ) The mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episode syndrome-associated human mitochondrial tRNALeu(UUR) mutation causes aminoacylation deficiency and concomitant reduced association of mRNA with ribosomes. J. Biol. Chem., , 275, , 19198–19209.
Florentz,C. and Sissler,M. ( (2001) ) Disease-related versus polymorphic mutations in human mitochondrial tRNAs: where is the difference? EMBO Rep., , 2, , 481–486.
Leontis,N.B. and Westhof,E. ( (1998) ) Conserved geometrical base-pairing patterns in RNA. Q. Rev. Biophys., , 31, , 399–455.
Agris,P.F. ( (1996) ) The importance of being modified: roles of modified nucleosides and Mg2+ in RNA structure and function. Prog. Nucleic Acid Res. Mol. Biol., , 53, , 79–129.
M?rl,M., D?rner,M. and P??bo,S. ( (1995) ) C to U editing and modifications during the maturation of the mitochondrial tRNAAsp in marsupials. Nucleic Acids Res., , 23, , 3380–3384.
Hino,N., Suzuki,T., Yasukawa,T., Seio,K., Watanabe,K. and Ueda,T. ( (2004) ) The pathogenic A4269G mutation in human mitochondrial tRNAIle alters the T-stem structure and decreases the binding affinity for elongation factor Tu. Genes Cells, , 9, , 243–252.
Wittenhagen,L.M. and Kelley,S.O. ( (2002) ) Dimerization of a pathogenic human mitochondrial tRNA. Nature Struct. Biol., , 9, , 586–590.(Louis Levinger1,2,*, Mario M?rl3 and Cat)
* To whom correspondence should be addressed. Tel: +1 718 262 2704; Fax: +1 718 262 2652; Email: louie@york.cuny.edu
ABSTRACT
Over 150 mutations in the mitochondrial genome have been shown to be associated with human disease. Remarkably, two-thirds of them are found in tRNA genes, which constitute only one-tenth of the mitochondrial genome. A total of 22 tRNAs punctuate the genome and are produced together with 11 mRNAs and 2 rRNAs from long polycistronic primary transcripts with almost no spacers. Pre-tRNAs thus require precise endonucleolytic excision. Furthermore, the CCA triplet which forms the 3' end of all tRNAs is not encoded, but must be synthesized by the CCA-adding enzyme after 3' end cleavage. Amino acid attachment to the CCA of mature tRNA is performed by aminoacyl-tRNA synthetases, which, like the preceding processing enzymes, are nuclear-encoded and imported into mitochondria. Here, we critically review the effectiveness and reliability of evidence obtained from reactions with in vitro transcripts that pathogenesis-associated mutant mitochondrial tRNAs can lead to deficiencies in tRNA 3' end metabolism (3' end cleavage, CCA addition and aminoacylation) toward an understanding of molecular mechanisms underlying human tRNA disorders. These defects probably contribute, individually and cumulatively, to the progression of human mitochondrial diseases.
INTRODUCTION
The human mitochondrial genome encodes 13 out of more than 80 polypeptide subunits of the mitochondrial respiratory chain complexes and contains 24 additional genes required for mitochondrial protein biosynthesis (1,2). These include 2 rRNA genes and 22 tRNA genes, one for each of 18 amino acids, and 2 each for tRNALeu (which read UUR and CUN codons) and tRNASer (which read UCN and AGY codons). The mitochondrial genome undergoes a 10- to 17-fold higher rate of mutation than the nuclear genome (3). Most mitochondrial mutations lead to polymorphisms (considered to be harmless), while a growing number correlate with various disorders. Since 1988 when the first pathogenesis-associated mutations in the mitochondrial genome were characterized (4,5), over 150 disease-correlated mutations have been found, about 100 of them located in tRNA genes . The tRNA genes make up only 10% of the mitochondrial genome (Figure 1A); these genes are thus ‘hot-spots’ for mutation-based mitochondrial pathogenesis.
Figure 1. (A) Human mitochondrial gene map. H with counterclockwise arrow and L with clockwise arrow designate direction of heavy and light DNA strand transcription. Outer ring is the light transcript, and inner ring is the heavy transcript. rRNAs and mRNAs are filled in with colors. Black bars designate tRNAs which punctuate the mitochondrial genome. (B) Heteroplasmy. Mixed (heteroplasmic) populations of wild-type and mutant mitochondrial genomes are present. Filled circles indicate mutant mitochondrial genomes and open circles indicate wild-type. (C) Thresholds. The thresholds for pathology are typically between 15 and 50% of mitochondrial tRNA function, affected by the extent of heteroplasmy (B) combined with effectiveness of the mutant tRNA in the series of reactions illustrated (Figure 2), in translation, and with its stability. A lower functional level would be lethal and a higher level would be without a phenotype.
Figure 2. The tRNA end processing pathway followed by aminoacylation. (A) tRNA is transcribed as a precursor, with a 5' end leader and a 3' end trailer. (B) RNase P has endonucleolytically cleaved the tRNA at +1. (C) tRNAse Z endonucleolytically cleaves the precursor on the 3' side of the discriminator base (N; +73). (D) CCA-adding enzyme (CCAse) adds CCA to the 3' end of the tRNA (N) produced by tRNase Z cleavage. (E) tRNA is charged with the cognate amino acid by a specific aminoacyl-tRNA synthetase (aaRS). Dashed line from CCA in (D) to tRNAse Z between (B) and (C) with an X through it indicates that 3'-CCA of mature tRNA is a tRNase Z anti-determinant (39).
The tRNA-linked disorders cover a wide range of symptoms and syndromes, including myopathies, encephalopathies, cardiopathies, deafness, ophthalmoplegia, diabetes and others. The relationship between mutations and disorders is complex; a single mutation can lead to different symptoms and different mutations to the same disorder. All these mitochondrial tRNA disorders more or less severely affect the energy metabolism of the cell. Efforts to understand the molecular mechanisms underlying their genotype/phenotype relationships have been made over the last 15 years; numerous reviews summarize the clinical and biological aspects (6–13) or the molecular mechanisms affected during the tRNA life cycles (14–18). Owing to their central function in protein synthesis, the malfunction of mutant tRNAs is thought to play a key role in the corresponding mitochondrial pathologies and has therefore been extensively investigated. Various aspects were explored, emphasizing aminoacylation properties and steady-state levels of aminoacyl-tRNA in mitochondria, but also including tRNA synthesis, structure, stability and interaction with partners of the translational machinery. However, so far only a limited number of mutations and of tRNAs have been explored, and for each, only one or a subset of possibilities were investigated. Neither general rules nor common features have emerged, and further studies are thus necessary.
Studies are complicated due to intrinsic properties of mitochondria and mitochondrial tRNAs (heteroplasmy, limited access to natural material and almost no possibility for genetic manipulation). Mitochondria are present in hundreds of copies per cell, each of them carrying more than one copy of DNA. A mitochondrial mutation can thus be heteroplasmic, varying in proportion to wild-type from 0 to 100% mutated DNA in the different tissues of an affected patient (Figure 1B). The threshold for effect of a mitochondrial mutation is usually between 15 and 50% of the normal functional level (referring to a combination of the proportion of the mutant tRNA in the heteroplasmic population with the effect of the mutation on processing, aminoacylation and function in translation; Figure 1C). Above the upper threshold, no pathology is observed, and below the lower threshold, the condition would be lethal. The heteroplasmic state also complicates analysis and interpretation of mutational effects. To minimize this problem, cybrid cells homoplasmic for a specific mutation can be produced by fusion of enucleated patient cells with a 0 cell line lacking mitochondria (19). Although such cells allowed the evaluation of several properties of mutated tRNAs (e.g. steady-state level of aminoacylation, stability and post-transcriptional modification), their construction is difficult and time consuming, so that only a handful of mutations have been analyzed so far. Moreover, the variety of cell types used to generate cybrids with different nuclear backgrounds (e.g. HeLa, osteosarcoma and lymphoblasts) does not allow straightforward comparisons (14). Ex vivo analyses, based on purification of a specific tRNA from cultured cells or from biopsies, are very powerful but have quantitative limitations (20,21).
The alternative approach, which avoids heteroplasmic effects in the study of pathological tRNAs and gives access to a sufficient amount of material, uses enzymes that are either recombinantly expressed or extracted from cultured cells without mitochondrial diseases, and tRNA substrates or precursors that are produced by in vitro transcription. This approach has recently made it possible to investigate the effects of several point mutations on various aspects of tRNA metabolism (Figure 2). Here, we review the defects in mitochondrial tRNA 3' end metabolism revealed by in vitro studies, including processing (cleavage of the 3'-trailer and CCA addition) and aminoacylation. Point mutations in mitochondrial tRNAs may contribute to tRNA malfunction by affecting the efficiency of one or a combination of these reactions. Advantages and limitations of the in vitro approach within the framework of investigating molecular mechanisms of human mitochondrial tRNA disorders will be discussed below.
tRNA 3' END METABOLISM
The 22 human mitochondrial tRNAs are synthesized from polycistronic precursor transcripts where they are interspersed among 11 mRNAs and two rRNAs (2,22,23), often with no spacers (Figure 1A). Table 1 provides details on the neighborhood of each tRNA gene. In about half the cases, tRNA excision from primary transcripts leads to mature mRNAs and rRNAs with proper ends, others are either flanked by a small (1–2 nt) spacer and several even overlap by one residue with the adjacent RNA, as in the case of tRNATyr/tRNACys. Therefore, the processing of mitochondrial tRNAs requires precise endonucleolytic cleavage at both 5' and 3' ends (Figure 2). Pre-tRNA 3' ends are subsequently completed by the addition of the CCA triplet, and a subset of nucleotides becomes post-transcriptionally modified. Finally, mature and correctly folded tRNAs are esterified at the 3' end with the cognate amino acid by the corresponding aminoacyl-tRNA synthetase. They are subsequently carried to the ribosome by translation factor , allowing protein synthesis. This brief overview illustrates the importance of 3' end metabolism for both tRNA synthesis and tRNA function.
Table 1. tRNA punctuation and 3' end sequence
Mutations in tRNAs may affect the reaction efficiency of the enzymes involved in tRNA 3' end metabolism in a direct way, or indirectly via changes in structural properties of the precursor or mature tRNA. Importantly, in this context, human mitochondrial tRNAs deviate structurally from classical tRNAs (24). Most extreme is tRNASer(AGY), in which the entire D-domain is missing, and an extended anticodon stem is present. Other tRNAs possess all expected secondary structural domains and can be folded into a cloverleaf. Large variations in the size of the D-loop, and especially T-loop (which is strictly maintained at 7 nt in canonical, nuclear-encoded tRNAs), are observed, and classically conserved or semi-conserved nucleotides (e.g. GG in the D-loop, UUC converted into TC in the T-loop), however, are absent. These common deviations from canonical secondary structure lead, in turn, to unpredictable tertiary interactions, suggesting greater flexibility of mitochondrial tRNAs. Owing to the nucleotide composition bias of the two mitochondrial DNA strands (1), 14 tRNAs (transcribed from the heavy strand; Figure 1A) are A-rich (36%) and G-poor (14%), and eight tRNAs (transcribed from the light strand) are U-rich (31.6%) and C-poor (16.4%), weakening the helical domains of tRNA secondary structures (24). These structural peculiarities may influence all aspects of the life-cycle of human mitochondrial tRNAs and contribute to the effects of pathogenic mutations.
CLEAVAGE OF tRNA 3'-TRAILERS
tRNA end processing
Excision of tRNAs from primary polycistronic mitochondrial transcripts is catalyzed by two specialized enzymes, RNase P and tRNase Z (the tRNA 3' end processing endonuclease, previously known as 3'-tRNase or RNase Z). Although not obligatory, a reaction order favoring cleavage by RNase P at the 5' end of a pre-tRNA followed by tRNase Z 3' end cleavage (Figure 2A–C) is observed in many systems, including higher plants (25), human mitochondria (26) and budding yeast (27). While RNase P efficiently cleaves substrates with long 3' end trailers, tRNase Z is apparently sensitive to the length of the 5'-leader (28). Experiments discussed below have therefore been performed on simplified model substrates.
Characterization of tRNase Z
Interestingly, the single fruit fly tRNase Z gene encodes both the nuclear and mitochondrial enzymes (29), and the ratio of nuclear/mitochondrial distribution in such cases is often 20:1 (A. Hopper, personal communication). tRNase Z RNAi knockdowns cause the accumulation of both nuclear-encoded and mitochondrial pre-tRNAs with 3' end extensions, but did not produce a phenotype or reduce the amount of mature cytoplasmic tRNA. The amount of mature mitochondrial tRNAIle and tRNATyr, however, was clearly reduced (29). The concentration of tRNase Z is thus more likely to be limiting in mitochondria than in nuclei, but issues of steady-state concentration and half-life of both the processing enzyme and the tRNA product must also be considered.
Human tRNase Z has also been cloned, recombinantly expressed and partially characterized, leading to the conclusion that these enzymes belong to the ELAC family (30). Both human homologs ELAC1 and ELAC2 (the short and long forms, respectively) possess the tRNase Z activity (31). It is unclear why humans and a number of other metazoans encode multiple forms of tRNase Z. Sorting servers (e.g. Mitoprot; http://ihg.gsf.de/ihg/mitoprot.html) predict that only the long form (ELAC2) carries an N-terminal mitochondrial targeting sequence, and ELAC2 has a lower KM and a wider substrate range than ELAC1 (32), consistent with ELAC2 function as mitochondrial tRNase Z. Since the human mitochondrial form of tRNase Z has not yet been proven, however, tRNase Z activity was obtained from HeLa mitoplast extracts for the analysis of mitochondrial tRNA 3' end processing (26,33,34).
ELAC2 was first characterized as a prostate cancer susceptibility gene (35). Two of the identified mutations which are weakly linked to prostate cancer are located in or close to the phosphodiesterase II domain, the active site of tRNase Z. ELAC2 is associated with mitotic spindles (36), and the RNAi knockdown of the Caenorhabditis elegans homolog of ELAC2 interferes with germline proliferation (37), but a possible function in mitochondrial tRNA processing, in which certain mutations increase the risk of human prostate cancer, is also worth considering.
CCA, a tRNase Z anti-determinant
The CCA triplet needed at the 3' end of all mature tRNAs is not mitochondrially encoded (rightmost column in Table 1), and must thus be added to the discriminator base (the last encoded 3' end nucleotide) by CCA-adding enzyme after 3' end processing. Mature in vitro transcribed tRNA with CCA at its 3' end is a tRNase Z anti-determinant (38,39); dashed line with an X through it; Figure 2DB,C; , being neither a substrate for nor an inhibitor of tRNase Z. This ensures that mature tRNA proceeds smoothly to aminoacylation without recycling through tRNase Z. Interestingly, a partial CCA terminus consisting of –CC is sufficient to produce the anti-determinant effect.
The CCA terminus is also a tRNase Z anti-determinant in Bacillus subtilis, in which the CCA triplet is encoded in some tRNA genes, while it must be added post-transcriptionally to the transcripts of others. Knockout of B.subtilis tRNase Z shows that most of the CCA-containing pre-tRNAs are processed by a 3'-exonuclease, while those without encoded CCA are processed endonucleolytically (41). In Thermotoga maritima, CCA is transcriptionally encoded at the 3' end of tRNA genes. Surprisingly, CCA is a cleavage site determinant in this organism, guiding tRNase Z to efficiently cleave downstream of the CCA triplet instead of after the discriminator base (42).
Defective tRNA end processing as a model for pathogenesis in vivo
The evidence which led to a model linking defective tRNA processing to pathogenesis was first obtained from studies of mitochondrial RNA in cells harboring disease-associated mutations in tRNALeu(UUR) (43,44). This tRNA gene is located between the genes for 16S rRNA and the ND1 subunit (Figure 1A). There is no spacer between the 3' end of 16S rRNA and position +1 of tRNALeu(UUR), and only a 2 nt spacer is present between the discriminator base of tRNALeu(UUR) and the initiation codon of the ND1 message (Table 1). An increase in the steady-state concentration of a discrete precursor (RNA 19) consisting of the linked 16S rRNA-tRNALeu(UUR)-ND1 RNA was observed with the A3243G and A3302G mutations (43,44), which are located in the D-loop and acceptor stem of the gene product, respectively (see Figure 3).
Figure 3. Mitochondrial tRNAs and pathogenesis-associated mutations analyzed for their effects on 3' end processing, CCA addition and/or aminoacylation. (A) tRNASer(UCN), (B) tRNAGly, (C) tRNAIle, (D) tRNALeu(UUR) and (E) tRNALys. tRNAs are represented in their secondary cloverleaf structures. Mutations detected so far in the corresponding genes are indicated by arrows, and numbering is according to the mitochondrial genome (1). Black arrows and numbers refer to pathology-related mutations for which one or more of the enzymatic activities described here have been studied (see Table 2); grey numbers and arrows identify those pathology-related mutations which have not yet been characterized as to their effects on any of these reactions (MITOMAP).
Table 2. Effects of mutations in human mitochondrial tRNAs on 3' end metabolism
Reduced tRNase Z processing efficiency of pathogenesis-associated tRNA precursors
The hypothesis that pathogenesis-associated mutations could cause defective tRNA end processing, thereby contributing to disease, was further investigated in vitro using precursors with both 5' and 3' end extensions (26,33,45). We have tested the hypothesis that pathogenesis-associated mutations in tRNAs could reduce the efficiency of precursor 3' end cleavage using wild-type and pathogenesis-associated mutations in tRNASer(UCN) (26), tRNAIle (46) and tRNALeu(UUR) (34) (Figure 3 and Table 2). According to this hypothesis, poor excision of tRNA could contribute to pathology. This would only be the case, however, if the intra-mitochondrial concentration of the processing enzyme is limiting.
The clearest example of a pathogenesis-associated mutation which could cause defective tRNA 3' end processing is the T7445C substitution in tRNASer(UCN), which causes non-syndromic deafness (47,48). T7445C falls just on the 3'-side of the discriminator base of tRNASer(UCN) (Figure 3), outside of the mature tRNA, suggesting that processing reactions depending on tRNA structure or sequence would be unaffected. The mutation, however, changes 3' end processing, leading to defective 3' ends (47). The cell lines homoplasmic for T7445C display approximately one-third of the steady-state level of tRNASer(UCN) observed in homoplasmic wild-type sibling cell lines (48). Interestingly, the T7445C substitution changes the sequence following the discriminator base (underlined) from G/UCU– to G/CCU–, a tRNase Z anti-determinant . The pathogenesis-associated T7445C tRNA precursor could not be cleaved in vitro, while wild-type tRNASer(UCN) could be processed efficiently (26). Additionally, tRNASer(UCN) with an extra G at the boundary between the V-loop and the T-stem (7472insG) seems to be processed and aminoacylated less efficiently than the normal tRNA (49,50). A possible rescue function for 3'-exonucleases in mitochondria has been discussed previously (26,50). In principle, any endonucleolytic cleavage which unlinks tRNASer(UCN) from the neighboring tRNATyr located 2.5 kb downstream (Figure 1A) could be followed by 3'-exonucleolytic cleavage to the discriminator.
Among trinucleotides found at the 3' end of human mitochondrial tRNAs following the discriminator base, /CCA is not observed, nor is /CC, and /C is found only once (in tRNACys; rightmost column in Table 1). In six cases , the trinucleotide following the discriminator base is a mitochondrial translation initiation codon (/AUG, /AUA, /AUU or /AAU). Genes for tRNAPhe and tRNAThr, like tRNALys, also have the trinucleotide AAU following the discriminator base, although the sequence immediately downstream does not encode a polypeptide in these cases. In conclusion, there is no natural tRNase Z anti-determinant in the sequences of 21 tRNA genes, and /C in tRNACys would not be expected to have much effect.
The kinetics of tRNase Z processing was performed using eight substitutions in tRNAIle associated with cardiomyopathy or ophthalmoplegia (46), and three in tRNALeu(UUR) correlated with mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes (MELAS) and assorted myopathies (34) (Figure 3). Four of the mutations (A4269G, A4295G, G4309A and A4317G) reduce the tRNase Z reaction efficiency up to 10-fold, and three tRNALeu(UUR) mutations decrease the efficiency up to 3-fold relative to wild type (Table 2), all due to lower Vmax. Interestingly, in tRNALeu(UUR), the reduction in processing efficiency increases with proximity to the 3'cleavage site (WT > A3243G > A3302G C3303U). These results support the hypothesis that 3' end processing defects are among the molecular events involving mutant mitochondrial tRNAs which contribute to mitochondrial tRNA disorders. Note, however, that there can be discrepancies between results obtained with wild-type and mutant tRNAs from cybrid cells and those from in vitro transcripts when it is possible to directly compare them. For example, the tRNAIle mutation A4269G was demonstrated to affect stability without any evident effect on maturation when the tRNAs were extracted from homoplasmic cybrid cells (51).
RNase P
This review is principally concerned with reactions at the 3' end of tRNA and mitochondrial pathogenesis, but for the sake of completeness, 5' end cleavage by RNase P should also be mentioned. Human mitochondrial RNase P activity was found by some to be indistinguishable from nuclear RNase P (52,53), while others found rat and human mitochondrial and nuclear RNase P to have different characteristics (33,54). The composition of human mitochondrial RNase P has not been completely resolved because it is difficult to purify mitochondria free of cytosolic or nuclear contamination (55). The number of RNase P molecules per mitochondrion required to process the mitochondrial transcripts could be very low (53), and if the intra-mitochondrial concentration of enzymes involved in tRNA metabolism is limiting, the modest reductions in processing efficiency arising from mutations in tRNA could lead to the accumulation of precursors and reduce the amount of functional mature tRNA.
SYNTHESIS OF THE CCA TERMINUS
Cloning, expression and characterization of tRNA nucleotidyltransferase
tRNA requires the base triplet CCA at the 3'-terminus for its aminoacylation and participation in protein synthesis. In mitochondrial tRNAs, this sequence must be added post-transcriptionally as an essential step in tRNA maturation (Figure 2). CCA addition is catalyzed by CCA-adding enzymes which are conserved in all organisms of all kingdoms. In humans, the mitochondrial, nuclear and cytosolic forms of this enzyme are encoded by a single gene located on chromosome 3 (56). A mitochondrial target sequence encoded in exon 1 ensures delivery of a proportion of the CCA-adding enzyme to mitochondria. Importantly, both the tRNase Z and CCA-adding enzyme are encoded in both forms by using two or three in-frame start codons separated by a short distance, with the mitochondrial targeting sequence being present only when translation initiates at the upstream AUG. Since CCA-adding enzymes are closely related to poly(A) polymerases, these enzymes were sometimes mistakenly identified by sequence analysis (57,58). The recombinant enzymes, however, can easily be distinguished based on their biochemical activity. The human CCA-adding enzyme was therefore cloned, recombinantly expressed and characterized (56,59). While showing kinetic properties comparable to those of other CCA-adding enzymes, the human enzyme has a specific affinity for mitochondrial tRNAs which lack some of the canonical elements, such as the TC-loop. The N-terminal domain of this enzyme contains the catalytic core as well as features such as nucleotide recognition/discrimination and specificity for tRNA (60,61). Recent domain swap experiments between the CCA-adding enzyme and poly(A) polymerase indicate that C-terminal domains also contribute to the discriminatory power of CCA addition (62).
Influence of pathology-related mutations on CCA addition
The activity of CCA-adding enzyme can be influenced by changes in sequence and/or structure of substrate tRNAs which can consequently contribute to pathogenicity, as first reported in certain disease-associated mutations in mitochondrial tRNAIle and tRNAGly (63) as well as tRNALeu(UUR) (34). In both tRNAIle and tRNAGly, an A was replaced by G in the TC-loop (A4317G in tRNAIle and A10044G in tRNAGly; Figure 3). These two domains combine to form the ‘elbow’ of the L-shaped three-dimensional (3D) structure of tRNA, a recognition element for CCA-adding enzymes (64–67). Structure probing experiments on in vitro transcripts carrying these replacements revealed that they lead to an aberrantly stable TC-arm or to a weakened interaction between the TC- and D-loop, respectively (63). In the kinetic analysis, KM values were only slightly affected, indicating that the enzyme's affinity for the mutated tRNAs was not reduced. The observed decrease in kcat suggests that the chemical step of catalysis is affected, possibly due to an inappropriate orientation of the tRNA substrate in the catalytic core of the enzyme (63). Besides tRNAIle and tRNAGly, five pathogenic mutations in tRNALeu(UUR) were investigated for their effects on CCA addition (Figure 3) (34). Only the substitution closest to the 3'-terminus, C3303T (causing cardiomyopathy), showed a significant effect on CCA addition (Table 2), reducing nucleotide incorporation by a factor of 5.5. While Vmax was unaffected, KM increased, indicating that C3303T reduces the affinity of the CCA-adding enzyme for tRNA without interfering with catalysis.
These examples show that pathogenesis-associated mutations which affect the same enzymatic reaction can do so by different means, either interfering with catalysis (as in A4317G in tRNAIle and A10044G in tRNAGly) or affecting the strength of enzyme–substrate interaction . Furthermore, the same mutation affects two different enzymes differently, causing a decrease in the Vmax of tRNase Z and an increase in the KM of CCA-adding enzyme (34).
AMINOACYLATION
Once the tRNA is synthesized (i.e. the steps of 5' and 3' end cleavage and CCA addition are finished as in Figure 2AD, post-transcriptional modifications are complete and 3D folding has taken place), the amino acid can be attached by the cognate aminoacyl-tRNA synthetase (aaRS; Figure 2DE). It can then be brought to the ribosome by transient binding to translation factors, allowing transfer of the amino acid to the growing polypeptide chain, thereby fulfilling its fundamental role in protein synthesis.
Mitochondrial aminoacyl-tRNA synthetases
Eukaryotic cells contain at least two sets of aminoacyl-tRNA synthetases, one devoted to cytoplasmic protein synthesis, the other to translation of the mitochondrial genetic information. In humans, cytosolic and mitochondrial enzymes are generally encoded by distinct genes, with the exception of GlyRS and LysRS . Enzymes targeted to mitochondria are synthesized with a mitochondrial signal sequence which is cleaved after import. Interestingly, mitochondrial synthetases recognize not only mitochondrial but also cytosolic tRNAs and tRNAs from other organisms; on the other hand, mitochondrial tRNAs are not recognized by non-cognate enzymes (69), probably due to their unusual structural properties.
Specific recognition of a tRNA by its cognate aaRS is based on the presence of ‘identity elements’, usually a set of bases or base pairs in the tRNA. Such elements have been defined precisely for a large number of bacterial, archaeal and lower eukaryotic tRNAs (70,71) but remain largely unknown for human mitochondrial tRNAs. Efficient aminoacylation also depends on structural features within the tRNA and on more global conformational properties.
Mutations in mitochondrial tRNAs reduce aminoacylation efficiency
To evaluate the impact of the mutations in pathology-related tRNAs on aminoacylation, wild-type and mutant mitochondrial tRNAs have been prepared by in vitro transcription and aminoacylated by the corresponding recombinantly expressed aminoacyl-tRNA synthetases. The first aminoacylations tested in such an in vitro system were those specific for serine and isoleucine. The serine system contributed to a fundamental understanding of tRNA/aaRS recognition, but pathology-related mutations were not analyzed (72,73). In contrast, 10 pathology-related mutations within tRNAIle associated with either ophthalmoplegia or cardiomyopathy were tested in the presence of a crude mitochondrial enzymatic extract (74) or with cloned and over-expressed IleRS (75,76) (Figure 3). Some mutations strongly reduce the ability of the tRNA to be aminoacylated, while others have only a very limited effect (Table 2). Mutations correlated with ophthalmoplegia reduced the isoleucylation efficiencies (as measured by the ratio kcat/KM compared to wild-type) from between 25- and 50-fold to more than 1000-fold. More modest reductions in aminoacylation efficiencies (or no effect) were reported for mutations correlated with cardiomyopathies (75,76). Mutants with compensatory mutations that restore disrupted base pairing regain wild-type aminoacylability. The structural effects of the mutations thus provide a stronger explanation for the reductions in aminoacylation efficiency than direct interference in tRNA/synthetase recognition. The analysis of tRNAIle variants also revealed that a mutant tRNA can potentially compete for the corresponding wild-type tRNA by inhibiting its aminoacylation. This situation could mimic the in vivo case, in which wild-type and mutant tRNAs co-exist in heteroplasmic patient cells (Figure 1B).
The use of in vitro transcripts leads to difficulties for investigation of aminoacylation properties of several human mitochondrial tRNAs other than tRNAIle and tRNASer because the corresponding wild-type transcripts do not fold into the expected cloverleaf secondary structures. The cases of tRNALys and tRNALeu(UUR) are of particular interest.
Wild-type tRNALys folds into an extended hairpin (77) due to the absence of a particular post-transcriptional modification, the N1-methyl group on adenine 9 (78). This methyl group prevents base pairing of nucleotides 9 and 64 (position 8355) from producing an extended acceptor stem in native tRNALys. The role of this methyl group has been mimicked in an in vitro transcript by mutating either nucleotide 9 or 64 to prevent them from base pairing (77). The corresponding ‘pseudo-wild-type’ transcripts fold into the cloverleaf structure and become active in aminoacylation with kinetic properties close to those of cytoplasmic human tRNALys (79). By second-site mutation of such ‘pseudo-wild-type’ transcripts, it became possible to investigate the effect on aminoacylation of all pathology-related mutations described so far in the mitochondrial tRNALys gene (79). Two (G8313A and G8328A) out of the eleven pathology-related mutations lead to a very strong loss (5000- to 7000-fold) in aminoacylation efficiency (kcat/KM); the other mutations had mild (2- to 10-fold losses in efficiency) or no effects. Thus, as in the case of tRNAIle, only a subset of mutations affects aminoacylation. The mechanisms by which the G8313A and G8328A mutations affect lysylation are unknown. Both mutations are located in the anticodon branch (anticodon domain and D-domain) on the same side on a model L-shaped 3D tRNA structure where they could directly interfere with LysRS interaction (79), while mutations without effect are located on the opposite side or in the amino acid accepting branch. Alternatively, these mutations could cause large structural rearrangements of the tRNA, leading indirectly to failure of synthetase recognition.
tRNALeu(UUR) has also been prepared by in vitro transcription. The cloned and purified human mitochondrial LeuRS accepts this transcript as a substrate at a 16- to 50-fold lower efficiency than native tRNALeu(UUR) extracted from cell lines (21,80). Here, the transcript only folds partially into a tRNA structure, despite the presence of sequence elements typical of classical tRNAs required for correct folding. Indeed, the amino acid acceptor stem and the TC-arm fold normally, but a large ‘floppy’ domain is observed instead of well-defined D and anticodon domains (34,80,81). Footprinting experiments suggest that the synthetase induces a structural change within the tRNA, leading to a productive interaction. Several mutations were analyzed to determine their effects on leucylation (Figure 3 and Table 2). For example, the effects of MELAS mutation A3243G and a second mutation at the same position correlated with encephalomyopathy (A3243T) lead to different effects (21,80). While mutation A3243G displays 10-fold reduced efficiency of aminoacylation, A3243T leads to a 300-fold decrease. The T3271C substitution in the anticodon stem has only a 3-fold negative effect, which can be reversed by a compensatory mutation (82). As was found for tRNAIle and tRNALys, mutations can affect aminoacylation efficiency either through kcat or through KM. Local structural perturbations were observed between some pathogenesis-associated mutations and wild-type tRNALeu(UUR) (80,82).
The structurally unorganized domain in tRNALeu(UUR) prepared by in vitro transcription, as opposed to the well-folded native tRNA extracted from cell lines (83), can be stabilized so as to more closely resemble a canonical structure. The site-directed mutagenesis that replaces C?A mismatches in either the D-stem or the anticodon stem by G-C pairs leads to cloverleaf structures, thus improving aminoacylation efficiency (81,83). These ‘pseudo-wild-types’ represent valuable reference molecules for a systematic investigation of pathology-related mutations on tRNALeu(UUR) functions.
CHANGE IN 3' END METABOLISM OF MITOCHONDRIAL tRNAs IS AN IMPORTANT CONSEQUENCE OF MUTATION
Pathology-related mutations affect 3' end metabolism
Analysis of transcribed wild-type or engineered ‘pseudo-wild-type’ substrates (tRNAs and tRNA precursors), or mutated derivatives, allow the effects of mutations to be screened in vitro. Mutations were found to affect any combination of the activities involved with 3' end metabolism of tRNAs (Figure 2B–E): cleavage of the 3'-trailer, CCA addition and aminoacylation. Although data summarized in Table 2 are not yet complete, it is clear that for many mutations, at least one of the three activities is significantly affected, making the 3' end of tRNAs a reasonable place to look for the effect of pathogenesis-associated mutations. Only three mutations have been tested for their effect on all three functions. Among these, mutation A3243G in tRNALeu(UUR) leads to a weak effect on 3' end cleavage (2.2-fold decrease in efficiency), no effect on CCA addition and a mild effect on aminoacylation (2- to 12-fold decrease in efficiency). The other two, A4317G in tRNAIle and C3303T in tRNALeu(UUR), substantially reduce the efficiency of 3' end cleavage and CCA addition but have little or no effect on aminoacylation. Nine mutations in tRNAIle have been tested for their effect on both 3' end cleavage and aminoacylation. For some of them (4269, 4295 and 4320), cleavage of the 3' end trailer of the primary tRNA transcript is reduced but aminoacylation is not affected; for others (4274, 4284 and 4300), 3' end maturation is unaffected, but reduced efficiency of aminoacylation is observed.
Whatever the affected activity, the end result is expected to be the same, a decrease in aminoacyl-tRNA available for protein synthesis. While the effect can differ at every level, an accumulation of small effects on several aspects of tRNA metabolism could lead to a strong global negative effect, sufficient to place the level of mitochondrial protein synthesis within the zone of pathology (Figure 1C). Indeed, limiting the amount of aminoacyl-tRNA, by any means, reduces the rate of translation (84,85). Such reductions have been observed in vivo in many cases involving pathogenesis-associated mitochondrial tRNAs (49,86–89). For some mutations, the 3' end metabolism was not affected (see Table 2); in these cases, other levels are presumably responsible for pathology, as suggested by the results of in vitro and in vivo studies by numerous investigators . The defects were observed in tRNA synthesis and maturation, including impaired termination of transcription, impaired RNase P processing and absence of post-transcriptional modifications (especially of the anticodon triplet, hindering codon reading). Post-aminoacylation functions of tRNA, including interaction with EF-Tu, codon reading or polysome formation, were sometimes affected. Decreased stability is the most common feature of mutant mitochondrial tRNAs associated with pathogenesis.
Structural aspects
As discussed above, the mechanisms by which mutations in tRNA affect the enzymatic activities of tRNA metabolism remain largely unknown. Mutations may interfere directly with recognition by the concerned enzyme, or change the overall structure enough to perturb the recognition. Owing to the intrinsic structural weakness of human mitochondrial tRNAs, the effect of mutations on tRNA structure may be amplified. Most of the pathology-associated mutations are conservative nucleotide transitions (90). When found in stems, canonical base pairs are often changed to weak G-U (U-G) wobble pairs or to C?A (A?C) mismatches (see Figure 3). In many cases, the effect on stem structure would be expected to be mild, because G-U and C?A can hydrogen bond with strength comparable to the A-U pair (91). Their structural effect could be magnified, however, since they are often found in regions which are already thermodynamically weak (18). Further illustrating this point, A4309G in the T-stem of tRNAIle displays a weakened secondary structure (46), and T3271C in the anticodon stem of tRNALeu(UUR) loses its negative effect on aminoacylation when combined with a stabilizing substitution in the D-stem (82).
EFFECTIVENESS AND CAVEATES CONCERNING USE OF IN VITRO TRANSCRIPTS AS TOOLS FOR SYSTEMATIC SCREENING
The results summarized here on the potential impact of mitochondrial tRNA mutations on 3' end metabolism were obtained using in vitro transcribed substrate tRNAs or tRNA precursors. The in vitro strategy of RNA preparation is of particular interest for human mitochondrial tRNAs since it allows the preparation of a large amount of any tRNA (either wild-type or mutant). Furthermore, in vitro kinetic analysis can lead to mechanistic insight. The structural flexibility of these tRNAs leads, in some cases, to the need to create ‘pseudo-wild-type’ molecules which have a single mutation introduced to stabilize structure and allow correct folding. Thoroughly documented in at least two cases to react in a manner comparable to that of the corresponding natural tRNAs extracted from cells, we believe they are relevant substitutes which allow useful analyses of second-site (pathogenesis-associated) mutations.
The lack of post-transcriptional modifications can be a drawback since such modifications can affect the structure and function of tRNA (92). Post-transcriptional modifications in classical tRNAs play either of two major roles, stabilizing tRNA structure and/or providing a recognition element for interacting partners such as aminoacyl-tRNA synthetases or mRNA codons. Therefore, aminoacylation properties of in vitro transcripts may be different from those of fully modified tRNAs in vivo. However, mutations introduced into the ‘pseudo-wild-type’ molecules are designed to stabilize the tRNA structure, replacing the structural role of the naturally occurring modifications to the extent that is feasible and allowing for efficient aminoacylation.
The in vitro approach is also valid when testing 3' end maturation activities such as tRNase Z activity, because this enzyme processes primary transcripts which may not yet be modified. Concerning CCA addition, this reaction can take place before all modifications are introduced into the tRNA (93). CCA-adding enzyme requires a properly folded tRNA acceptor stem to synthesize the CCA triplet; this domain has been shown to be correctly folded in the case of in vitro transcripts which have been analyzed for CCA addition . Modified nucleotides may thus be unnecessary for CCA addition and by inference, for the preceding tRNase Z reaction.
Significance of small effects of mutations on enzyme activities measured in vitro
Two major questions affect the interpretation of measured kinetic values. (i) To what extent do in vitro effects reflect the in vivo situation? (ii) Do small effects on reaction efficiency contribute to the expression of disorders? The first question is difficult to answer since in vitro assays take into account neither possible competition events nor the cellular environment, which the in vivo approach does. Moreover, one approach measures kinetic parameters of an enzymatic reaction leading to a value for efficiency, while the other measures the steady-state level of reaction products (e.g. aminoacyl-tRNA). These parameters cannot be compared directly, leading to apparent discrepancies between some of the in vitro and in vivo results. A valuable approach would be to measure reaction kinetics using in vitro wild-type and mutated transcripts and corresponding native wild-type and mutated tRNAs extracted from cybrid cell lines. In the case of tRNALys, results were the same (20,79). In the case of tRNALeu(UUR), however, results were significantly different (21). The native mutated tRNALeu(UUR) (A3243G) showed a 25-fold loss in aminoacylation efficiency when compared to the corresponding wild-type tRNA, while the mutated in vitro transcript only led to a 2.5- to 3-fold loss compared to the wild-type transcript. The transcripts can thus underestimate the impact of mutations; they can nonetheless usefully serve as indicators/detectors of the effects of mutation on aminoacylation.
Concerning the magnitude of effects and their relevance to pathology, Table 2 shows between 2.1- and 12.5-fold effects of mutations on cleavage of the 3'-trailer and at most a 5.5-fold decrease in efficiency of CCA addition. On the other hand, effects on aminoacylation properties of transcripts can be much greater, ranging from a 2-fold to a several 1000-fold decrease in efficiency. While several 1000-fold reductions in aminoacylation would be expected to contribute strongly to pathogenicity, even small reductions in individual maturation steps can contribute to the onset of disease, amplified by the cumulative effects of these impairments. Along these lines, the A4269G mutation in tRNAIle displays a 2-fold decrease in affinity for EF-Tu (94).
FUTURE PROSPECTS
Pathogenesis-associated tRNAs may contribute to mitochondrial disease due to individual effects or to accumulation of a number of deficiencies in end maturation, including 3' end cleavage by tRNase Z and CCA addition by tRNA nucleotidyltransferase, and in aminoacylation. Effects of the pathogenesis-associated substitutions on these reactions, analyzed using in vitro transcripts, does not preclude other events, including 5' end processing by RNase P, post-transcriptional modifications, elongation factor binding, ribosome binding, etc.
One promising direction is the analysis of nuclear-encoded enzymes involved in mitochondrial tRNA metabolism. In the case of enzymes encoded by single genes which are targeted to the mitochondria, cytoplasm and/or nucleus, it will be important to establish the relative enzyme concentrations in the individual compartments, which will greatly influence tRNA metabolism.
Virtually all pathogenesis-associated tRNA mutations can be investigated using in vitro transcripts. It will thus be possible to quantify and classify the effect(s) of individual mutations on metabolism as it relates to pathogenicity and to accumulate fundamental information concerning structure/function relations in human mitochondrial tRNAs. Although the in vitro strategy is not universally applicable, it is a valuable approach for unraveling many aspects of tRNA-based mitochondrial pathology before undertaking more costly in vivo analyses.
ACKNOWLEDGEMENTS
We gratefully acknowledge helpful conversations with M. Sissler and her help in the preparation of Figure 3. This work was supported by NIH grants F33GM64266, S06GM08153 and T34GM08498 (L.L.), by DFG grant Mo 634/2 and MPG (M.M.) and by CNRS, AFM, EC-QLG2-CT-1999-00660 (C.F.).
REFERENCES
Anderson,S., Bankier,A.T., Barrel,B.G., de Bruijn,M.H.L., Coulson,A.R., Drouin,J., Eperon,J.C., Nierlich,D.P., Roe,B.A., Sanger,F. et al. ( (1981) ) Sequence and organization of the human mitochondrial genome. Nature, , 290, , 457–465.
Montoya,J., Ojala,D. and Attardi,G. ( (1981) ) Distinctive features of the 5'-terminal sequences of the human mitochondrial mRNAs. Nature, , 290, , 465–470.
Pesole,G., Gissi,C., DeChirico,A. and Saccone,C. ( (1999) ) Nucleotide substitution rate of mammalian mitochondrial genomes. J. Mol. Evol., , 48, , 427–434.
Holt,I.J., Harding,A.E. and Morgan-Hughes,J.A. ( (1988) ) Deletions of muscle mitochondrial DNA in patients with mitochondrial myopathies. Nature, , 331, , 717–719.
Lestienne,P. and Ponsot,G. ( (1988) ) Kearns–Sayre syndrome with muscle mitochondrial DNA deletion. Lancet, , 1, , 885.
Larsson,N.-G. and Clayton,D.A. ( (1995) ) Molecular genetic aspects of human mitochondrial disorders. Annu. Rev. Genet., , 29, , 151–178.
Schon,E.A., Bonilla,E. and DiMauro,S. ( (1997) ) Mitochondrial DNA mutations and pathogenesis. J. Bioenerg. Biomembr., , 29, , 131–149.
DiMauro,S. and Schon,E. ( (1998) ) Nuclear power and mitochondrial disease. Nature Genet., , 19, , 214–215.
Wallace,D.C. ( (1999) ) Mitochondrial diseases in man and mouse. Science, , 283, , 1482–1488.
Chinnery,P.F. and Turnbull,D.M. ( (2000) ) Mitochondrial DNA mutations in the pathogenesis of human disease. Mol. Med. Today, , 6, , 425–432.
Shapira,A. ( (2000) ) Mitochondrial disorders. Curr. Opin. Neurol., , 13, , 527–532.
Dimauro,S. and Schon,E. ( (2001) ) Mitochondrial DNA mutations in human disease. Am. J. Med. Genet., , 106, , 18–26.
Betts,J., Lightowlers,R.N. and Turnbull,D. ( (2004) ) Neuropathological aspects of mitochondrial DNA disease. Neurochem. Res., , 29, , 505–511.
Jacobs,H.T. and Holt,I.J. ( (2000) ) The np 3243 MELAS mutation: damned if you aminoacylate, damned if you don't. Hum. Mol. Genet., , 9, , 463–465.
Florentz,C. ( (2002) ) Molecular investigations on tRNAs involved in human mitochondrial disorders. Biosci. Rep., , 22, , 81–98.
Jacobs,H.T. ( (2003) ) Disorders of mitochondrial protein synthesis. Hum. Mol. Genet., , 12, , R293–R301.
Florentz,C., Sohm,B., Tryoen-Tóth,P., Pütz,J. and Sissler,M. ( (2003) ) Human mitochondrial tRNAs in health and disease. Cell. Mol. Life Sci., , 60, , 1356–1375.
Wittenhagen,L.M. and Kelley,S.O. ( (2003) ) Impact of disease-related mitochondrial mutations on tRNA structure and function. Trends Biochem. Sci., , 28, , 605–611.
King,M.P. and Attardi,G. ( (1989) ) Human cells lacking mtDNA: repopulation with exogenous mitochondria by complementation. Science, , 246, , 500–503.
Yasukawa,T., Suzuki,T., Ishii,N., Ohta,S. and Watanabe,K. ( (2001) ) Wobble modification defect in tRNA disturbs codon-anticodon interaction in a mitochondrial disease. EMBO J., , 20, , 4794–4802.
Park,H., Davidson,E. and King,M.P. ( (2003) ) The pathogenic A3243G mutation in human mitochondrial tRNALeu(UUR) decreases the efficiency of aminoacylation. Biochemistry, , 4, , 958–64.
Ojala,D., Montoya,J. and Attardi,G. ( (1981) ) tRNA punctuation model of RNA processing in human mitochondria. Nature, , 290, , 470–474.
Chomyn,A., Cleeter,M.W., Ragan,C.I., Riley,M., Doolittle,R.F. and Attardi,G. ( (1986) ) URF6, last unidentified reading frame of human mtDNA, codes for an NADH dehydrogenase subunit. Science, , 234, , 614–618.
Helm,M., Brulé,H., Friede,D., Giegé,R., Pütz,J. and Florentz,C. ( (2000) ) Search for characteristic structural features of mammalian mitochondrial tRNAs. RNA, , 6, , 1356–1379.
Kunzmann,A., Brennicke,A. and Marchfelder,A. ( (1998) ) 5' end maturation and RNA editing have to preceed tRNA 3' processing in plant mitochondria. Proc. Natl Acad. Sci. USA, , 95, , 108–113.
Levinger,L., Jacobs,O. and James,M. ( (2001) ) In vitro 3' end endonucleolytic processing defect in a human mitochondrial tRNASer(UCN) precursor with the U7445C substitution, which causes non-syndromic deafness. Nucleic Acids Res., , 29, , 4334–4340.
Kuffel,J. and Tollervey,D. ( (2003) ) 3'-processing of yeast tRNATrp precedes 5'-processing. RNA, , 9, , 202–208.
Nashimoto,M., Wesemann,D.R., Geary,S., Tamura,M. and Kaspar,R.L. ( (1999) ) Long 5' leaders inhibit removal of a 3' trailer from a precursor tRNA by mammalian tRNA 3' processing endoribonuclease. Nucleic Acids Res., , 27, , 2770–2776.
Dubrovsky,E.B., Dubrovskaya,V.A., Levinger,L., Schiffer,S. and Marchfelder,A. ( (2004) ) Drosophila RNase Z processes mitochondrial and nuclear pre-tRNA 3' ends in vivo. Nucleic Acid Res., , 32, , 255–262.
Schiffer,S., Rosch,S. and Marchfelder,A. ( (2002) ) Assigning a function to a conserved group of proteins: the tRNA 3'-processing enzymes. EMBO J., , 21, , 2769–2777.
Takaku,H., Minegawa,A., Takagi,M. and Nashimoto,M. ( (2003) ) A candidate prostate cancer susceptibility gene encodes tRNA 3' processing endoribonuclease. Nucleic Acid Res., , 31, , 2272–2278.
Takaku,H., Minegawa,A., Takagi,M. and Nashimoto,M. ( (2004) ) T N-terminal half domain of the long form of tRNase Z is required for the RNase 65 activity. Nucleic Acid Res., , 32, , 4429–4438.
Rossmanith,W., Tullo,A., Potuschak,T., Karwan,R. and Sbisa,E. ( (1995) ) Human mitochondrial tRNA processing. J. Biol. Chem., , 270, , 12885–12891.
Levinger,L., Oestereich,I., Florentz,C. and M?rl,M. ( (2004) ) A pathogenesis-associated mutation in human mitochondrial tRNALeu(UUR) leads to reduced 3'-end processing and CCA addition. J. Mol. Biol., , 337, , 535–544.
Tavtigian,S.V., Simard,J., Teng,D.H., Abtin,V., Baumgard,M., Beck,A., Camp,N.J., Carillo,A.R., Chen,Y., Dayananth,P. et al. ( (2001) ) A candidate prostate cancer susceptibility gene at chromosome 17p. Nature Genet., , 27, , 172–180.
Korver,W., Guevara,C., Chen,Y., Neutoboom,S., Bookstein,R., Tavtigian,S. and Lees,E. ( (2003) ) The product of the candidate prostate cancer susceptibility gene ELAC2 interacts with the gamma-tubulin complex. Int. J. Cancer, , 104, , 283–288.
Smith,M.M. and Levitan,D.J. ( (2004) ) The Caenorhabditis elegans homolog of the putative prostate cancer susceptibility gene ELAC2, hoe-1, plays a role in germline proliferation. Dev. Biol., , 266, , 151–160.
Nashimoto,M. ( (1997) ) Distribution of both length and 5' terminal nucleotides of mammalian pre-tRNA 3' trailers reflect properties of 3' processing endoribonuclase. Nucleic Acids Res., , 25, , 1148–1154.
Mohan,A., Whyte,S., Wang,X., Nashimoto,M. and Levinger,L. ( (1999) ) The 3' end CCA of mature tRNA is an antideterminant for eukaryotic 3'-tRNase. RNA, , 5, , 245–256.
Schiffer,S., Rosch,S. and Marchfelder,A. ( (2003) ) Recombinant RNase Z does not recognize CCA as part of the tRNA and its cleavage efficiency is influenced by acceptor stem length. Biol. Chem., , 384, , 333–342.
Pelligrini,O., Nezzar,J., Marchfelder,A., Putzer,H. and Condon,C. ( (2003) ) Endonucleolytic processing of CCA-less tRNA precursors by RNase Z in Bacillus subtilis. EMBO J., , 22, , 4534–4543.
Minegawa,A., Takaku,H., Takagi,M. and Nashimoto,M. ( (2004) ) A novel endonucleolytic mechanism to generate the CCA 3'-termini of tRNA molecules in Thermotoga maritima. J. Biol. Chem., , 279, , 15688–15697.
King,M.P., Koga,Y., Davidson,M. and Schon,E.A. ( (1992) ) Defects in mitochondrial protein synthesis and respiratory chain activity segregate with the tRNALeu(UUR) mutation associated with mitochondrial myopathy, encephalopathy, lactic acidosis and strokelike episodes. Mol. Cell. Biol., , 12, , 480–490.
Bindoff,L.A., Howell,N., Poulton,J., McCullough,D.A., Morten,K.J., Lightowlers,R.N., Turnbull,D.M. and Weber,K. ( (1993) ) Abnormal RNA processing associated with a novel tRNA mutation in mitochondrial DNA. A potential disease mechanism. J. Biol. Chem., , 268, , 19559–19564.
Rossmanith,W. and Karwan,R.M. ( (1998) ) Impairment of tRNA processing by point mutations in mitochondrial tRNALeu(UUR) associated with mitochondrial diseases. FEBS Lett., , 433, , 269–274.
Levinger,L., Giegé,R. and Florentz,C. ( (2003) ) Pathology-related substitutions in human mitochondrial tRNAIle reduce precursor 3'-end processing efficiency in vitro. Nucleic Acids Res., , 31, , 1904–1912.
Reid,F., Rovio,A.T., Holt,I.J. and Jacobs,H.T. ( (1997) ) Molecular phenotype of a human lymphoblastoid cell-line homoplasmic for the np-7445 deafness-associated mitochondrial mutation. Hum. Mol. Genet., , 6, , 443–449.
Guan,M.-X., Enriquez,J.A., Fishel-Ghodsian,N., Puranam,R.S., Lin,C.P., Maw,M.A. and Attardi,G. ( (1998) ) The deafness-associated mitochondrial DNA mutation at position 7445, which affects tRNASer(UCN) precursor processing, has long-range effects on NADH dehydrogenase subunit ND6 gene expression. Mol. Cell. Biol., , 18, , 5868–5879.
Toompuu,M., Yasukawa,T., Suzuki,T., Hakkinen,T., Spelbrink,J., Watanabe,K. and Jacobs,H. ( (2002) ) The 7472insC mitochondrial DNA mutation impairs the synthesis and extent of aminoacylation of tRNASer(UCN) but not its structure or rate of turnover. J. Biol. Chem., , 22240–22250.
Toompuu,M., Levinger,L., Nadal,A. and Jacobs,H. ( (2004) ) The 7472insG mutation impairs 5' and 3' processing of tRNASer(UCN). Biochem. Biophys. Res. Commun., , 322, , 803–813.
Yasukawa,T., Hino,N., Suzuki,T., Watanabe,K., Ueda,T. and Ohta,S. ( (2000) ) A pathogenic point mutation reduces stability of mitochondrial mutant tRNAIle. Nucleic Acids Res., , 28, , 3779–3784.
Doersen,C.J., Guerrier-Takada,C., Altman,S. and Attardi,G. ( (1985) ) Characterization of an RNase P activity from HeLa cell mitochondria. Comparison with the cytosol RNase P activity. J. Biol. Chem., , 260, , 5942–5949.
Puranam,R.S. and Attardi,G. ( (2001) ) The RNase P associated with HeLa cell mitochondria contains an essantial RNA component identical in sequence to that of the nuclear RNase P. Mol. Cell. Biol., , 21, , 548–561.
Manam,S. and Van Tuyle,G.C. ( (1987) ) Separation and characterization of 5'- and 3'-tRNA processing nucleases from rat liver mitochondria. J. Biol. Chem., , 262, , 10272–10279.
Rossmanith,W. and Potuschak,T. ( (2001) ) Difference between Mitochondrial RNase P and Nuclear RNase P. Mol. Cell. Biol., , 21, , 8236–8237.
Reichert,A., Thurlow,D. and M?rl,M. ( (2001) ) A eubacterial origin for the human tRNA nucleotidyltransferase? Biol. Chem., , 382, , 1431–1438.
Raynal,L., Krisch,H. and Carpousis,A. ( (1998) ) The Bacillus subtilis nucleotidyltransferase is a tRNA CCA-adding enzyme. J. Bacteriol., , 180, , 6276–6282.
Sohlberg,B., Huang,J. and Cohen,S. ( (2003) ) The Streptomyces coelicolor polynucleotide phosphorylase homologue, and not the putative poly(A) polymerase, can polyadenylate RNA. J. Bacteriol., , 185, , 7273–7278.
Nagaike,T., Suzuki,T., Tomari,Y., Takemoto-Hori,C., Negayama,F., Watanabe,K. and Ueda,T. ( (2001) ) Identification and characterization of mammalian mitochondrial tRNA nucleotidyltransferase. J. Biol. Chem., , 276, , 40041–40049.
Li,F., Xiong,Y., Wang,J., Cho,H., Tomita,K., Weiner,A. and Steitz,T. ( (2002) ) Crystal structures of the Bacillus stearothermophilus CCA-adding enzyme and its complexes with ATP or CTP. Cell, , 111, , 815–824.
Augustin,M., Reichert,A., Betat,H., Huber,R., Morl,M. and Steegborn,C. ( (2003) ) Crystal structure of the human CCA-adding enzyme: insights into template-independent polymerization. J. Mol. Biol., , 328, , 985–994.
Betat,H., Rammelt,C., Martin,G. and Moerl,M. ( (2004) ) Exchange of regions between bacterial poly(A) polymerase and CCA adding enzyme generates altered specificities. Mol. Cell, , 15, , 389–398.
Tomari,Y., Hino,N., Nagaike,T., Suzuki,T. and Ueda,T. ( (2003) ) Decreased CCA-addition in human mitochondrial tRNAs bearing a pathogenic A4317G or A10044G mutation. J. Biol. Chem., , 278, , 16828–16833.
Spacciapoli,P., Doviken,L., Mulero,J.J. and Thurlow,D.L. ( (1989) ) Recognition of tRNA by the enzyme ATP/CTP:tRNA nucleotidyl transferase. J. Biol. Chem., , 264, , 3799–3805.
Spacciapoli,P. and Thurlow,D.L. ( (1990) ) Purines in tRNAs required for recognition by ATP/CTP:tRNA nucleotidyltransferase from rabbit liver. J. Mol. Recognit., , 3, , 149–155.
Hegg,L.A. and Thurlow,D.L. ( (1990) ) Cytidines in tRNAs that are required intact by ATP/CTP:tRNA nucleotidyltransferases from Escherichia coli and Saccharomyces cerevisiae. Nucleic Acids Res., , 18, , 5975–5979.
Shi,P.-Y., Maizels,N. and Weiner,A.M. ( (1998) ) CCA addition by tRNA nucleotidyltransferase: polymerization without translocation? EMBO J., , 17, , 3197–3206.
Sissler,M., Pütz,J., Fasiolo,F. and Florentz,C. ( (2004) ) Chapter 28. In Ibba,M., Francklyn,C. and Cusack,S. (eds), The Aminoacyl-tRNA Synthetases. Landes Biosciences, Georgetown, TX.
Kumazawa,Y., Himeno,H., Miura,K.-I. and Watanabe,K. ( (1991) ) Unilateral aminoacylation specificity between bovine mitochondria and eubacteria. J. Biochem., , 109, , 421–427.
Giegé,R., Sissler,M. and Florentz,C. ( (1998) ) Universal rules and idiosyncratic features in tRNA identity. Nucleic Acids Res., , 26, , 5017–5035.
Giegé,R. and Frugier,M. ( (2003) ) In Lapointe,J. and Brakier-Gringas,L. (eds), Translation Mechanisms. Landes Sciences, Georgetown, TX, pp. 1–24.
Yokogawa,T., Shimada,N., Takeuchi,N., Benkowski,L., Suzuki,T., Omori,A., Ueda,T., Nishikawa,K., Spremulli,L.L. and Watanabe,K. ( (2000) ) Characterization and tRNA recognition of mammalian mitochondrial seryl-tRNA synthetase. J. Biol. Chem., , 275, , 19913–19920.
Shimada,N., Suzuki,T. and Watanabe,K. ( (2001) ) Dual mode of recognition of two isoacceptor tRNAs by mammalian mitochondrial seryl-tRNA synthetase. J. Biol. Chem., , 276, , 46770–46778.
Degoul,F., Brulé,H., Cepanec,C., Helm,M., Marsac,C., Leroux,J.-P., Giegé,R. and Florentz,C. ( (1998) ) Isoleucylation properties of native human mitochondrial tRNAIle and tRNAIle transcripts. Implications for cardiomyopathy-related point mutations (4269, 4317) in the tRNAIle gene. Hum. Mol. Gen., , 7, , 347–354.
Kelley,S., Steinberg,S. and Schimmel,P. ( (2000) ) Functional defects of pathogenic human mitochondrial tRNAs related to structural fragility. Nature Struct. Biol., , 7, , 862–865.
Kelley,S.O., Steinberg,S.V. and Schimmel,P. ( (2001) ) Fragile T-stem in disease-associated human mitochondrial tRNA sensitizes structure to local and distant mutations. J. Biol. Chem., , 276, , 10607–10611.
Helm,M., Brulé,H., Degoul,F., Cepanec,C., Leroux,J.-P., Giegé,R. and Florentz,C. ( (1998) ) The presence of modified nucleotides is required for cloverleaf folding of a human mitochondrial tRNA. Nucleic Acids Res., , 26, , 1636–1643.
Helm,M., Giegé,R. and Florentz,C. ( (1999) ) A Watson–Crick base-pair-disrupting methyl group (m1A9) is sufficient for cloverleaf folding of human mitochondrial tRNALys. Biochemistry, , 38, , 13338–13346.
Sissler,M., Helm,M., Frugier,M., Giegé,R. and Florentz,C. ( (2004) ) Aminoacylation properties of pathology-related variants of human mitochondrial tRNALys variants. RNA, , 10, , 841–853.
Sohm,B., Frugier,M., Brulé,H., Olszak,K., Przykorska,A. and Florentz,C. ( (2003) ) Towards understanding human mitochondrial leucine aminoacylation identity. J. Mol. Biol., , 328, , 995–1010.
Roy,M.D., Wittenhagen,L.M., Vozzella,B.E. and Kelley,S.O. ( (2004) ) Interdomain communication between weak structural elements within a disease-related human tRNA. Biochemistry, , 43, , 384–392.
Wittenhagen,L.M., Roy,M.D. and Kelley,S.O. ( (2003) ) The pathogenic U3271C human mitochondrial tRNALeu(UUR) mutation disrupts a fragile anticodon stem. Nucleic Acid Res., , 31, , 596–601.
Sohm,B., Sissler,M., Park,H., King,M.P. and Florentz,C. ( (2004) ) Recognition of human mitochondrial tRNALeu(UUR) by its cognate leucyl-tRNA synthetase. J. Mol. Biol., , 339, , 17–29.
Brandsma,M., Janssen,M. and Moller,W. ( (1997) ) Termination of Quiescence in Crustacea. The role of transfer RNA aminoacylation in the brine shrimp Artemia. J. Biol. Chem., , 272, , 28912–28917.
Sorensen,M.A. ( (2001) ) Charging levels of four tRNA species in Escherichia coli Rel(+) and Rel(–) strains during amino acid starvation: a simple model for the effect of ppGpp on translational accuracy. J. Mol. Biol., , 307, , 785–798.
Enriquez,J.A., Chomyn,A. and Attardi,G. ( (1995) ) MtDNA mutation in MERRF syndrome causes defective aminoacylation of tRNALys and premature translation termination. Nature Genet., , 10, , 47–55.
Hao,H. and Moraes,C.T. ( (1997) ) A disease-associated G5703A mutation in human mitochondrial DNA causes a conformational change and a marked decrease in steady-state levels of mitochondrial tRNAAsn. Mol. Cell. Biol., , 17, , 6831–6837.
B?rner,G.V., Zeviani,M., Tiranti,V., Carrara,F., Hoffmann,S., Gerbitz,K.D., Lochmuller,H., Pongratz,D., Klopstock,T., Melberg,A. et al. ( (2000) ) Decreased aminoacylation of mutant tRNAs in MELAS but not in MERRF patients. Hum. Mol. Genet., , 9, , 467–475.
Chomyn,A., Enriquez,J.A., Micol,V., Fernandez-Silva,P. and Attardi,G. ( (2000) ) The mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episode syndrome-associated human mitochondrial tRNALeu(UUR) mutation causes aminoacylation deficiency and concomitant reduced association of mRNA with ribosomes. J. Biol. Chem., , 275, , 19198–19209.
Florentz,C. and Sissler,M. ( (2001) ) Disease-related versus polymorphic mutations in human mitochondrial tRNAs: where is the difference? EMBO Rep., , 2, , 481–486.
Leontis,N.B. and Westhof,E. ( (1998) ) Conserved geometrical base-pairing patterns in RNA. Q. Rev. Biophys., , 31, , 399–455.
Agris,P.F. ( (1996) ) The importance of being modified: roles of modified nucleosides and Mg2+ in RNA structure and function. Prog. Nucleic Acid Res. Mol. Biol., , 53, , 79–129.
M?rl,M., D?rner,M. and P??bo,S. ( (1995) ) C to U editing and modifications during the maturation of the mitochondrial tRNAAsp in marsupials. Nucleic Acids Res., , 23, , 3380–3384.
Hino,N., Suzuki,T., Yasukawa,T., Seio,K., Watanabe,K. and Ueda,T. ( (2004) ) The pathogenic A4269G mutation in human mitochondrial tRNAIle alters the T-stem structure and decreases the binding affinity for elongation factor Tu. Genes Cells, , 9, , 243–252.
Wittenhagen,L.M. and Kelley,S.O. ( (2002) ) Dimerization of a pathogenic human mitochondrial tRNA. Nature Struct. Biol., , 9, , 586–590.(Louis Levinger1,2,*, Mario M?rl3 and Cat)