**1. Introduction**

Transfer RNAs are key partners in the ribosome-translation machinery. Generally, they are composed of c.70–90 nucleotides (nts). Moreover, they are the most abundant nucleic acid

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2018 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

species, constituting up to 10% of all cellular RNAs [1]. Therewith, the number of tRNA molecules is, e.g., about 2 × 105 in *Escherichia coli* and 3 × 10<sup>6</sup> in yeast cell [2]. Due to their anticodon, they read genetic information on mRNAs and deliver codon specified amino acids attached to their distal 3′-extremity for peptide bond synthesis on the ribosome. In this sense, tRNA is a key molecule which makes it possible to pass from a covalent bond between a RNA and an amino acid (fossil trace of the RNA world to the RNA/protein world transition) to peptide bonds (RNA/protein world). Genes specifying tRNAs (noted *trn*) are present in prokaryotic and nuclear genomes and in most of the DNAs of organelles (chloroplasts and mitochondria). Usually, tRNAs have a characteristic canonical cloverleaf secondary structure made up of the aminoacyl acceptor-stem and the D-arms (as it contains dihydrouridine), anticodon-arms, and T-arms (for the sequence TΨC where Ψ is pseudouridine), the hairpins, or "arms" consisting of a stem (helicoidal region in 3D) ending in a loop (**Figure 1**). The lengths of each arm, as well as the loop "diameter," vary from the tRNA type and from species to species. Furthermore, deduced *trn* sequences and even sequenced mature tRNAs exhibit reduced D-arms or T-arms or even lacking at least one of them, and in the extreme situation such as in *Enoplea* (nematodes) mitochondrial (mt)-*trn* genes are totally armless [3]. However, around 90% of the mttRNAs fold into the canonical cloverleaf structure [4]. In all the genetic systems, the tRNAs can carry a myriad of idiosyncratic posttranscriptional chemical modifications (e.g., http:// modomics.genesilico.pl/; http://www.genesilico.pl/rnapathwaysdb/), and the total number of modified nts is nearly 120 [1]. Moreover, tRNAs become functional by postprocessing addition of the 3′-terminal CCA sequence. Modifications can also be necessary to ensure correct folding [4]. The tRNA folds into an L-shaped 3D structure in which two helical domains (acceptor/T and D/anticodon) are perpendicularly arranged. This particular juxtaposition of the two functional centers, the anticodon and the acceptor terminus, is essential for tRNA function. The two domains are linked together by connector regions, one between the acceptor- and D-stems (connector 1) and the second between the anticodon- and T-stems (connector 2 which has a

True Mitochondrial tRNA Punctuation and Initiation Using Overlapping Stop and Start Codons…

http://dx.doi.org/10.5772/intechopen.75555

5

The ribosome allows the best possible spatial arrangement of the various partners and ensures catalysis, but the adaptor molecule which acts as a link between codes of mRNA and amino acids of polypeptides is the tRNA. In order to fill this major role, tRNAs have two distinct characteristics corresponding to two different genetic codes, the anticodon and the operational codes. The latter which is mainly embodied in the acceptor-stem allows to bind covalently and with high specificity an amino acid to a tRNA, a reaction catalyzed by a specific aminoacyl-tRNA synthetase (aaRS) [6]. The operational code might have actually predated the "classic" code associated with anticodons [7]. Moreover, the tRNAs exhibit diversity in uniqueness, all of them must be similar for entering the ribosome machinery; therefore, they generally look structurally homogeneous, especially in their secondary and tertiary structures even if "non-classical" tRNAs are known [3]. Moreover, cloverleaf structure and especially the tertiary interaction network governing the L-shaped tRNA architecture imply conserved and semiconserved bps and nts. On the other side, each type of tRNA structures must interact specifically with aaRSs and posttranscriptional modification enzymes, which implies that parts of their sequences and of their structures (as the V-R size) allow to distinguish them.

Reduced bacterial and most organelle genomes do not encode the full set of 32 tRNA species required to read all triplets of the standard genetic code according to the conventional wobble rules. Superwobbling where a single tRNA species contains modifications of the anticodon-loop, such as an hypermodified uridine at the wobble position 34 of the anticodon, reads all 4 nts at third codon position and has been suggested as a possible mechanism for how reduced tRNA sets may be functional [8]. Indeed, many metazoan mtDNAs have only a total of 22 tRNAs, apparently sufficient to recognize all codons (two tRNAs each for serine and leucine and one tRNA for each of the other 18 amino acids). However, superwobbling induces a reduced translational efficiency, which could explain why most organisms have adopted pairs of isoaccepting tRNAs over the superwobbling mechanism [9]. Moreover, e.g., in Cnidaria (sea anemones, corals, etc.) or Chaetognatha (marine invertebrates), current mtD-NAs have lost several of their *trn* genes, and the absence of an apparently full set of mt-*trn* genes has also been mentioned [10]. Studies have investigated the fate of missing tRNAs and their corresponding aaRSs [11], and in many cases, the lost tRNAs are functionally replaced by imported nucleus-encoded tRNAs [10]. However, recent search strategies suggest that efficient reanalyzes detect several tRNA-like structures (TLS), which can be efficient tRNAs [12]. Compared to mitochondria found in other eukaryotic kingdoms, those of metazoa are massively reduced in their genetic structure [4]. Their mtDNA is a short, circular molecule that generally contains about 13 intronless, protein-coding genes, all of which are involved in aerobic respiration (also called oxidative phosphorylation) [13]. Moreover, the coding sequences of genes are usually separated by at most a few nts and long polycistronic precursor transcripts may be processed into mature mRNA and rRNA by precise cleavage of the 5′ and 3′-termini of the flanking tRNAs. This processing, which is known as the tRNA punctuation model [14], is mediated by RNase P and Z endonucleases, respectively [15]. However, this model is not always applicable, genes are not bound by *trn* genes or these latter may not be involved in the

variable length from 0 to 21 nts and is also named variable region hereafter V-R) [5].

**Figure 1.** Typical cloverleaf secondary structure of a metazoan mt-ss-tRNA (left) with 3D image of an L-shaped tRNA (right). In 2D structure, the standard numbering was applied [5]. The first two nucleotides of the variable region and those of the D-loops and T-loops were represented by circles. The diagonal dashed line indicates the approximate separation between the "top half" and the "cherry-bob"/"bottom half". Nucleotide types were given for UAG10 and AUG49 triplets, the discriminator base (which is preferentially an A), and the CCA tail at the 3′-end. Short lines connect nucleotides forming Watson-Crick pairing within stems. Coloring: acceptor-stem in purple, D-arm in red, anticodon-arm in blue with the anticodon in black, T-arm in green, and CCA tail in orange. The yellow segments represented respectively in descending order of size, the variable region (connector 2), the connector 1 and the nt 26. 3D structure reproduced with the kind permission of Prof. N.R. Voss (Roosevelt University, Ill.) https://commons.wikimedia.org/wiki/File:3d\_tRNA.png.

two domains are linked together by connector regions, one between the acceptor- and D-stems (connector 1) and the second between the anticodon- and T-stems (connector 2 which has a variable length from 0 to 21 nts and is also named variable region hereafter V-R) [5].

species, constituting up to 10% of all cellular RNAs [1]. Therewith, the number of tRNA mol-

they read genetic information on mRNAs and deliver codon specified amino acids attached to their distal 3′-extremity for peptide bond synthesis on the ribosome. In this sense, tRNA is a key molecule which makes it possible to pass from a covalent bond between a RNA and an amino acid (fossil trace of the RNA world to the RNA/protein world transition) to peptide bonds (RNA/protein world). Genes specifying tRNAs (noted *trn*) are present in prokaryotic and nuclear genomes and in most of the DNAs of organelles (chloroplasts and mitochondria). Usually, tRNAs have a characteristic canonical cloverleaf secondary structure made up of the aminoacyl acceptor-stem and the D-arms (as it contains dihydrouridine), anticodon-arms, and T-arms (for the sequence TΨC where Ψ is pseudouridine), the hairpins, or "arms" consisting of a stem (helicoidal region in 3D) ending in a loop (**Figure 1**). The lengths of each arm, as well as the loop "diameter," vary from the tRNA type and from species to species. Furthermore, deduced *trn* sequences and even sequenced mature tRNAs exhibit reduced D-arms or T-arms or even lacking at least one of them, and in the extreme situation such as in *Enoplea* (nematodes) mitochondrial (mt)-*trn* genes are totally armless [3]. However, around 90% of the mttRNAs fold into the canonical cloverleaf structure [4]. In all the genetic systems, the tRNAs can carry a myriad of idiosyncratic posttranscriptional chemical modifications (e.g., http:// modomics.genesilico.pl/; http://www.genesilico.pl/rnapathwaysdb/), and the total number of modified nts is nearly 120 [1]. Moreover, tRNAs become functional by postprocessing addition of the 3′-terminal CCA sequence. Modifications can also be necessary to ensure correct folding [4]. The tRNA folds into an L-shaped 3D structure in which two helical domains (acceptor/T and D/anticodon) are perpendicularly arranged. This particular juxtaposition of the two functional centers, the anticodon and the acceptor terminus, is essential for tRNA function. The

**Figure 1.** Typical cloverleaf secondary structure of a metazoan mt-ss-tRNA (left) with 3D image of an L-shaped tRNA (right). In 2D structure, the standard numbering was applied [5]. The first two nucleotides of the variable region and those of the D-loops and T-loops were represented by circles. The diagonal dashed line indicates the approximate separation between the "top half" and the "cherry-bob"/"bottom half". Nucleotide types were given for UAG10 and AUG49 triplets, the discriminator base (which is preferentially an A), and the CCA tail at the 3′-end. Short lines connect nucleotides forming Watson-Crick pairing within stems. Coloring: acceptor-stem in purple, D-arm in red, anticodon-arm in blue with the anticodon in black, T-arm in green, and CCA tail in orange. The yellow segments represented respectively in descending order of size, the variable region (connector 2), the connector 1 and the nt 26. 3D structure reproduced with the kind permission of Prof. N.R. Voss (Roosevelt University, Ill.) https://commons.wikimedia.org/wiki/File:3d\_tRNA.png.

in yeast cell [2]. Due to their anticodon,

in *Escherichia coli* and 3 × 10<sup>6</sup>

ecules is, e.g., about 2 × 105

4 Mitochondrial DNA - New Insights

The ribosome allows the best possible spatial arrangement of the various partners and ensures catalysis, but the adaptor molecule which acts as a link between codes of mRNA and amino acids of polypeptides is the tRNA. In order to fill this major role, tRNAs have two distinct characteristics corresponding to two different genetic codes, the anticodon and the operational codes. The latter which is mainly embodied in the acceptor-stem allows to bind covalently and with high specificity an amino acid to a tRNA, a reaction catalyzed by a specific aminoacyl-tRNA synthetase (aaRS) [6]. The operational code might have actually predated the "classic" code associated with anticodons [7]. Moreover, the tRNAs exhibit diversity in uniqueness, all of them must be similar for entering the ribosome machinery; therefore, they generally look structurally homogeneous, especially in their secondary and tertiary structures even if "non-classical" tRNAs are known [3]. Moreover, cloverleaf structure and especially the tertiary interaction network governing the L-shaped tRNA architecture imply conserved and semiconserved bps and nts. On the other side, each type of tRNA structures must interact specifically with aaRSs and posttranscriptional modification enzymes, which implies that parts of their sequences and of their structures (as the V-R size) allow to distinguish them.

Reduced bacterial and most organelle genomes do not encode the full set of 32 tRNA species required to read all triplets of the standard genetic code according to the conventional wobble rules. Superwobbling where a single tRNA species contains modifications of the anticodon-loop, such as an hypermodified uridine at the wobble position 34 of the anticodon, reads all 4 nts at third codon position and has been suggested as a possible mechanism for how reduced tRNA sets may be functional [8]. Indeed, many metazoan mtDNAs have only a total of 22 tRNAs, apparently sufficient to recognize all codons (two tRNAs each for serine and leucine and one tRNA for each of the other 18 amino acids). However, superwobbling induces a reduced translational efficiency, which could explain why most organisms have adopted pairs of isoaccepting tRNAs over the superwobbling mechanism [9]. Moreover, e.g., in Cnidaria (sea anemones, corals, etc.) or Chaetognatha (marine invertebrates), current mtD-NAs have lost several of their *trn* genes, and the absence of an apparently full set of mt-*trn* genes has also been mentioned [10]. Studies have investigated the fate of missing tRNAs and their corresponding aaRSs [11], and in many cases, the lost tRNAs are functionally replaced by imported nucleus-encoded tRNAs [10]. However, recent search strategies suggest that efficient reanalyzes detect several tRNA-like structures (TLS), which can be efficient tRNAs [12].

Compared to mitochondria found in other eukaryotic kingdoms, those of metazoa are massively reduced in their genetic structure [4]. Their mtDNA is a short, circular molecule that generally contains about 13 intronless, protein-coding genes, all of which are involved in aerobic respiration (also called oxidative phosphorylation) [13]. Moreover, the coding sequences of genes are usually separated by at most a few nts and long polycistronic precursor transcripts may be processed into mature mRNA and rRNA by precise cleavage of the 5′ and 3′-termini of the flanking tRNAs. This processing, which is known as the tRNA punctuation model [14], is mediated by RNase P and Z endonucleases, respectively [15]. However, this model is not always applicable, genes are not bound by *trn* genes or these latter may not be involved in the processing of precursor RNAs. Besides, in several taxa mt-mRNAs, rRNAs and even tRNAs may be oligoadenylated or polyadenylated [16]. This has numerous consequences with potentially dual and opposite roles: this promotes transcript stability or offers a target for initiating degradation. Overlapping genes on the same DNA strand occur throughout metazoa [17]. Therefore, the termination points of the protein-encoding genes could be difficult to infer as stop codons (generally UAA or UAG) may be absent. It is accepted that abbreviated stop codons (U or UA) are converted to UAA codons by polyadenylation after transcript cleavage, and this has been confirmed by analyzes of transcripts in some cases [18]. Sometimes, the initiation codon may also not have been detected. For several protein-encoding genes, the question of a possible overlapping with adjacent downstream or upstream *trn* genes is often raised [19]. Moreover, overlaps between adjacent mt-*trn* genes are frequent, but it is out of our topic [19, 20].

Incidentally, in 2004, searching for chaetognath mt-*trn* genes [21], it was observed that tRNAs bear nt triplets corresponding to stop or start codons at precise conserved positions, and this constitutes the original topic of this chapter.
