Comparative genomics of Mycobacterium mucogenicum and Mycobacterium neoaurum clade members emphasizing tRNA and non-coding RNA

Background Mycobacteria occupy various ecological niches and can be isolated from soil, tap water and ground water. Several cause diseases in humans and animals. To get deeper insight into our understanding of mycobacterial evolution focusing on tRNA and non-coding (nc)RNA, we conducted a comparative genome analysis of Mycobacterium mucogenicum (Mmuc) and Mycobacterium neoaurum (Mneo) clade members. Results Genome sizes for Mmuc- and Mneo-clade members vary between 5.4 and 6.5 Mbps with the complete MmucT (type strain) genome encompassing 6.1 Mbp. The number of tRNA genes range between 46 and 79 (including one pseudo tRNA gene) with 39 tRNA genes common among the members of these clades, while additional tRNA genes were probably acquired through horizontal gene transfer. Selected tRNAs and ncRNAs (RNase P RNA, tmRNA, 4.5S RNA, Ms1 RNA and 6C RNA) are expressed, and the levels for several of these are higher in stationary phase compared to exponentially growing cells. The rare tRNAIleTAT isoacceptor and two for mycobacteria novel ncRNAs: the Lactobacillales-derived GOLLD RNA and a homolog to the antisense Salmonella typhimurium phage Sar RNA, were shown to be present and expressed in certain Mmuc-clade members. Conclusions Phages, IS elements, horizontally transferred tRNA gene clusters, and phage-derived ncRNAs appears to have influenced the evolution of the Mmuc- and Mneo-clades. While the number of predicted coding sequences correlates with genome size, the number of tRNA coding genes does not. The majority of the tRNA genes in mycobacteria are transcribed mainly from single genes and the levels of certain ncRNAs, including RNase P RNA (essential for the processing of tRNAs), are higher at stationary phase compared to exponentially growing cells. We provide supporting evidence that Ms1 RNA represents a mycobacterial 6S RNA variant. The evolutionary routes for the ncRNAs RNase P RNA, tmRNA and Ms1 RNA are different from that of the core genes. Electronic supplementary material The online version of this article (10.1186/s12862-019-1447-7) contains supplementary material, which is available to authorized users.


Background
Mycobacteria are divided into rapid and slow growing mycobacteria, RGM and SGM. Among SGM, Mycobacterium tuberculosis (Mtb) causes tuberculosis (TB) while the non-pathogenic RGM Mycobacterium smegmatis (Msmeg) is frequently used as a mycobacterial model system. Both SGM and RGM can inhabit various environmental reservoirs such as ground and tap water, soil, animals and humans. They can form aggregates and biofilms and appear to have a growth advantage in water that contains disinfecting agents [1][2][3][4].
The RGM Mycobacterium mucogenicum (Mmuc) was first reported as a Mycobacterium chelonae-like bacteria in connection with a peritonitis outbreak in 1982 [5][6][7]. On the basis of partial sequencing of the 16S rRNA gene Mmuc was later proposed to be a new taxon and the name reflects that it is highly mucoid when grown on solid media [3,[8][9][10]. M. mucogenicum together with Mycobacterium phocaicum (Mpho) and Mycobacterium aubagnense (Maub) constitute the Mmuc-clade [3,[9][10][11]. These RGM are water-borne and Mmuc is claimed to be one of the most abundant nontuberculosis mycobacteria (NTM) in tap water and it is found in sewage and in hospital water systems. They are opportunistic pathogens and have been demonstrated to be associated with various infections and they show high tolerance against first-line anti-tuberculosis drugs such as isoniazid, rifampin and pyrazinamide [3,5,7,[9][10][11][12][13][14][15][16][17][18][19][20]. Moreover, strains of the phylogenetically close mycobacteria Mycobacterium neoaurum (Mneo) and Mycobacterium cosmeticum (Mcos), which both belong to the Mneo-clade, have also been isolated from patients suffering from various infections [21][22][23][24]. Together this emphasizes the importance of this group of NTMs and provided an incentive for a comparative genomic analysis of these closely related species.
The tRNA genes are implicated to be targets for integration of foreign DNA [25,26] and in bacteria their number and gene synteny vary. For example, the 4.6 Mbp long Escherichia coli K12 MG1655 genome carries 86 while Streptomyces coelicolor (genome size, 8.7 Mbp) encodes 65 tRNA genes [27,28]. On the basis of available data, it appears that the number of tRNA genes in several mycobacteria such as MtbH37Rv (genome size, 4.4 Mbp) and Msmeg (genome size, 7 Mbp) do not exceed 50 [29][30][31][32]. Given the seemingly low number of tRNA genes among mycobacteria we were interested to understand whether this is also the case for other mycobacteria. Moreover, a preliminary survey of tRNA gene organization in mycobacteria such as MtbH37Rv and Msmeg indicates that many tRNA are transcribed from single genes, which is in contrast to other bacteria, e.g., E. coli and Bacillus subtilis [27,29,30,33]. We were therefore interested in finding whether this might also apply to other mycobacteria focusing on the NTM discussed above. In this context, our understanding of non-coding (nc) RNA in mycobacteria, apart from MtbH37Rv and Msmeg [34], is scarce. Hence, we also decided to conduct a comparative genomic analysis focusing on ncRNA genes in Mmucand Mneo-clade members to understand absence and presence of ncRNA genes with respect to phylogeny and their evolution. We were particularly interested to study ncRNAs with implicated functions in RNA processing, gene regulation and gene expression such as RNase P RNA, the catalytic subunit of the tRNA processing endoribonuclease P, RNase P [35].
We report that genome sizes among Mmucand Mneo-clade members, including the complete genome of the Mmuc type strain, vary between approximately 5.4 to 6.5 Mbp (see also the NCBI database). Comparative genomic analysis (17 genomes in total) provide some insight into how some of the characteristics of these genomes such as genome size, common and unique genes, and horizontal gene transfer (HGT) might be manifested as phenotypic differences. Specifically, the number of tRNA genes vary between 46 and 79 (including one pseudo tRNA gene) with Maub having the highest number; 39 tRNA genes were predicted to be present in all studied mycobacteria. For those species having higher number of tRNA genes our data suggest that several have been acquired through HGT and we show that some of these are expressed. Moreover, predicted ncRNA genes range between 28 and 46, and while some are species specific others are also present in MtbH37Rv and Msmeg, and for some ncRNAs, e.g. RNase P RNA, the evolutionary routes are different from that of the core genes. Finally, we provide supportive data suggesting that one of the ncRNAs, Ms1 RNA, represents a mycobacterial 6S RNA variant, which in other bacteria acts as a global regulator.

Strains and cultivation
The M. mucogenicum DSM44124 (Mmuc T ), M. phocaicum DSM45104 (Mpho T ), M. aubagnense DSM45150 (Maub T ), M. neoaurum DSM44074 (Mneo T ) and M. cosmeticum DSM44829 (Mcos T ) type strains were obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen in Germany and grown under conditions as recommended by the supplier. The DNA was isolated and purified as previously described [31].

Genome sequencing, assembly and annotation
The Mmuc T whole genome sequencing was performed at the NGI-Uppsala Genome Center using the PacBio technology and the average read length was 10936 bp and the median coverage ≈101x. The genomes of the other four Mycobacterium spp., Mpho T , Maub T , Mneo T and Mcos T , were sequenced at the SNP&SEQ Technology Platform using the Illumina-Hiseq2000 platform at Uppsala University, Uppsala with an average read length of 100 bp and a median coverage of ≈176x. The PacBiogenerated reads were assembled using the SMARTanalysis HGAP3 assembly pipeline [36] and polished using Quiver (Pacific Biosciences, Menlo Park, CA, USA). The Illumina generated reads were assembled using the A5-assembly pipeline [37,38] as previously described [31]. For the annotation of rRNA genes we used RNAmmer version 1.2 [39] and for annotation of tRNA genes the tRNAScan-SE program version 1.23 [40]. Small ncRNA genes were predicted using the Rfam version 12.1 database and INFERNAL (version 1.1.2) with a threshold cut off value of ≥34 [41][42][43][44][45].

Plasmids, Phage DNA and IS-elements
To predict the presence of plasmid sequences, the scaffolds of the different mycobacterial draft genomes were sorted and aligned using the closest reference genome (Mmuc T ; complete genome) and the Mauve software (version 2.3.1 [49]). The aligned scaffolds were subsequently concatenated and subjected to BLAST search against the NCBI plasmid database (ftp://ftp.ncbi. nlm.nih.gov/refseq/release/plasmid/, accessed March 2015; May 2019).
Phage sequences were predicted using the PHAGE database at the PHASTER server (PHAge Searching Tool server; [50,51]).
The presence of IS elements in the complete Mmuc T genome was predicted using the ISsaga server and manually inspected searching for inverted repeats and presence of transposase genes [52].

Core gene analysis
Translated protein coding sequences (CDS) were extracted for the target genomes and combined with CDSs from available Mycobacterium spp. to generate a BLAST database. A reciprocal BLAST was performed using the BLASTp and the PanOCT (version 1.9) with the following parameter settings, identity of at least 45%, query coverage >70% and e-value cutoff 1e-05 [32]. The predicted orthologous genes were also subjected to gene synteny analysis using PanOCT [53].

Phylogenetic analysis and average nucleotide identity
Single gene phylogenies were generated based on the 16S rRNA gene, rnpB, ssrA, ffh and the Ms1 RNA gene.
The respective gene sequences were extracted from the five genomes (Mmuc T , Mpho T , Maub T , Mneo T and Mcos T ) and 104 publicly available mycobacterial genomes from the NCBI data base. The homologous genes were identified using the PanOCT ortholog. Single (16S rDNA, rnpB, ssrA, ffh and the Ms1 RNA gene) and core genes were aligned using the MAFFT (version 7.147b) software [54]. Phylogenetic trees based on 16S rDNA, core genes and multiple sequence alignments were computed using the FastTree version 2.1.7 [55] and figures were generated using the FigTree Software version 1.3.1 (http://tree.bio.ed.ac.uk/software/figtree/).
The Jspecies tool was used to calculate the average nucleotide identity (ANI) values. ANI values were clustered using an unsupervised hierarchical clustering algorithm and plotted using "R" environment [56][57][58].

RNA extraction and northern analysis
Mmuc T and Maub T were grown in liquid 7H9 medium (including 10% OADC supplement and 0.25% Tween 80) at 37°C and 100 rpm rotation. At exponential (OD 600 = 0.3-0.5) and stationary (OD 600 = 5-6) growth phases, cells were pelleted and snap frozen in liquid nitrogen. RNA was extracted following the protocol described elsewhere [59]. Triplicate samples/independent cultures (10 μg RNA from Maub T and 15 μg from Mmuc T per sample) from each growth phase were then separated on an 8% polyacrylamide gel and electroblotted to a Hybond N+ membrane as previously described [60]. The membranes were then probed with 32 P-5'-end-labelled DNA oligonucleotide, washed, and exposed to a phosphorimager screen. See Additional file 1: Table S1 for DNA oligonucleotide sequences and hybridization temperatures (for specifics see [60]).

Genome assembly and annotation
The size of the complete Mmuc T genome is 6,099,273 while the sizes for the draft Mpho T , Maub T , Mneo T and Mcos T genomes are 5,771,543; 6,191,633; 5,477,923 and 6, 446,208 base pairs, respectively ( Fig. 1a; Table 1). The GCcontent for the five species was calculated to be in the range 66.3% to 68.3%. The complete Mmuc T genome includes 5881 protein-coding sequences (CDS), two ribosomal RNA operons, 57 tRNA genes and 33 non-coding (nc) RNA genes. For the other four species, the number of CDS range from 5199 to 6270 genes, which correlates with their respective genome size (Table 1). Moreover, for Mneo T and Mcos T 46 tRNA genes were identified, while Mpho T and Maub T carry 60 and 79 (among these one pseudo tRNA gene) tRNA genes, respectively. The Rfam annotated ncRNA genes in Mpho T , Maub T , Mneo T and Mcos T were 45, 41, 34 and 28, respectively (Table 1; in total 14 mycobacteria analyzed). In Table 1 we give a basic  Table 1 and Additional file 1: Table S2.  (Fig. 1b). However, we cannot exclude that this is due to draft genome status. Alignment of the 16 genomes (except M. sp. URHB0044) is presented in Additional file 1: Figure  S1 (note that we analyzed two versions of Mcos DSM44829 and Mneo DSM44074 ). Unless otherwise stated, below we focus on the five type strains, Mmuc T , Mpho T , Maub T , Mneo T and Mcos T . We did not detect any plasmids or plasmid fragments in any of the five type strains (Mmuc T , Mpho T , Maub T , Mneo T and Mcos T ) while phage sequence stretches were predicted to be present (Table 1; for details see Additional file 1: Table S3). In Mmuc T three intact prophage regions were predicted using PHASTER (score, >90; see Methods). These regions were not detected in any of the other four species. Two additional phage segments, classified as "questionable" (score, 70 to (See figure on previous page.) Fig. 1 Complete genome of Mmuc T . a Circos plot showing the complete genome sequence of Mmuc T . The outer to inner circles represent: Red and blue dots represent genome wide distribution of tRNA and ncRNA genes, while the two black lines represent the rRNA operons. Green histogram represents the average sequencing read depth for Mmuc T , while the overlapping grey circle gives the read depth scale (the distance between the two circles is 50x). The blue and green blocks in the two subsequent circles represent genes that are predicted to be transcribed from the leading and lagging strands, respectively. Dark blue and brown blocks indicate core and unique genes comparing Mmucand Mneo-clade members. Next inner circle, black line marks the positioning of the IS-elements. The four blocks represent the phage sequences. The GC-content is represented by the blue and grey "spikes". It was calculated using a sliding window of 5000 bp and the grey circles give the scale, which oscillates between ±10% of the mean value 67.18%. Next track in red (positive) and green (negative) shows the GC skew using a sliding window of 5000 bp. The inner circle shows the size of the Mmuc T genome. Generation of circos plots, see http://circos.ca. b Whole-genome alignment for the five type strains Mmuc T , Mpho T , Maub T , Mneo T and Mcos T . Each horizontal block represents one genome and vertical lines between the genome correspond to homologous regions whereas diagonal lines correspond to genome rearrangements (blue lines genomic inversions and white gaps represent insertions/deletions). The black boxes/vertical lines represent phage sequences; single stars in red mark the presence of an intact phage and black stars represent incomplete/questionable pro-phage sequence (see text for details)  [50,51]). For the complete Mmuc T genome we predicted 23 IS elements (for two of these we were unable to identify inverted repeat sequences) representing eight different types (Additional file 1: Table S4a). IS-elements were also predicted in the Mpho T , Maub T , Mneo T and Mcos T draft genomes (Additional file 1: Table S4b), but their precise locations were not defined. For these four strains the number of IS elements vary and Mpho T carries 65 predicted IS elements. The highest number was detected in Mmuc LZSF01 for which we predicted 85 IS elements (Additional file 1: Table S4b).
Taken together this suggests that phage and IS elements likely have had an impact on the evolution of these type strains.

Identification of core genes and phylogenetic analysis
To construct phylogenetic trees, we first used 291 core genes present in Mmuc T , Mpho T , Maub T , Mneo T and Mcos T and 104 mycobacterial genomes available at the NCBI genome database (ftp://ftp.ncbi.nlm.nih.gov/geno mes/). Second, to get deeper insights into the phylogenetic relationship within the Mmucand Mneo-clades we identified 2226 core genes predicted to be present in the members of these two clades, including the recently released draft Mmuc genomes (strains Mmuc LZSF01 , Mmuc LZLC01 and Mmuc CSURP2099 ; Additional file 1: Figure S1), and in M. sp. URHB0044 (in total 17 species; Table 1; M. sp. URHB0044 was chosen as an outgroup). Functional classification of the 291 and 2226 core genes revealed that the majority (>60%) of the classified genes belongs to the subsystems "Amino Acids and Derivatives", "Cofactors, Vitamins…Pigments", "Carbohydrates", "Protein Metabolism" and "Fatty Acids, Lipids, and Isoprenoids" (Additional file 2: Figure S2a, b).
Both core phylogenetic trees are in agreement and suggested that Mmuc T , Mpho T and Maub T are close neighbors to four other Mycobacterium spp., M. sp. 360MFT, M. sp. UNC410, M. llatzerense strain CLUC14 and M. sp. UNC280 (referred to as Msp360, Msp410, Mllat, and Msp280, respectively; Table 1). Hence it appears that these mycobacteria belong to the Mmuc-clade [3]. The Mneo T and Mcos T cluster together with other Mcos and Mneo strains in the Mneo-clade (Fig. 2a, b; Additional file 2: Figure S3a; [3]). Moreover, Mmuc LZLC01 and Mmuc CSURP2099 are positioned close to Mmuc T except Mmuc LZSF01 , which is closer to Mpho T (Fig. 2a, b). Also, inspection of the Mneo-clade suggested that the Mneo VK-MAc-1815D constitutes a separate branch of Mneo ( Fig. 2a; see Discussion; for comparison with 16S rDNA phylogeny see Additional file 2: Figure S3a, b).
Analysis of the average nucleotide identity, ANI, also grouped the 17 mycobacteria in accordance with their position in the core (2226 genes) phylogenetic tree (Fig. 2b, c).
Taken together, the phylogenetic data revealed that the positioning of these mycobacteria was in accordance with our current understanding of the mycobacterial phylogeny [3, 61-63; and our unpublished data]. The data further suggested that Mmuc LZSF01 is a Mpho strain and that Mneo VKMAc-1815D should be considered as a separate species or Mneo subspecies (see Discussion).

Number of tRNA genes, their organization and expression of selected tRNA genes
The number of predicted tRNA genes in Mmuc T , Mpho T , Maub T , Mneo T and Mcos T were 57, 60, 79 (including one pseudo tRNA gene), 46 and 46, respectively, with Maub T having the highest number (Table 1; Fig. 3; Additional file 3: Figure S4a). However, this is not related to which clade these species belong to since Msp280 and Mllat have 46 and 47, respectively, which is the same as observed for members of the Mneo-clade (Fig. 2). The different Mmuc strains (Mmuc T , Mmuc LZSF01 , Mmuc LZLC01 and Mmuc CSURP2099 ) all harbor the same number of predicted tRNA genes with the exception of Mmuc LZSF01 , which is positioned close to Mpho T (Table 1; Fig. 2). Of the 119 predicted tRNA genes, 39 are common to members of the Mmucand Mneo-clades and they cover all 20 amino acids (Additional file 3: Figure S4a).
Larger clusters of tRNA genes with more than four genes as in Mmuc T , Mpho T and Maub T are uncommon among mycobacteria ( Fig. 3; Additional file 3: Figure  S4b-e; Table S5; of note, clusters of tRNA genes with more than four genes were not found to be present in the Mneo-clade members; see Discussion). Among the 78 tRNA genes in Maub T 33 are unique and 32 were predicted to cluster in one region (the "32 tRNA gene" cluster) and transcribed from the same DNA (lagging) strand (tRNA genes constituting this cluster were present on three scaffolds but see Discussion; Additional file 3: Figure S4c).
Mmuc T and Mpho T carry a structurally different tRNA gene cluster with 11 and 14 tRNA genes, respectively (cf. Fig. 3 and Additional file 3: Figure S4b). Apart from the presence of the three extra genes in Mpho T , which encode for tRNA Trp CCA, tRNA Gly TCC and tRNA Leu CAA, the gene synteny of this tRNA gene cluster is the same in these two mycobacteria (Fig. 3b). Moreover, the structures (based on their gene sequences) for the majority of the tRNA genes comprising this cluster are the same with one notable exception, the tRNA His gene. In Mpho T this tRNA gene carries one additional nucleobase and "mutations" in the region corresponding to the anticodon stem, which results in a distorted stem (Additional file 3: Figure S5; alignment of all predicted tRNA genes). Compared to their corresponding tRNA isoacceptor genes located elsewhere in the genome predicted that all the tRNA isoacceptors originating from the cluster were structurally different. For example, the cluster encoded tRNA Leu TAG isoacceptor has a short variable loop whereas the isoacceptor present in all Mmucand  Figure S2a). *Mark the mycobacterial type strains for which the genomes were sequenced in the present study. b Phylogenetic tree based on 2226 core genes predicted to be present in all 17 mycobacteria. In (a) and (b) the percentage values in the nodes represent boot strap values generated by 1000 cycles. c Classification of the 17 mycobacteria using Jspecies. The dendogram show ANI values after clustering using unsupervised hierarchical clustering Mneo-clade members has an expected "normal" larger variable loop (Additional file 3: Figure S5).
Genes encoding the tRNA isoacceptors tRNA Ile TAT and tRNA Arg TCG were detected to be present (and part of a larger tRNA gene cluster) only in Mmuc T and Mpho T (see above), and in Msp360, Msp410, Mmuc CSURP2099 , Mmuc LZSF01 and Mmuc LZSC01 . Analysis of the gene synteny of this cluster in these mycobacteria (not shown for Mmuc LZSF01 and Mmuc LZLC01 ) predicted the presence of an HNH endonuclease gene within this cluster while this ortholog is missing in the other ten mycobacteria (Fig. 3b). The finding that Maub T lacks this tRNA gene cluster, in combination with that it is positioned closer to the ancestor (Fig. 2a, b; Additional file 2: Figure S3a) than the other Mmuc-clade members argues for the possibility that the tRNA gene cluster might have been acquired after Maub T diverged from the other Mmuc-clade members.
To understand whether the tRNA isoacceptors encoded from the tRNA gene cluster are expressed we focused on tRNA Ile TAT, tRNA Arg TCG and tRNA Ile GAT and included the tRNA Ile GAT isoacceptor encoded a b Fig. 3 tRNA gene organization and gene synteny. a Mapping of the 57 tRNA genes in Mmuc T . Blue boxes and arrows mark tRNA genes that are transcribed from the leading (+) strand, while red marks those that are transcribed from the lagging (-) strand. Numbers given in parenthesis indicate the numbers of nucleotides separating two tRNA genes. For orientation, we also marked the 40 predicted "transcriptional units", 1 to 40. b Gene synteny for "transcription units" 11 and 12, see (a), for selected mycobacteria carrying these tRNA genes. The tRNA genes are marked in green except the tRNA Ile TAT gene, which is marked in red. The vertical boxes marked in brown highlight homologous genes. Note the presence of the HNH endonuclease gene located between the two tRNA gene clusters (see main text). Generation of gene synteny plots, see http://genoplotr.r-forge.r-project.org/ from another location in Mmuc T . On the basis of the data presented in Fig. 4 we conclude that these four tRNA isoacceptors are expressed and their levels are higher in cells in stationary phase than in exponentially growing cells.
We previously reported the presence of two tRNA-Cys GCA genes, one of which is commonly present in mycobacteria while the other is present in some RGM [31]. Two tRNA Cys GCA genes were also predicted to be present in the members of the Mmucand Mneoclades while three were predicted for Maub T with one located within the "32 tRNA gene cluster". Gene alignment revealed that the structure of the third tRNA-Cys GCA gene differs from the other two isoacceptors (see above; Additional file 3: Figures S4c and S5). All three Maub T tRNA Cys isoacceptor genes were expressed and the levels were similar in exponentially growing and stationary cells, if anything, slightly increased in stationary phase cells ( Fig. 4; see also the Discussion).
It should be noted that genes encoding tRNA SelCys or selB appear not to be present in any of the Mmucand Mneo-clade member genomes. Also, no tRNA genes were predicted to be positioned in the ribosomal RNA operons.

Aminoacyl-tRNA synthetases, AARS
The tRNAs depend on the aminoacyl-tRNA synthetases (AARSs) for their function. We therefore analyzed members of the Mmucand Mneo-clades for the presence of AARS genes ( Fig. 5a; Additional file 3: Table S6a). While 18 AARS genes were predicted to be present in these genomes we were unable to identify any homologous genes for asparaginyl-tRNA synthetase (AsnRS) and glutaminyl-tRNA synthetase (GlnRS) in keeping with  Figure S4c; panels, IV-VI]. The RNA was extracted from stationary and exponential growing cells as described in Methods. M indicates a pUC19/MspI size marker with selected sizes indicated to the left of (I). Lanes 1-3 and 7-8 represent RNA from exponentially grown cells and lanes 4-6 and 9-11 represent RNA from stationary cells. The identities of the detected tRNAs are marked below the images. The boxed insets show the respective blots probed for 5S rRNA as indicated to the left of panel I. The lower panel I inset represents the same blot probed against tRNA Arg TCG. b Levels of tRNAs in Mmuc T and Maub T in stationary phase relative to exponentially growing cells. The bar plot shows a quantification of the Northern blot data shown in Fig. 5a. The amino acid three-letter code and the corresponding anticodon triplet mark the tRNA isoacceptor variants. The numbers on the x-axis corresponds to the numbering in (a), e.g. 17 correspond to Ile_GAT_17 and as indicated. The experiment was performed as outlined in Methods and values represent the mean and errors based on three independent experiments. Statistical significance using Student's unpaired, two-tailed t-test; *: p< 0.05; **: p < 0.01; ***: p< 0.001 previous findings [64,65]. For bacteria that lack AsnRS and GlnRS genes, tRNA charging of asparagine and glutamine can be accommodated using the tRNAdependent amidotransferase pathway (Adt), which includes GatCAB [66]. Accordingly, gatCAB homologs were predicted in the 16 genomes suggesting that the Adt pathway is operating in these mycobacteria (Additional file 3: Table S6b; of note, gatA was not predicted to be present in Msp410 possibly due to draft genome status).
The Mtb isoleucyl-tRNA synthetase (IleRS) is eukaryotic like [67]. Hence, we were interested in understanding whether this is also the case for the members of the Mmucand Mneo-clades. Only a single IleRS gene copy was predicted to be present in these mycobacteria and our analysis suggested that all encode a eukaryotic like IleRS (Additional file 3: Figure S6a).
We also predicted genes encoding AARS paralogs in some of the mycobacteria (Additional file 3: Supplementary text; Figure S6b-f; Table S6a, c) and genes encoding cyclodipeptide synthetases (CDPSs) in the Mneo strains (Additional file 3: Figure S7a, b). In MtbH37Rv this enzyme can hijack aminoacylated tRNAs resulting in formation of cyclodipeptides that subsequently are used for mycocyclosin biosynthesis [68][69][70][71]. Prediction of the CDPS gene in Mneo strains suggests that it can be present in both SGM and RGM.
RNase P RNA, tmRNA and 4.5S RNA As expected genes encoding RNase P RNA (rnpB), tmRNA (ssrA) and 4.5S RNA (ffs) were identified in the genomes for all Mmucand Mneo-clade members. Alignments suggest that ssrA and ffs are well conserved whereas rnpB show more diversity (Additional file 4: Figure S9a-f ) consistent with it being a good biomarker for mycobacterial species identification [73]. We also noted that the ssrA (tmRNA) encoded C-terminal protein tag is well conserved compared to MtbH37Rv and MsmegMC 2 -155 with variation only at the fourth  ; Fig. 5b). Analyzing the location of the genes encoding proteins that interact with these RNAs [the C5 protein (rnpA, RNase P), SmpB (smpB, tmRNA) and the Ffh protein (ffh, 4.5S RNA)] revealed that, as in other bacteria (e.g., E. coli), the protein coding genes are separated relative to the chromosomal locations for the respective RNA coding genes rnpB, ssrA and ffs (Additional file 4: Figure S9a, c, e). This raises the question how the expression of these gene pairs is regulated, which warrants further studies. In the meantime, measuring the levels of expression of RNase P RNA, tmRNA and 4.5S RNA by northern analysis showed higher levels in stationary Mmuc T and Maub T cells than in exponentially growing cells ( Fig. 6; Additional file 4: Figure S10; for Maub T we only measured the levels for tmRNA).
We generated phylogenetic trees for rnpB, ssrA and ffs extracted from 113 mycobacterial genomes (Additional file 4: Figure S9g-i). Interestingly, in comparison to the core phylogeny (Additional file 2: Figure S3a  In conclusion, it appears that the evolutionary route for these ncRNAs is different compared to the core genes.

GOLLD RNA
The large GOLLD (>400 nucleotides long; Additional file 4: Figure S11a) RNA was only predicted to be present in Maub T . The predicted gene is part of two scaffolds. However, its gene synteny is supported by the presence of the GOLLD RNA gene on one contig in two other Mycobacterium spp., e.g. Mycobacterium abscessus M24 and Mycobacterium conceptionense strain MLE (Additional file 4: Figure S11a; NCBI database; we cannot conclusively exclude the presence of multiple GOLLD RNA gene copies; see also the Discussion). The gene is located within the "32 tRNA cluster" with tRNA genes present both up-and downstream of the GOLLD RNA gene. In addition, a pseudo tRNA gene is predicted to be located within the GOLLD RNA gene (Additional file 4: Figure S11b).
Northern blot analysis suggested that GOLLD RNA is expressed in Maub T at a low level, as judged from the weak signal. The level was two-fold higher in stationary phase compared to exponentially growing cells (Fig. 6b, c; Additional file 4: Figure S10) following the same trend as for RNase P RNA, tmRNA and 4.5S RNA (see above). No expression of the predicted pseudo tRNA gene located within the GOLLD RNA gene could be detected, in spite of putative σ B promoters upstream of the pseudo tRNA gene (Additional file 4: Figure S11a; see Discussion).

Ms1 RNA ("6S RNA") and 6C RNA
The gene encoding Ms1 RNA was predicted to be present in all members of the Mmucand Mneo-clades as well as in other available mycobacterial and Actinobacteria genomes ( Fig. 7a; Additional file 4: Figure S12). Comparing gene synteny for the Ms1 RNA region (and sequence alignment) in mycobacteria and Streptomyces coelicolor   Figure S12b). Helices and loops are named P1 to P7 and L1 to L3 as indicated. e Predicted secondary structure for Mmuc T 6C RNA. The helices and loops are named as in (d), see also Additional file 4: Figure S12d revealed that the Ms1 RNA gene is positioned where the 6S RNA gene maps in S. coelicolor [74] (Additional file 4: Figure S12a, b). This supports the notion that Ms1 RNA might be a mycobacterial variant of 6S RNA [75]; see Discussion. A phylogenetic Ms1 DNA tree further suggests that the gene present in Mneo strains and Mmucclade members is close to the mycobacterial ancestral gene (Additional file 4: Figure S12c). We also note that in this tree the M. chelonaeand Mmuc-clades are positioned close together indicating the evolutionary path of the Ms1 RNA gene.
Northern analysis suggested that Ms1 RNA is expressed in Mmuc T and that its level of expression is higher in stationary cells (Fig. 7b, c; Additional file 4: Figure S10; the reason for two bands in one of the samples is unknown but might be related to degradation).
Secondary structural models of the Ms1 RNA were generated based on the Mmuc T Ms1 RNA sequence using the M-fold tool/ RNA-fold (Fig. 7d). Comparing the predicted Ms1 RNA structures of Mmuc T , Mpho T , Maub T , Mneo T and Mcos T revealed nucleotide variations in several regions marked with dashed boxes (Fig. 7d). Several of these changes are consistent with the predicted Ms1 RNA secondary structure. Noteworthy, unconventional base pairing such as G•A base pairs distort the helical structure and these distortions might be part of a protein binding site(s).
Close to the Ms1 RNA gene we identified the presence of the gene encoding 6C RNA, which is widespread among Actinobacteria [76,77]. Its location relative to the Ms1 RNA gene is conserved, with one gene in between, comparing the mycobacteria studied here and S. coelicolor ( Fig. 7a; Additional file 4: Figure S12a). Sequence alignment also suggested that 6C RNA is highly conserved among mycobacteria belonging to the Mmucand Mneo-clades (Additional file 4: Figure S12d). Although its function in mycobacteria is unclear it is expressed in exponentially growing Mmuc T cells and stationary cells with a significantly higher level in stationary cells (Fig. 7b, c; see Discussion).
The 6C RNA secondary structure was manually folded and resulted in a stable structure composed of three helices, P1 to P3, with two C-rich loop structures (Fig. 7e).

Sar antisense RNA
The ncRNA categorized as Sar antisense RNA was predicted to be unique to Mmuc T (and Mmuc CSURP2099 ) and located in a phage region (0.65 Mbp region; Fig. 1b; Additional file 4: Figure S8a, b; Table S7). The homology to other Sar RNA genes is low [78]. The Mmuc T Sar RNA gene is expressed and the level was modestly higher in stationary cells compared to exponentially growing cells (Fig. 7a, c). In Salmonella typhimurium Sar RNA was identified as a regulator during development of the temperate bacteriophage P22 [78]. Its function in Mmuc T is unknown. But, the gene is located in the intergenic region between two genes (MUCO_DSM00642 and MUCO_DSM00643) encoding proteins of unknown function where the former encodes a signal peptide, while the Sar RNA gene is transcribed from the opposite strand. It is therefore conceivable that this RNA is involved in regulating the expression of either of these genes or both.

ncRNAs, comparison with other mycobacteria
We also predicted ncRNA genes in MtbH37Rv and MsmegMC 2 -155 using the Rfam data base (Additional file 4: Figure S8; Table S7; see also [34] and Refs therein). This comparison revealed that 18 (out of 20) ncRNA genes present in the Mmucand Mneo-clades are also present in MtbH37Rv and MsmegMC 2 -155. There are also seven ncRNA genes in MtbH37Rv and MsmegMC 2 -155 for which no orthologs could be found in the 17 genomes (Table 1; Additional file 4: Figure S8). Moreover, the In-tron_gpII RNA (Rfam annotation, RF00029), which is present in two copies in MtbH37Rv, is also predicted to be present in Mpho T with three copies (Additional file 4: Figure S13a, b). The sequences of the three Mpho T gene copies are similar but they differ from the two genes in MtbH37Rv. Comparison of their gene synteny revealed that for MtbH37Rv both Intron_gpII RNA genes are located within genes while for Mpho T the three Intron_gpII RNA genes were predicted to be present in intergenic regions (Additional file 4: Figure S13a). Together this suggested that the Intron_gpII RNA ncRNA genes in these two mycobacteria likely are of different origin.

Discussion
The size of the genomes for Mmucand Mneo-clade members including type strains range between 5.4 to 6.5 Mbp. Core gene phylogeny based on 291 core genes present in 109 mycobacterial genomes and genes common to the Mmucand Mneo-clade members and M. sp. URHB0044 positioned the Mmucand Mneoclades next to one another ( Fig. 2a; Additional file 2: Figure S3). The tree based on 2226 genes present in 17 mycobacteria and ANI values cluster the Mmuc-clade members (Fig. 2) and indicated that Mmuc LZSF01 is closer to Mpho T than it is to Mmuc T . Hence, Mmuc LZSF01 should be considered to be a Mpho strain. Consistent with this is also the finding that the numbers of tRNA genes in Mpho T and Mmuc LZSF01 are conserved, as are the structures of 14 tRNA genes encompassing this cluster, while the corresponding cluster in Mmuc T lacks three tRNA genes (see below). This analysis also indicates that Msp360 and Msp410 are likely to be two Mmuc strains. Considering the Mneo-clade, the positioning of Mneo VKMAc-1815D deviates from the other Mneo strains including the type strain Mneo T . This raises the question whether Mneo VKMAc-1815D should be considered as a subspecies of Mneo or perhaps a different species. For M. sp. URHB0044, our data does not provide any information about which clade it belongs to and possibly it is a new mycobacterial species.

Variation in tRNA among Mmuc-and Mneo-clade members
The number of predicted coding sequences (CDSs) correlates with genome size while the number of tRNA coding genes does not (Table 1). This is also apparent by comparing the number of tRNA genes in the SGM MtbH37Rv (4.4 Mbp; [29]) and RGM MsmegMC 2 -155 (7 Mbp; [30]), which carry 45 and 47 tRNA genes, respectively. Evidently, the number of tRNA genes does not relate to differences in growth rates. Analyzing the tRNA gene organization in the complete Mmuc T genome, in total 57 tRNA genes, suggests that the majority of the tRNA genes are transcribed from single genes (32 of 46; not considering the cluster having 11 tRNA genes, which are predicted to be transcribed from two operons, one encompassing eight and the other three genes; Fig. 3). The remaining predicted tRNA operons encode two (four operons) and three (two operons) tRNA genes ( Fig.  3a; Additional file 3: Figure S4b-e). This is similar to MtbH37Rv and MsmegMC 2 -155 where 31 of 45 and 33 of 47, respectively, appear to be transcribed from single tRNA genes. The remaining tRNA genes in these two mycobacteria are predicted to be transcribed from operons with two or three tRNA genes and these are the same as those predicted in Mmuc T (Additional file 3: Figure S4a, b). In other bacteria, such as the Gram negative E. coli K12 and the Gram positive Firmicutes (e.g., B. subtilis and Staphylococcus aureus), the tRNA genes are transcribed from large operons [27,33,[79][80][81][82][83][84]. For example, a cluster of as many as 27 tRNA was identified in S. aureus [82]. This is in contrast to mycobacteria in which tRNAs in general appear to be transcribed mainly from single genes. This is also supported from our analysis of the tRNA genes in the draft genomes of Mpho T , Maub T , Mneo T and Mcos T (Additional file 3: Figure S4b-e).
Several of the tRNA genes that are common to Mmuc T , MtbH37Rv and MsmegMC 2 155 (for which complete genomes are available) are positioned at different locations on the circular chromosome suggesting rearrangements of these tRNA genes after these mycobacteria diverged (Additional file 5: Figure S14a, b). However, the positioning of the tRNA genes near the origin of replication, oriC, appears to be conserved comparing Mmuc T , MtbH37Rv and MsmegMC 2 -155 (oriC, its location is predicted on the basis of the positioning of dnaA, dnaN and rpmH; [32,85]; positioning of oriC is also supported by the GC skew, see Fig. 1). In this context, certain intergenic regions in MtbH37Rv have been identified as necessary for optimal growth and 11 specific tRNA genes are located in these regions [86]. The same tRNA genes are located at roughly the same positions (with the same orientation) on the Mmuc T chromomosome except for the tRNA Thr (GGT) and tRNA Met (CAT) genes, which have been re-located (Additional file 5: Figure S14a, b). This raises interesting questions with respect to whether it is the tRNA genes and their expression that are needed for optimal growth or if it is their location on the chromosome in relation to oriC. Thus, it would be interesting to follow the same protocol as Zhang et al. [86] using other mycobacteria such as Mmuc T to address these questions.

Horizontally transferred tRNA gene clusters
For Mmuc T and Mpho T (and Mmuc CSURP2099 , Msp360 and Msp410) we identified tRNA gene clusters carrying 11 and 14 tRNA genes, respectively, while Maub T carries a different cluster encoding 32 tRNA genes. On the basis of gene synteny and tRNA gene sequence this latter 32 tRNA gene cluster is likely to be of a different origin than the clusters in Mmuc T and Mpho T (see also below), which are the same apart from that Mmuc T lacks three tRNA genes. Moreover, compared to the commonly present tRNAs the sequences for the genes encompassing these clusters suggest that they encode for structurally different tRNAs. Hence, it is likely that these tRNA gene clusters are the result of horizontal gene transfer and that this occurred after Mmuc T and Mpho T diverged from Maub T . This is supported by the presence of homing HNH endonuclease genes within these tRNA gene clusters. Homing endonucleases are known to be involved in moving genes [87,88] including tRNA genes as in E. coli bacteriophages [89]. It is therefore plausible that these large tRNA gene clusters present in Mmuc T , Mpho T and Maub T originate from phages.
In the Mmuc T -Mpho T tRNA cluster, the HNH endonuclease gene is positioned between blocks of eight and three genes (six in Mpho T ; Fig. 3b). Among the eight tRNA genes, one encodes tRNA Ile (TAT). This tRNA gene is not present in members of the Mneo-clade, Maub T or MtbH37Rv and MsmegMC 2 -155 and it reads AUA codons. The AUA codon is rare among members of the Mmucand Mneo-clades, while it is more abundant in MtbH37Rv (unpublished data). When the tRNA Ile (TAT) gene is missing the AUA codon is read by tRNA Ile (CAT) carrying a lysidine (L, 2-lysyl-cytidine) instead of C at position 34 (underlined). The enzyme responsible for this modification is TilS [90] and all the Mmucand Mneo-clade members (and MtbH37Rv) have TilS homologs. The Mmuc T tRNA Ile (TAT) is expressed and the level is higher in stationary cells. Taken together, for those mycobacteria having the tRNA Ile (TAT) gene there appears to be two ways to decode AUA; whether this arrangement gives any advantage remains to be investigated. Also, since the number of tRNA genes in the Mmuc T -Mpho T tRNA cluster differ makes it possible to use this as a marker to differentiate these two species.
In contrast to the Mmuc T -Mpho T tRNA cluster, the Maub T gene cluster encompasses both tRNA and protein genes. Large tRNA gene clusters have previously been reported to be present in M. abscessus M24 (34 tRNA genes; [91]) and M. conceptionense MLE (37 tRNA genes; NCBI bioproject id PRJNA288077; see also [92]). The majority of the tRNA genes encompassing the Maub T cluster are also present in the M. abscessus M24 and M. conceptionense MLE clusters (Additional file 3: Figure  S11b; Table S5). In these three mycobacteria, this cluster also encodes GOLLD RNA and HNH endonuclease. This large ncRNA (>400 nucleotides long) was first identified in Lactobacillus brevis ATCC 367 [72]. The GOLLD RNA is expressed in Maub T albeit at a low level. Its function, however, is not known but it is associated with a prophage and its expression is linked to phage production. Interestingly, in many instances the GOLLD RNA gene is present close to the location of tRNA genes and it also encodes a predicted tRNA gene [72] and for Maub T our data suggests that this gene corresponds to a pseudo tRNA gene. Moreover, our phylogenetic analysis indicates that the mycobacterial GOLLD RNA genes likely are of phage origin (unpublished data). Together, this suggests that this large "32 tRNA gene" cluster in Maub T , which includes the GOLLD RNA gene and a HNH endonuclease gene, is the result of horizontal gene transfer.
The Maub T "32 tRNA gene" cluster harbors an additional tRNA Cys gene equipping Maub T with three tRNA Cys genes (that are expressed; Fig. 4). Noteworthy, Mmucand Mneo-clade members as well as MsmegMC 2 -155 all have two while MtbH37Rv has one tRNA Cys gene (for other RGM, see [31,32]; Additional files 3 and 5: Figures S4, S5 and S14). Inspection of their secondary structures reveals that the two extra tRNA Cys both have four base pairs long anticodon stems with the potential to form non-Watson-Crick base pairs, C27-U43 (additional tRNA Cys gene present in all Mmucand Mneo-clade members) and A27-C43 (tRNA Cys gene present only in Maub T ). Whether this influences function or if these extra tRNA Cys isoacceptors indeed are functional or influence growth remains to be investigated.
RNase P processing and tRNA gene structure The tRNA genes are transcribed as precursors (pre-tRNAs) and the endoribonuclease P (RNase P) is responsible for generating the tRNA 5' termini. The majority of tRNAs carry a guanosine at their 5' ends, referred to as G +1 , and seven base pairs long amino acid acceptor (aa-) stems. The residue immediately 5' (the -1 position; N -1 ) of G +1 at the RNase P cleavage site in pre-tRNAs has been suggested to interact with the conserved residue A248 (E. coli numbering) in the RNA subunit of RNase P (RPR) while G +1 is suggested to act as a guiding nucleotide ( [35] and Refs therein). The majority (≈60%) of the tRNA genes in E. coli have a U at -1, while in mycobacteria the occurrence of U at this position is below 20%. Interestingly, G at -1 is frequent in mycobacteria and can be as high as 30% (Additional file 5: Figure  S15). These observations lead to two important points. First, mycobacteria have A at the corresponding "248 position" in the RPR. It is therefore unlikely that the majority of tRNA precursors form a Watson-Crick base pair between residue -1 and "A248" in mycobacterial RPRs ( [35] and Refs therein). Second, since G at the cleavage site functions as a guiding nucleotide the question is how mycobacterial RNase P handle the processing of those tRNA transcripts with G at -1. The tRNA His transcript carries G at -1 and is cleaved at -1 instead of +1 generating a functional tRNA with an eight base pair long aa-stem ( [35] and Refs therein). For the other mycobacterial tRNA transcripts carrying G at -1, RNase P might first cleave at -1 and then at the correct site, +1, generating functional tRNAs with seven base pairs long aa-stems. Another possibility, mycobacterial RNase P has an intrinsic capacity to efficiently cleave tRNA transcripts at the correct site +1 despite the presence of G at -1. In this context, we note that the frequency of G -1 for the "HGT tRNA genes" in Mmuc T , Mpho T and Maub T discussed above is low (or absent), while for U -1 it is ≈50% (Additional file 5: Figure S15). This further supports the notion that these tRNA genes have been acquired through horizontal gene transfer.
Stable ncRNAs -RPR, tmRNA, 4.5S RNA, Ms1 RNA ("6S RNA") and 6C RNA A number of functional ncRNAs have been identified in bacteria and recently several reports document the presence of significant numbers of ncRNAs in MtbH37Rv and MsmegMC 2 -155 using both in silico and experimental approaches (reviewed by [34,93]). Here we identified several putative ncRNAs in Mmucand Mneo-clade members. Focusing on the "classical" ncRNAs such as the catalytic RNase P RNA, the scavenging tmRNA and the signal recognition particle 4.5S RNA we showed that their levels in Mmuc T (and tmRNA in Maub T ) are higher in stationary phase compared to exponential phase. This is in keeping with their levels of expression being influenced by growth conditions (see e.g. [94][95][96]). Interestingly, tmRNA is involved in the control of the Caulobacter cell-cycle, sporulation in Bacillus and cell development in Streptomyces species [96]. Given that mycobacteria undergo changes in their cell morphology, including spore formation, upon exposure to environmental changes [97][98][99] it would be interesting to study the impact of tmRNA on cell morphology among mycobacteria.
The 6S RNA is a global transcriptional regulator in bacteria with a role in the adaptation to stationary phase. It binds to the housekeeping σ factor (σ A ) in the RNA polymerase holoenzyme and thereby preventing the polymerase from binding to DNA promoters that require σ A for transcription. The identification of 6S RNA in mycobacteria has been an enigma [100]. We predicted the presence of genes homologous to the Ms1 RNA gene in all Mmucand Mneo-clade members. In other bacteria the level of 6S RNA increases when the cells enter stationary phase [100]. Our data showed that the level of Ms1 RNA in Mmuc T cells is significantly higher in stationary phase compared to exponential phase in keeping with what has been reported for MtbH37Rv and MsmegMC 2 -155 [75,101]. Moreover, Hnilicová et al. [75] provided data suggesting that Ms1 RNA binds to the RNA polymerase core enzyme and not to any of the σ factors, including σ A . Searching for the presence of the Ms1 RNA gene among the Actinobacteria reveals that homologs are present in several species (Fig. 7a; Additional file 4: Figure S12). The gene synteny for 6S RNA in S. coelicolor [74,102] and the Ms1 RNA gene in mycobacteria are similar. Together this suggests that Ms1 RNA might be a 6S RNA variant that interacts with RNA polymerase and thereby affect the expression of genes; however, whether it has the same function as 6S RNA in other bacteria remains to be deciphered (see [75]). In this context, the level of Ms1 RNA accumulates upon infection of mice lungs. Also, overexpression of Ms1 RNA in MtbH37Rv affects the levels of roughly 300 genes [93,101]. Hence, like 6S RNA, Ms1 RNA appears to be a global regulator with a role in adaptation to stationary phase and growth inside the host cells.
We also identified another known and conserved gene close to the Ms1 RNA gene in the Mmucand Mneo-clade members, the 6C RNA gene for which we have little functional information. Its location is conserved relative to the Ms1 RNA gene among different Actinobacteria and the level of 6C RNA is higher in Mmuc T cells in stationary phase compared to exponentially growing cells, albeit not as pronounced as for Ms1 RNA. The expression of 6C RNA has also been detected in other mycobacteria, Cornynebacterium glutamicum and S. coelicolor [76,77,103,104]. In S. coelicolor, 6C RNA is upregulated during development/sporulation and it has been discussed that it reflects dormancy and slower metabolism [76]. For C. glutamicum, it is suggested to be involved in the SOS response and possibly also in the control of cell division as well as the GlxR, a regulator of carbon source metabolism and energy conversion, regulatory network [77,105]. The role of mycobacterial 6C RNA is unknown but from the discussion above it is conceivable that it is involved in regulating cell shape and development including dormancy and spore formation in mycobacteria [97,98], for a review see [99]. Clearly, deciphering the functions of these (and other) ncRNAs will provide information about the way mycobacteria adapts to different growth conditions and perhaps also in relation to cell differentiation.

Conclusions
The genome sizes for Mmucand Mneo-clade members range between 5.4 to 6.5 Mbp and phages, IS elements, horizontally transferred tRNA gene clusters, and phagederived ncRNAs have likely influenced the evolution of the Mmucand Mneo-clades. While the number of predicted coding sequences correlates with genome size, the number of tRNA coding genes does not. Among the Mmucand Mneo-clade members, 39 tRNA genes are common. Members of the Mmuc-clade harbor several tRNA genes that have been acquired through HGT. The majority of the tRNA genes in mycobacteria are transcribed from single genes. As in other mycobacteria tRNA charging of asparagine and glutamine is accommodated using the tRNA-dependent amidotransferase pathway (Adt) in the Mmucand Mneo-clade members. The levels of RNase P RNA (essential for the processing of tRNAs), tmRNA, and 4.5S RNA are higher at stationary phase compared to exponentially growing cells. Two new and horizontally transferred ncRNAs are present and expressed in Mmuc-clade members. The levels of the ncRNAs, 6C RNA and Ms1 RNA, are higher in cells at stationary phase compared to exponentially growing cells. The Ms1 RNA likely represents a mycobacterial 6S RNA variant. The evolutionary routes for the ncRNAs RNase P RNA, tmRNA and Ms1 RNA are different from that of the core genes.