Molecular Evidence of Lateral Gene Transfer in rpoB Gene of Mycobacterium yongonense Strains via Multilocus Sequence Analysis

Recently, a novel species, Mycobacterium yongonense (DSM 45126T), was introduced and while it is phylogenetically related to Mycobacterium intracellulare, it has a distinct RNA polymerase β-subunit gene (rpoB) sequence that is identical to that of Mycobacterium parascrofulaceum, which is a distantly related scotochromogen, which suggests the acquisition of the rpoB gene via a potential lateral gene transfer (LGT) event. The aims of this study are to prove the presence of the LGT event in the rpoB gene of the M. yongonense strains via multilocus sequence analysis (MLSA). In order to determine the potential of an LGT event in the rpoB gene of the M. yongonense, the MLSA based on full rpoB sequences (3447 or 3450 bp) and on partial sequences of five other targets [16S rRNA (1383 or 1395 bp), hsp65 (603 bp), dnaJ (192 bp), recA (1053 bp), and sodA (501 bp)] were conducted. Incongruences between the phylogenetic analysis of the full rpoB and the five other genes in a total of three M. yongonense strains [two clinical strains (MOTT-12 and MOTT-27) and one type strain (DSM 45126T)] were observed, suggesting that rpoB gene of three M. yongonense strains may have been acquired very recently via an LGT event from M. parascrofulaceum, which is a distantly related scotochromogen.


Introduction
From a clinical and epidemiological perspective, the members of the Mycobacterium avium complex (MAC) are the most important nontuberculous mycobacteria (NTM). Traditionally, MAC includes two species, M. avium and M. intracellulare [1,2,3]; in Korea, the prevalence of M. intracellulare infections is higher than that of M. avium [4]. Recently, it was reported that M. intracellulare related strains from Korean patients showed more genetic diversity; the strains can be divided into a total of five distinct groups using the sequence analysis of hsp65, internal transcribed spacer and 16S rRNA genes [5].
Generally, the informative genes associated with the central dogma of bacteria, such as the 16S rRNA gene or the RNA polymerase gene (rpoB), have been reported to be recalcitrant to lateral gene transfer (LGT) events. However, the LGT events of informative genes within the genus Mycobacterium have been disclosed in two recent reports. One report described the potential LGT event of the rpoB gene between three groups of strains belonging to Mycobacterium abscessus (M. abscessus sensu stricto, Mycobacterium massiliense and Mycobacterium bolletii) [6]; the other report described the potential LGT event of the 16S rRNA gene between Mycobacterium franklinii and Mycobacterium chelonae [7]. Moreover, a novel species, M. yongonense, which is phylogenetically related to M. intracellulare, was introduced from studies of a Korean patient with pulmonary symptoms. Notably, M. yongonense proved to have a distinct RNA polymerase gene (rpoB) sequence identical to that of M. parascrofulaceum, which is a distantly related scotochromogen, suggesting that the rpoB gene was acquired via a potential LGT event [8].
The aims of the current study are two-fold: the first is to discover the epidemiologic features of M. yongonense from an infection cohort previously identified as M. intracellulare and the second is to prove the presence of the LGT event in the rpoB gene of the M. yongonense strains via multilocus sequence analysis (MLSA). In order to determine the potential of an LGT event in the rpoB gene of M. yongonense, the MLSA based on full rpoB sequences (3447 or 3450 bp) and partial sequences of the other five targets [16S rRNA (1383 or 1395 bp), hsp65 (603 bp), dnaJ (192 bp), recA (1053 bp), and sodA (501 bp)] were applied to a total of seven mycobacteria strains: three M. yongonense (MOTT-12, MOTT-27 and DSM 45126 T ), two M. intracellulare strains (MOTT-02 and ATCC 13950 T ), and two M. parascrofulaceum strains (MOTT-01 and ATCC BAA-614 T ).

Mycobacterial isolates
Seven mycobacteria strains, including three reference strains (M. intracellulare ATCC 13950 T , M. parascrofulaceum ATCC BAA-614 T and M. yongonense DSM 45126 T ) and four clinical isolates (MOTT-01, MOTT-02, MOTT-12, and MOTT-27) were analyzed using the MLSA (Table S1). Of the four clinical isolates, one (MOTT-01) was identified as M. parascrofulaceum, one (MOTT-02) as M. intracellulare and two (MOTT-12 and MOTT-27) as M. yongonense, using a combination of the hsp65 and rpoB sequence based analyses. The experiment was based entirely on the extracted genomic DNA from the isolates, so the research was undertaken without informed consent and a waiver of informed consent was obtained from the Institutional Review Board (IRB) of Seoul National University Hospital. This work was approved by the IRB of Seoul National University Hospital (C-1204-003-403).

Biochemical tests
In order to identify and differentiate the two M. yongonense clinical isolates (MOTT-12 and MOTT-27), their biochemical test profiles were compared with those of three type reference strains: M. intracellulare ATCC 13950 T , M. yongonense DSM 45126 T and M. parascrofulaceum ATCC BAA-614 T . The colony morphology, pigmentation in the dark, photo-induction and growth at different temperatures (25uC, 37uC and 45uC) were tested on 7H10 agar plates with OADC over a six-week incubation period. The acidalcohol-fastness was examined via Ziehl-Neelsen and auramine O staining. The biochemical characteristics of niacin accumulation, nitrate reductase, arylsulfatase on Days 3 and 14, and the heatstable catalase (pH 7, 68uC), tellurite reductase, Tween 80 hydrolysis, urease and pyrazinamidase were tested [10]. The inhibition tests including the tolerance to thiophene-2-carboxylic acid hydrazide (TCH), p-nitrobenzoate (PNB), 5% sodium chloride, ethambutol (EMB), and picric acid were performed; and the ability to grow on MacConkey agar without crystal violet was also examined. and sodA (501 bp)] of the four clinical and three reference stains were analyzed. The bead beater-phenol extraction method was used to extract the chromosomal DNA of these strains, as previously reported [9]; the extracted DNA samples were then used as templates for the polymerase chain reaction (PCR) amplifications of the six independent sequence targets [rpoB (partial and complete), 16S rRNA, hsp65, dnaJ, recA, and sodA]. The PCR amplifications were bi-directionally sequenced using the same primers as those used in the PCR. The PCR amplification and sequence analysis of the rpoB (partial and complete), 16S rRNA, hsp65, dnaJ, recA, and sodA genes were performed as described previously [5,9,11,12,13,14]. A total of six primer sets were used for the amplification of the full rpoB gene (3447 or 3450 bp) sequence. The locations and sequences of the primers for the rpoB amplification are shown in Figure S1 and Table S2, respectively. These primer sets were designed using the whole genome sequence database of M. intracellulare ATCC 13950 T (GenBank no. ZP_05227774) and M. avium 104 (GenBank no. NC_008595). The sequences of the primers for the amplification and sequencing of the rpoB (partial and complete), 16S rRNA, hsp65, dnaJ, recA, and sodA genes are also shown in Table S2. For the phylogenetic analysis of the rpoB (partial and complete), 16S rRNA, hsp65, dnaJ, recA, and sodA genes, the nucleotide sequence similarities of each gene were determined using the MegAlign package (DNASTAR) software. The phylogenetic trees were constructed from the full sequences of the rpoB gene (3447 or 3450 bp), the partial sequences of four genes [hsp65 (603-bp), dnaJ (192 bp), recA (1053 bp), and sodA (501 bp)] and 16S rRNA (1383 or 1395 bp) sequences using the neighbor-joining method [15] in the MEGA 4 software; the bootstrap values were calculated from 1,000 replications [16].  Table S1. Among these, the hsp65 (FJ849777) gene sequences of the MOTT-27 strain were retrieved from a previous report by Park et al. [5] (Table S1).

Results and Discussion
Characterization of the phenotypic traits of the two M. yongonense clinical strains (MOTT-12 and MOTT-27) based on conventional biochemical tests The conventional taxonomic approaches based on biochemical traits demonstrated that all strains shared similar growth patterns. Pigmentation is known to be the most pronounced difference between M. intracellulare and M. parascrofulaceum; the former is a nonphotochromogen; however, the latter is a scotochromogen [17]. The two M. yongonense clinical strains in the current study (MOTT-12 and MOTT-27) proved to be nonchromogens, suggesting that they are phenotypically closer to M. intracellulare, rather than M. parascrofulaceum as described previously [8]. However, the differences in some biochemical traits such as nitrate reductase, arylsulfatase and tellurite reductase were found between M. yongonense DSM 45126 T and the two clinical strains (MOTT-12 and MOTT-27) (see Table S3). The full rpoB gene sequence proved useful for the delineation of the bacterial species [18]. A rpoB gene sequence similarity of ,97.0% is reported to be significantly correlated with a DNA-DNA hybridization (DDH) value of ,70%, which is the universal cut-off value for the delineation of a bacterial species [18]. All full length rpoB sequences obtained in the current study were verified to be encoded in the proper deduced RpoB amino acids in the in silico translation. The phylogenetic analysis based on the full rpoB sequences (3450 bp) demonstrated that the three M. yongonense isolates (MOTT-12, MOTT-27, and DSM 45126 T ) formed a tight cluster with the M. parascrofulaceum strains (ATCC BAA-614 T and MOTT-01) rather than with the M. intracellulare strains (ATCC 13950 T and MOTT-02). Also their phylogenetic relationship was supported by a high bootstrap value (100.0; Figure 1). The sequence similarity value of the full rpoB sequences between the three M. yongonense strains and two M. parascrofulaceum strains ranged from 99.7% to 99.8%, which presented eight to nine bp mismatches among 3450 bp. However, the sequence similarity values between the three M. yongonense strains and two M. intracellulare strains ranged from 94.7% to 94.9%, which presented 181 to 196 bp mismatches from 3450 bp ( Table 1). The high similarity value observed between the M. yongonense and M. parascrofulaceum strains indicates that these two different species share almost identical rpoB sequences. Furthermore, the similarity values observed between the M. yongonense and M. intracellulare strains are lower than that of the cut-off value (97.0%) for the delineation of bacterial species [18] (Table 1).

Phylogenetic analysis based on the 16S rRNA and hsp65 gene
In order to verify the above hypothesis, a phylogenetic analysis of the three M. yongonense strains was performed using two other genes (16S rRNA and hsp65 genes), which have been used widely for mycobacteria taxonomies and diagnostics [13,19,20,21]. Despite some problems in the bacteria taxonomy, the 16S rRNA gene sequence-based comparisons have been and remain invaluable in describing the prokaryotic diversity; they are indispensable in the delineation of bacterial species [22]. The phylogenetic analysis based on the 16S rRNA sequence (1383 or 1395 bp) indicated that the three M. yongonense strains belonged to the M. intracellulare group, exhibiting a sequence similarity ranging from 99.8% to 100% with two other M. intracellulare strains (ATCC 13950 T and MOTT-02; data not shown). The three M. yongonense strains exhibited a relatively low level of similarity value (96.8%) with the M. parascrofulaceum strains, which was lower than the universally accepted cut-off value for the 16S rRNA gene (97.0%) for bacteria species delineation (data not shown) [23]. This strongly suggests that the three M. yongonense strains are phylogenetically related to M. intracellulare.
The hsp65 gene sequence based methods have been the most widely used methods for mycobacteria taxonomies as alternatives to the 16S rRNA based methods [9,13]. The three M. yongonense strains exhibited some minor variations (99.3% similarity value with four base pair mismatches of the 603 bp hsp65 sequences) compared with the other two M. intracellulare strains (ATCC 13950 T and MOTT-02). The phylogenetic analysis based on the hsp65 gene sequence (603 bp) indicated that the three M. yongonense strains belonged to the M. intracellulare group, rather than to the M. parascrofulaceum group, which indicates a low level of sequence similarity value of 94.9% with the two M. parascrofulaceum strains (ATCC BAA-614 T and MOTT-01; data not shown). This also strongly supports their phylogenetic location in M. intracellulare.
The phylogenetic analyses based on the dnaJ gene sequence    Table 1).
The effect of adding the rpoB sequence to the concatenated sequences of the five MLSA genes (7182-7194 bp) was also studied. The topology of the obtained tree ( Figure 2B) was  radically different from only that constructed from the MLSA gene sequences (Figure 2A). The branch of the M. yongonense strains forming the same cluster with that of the M. intracellulare strains in the MLSA tree were transferred into a cluster belonging to the M. parascrofulaceum strains in the MLSA + rpoB tree, which was strongly supported with a high level of bootstrap values (100%). The discrepancy observed between the topology structures of both trees suggests the potential LGT event of the rpoB gene from the M. parascrofulaceum strain into the M. yongonense strain. From a clinical perspective, these results emphasize the importance of the MLSA for mycobacteria identification. Currently, the rpoB gene has been used widely as a target gene for bacterial identification, particularly for mycobacteria identification [9,24,25]. However, the data in this study implies that some strains of M. yongonense could be misidentified as M. parascrofulaceum when only a single rpoB gene is used in the identification or as M. intracellulare with use of chronometers other than the rpoB gene.
In conclusion, collective consideration of the molecular taxonomic data based on the full rpoB and five other genes, which have been used widely for mycobacterial identification has led to the conclusion that the three M. yongonense strains with the signature rpoB gene have potentially acquired their rpoB gene via a very recent LGT event from M. parascrofulaceum. However, the details of the LGT events between M. parascrofulaceum and M. yongonense strains must be further elucidated in a future study. Furthermore, the data presented here also suggests that the rpoB gene analysis alone may have potential for misidentification in mycobacteria diagnostics. Thus, an approach using multilocus genes should be conducted for mycobacteria identification. Figure S1 Locations of primers used for amplification of the full rpoB (3450 bp) gene sequence in this study.

(DOCX)
Table S1 Mycobacteria strains used in this study.
(DOC) Table S2 The primer sets used for amplification of the full rpoB, partial rpoB, hsp65, 16S rRNA, dnaJ, recA, and sodA in this study. (DOC)