Molecular analysis of clinical isolates previously diagnosed as Mycobacterium intracellulare reveals incidental findings of “Mycobacterium indicus pranii” genotypes in human lung infection

Background Mycobacterium intracellulare is a major cause of Mycobacterium avium complex lung disease in many countries. Molecular studies have revealed several new Mycobacteria species that are closely related to M. intracellulare. The aim of this study was to re-identify and characterize clinical isolates from patients previously diagnosed with M. intracellulare lung disease at the molecular level. Methods Mycobacterial isolates from 77 patients, initially diagnosed with M. intracellulare lung disease were re-analyzed by multi-locus sequencing and pattern of insertion sequences. Results Among the 77 isolates, 74 (96 %) isolates were designated as M. intracellulare based on multigene sequence-based analysis. Interestingly, the three remaining strains (4 %) were re-identified as “Mycobacterium indicus pranii” according to distinct molecular phylogenetic positions in rpoB and hsp65 sequence-based typing. In hsp65 sequevar analysis, code 13 was found in the majority of cases and three unreported codes were identified. In 16S–23S rRNA internal transcribed spacer (ITS) sequevar analysis, all isolates of both species were classified within the Min-A ITS sequevar. Interestingly, four of the M. intracellulare isolates harbored IS1311, a M. avium-specific element. Two of three patients infected with “M. indicus pranii” had persistent positive sputum cultures after antibiotic therapy, indicating the clinical relevance of this study. Conclusions This analysis highlights the importance of precise identification of clinical isolates genetically close to Mycobacterium species, and suggests that greater attention should be paid to nontuberculous mycobacteria lung disease caused by “M. indicus pranii”.


Background
Mycobacterium avium complex (MAC) lung disease is the most common lung disease caused by nontuberculous mycobacteria (NTM) and its prevalence has been increasing worldwide [1][2][3][4][5][6][7][8][9][10]. MAC was originally composed of two species, M. avium and Mycobacterium intracellulare [11]. Many laboratories and studies reported these species as MAC because they are highly similar, and the clinical features of patients who are infected with these two species are considered indistinguishable [11,12]. However, some studies have suggested that differentiation between M. avium and M. intracellulare may have epidemiologic and clinical relevance [13,14].
M. intracellulare is a major cause of MAC lung disease in many countries [2]. Molecular studies have revealed the presence of additional taxonomic groups or sequence variants (sequevars) that are closely related to M. intracellulare. Several new species were recently identified including Mycobacterium chimaera, Mycobacterium colombiense, Mycobacterium arosiense, Mycobacterium vulneris, Mycobacterium marseillense, Mycobacterium timonense, Mycobacterium bouchedurhonense, Mycobacterium mantenii, and Mycobacterium yongonense [15][16][17][18][19][20]. However, data regarding the proportion of these new species that are etiologic organisms in patients with previously diagnosed M. intracellulare lung disease are very limited.
The methods of correct identification for mycobacterial species in clinical laboratories have changed dramatically over the past two decades. Molecular methods have now surpassed biochemical tests and high-performance liquid chromatography as the method of choice for identifying NTM [21]. Polymerase chain reaction (PCR) restriction fragment length polymorphism analysis (PRA) based on restriction digestion products of specific genes such as hsp65, 16S rRNA, rpoB, and 16S-23S rRNA internal transcribed spacer (ITS) has been reported as a rapid, feasible, and inexpensive diagnostic method [22][23][24][25]. The gold standard identification method of 16S rRNA gene sequencing and sequencing of each loci as a single identification target has failed to discriminate closely related Mycobacterium species such as MAC, Mycobacterium abscessus-Mycobacterium chelonae, Mycobacterium farcinogenes-Mycobacterium senegalense, Mycobacterium kansasii-Mycobacterium gastri, and Mycobacterium marinum-Mycobacterium ulcerans [26,27]. Recently, multigene sequence-based typing has been suggested as the new standard method for identifying Mycobacterium species that are not well discriminated by 16S rRNA gene sequences alone [8,[28][29][30].
In our institution, the rpoB-PRA method had been used for species identification and diagnosis of MAC lung disease [14]. Recently published papers have emphasized the importance of taxonomy in distinguishing the many species and subspecies of MAC. Nonsequencing methods or 16S rRNA sequencing might fail to distinguish closely related species [31,32], indicating that nonsequencing-based approaches or analysis of a single target are not suitable for the accurate identification of (sub-)species belonging to MAC.
Thus, the aim of this study was to re-identify clinical isolates from patients previously diagnosed with M. intracellulare lung disease and to characterize their molecular pattern. For this purpose, the following methods were used: (1) multigene sequence-based typing of 16S rRNA, rpoB, hsp65 and ITS genes, (2) hsp65 and ITS sequevar-based classification, and (3) insertion element analysis. Finally, three "Mycobacterium indicus pranii" strains that were previously identified as M. intracellulare were re-identified. The clinical characterization of lung disease caused by these three "M. indicus pranii" infections was described. "M. indicus pranii" is of specific interest due to its evolutionary significance and therapeutic potential in various disease processes. This study raises the possibility of "M. indicus pranii" as a pathogenic organism in the appropriate host and clinical situation, a notion not previously suggested in prior publications.

Study subjects
Clinical isolates from 77 consecutive patients who were newly diagnosed with M. intracellulare lung disease from Jan. 2008 to Dec. 2009 at Samsung Medical Center (a 1,961-bed referral hospital in Seoul, Korea) were collected and stored. This study was approved by the Institutional Review Board of Samsung Medical Center (File No. 2008-09-016). All patients' record and information was anonymized and de-identified prior to analysis. All patients met the diagnostic criteria for NTM lung disease [11]. All patients were immunocompetent and none of the patients tested positive for human immunodeficiency virus. Baseline patient characteristics are summarized in Table 1.
The isolates were collected before initiating antibiotic treatment for M. intracellulare lung disease. NTM species were identified as M. intracellulare by PRA based on the rpoB gene at time of diagnosis [14].
Identification of patient isolates by multigene sequence-based typing NTM were propagated in Middlebrook 7H9 broth (Difco Laboratories, Detroit, MI, USA) supplemented with 10 % (vol/vol) oleic acid-albumin-dextrose-catalase (OADC; BD Diagnostics). Mycobacterial DNA was extracted using a DNeasy Blood and Tissue Kit according to the manufacturer's instructions (Qiagen, Valencia, CA). Multigene sequence-based typing including hsp65, rpoB, ITS and 16S rRNA fragments was carried out using PCR primer sets as described previously ( Table 2). The PCR products of target genes were subjected to sequence analysis. hsp65 and ITS sequevar analysis were performed as previously described [33,34]. The nucleotide sequences of these genes were compared with data reported by BLAST analysis (http://www.ncbi.nlm.nih.gov/) against sequences from M. intracellulare ATCC13950 T , M. intracellulare ATCC15985, M. intracellulare MOTT-64, M. intracellulare MOTT-02, M. intracellulare MOTT36Y, and "M. indicus pranii" MTCC9506. For phylogenetic analysis, sequences were trimmed using the CLUSTAL-W multiple sequence alignment program [35]. Phylogenetic trees were obtained from DNA sequences utilizing the neighbor-joining method and Kimura's two parameter distance correction model with 1000 bootstrap replications supported by MEGA 6.0 software [36].

Insertion element analysis
Multiplex PCR was performed to detect four target genes, IS900, IS901, IS1311 and DT1, using previously described methods [37,38]. PCR product sizes of 398 bp, 754 bp, 608 bp, and 296 bp corresponded to amplification of IS900, IS901, IS1311, and DT1 targets, respectively (data not shown). Amplification of only the DT1 gene indicated M. intracellulare. PCR products of insertion elements were sequenced and the existence of a specific insertion element in each strain was confirmed. DNA isolated from Mycobacterium abscessus ATCC19977, Mycobacterium tuberculosis H37Rv ATCC27294, and Mycobacterium gastri ATCC15754 were used as negative controls for each primer set in each PCR run.

Re-identification of clinical isolates by multigene sequence-based typing
Isolates from 77 patients diagnosed with M. intracellulare lung disease were re-identified. Clinical isolates from 74 (96 %) patients were identified as M. intracellulare and those from three (4 %) patients were identified as "M. indicus pranii" using multiple gene sequencing analysis ( Table 3). The 16S rRNA and ITS sequences of "M. indicus pranii" isolates were identical to those of the "M. indicus pranii" type strain (GenBank accession no. CP002275) and the M. intracellulare type strain (GenBank accession nos. GQ153276 and CP003322, respectively). However, the rpoB and nearly complete hsp65 sequences (PCR with hsp65-sequevar primer sets) of "M. indicus pranii" isolates (isolate 01, 46 and 70) were only identical to those of the "M. indicus pranii" type strain (GenBank accession no. CP002275). They were 99.6 % (708/711) and 99.8 % (1413/1416) similar to the rpoB and hsp65 sequences of the M. intracellulare type strain (GenBank accession nos. JQ411539 and DQ284774, respectively). The phylogenetic tree of all isolates with M. intracellulare and "M. indicus pranii" type strains is shown in Figs. 1, 2 and 3.
In all, the 77 isolates were classified to seven different hsp65 sequevars according to the method described by Turenne et al. [33]. Four of these sequevars were well recognized as M. intracellulare type and related strains, and 3 were newly identified in this study. The new sequevars were coded N4, N5 and N6 [followed by the code name given in the previous paper [39]. Ten single nucleotide polymorphisms (SNPs) excluding SNPs reported in the previous study were identified in this study ( Table 4). The distribution of hsp65 sequevars in the 77 isolates is shown in Tables 3 and 4. In all, 74 M. intracellulare and three "M. indicus pranii" isolates were classified as the Min-A ITS sequevar. GCT AGT TGG ATC GCG CCG AAC ACC GG Three patients were re-diagnosed as having "M. indicus pranii" lung disease (Table 5 and Fig. 4). Two patients received combination antibiotic therapy including clarithromycin, ethambutol, rifampin, and streptomycin. Three isolates from each patient were identified as "M. indicus pranii" using multigene sequence-based typing and had no mutations in rrl (23S rRNA gene) according to sequencing analysis, which is known as main mechanism of acquired macrolide resistance in MAC [40]. Patient 1 died of an accident after five months of antibiotic therapy, and patient 2 showed persistent positive sputum cultures after 24 months of antibiotic therapy. Patient 3 was followed up without antibiotic treatment for 5.5 years because of mild symptoms.

Discussion and conclusions
In this study, clinical isolates from 77 patients that were previously diagnosed with M. intracellulare lung disease over a two-year period were re-identified. Species identification was initially performed by a non-sequencing method and then species were re-identified using a sequencing method. Among the 77 isolates identified as M. intracellulare by PRA at the time of diagnosis, 74 isolates were repeatedly identified as M. intracellulare. The remaining three were re-identified as "M. indicus pranii" by multigene sequence-based typing. However, hsp65 and ITS sequevar analyses were not precise enough to discriminate between M. intracellulare and "M. indicus pranii" in this study. To our knowledge, this is the first study to report documented cases of "M. indicus pranii" lung disease in humans. M. intracellulare isolate 64 with code N4 and isolate 53 with code N5 were negative for IS900, IS901, and IS1311, and positive for DT1. Since classification among MAC subsets based on the hsp65 sequevar has been proposed [33], there have been several studies published on M. avium hsp65 sequevars, but none on M. intracellulare hsp65 sequevars. The distribution of M. intracellulare hsp65 sequevars in other countries is unknown. However, two-thirds of the strains from this Koreanbased study were code 13 type, indicating that further studies to characterize this species are needed.
In general, IS1311 is present in all members of the M. avium subspecies and is not present in M. intracellulare. Four M. intracellulare isolates possessed identical IS1311 to that of the M. avium in this study, which is a novel observation. Since a number of different IS elements have been described in various NTM species, the    species-specific IS elements have been revisited for MAC identification [37,41,42]. IS elements are mobile by nature, so there is a risk that similar elements are found in unrelated bacteria because of mobility to or from MAC organisms. Therefore, IS-based PCR differentiation of MAC must be performed in combination with other genetic analyses. Based on this study, DT1 is the optimal candidate marker gene for identification of M. intracellulare and "M. indicus pranii". Sequences analysis of hsp65 and rpoB provides phylogenetic placement, allowing discrimination between the two species. "M. indicus pranii" is initially named "Mycobacterium w (Mw)" and used as a potential leprosy vaccine [43]. The use of the name "Mycobacterium w" gives an impression that Mw is related to the hypervirulent M. tuberculosis-W (Beijing strain) strain. To avoid confusion, Talwar et al. suggested using the nomenclature "Mycobacterium indicus pranii" [44]. However, neither of its name is found on the List of Prokaryotic Names with Standing in Nomenclature, and the designation "Mycobacterium indicus pranii" does not conform to the binomial naming convention used for bacterial species [45]. In a recent publication, Alexander et al. suggested that "M. indicus pranii" is a strain of M. intracellulare [46]. "M. indicus pranii" is considered to be a non-pathogenic microorganism and no human infections have been reported to date [47]. Use of the "M. indicus pranii" vaccine is based on the assumption that antigens shared between M. tuberculosis and this saprophytic mycobacterium is relevant for protective immunity and that "M. indicus pranii" lacks many of the harmful components present in M. tuberculosis [48]. "M. indicus pranii" immunotherapy did demonstrate protective efficacy against tuberculosis [49]. However, patients with pericardial tuberculosis who received "M. indicus pranii" injections demonstrated no significant benefit with respect to any reported outcomes in recent papers [50,51]. The efficacy of "M. indicus pranii" in severe sepsis has recently been reported [52].
On the basis of our findings, "M. indicus pranii" should be considered a cause of pulmonary disease in humans with pre-existing lung disease, such as tuberculosis and bronchiectasis. In addition, the virulence of Classification according to Turenne et al. [33] b • indicates the same base pair as in code 10; New base pair position found in this study are indicated by bold font c New code types found in this study are designated by code N4, N5, and N6 Unlike M. tuberculosis, which has no environmental reservoir, NTM are ubiquitous microorganisms readily isolated from environmental sources, including soil and water. Despite the reportedly low virulence of NTM in immunocompetent human hosts, an increase in their isolation frequency has been seen in the last decade. Genetic analyses have greatly improved our understanding of the phylogeny and evolutionary diversity of NTM. Our study suggests that precise differentiation of M. intracellulare isolates may provide clinically relevant data including ecology, epidemiology, virulence, and treatment outcomes [32,53].
The precise re-identification of clinical isolates initially identified as M. intracellulare by a non-sequencing method in patients with M. intracellulare lung disease revealed that most cases were caused by M. intracellulare. However, some were caused by "M. indicus pranii". Our study indicates the role of "M. indicus pranii" as an agent of severe and chronic lung disease in immunocompetent patients, suggesting that further study is needed to investigate its pathogenicity.

Funding
The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. The views expressed in this article are those of the authors and do not necessarily represent the views of the Korean Health Technology R&D Project and NRF.