Pulmonary samples
From January 2002 to November 2016, a total of 10 466 pulmonary samples (7699 sputum, 2093 bronchial washes, 674 pleural fluids) obtained from TB suspects, originating mainly from Northern Tunisia (Tunis, Bizerte, and Zaghouan), was processed for mycobacterial isolation. For those patients with serial samples, only the very first specimen was taken into consideration.
Mycobacterial isolation and phenotypic characterization
Briefly, pulmonary specimens (sputum, pleural fluid, bronchial wash) were processed by N-acetyl L cysteine-sodium hydroxide (NALC-NaOH) method, and inoculated into Löwenstein-Jensen (LJ) medium. Inoculated LJ slants were incubated at 37°C for eight weeks and examined every week.
Differentiation between MTBC and NTM colonies were initially performed by four standard biochemical tests: niacin, nitrate, heat-resistant catalase test (HRCT) and para- nitro benzoic acid (PNB). The NTMs were confirmed by PCR targeting the recA intein, as previously described [9], and were further characterized phenotypically by morphological character and biochemical tests, i.e colony morphology, growth rate, growth at various temperatures (25°C, 37°C and 44°C), pigment production in dark (schotochromogen), pigment production on exposure of light (photochromogen), absence of pigment production (non-chromogen), semi-quantitative Catalase test (SQCT), Thiophene2-carboxylic acid hydrazide (TCH) Susceptibility Test, Tween 80 hydrolysis, Aryl sulphatase test (3 days and 14 days), and β-Gal activities.
PCR amplification and nucleotide sequencing
From a few colonies cultivated on LJ slants, the DNA was extracted by thermolysis. For this purpose, the colonies are suspended in 100 μl of 1% EDTA / Triton Tris (TE / Triton) and inactivated at 80°C for 30 minutes.
PCR amplifications and sequencing of the rpoB, 16 S rRNA, hsp65, and sodA gene sequences were performed using the primer pairs listed in additional file 1.
PCR reactions consisted of a15-min inactivation period followed by 35 cycles of 95 °C for 30s, 60 °C for 1min (64°C for 1 min for rpoB, 52°C for 30 sec for 16S rRNA) and 72 °C for 2 min, with a final extension step at 72 °C for 5-7 min.
PCR amplifications were performed in a 50-μl PCR mixture containing 5 μl 10x buffer (Qiagen, Courtaboeuf, France), 200 mM each dNTP, 1.5 mM MgCl2, 1.25 U HotStarTaq polymerase (Qiagen), 1 μl each primer (10 pM), 33 μl nuclease-free water and 5 μl DNA template. Negative controls consisting of PCR mixture without DNA template were included in each PCR run. The PCR products were visualized by gel electrophoresis, treated with Shrimp Alkaline Phosphatase (SAP) and Exonuclease I (ExoI) (Sigma-Aldrich,USA), and sequenced in both directions using the BigDye Terminator sequencing kit (Applied Biosystems, Villebon-sur-Yvette, France) with an ABI PRISM 3100 automatic sequencer (ABI, USA). The sequences generated in the present study were deposited to the GeneBank. Their accession numbers are provided in additional file 2.
Nucleotide sequences were processed using BioEdit (BioEdit software, version 7.0.5), and aligned with the GenBank database (NCBI) using the Basic Local Alignment Search Tool (BLAST).
Species identification
Species identification was primarily based on similarity rates with rpoB gene sequence of reference strains (http://www.bacterio.net/mycobacterium.html) [10]. Briefly, assignment of an isolate to a particular species was allowed if its rpoB sequence displayed at least 99.3% similarity to type strain for slow growers and 98.3% for rapidly growing mycobacteria. Isolates showing similarity rates below these thresholds but ≥ 94% were identified to complex level. Isolates displaying less than 94% similarity were referred to as rpoB unidentified Mycobacterium species (rUMS). With regard to 16S rRNA gene sequencing, identification was based on 100% matches.
Subtyping of Mycobacterium kansasii was performed by PCR-restriction fragment length polymorphism (PCR-RFLP) analysis of the hsp65 gene as described previously [11].
Phylogenetic analyses
Phylogenetic analyses were carried out with MEGA (Molecular Evolutionary Genetics Analysis, version 7.0) [12]. Phylogenetic tree analyses were performed using the neighbor-joining method based on the Kimura two-parameter model with 1000 bootstrap replicates. Nocardia nova was used as the outgroup.