Inverse correlation between salt tolerance and host-adaptation in mycobacteria

The genus Mycobacterium includes host-adapted organisms regarded as obligate and opportunistic pathogens and environmental organisms. Factors contributing to this wide range of adaptations are poorly known. We studied the salt tolerance of 46 Mycobacterium species of medical interest. Representative strains of the Mycobacterium tuberculosis complex, Mycobacterium avium complex, Mycobacterium chelonae-abscessus complex, Mycobacterium ulcerans, Mycobacterium marinum, Mycobacterium lentiflavum, Mycobacterium fortuitum and Mycobacterium conceptionense were inoculated on Middlebrook 7H10 medium supplemented with 0–10 % sodium chloride. Colonies were counted after 2–4 week incubation at the appropriate 30–37 °C temperature depending on the tested strain. Further comparative genomics was done on 15 Mycobacterium strains representing the spectrum of salt-tolerance of mycobacteria. Based on the results the different species were grouped according to their salt tolerance into a “salt-sensitive” group (growth up to ≤3 % salt) containing the M. tuberculosis complex, Mycobacterium chelonae, Mycobacterium lentiflavum, Mycobacterium ulcerans and Mycobacterium marinum; a “salt-intermediate” group (growth between 4 and 6 % salt) comprising Mycobacterium avium, Mycobacterium intracellulare, Mycobacterium chimaera and a “salt-resistant” group (growth up to >6 %) comprising Mycobacterium homonissuis, Mycobacterium bolettii, Mycobacterium fortuitum and Mycobacterium conceptionense. Genomic analysis revealed that 290 genes were unique to species belonging to the salt-sensitive group; and that 15 % were annotated as being functionally associated with the ESX secretion systems Pro-Glu and Pro–Pro-Glu family proteins. In this work we found an inverse correlation between salt tolerance and host adaptation. We thus propose that salinity is one of the multiple factors determining the ecological niches of mycobacteria.


Background
The genus Mycobacterium comprises more than 150 species [1]. The vast majority of these mycobacteria are environmental organisms found in soil and aquatic environments, with a few exhibiting some degree of host-adaptation, illustrated by their intra-amoebal survival [2] and variable pathogenicity in mammals and humans, culminating in well-adapted Mycobacterium leprae responsible for animal and human leprosy [3,4] and Mycobacterium tuberculosis, complex organisms responsible for animal and human tuberculosis [5]. Later organisms exhibit the widest spectrum of niches from soil [6,7], to amoeba [2] and mammals including humans [5]. The factors contributing to the survival of mycobacteria in one particular ecological niche are not yet fully understood.
Here, we explored salinity as one of the factors which could potentially affect the survival of mycobacteria in their ecological niches. More precisely, we focused our study on a few species of veterinary and medical interest, as they exhibit the broadest spectrum of ecological styles, from inanimate environments to amoeba and hosts.  [8]. M. tuberculosis, "M. canettii" and M. ulcerans strains were handled in a biosafety class 3 laboratory, while other mycobacteria were handled in a biosafety class 2 laboratory. Except for "M. canettii", all clinical strains have been isolated and cultured by the authors. No ethical approval was required for this study. All mycobacteria were subcultured on Löwenstein-Jensen medium; (bioMérieux, Craponne, France), suspended in sterile phosphate buffered saline (PBS) and vortexed to complete homogenization of the suspension. As for clumping M. tuberculosis, M. bovis BCG, M. abscessus and M. ulcerans, homogenization was achieved by rigorously vortexing with 3-mm glass beads (Sigma-Aldrich, Saint-Quentin-Fallavier, France) followed by four bypasses through a 25-G needle to disperse the remaining clumped bacilli. Homogenized suspensions were calibrated using spectrophotometry to a 1 Mc Farland (McF) unit, equivalent to a 10 7 colonyforming units (CFU)/mL inoculum for non-clumping mycobacteria and a 10 6 CFU/mL inoculum for clumping mycobacteria.

Culture
Middlebrook 7H10 medium (Becton-Dickinson, Le Pont-de-Claix, France) was supplemented with sodium chloride (NaCl) (Sigma-Aldrich, Saint-Quentin-Fallavier, France) in 1 % increasing salt concentrations from 0 (weight/volume) to 10 %. A 100 µL-volume containing 10 4 CFUs was inoculated in triplicate on 90-mm sterile lentiflavum were incubated at 35 °C in ambient air as previously described [9]. Plates inoculated with M. ulcerans and M. marinum were incubated at 30 °C in ambient air. Plates were checked by visual inspection weekly for colonies for 4 weeks. Colonies were counted regardless of their size and counting was considered interpretable when >10 3 colonies were observable on the control (0 %-NaCl) Middlebrook 7H10 medium plates. A strain was considered salt tolerant when more than 50 colonies developed on the 7H10-NaCl supplemented media. Image J program [10] was used to measure the average size of colonies after 50 colonies were randomly chosen from each plate. The morphology of colonies was observed by the naked-eye. Ziehl-Neelsen staining was conducted to confirm the identity of the colonies.

Genome analyses
The whole genome and proteome of 15 mycobacterial species under investigation were downloaded from Genbank (Table 1). The proteins were clustered into orthologous groups using orthoMCL [11] with a conservative parameter value of 60 % sequence identity. Homologous sequences were selected using the all-against-all BlastP algorithm [12] with an E value of <10 −5 . Clustering of the orthologous sequences was then analyzed using the Markov Cluster algorithm [13]. Determination of the different unique core genomes was based on the homology clusters found by orthoMCL. The resulting orthologous groups were used to construct a whole-genome tree using the Neighbor-Net algorithm based on a gene content matrix using splitree [14]. The similarity between two species was defined as the number of genes in common divided by the total number of genes of the two species [15].  (Tables 2, 3). Based on these results, we defined three groups of mycobacteria: Group 1 includes "salt-sensitive" species (growth up to ≤3 % salt); Group 2 includes the "salt-intermediate" species (growth up between to 4 and 6 % salt), while Group 3 includes "salt-resistant" species (growth up to >6 % salt).

Colony size and morphology
Except for M. bolettii, we observed that the size of colonies significantly decreased as salinity increased ( Table 3). As for M. bolettii, the size of colonies increased from 1.16 ± 0.4 mm in the Middlebrook 7H10 control medium up to 2.95 ± 0.9 mm in 3 % NaCl Middlebrook 7H10 medium (P < 0.05, student's t test), then decreased down to 0.4 ± 0.2 mm in 7 % NaCl Middlebrook 7H10 medium (P < 0.05, student's t test) ( Fig. 1; Table 4).

Discussion
We observed a previously unrated wide range of salt tolerance for mycobacteria, from 1 % (M. lentiflavum) to 8 % (M. fortuitum) among mycobacteria of veterinary and medical interests comprising of obligate and opportunistic pathogens. Data here reported were authenticated by being reproduced in triplicate. Furthermore, data here obtained for M. abscessus and M. chelonae agree with the previously reported 50 g/L salt tolerance for the same species using a comparable methodology [9]. Likewise, the 5 % limit here found for M. intracellulare, has been previously reported using a slightly different methodology [9]. The data here reported for M. tuberculosis also agree with those previously reported for M. tuberculosis and Mycobacterium bovis using the BioLog technique [19]. In a later study, M. tuberculosis H37Rv as well as a M. tuberculosis Beijing strain were shown to be highly susceptible to salt with metabolic activity exponentially dropping as salt concentration increased from 0 to 3 % [19]. Accordingly, we observed that M. tuberculosis growth was inhibited by salinity rate >3 %.
In this study, we observed that M. tuberculosis complex members tolerate up to 3 %, most M. avium complex members between 4 and 5 % and M. abscessus and M. bolettii up to 7 %. Furthermore, variations were observed within these phylogenetic complexes, as previously reported for M. tuberculosis and M. bovis in the M. tuberculosis complex [19]. In the M. avium complex, M. chimaera exhibits a salt tolerance limited to 3 %, which is much lower than that of M. avium subsp. homnissuis which is limited to 7 %-NaCl. The same observation holds true for the M. chelonae-abscessus complex where M. chelonae exhibits a 3 % tolerance whereas M. massiliense grew up to 4 % and M. bolettii up to 7 %. It is worth noting that these two species yielded quite different colony morphology, which was smooth for M. bolettii and rough for M. abscessus. Such notable differences in salt-tolerance could be further incorporated in the easy identification of colonies in clinical microbiology, as previously reported [9].
Rather than observing a correlation with the phylogenetic position, we observed a correlation with the  genome content and encoding capacity. Indeed, we found that Group 1 salt-susceptible mycobacteria contained a 290-gene core that is absent in salt-tolerant groups. It is worth noting that more than one-third of these genes encode for PE/PPE family proteins. PE/PPEs are functionally associated with type VII or ESX secretion systems and could act as virulence factors helping the bacteria to establish a successful infection inside the host [20][21][22]. We therefore observed an inverse correlation between salt tolerance and host adaptation (Fig. 3)   Denotes P < 0.05 when compared to NaCl-0 % terrestrial mammals [24]. General stress tolerance proteins [16], Na+/H+ efflux pumps [17], porins and genes implicated in porin regulation [18] are of utmost importance for bacteria in order to adapt to environmental changes such as the presence of toxic compounds including salt. Strangely, only three proteins harboring the GsiB domain and a putative enoyl-CoA hydratase (EchM) previously reported to be responsible for salt tolerance [16] were found with low identity <30 % in M. fortuitum, M. avium and M. hominissuis, which tolerate salt. Na+/ H+ efflux pumps proteins and porins were found only in strains which are intermediate or resistant to salt.

Conclusions
In conclusion, we propose that salinity is one of the multiple factors which determine the ecological niches of mycobacteria, with tolerance to salt being roughly inversely correlated with host adaptation.

Table 5 Clusters of orthologous genes (COG) classification of 290 core genes specifically found in Group 1 mycobacteria
There was no unique gene for Group 2 and Group 3 mycobacteria