Antimicrobial Treatment Improves Mycobacterial Survival in Nonpermissive Growth Conditions

Antimicrobials targeting cell wall biosynthesis are generally considered inactive against nonreplicating bacteria. Paradoxically, we found that under nonpermissive growth conditions, exposure of Mycobacterium bovis BCG bacilli to such antimicrobials enhanced their survival. We identified a transcriptional regulator, RaaS (for regulator of antimicrobial-assisted survival), encoded by bcg1279 (rv1219c) as being responsible for the observed phenomenon. Induction of this transcriptional regulator resulted in reduced expression of specific ATP-dependent efflux pumps and promoted long-term survival of mycobacteria, while its deletion accelerated bacterial death under nonpermissive growth conditions in vitro and during macrophage or mouse infection. These findings have implications for the design of antimicrobial drug combination therapies for persistent infectious diseases, such as tuberculosis.

Benjamini and Hochberg multiple testing correction and fold change >2) comparing treatments with a protective effect at 2 months (C/E/I) to treatments with no effect on long-term survival (C/STR). Bcg1279c (raas) was significantly induced in all samples treated with C/E/I.

pMind-raasmtb pMind
FIG S3 Effect of Raas over-expression on re-growth of M. bovis BCG. One month old pMind or pMind-raas mtb cells were serially diluted (10 -1 to 10 -5 ), inoculated onto 7H10 agar plates and incubated for 4 weeks. The colony sizes reflect different dynamics of recovery, with raas over-expressing bacilli recovering more quickly than bacilli containing an empty vector.

A.
B.

FIG S4 Growth of M. bovis BCG and M. tuberculosis Δraas mutants in vitro.
Mycobacterial strains were inoculated in supplemented Sautonʼs medium and incubated at 37°C with shaking (M. tuberculosis) or without shaking M. bovis BCG. Optical density was measured at 580 nm using a Jenway spectrophotometer.

Supplemented Sautonʼs medium
The following chemicals (g/litre) KH2PO4 0.5 g, MgSO4 0.5 g, asparagine 4 g, citric acid 2 g, ferric ammonium citrate 0.05 g were dissolved in Milli-Q water. Glycerol (1 ml) and 100 µl of 1 % (w/v) ZnSO4 were added and pH was adjusted to 7.5 before autoclaving at 121°C for 15 min. Sterilised Sautonʼs medium was kept in dark place and used within 2 weeks. 900 ml of autoclaved Sautonʼs base were mixed with 100 ml of ADC supplement and 10 ml of Tween 80 10% (w/v) solution. ADC supplement contained (per 1 Litre) 50 g of Bovine serum albumin, 20 g of glucose and 8.5 g NaCl. ADC supplement and Tween 80 solutions were filter-sterilised.

Concentrations
To measure antibiotic sensitivity, serially diluted cells were spotted onto agar containing different concentrations of antimicrobials and the plates incubated at 37°C for up to 8 weeks, before counting colony forming units. Serially diluted cells were also inoculated into liquid medium containing drugs as a further measure of drug sensitivity. For isoniazid, ethambutol and reserpine, filter disc assays were also used. Briefly, 6 mm paper discs (Whatman), soaked with a range of antibiotic concentrations were applied to agar plates inoculated with a lawn of 10 7 M. bovis BCG cells. Plates were incubated at 37°C for 4 weeks, before measuring the diameters of the resultant inhibition zones. Comparative spot intensities from the images were calculated using Imagene 5.5 (BioDiscovery) and imported into GeneSpring GX 7.3.1 (Agilent Technologies). The array data were normalized to the 50 th percentile of all genes detected to be present on the array, and filtered to include only genes flagged to be present on 80% of the arrays.
Significantly differentially expressed genes were identified in pair-wise comparisons between control and antibiotic-treated cells using a t-test (p-value <0.05 with Benjamini and Hochberg multiple testing correction) and a fold change >2. Five genes were significantly induced by cerulenin, and ethambutol and isoniazid relative to drug-free bacilli but were not induced by streptomycin exposure (Table S1). Similar clusters of genes (all including Rv1219c) were identified in alternative analysis strategies by grouping antibiotic treatments by protective effect after prolonged stationary phase ( Figure S2).
Microarray data were confirmed by quantitative RT-PCR. DNA contamination was removed with Turbo DNA-free DNAase (Ambion) before cDNA was generated using Superscript Reverse Transcriptase II (Invitrogen) and gene-specific primers (Table S2).
Q-PCR was performed in a Corbett Rotor Gene 6000 real time thermocycler using Absolute QPCR SYBR Green mix (Thermo) and gene-specific primers. Levels of expression were normalized to 16s rRNA (2).

Transcriptomics of Raas Knockout and Complemented M. bovis BCG Strains
Mycobacterial RNA was extracted from WT, raas deleted and complemented M. bovis BCG strains using the GTC/Trizol method (3), then DNase-treated and purified using RNeasy columns (Qiagen). Microarray hybridizations were conducted as described above, hybridizing RNA derived from three biological replicates. Significantly differentially expressed genes were identified comparing knockout mutant to both WT