Bacteria strains, media and growth conditions

The mycobacterial strains used in this study are listed in Table S1. The strains were routinely cultured in standard Middlebrook 7H9 and 7H10 supplemented with 0.05% Tween 80 and 0.5% and 0.2% of glycerol respectively. The M. tuberculosis cultures were further supplemented with 0.5% bovine serum albumin, 0.085% sodium chloride and 0.2% dextrose (ADN).

The growth in defined iron and zinc conditions was performed in Minimal Medium (MM): 0.5% w/v KH₂PO₄; 0.5 % w/v asparagine; 0.2% glycerol; 0.05% tween 80; ADN 10%, pH 6.8 treated with Chelex100 (Sigma) at 4°C for 36 h, replacing Chelex 100 every 8-12 h. Chelex 100 was removed by filtration and the medium was supplemented with 40mg/L MgSO₄; 0.1mg/L MnSO₄ and the desired amounts of iron and zinc in the form of FeCl₃ and ZnSO₄, respectively: 50 µM iron and 3,67 µM zinc for complete MM (CMM); 2 µM iron and 3,67 µM zinc for low-iron MM (LIMM) or 0 µM for no iron supplemented MM (NIMM); 50 µM iron and no zinc for low-zinc MM (LZMM). When required, antibiotics were added at the following concentrations: 20 µg/ml kanamycin (Km); 20 µg/ml streptomycin (Sm); 50 µg/ml hygromycin (Hyg).

Escherichia coli DH5α was routinely used in the DNA-cloning procedures. The cells were grown in Luria Bertani (LB) broth and LB-agar. When indicated, antibiotics were added at the following concentration: 50 µg/ml Km; 20 µg/ml (Sm); 200 µg/ml Hyg.

For growth analyses, M. tuberculosis strains were pre-grown in rolling bottles in 20 ml of MM or Middlebrook 7H9 with or without 200 ng/ml of anhydrotetracycline (ATc) for 48 h. This concentration has been previously determined to stop the growth of the esx-3 conditional mutant [26]. Subsequently, the cultures were diluted 1:10 in fresh medium plus or minus 200 ng/ml ATc and the growth was monitored measuring the increase in optical density (OD) at 540 nM. M. smegmatis strains were pre-grown overnight in 5 ml of MM or Middlebrook 7H9 in shaking cultures until late exponential phase, diluted to an OD₅₄₀ of 0.025 and the growth was monitored for three days.

Construction of M. smegmatis mutants

To create the unmarked esx-3 deletion in M. smegmatis, we used the two-step cross-over strategy based on pNIL-pGOAL [37]. Plasmids and primers used for mutants construction are shown in Tables S1 and S2. DNA regions of about 1 kb flanking msmeg_0615 and 3’ to msmeg_0624 were amplified from M. smegmatis mc² 155 genomic DNA using the RP1215/RP1216 and RP1217/RP1218 primer pairs. The two PCR fragments were cloned in pCRBluntII-TOPO (Invitrogen) and then subcloned together in p2NIL obtaining the plasmid pAGN25 (Figure S1). The sacB/lacZ/hyg cassette from pGOAL19 [37] was then cloned in the unique PacI restriction site of pAGN25 to obtain the final suicide plasmid pAGN27 (Figure S1). This plasmid was introduced by electroporation in M. smegmatis mc²155 and the transformants were selected on medium containing Hyg. Blue colonies expressing β-galactosidase were selected. The first cross over event was verified by PCR using the primer pair RP1277/1278 and RP1258/1279 flanking the homologous regions used for recombination (Figure S2A-B). Transformants were grown in medium without antibiotic to allow the second cross over event, and then selected on Middlebrook 7H10 supplemented with 8% sucrose and X-gal (40 µg/ml). Sucrose resistant and white colonies were screened by PCR using the primer pair RP1277/RP1279 flanking the deleted region to confirm esx-3 deletion (Figure S2C). The esx-3 null mutant was named MS190.

To generate the unmarked deletion of the fxbA gene in M. smegmatis we used the strategy showed in Figure S3. Briefly, 0.5-kb regions flanking the fxbA gene were amplified from M. smegmatis genomic DNA using the primers couples RP1261/RP1262 and RP1263/RP1264. The PCR products were then subcloned together in pCRBluntII-TOPO (Invitrogen) obtaining the plasmid pAGN29. The Hyg cassette flanked by dif sites from pAL70 [38] was then inserted in the unique BglII site between the two fxbA-flanking fragments in pAGN29 obtaining pAGN30. The cassette including the fxbA flanking regions and the Hyg-resistance determinant flanked by dif sites was extracted from pAGN30 and integrated using the recombineering technology [39] in the genomes of MS192 and MS81.1 (MS190 and mc²155 derivatives harboring the pJV53 vector carrying the genes encoding the two recombinases that mediate homologous recombination). Transformants were selected on Hyg and the integration was verified using the primer pair RP472/RP1308 and RP747/RP1309. Then, the transformants were grown in medium not supplemented with hygromycin to allow the excision of the Hyg cassette due to the recombination between the dif sequences. The excision was verified by the inability of the clones to grow in presence of Hyg and by PCR using the primer pair RP1308/RP1309. The deletion was further confirmed by sequencing. The esx-3/fxbA double mutant and the fxbA single mutant were named MS195 and MS197, respectively.

RNA extraction

Total RNA was extracted from bacteria in early exponential phase grown in rolling bottles. Cells were harvested at 1000 g for 5 min. The total RNA was extracted as previously described [27]. Briefly, the cells were resuspended in 1 ml Trizol (Invitrogen) and added with 0.5 ml of 0.1-mm-diameter zirconia/silica beads (BioSpec Products) and disrupted with two 1-min pulses in a Mini-Beat Beater (BioSpec Products). The nucleic acids were extracted with a solution of chloroform: isoamyl alchol (24:1) and precipitated overnight at 4°C with isopropanol. After centrifugation, the pellet was washed with 75% ethanol and resuspended in water. DNA was removed treating samples with 4 units DNase (Ambion) for 1 h and the RNA was purified using the RNeasy kit (Quiagen). The treatment with DNase and the following clean up were performed twice to completely remove DNA. The concentration and purity of RNA were determined spectrophotometrically, integrity was verified on 1.3% agarose gel and by Agilent 2100 Bioanalyzer.

cDNA labeling and microarray analysis

DNA array analysis was performed on RNA samples extracted from the esx-3 conditional mutant TB79 and its parental strain TB38 grown in 35 ml of Middlebrook 7H9 added with 100 ng/ml of ATc for 48 h. The time and ATc concentration were chosen to ensure that TB79 cells were still actively dividing even if esx-3 mRNA was at least 80% lower than in TB38 (data not shown).

Fluorescently-labeled cDNA was produced as previously described [40] by reverse transcription (RT) of total RNA with Superscript II (Invitrogen LifeTechnologies) in the presence of Cy3-dUTP or Cy5-dUTP (Amersham Pharmacia) by using random hexamers to prime cDNA synthesis (Invitrogen Life Technologies).

M. tuberculosis oligoarrays from the Center for Applied Genomics, International Center for Public Health (Newark, NJ, USA), consisting of 4295 70-mer oligonucleotides representing 3924 open reading frames (ORFs) from M. tuberculosis strain H37Rv and 371 unique ORFs from strain CDC 1551 that are not present in H37Rv were used for the hybridization.

Hybridization on DNA microarrays was performed as previously described [41]. Briefly, the microarrays were pre-hybridized in 0.1% SDS and 3% BSA, washed in water and in isopropanol and finally air-dried. Competitive hybridizations with equal amounts of purified Cy3/Cy5-labelled cDNA (approx. 60 pmol) were performed in duplicate with both dye arrangements.

Microarrays were scanned using an Agilent G2565CA microarray scanner system [with Scan Control 8.1 software (Agilent Technologies)], and fluorescence intensities of the two channels at each spot were quantified using Agilent Feature Extraction 10.1 software (Agilent Technologies).

Data were normalized with the Web-based tool DNMAD [42], using the print-tip lowess method after background subtraction. Genes significantly differentially expressed were identified using the significance analysis of microarrays (SAM) tool [43], part of the Tiger MultiExperiment Viewer package, version 4.6 (TMeV), available at http://www.tm4.org/mev/ [40].

Differently expressed genes were defined by a q-value equal to 0 and a fold difference ≥1.8 (induced) or ≤0.6 (repressed). Differentially regulated genes with q-value until 5.58 were included if part of an operon or regulon differentially regulated or if their up regulation was observed by quantitative real-time RT-PCR (i.e. irtA and irtB).

DNA array data have been deposited in NCBI's Gene Expression Omnibus [44] and are accessible through GEO Series accession number GSE47834 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE47834).

Quantitative real-time RT-PCR

Quantitative real-time RT-PCR to confirm alterations in the expression of genes identified by microarray was performed on RNA samples collected in the following ways: i) TB38 and TB79 grown in 35 ml of LIMM and Middlebrook 7H9 media with 100 ng/ml ATc for 48h; ii) TB38 and TB79 were pre-grown in CMM with or without 200 ng/ml of ATc, after 48h the cultures were diluted 1:10 with fresh medium containing 100 ng/ml of ATc and the RNA samples were collected after 48 h. In CMM two passes were performed since after a single pass, even if repression of esx-3 was clear, up-regulation of the iron and zinc repressed genes was very weak.

500 ng of the total RNA were retro-transcribed using random primers and murine leucoblastoma virus retro-transcriptase (MULV) (Applied Biosystem) as previously described [27]. The RNA was denatured 2 min at 98°C and added to a solution containing 5.5 mM MgCl₂, 1X MULV Buffer (Applied Biosystem), 250µM dNTPs, 2.5µM random hexamers, 10 units RNAse inhibitor, 31 units MULV enzyme (Applied Biosystem) in a final volume of 25 µl; the retro-transcription was carried out with the following cycle: 10 min at 25°C for random hexamers annealing, 50 min at 45°C for retro-transcription, 5 min at 95°C to inactive the retro-transcriptase.

Fifty ng of retro-transcribed RNA were used in each PCR reaction in a final volume of 25 µl. The quantitative real-time RT-PCR was performed with SYBR Green master mix 2X (Applied Biosystem) in Applied Biosystems 7700 Prism spectrofluorometric thermal cycler (Perkin-Elmer): a first cycle of denaturation at 95°C for 10 min to activate the enzyme; 40 cycles of 1 min denaturation, 30 sec annealing at 60°C, 30 sec elongation at 72°C. Fluorescence was measured during the annealing step and plotted automatically for each sample. The results were normalized to the amount of sigA mRNA [45], whose expression levels are not affected by the concentrations of intracellular iron or zinc (data not shown). RNA samples that had not been reverse transcribed were included in all experiments to exclude significant DNA contamination. For each sample, melting curves were performed to confirm the purity of the PCR products.

The fold induction was showed as ratio between sigA cDNA-normalized cDNA levels of: i) conditional mutant/parental strain both grown with ATc; ii) conditional mutant grown with ATc/conditional mutant grown without ATc. The primers used in these experiments are listed in Table S3.

Streptonigrin (STN) sensitivity assay

M. tuberculosis was grown in Middlebrook 7H9 in rolling bottles with or without 100 ng/ml ATc for 72 h. After the first 24 h the cultures were divided in two tubes and 5 µg/ml of STN was added to one tube. After 72 h of incubation the viability of both cultures was determined by plating onto Middlebrook 7H10 plates.

M. smegmatis was grown in LIMM overnight, the cultures were diluted to OD₆₀₀ of 0.1 in 20 ml of same type of medium and incubated with STN 5 µg/ml for 30 min. Viability, determined by plating serial dilution of the cultures on Middlebrook 7H10 plates, was evaluated at time 0 and 30 minutes.

Preparation of ⁵⁵Fe- carboxymycobactin and iron uptake assays

Ferri-carboxymycobactin (Fe⁺³-CM) labeling and iron uptake was performed as previously described [34]. Briefly, purified Fe-CM was deferrated by incubation with one volume of 50mM EDTA, pH 4.0 for 18 h or until the absorbance at 450 nm decreased to less than 10% of its initial value. The solution was then extracted into chloroform, the chloroform evaporated, and the residue resuspended in 50% ethanol. Desferri-CM was then mixed with ⁵⁵FeCl₃ (PerkinElmer) in a ratio of 2:1 CM to iron, and the mixture was incubated 10 min at room temperature. ⁵⁵Fe³⁺-CM was then extracted into chloroform and washed twice with water to remove free unbound ⁵⁵Fe³⁺. The chloroform was evaporated, and ⁵⁵Fe³⁺-CM was suspended in 50% ethanol and kept in small aliquots at -20°C.

For iron uptake experiments mycobacterial strains were iron depleted by pre-growing them in NIMM for two generations and then in the presence or absence of 200 ng/ml ATc for two more generations before assessing iron uptake. At this time the cells with and without ATc were replicating at the same rate. Cultures were collected by centrifugation, diluted to 2X10⁸ bacilli per ml in NIMM and 1 ml of cells the cell suspension was incubated with ⁵⁵Fe⁺³-CM (10⁶ cpm) at 0 or 37°C. Samples of 0.5 ml were removed at indicated time points, immediately placed in ice and uptake stopped with cold 0.1 M LiCl. Bacteria were collected by centrifugation and washed twice with cold NIMM. The bacterial pellet was resuspended in scintillation fluid and radioactivity determined using a Beckman LS6500 scintillation counter set at the wide-open window setting. The data are expressed as cpm incorporated by 1X10⁸ cells at each time point after subtracting the radioactivity incorporated in cells incubated on ice, which was less than 10% of total incorporation.

Growth in minimal medium with hemin as iron source

Stock solutions 10 mM hemin (Sigma) and 50 mM of the iron chelator deferoxamine mesylate salt (DFO) (Sigma) were prepared fresh before each experiment dissolving them in 10mM NaOH and water respectively. M. tuberculosis strains were grown in NIMM, in NIMM supplemented with 50 µM DFO, in NIMM supplemented with 5 or 10 µM of hemin and 50 µM DFO. In these conditions the strains were pre-grown for 48 h with or without 200 ng/ml of ATc and subsequently the cultures were diluted 1:10 in fresh media with the same composition and the optical density was monitored for a week. M. smegmatis strains were pre-grown overnight in 5 ml of LIMM and next diluted to OD₅₄₀ of 0.025 in NIMM supplemented as in M. tuberculosis. The growth was monitored for about 72 h.

MIC determination by microplate Alamar Blue assay

M. smegmatis cultures were grown overnight in 5 ml of 7H9, diluted until final concentration of 10⁴ cells/ml. 5µl of antibiotic stocks were added to the first well of black 96-micro well plate (NUNC) containing 200 µl of diluted culture and 1:2 serial dilutions were performed in the next wells containing 100 µl of the same diluted culture. The microplate was incubated at 37°C for 48 h and then 10 µl of Alamar Blue dye (Invitrogen) was added to each well and analyzed by fluorescent spectrophotometer (Infinite F200-PRO Filter/Tecan) after 2 h. As a control Alamar Blue was also added to a well containing cells not treated with drugs and a well containing only medium (background value). After background subtraction the fluorescent value obtained from each well was normalized to the fluorescence value obtained from the well containing cells without antibiotic. The concentration able to reduce the fluorescence signal of 90% was considered the minimal inhibiting concentration (MIC). Each experiment was performed in duplicated.

SDS sensitivity

M. smegmatis cultures were grown overnight in 5 ml of Middlebrook 7H9, diluted to an OD₆₀₀ of 0.1 in 20 ml of Middlebrook 7H9 and incubated with 0.1% SDS for 120 min at 37°C in shaking flasks. Culture samples were taken at 0, 30, 60 and 120 min and serial dilutions plated on Middlebrook 7H10 plates for enumerating CFUs.

THP1-derived macrophages infection

THP-1 monocytes (American Type Culture Collection) were grown in suspension at 37°C in 5% CO2 in RPMI (Gibco) supplemented with 10% (vol/vol) fetal bovine serum (FBS) (Gibco) and 50 μM β-mercaptoethanol. The infection was performed as previously described [46]. Briefly, the cells were plated in 96-well plates at a density of 7.5 × 10⁴ cells/well and incubated with 50 ng/ml of phorbol 12-myristate 13-acetate (PMA) (Sigma) for 24 h to induce the differentiation in macrophages. THP-1-derived macrophages were infected with M. tuberculosis at a multiplicity of infection of 1:20 (CFU:macrophages) for 90 min at 37°C with mycobacteria previously grown in Middlebrook 7H9. Cells were washed with PBS to remove extracellular bacteria, and fresh medium with or without 200 ng/ml ATc was added. The medium was replaced every 48 h. At the time 0 and every 24 h for 6 days the medium was removed and then intracellular bacteria were released by lysing the macrophages with 0.05% SDS. The suspensions obtained from the lysed macrophages were immediately diluted in Middlebrook 7H9 and plated on Middlebrook 7H10 to determine viable counts. During the experiments, the macrophage viability was evaluated by staining with Trypan blue.

Metal-dependent growth of ESX-3-depleted M. tuberculosis

In a previous work we constructed an M. tuberculosis conditional mutant (TB79) in which the expression of the esx-3 gene cluster is repressed by ATc and showed that its growth in Middlebrook 7H9 stopped after few generations upon esx-3 repression [26].

In M. tuberculosis the esx-3 gene cluster is regulated by the iron dependent repressor IdeR and by the zinc uptake regulator Zur [27,47], suggesting its involvement in the adaptive response to the availability of these two metals. To challenge this hypothesis the esx-3 conditional mutant (TB79) and its parental strain (TB38) were cultured in defined MM with different concentrations of iron or zinc. In our previous work we found that to detect the TB79 growth phenotype in Middlebrook 7H9, liquid cultures had to be pre-grown in 7H9 in the presence of ATc before being diluted 1:10 in fresh medium containing ATc (Figure 1 E). Applying the same protocol, we found that the conditional mutant TB79 cultured in low iron (LIMM) and low zinc (LZMM) with ATc, stopped dividing after the first dilution (Figure 1 C-D), while in complete MM (CMM) plus ATc its growth was arrested only after a second dilution (Figure 1 A-B). These results indicate that the repression of esx-3 reduces growth in both low iron and low zinc medium and that sustained repression of this system prevents normal growth even under iron and zinc sufficient conditions. Surprisingly, the growth defect experienced by maintaining ESX-3 repression was more evident in Middlebrook 7H9 than in CMM, despite the presence of an iron concentration three fold higher in Middlebrook 7H9 than in CMM and an equal zinc concentration in the two media.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Growth of ESX-3 depleted M. tuberculosis.

Strains were grown in rolling bottles with or without ATc 200 ng/ml in A) CMM (50μM FeCl₃ and 3,67 μM ZnSO₄) first dilution; B) CMM second dilution; C) LIMM (2μM FeCl₃ and 3,67μM ZnSO₄) first dilution; D) LZMM (50μM FeCl₃ and no ZnSO₄) first dilution; E) Middlebrook 7H9 (^≈150μM ferric ammonium citrate and ^≈3,67μM ZnSO₄) first dilution. The arrow indicates the time point at which the second dilution was performed. Growth of the parental strain in all tested conditions was the same with or without ATc: only growth curve of cultures with ATc are shown. Values represent the average of three independent experiments with relative standard errors.

https://doi.org/10.1371/journal.pone.0078351.g001

Transcriptional profile of ESX-3-depleted M. tuberculosis

To better characterize the impact of esx-3 down-modulation on M. tuberculosis physiology, we compared the global transcriptional profiles of the esx-3 conditional mutant and its parental strain grown in Middlebrook 7H9 supplemented with ATc. Cells were collected after 48 h of exposure to 100 ng/ml of ATc. In these conditions esx-3 transcription was repressed about 35 fold (Figure S4), while growth and vitality of the mutant were not drastically affected (data not shown). In addition to the genes belonging to the esx-3 cluster, the expression of 70 genes was significantly altered in the conditional mutant cultured in the presence of ATc. A total of 52 genes were up-regulated and 18 were down-regulated (Table S4). Forty of these genes are regulated by iron and IdeR and encode proteins involved in adaptation to low iron conditions. Up-regulated genes included those encoding the mycobactin synthesis proteins (Mbt) [48], the membrane proteins MmpS4/L4, involved in siderophore export [49], the ABC transporter IrtAB, responsible for Fe⁺³-carboxymycobactin uptake [33] and the SUF [Fe-S] cluster assembly system (Rv1460-Rv1466) [50]. However, the nuo operon, encoding the different components of an NADH:ubiquinone oxidoreductase (rv3145-rv3158) known to be induced in high iron [47], was down regulated in ESX-3-depleted cells. Eight genes belonged to the Zur regulon, in particular the rv2055c-rv2058c operon, encoding ribosomal proteins hypothesized to be involved in the release of zinc from the cytoplasmic stores and the rv2059-rv2060 operon, encoding a putative zinc transport system [27,51,52] were all up-regulated in the absence of ESX-3. Finally, 22 genes not known to be iron or zinc regulated and encoding proteins of unknown function, proteins involved in stress response as AhpD or proteins containing iron as the NADH dehydrogenase Ndh were either up or down-regulated (Table S4).

Taken together, these results indicate that the conditional mutant grown in Middlebrook 7H9 in presence of ATc (causing the repression of esx-3) experienced iron and zinc starvation regardless of the high concentration of these two metals in the medium.

Expression analysis of iron- and zinc-regulated genes in ESX-3-depleted M. tuberculosis during growth in defined metals concentrations

To better characterize the transcriptional response of the ESX-3-deficient M. tuberculosis to iron and zinc availability, we examined the expression of three iron-regulated genes (mbtB, irtA, and rv1460) and 2 zinc-regulated genes (rv2059 and rpmB2) by quantitative real-time RT-PCR in bacteria grown in MM with defined concentrations of iron and zinc. Figure 2A shows the ratio of the expression levels of the 5 genes between the conditional mutant and its parental strain grown in CMM in the presence of ATc. The 5 genes were expressed at a higher level in the conditional mutant validating the microarray results and supporting the idea that esx-3 deficient M. tuberculosis experiences iron and zinc limitation despite the abundance of these metals in the medium. Moreover, we also measured the ratio of the expression levels of the same genes in the conditional mutant grown in the presence and absence of ATc. Also in this case, the genes analyzed were expressed at a higher level when esx-3 was down-regulated by ATc (Figure 2A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Expression analysis in ESX-3 depleted M. tuberculosis in defined medium.

Quantitative real-time RT-PCR performed on total RNA extracted from cells grown in CMM (50μM FeCl₃ and 3,67 μM ZnSO₄) and in LIMM (2μM FeCl₃ and 3,67μM ZnSO₄). The graphs represent the ratio between expression levels (normalized to sigA mRNA) in the conditional esx-3 mutant (TB79) and its parental strain (TB38) grown with or without ATc (+ symbol) in A) CMM and B) LIMM. Values represent the average of three independent experiments with the relative standard errors.

https://doi.org/10.1371/journal.pone.0078351.g002

The same experiment was performed in LIMM and again when esx-3 was down-regulated by ATc the expression of iron-regulated genes increased more than in the parental strain and in the conditional mutant grown without ATc (Figure 2B).

The results of the arrays and quantitative real-time RT-PCR clearly demonstrate that esx-3 deficiency results in a response that is typical of iron/zinc limitation.

ESX-3-depleted M. tuberculosis contains lower levels of intracellular iron

Streptonigrin (STN) can be used as an indirect measure of intracellular iron levels, as low intracellular concentration of this metal reduces the sensitivity to this antibiotic [53]. To further determine whether M. tuberculosis is experiencing iron limitation in Middlebrook 7H9 when esx-3 is repressed, we compared STN sensitivity of the esx-3 conditional mutant TB79 and its parental strains TB38 in the presence or absence of ATc. As shown in Figure 3 the resistance to STN of the conditional mutant increased about 100 fold in the presence of ATc, suggesting lower intracellular levels of iron in these cells following ESX-3 depletion.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Streptonigrin sensitivity of the M. tuberculosis esx-3 conditional mutant.

Cells in exponential phase were treated with 5μg/ml STN (5X MIC) for 48h. The graphs show the survival rate in cultures treated 48 h with STN in respect to untreated cultures (see methods). The values represent the average of three independent experiments with relative standard errors. Asterisks represent statistically significant differences (p<0.05) respect to the conditional mutant (TB79) grown with ATc.

https://doi.org/10.1371/journal.pone.0078351.g003

Repression of esx-3 decreases uptake of Fe⁺³-carboxymycobactin

In order to determine whether the iron deficiency phenotype exhibited by TB79 is the result of decreased iron import when esx-3 is repressed, we measured uptake of ⁵⁵Fe-CM by TB79 and its parental strain in the presence and absence of ATc. No significant difference in the amount of iron taken up by the parental strain in the presence or absence of ATc was observed (data not shown), however, repression of esx-3 by ATc drastically decreased iron incorporation in TB79 (Figure 4) supporting a role of esx-3 in iron uptake.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Fe³ carboxymycobactin (CM) uptake in ESX-3-depleted M. tuberculosis.

Incorporation of Fe⁵⁵-CM in TB79 cultured in the presence or absence of ATc was measured as described in materials and methods. The results are the average of duplicates in a representative experiment that was repeated 4 times.

https://doi.org/10.1371/journal.pone.0078351.g004

Iron dependent growth and low intracellular iron concentration of an M. smegmatis esx-3 null mutant

Since in M. smegmatis esx-3 gene cluster is regulated by IdeR and not by Zur, we hypothesized that in this specie this secretion system is only involved in iron uptake. It has been previously shown that in M. smegmatis esx-3 is essential in iron-limiting conditions when the biosynthetic pathway for exhochelin (the major siderophore in this species) is inactivated [36]. However, the involvement of ESX-3 in zinc uptake has not been investigated. To test our hypothesis, we constructed an esx-3 and fxbA (gene essential for exochelin biosynthesis) null mutants, as well as an esx-3/fxbA double mutant in M. smegmatis mc²155, and analyzed the growth of these mutants in MM with or without iron or zinc (Figure 5). The single mutants did not display any growth defect in all conditions tested, while iron limitation resulted in delayed growth of the double mutant, confirming previous observations [36] (Figure 5B). The growth of the esx-3-fxbA double mutant was not affected in medium lacking zinc (Figure 5C), indicating that in M. smegmatis ESX-3 is dispensable for growth in low zinc. Sensitivity to STN was also evaluated in these mutants grown in LIMM. All mutants showed an increased resistance to STN compared to their wild type strain: in particular, the double esx-3/fxbA mutant was the most resistant and the single esx-3 mutant was the most sensitive (Figure 5D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Growth in different metals concentrations and streptonigrin sensitivity of the M. smegmatis esx-3 null mutants.

The wild type strain mc²155 and its derivatives mutants esx-3/fxbA, esx-3, and fxbA were grown in shaking cultures in MM with different iron and zinc concentrations: (A) CMM (50μM FeCl₃ and 3,67 μM ZnSO₄); (B) LIMM (2μM FeCl₃ and 3,67μM ZnSO₄), (C) LZMM (50μM FeCl₃ and no ZnSO₄ added). Cells were pre-grown in CMM. The experiment was repeated three times. Data from one representative experiment are shown. D) Sensitivity to STN: cells in exponential phase were treated with 5μg/ml STN (5X MIC) for 30’; the graphs shows the survival rate in cultures treated 30 min with STN compared with untreated cultures (see methods). The values represent the average of three independent experiments with relative standard errors. Asterisks represent statistically significant differences (p<0.05) respect to the wild type strain.

https://doi.org/10.1371/journal.pone.0078351.g005

Growth of the ESX-3-depleted M. tuberculosis and M. smegmatis in the presence of heme as iron source

In addition to siderophores, mycobacteria have systems to acquire iron from the heme prosthetic group [54,55]. To determine whether ESX-3 is necessary for utilization of heme as iron source we grew the M. tuberculosis esx-3 conditional mutant and the M. smegmatis esx-3/fxbA null mutant in MM with hemin as the only iron source. As showed in Figure 6A, the growth of the M. tuberculosis esx-3 conditional mutant was not rescued by hemin when it was grown in the presence of ATc, while its growth improved in the presence of hemin if ATc was not present. However, hemin was able to clearly enhance the growth of the M. smegmatis esx-3/fxbA double mutant (Figure 6B). These results clearly demonstrate that M. smegmatis does not require ESX-3 to utilize hemin. The lack of rescue by hemin in M. tuberculosis suggests that ESX-3 is necessary for hemin utilization or, alternatively that deficient zinc acquisition prevented normal growth even in the presence of heme.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Growth in hemin-supplemented medium.

The strains were grown in NIMM (3,67μM ZnSO₄ and no FeCl₃) with or without 50 µM DFO and 5 or 10 µM hemin. (A) M. tuberculosis esx-3 conditional mutant (TB79) and its parental strain (TB38) grown with or without 200ng/ml of ATc. The growth of the parental strain in all tested conditions was the same with or without ATc and only growth curve of cultures with ATc are shown. (B) M. smegmatis esx-3/fxbA mutant and its parental wild type strain. Values represent one representative experiment out of three performed.

https://doi.org/10.1371/journal.pone.0078351.g006

ESX-3 depletion doesn’t affect cell wall permeability

It is known that depletion of some components of ESX-1 and ESX-5 cause defects in cell wall permeability [15,56]. This might indicate a common role of T7SS in the regulation of cell wall integrity. To determine whether ESX-3 deficiency causes changes in cell wall permeability, we measured the resistance to antibiotics with different structures as well as sensitivity to SDS in the M. smegmatis esx-3 mutant. The MICs to ampicillin, rifampicin and Km in Middlebrook 7H9, as well as sensitivity to SDS were the same in the esx-3 null mutant and in its parental strain (Figures S5 and S6). Since ESX-3 is induced in low iron we evaluated the MICs to these antibiotics also in this condition; however, even in these conditions we could not observe any difference between the two strains (data not shown). The same experiments were performed in the M. tuberculosis esx-3 conditional mutant grown with or without ATc. Even in this case we could not observe any difference in MIC or SDS sensitivity due to the absence of ESX-3 (data not shown). Thus, these results suggest that no drastic effects in cell wall permeability occur as a result of ESX-3 depletion in M. smegmatis or in M. tuberculosis.

ESX-3 depletion prevents M. tuberculosis replication in THP-1-derived macrophages

To determine whether ESX-3 is essential for intracellular replication of M. tuberculosis, we infected THP-1-derived macrophages with the esx-3 conditional mutant in the presence or absence of ATc. As shown in Figure 7 the mutant was not able to replicate in these macrophages when ATc was added to the culture medium indicating that ESX-3 is necessary for intracellular growth of M. tuberculosis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 7. Infection of THP1-derived macrophages.

THP1-derived macrophages were infected with conditional mutant (TB79) and parental strain (TB38) at an MOI of 1:20 in presence or absence of 200 ng/ml of ATc. The growth curve was measured as CFU/ml after release of intracellular bacteria in 100 µl of 0.05% SDS . The values represent the average of two independent experiments with relative standard error.

https://doi.org/10.1371/journal.pone.0078351.g007