The senX3regX3 two-component regulatory system of Mycobacterium tuberculosis is required for virulence

Tanya Parish1,2, Debbie A. Smith2, Gretta Roberts1, Joanna Betts3 and Neil G. Stoker2,4

1 Department of Medical Microbiology, Barts and the London, Queen Mary's School of Medicine and Dentistry, Turner Street, London E1 2AD, UK
2 Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
3 GlaxoSmithKline, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, UK
4 Department of Pathology and Infectious Diseases, Royal Veterinary College, Royal College Street, London NW1 0TU, UK

Correspondence
Tanya Parish
t.parish{at}qmul.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Two-component regulatory systems have been widely implicated in bacterial virulence. To investigate the role of one such system in Mycobacterium tuberculosis, a strain was constructed in which the senX3regX3 system was deleted by homologous recombination. The mutant strain (Tame15) showed a growth defect after infection of macrophages and was attenuated in both immunodeficient and immunocompetent mice. Competitive hybridization of total RNA from the wild-type and mutant strains to a whole-genome microarray was used to identify changes in gene expression resulting from the deletion. One operon was highly up-regulated in the mutant, indicating that regX3 probably has a role as a repressor of this operon. Other genes which were up- or down-regulated were also identified. Many of the genes showing down-regulation are involved in normal growth of the bacterium, indicating that the mutant strain is subject to some type of growth slow-down or stress. Genes showing differential expression were further grouped according to their pattern of gene expression under other stress conditions. From this analysis 50 genes were identified which are the most likely to be controlled by RegX3. Most of these genes are of unknown function and no obvious motifs were found upstream of the genes identified. Thus, it has been demonstrated that the senX3regX3 two-component system is involved in the virulence of M. tuberculosis and a number of genes controlled by this system have been identified.


Abbreviations: BMDM, bone-marrow-derived macrophage; 2CR, two-component regulatory system


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Bacteria sense their environment by means of two-component regulatory systems (2CRs) (Hoch, 2000; Smith et al., 2001; Urao et al., 2000). These systems enable organisms to respond to changing environmental conditions by co-ordinated gene regulation. Each system consists of a sensor which responds to a particular stimulus and a regulator which controls the expression of a set of genes. The sensory protein autophosphorylates in response to the stimulus and this phosphate is then transferred to a conserved aspartate residue in the regulatory protein. The regulator is a DNA-binding protein that acts as a transcriptional regulator and can turn sets of genes on or off. Genes that are directly controlled by the same regulator in this fashion are termed the regulon. In addition to direct control, regulatory cascades can occur where one regulator controls the expression of another thus magnifying the number of genes controlled in response to the original signal. 2CRs have been widely implicated in the virulence of pathogenic bacteria, since they can control sets of virulence genes (Dziejman & Mekalanos, 1995). 2CRs have been suggested as potential drug targets as they are not found in higher eukaryotes (Barrett & Hoch, 1998) and attenuated 2CR mutant strains have also been proposed as vaccine candidates (Garcia Vescovi et al., 1996; Groisman & Heffron, 1995). Since each 2CR can control many different genes, it is possible that inactivating one of these systems, either by way of deletion or by use of inhibitors, would be a more effective approach than targeting an individual enzyme and would be less likely to lead to the development of resistance problems.

Mycobacterium tuberculosis is a sophisticated pathogen that can persist in the human host for many years. The bacteria are exposed to many different conditions during the infection and disease process and must have well-tuned mechanisms to respond to any given environment. For example, the bacteria can be found growing both intracellularly (in macrophages) or extracellularly (in the granuloma). Since the function of 2CRs is to adapt to different external conditions, they are likely to play an important role in the ability of M. tuberculosis to sense and respond to different host environments.

The current TB vaccine (Mycobacterium bovis BCG), a live attenuated strain, is not ideal due to its variable efficacy. Several approaches are being taken to try to improve the vaccine, of which one is to rationally attenuate M. tuberculosis. A genetic knock-out in one of the 2CRs may be a valuable component of such a vaccine strain, as many relevant genes may be affected at the same time. In addition, identifying genes that are regulated by these systems may lead to the identification of new virulence factors. This will increase our knowledge of pathogenesis and may allow for the development of novel interventions.

M. tuberculosis has 11 2CRs, as identified from the genome sequence (Cole et al., 1998), and several of these have been implicated in virulence. For example, PhoP and Prr mutants are attenuated (Ewann et al., 2002; Perez et al., 2001), whereas DevR, TcrXY, TrcS and KdpDE mutants are hypervirulent (Parish et al., 2003), showing that these systems do play an important role during infection. The senX3regX3 system, originally identified by degenerate PCR, was the first reported example of an M. tuberculosis 2CR (Wren et al., 1992). The two genes are separated by a small intergenic region which contains three repeats of a MIRU (mycobacterial interspersed repeat unit), although they are very likely to be co-transcribed. Phosphorylation of the regulator by the sensor has been demonstrated and there is some evidence that the system is auto-regulated with RegX3 binding to its own promoter (Himpens et al., 2000). However, the genes that this system controls have not been identified. We have investigated the role of this system in virulence and used a mutant strain lacking a functional senX3regX3 system in order to characterize the regulon.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Strains and growth of bacteria.
Wild-type M. tuberculosis H37Rv (ATCC 25618) and the senX3regX3 deletion strain (Tame15) were grown at 37 °C in Middlebrook 7H9 medium (Difco) supplemented with 10 % (v/v) OADC (Becton Dickinson) and 0·05 % (w/v) Tween 80, or on Middlebrook 7H10 agar (Difco) supplemented with 10 % (v/v) OADC. Hygromycin was used at 100 µg ml-1, kanamycin at 20 µg ml-1, streptomycin at 20 µg ml-1, X-Gal at 50 µg ml-1 and sucrose at 2 % (w/v) where appropriate.

Construction and confirmation of the senX3 deletion strain.
The senX3regX3 deletion was constructed using previously published methods (Parish & Stoker, 2000). Plasmids used in this study are described in Table 1. A deletion delivery construct (pSOUP25) was made using the pNIL and pGOAL series vectors (maps available on request) (Parish & Stoker, 2000). The deletion constructed encompassed the 3' end of the senX3 gene, the intergenic region and the 5' end of the regX3 gene (Fig. 1). Mutants were constructed using a two-step strategy as described previously. Briefly, 1–5 µg of vector DNA was pre-treated with UV to stimulate homologous recombination and used to electroporate M. tuberculosis. Single cross-over strains were selected on agar containing hygromycin, kanamycin and X-Gal. An individual colony was streaked out onto agar (without antibiotics) to allow the second cross-over to occur. Cells were resuspended in media and serial dilutions were plated onto X-Gal and sucrose. Sucrose-resistant, white colonies were tested for kanamycin sensitivity and analysed by PCR and Southern hybridization. PCR primers regP1 (5'-GGTAATTGTTTGAGATCCCAC-3') and regP3 (5'-GTCCGCTAGCCCTCGAGTTTG-3') were used to PCR-amplify the whole operon to distinguish strains carrying the wild-type (2·3 kb) from the deletion allele (1·4 kb) (Fig. 2). The PCR product from the deletion strain was sequence-verified. Genomic DNA was prepared according to the method of Belisle & Sonnenberg (1998). Southern hybridization was carried out using the AlkPhos Direct kit (Amersham) according to the manufacturer's instructions in order to confirm the expected genotype (Fig. 2).


View this table:
[in this window]
[in a new window]
 
Table 1. Plasmids used in this study

 


View larger version (14K):
[in this window]
[in a new window]
 
Fig. 1. Construction of the senX3regX3 deletion mutant strain. (a) Arrangement of genes in the M. tuberculosis chromosome showing the regions used in the construction of pSOUP25. Open arrow, response regulator; hatched arrow, sensor; solid arrows, other genes; open box, region deleted in the mutant. Relevant restriction sites are shown. regP1 and regP3 were the PCR primers used for characterizing strains (see Fig. 2). (b) pSOUP25 delivery vector used for mutagenesis. The 4·3 kb BglII–BamHI M. tuberculosis fragment indicated in (a) was cloned into p2NIL and a 0·9 kbp deletion was subsequently made. The marker gene cassette from pGOAL19 was added to make the final delivery vector. kan, kanamycin resistance gene; hyg, hygromycin resistance gene; PAg85a–lacZ, {beta}-galactosidase driven by the mycobacterial antigen 85 A promoter; Phsp60–sacB, sucrose sensitivity gene driven by the mycobacterial Hsp60 promoter.

 


View larger version (32K):
[in this window]
[in a new window]
 
Fig. 2. Genetic analysis of the senX3regX3 mutant. (a) Southern analysis. Genomic DNA was digested with PstI and then hybridized to the senX3regX3 region probe. The wild-type (3·0 kbp) and deletion bands (2·1 kbp) are indicated. (b) PCR analysis. The PCR primers regP1 and regP3 were used to amplify the region from the wild-type and mutant strains. The expected sizes for wild-type (2·3 kbp) and deletion (1·4 kbp) bands are indicated. Lanes: 1, wild-type; 2, Tame15 deletion strain; M, lambda HindIII markers.

 
In vitro growth curve.
Strains were inoculated into twenty 50 ml capacity tubes each containing 10 ml of liquid media to a theoretical OD600 value of 0·05. Cultures were incubated standing at 37 °C and growth was monitored by measuring the OD600 value. A fresh tube was used for each time point.

Infection assays.
Viable stocks of wild-type M. tuberculosis and mutants were grown in 10 ml of liquid medium until an OD600 value of between 0·5 and 1·0 was reached. Bacteria were washed once in Dulbecco's PBS (Sigma) and resuspended in 5 ml sterile PBS. Aliquots were used fresh for the THP1 infections or stored at -70 °C until use. At the time of inoculation, serial dilutions were plated to determine the input c.f.u. value.

Macrophage infection assays
The THP1 macrophage-like human cell line and bone-marrow-derived macrophages (BMDMs) were used for in vitro assays. Macrophage viability over the assay time was typically greater than 95 %. Antibiotics were not added to the cells, since M. tuberculosis does not replicate in the medium in this assay.

THP1 infection.
THP1 cells were maintained in culture, treated with phorbol 12-myristate 13-acetate to induce differentiation, washed and then infected as described by Lukey (2001). Extracellular bacteria were removed by washing several times. Determination of the initial inoculum was assessed by plating serial dilutions, and the number of intracellular bacteria was monitored over 7 days.

BMDMs.
BMDMs from BALB/c mice were isolated and infected in the absence of antibiotics as described previously (Smith et al., 2001). Macrophage monolayers were pre-stimulated with IFN{gamma} (Gibco) at a concentration of 200 units ml-1 for 4 h prior to infection. Cells were infected for 4 h and washed six times in warm tissue culture medium to remove extracellular bacteria. The infection dose was assayed independently by plating the inoculum. The number of viable mycobacteria was assessed by lysis of the macrophage monolayer with 1 ml sterile distilled water containing 0·1 % Triton X-100 per well, followed by plating serial dilutions.

Infection of mice and tissue analysis.
Mice were infected with 1x106 viable mycobacteria in 200 µl pyrogen-free saline via a lateral tail vein. Where appropriate, infected mice were killed by cervical dislocation in accordance with humane end point protocols under the Animals Scientific Procedures Act, 1986 (UK). Median survival times were calculated for each group and statistical analysis was performed using Kaplan–Meier plots and Log Rank tests of survival. For tissue analysis, lungs, livers and spleens were collected aseptically and passed through a 100 micron pore-size sieve (Falcon) in 7H9 medium containing 0·05 % (w/v) Tween 80. Serial 10-fold dilutions were plated and c.f.u. were counted after 4 weeks. Statistical analysis was performed using Student's t-test.

Microarray analysis.
Wild-type and mutant M. tuberculosis were grown in 100 ml media in roller bottles to late-exponential phase (7 days). Cells were harvested by centrifugation and RNA was prepared according to the method of Movahedzadeh et al. (2001). Fluorescently labelled cDNA was prepared from total RNA by direct incorporation of fluorescent nucleotide analogues during a first-strand reverse transcription (RT) reaction as described previously (Betts et al., 2002). Wild-type RNA was labelled with Cy3-dCTP and mutant RNA was labelled with Cy5-dCTP, and they were compared directly by competitive hybridization. DNA microarrays used consisted of 3649 PCR-amplified ORF-specific DNA fragments, representing 93 % of the predicted 3924 M. tuberculosis H37Rv ORFs, and hybridizations were performed as described previously (Betts et al., 2002). Slides were scanned using a ScanArray 3000 instrument (GSI Lumonics) and the resulting images were analysed using GENEPIX PRO 3.0 software (Axon Instruments). RNA was isolated from three separate cultures and duplicate hybridizations were carried out for each, making a total of six hybridizations. Data from GENEPIX were analysed in GENESPRING (Silicon Genetics). Data points were excluded from the analysis if the spots were flagged as absent or marginal by GENEPIX. For each slide, the normalization was as follows: each gene's measured intensity (mutant) was divided by the control channel value (wild-type) in each sample to give the ratio of expression; when the control channel value was below 10·0 the datum point was considered bad; the 50th percentile of all measurements was used as a positive control for each sample; each measurement for each gene was divided by this synthetic positive control (assuming that this was at least 0·01); normalized values below 0 were set to 0. Genes were defined as being differentially regulated where there was a greater than twofold change in at least four of the hybridizations and where P<0·05 by Student's t-test. The t-test was conducted on the six experiments as a group. Data from published array experiments were imported into GENESPRING. Cluster analysis was then performed to group the genes depending upon their expression pattern using GENESPRING.

Promoter activity assays.
The promoter region of the senX3regX3 operon was amplified using primers regP1 and regP2 (5'-CAGCGCCGAGAACACAGTCAC-3') and cloned into pGEM-EasyT vector (Promega). The promoter region was subcloned as a blunt-ended fragment in the forward orientation into the ScaI site of the promoter-probe shuttle vector pSM128 to make plasmid pIKL-R1. The plasmid was electroporated into wild-type and Tame15 strains, and transformants were selected on streptomycin. Transformants were grown in 10 ml liquid medium to late-exponential phase before assaying for {beta}-galactosidase activity as described previously (Parish et al., 2001). Three independent transformants were each assayed in duplicate.


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Construction of the senX3regX3 mutant, Tame15
We have previously examined the role of 2CRs in the virulence of M. tuberculosis by constructing several deletion strains. Of five deletion strains studied, four (devR{Delta}, tcrXY{Delta}, trcS{Delta} and kdpDE{Delta}) were found to be significantly more virulent in a mouse model of infection (Parish et al., 2003). We extended these studies to look at another system (senX3regX3) in more detail. We constructed a mutant strain of M. tuberculosis containing a deletion of the senX3regX3 system by homologous recombination. A strain containing a 0·9 kbp deletion encompassing the 3' end of senX3, the intergenic region and the majority of regX3 was constructed using a two-step homologous recombination method (Parish & Stoker, 2000). The deletion is shown in Fig. 1. The replacement of the wild-type gene by the deleted version was confirmed by both PCR and Southern analysis (Fig. 2). Motif analysis shows that the sensor protein has a histidine kinase domain located near the centre of the protein and a phospho-acceptor domain at the C-terminal end. The regulator has a response regulator phospho-receiver domain in the N-terminal region and the DNA-binding effector domain in the C-terminal end. The deletion we constructed completely removes the phospho-receiver domain and the majority of the DNA-binding domain from RegX3 whilst both the sensor protein domains are left intact.

Growth characteristics
We first looked at the growth characteristics of the mutant in axenic culture (Fig. 3). The mutant behaved erratically, in that on some occasions growth resembled the wild-type strain (Fig. 3a) whilst on other occasions there seemed to be a clear defect (Fig. 3b). Small differences in the inoculum with respect to growth phase, number of bacteria or other factors may be responsible for the inconsistent pattern of growth. Since M. tuberculosis is such a slow-growing organism, very small differences in the initial inoculum could be magnified during prolonged growth. The failure of the mutant to grow to the same optical density value as the wild-type was noted on numerous occasions but was not perfectly reproducible. Thus, we concluded that the mutant had a subtle growth defect which we could not quantify precisely.



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 3. Growth characteristics of Tame15 in axenic culture. Two growth curves are shown for the mutant and deletion strains representing the two different types of growth observed. (a) Normal growth; (b) restricted growth. {circ}, Wild-type; {bullet}, Tame15.

 
Macrophage assays
To determine whether the mutant was attenuated in terms of its ability to survive and grow intracellularly, we looked at growth within two different cell types. The growth of the mutant was assayed in the macrophage-like THP1 cell line (Fig. 4a) over several days. As can be seen in Fig. 4(a), the mutant showed a high degree of attenuation at both low and high m.o.i. In the low m.o.i. infection, the mutant was killed by the macrophages and completely cleared by day 7, whereas the wild-type remained viable and increased in number. In the high m.o.i. infection, the mutant survived and replicated but not to the same extent as the wild-type. We then looked at growth within IFN{gamma}-primed murine BMDMs over 3 days (Fig. 4b). The mutant showed attenuation in this system as well. There was a significantly higher rate of killing, which was apparent at the earliest time point post-infection (24 h, P<0·0007). Thus, the mutant showed a clearly reduced ability to survive within a macrophage cell line and activated primary BMDMs.



View larger version (15K):
[in this window]
[in a new window]
 
Fig. 4. Intracellular survival. Macrophages were infected at the m.o.i. indicated and the number of intracellular viable bacteria was measured over several days. (a) THP1 macrophage-like cell line. (b) IFN{gamma}-treated murine macrophages. Results are given as the mean±SE of triplicate wells. {blacksquare}, Wild-type; {blacktriangleup}, Tame15.

 
Virulence in two mouse models
Since the mutant showed attenuation in both in vitro models used, we used two different mouse models of infection to establish whether the mutation in the senX3regX3 system also altered the virulence of M. tuberculosis in vivo. To look at infection in the absence of acquired immunity, SCID mice were infected intravenously and monitored for survival (Fig. 5). Infection with wild-type M. tuberculosis H37Rv led to death with a median survival time (MST) of 40·5 days. The mutant strain was significantly less virulent with a MST of 47 days (P<0·0005). We then looked at the kinetics of bacterial growth in the immunocompetent host using DBA/2 mice. Bacterial loads in the organs were measured on days 15, 30 and 59 (Fig. 6). The bacterial loads in the livers of mice infected with the mutant were significantly lower on days 15 and 30, and by day 59 a reduction in numbers was also apparent in the lungs. Thus, the mutant had mildly reduced virulence in immunocompetent mice in the early stages of infection.



View larger version (10K):
[in this window]
[in a new window]
 
Fig. 5. Virulence of Tame15 in SCID mice. Survival of SCID mice after infection with 1x106 bacteria. Each group contained six mice and results are representative of two separate experiments. {blacksquare}, Wild-type; {blacktriangleup}, Tame15; {circ}, control (PBS).

 


View larger version (13K):
[in this window]
[in a new window]
 
Fig. 6. Virulence of Tame15 in DBA mice. Mice were infected intravenously with 1x106 wild-type or mutant bacteria and organ loads were measured. The results represent means±SE for three mice per group with significance measured using Student's t-test. An asterisk indicates where P<0·05. Solid bars, wild-type; open bars, Tame15. (a) Lung; (b) liver; (c) spleen.

 
A regX3 transposon mutant of strain Mt103 has previously been shown not to be attenuated using a low dose (103 bacteria) aerosol infection model of C57BL/6J mice, as assessed by bacterial loads in the lungs over 40 days (Ewann et al., 2002). The difference between these results and ours may be due to differences in the model, i.e. the route of infection, infection dose and mouse strain. For example, we used DBA mice which are more susceptible to infection than C57BL/6J mice. Alternatively, it may reflect a real difference in the mutants themselves. The transposon mutant is an insertion into the 5' end of regX3 and may not completely abrogate RegX3 function, whereas our mutant has the majority of the gene and both domains deleted. Tame15 also has the 3' end of senX3 deleted and although it seems unlikely that this would abrogate SenX3 function, it cannot be ruled out. There are precedents for a sensory protein interacting with more than one regulator and if this were the case then our mutant may have a different phenotype due to the additional absence of SenX3 function. Alternatively, it may be the fact that the transposon mutant was constructed in a different genetic background (strain Mt103 as opposed to H37Rv). Tame15 was attenuated in all four systems tested, so we are confident that this is a real phenotype. The in vivo results are also consistent with the defect in axenic growth.

The reason for the attenuation of Tame15 is not yet known. It is possible that the attenuation may result from a general decrease in the growth rate; however, this is true of many strains, e.g. auxotrophic strains are attenuated because their growth is severely restricted by amino acid availability (Smith et al., 2001). Nevertheless, Tame15 is an attenuated strain, whatever the root cause. The difference seen in the SCID mouse model was not due to small variations in inoculum size, since the experiment was conducted twice with similar results (inocula sizes 1x106 vs 2x106 and 4x106 vs 2x106, respectively).

Promoter activity
To look at promoter activity for the senX3regX3 system in the wild-type and mutant strain we cloned the upstream promoter region into the integrating vector pSM128 (Dussurget et al., 1999) (Fig. 7). Using the {beta}-galactosidase reporter gene, we were able to assess promoter activity under growth in aerobic cultures. Promoter activity was low in both strains, but unexpectedly the promoter activity was 2·8-fold increased in the mutant strain (wild-type, 2±0·4 units; Tame15, 5·6±0·7 units; P<0·00001).



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 7. Promoter activity analysis. (a) The upstream promoter region of the senX3regX3 operon was amplified using the primers regP1 and regP2 and cloned into the shuttle vector pSM128 upstream of the lacZ reporter gene making pIKL-R1 (b). pIKL-R1 was transformed into wild-type and Tame15 strains.

 
There is some indirect evidence for auto-regulation of RegX3. Himpens et al. (2000) showed that there was twofold higher promoter activity of the M. tuberculosis promoter in Mycobacterium smegmatis when the strain also carried the M. tuberculosis senX3regX3 operon. In addition, they showed that RegX3 binds to its own promoter in the absence of phosphorylation. In contrast, we see that promoter activity is actually higher in the mutant strain. We can explain this apparent discrepancy if RegX3 actually represses promoter activity in the unphosphorylated state by physically blocking RNA polymerase access to the promoter. In the phosphorylated state, it would be expected to undergo a conformational change which would lead to promoter activation (possibly by recruiting RNA polymerase). In our mutant in the absence of RegX3, promoter activity would be slightly higher as there is no unphosphorylated RegX3 bound and therefore no steric hindrance to RNA polymerase. Alternatively, the auto-regulation seen in M. smegmatis may not occur in M. tuberculosis.

Analysis of global gene expression
Two-component systems generally function as global regulators of gene expression in response to environmental conditions. Each system controls a set of genes, termed the regulon, in response to a particular signal. To identify potential members of the senX3regX3 regulon, we compared global gene expression in the mutant with that of the wild-type strain. The reporter assays confirmed that the system is expressed under aerobically grown conditions, so we would expect to see changes in expression of genes controlled by this 2CR under these conditions. Therefore, RNA was prepared from both strains grown in roller bottles (aerobic) and competitively hybridized to a whole-genome microarray. Genes showing a significant difference in expression in the mutant were identified and are given in Table 2. As expected from the deletion constructed, regX3 transcripts were significantly reduced; however, the senX3 transcript was unchanged (expression ratio not significantly different from 1). Thus, the change in senX3 promoter activity observed in the reporter gene was not reflected in the array data. This may be because of the low absolute activity of the promoter, so that the twofold increase was below the limit of detection of the microarray. We confirmed the absence of regX3 mRNA by quantitative RT-PCR and found that there was 100-fold less regX3 mRNA in the mutant strain (data not shown).


View this table:
[in this window]
[in a new window]
 
Table 2. Whole-genome microarray analysis of gene expression in Tame15

Genes that were identified as being differentially expressed in the senX3regX3 mutant are listed. The Rv numbers from Cole et al. (1998) are given for reference, together with any gene designation, the functional classification assigned, the mean fold change in expression, the P value and the predicted function.

 
In total, 30 genes were up-regulated and 68 genes were down-regulated. From these results it is not possible to determine which effects are direct and which are indirect. However, in yeast, a survey of 106 regulatory proteins showed that they interacted with between 0 and 180 promoter regions with a mean of 38, so our tally of genes is well within the expected range (Lee et al., 2002).

Differential gene expression
At this stage, it is not possible to determine which genes are directly controlled at the transcriptional level by RegX3 and which genes may be indirectly controlled, for example, via other regulators. However, we can say that all of the genes whose expression changes must rely on the senX3regX3 system in some way for normal expression. If RegX3 directly controls the transcription of these genes, then it must be a negative regulator of up-regulated genes, i.e. it represses the expression of such genes. In contrast, for those genes that are down-regulated in the mutant, RegX3 would act as a positive regulator (inducer). More genes were expressed at a lower level in the mutant than were de-repressed, indicating that a larger number of genes rely on the senX3regX3 system for normal expression. The possibility of regulatory cascades is raised by the array data since there are four potential transcriptional regulators whose expression changes, two going up in the mutant (Rv1990c and Rv2669) and two going down (Rv2488c and Rv2308). In addition, two other up-regulated genes show some similarity to anti-anti-sigma factors (Rv2638, Rv0516c). In yeast, several different types of networks of regulatory proteins have been identified (Lee et al., 2002) and it is likely that similar networks exist in bacteria.

Down-regulated genes
Most genes that were significantly down-regulated decreased by a factor of two- to threefold. More genes appear to be repressed than induced in the mutant and these fall into several categories. Many of the genes are involved in basic macromolecule biosynthesis, particularly DNA and RNA synthesis. We have previously seen that deletion of a 2CR can have an indirect effect on gene expression. With TrcS the cells appeared to be stressed and several genes were differentially expressed in response to the stress rather than as a direct result of the regulator deletion (Wernisch et al., 2003). The data for this strain look as if a similar stress response is occurring.

Several ribosomal proteins are down-regulated as are genes involved in DNA replication, repair and recombination (dnaB, ruvC, dnaQ, Rv3644c) as well as insertion elements. This would seem to indicate that the mutant would have a slower growth rate than the wild-type since it is less capable of synthesizing new DNA, RNA and protein. This is consistent with the growth phenotype observed earlier.

A number of genes involved in fatty acid degradation are repressed (fadE6, accD2, fadE14, fadE23) as are several probable oxidoreductases and dehydrogenases (Rv0183, Rv1714, Rv1812c). There are also a number of genes involved in cell-wall biosynthesis, including lipid biosynthesis (acpM, desA3, fbpC2) and the cell envelope (membrane and exported proteins; lprE, Rv0867c, Rv1433, Rv1457c). The alkyl-hydroperoxidases ahpC and ahpD are both down-regulated, whereas superoxide dismutase (sodA) is up-regulated, suggesting that there is a change in the type of oxidative stress the cells are facing internally, rather than a general increase in stress. Several members of the PE/PPE family are down-regulated as are a large number of conserved hypothetical and unknown proteins.

Up-regulated genes
Most genes that were significantly up-regulated increased by a factor of two- to threefold. However, the expression of one operon (Rv0096 to Rv0101) was highly elevated. The level of induction in the mutant was five- to 17-fold over that in the wild-type (Fig. 8). Since the Rv0096 operon was so highly induced, it seems likely that this operon is controlled directly by the RegX3 regulator rather than via an indirect effect, due to, for example, slowed growth rate or increased intracellular stress levels. The operon looks to consist of several genes comprising at least Rv0096 to Rv0100 and possibly Rv0101 and Rv0102 as well (Fig. 8). Four of these genes are significantly up-regulated in the mutant. Rv0096 encodes a member of the PE/PPE family whose function is unclear, although these proteins have been proposed to be involved in antigenic variation (Banu et al., 2002), and other members of this family have been shown to be up-regulated in the frog model of mycobacterial-induced granuloma formation (Ramakrishnan et al., 2000), suggesting a role in pathogenicity. Rv0097 encodes a possible oxidoreductase; Rv0098 and Rv0099 both encode conserved hypothetical proteins of unknown function. FadD10 is one of a large number of fatty acid CoA-ligases proposed to be involved in lipid metabolism. The nrp gene encodes a non-ribosomal peptide synthase whose biological role is yet to be firmly elucidated, but it has been proposed to be involved in lipid metabolism due to its location in this operon (Cole et al., 1998).



View larger version (13K):
[in this window]
[in a new window]
 
Fig. 8. Operon structure of differentially expressed genes. The chromosomal arrangement of the Rv0096 proposed operon is shown. Values below the genes show the mean fold increase in gene expression in the mutant strain and the P value from Student's t-test.

 
The trxA and sodA genes are both up-regulated in the mutant. The enzymes these genes encode play a role in maintaining a suitable intracellular environment: thioredoxin participates in many redox reactions and maintains the redox potential of the cell, whilst superoxide dismutase is involved in the removal of free radicals generated during normal metabolism. The increase in these enzymes may indicate that the cell is under more stress than normal. Two genes associated with low oxygen environments are also induced, cydB encoding the cytochrome d ubiquinol oxidase subunit II and ald encoding L-alanine dehydrogenase. CydB has been proposed as the terminal oxidase complex which is used during low oxygen growth. Ald is well known to be induced in the Wayne model of hypoxic growth (Wayne & Sohaskey, 2001), under anaerobic conditions, and in stationary phase (Feng et al., 2002) and may also be important in cell-wall biosynthesis where L-alanine is required for the peptidoglycan.

Of the other genes that show increased expression, citrate synthase 3 is involved in energy metabolism (tricarboxylic acid cycle) at an important control point of the cycle, Rv0103c encodes a probable copper cation transporter, three others are insertion sequence or phage elements (Rv2424c, Rv2647, Rv3750c), one is a bacteriocin-like protein (Rv3660c) and the remainder encode conserved hypotheticals or unknowns.

Taken together, these differences seem to suggest a subtle change in the normal metabolism (and growth) of the mutant bacteria and also a change in the type of intracellular stress that the bacteria are facing. A secondary effect of the mutation may be to reduce the growth rate, but it is not clear why.

Comparison of expression data
Several genes that were down-regulated appeared to be involved in normal growth; for example, several ribosomal proteins and dnaB. Several genes identified as stress proteins were also differentially expressed, either down-regulated (grpE, ahpC and ahpD) or up-regulated (sodA). This fits in with our previous observation that the mutant had a slight growth defect and indicated that the deletion of the system is probably causing some type of stress within the cells. It is possible that these genes are not directly controlled by RegX3 itself, but are down-regulated as a secondary effect of the mutation. To refine the list of potential regulon members, we conducted a meta-analysis of previously published data to determine if there were any significant patterns of expression relating to stress.

By comparing genes that were differentially expressed under various stress conditions, we looked for patterns of expression which would indicate that certain groups of genes are being co-ordinately regulated in response to any type of stress. We looked at their expression patterns in other published array data representing several different types of stress: heat shock (Stewart et al., 2002), carbon limitation (Betts et al., 2002), SDS treatment (Manganelli et al., 2001), diamide treatment (Manganelli et al., 2002), low oxygen tension (Sherman et al., 2001) and low iron (Rodriguez et al., 2002). Cluster analysis was used to group genes with similar patterns of expression. Genes which showed the same pattern of expression in more than one stress condition were in the same cluster and we considered them to be part of a general stress response. This made them unlikely to be directly controlled by RegX3, so we excluded them from our list. The remaining genes fell into four clusters.

Groups 1 and 2 were the up-regulated genes and groups 3 and 4 were the down-regulated genes. Two groups represented genes that only changed in the senX3regX3 mutant (1 and 3). The only condition in which senX3 or regX3 had been seen to change significantly was carbon starvation, where senX3 was down-regulated by 2·2-fold, and one group (2) contained the genes that went up in both conditions. Group 4 contained those genes that were down-regulated in the mutant, but up-regulated under the other stress conditions (indicating that the change in expression was not due to stress alone). Of the 98 genes originally identified, we narrowed our list down to 50 (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 3. Cluster analysis of differentially regulated genes

Genes were grouped according to their behaviour under several different conditions. Group 1, up-regulated in Tame15 only; group 2, up-regulated in Tame15 and under carbon starvation; group 3, down-regulated in Tame15 only; group 4, down-regulated in Tame15, up-regulated in at least one other stress condition.

 
Thus, we predict that the genes in Table 3 are the most likely members of the senX3regX3 regulon. The majority of these genes are of unknown function, so at this stage it is difficult to speculate about the likely stimulus for this system.

Motif analysis
We further analysed the genes from Table 3 to see if there were any common motifs present in the regions immediately upstream of the genes that might be good candidates for a DNA-binding region for RegX3. We looked at a subset of those genes which showed the greatest-fold difference in expression. Sequences were analysed for the presence of tandem and inverted repeats and multiple alignments were carried out, but no significant patterns emerged. The sequences did not show any significant similarities with the previously identified binding site in the RegX3 promoter region (Himpens et al., 2000). It may be that there are different DNA-binding recognition sites for RegX3 in alternative conformational states. Alternatively, it may be that these genes are controlled by an indirect effect and we would not expect to see RegX3 binding in that case.

Identification of the stimulus
To identify the stimulus to which the senX3regX3 system responds, we tried two approaches. First, we assayed the ability of the mutant to survive different in vitro conditions and stresses. We looked at viability during exposure to extremes of pH (2 and 12), ability to survive extended stationary phase in standing culture and ability to withstand complete nutrient starvation. The mutant showed no significant difference compared to the wild-type strain (data not shown). To survey a larger number of conditions, we then looked at promoter activity from pIKL-R1 under a variety of conditions to determine if it was induced. Conditions tested included several antibiotics (kanamycin, tetracycline, gentamicin, isoniazid, ampicillin, rifampicin), lysozyme, HCl, NaOH, DMSO, SDS and H2O2, but no conditions were found to up-regulate promoter activity. Thus, the stimulus for this system stills remains unknown.

Conclusion
We have constructed a strain with a deletion of the senX3regX3 2CR. We used several different models to determine if the deletion of the senX3regX3 system had any effect on the virulence of the bacterium. The mutant showed significant attenuation in both activated and resting macrophages and in immunocompromised and immunocompetent mice. Thus, it seems that the mutant is less able to cause disease regardless of the involvement of the immune system. This attenuation was not as large as that seen previously for other mutants (e.g. the complete attenuation of the trpD mutant; Smith et al., 2001), but it was a significant reduction. We used whole-genome microarrays to identify genes that are differentially expressed in the mutant, and based on these data we have identified 50 potential members of the senX3regX3 regulon. The construction of further mutants in these genes should lead to the identification of the genes whose roles are required during infection.


   ACKNOWLEDGEMENTS
 
This work was funded by the GlaxoSmithKline Action TB Programme. We thank Ruth McAdam and Ken Duncan for useful comments and Heidi Alderton for technical assistance.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Banu, S., Honore, N., Saint-Joanis, B., Philpott, D., Prevost, M. C. & Cole, S. T. (2002). Are the PE-PGRS proteins of Mycobacterium tuberculosis variable surface antigens? Mol Microbiol 44, 9–19.[CrossRef][Medline]

Barrett, J. F. & Hoch, J. A. (1998). Two-component signal transduction as a target for microbial anti-infective therapy. Antimicrob Agents Chemother 42, 1529–1536.[Free Full Text]

Belisle, J. T. & Sonnenberg, M. G. (1998). Isolation of genomic DNA from mycobacteria. In Mycobacteria Protocols, pp. 31–44. Edited by T. Parish & N. G. Stoker. Totowa, NJ: Humana Press.

Betts, J. C., Lukey, P. T., Robb, L. C., McAdam, R. A. & Duncan, K. (2002). Evaluation of a nutrient starvation model of Mycobacterium tuberculosis persistence by gene and protein expression profiling. Mol Microbiol 43, 717–731.[CrossRef][Medline]

Cole, S. T., Brosch, R., Parkhill, J. & 39 other authors (1998). Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544.[CrossRef][Medline]

Dussurget, O., Timm, J., Gomez, M., Gold, B., Yu, S. W., Sabol, S. Z., Holmes, R. K., Jacobs, W. R. & Smith, I. (1999). Transcriptional control of the iron-responsive fxbA gene by the mycobacterial regulator IdeR. J Bacteriol 181, 3402–3408.[Abstract/Free Full Text]

Dziejman, M. & Mekalanos, J. J. (1995). Two-component signal transduction and its role in the expression of bacterial virulence factors. In Two-component Signal Transduction, pp. 305–317. Edited by J. A. Hoch & T. J. Silhavy. Washington, DC: American Society for Microbiology.

Ewann, F., Jackson, M., Pethe, K., Cooper, A., Mielcarek, N., Ensergueix, D., Gicquel, B., Locht, C. & Supply, P. (2002). Transient requirement of the PrrA–PrrB two-component system for early intracellular multiplication of Mycobacterium tuberculosis. Infect Immun 70, 2256–2263.[Abstract/Free Full Text]

Feng, Z. Y., Caceres, N. E., Sarath, G. & Barletta, R. G. (2002). Mycobacterium smegmatis L-alanine dehydrogenase (Ald) is required for proficient utilization of alanine as a sole nitrogen source and sustained anaerobic growth. J Bacteriol 184, 5001–5010.[Abstract/Free Full Text]

Garcia Vescovi, E., Soncini, F. C. & Groisman, E. A. (1996). Mg2+ as an extracellular signal: environmental regulation of Salmonella virulence. Cell 84, 165–174.[Medline]

Groisman, E. A. & Heffron, F. (1995). Regulation of Salmonella virulence by two-component regulatory systems. In Two-component Signal Transduction, pp. 319–332. Edited by J. A. Hoch & T. J. Silhavy. Washington, DC: American Society for Microbiology.

Himpens, S., Locht, C. & Supply, P. (2000). Molecular characterization of the mycobacterial SenX3–RegX3 two-component system: evidence for autoregulation. Microbiology 146, 3091–3098.[Abstract/Free Full Text]

Hoch, J. A. (2000). Two-component and phosphorelay signal transduction. Curr Opin Microbiol 3, 165–170.[CrossRef][Medline]

Lee, T. I., Rinaldi, N. J., Robert, F. & 18 other authors (2002). Transcriptional regulatory networks in Saccharomyces cerevisiae. Science 298, 799–804.[Abstract/Free Full Text]

Lukey, P. T. (2001). Macrophage virulence assays. In Mycobacterium tuberculosis Protocols, pp. 271–280. Edited by T. Parish & N. G. Stoker. Totowa, NJ: Humana Press.

Manganelli, R., Voskuil, M. I., Schoolnik, G. K. & Smith, I. (2001). The Mycobacterium tuberculosis ECF sigma factor sigma(E): role in global gene expression and survival in macrophages. Mol Microbiol 41, 423–437.[CrossRef][Medline]

Manganelli, R., Voskuil, M. I., Schoolnik, G. K., Dubnau, E., Gomez, M. & Smith, I. (2002). Role of the extracytoplasmic-function sigma factor sigma(H) in Mycobacterium tuberculosis global gene expression. Mol Microbiol 45, 365–374.[CrossRef][Medline]

Movahedzadeh, F., Gonzalez-Y-Merchand, J. A. & Cox, R. A. (2001). Transcription start site mapping. In Mycobacterium tuberculosis Protocols, pp. 105–124. Edited by T. Parish & N. G. Stoker. Totowa, NJ: Humana Press.

Parish, T. & Stoker, N. G. (2000). Use of a flexible cassette method to generate a double unmarked Mycobacterium tuberculosis tlyA plcABC mutant by gene replacement. Microbiology 146, 1969–1975.[Abstract/Free Full Text]

Parish, T., Turner, J. & Stoker, N. G. (2001). amiA is a negative regulator of acetamidase expression in Mycobacterium smegmatis. BMC Microbiol 1, 19.[CrossRef][Medline]

Parish, T., Smith, D. A., Kendall, S., Casali, N., Bancroft, G. J. & Stoker, N. G. (2003). Deletion of two-component regulatory systems increases virulence of Mycobacterium tuberculosis. Infect Immun 71, 1134–1140.[Abstract/Free Full Text]

Perez, E., Samper, S., Bordas, Y., Guilhot, C., Gicquel, B. & Martin, C. (2001). An essential role for phoP in Mycobacterium tuberculosis virulence. Mol Microbiol 41, 179–187.[CrossRef][Medline]

Ramakrishnan, L., Federspiel, N. A. & Falkow, S. (2000). Granuloma-specific expression of Mycobacterium virulence proteins from the glycine-rich PE-PGRS family. Science 288, 1436–1439.[Abstract/Free Full Text]

Rodriguez, G. M., Voskuil, M. I., Gold, B., Schoolnik, G. K. & Smith, I. (2002). ideR, an essential gene in Mycobacterium tuberculosis: role of IdeR in iron-dependent gene expression, iron metabolism, and oxidative stress response. Infect Immun 70, 3371–3381.[Abstract/Free Full Text]

Sherman, D. R., Voskuil, M., Schnappinger, D., Liao, R. L., Harrell, M. I. & Schoolnik, G. K. (2001). Regulation of the Mycobacterium tuberculosis hypoxic response gene encoding alpha-crystallin. Proc Natl Acad Sci U S A 98, 7534–7539.[Abstract/Free Full Text]

Smith, D. A., Parish, T., Stoker, N. G. & Bancroft, G. J. (2001). Characterization of auxotrophic mutants of Mycobacterium tuberculosis and their potential as vaccine candidates. Infect Immun 69, 1142–1150.[Abstract/Free Full Text]

Stewart, G. R., Wernisch, L., Stabler, R., Mangan, J. A., Hinds, J., Laing, K. G., Young, D. B. & Butcher, P. D. (2002). Dissection of the heat-shock response in Mycobacterium tuberculosis using mutants and microarrays. Microbiology 148, 3129–3138.[Abstract/Free Full Text]

Urao, T., Yamaguchi-Shinozaki, K. & Shinozaki, K. (2000). Two-component systems in plant signal transduction. Trends Plant Sci 5, 67–74.[CrossRef][Medline]

Wayne, L. G. & Sohaskey, C. D. (2001). Nonreplicating persistence of Mycobacterium tuberculosis. Annu Rev Microbiol 55, 139–163.[CrossRef][Medline]

Wernisch, L., Soneji, S., Wietzorrek, A., Parish, T., Hinds, J., Butcher, P. D. & Stoker, N. G. (2003). Analysis of whole-genome microarray replicates using mixed models. Bioinformatics 19, 53–61.[Abstract/Free Full Text]

Wren, B. W., Colby, S. M., Cubberley, R. R. & Pallen, M. J. (1992). Degenerate PCR primers for the amplification of fragments from genes encoding response regulators from a range of pathogenic bacteria. FEMS Microbiol Lett 99, 287–292.

Received 17 January 2003; revised 6 February 2003; accepted 20 February 2003.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Parish, T.
Articles by Stoker, N. G.
Articles citing this Article
PubMed
PubMed Citation
Articles by Parish, T.
Articles by Stoker, N. G.
Agricola
Articles by Parish, T.
Articles by Stoker, N. G.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2003 Society for General Microbiology.