Characterization of a Mycobacterium tuberculosis H37Rv transposon library reveals insertions in 351 ORFs and mutants with altered virulenceb

Ruth A. McAdam1, Selwyn Quana,1, Debbie A. Smith2, Stoyan Bardarov3, Joanna C. Betts1, Fiona C. Cook1, Elizabeth U. Hooker1, Alan P. Lewis1, Peter Woollard1, Martin J. Everett1, Pauline T. Lukey1, Gregory J. Bancroft2, William R. Jacobs, Jr3 and Ken Duncan1

GlaxoSmithKline, Medicines Research Centre, Gunnels Wood Road, Stevenage SG1 2NY, UK1
London School of Hygiene & Tropical Medicine, Keppel St, London WC1E 7HT, UK2
Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA3

Author for correspondence: Ken Duncan. Tel: +44 1438 763841. Fax: +44 1438 764799. e-mail: kd9430{at}gsk.com


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
A library of Mycobacterium tuberculosis insertional mutants was generated with the transposon Tn5370. The junction sequence between the transposon and the mycobacterial chromosome was determined, revealing the positions of 1329 unique insertions, 1189 of which were located in 351 different ORFs. Transposition was not completely random and examination of the most susceptible genome regions revealed a lower-than-average G+C content ranging from 54 to 62 mol%. Mutants were obtained in all of the recognized M. tuberculosis functional protein-coding gene classes. About 30% of the disrupted ORFs had matches elsewhere in the genome that suggested redundancy of function. The effect of gene disruption on the virulence of a selected set of defined mutants was investigated in a severe combined immune deficiency (SCID) mouse model. A range of phenotypes was observed in these mutants, the most notable being the severe attenuation in virulence of a strain disrupted in the Rv1290c gene, which encodes a protein of unknown function. The library described in this study provides a resource of defined mutant strains for use in functional analyses aimed at investigating the role of particular M. tuberculosis genes in virulence and defining their potential as targets for new anti-mycobacterial drugs or as candidates for deletion in a rationally attenuated live vaccine.

Keywords: gene disruption, severe combined immune deficiency (SCID) mouse, attenuation

Abbreviations: BCG, Bacille Calmette–Guérin; SCID, severe combined immune deficiency

b The precise locations of all of the insertions examined in this study can be found as supplementary data in Microbiology Online (http://mic.sgmjournals.org).

a Present address: Department of Molecular Biology and Microbiology, Tufts University, Boston, MA 02111, USA.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
There is an urgent need for new drugs to improve tuberculosis therapy and for a better vaccine for prophylaxis (O’Brien & Nunn, 2001 ; Bishai & Mercer, 2001 ). The complete genome sequence of the pathogen Mycobacterium tuberculosis (Cole et al., 1998 ) reveals the primary sequence of every potential drug and vaccine target within this organism. The challenge now is to identify those targets that are essential for survival of the organism under the conditions encountered in the host, or which contribute to the organism’s virulence. Although the genome sequence of M. tuberculosis has been of great value in understanding more about the bacterium, few virulence mechanisms comparable with other pathogens have been found. About 40% of the predicted ORFs in the genome of M. tuberculosis are designated ‘unknowns’, yet many may represent excellent targets for intervention.

A novel therapeutic agent with the potential to reduce the duration of chemotherapy from the current minimum of six months to a more acceptable period would make a very significant impact on treatment outcomes in the clinic (O’Brien & Nunn, 2001 ). However, this strategy relies upon killing a poorly understood population of bacteria that persists for long periods of time despite the presence of highly active drugs. The enzyme isocitrate lyase has been shown to be required for the survival of M. tuberculosis in the persistence phase in a murine model of infection (McKinney et al., 2000 ). Identification of more targets within M. tuberculosis with similar properties to isocitrate lyase would increase the likelihood of new drugs against this organism being found.

The only available M. tuberculosis vaccine, the Bacille Calmette–Guérin (BCG), is an attenuated strain of the M. tuberculosis-related species Mycobacterium bovis. Its protective efficacy varies widely (from 0 to 80%) in different parts of the world (Fine, 1995 ) and is thought to be waning with time (Behr & Small, 1997 ). A potential route to generating a new vaccine candidate is rational attenuation of M. tuberculosis. This approach requires the identification and characterization of the genes that contribute to the virulence of M. tuberculosis.

Random mutagenesis using mobile genetic elements such as transposons has been a highly successful approach for identifying virulence factors in many bacterial pathogens (Manoil & Beckwith, 1985 ; Bowe et al., 1998 ). The transposon inserts into host DNA thus inactivating individual genes and operons. Bacterial strains with mutations in their metabolic pathways can then be identified using selective medium in vitro, and genes which are not essential for growth in vitro can be screened for a role in vivo using virulence assays. The location of the transposon in the chromosome of a mutant may be determined by sequencing the junction of the transposon and chromosomal DNA. After saturating mutagenesis, essential genes or pathways can, in theory, be identified as those without transposon insertions by using genomic footprinting techniques (Akerley et al., 1998 ).

In mycobacteria, transposons Tn5367 and Tn5368 (McAdam et al., 1995 ), and Tn5370 (Cox et al., 1999 ) derived from IS1096 (Cirillo et al., 1991 ; McAdam et al., 1995 ) can be efficiently delivered to the bacteria on a temperature-sensitive mycobacteriophage (Bardarov et al., 1997 ) or by using a temperature-sensitive, counter-selectable plasmid (Pelicic et al., 1997 ). Transposons derived from Mariner can also be used to generate mutants in mycobacteria (Rubin et al., 1999 ; Sassetti et al., 2001 ). The studies detailed above have been successful in isolating mycobacterial strains with mutations in their metabolic pathways.

To identify virulence factors in M. tuberculosis, the technique of signature-tagged mutagenesis (Hensel et al., 1995 ) has been applied (Camacho et al., 1999 ; Cox et al., 1999 ). Pools of mutant bacteria generated using a number of transposons, each having different sequence tags, were inoculated into mice and the clearance of individual mutants was monitored by hybridization of the surviving pool to the tags; the identity of mutants that did not grow in the mice was subsequently determined. The advantage of this approach is the direct selection for the virulence phenotype in an animal model, but its success relies on the quality of the mutants in the input pool, the nature of which is not known in advance. Two groups working independently of each other used signature-tagged mutagenesis to screen mutant libraries generated by transposons Tn5367 and Tn5370 in M. tuberculosis; both groups identified mutants in a limited set of genes involved in lipid metabolism (Camacho et al., 1999 ; Cox et al., 1999 ). Therefore, it is possible that the number of virulence factors that can be detected in M. tuberculosis by this method is low or, alternatively, the libraries created by IS1096-derived transposons are limited in their degree of complexity.

We have taken a different approach to using transposon mutagenesis to identify genes essential for the virulence of M. tuberculosis. We generated a library of mutant M. tuberculosis strains by delivery of transposon Tn5370 in pJSC84 to the organism using the temperature-sensitive bacteriophage phAE87 (Cox et al., 1999 ) and subsequently identified the transposon position in 1474 of the mutants. This approach revealed the precise identity of each mutant and provided a choice of candidates to screen in vivo. We found that Tn5370 did not insert into the genome of M. tuberculosis in a completely random fashion, which limited the number of ORFs mutated. A pilot experiment to test the virulence of a random selection of the mutants in severe combined immune deficiency (SCID) mice identified mutants with either significantly increased or decreased virulence, and singled out one ORF of unknown function that contributed strongly to the loss of virulence of M. tuberculosis in the SCID mouse.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial culture.
M. tuberculosis ‘wild-type’ strain H37Rv (ATCC 25618) was grown at 37 °C in Middlebrook 7H9 medium supplemented with 10% (v/v) albumin/dextrose/catalase (ADC) enrichment and 0·05% (v/v) Tween 80 (Jacobs et al., 1991 ). After phage transduction, mutants were selected on Middlebrook 7H10 medium plates containing 0·5% glycerol, 10% (v/v) ADC and 50 µg hygromycin ml-1. For in vivo experiments, mycobacteria were grown in 10 ml liquid medium supplemented with oleic acid/albumin/dextrose/catalase (OADC) until an OD600 value of between 0·5 and 1·0 was reached. The cultures were then washed once in PBS. At the time of their inoculation into the SCID mice, aliquots of the mycobacterial cultures were also plated onto 7H10 medium supplemented with OADC to verify their input c.f.u. values.

Transposon mutagenesis.
Lysates of temperature-sensitive phage phAE87 carrying Tn5370 were prepared at 30 °C in Mycobacterium smegmatis mc2155, as described previously (Bardarov et al., 1997 ). M. tuberculosis H37Rv was grown to late exponential phase in 100 ml M-ADC-TW. It was then harvested by centrifugation, washed in 10 ml of a pre-warmed buffer (10 mM Tris/HCl, pH 7·6, 100 mM NaCl, 10 mM MgSO4, 2 mM CaCl2) and resuspended in 10 ml of the same buffer. Cells of M. tuberculosis H37Rv were infected with a high-titre phage lysate (>1x1011 p.f.u. ml-1) at an m.o.i. of 10, and incubated at the non-permissive temperature of 39 °C for 4 h to allow phage attachment. Cells were then plated directly onto selective medium. Transposition frequency, as measured by the number of hygromycin-resistant mutants (total number of input cells)-1, was generally <1x10-6 mutants (input cell)-1. Spontaneous hygromycin resistance was seen in <1x10-9 mutants (input cell)-1.

Ten-thousand colonies were picked from the selective medium and inoculated into individual wells in 96-well microtitre plates containing Middlebrook 7H9 medium, and grown to high cell density at 37 °C. Replica plates were generated and stored at -80 °C.

Genomic DNA extraction.
A protocol based on the method of Pospiech & Neumann (1995) for the extraction of genomic DNA was adapted for a 96-well plate format. For each mutant strain, a 3 ml culture was grown and harvested by centrifugation. The cell pellet was then transferred to a 96-well deep-well plate (Advance Biotech) containing 200 µl lysozyme (5 mg ml-1) in 20 mM Tris/HCl (pH 9·0) and incubated overnight at 37 °C. The cells were then lysed by adding SDS to 2% and proteinase K to 33 µg ml-1 to the wells, followed by further incubation at 37 °C for 3 h. The lysate was transferred to a screw-cap microfuge tube containing 90 µl of 5 M NaCl and 200 µl CHCl3, and the sample was centrifuged at 14000 r.p.m. for 15 min to separate the aqueous and organic phases. The upper aqueous phase was transferred to a 96-well plate; DNA was precipitated from the sample by the addition of 0·6 vol. 2-propanol. The sample was then spun at 2000 r.p.m. for 30 min, and the resulting pellet was resuspended in 50 µl of 5 mM Tris/HCl (pH 8·0) buffer.

Sequencing of the transposon-insertion site.
A restriction-site PCR method adapted from Sarkar et al. (1993) was used for sequencing the transposon-insertion sites. PCRs were performed in 96-well plates (Costar). The first round of PCR used a transposon-specific primer (SP1, 5'-TGCAGCAACGCCAGGTCCACACT-3') and an arbitrarily-degenerate primer (RS6-4, 5'-GTAATACGACTCACTATAGGGCNNNNCATG-3') to amplify the chromosomal sequence flanking the transposon-insertion site. PCRs were carried out in a total volume of 20 µl in 10 mM Tris/HCl (pH 8·3), 50 mM KCl, 1·5 mM MgCl2, 0·01% (w/v) gelatin, 0·25 mM dNTPs, 0·1 µM SP1, 0·5 µM RS6-4 and 0·25 U AmpliTaq (Perkin Elmer). First-round cycling was performed with an initial denaturation step at 94 °C for 5 min, followed by 40 cycles of denaturation at 94 °C for 5 min, annealing at 50 °C for 30 s and extension at 72 °C for 90 s, with a final extension step at 72 °C for 7 min.

A second round of PCR was done using 1 µl of the first-round products as template and the nested primers SP2 (5'-CTCTTGCTCTTCCGCTTCTTCTCC-3') and T7 (5'-TAATACGACTCACTATAGGG-3') to generate more specific products for sequencing. Reactions were carried out in a total volume of 20 µl in 10 mM Tris/HCl (pH 8·3), 50 mM KCl, 1·5 mM MgCl2, 0·01% (w/v) gelatin, 0·25 mM dNTPs, 0·5 µM SP2, 0·5 µM T7, 5% (v/v) DMSO and 0·25 U AmpliTaq. Second-round cycling was performed with a denaturation step at 94 °C for 5 min, followed by 30 cycles of denaturation at 94 °C for 15 s, annealing at 55 °C for 15 s and extension at 72 °C for 30 s, with a final extension step at 72 °C for 7 min.

Sequence determination.
PCR products were treated with Exonuclease I (USB) and shrimp alkaline phosphatase (USB) for 30 min at 37 °C, to degrade residual primers. This treatment was followed by heat inactivation at 65 °C for 20 min. A further nested primer (SP3, 5'-GCACACCCAAGCCAACCAGACC-3') was used as the sequencing primer. PCR products were sequenced on an ABI 3700 sequencer (PE Biosystems) using the BigDye terminator chemistry. Cycle sequencing was performed with 30 cycles of denaturation at 95 °C for 5 min, annealing at 55 °C for 15 s and extension at 64 °C for 4 min.

Location of the transposon-insertion site.
Customized PERL scripts that made extensive use of BioPerl modules (http://bioperl.org) were used to analyse the sequencing results. To identify the precise genome insertion position of the transposon in each mutant, the transposon sequence was trimmed and a BLASTN search of the sequence against the complete H37Rv genome sequence (obtained from http://www.sanger.ac.uk/Projects/M_tuberculosis/) was used to align the two sequences. Only alignments in which at least 20 bp of the chromosomal sequence completely matched the H37Rv genome sequence were retained for further analysis; sequences without transposon sequences were discarded.

Infection of mice and tissue analysis.
For the first in vivo screens, CB-17/Icr SCID mice obtained from Dr C. M. Hetherington (National Institute for Medical Research, Mill Hill, London) were bred under aseptic conditions at the London School of Hygiene & Tropical Medicine. Female mice of between 8 and 10 weeks of age were used. Mice were infected with 1x106 viable mycobacteria in 200 µl of pyrogen-free saline via a lateral tail vein. Where appropriate, infected mice were killed by cervical dislocation in accordance with the humane end-point protocols covered under the Animals Scientific Procedures Act, 1986 (UK).

In a parallel set of in vivo screens, 6-week-old female SCID mice were obtained from the Albert Einstein College of Medicine Animal Institute and infected by the intravenous route with 1x103 viable mycobacteria.

Median survival times were calculated for each group of mice; statistical analyses were performed using Kaplan–Meier plots and log rank tests of survival.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Generation and characterization of the M. tuberculosis library
A library of more than 10000 mutants of M. tuberculosis H37Rv was constructed. Mutagenesis was carried out using the mini-transposon Tn5370 (Cox et al., 1999 ), which was delivered to the bacterium by using the temperature-sensitive mycobacteriophage phAE87 (Bardarov et al., 1997 ). Tn5370 is a derivative of Tn5367 (Bardarov et al., 1997 ) with a number of features that improve its utility (Fig. 1). It consists of a hygromycin-resistance gene that is flanked by resolution (res) sites which enable removal of the marker with Tn{gamma}{delta} resolvase at a later stage. The res sites are in turn flanked by the inverted repeats (IRs) that are required for transposition. The transposase gene (tnpA) is outside of the IRs so it is not carried to the insertion location, which is essential for the construction of stable mutants that have one transposon insertion mutation.



View larger version (6K):
[in this window]
[in a new window]
 
Fig. 1. Structure of transposon Tn5370. IR, inverted repeat; res, resolution site; hyg, hygromycin-resistance gene; tnpA, transposase gene.

 
DNA was extracted from 2280 individual mutants and the sequence at the junction between the 5' end of the transposon element and the chromosomal DNA of each mutant was determined using restriction-site PCR. The location of the transposon insertion was then mapped to the M. tuberculosis H37Rv genome sequence.

In total, 1474 insertion sites were mapped. Unique insertion sites (1329 in total) were identified, 116 insertions appeared in the exact same position twice, 13 appeared in the exact same position three times and one appeared in the exact same position four times. These identical mutants could have arisen from either sibling colonies or cross-contamination. On 20 occasions mutants with the same insertion position were located on the same sequencing plate; on five occasions mutants with the same insertion position were found in adjacent wells. The precise locations of all of the insertions examined in this study can be found as supplementary data in Microbiology Online (http://mic.sgmjournals.org). Eighty-one percent (1189) of the insertions were located within ORFs and 19% (285) were intergenic; the proportion of insertions into ORFs (81%) was slightly less than the proportion of the M. tuberculosis genome that encodes proteins (90%; Cole et al., 1998 ). In total, 351 ORFs were mutated; the characteristics of these ORFs are shown in Table 1. The distribution of the transposon mutants amongst the TubercuList (http://genolist.pasteur.fr/TubercuList/) functional classes of M. tuberculosis is shown in Table 2; the percentage of mutated ORFs within each functional class was between 5 and 12%, with the exception of the PE/PPE class [acidic, glycine-rich proteins of the PE (Pro–Glu) and PPE (Pro–Pro–Glu) protein families], which was overrepresented. Sequencing efforts were terminated at this stage, as the number of insertions into ORFs that had not previously been interrupted had dropped to less than 10% on each microtitre plate.


View this table:
[in this window]
[in a new window]
 
Table 1. Characteristics of the 351 ORFs containing transposon insertions

 

View this table:
[in this window]
[in a new window]
 
Table 1 (cont.)

 

View this table:
[in this window]
[in a new window]
 
Table 2. Distribution of the transposon mutants amongst the functional classes of M. tuberculosis

 
Randomness of insertion
Although the representation of mutants in most of the TubercuList functional classes and the proportion of insertions into ORFs and intergenic DNA suggested that transposition was random, the insertion positions of the transposon were further analysed to determine the distribution of mutations in the M. tuberculosis genome. The position of the transposon in each mutant was plotted on a circular map representing the M. tuberculosis genome (Fig. 2). In the centre of the circle, the number of insertions within 10 kbp intervals are shown; the relative lengths of the lines are proportional to the number of insertions. Fig. 2 demonstrates that the transposon can insert into all sectors of the M. tuberculosis genome, but that some areas are favoured more than others.



View larger version (33K):
[in this window]
[in a new window]
 
Fig. 2. Representation of the distribution of transposon-insertions sites on the chromosome of M. tuberculosis H37Rv, which are indicated by bars on the circle. The internal lines indicate the number of insertions within 10 kbp segments of the chromosome; the relative lengths of the lines are proportional to the number of insertions.

 
The phemonenon of ‘favourable insertion’ was examined further by scrutinizing the insertions in the mutated ORFs. When the number of mutants having insertions in the same ORF was examined, it was seen that some ORFs were more susceptible to transposon insertion than others. Close examination of the DNA sequence in 20 of the regions most susceptible to insertion revealed that the G+C content ranged from 54 to 62 mol% (Table 3), values considerably lower than the G+C content of the whole M. tuberculosis genome (65·6 mol%; Cole et al., 1998 ). No identifiable target site could be found in the DNA adjacent to the transposon, but an A/T or T/A pair was present at positions 4 and 5 in the flanking DNA as reported previously (McAdam et al., 2000 ).


View this table:
[in this window]
[in a new window]
 
Table 3. G+C content for M. tuberculosis genome regions with high insertion rates

 
Analysis of insertions
Fifty-two-percent of the ORFs within the M. tuberculosis genome have homologues elsewhere in the genome. Therefore, it is possible that mutants were selected because the function of the mutated ORF could be carried out by homologous genes. To assess the redundancy of the 351 ORFs containing insertions, BLASTP analysis was performed on the amino-acid sequence of each ORF against the whole M. tuberculosis H37Rv genome. Using a cut-off value of at least 30% identity over 80% of the query length, 107 (30%) of the disrupted ORFs had matches elsewhere in the genome. Further interrogation of the BLASTP output showed that 192 (55%) of the disrupted ORFs had matches elsewhere in the genome with 30% identity over a stretch of at least 100 aa, the average length of a domain. As expected, the TubercuList functional classes that showed the most redundancy included PE/PPEs, and insertion sequences and phages. The functional classes lipid metabolism (fad and pks genes), and cell-wall and cell processes (mmpL genes) also showed redundancy. A similar observation was made in a study of naturally occurring IS6110 insertion loci in clinical isolates of M. tuberculosis (Sampson et al., 1999 ).

We compared the insertion positions of Tn5370 in our library to literature reports of IS6110 insertion loci and deletions occurring in clinical strains of M. tuberculosis. Of the 45 ORFs in the clinical strains of M. tuberculosis described as being disrupted by IS6110 (Sampson et al., 2001 ), 14 matched ORFs containing insertions obtained in this study. In addition, three of the 22 intergenic insertion regions were common to both sets of data. In a study to detect small-scale genomic deletions in 19 clinical isolates of M. tuberculosis, Kato-Maeda et al. (2001 ) found that 11 of the 25 regions deleted within these isolates contained transposon insertions. These insertion regions include three of the five regions described as being associated with IS6110. Six of the 23 regions deleted in the M. tuberculosis clinical strain CDC 1551 compared to H37Rv (Betts et al., 2000 ) were disrupted by Tn5370, and of the 19 regions deleted from M. bovis BCG (Salamon et al., 2000 ; Domenech et al., 2001 ), 11 were found to contain transposon insertions. It is of note that a high proportion of these previously reported unstable regions of the genome are susceptible to Tn5370 insertion. However, only four of the regions that were found to have high transposon-insertion rates in our library (Rv1917c, Rv1500–Rv1507c, Rv2351c–Rv2353c and Rv3343c) contained IS6110 insertion loci or were deleted from M. tuberculosis clinical strains or M. bovis BCG, suggesting that transposition was not biased to these regions.

Approximately 173 (49%) of the disrupted ORFs could lie within operons. Fifty-eight of these were within the last gene of the operon and were, therefore, unlikely to affect expression of other genes. In total, an estimated 90 operons were disrupted. This includes seven insertions in the phosphophthiocerol operon, which has been identified by others as a virulence gene cluster using signature-tagged transposon mutagenesis (Camacho et al., 1999 ; Cox et al., 1999 ).

Virulence of mutants in SCID mice
Since strains were obtained with mutations in ORFs covering a variety of functional classes, the first screen for virulence in vivo was designed to encompass mutations in a broad range of metabolic pathways to determine whether any played an essential role in vivo. Strains with mutations in ORFs from functional classes including DNA repair, DNA transcription, cell division, cell processes, regulators, degradation of macromolecules, small-molecule degradation, energy metabolism and lipid metabolism were chosen (Table 4). A random selection of mutants was made, to avoid biasing the outcome by testing those mutants that we predicted would be attenuated because of some pre-existing knowledge. None of the mutants chosen displayed any obvious phenotype in vitro. The SCID mouse has previously proven very sensitive in revealing changes in M. tuberculosis virulence in the absence of acquired immunity (Smith et al., 2001 ). Mice were infected intravenously with either mutants or wild-type M. tuberculosis H37Rv and their survival was monitored (Table 4). Fig. 3 shows the statistical analysis of median survival times for the transposon mutants compared to the wild-type strain. This experiment was conducted in two groups of mice. Mice infected with M. tuberculosis H37Rv died with a median survival time of 32·5 (Fig. 3a) and 32 days (Fig. 3b). Of the 11 mutants tested, five were unchanged in their virulence when compared to the wild-type strain. One mutant was marginally attenuated in its virulence, as defined by statistical analysis (P<0·05); this strain had a mutation in pvdS, which encodes a putative regulator. The strain with a mutation in Rv1773c, encoding a transcriptional regulator belonging to the IclR family, was marginally more virulent (P<0·05) than the pvdS mutant. Two further mutant strains were more virulent at the P<0·01 level than the pvdS or Rv1773c mutants; the increased virulence of these strains was due to mutations in Rv3252c, encoding a possible monooxygenase, and in Rv3213c, encoding a protein with similarity to Soj, a sporulation inhibitor from Bacillus subtilis, respectively. Interestingly, two of the mutants tested, Rv2467 and Rv0469, were significantly more virulent (P<0·001) than the other mutants, and had disruptions in pepD (encoding an aminopeptidase) and umaA1 (encoding an unknown mycolic acid methyltransferase), respectively.


View this table:
[in this window]
[in a new window]
 
Table 4. Median survival times for SCID mice infected with either selected mutants with transposon insertions in ORFs of known or predicted function, or selected mutants with transposon insertions in ORFs of unknown function

 


View larger version (11K):
[in this window]
[in a new window]
 
Fig. 3. (a, b) Virulence of transposon mutants of M. tuberculosis in SCID mice. SCID mice were infected with wild-type or mutant M. tuberculosis cells (1x106 cells) and their survival was monitored. Data are representative of six mice per group and indicate the median time to death. The bars indicate the range of time to death. *, P<0·05; **, P<0·01; ***, P<0·001.

 
In a parallel study, mutants with transposon insertions exclusively in ORFs of unknown function were selected and tested at a lower infecting dose in SCID mice. Survival times for the mice are given in Table 4 and results are represented graphically in Fig. 4. In Fig. 4, the percentage of survivors over the time-course is plotted, as this best illustrates the extreme phenotype of one mutant (Rv1290). Again, the data show that a range of virulence phenotypes was obtained. Three mutants (Rv1290c, Rv1891c and Rv3404c) gave rise to a highly significant increase in survival time for the mice (P<0·001). After infection with the strain carrying a mutation in Rv1290c, the survival time of the mice increased from a median of 25 to 62 days. Rv1290c encodes an integral membrane protein with homology to unknown proteins in Anabaena sp. and Streptomyces coelicolor. An equivalent gene is present in M. bovis (http://www.sanger.ac.uk/Projects/M_bovis/) and M. tuberculosis CDC 1551 (http://tigrblast.tigr.org/ufmg/), but not in Mycobacterium leprae (Cole et al., 2001 ), M. smegmatis or Mycobacterium avium (http://tigrblast.tigr.org/ufmg/).



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 4. Survival of SCID mice infected with M. tuberculosis mutants with transposon insertions in genes of unknown function. {bullet}, Wild-type; {circ}, Rv0104; {blacktriangledown}, Rv0942; {bigtriangledown}, Rv1046; {blacksquare}, Rv1115; {square}, Rv1290; {diamondsuit}, Rv1891; {lozenge}, Rv3404.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
A better understanding of the virulence mechanisms of M. tuberculosis is required if we are to discover a new generation of drugs against tuberculosis. Despite the sequencing of the genome of several mycobacterial species, including M. tuberculosis and M. leprae, we have few clues as to how M. tuberculosis is able to subvert host defences and invade, survive and grow in the host. It has been shown that for infection of C57Bl/6 mice with M. tuberculosis to persist, a functioning icl (isocitrate lyase) gene is required (McKinney et al., 2000 ). This suggests that metabolic activity contributes to the virulence of M. tuberculosis and, in the absence of homologues to virulence factors found in other bacterial pathogens, it is increasingly likely that regulation of metabolism in this organism in response to its environment makes large contributions to its virulence. Other genes known to contribute to the virulence of M. tuberculosis include those involved in the biosynthesis of complex lipids, such as phthiocerol and phenylphthiocerol dimycocerosates (Camacho et al., 1999 ; Cox et al., 1999 ), the cyclopropane synthetase gene pcaA (Glickman et al., 2000 ) and amino acid biosynthetic pathway genes (Jackson et al., 1999 ; Hondalus et al., 2000 ; Smith et al., 2001 ). A second class of ORFs of interest are those in genes of unknown function – these have no regions homologous to functional motifs and often have no homologues in other bacteria. One such gene, erp, encodes an exported repetitive protein (Berthet et al., 1998 ). The construction and study of mutants of M. tuberculosis has given us useful insights into the metabolic and other processes of this pathogen. A consortium was recently formed with the aim of building a database of known and unknown M. tuberculosis protein structures (http://www.doe-mbi.ucla.edu/TB/). Our study (or this type of analysis) would support this structural genomics approach with functional analysis of the suggested protein activity.

The advance in homologous recombination techniques for M. tuberculosis now means that it is possible to target any gene of choice. However, the process of selecting and inactivating individual genes on a large scale is an awesome one. Transposon mutagenesis offers a convenient way of generating large numbers of mutant bacteria. Isolation of a mutant strain implies that disruption of the gene into which the transposon has inserted has not had a lethal effect on survival of the bacterium in culture. This technique is, therefore, unsuitable for generating mutations in genes essential for growth. Thus, we set out to saturate the chromosome of M. tuberculosis with insertions of transposon Tn5370 by generating a library of mutants, and then to identify the insertion points, and thus the genes disrupted by these insertions, by sequencing the junction between the transposon and chromosomal DNA. It was hoped that this genome-scale approach would suggest which genes were essential for growth of M. tuberculosis by identifying those that did not have a transposon insertion. This approach was limited by the randomness of transposition if there was any insertion bias and by polar effects caused when the insertion was within an operon that had an essential gene or genes distal to the site of disruption. Our use of the mini-transposon Tn5370 ensured that the resulting M. tuberculosis mutants were stable. The presence of the res sites in the res–hyg–res gene cassette facilitated removal of the hygromycin-resistance marker, by transient expression of the Tn{gamma}{delta} resolvase gene in the mutated strain. This would allow us to further analyse phenotypically the unmarked mutant strains as well as to generate multiple mutations in isogenic backgrounds.

Our sequencing effort has so far revealed 1474 Tn5370 transposon-insertion sites within the M. tuberculosis H37Rv genome, disrupting 351 genes. As expected, the proportion of insertions into ORFs (81%) is less than the proportion of the M. tuberculosis genome that is protein-coding (90%). Analysis of clusters of insertions showed that Tn5370 insertion generally favours regions of the genome that have a lower-than-average G+C content. This limits our ability to identify those rare insertions into regions of the chromosome with an above-average G+C content. That 25% of the regions containing IS6110 loci in M. tuberculosis, and 42% of the regions deleted from M. tuberculosis clinical strains or M. bovis BCG, were found to have Tn5370 insertions suggests that these regions of the genome are susceptible to transposon insertion. It is thought that recombination between adjacent insertion elements is a common mechanism through which genomic regions become deleted (Brosch et al., 2001 ). Insertions into these regions, however, were not the most frequently represented insertions in the Tn5370 library and did not bias the distribution of the disrupted genes. With the exceptions of the plcABC cluster and pks2 (Parish & Stoker, 2000 ; Sirakova et al., 2001 ), none of the insertions we generated were into genes that have already been disrupted by homologous recombination methods; therefore, we expect that very many more genes can be disrupted in the M. tuberculosis genome. It should, however, be noted that if a mutant in a particular gene of interest is required homologous recombination methods are more appropriate than transposon mutagenesis.

We randomly selected strains from our library that had mutations in a variety of metabolic pathways and in some ORFs of unknown function and tested their virulence in SCID mice, as we were interested to discover whether any of our mutants had an altered phenotype in vivo. We, and others, have used the SCID mouse model successfully to define the strong attenuation in virulence seen in a number of auxotrophic mutants of M. tuberculosis unable to synthesize tryptophan (trpD), proline (proC) or arginine (argF) (Smith et al., 2001 ; Gordhan et al., 2002 ) – these mutants also exhibited a very strong attenuation in virulence, reflected in a reduction in c.f.u. values and survival in immunocompetent mice. In the screening of the transposon mutants generated in this study, we found only marginal attenuation of virulence in some of the strains and 39% of the mutants screened showed virulence equal to that of wild-type M. tuberculosis. The latter included M. tuberculosis strains with mutations in fpg, responsible for the repair endonuclease protein, and in lsr2, which encodes a protein that is a putative cell-wall antigen recognized by T-cells from leprosy patients (Laal et al., 1991 ). Surprisingly, a significant proportion of the mutants were hypervirulent – these comprised strains with mutations in genes encoding members of several of the functional classification groups studied and included strains with mutations in pepD, an aminopeptidase or metalloprotease gene, and umaA1, a gene encoding a mycolic acid methyltransferase involved in cell-wall biosynthesis.

At present, we are unable to explain the mechanism for the strong attenuation of virulence observed with the Rv1290c mutant because of the unknown function of the protein that this gene encodes. The fact that this gene is present in M. tuberculosis and M. bovis strains but is missing from M. leprae, M. smegmatis and M. avium clearly shows that the Rv1290c gene product does not play a role in determining the cell-division rate. Its insertional inactivation, however, shows that it is not essential for the in vitro growth of virulent M. tuberculosis. The strong attenuation of virulence of the Rv1290c mutant in the SCID mouse model clearly shows that this gene has an important in vivo function in determining the extent of the virulence of M. tuberculosis. This makes this mutant strain a good candidate for further study of the in vivo growth kinetics of M. tuberculosis and any effect on the immunopathology of infection by this organism in an immunocompetent mouse model.

In conclusion, we have demonstrated that it is possible to generate and identify large numbers of M. tuberculosis mutants using transposon mutagenesis. We have also shown that transposon Tn5370 favours insertion into low-G+C regions of the M. tuberculosis chromosome. In a pilot experiment in a highly immunocompromised mouse model, we have shown that several of the mutants we generated have altered phenotypes in vivo. This suggests to us that M. tuberculosis is a pathogen well-adapted to its niche, and that inactivation of individual genes within this organism can slightly increase or decrease its virulence. However, inactivation of some genes of M. tuberculosis, notably Rv1290c, can have a significant impact on the virulence of this bacterium in the absence of acquired immunity. Hence, the Rv1290c gene is a candidate for further investigation of its role in M. tuberculosis virulence. Further studies using immunocompetent mice will identify genes within M. tuberculosis which regulate its survival in the presence of adaptive immunity.


   ACKNOWLEDGEMENTS
 
We are grateful to Heidi Alderton for her expert technical assistance.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Akerley, B. J., Rubin, E. J., Camilli, A., Lampe, D. J., Robertson, H. M. & Mekalanos, J. J. (1998). Systematic identification of essential genes by in vitro mariner mutagenesis. Proc Natl Acad Sci USA 95, 8927-8932.[Abstract/Free Full Text]

Bardarov, S., Kriakov, J., Carriere, C., Yu, S., Vaamonde, C., McAdam, R. A., Bloom, B. R., Hatfull, G. F. & Jacobs, W. R.Jr (1997). Conditionally replicating mycobacteriophages: a system for transposon delivery to Mycobacterium tuberculosis. Proc Natl Acad Sci USA 94, 10961-10966.[Abstract/Free Full Text]

Behr, M. A. & Small, P. M. (1997). Has BCG attenuated to impotence? Nature 389, 133-134.[Medline]

Berthet, F. X., Lagranderie, M., Gounon, P. & 9 other authors (1998). Attenuation of virulence by disruption of the Mycobacterium tuberculosis erp gene. Science 282, 759–762.[Abstract/Free Full Text]

Betts, J. C., Dodson, P., Quan, S., Lewis, A. P., Thomas, P. J., Duncan, K. & McAdam, R. A. (2000). Comparison of the proteome of Mycobacterium tuberculosis strain H37Rv with clinical isolate CDC 1551. Microbiology 146, 3205-3216.[Abstract/Free Full Text]

Bishai, D. M. & Mercer, D. (2001). Modeling the economic benefits of better TB vaccines. Int J Tuberc Lung Dis 5, 984-993.[Medline]

Bowe, F., Lipps, C. J., Tsolis, R. M., Groisman, E., Heffron, F. & Kusters, J. G. (1998). At least four percent of the Salmonella typhimurium genome is required for fatal infection of mice. Infect Immun 66, 3372-3377.[Abstract/Free Full Text]

Brosch, R., Pym, A. S., Gordon, S. V. & Cole, S. T. (2001). The evolution of mycobacterial pathogenicity: clues from comparative genomics. Trends Microbiol 9, 452-458.[Medline]

Camacho, L. R., Ensergueix, D., Perez, E., Gicquel, B. & Guilhot, C. (1999). Identification of a virulence gene cluster of Mycobacterium tuberculosis by signature-tagged transposon mutagenesis. Mol Microbiol 34, 257-267.[Medline]

Cirillo, J. D., Barletta, R. G., Bloom, B. R. & Jacobs, W. R.Jr (1991). A novel transposon trap for mycobacteria: isolation and characterization of IS1096. J Bacteriol 173, 7772-7780.[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.[Medline]

Cole, S. T., Eiglmeier, K., Parkhill, J. & 41 other authors (2001). Massive gene decay in the leprosy bacillus. Nature 409, 1007–1011.[Medline]

Cox, J. S., Chen, B., McNeil, M. & Jacobs, W. R.Jr (1999). Complex lipid determines tissue-specific replication of Mycobacterium tuberculosis in mice. Nature 402, 79-83.[Medline]

Domenech, P., Barry, C. E.3rd & Cole, S. T. (2001). Mycobacterium tuberculosis in the post-genomic age. Curr Opin Microbiol 4, 28-34.[Medline]

Fine, P. E. (1995). Variation in protection by BCG: implications of and for heterologous immunity. Lancet 346, 1339-1345.[Medline]

Glickman, M. S., Cox, J. S. & Jacobs, W. R.Jr (2000). A novel mycolic acid cyclopropane synthetase is required for coding, persistence, and virulence of Mycobacterium tuberculosis. Mol Cell 5, 717-727.[Medline]

Gordhan, B. G., Smith, D. A., Alderton, H., McAdam, R. A., Bancroft, G. J. & Mizrahi, V. (2002). Construction and phenotypic characterization of an auxotrophic mutant of Mycobacterium tuberculosis defective in L-arginine biosynthesis. Infect Immun 70, 3080-3084.[Abstract/Free Full Text]

Hensel, M., Shea, J. E., Gleeson, C., Jones, M. D., Dalton, E. & Holden, D. W. (1995). Simultaneous identification of bacterial virulence genes by negative selection. Science 269, 400-403.[Medline]

Hondalus, M. K., Bardarov, S., Russell, R., Chan, J., Jacobs, W. R.Jr & Bloom, B. R. (2000). Attenuation of and protection induced by a leucine auxotroph of Mycobacterium tuberculosis. Infect Immun 68, 2888-2898.[Abstract/Free Full Text]

Jackson, M., Phalen, S. W., Lagranderie, M., Ensergueix, D., Chavarot, P., Marchal, G., McMurray, D. N., Gicquel, B. & Guilhot, C. (1999). Persistence and protective efficacy of a Mycobacterium tuberculosis auxotroph vaccine. Infect Immun 67, 2867-2873.[Abstract/Free Full Text]

Jacobs, W. R.Jr, Kalpana, G. V., Cirillo, J. D., Pascopella, L., Snapper, S. B., Udani, R. A., Jones, W., Barletta, R. G. & Bloom, B. R. (1991). Genetic systems for mycobacteria. Methods Enzymol 204, 537-555.[Medline]

Kato-Maeda, M., Rhee, J. T., Gingeras, T. R., Salamon, H., Drenkow, J., Smittipat, N. & Small, P. M. (2001). Comparing genomes within the species Mycobacterium tuberculosis. Genome Res 11, 547-554.[Abstract/Free Full Text]

Laal, S., Sharma, Y. D., Prasad, H. K., Murtaza, A., Singh, S., Tangri, S., Misra, R. S. & Nath, I. (1991). Recombinant fusion protein identified by lepromatous sera mimics native Mycobacterium leprae in T-cell responses across the leprosy spectrum. Proc Natl Acad Sci USA 88, 1054-1058.[Abstract]

Manoil, C. & Beckwith, J. (1985). TnphoA: a transposon probe for protein export signals. Proc Natl Acad Sci USA 82, 8129-8133.[Abstract]

McAdam, R. A., Weisbrod, T. R., Martin, J., Scuderi, J. D., Brown, A. M., Cirillo, J. D., Bloom, B. R. & Jacobs, W. R.Jr (1995). In vivo growth characteristics of leucine and methionine auxotrophic mutants of Mycobacterium bovis BCG generated by transposon mutagenesis. Infect Immun 63, 1004-1012.[Abstract]

McAdam, R. A., Quan, S. & Guilhot, C. (2000). Mycobacterial transposons and their applications. In Molecular Genetics of Mycobacteria, pp. 69–84. Edited by G. F. Hatfull & W. J. Jacobs, Jr. Washington, DC: American Society for Microbiology.

McKinney, J. D., Honer Zu Bentrup, K., Munoz-Elias, E. J. & 7 other authors (2000). Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase. Nature 406, 735–738.[Medline]

O’Brien, R. J. & Nunn, P. P. (2001). The need for new drugs against tuberculosis. Obstacles, opportunities, and next steps. Am J Respir Crit Care Med 163, 1055-1058.[Free Full Text]

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]

Pelicic, V., Jackson, M., Reyrat, J. M., Jacobs, W. R.Jr, Gicquel, B. & Guilhot, C. (1997). Efficient allelic exchange and transposon mutagenesis in Mycobacterium tuberculosis. Proc Natl Acad Sci USA 94, 10955-10960.[Abstract/Free Full Text]

Pospiech, A. & Neumann, B. (1995). A versatile quick-prep of genomic DNA from gram-positive bacteria. Trends Genet 11, 217-218.[Medline]

Rubin, E. J., Akerley, B. J., Novik, V. N., Lampe, D. J., Husson, R. N. & Mekalanos, J. J. (1999). In vivo transposition of mariner-based elements in enteric bacteria and mycobacteria. Proc Natl Acad Sci USA 96, 1645-1650.[Abstract/Free Full Text]

Salamon, H., Kato-Maeda, M., Small, P. M., Drenkow, J. & Gingeras, T. R. (2000). Detection of deleted genomic DNA using a semiautomated computational analysis of GeneChip data. Genome Res 10, 2044-2054.[Abstract/Free Full Text]

Sampson, S., Warren, R., Richardson, M., van der Spuy, G. & van Helden, P. (2001). IS6110 insertions in Mycobacterium tuberculosis: predominantly into coding regions. J Clin Microbiol 39, 3423-3424.[Free Full Text]

Sampson, S. L., Warren, R. M., Richardson, M., van der Spuy, G. D. & van Helden, P. D. (1999). Disruption of coding regions by IS6110 insertion in Mycobacterium tuberculosis. Tuber Lung Dis 79, 349-359.[Medline]

Sarkar, G., Turner, R. T. & Bolander, M. E. (1993). Restriction-site PCR: a direct method of unknown sequence retrieval adjacent to a known locus by using universal primers. PCR Methods Appl 2, 318-322.[Medline]

Sassetti, C. M., Boyd, D. H. & Rubin, E. J. (2001). Comprehensive identification of conditionally essential genes in mycobacteria. Proc Natl Acad Sci USA 98, 12712-12717.[Abstract/Free Full Text]

Sirakova, T. D., Thirumala, A. K., Dubey, V. S., Sprecher, H. & Kolattukudy, P. E. (2001). The Mycobacterium tuberculosis pks2 gene encodes the synthase for the hepta- and octamethyl-branched fatty acids required for sulfolipid synthesis. J Biol Chem 276, 16833-16839.[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]

Received 26 April 2002; revised 5 June 2002; accepted 25 June 2002.