SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Although large human populations have been safely immunized against tuberculosis with two live vaccines, Mycobacterium bovis BCG or Mycobacterium microti, the vole bacillus, the molecular basis for the avirulence of these vaccine strains remains unknown. Comparative genomics has identified a series of chromosomal deletions common to both virulent and avirulent species but only a single locus, RD1, that has been deleted from M. bovis BCG and M. microti. Restoration of RD1, by gene knock-in, resulted in a marked change in colonial morphology towards that of virulent tubercle bacilli. Three RD1-encoded proteins were localized in the cell wall, and two of them, the immunodominant T-cell antigens ESAT-6 and CFP-10, were also found in culture supernatants. The BCG::RD1 and M. microti::RD1 knock-ins grew more vigorously than controls in immunodeficient mice, inducing extensive splenomegaly and granuloma formation. Increased persistence and partial reversal of attenuation were observed when immunocompetent mice were infected with the BCG::RD1 knock-in, whereas BCG controls were cleared. Knocking-in five other RD loci did not affect the virulence of BCG. This study describes a genetic lesion that contributes to safety and opens new avenues for vaccine development.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Protection against Mycobacterium tuberculosis, the principal aetiological agent of tuberculosis in humans, can be induced by immunizing with either the BCG vaccine, Mycobacterium bovis BCG (Calmette, 1927), or the vole bacillus, Mycobacterium microti (Hart and Sutherland, 1977), members of the tightly knit M. tuberculosis complex. Both vaccines prevent tuberculosis although, for unknown reasons, they are of limited efficacy against adult pulmonary disease in highly endemic areas (Tuberculosis Prevention Trial, 1980; Ponnighaus et al., 1992; Fine, 1995). M. bovis BCG was derived from a fully virulent isolate of M. bovis by prolonged serial passage that culminated in its attenuation (Calmette, 1927), whereas M. microti is naturally attenuated for humans but causes fulminant tuberculosis in voles and shrews (Wells, 1937; Hart and Sutherland, 1977). In neither case has the molecular basis of the attenuation process been understood, although the safety record of these vaccines is exemplary. Over three billion individuals have been immunized with BCG without major side-effects (Bloom and Murray, 1992), while nearly one million humans have received live vole bacillus vaccine (Sula and Radkovsky, 1976; Hart and Sutherland, 1977). The stability of BCG and its lack of reversion to virulence suggests that an irreversible genetic event such as gene deletion could have contributed to the original attenuation process (Mahairas et al., 1996).

The availability of the complete genome sequence of the paradigm strain, M. tuberculosis H37Rv, has catalysed comparative genomics of mycobacteria (Cole et al., 1998). This new discipline (Cole, 1998) is providing deep insight into the evolution of the M. tuberculosis complex (Brosch et al., 2002) and starting to explain the biological differences observed between its six members. Comparative genomics studies have identified more than 140 genes whose presence is facultative, that may confer differences in phenotype, host range and virulence. Relative to the genome sequence of M. tuberculosis H37Rv (Cole et al., 1998), many of these genes occur in chromosomal regions of difference (RD) that have been deleted from certain species [RD1–RD16, RvD1–RvD5, MiD1–MiD3 (Mahairas et al., 1996; Brosch et al., 1998; 2002; Behr et al., 1999; Gordon et al., 1999; Salamon et al., 2000; Brodin et al., 2002)] but only one, RD1, is restricted to the avirulent strains M. bovis BCG and M. microti (Calmette, 1927; Hart and Sutherland, 1977).

The RD1 end-points are not identical in BCG and vole bacilli (Brodin et al., 2002), but the deletions have removed from both vaccine strains a cluster of six genes (Rv3871–Rv3876) that are part of the ESAT-6 locus (Fig. 1A). ESAT-6, or early secreted antigenic target (Sorensen et al., 1995), is the prototype of a 22-membered protein family whose genes occur in pairs at 11 loci in M. tuberculosis H37Rv (Cole et al., 1998; Tekaia et al., 1999). Three ESAT-6 loci are variable among the species comprising the M. tuberculosis complex (RD1, RD5 and RD8). The missing products are members of the PE, PPE and ESAT-6 protein families that are characteristic of mycobacteria in particular (PE and PPE) and actinobacteria in general (ESAT-6), but are nevertheless of unknown function (Cole et al., 1998; Tekaia et al., 1999; Banu et al., 2002; Brennan and Delogu, 2002). Despite lacking obvious secretion signals, the RD1-encoded ESAT-6 protein (Rv3875) is an abundant component of short-term culture filtrate, and an immunodominant T-cell antigen that induces potent Th1 responses (Sorensen et al., 1995), as is CFP-10 (Berthet et al., 1998a), another member of the ESAT-6 family, encoded by the neighbouring gene (Rv3874).

image

Figure 1. M. bovis BCG and M. microti have chromosomal deletions, RD1, spanning the cfp10-esat6 locus.

A. Map of the cfp10-esat6 region showing the six possible reading frames and the M. tuberculosis H37Rv gene predictions (http:genolist.pasteur.frTubercuList). Stop codons are shown as small vertical bars. The deleted regions are shown for BCG and M. microti with their respective H37Rv genome co-ordinates, and the extent of the conserved ESAT-6 locus, defined by Tekaia et al. (1999), is indicated by the grey bar.

B. Table showing characteristics of deleted regions selected for complementation analysis. Potential virulence factors and their putative functions disrupted by each deletion are shown. The co-ordinates are for the M. tuberculosis H37Rv genome.

C. Diagram of cloned genomic DNA fragments used to complement BCG for the RD1 deletion. The RD1-I106 and RD1-2F9 fragments were cloned in the mycobacterial shuttle vector pYUB412. pAP34 was constructed by subcloning an EcoRI–XbaI fragment into the integrative vector pKINT (Pym et al., 2001). The ends of each fragment are related to the BCG RD1 deletion (shaded box) with black lines, and the H37Rv co-ordinates for the other fragment ends are given in kilobases.

D. Immunoblot analysis, using an ESAT-6 monoclonal antibody, of whole-cell protein extracts from log-phase cultures of M. tuberculosis H37Rv, BCG::pYUB412, BCG::RD1-I106, BCG::RD1–2F9, M. bovis, Mycobacterium smegmatis, M. smegmatis::pYUB412 and M. smegmatis::RD1-2F9.

Download figure to PowerPoint

In this study, we have systematically examined the effects of knocking-in selected RD regions that have been lost since the divergence of M. bovis from the last common ancestor of the M. tuberculosis complex or that may have been deleted as part of the attenuation process that gave rise to M. bovis BCG. Several of these regions are predicted to code for a variety of putative virulence factors. However, the experiments presented here demonstrate that, among the RD regions tested, only the absence of RD1 from M. bovis BCG and M. microti contributes to their avirulence.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Comparative genomics and rationale for knock-ins

Based on their distribution among the tubercle bacilli and their potential to encode virulence functions, RD1, RD3, RD4, RD5, RD7 and RD9 (Fig. 1A and B) were accorded highest priority for functional genomic analysis using ‘knock-ins’ of M. bovis BCG to assess their possible contribution to the attenuation process. RD3, RD4, RD5 and RD7 are predicted to encode proteins involved in phage biogenesis, exopolysaccharide synthesis, membrane de-gradation and cell invasion respectively. Clones spanning these RD regions were selected from an ordered M. tuberculosis H37Rv library of integrating shuttle cosmids (Cole et al., 1998; Bange et al., 1999), and individually electroporated into BCG Pasteur, where they inserted stably into the attB site (Lee et al., 1991). Expression of the knocked-in genes was monitored by reverse transcription polymerase chain reaction (RT-PCR), microarray analysis or Western blotting as appropriate. Only the reintroduction of RD1 led to profound phenotypic alteration, and maps of the clones used are shown (Fig. 1C). These integrating shuttle constructs were also used in parallel experiments conducted with M. microti, which has the same junction sequences as M. bovis BCG in its RD3, RD7 and RD9 loci but differs at RD1 and RD5 (Brodin et al., 2002). Of particular interest is the finding that the RD1 deletion in M. microti is more extensive and has removed 13 genes (Rv3864–Rv3876; Fig. 1A), but these are all present on the integrating clone RD1-2F9 (Rv3860–Rv3885c) used in the experiments reported below.

RD1 expression studies

The functional integrity of the RD1 locus was confirmed by immunoblotting whole-cell protein extracts of the M. bovis BCG::RD1 and M. microti recombinants using an ESAT-6 monoclonal antibody (Harboe et al., 1998) and polyclonal sera for the PPE protein, Rv3873, and CFP-10 (Rv3874) (Colangelli et al., 2000). This demonstrated that the genes were expressed from these constructs at levels comparable with those of M. tuberculosis and M. bovis (Fig. 1D) and that, as reported for M. tuberculosis, culture supernatants of BCG::RD1-2F9 contained CFP-10 and ESAT-6 (data not shown), but not Rv3873, which was strictly cell associated (see below). Essentially similar results were obtained with the recombinant M. microti strains (data not shown).

Colonial morphology

Complementation of BCG with RD1 was accompanied by a change in colonial appearance as the BCG Pasteur ‘knock-in’ strains developed a strikingly different morphotype (Fig. 2A). The RD1-complemented strains adopted a spreading, less rugose morphology, closer to that characteristic of M. tuberculosis and M. bovis, and this was more apparent when the colonies were inspected by light microscopy (Fig. 2B). These changes were seen after complementation with three different RD1 constructs (pAP34, RD1-I106 and RD1-2F9, Fig. 1C) and on complementing M. microti (data not shown). Pertinently, Calmette and Guérin (Calmette, 1927) observed a change in colony morphology during their initial passaging of M. bovis, and our experiments now demonstrate that this change, corresponding to loss of RD1, may have contributed to attenuating this virulent strain. The integrity of the cell wall is known to be a key virulence determinant for M. tuberculosis (Barry, 2001), and changes in both cell wall lipids (Glickman et al., 2000) and proteins (Berthet et al., 1998b) have been shown to alter colony morphology and diminish persistence in animal models.

image

Figure 2. Complementation of BCG Pasteur with the RD1 region alters the colony morphology and leads to the accumulation of Rv3873 and ESAT-6 in the cell wall.

A. Three-week-old cultures of BCG::pYUB412, BCG::I106 or BCG::RD1-2F9 growing on agar. Bacterial cells were washed, resuspended and serially diluted in PBS. A sample of 20 µl of each 10-fold dilution was spotted in five or six drops onto Middlebrook 7H10 agar plates. The white square shows the area of the plate magnified in the image to the right.

B. Light microscope image at 50-fold magnification of BCG::pYUB412 and BCG::RD1 colonies. Drops (5 µl) of bacterial suspensions of each strain were spotted adjacently onto 7H10 plates and imaged after 10 days growth, illuminating the colonies through the agar.

C and D. Immunoblot analysis of different cell fractions of H37Rv obtained from http:www.cvmbs.colostate.edumicrobiologytbResearchMA.html using either an anti-ESAT-6 antibody (C) or anti-Rv3873 (PPE) rabbit serum (D). H37Rv and BCG signify whole-cell extracts from the respective bacteria, and Cyt, Mem and CW correspond to the cytosolic, membrane and cell wall fractions of M. tuberculosis H37Rv.

Download figure to PowerPoint

To provide further insight into these morphological changes, antibodies recognizing three RD1 proteins (Rv3873, CFP10 and ESAT-6) were used in immunocytological and subcellular fractionation analysis. When the different cell fractions from M. tuberculosis were immunoblotted, all three proteins were localized in the cell wall fraction (Fig. 2C), although significant quantities of Rv3873, a PPE protein, were also detected in the membrane and cytosolic fractions (Fig. 2D). Using immunogold staining and electron microscopy, the presence of ESAT-6 in the envelope of M. tuberculosis was confirmed, but no alteration in capsular ultrastructure could be detected (data not shown). Previously, CFP-10 and ESAT-6 have been considered as secreted proteins (Berthet et al., 1998a), but our results suggest that their biological functions are linked directly with the cell wall.

Virulence studies of BCG knock-ins

Changes in colonial morphology are often accompanied by altered virulence (Glickman et al., 2000). Initial assessment of the growth of different BCG::RD1 ‘knock-ins’ in C57BL/6 or BALB/c mice after intravenous infection revealed that complementation did not restore levels of virulence to those of the reference strain M. tuberculosis H37Rv (Fig. 3A). Fifty days after infection, the BCG::RD1 knock-in was over 100-fold less abundant than M. tuberculosis, despite the use of a larger inoculum, whereas BCG alone was present at even lower levels. In longer term experiments, modest yet significant differences were detected in the persistence of the BCG::RD1 ‘knock-ins’ in comparison with BCG controls. After intravenous infection of C57BL/6 mice, only the RD1 ‘knock-ins’ were still detectable in the lungs after 106 days (Fig. 3B). This difference in virulence between the RD1 recombinants and the BCG vector control was far more pronounced in severe combined immunodeficiency (SCID) mice, as can be seen in the lungs and spleens, where BCG::RD1 accumulated to levels between 100- and 1000-fold higher than those seen with BCG (Fig. 3C). These findings were reproduced in five independent experiments of different duration.

image

Figure 3. Complementation of BCG Pasteur with the RD1 region increases bacterial persistence and pathogenicity in mice.

A. Bacteria in the spleen and lungs of BALB/c mice after intravenous (i.v.) infection with 106 colony-forming units (cfu) of M. tuberculosis H37Rv (black) or 107 cfu of either BCG::pYUB412 (white) or BCG::RD1-I106 (light grey).

B. Bacterial persistence in the spleen and lungs of C57BL/6 mice after i.v. infection with 105 cfu of BCG::pYUB412 (white), BCG::RD1-I106 (light grey) or BCG::RD1-2F9 (dark grey).

C. Bacterial multiplication after i.v. infection with 106 cfu of BCG::pYUB412 (white) and BCG::RD1-2F9 (dark grey) in severe combined immunodeficiency (SCID) mice.

For (A)–(C), each time point is the mean of three or four mice, and the error bars represent standard deviations.

D. Spleens from SCID mice 3 weeks after i.v. infection with 106 cfu of BCG::pYUB412, BCG::RD1-2F9 or BCG::I301 (an RD3 ‘knock-in’, Fig. 1B). The scale is in centimetres.

Download figure to PowerPoint

The inability to restore full virulence to BCG Pasteur was not caused by instability of our constructs, as all clones isolated from mice 50 days after infection still expressed ESAT-6 and Rv3873 proteins and displayed spreading colony morphology. Complementing with RD1, we were also unable to render BCG Russia as virulent as M. tuberculosis H37Rv (data not shown). BCG Russia has been passaged less than BCG Pasteur and is therefore presumed to be closer to the original attenuated ancestor (Behr et al., 1999). This indicates that the attenuation of BCG was a polymutational process, and loss of residual virulence for animals was documented in the late 1920s (Oettinger et al., 1999).

Organomegaly and histology

The BCG::RD1 ‘knock-in’ was markedly more virulent, as evidenced by the growth rate in lungs and spleen in SCID mice (Fig. 3C) and also by an increased degree of splenomegaly (Fig. 3D) compared with controls and a BCG::RD3 ‘knock-in’. These effects were also seen in immunocompetent mice, but to a lesser extent. Other organs including the liver and kidney showed signs of inflammation and abscesses. Histological examination of lung sections from SCID mice infected with BCG revealed normal airways and very few acid-fast bacilli (Fig. 4A and C). No epithelioid or giant cells were seen. In contrast, the airways were obstructed and inflamed in the lungs from animals infected with BCG::RD1. The surface of the lung was infiltrated by a few polymorphonuclear cells and numerous macrophages/monocytes (80–90% of the cells) with the formation of granuloma-like structures containing many acid-fast bacilli (Fig. 4B and D). These increases in virulence, after complementation with the RD1 region, demonstrate further that the loss of this genomic locus contributed to the attenuation of BCG.

image

Figure 4. Lung morphology. Histological examination of representative serial 4 µm sections in SCID mice 35 days after infection with BCG::pYUB412 (A and C) or BCG::RD1 (B and D) using tissue samples from the experiment shown in Fig. 3C. Tissue was stained with haematoxylin and eosin (A and B) or by the Ziehl–Neelsen method (C and D). The inset shows a microscopic field of (C) and (D) at 400-fold magnification; note the presence of numerous bacilli in BCG::RD1-infected tissue.

Download figure to PowerPoint

Virulence studies of M. microti knock-ins

To determine whether the findings obtained with the RD1 genes in M. bovis BCG could be replicated in another attenuated host, virulence studies were also performed with recombinant M. microti strains. As the vole bacillus OV254 showed little, if any, multiplication in the lungs and spleens of immunocompetent mice (data not shown), growing more slowly than BCG, experiments were conducted in SCID mice. The OV254::RD1 knock-in displayed a marked growth advantage over the vector control and reached higher numbers in both spleens and lungs (Fig. 5), although these effects were less pronounced than those observed with the equivalent BCG::RD1 knock-in (Fig. 3C). Again, splenomegaly was only observed in animals with M. microti OV254::RD1 infections, and histological examination revealed similar effects to those shown in Fig. 4.

image

Figure 5. Complementation of M. microti OV254 with the RD1 region increases pathogenicity in SCID mice. Bacterial persistence in the spleen and lungs after i.v. infection with 106 cfu of M. microti OV254::pYUB412 (white)or M. microti OV254::RD1-2F9 (dark grey). Each time point is the mean of three mice, and the error bars represent standard deviations.

Download figure to PowerPoint

Knocking-in other RDs

Using the same experimental strategy, we also tested the effects of complementing with RD3, RD4, RD5, RD7 and RD9 (Brosch et al., 1998; Cole et al., 1998; Behr et al., 1999; Gordon et al., 1999) encoding putative virulence factors (Fig. 1B). No difference in colony morphology, such as that shown in Fig. 2A and B, was seen after transformation. In contrast to RD1, the reintroduction of these deleted regions, which are not restricted to avirulent strains of the M. tuberculosis complex, did not affect virulence in immunocompetent mice or, in those cases tested (RD3, RD5, RD7), in SCID mice. It is still possible that deletion effects act synergistically, although other attenuating mechanisms may be at play.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

For over 80 years, BCG has been used as a live vaccine against tuberculosis, although the genetic basis of its attenuation has never been defined. Our data demonstrate that the loss of RD1 was one genetic event that contributed to the attenuation of BCG, as a significant increase in virulence was obtained after its reintroduction. BCG was obtained after 230 serial passages of a virulent isolate of M. bovis in a medium containing bile salts, glycerol and potato extract over a 14 year period from 1908 to 1921 (Calmette, 1927). As early as the 15th passage, Calmette reported a decrease in virulence in guinea pigs (Calmette and Guérin, 1911). Without access to the serial isolates obtained by Calmette and Guérin, it is impossible to determine whether the loss of RD1 was the primary attenuating mutation or if it occurred during the subsequent passaging of BCG. Thus, our attempts to elucidate the key events that led to attenuation are restricted to comparison of extant BCG isolates with virulent members of the M. tuberculosis complex. RD1, together with RD2 and RD3, was first identified by Stover and colleagues (Mahairas et al., 1996) and, in the same study, it was reported that complementation of BCG with the genes from the missing regions failed to increase virulence in the mouse. In contrast, we demonstrate here for the first time that reintroduction of RD1 into BCG does indeed result in a significant increase in virulence. The use of a stably integrating vector, containing the RD1 locus and its flanking regions, may account for this discrepancy as Mahairas et al. (1996) used a vector carrying only a subset of the genes present in RD1-2F9. However, complementation of neither BCG Pasteur nor the least-passaged strain, BCG Russia, with RD1 resulted in the restoration of virulence to levels characteristic of M. tuberculosis or M. bovis. This probably indicates that multiple attenuating mutations accrued during the initial isolation of BCG and also possibly during the period 1928–33 when loss of residual virulence for animals was documented (reviewed by Oettinger et al., 1999).

The results presented here, together with studies of genetic diversity among tubercle bacilli, weaken the possibility that genes contained in the RD3, RD4, RD5, RD7 and RD9 loci contribute extensively to virulence. Instead, it seems more likely that mutations at other unlinked sites were responsible for this further decrease in virulence, and these may correspond to point mutations, inversions, duplications, small insertions or deletions, as these events would have escaped detection by microarrays or related approaches that have been used in whole-genome screens (Behr et al., 1999; Gordon et al., 1999; Salamon et al., 2000). The complete genome sequence of BCG will greatly assist in determining the nature of these other mutations.

Mycobacterium microti is host restricted and naturally attenuated for humans, yet protects against tuberculosis as well as BCG does (Sula and Radkovsky, 1976; Hart and Sutherland, 1977). An independent deletion has occurred in the RD1 region of its genome and, when complemented, a striking increase in the virulence of M. microti OV254 was seen in SCID mice. Once again, this indicates that loss of the RD1 region had an attenuating effect, and preliminary results with ESAT-6 knock-out mutants of M. bovis support this finding (Wards et al., 2000). However, expression of the RD1 locus in both vaccine strains does not lead to full-blown pathogenesis, and this argues in favour of complete attenuation arising as the synergistic effect of multiple mutations. Interestingly, we have identified certain strains of M. microti with RD1 deletions that are associated with disease in humans, suggesting that the loss of RD1 can be compensated for genetically in this species (Brodin et al., 2002). The fact that BCG has never reverted to virulence (Bloom and Fine, 1994; Mahairas et al., 1996; Oettinger et al., 1999) is consistent with the notion that the RD1 deletion represents one of several steps to avirulence in this species and explains why BCG has been such a safe vaccine in the immunocompetent host.

The immunodominant T-cell antigens, ESAT-6 and CFP-10 (Sorensen et al., 1995; Berthet et al., 1998a; Skjøt et al., 2000), are the best known products of the RD1 region, and these small, related proteins accumulate in the culture supernatant. They were thought to be secreted in a signal peptide-independent manner, but the finding that these proteins, and the ESAT-6 orthologue of Mycobacterium leprae (Spencer et al., 2002), are located in the cell wall suggests that they may simply be sloughed off or released as part of a lipid-containing vesicular structure (Beatty et al., 2001). It is also of interest to note the presence of the bulk of the PPE protein, Rv3873, in this subcellular compartment. As this is the second time that a PPE protein has been located in the cell envelope fraction (Sampson et al., 2001), it suggests that others may also be found there. Evidence has been published recently that indicates a direct interaction between ESAT-6 and CFP-10 to form a heterodimer (Renshaw et al., 2002). It is not yet known whether this complex can also include Rv3873 or the PE protein, Rv3872, but it is clear from studies with pAP34 (Fig. 1C) that expression of these four proteins results in a change in colonial morphology (data not shown). The morphotype of the BCG::RD1 knock-in is typical of virulent tubercle bacilli, although it remains to be seen whether this ultrastructural difference is responsible for the increase in virulence or if this is mediated directly by RD1-encoded products. The role of each of the proteins from the conserved ESAT-6 region (Tekaia et al., 1999) will require precise definition.

Elucidation of the genetic events that were responsible for the attenuation of BCG is of considerable relevance for the production of new attenuated tuberculosis vaccines. This approach to vaccine development is driven by the premise that BCG is derived from M. bovis, which lacks 14 genomic regions present in M. tuberculosis, and that these genetic differences adversely affect its immunogenicity and therefore its efficacy as a live vaccine. Attempts to construct attenuated M. tuberculosis strains for use as vaccines are tempered by concerns over their safety. However, if the mutations responsible for the attenuation of BCG were reproduced in M. tuberculosis, this might result in a vaccine genetically close to M. tuberculosis but proven to be safe. The observation that RD3, RD4, RD5, RD7 and RD9, which remove or interrupt 61 open reading frames (ORFs), do not affect virulence means that these genomic fragments can be incorporated in a live vaccine without affecting safety detrimentally.

Recombinant BCG strains offer an alternative strategy for live vaccine development, and encouraging results have been obtained using recombinant BCG expressing cytokines or overproducing antigens (Murray et al., 1996; Horwitz et al., 2000). RD1 is of particular interest in this respect, as both ESAT-6 and CFP-10 are considered as potential candidates for inclusion in a subunit vaccine. However, the results of all immunization studies performed to date indicate that the protection they confer in this formulation is no better than that afforded by BCG (Brandt et al., 2000; Olsen et al., 2000; Louise et al., 2001). ESAT-6 has also been expressed as part of a novel recombinant Salmonella-based vaccine but, again, although the initial results were promising, better efficacy than BCG has not yet been obtained (Mollenkopf et al., 2001). As ESAT-6 and CFP-10 form a heterodimer and are likely to be part of a cell wall-associated complex, it is possible that the expression of ESAT-6 and CFP-10 in the context of BCG may be the optimal means of delivering these potent T-cell antigens, thereby leading to improved protection. Work is currently in progress to assess the potential of the BCG::RD1 strains as vaccine candidates.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Genetic constructs

Individual clones spanning RD1 and the other deletions were selected from an ordered M. tuberculosis H37Rv genomic library (R. Brosch, unpublished) in pYUB412 (Cole et al., 1998; Bange et al., 1999). The inserts in the identified cosmids were ‘end-sequenced’ and compared with the genome sequence of M. tuberculosis H37Rv to establish their size and integrity. Electrocompetent cells for M. bovis BCG Pasteur 1173, BCG Russia or M. microti OV254 were prepared from 400 ml of a 10-day-old Middlebrook 7H9 culture. In the case of M. microti, the 7H9 medium was supplemented with 2 mg ml−1 sodium pyruvate. Bacilli were harvested by centrifugation at 3000 g for 20 min at 16°C, washed with H2O at room temperature and resuspended in 1–2 ml of 10% glycerol at room temperature after recentrifugation. Bacilli (250 µl) were mixed with the respective cosmids and electroporated using a Bio-Rad gene pulser. After electroporation, bacilli were resuspended in medium and left overnight at 37°C. Transformants were selected on Middlebrook 7H11 medium (Difco) supplemented with oleic acid–albumin–dextrose–catalase (OADC; Difco) and 50 µg ml−1 hygromycin or 20 µg ml−1 kanamycin as appropriate. Hygromycin-resistant colonies appearing after 3 weeks were analysed for the presence of the integrated vector by PCR, using primers specific for the antibiotic resistance cassette or RD genes encoded by the cosmid inserts.

Virulence studies

For the virulence assays, 50 ml cultures of the individual mycobacterial strains were grown in parallel in Middlebrook 7H9-ADC medium supplemented with 0.05% Tween 80 and 50 µg ml−1 hygromycin. Bacteria were harvested, washed and resuspended in 50 mM sodium phosphate buffer (pH 7.0). Bacterial suspensions, obtained by brief sonication, were then aliquoted and frozen at −80°C. A single defrosted aliquot was used to quantify the cfus before inoculation. Six-week-old female BALB/cByJIco, C57BL/6 or SCID mice (IFFA Credo) were infected intravenously via the lateral tail vein. Organs from sacrificed mice were homogenized using a Mickle apparatus and 2.5 mm diameter glass beads. Serial fivefold dilutions in PBS were plated on 7H11 agar, and cfus were ascertained after 3 weeks growth at 37°C.

Immunological assays and histopathology

Antibodies to the purified His-tagged proteins Rv3872, Rv3873 and CFP-10 were produced in rabbits using Freund's incomplete adjuvant, whereas a monoclonal antibody to ESAT-6 was a kind gift from P. Andersen. Western blots were performed as described previously (Pym et al., 2001). Organs were fixed in 2% paraformaldehyde before paraffin embedding. Serial 4 µm sections were cut and stained with haematoxylin and eosin or by the Ziehl–Neelsen method according to standard procedures.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

We acknowledge the generous gifts of ESAT-6 monoclonal antibody from P. Andersen, and recombinant RD1 proteins from M. L. Gennaro. M.-C. Prevost and S. Shorte are thanked for help with microscopy, and L. Majlessi, C. Leclerc, C. Demangel, P. Winstanley and C. Gilks for their support and helpful suggestions. This work was funded by the Wellcome Trust, the Institut Pasteur, the European Community (CT-1999-01093) and the Association Française Raoul Follereau. A.S.P. was in receipt of a Wellcome Trust Training Fellowship in Clinical Tropical Medicine.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Bange, W.R., Collins, F.M., and Jacobs, W.R., Jr (1999) Survival of mice infected with Mycobacterium smegmatis containing large DNA fragments from Mycobacterium tuberculosis. Tubercle Lung Dis 79: 171180.
  • Banu, S., Honoré, N., Saint-Joanis, B., Philpott, D., Prévost, M.-C., and Cole, S.T. (2002) Are the PE-PGRS proteins of Mycobacterium tuberculosis variable surface antigens? Mol Microbiol 44: 919.
  • Barry, C.E. (2001) Interpreting cell wall ‘virulence factors’ of Mycobacterium tuberculosis. Trends Microbiol 9: 237241.
  • Beatty, W.L., Ullrich, H.J., and Russell, D.G. (2001) Mycobacterial surface moieties are released from infected macrophages by a constitutive exocytic event. Eur J Cell Biol 80: 3140.
  • Behr, M.A., Wilson, M.A., Gill, W.P., Salamon, H., Schoolnik, G.K., Rane, S., and Small, P.M. (1999) Comparative genomics of BCG vaccines by whole-genome DNA microarrays. Science 284: 15201523.
  • Berthet, F.X., Rasmussen, P.B., Rosenkrandt, I., Andersen, P., and Gicquel, B. (1998a) A Mycobacterium tuberculosis operon encoding ESAT-6 and a novel low-molecular-mass culture filtrate protein (CFP-10). Microbiology 144: 31953203 .
  • Berthet, F.X., Lagranderie, M., Gounon, P., Laurent-Winter, C., Ensergueix, D., Chavarot, P., et al. (1998b) Attenuation of virulence by disruption of the Mycobacterium tuberculosis erp gene. Science 282: 759762.
  • Bloom, B.R., and Fine, P.E.M. (1994) The BCG experience: implications for future vaccines against tuberculosis. In Tuberculosis: Pathogenesis, Protection, and Control. Bloom, B.R. (ed.). Washington, DC: American Society for Microbiology Press, pp. 531557.
  • Bloom, B.R., and Murray, C.J.L. (1992) Tuberculosis: commentary on a reemergent killer. Science 257: 10551064.
  • Brandt, L., Elhay, M., Rosenkrands, I., Lindblad, E.B., and Andersen, P. (2000) ESAT-6 subunit vaccination against Mycobacterium tuberculosis. Infect Immun 68: 791795.
  • Brennan, M.J., and Delogu, G. (2002) The PE multigene family: a molecular mantra for mycobacteria. Trends Microbiol 10: 246249.
  • Brodin, P., Eiglmeier, K., Marmiesse, M., Billault, A., Garnier, T., Niemann, S., et al. (2002) Bacterial artificial chromosome-based comparative genomic analysis identifies Mycobacterium microti as a natural ESAT-6 deletion mutant. Infect Immun 70: 55685578.
  • Brosch, R., Gordon, S.V., Billault, A., Garnier, T., Eiglmeier, K., Soravito, C., et al. (1998) Use of a Mycobacterium tuberculosis H37Rv bacterial artificial chromosome (BAC) library for genome mapping, sequencing and comparative genomics. Infect Immun 66: 22212229.
  • Brosch, R., Gordon, S.V., Marmiesse, M., Brodin, P., Buchrieser, C., Eiglmeier, K., et al. (2002) A new evolutionary scenario for the Mycobacterium tuberculosis complex. Proc Natl Acad Sci USA 99: 36843689.
  • Calmette, A. (1927) In La Vaccination Preventive Contre la Tuberculose. Paris: Masson et cie, pp. 250.
  • Calmette, A., and Guérin, C. (1911) Recherches expérimentales sur la défense de l’organisme contre l’infection tuberculeuse. Ann Inst Pasteur 25: 625641.
  • Colangelli, R., Spencer, J.S., Bifani, P., Williams, A., Lyaschenko, K., Keen, M.A., et al. (2000) MTSA-10, the product of the Rv3874 gene of Mycobacterium tuberculosis, elicits tuberculosis-specific, delayed-type hypersensitivity in guinea pigs. Infect Immun 68: 990993.
  • Cole, S.T. (1998) Comparative mycobacterial genomics. Curr Opin Microbiol 1: 567571.
  • Cole, S.T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris, D., et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393: 537544.
  • Fine, P.E.M. (1995) Variation in protection by BCG: implications of and for heterologous immunity. Lancet 346: 13391345.
  • Glickman, M.S., Cox, J.S., and Jacobs, W.R., Jr (2000) A novel mycolic acid cyclopropane synthetase is required for coding, persistence, and virulence of Mycobacterium tuberculosis. Mol Cell 5: 717727.
  • Gordon, S.V., Brosch, R., Billault, A., Garnier, T., Eiglmeier, K., and Cole, S.T. (1999) Identification of variable regions in the genomes of tubercle bacilli using bacterial artificial chromosome arrays. Mol Microbiol 32: 643656.
  • Harboe, M., Malin, A.S., Dockrell, H.S., Wiker, H.G., Ulvund, G., Holm, A., et al. (1998) B-cell epitopes and quantification of the ESAT-6 protein of Mycobacterium tuberculosis. Infect Immun 66: 717723.
  • Hart, P.D., and Sutherland, I. (1977) BCG and vole bacillus vaccines in the prevention of tuberculosis in adolescence and early adult life. Br Med J 2 (6082):293295.
  • Horwitz, M.A., Harth, G., Dillon, B.J., and Maslesa-Galic, S. (2000) Recombinant bacillus calmette-guerin (BCG) vaccines expressing the Mycobacterium tuberculosis 30-kDa major secretory protein induce greater protective immunity against tuberculosis than conventional BCG vaccines in a highly susceptible animal model. Proc Natl Acad Sci USA 97: 1385313858.
  • Lee, M.H., Pascopella, L., Jacobs, W.R., and Hatfull, G.F. (1991) Site-specific integration of mycobacteriophage L5: integration-proficient vectors for Mycobacterium smegmatis, Mycobacterium tuberculosis, and bacille Calmette-Guérin. Proc Natl Acad Sci USA 88: 31113115.
  • Louise, R., Skjot, V., Agger, E.M., and Andersen, P. (2001) Antigen discovery and tuberculosis vaccine development in the post-genomic era. Scand J Infect Dis 33: 643647.
  • Mahairas, G.G., Sabo, P.J., Hickey, M.J., Singh, D.C., and Stover, C.K. (1996) Molecular analysis of genetic differences between Mycobacterium bovis BCG and virulent M. bovis. J Bacteriol 178: 12741282.
  • Mollenkopf, H.J., Groine-Triebkorn, D., Andersen, P., Hess, J., and Kaufmann, S.H. (2001) Protective efficacy against tuberculosis of ESAT-6 secreted by a live Salmonella typhimurium vaccine carrier strain and expressed by naked DNA. Vaccine 19: 40284035.
  • Murray, P.J., Aldovini, A., and Young, R.A. (1996) Manipulation and potentiation of antimycobacterial immunity using recombinant bacille Calmette-Guerin strains that secrete cytokines. Proc Natl Acad Sci USA 93: 934939.
  • Oettinger, T., Jorgensen, M., Ladefoged, A., Haslov, K., and Andersen, P. (1999) Development of the Mycobacterium bovis BCG vaccine: review of the historical and biochemical evidence for a genealogical tree. Tubercle Lung Dis 79: 243250.
  • Olsen, A.W., Hansen, P.R., Holm, A., and Andersen, P. (2000) Efficient protection against Mycobacterium tuber-culosis by vaccination with a single subdominant epitope from the ESAT-6 antigen. Eur J Immunol 30: 17241732.
  • Ponnighaus, J.M., Fine, P.E., Sterne, J.A., Wilson, R.J., Msosa, E., Gruer, P.J., et al. (1992) Efficacy of BCG vaccine against leprosy and tuberculosis in northern Malawi. Lancet 339: 636639.
  • Pym, A.S., Domenech, P., Honoré, N., Song, J., Deretic, V., and Cole, S.T. (2001) Regulation of catalase-peroxidase (KatG) expression, isoniazid sensitivity and virulence by furA of Mycobacterium tuberculosis. Mol Microbiol 40: 879889.
  • Renshaw, P.S., Panagiotidou, P., Whelan, A., Gordon, S.V., Hewinson, G.R., Williamson, R.A., and Carr, M.D. (2002) Conclusive evidence that the major T-cell antigens of the M. tuberculosis complex ESAT-6 and CFP-10 form a tight, 1:1 complex and characterisation of the structural properties of ESAT-6, CFP-10 and the ESAT-6–CFP-10 complex: implications for pathogenesis and virulence. J Biol Chem 274: 2159821603.
  • Salamon, H., Kato-Maeda, M., Small, P.M., Drenkow, J., and Gingeras, T.R. (2000) Detection of deleted genomic DNA using a semiautomated computational analysis of GeneChip data. Genome Res 10: 20442054.
  • Sampson, S.L., Lukey, P., Warren, R.M., Van Helden, P.D., Richardson, M., and Everett, M.J. (2001) Expression, characterization and subcellular localization of the Mycobacterium tuberculosis PPE gene Rv1917c. Tuberculosis (Edinb) 81: 305317.
  • Skjøt, R.L.V., Oettinger, T., Rosenkrands, I., Ravn, P., Brock, I., Jacobsen, S., and Andersen, P. (2000) Comparative evaluation of low-molecular-mass proteins from Mycobacterium tuberculosis identifies members of the ESAT-6 family as immunodominant T-cell antigens. Infect Immun 68: 214220.
  • Sorensen, A.L., Nagai, S., Houen, G., Andersen, P., and Andersen, Å. (1995) Purification and characterization of a low molecular mass T-cell antigen secreted by Mycobacterium tuberculosis. Infect Immun 63: 17101717.
  • Spencer, J.S., Marques, M.A., Lima, M.C., Junqueira-Kipnis, A.P., Gregory, B.C., Truman, R.W., and Brennan, P.J. (2002) Antigenic specificity of the Mycobacterium leprae homologue of ESAT-6. Infect Immun 70: 10101013.
  • Sula, L., and Radkovsky, J. (1976) Protective effects of M. microti vaccine against tuberculosis. J Hyg Epidemiol Microbiol Immunol 20: 16.
  • Tekaia, F., Gordon, S.V., Garnier, T., Brosch, R., Barrell, B.G., and Cole, S.T. (1999) Analysis of the proteome of Mycobacterium tuberculosis in silico. Tubercle Lung Dis 79: 329342.
  • Tuberculosis Prevention Trial, Madras (1980) Trial of BCG vaccines in South India for tuberculosis prevention. Ind J Med Res 72 (Suppl.):174.
  • Wards, B.J., De Lisle, G.W., and Collins, D.M. (2000) An esat6 knockout mutant of Mycobacterium bovis produced by homologous recombination will contribute to the development of a live tuberculosis vaccine. Tubercle Lung Dis 80: 185189.
  • Wells, A.Q. (1937) Tuberculosis in wild voles. Lancet 232: 1221.