Vol. 1, No. 2 2004
Drug Discovery Today: Technologies Editors-in-Chief Kelvin Lam – Pfizer, Inc., USA Henk Timmerman – Vrije Universiteit, The Netherlands DRUG DISCOVERY
TODAY
TECHNOLOGIES
Target validation
Mycobacterial genetics in target validation Digby F. Warner, Valerie Mizrahi* MRC/NHLS/WITS Molecular Mycobacteriology Research Unit, School of Pathology of the University of the Witwatersrand and the National Health Laboratory Service, PO Box 1038, Johannesburg 2000, South Africa
One-third of the world’s population is infected with Mycobacterium tuberculosis and TB claims 2 million lives every year, yet no new first-line anti-TB drugs have been introduced since 1970. The urgent need to find new agents to reduce the duration and complexity of TB chemotherapy and to eradicate persistent infection
Section Editors Luis Menandez-Arias, Pierre Chatelain, Bernard Masareel Hundreds of lives are lost each day to Mycobacterium tuberculosis, the world’s most deadly bacterial pathogen, yet more than three decades have elapsed since a new TB drug was last developed. In this article, the authors relate how the sequenced genomes of two strains of M. tuberculosis are being used for target validation.
is reinforced by the emergence of multi-drug resistant strains and the HIV pandemic. In this review, we highlight the application of recent advances in mycobacterial genetics in drug target validation.
Introduction Although Mycobacterium tuberculosis presents an enormously challenging therapeutic target, the erosion of the efficacy of TB therapy through non-compliance, the emergence of multidrug resistant (MDR) TB and the HIV pandemic have added impetus to the call for improved drugs. The current anti-tubercular drugs were selected over 30 years ago for their well-defined bactericidal activity against actively replicating tubercle bacilli in the human host. However, the dire need for a sterilising drug to shorten the duration of chemotherapy and ultimately eradicate the massive reservoir of bacteria in the estimated 2 billion asymptomatically infected individuals worldwide has never been addressed. Detailed guidelines for the entire development process have been proposed by the Global Alliance for TB Drug Development (http://www.tballiance.org). In this review, we focus exclusively on the application of molecular genetics in drug target validation. *Corresponding author: (V. Mizrahi) [email protected] 1740-6749/$ ß 2004 Elsevier Ltd. All rights reserved.
DOI: 10.1016/j.ddtec.2004.07.001
Five key technologies The availability of the complete genome sequence of two strains of M. tuberculosis [1,2] underlies all current approaches to target validation and the genetic technologies highlighted below should be viewed as complementary. However, to place these technologies in context, it is necessary to consider briefly the properties required of a TB drug target. The attractiveness of essential genes is obvious – abrogation of function is lethal – and those proteins shown to be essential for growth in vitro have always been considered high-priority targets for the rational design of bactericidal compounds. However, the identification and classification of mutants of M. tuberculosis with phenotypes dependent on the nature and duration of host–bacterial interactions in vivo [3] has demonstrated convincingly that lack of essentiality in vitro should not automatically disqualify a gene for consideration as a TB drug target. On the contrary, the conditional essentiality exemplified by the failure of mutants in genes such as icl (encoding isocitrate lyase) to persist in vivo in the chronic mouse infection model [4] might be precisely the property required of targets for sterilising drugs.
Targeted gene knockout The development of robust methods for targeted gene disruption in M. tuberculosis has been driven by the desire to www.drugdiscoverytoday.com
93
Drug Discovery Today: Technologies | Target validation
create rationally attenuated mutants as potential vaccine candidates [5] and as a means of investigating the function of specific genes, including those encoding potential drug targets [4,6]. Using suicide or conditionally replicating phage or plasmid vectors to deliver mutant alleles, knockout mutants in non-essential genes can now be generated within a matter of weeks [7,8]. The availability of integrating and replicating vectors not only enables mutant strains to be genetically complemented, but also allows the essentiality of gene function to be established for cases in which knockout mutants cannot be recovered.
Random (transposon) mutagenesis Although targeted gene knockout is the method of choice for investigating individual gene function, large-scale approaches to target identification and validation required the development of techniques based on the random insertion of transposable elements throughout the genome. The immense power of transposon mutagenesis rests in the ability to generate and screen large numbers of mutants simultaneously. Although initially applied in the screening of mutant libraries in vitro, the potential of this technique for target validation was fully realised when adapted for use in vivo by the addition of DNA sequence tags [9,10]. The application of signature-tagged mutagenesis (STM; [9]) in the mouse infection model provides a powerful means of identifying and classifying mutants based on phenotypes revealed throughout the infection process [10] and revealing and validating persistence targets which would otherwise be impossible to predict. A particularly innovative tool for target identification and validation in M. tuberculosis is the transposon site hybridisation (TraSH) method, in which mutant pools can be mapped by a DNA microarray-based approach [11,12]. In the first of two studies demonstrating its applicability, the TraSH method was used to the identify genes required for optimal growth of M. tuberculosis in vitro [11]. Many of the genes identified as essential are conserved in the M. leprae genome, which might be considered the minimal mycobacterial gene set, and a large proportion of these have no identifiable bacterial orthologs. Moreover, many of the essential genes have no assigned function. The utility of the TraSH technique in vivo was confirmed by its successful application in a mouse model of infection [12], underscoring the enormous potential of TraSH and related methods for functional genomic screening during various stages of the infection process. However, the most exciting finding of the latter study was that a significant subset of the genes required for growth in vivo are unique to mycobacteria and closely related species and, as such, hold considerable promise for new drug development. In recognition of its value as an information resource, the TraSHbased classifications revealed by these studies have been 94
www.drugdiscoverytoday.com
Vol. 1, No. 2 2004
added to the individual gene descriptions on the TubercuList World Wide Web Server of the Pasteur Institute (http:// genolist.pasteur.fr/TubercuList/). These advances notwithstanding, there are, however, limitations inherent in each of the high-density mutagenesis methods such as potential ‘pooling’ effects, which can accentuate mutant phenotypes owing to competition, the possible complementation of certain mutants in trans and problems associated with the animal model (e.g. mode of infection, relevance to the disease in humans). Unlike STM, TraSH is further limited by the inability to recover individual mutants, necessitating a reverse genetic approach (e.g. targeted knockout) for further analysis. Nonetheless, the essential genes and targets revealed by these methods have already provided a solid foundation for drug discovery efforts. Moreover, modification of experimental conditions, models and application of TraSH in specific mutant backgrounds can be expected to yield many more persistence targets as well as other classes of conditionally essential genes, genes of complementary function, and genes in related pathways.
Conditional mutagenesis Although useful, the single-gene and high-density mutagenesis approaches described above are limited by the permanent abrogation of gene function. Although this might not always affect the ability of a mutant to compete (or infect), there is the possibility that disruption ab initio is unfairly prejudicial, immediately invalidating the observed phenotype. Validation of alternative persistence targets will require the development of techniques to allow genes to be switched off at a particular stage of infection. In addition, the availability of methods for conditional mutagenesis would facilitate the study of essential genes by allowing, for example, the identification of those genes whose functional inactivation leads to cell lysis. Inducible expression of anti-sense RNA provides one possible approach, which has been validated to some extent in preliminary studies [13]. However, this and other methods for generating conditional knockouts are critically limited by the current lack of tightly regulatable promoters for use in mycobacteria, the identification of which is considered an imperative outstanding issue (see Box).
Comparative genomics Mycobacterial genetics has been revolutionised by the advent of the post-genomic era, which was ushered in by the completion of the genome sequence of the M. tuberculosis strain H37Rv in 1998 [1]. Several other mycobacterial genomesequencing projects have subsequently been completed or are underway (M. bovis, M. leprae, M. avium, M. marinum, M. microti and M. smegmatis – http://www.sanger.ac.uk; http:// www.tigr.org). The availability of complete sequences enables whole-genome comparisons within and between
Vol. 1, No. 2 2004
closely related species [14]. The ability to identify patterns of evolutionary inheritance through the analysis of homologous or paralogous gene content, or to isolate those genes which have been inherited/lost together, can be used to infer essentiality for a host of metabolic and virulence functions, or simply to cluster genes based on participation in the same or related biological pathways. Beyond the identification of conserved genes and the assigning of putative function, inter-strain genome comparisons can provide a more dynamic sense of genome evolution than inter-species analyses. Two clinical isolates were recently selected for sequencing (M. tuberculosis strains CDC1551 [2] and 210 [http://www.tigr.org]). Together with the genome sequence of laboratory strain, H37Rv, these additional sequences are providing important insights into inter-strain variation [2], which are of significant value in drug discovery and vaccine development. Furthermore, the application of array-based technology has now made genome-wide comparisons possible in the absence of wholegenome sequence data. A recent study [15] comparing 100 clinical isolates of M. tuberculosis with the H37Rv reference strain uncovered a significant extent of genome plasticity and revealed those genes of H37Rv that have been lost in this group of clinical isolates. Of particular relevance to target validation is the potential to apply the knowledge gained from genomic studies of clinical isolates to further refine the list of genes comprising the essential virulence gene set, which could be used to restrict targets to those unique mycobacterial genes whose inactivation will not adversely affect normal intestinal flora – a problem that has plagued broad-spectrum antibiotics and is especially relevant in an HIV-prevalent context.
Bacterial mRNA analysis Treatment duration is determined by the sterilising activities of the drugs. The principal benefit of the current four-drug combination therapy accrues during the first two months when the compounds are bactericidal; the four to six month continuation phase is required to eliminate persisting bacilli and decrease the risk of relapse. A significant reduction in the duration of treatment will require the development of potent sterilising compounds. However, sterilising activity is difficult to measure and requires an appropriate model of persistence. In addition, conventional microbiological methods for resistance determination might be rendered inapplicable by the fact that persistence targets are likely not be essential for normal growth of the bacteria. All of the above considerations place an extremely high premium on the rapid development and validation of models for persistence in which sterilising compounds might be analysed. Recent advances in RNA-based technologies have shifted the focus of mycobacterial genetics to the analysis of gene expression profiles [16]. A fundamental assumption under-
Drug Discovery Today: Technologies | Target validation
lying the use of expression analysis for target validation is that the expression of a gene suggests that its encoded product is necessary under the given condition. Based on this assumption, sentinel genes can be used as surrogate markers to probe prevailing physiological or environmental conditions. Such markers have already been employed to provide some measure of the appropriateness of particular model systems [17] and to investigate the environments encountered by M. tuberculosis in vitro, in mice [18,19] and in humans [19] using samples containing relatively large numbers of bacilli. Sentinel genes could be of particular use in the rapid evaluation of the effectiveness of new treatment regimens for persistent or latent TB infection. This has the further potential of shortening the duration of clinical trials and thus removing a significant disincentive to new drug development. Realisation of the full potential of mRNA analysis tools is completely reliant on the ability to isolate RNA from bacteria grown in vivo. Studies reported to date have been limited to RNA analysis by RT-PCR-based methods because of the current technical limitations in extracting RNA of sufficient quality and quantity for whole-genome expression profiling by microarray [19] (see Table 1). A novel attempt to overcome this limitation involved the pooling of RNA extracted from some 50 isogenic mice [20]. However, a technological leap in RNA isolation methodologies and in detection sensitivity will be required before expression profiling of samples obtained from the paucibacillary lesions of humans with latent TB infection could be considered, even at the single-gene level.
Strategy comparison These technologies should not be considered as alternatives and it is likely that different approaches will be applied to aspects of the same problem. However, mycobacterial molecular genetics is currently performed almost exclusively in academic laboratories. As such, the technologies employed are dictated primarily by the availability of resources, a sizeable portion of which is allocated to the operation and maintenance of specialised, Biosafety Level 3 (BSL-3) containment facilities. Target validation requires the demonstration of gene essentiality at some stage of the infection process. Three methods for generating mutants and validating phenotypes were highlighted. Targeted gene knockout is a reverse genetic approach to the elucidation of gene function. Pre-selection of the target implies that this approach is hypothesis-driven, and therefore, potentially limited by prevailing dogma or preconception. Although techniques for targeted mutagenesis have advanced considerably in the past few years, the process remains time-consuming and inapplicable to large gene sets. However, the advantages are that the precise nature of the mutation introduced assures the abrogation of gene function and that the same techniques employed to generate www.drugdiscoverytoday.com
95
www.drugdiscoverytoday.com
Name of specific type of technology
Targeted gene knockout
Random (transposon) mutagenesis
Conditional mutagenesis
Comparative genomics
Bacterial mRNA analysis
Names of specific technologies with associated companies and company websites
Allelic exchange mutagenesis using suicide or conditionally replicating delivery vectors; anti-sense. Not commercialised
Signature-tagged mutagenesis (STM); transposon-site hybridisation (TraSH). Not commercialised
Regulatable gene expression; anti-sense. Not commercialised
Bioinformatics, DNA microarray-based technologies. Tuberculist WWW Server http:// genolist.pasteur.fr/ TubercuList/
Transcriptomics (whole-genome expression profiling); real-time PCR. Qiagen tuberculosis Genome array-ready oligo set http://oligos.qiagen.com/ arrays/oligosets_tb.php; real-time PCR machines see for example http://www.appliedbiosystems.com; molecular beacon design – http:// www.molecular-beacons.org
Pros
Direct; target specific; low-tech
Direct; rapid; genome-wide
Direct; target specific; regulatable
In silico
Real-time; dynamic
Cons
Laborious; time-consuming; hypothesis-driven; requires specialist skills and BSL-3 containment facility
Undirected; limited by insertion-site specificity of transposon; requires validation/follow-up; requires specialist skills and BLS-3 containment facility
Paucity of tightly regulatable mycobacterial promoters; requires development before assessment/implementation
Technologically intensive; requires specialist skills
Indirect; technologically intensive; technically demanding; limited by quality and quantity of RNA; requires access to clinical material, specialist skills and BSL-3 containment facility
Costs
Financial – low; time/labour – high
Financial – intermediate; time/labour – intermediate
Not yet established
Financial – high; time/labour – high
Financial – high; time/labour – intermediate
References
[7,8]
[9–12]
[13]
[1,2,14,15]
[17–20]
Drug Discovery Today: Technologies | Target validation
96
Table 1. Comparison of key mycobacterial target validation technologies
Vol. 1, No. 2 2004
Vol. 1, No. 2 2004
Drug Discovery Today: Technologies | Target validation
Links
Related articles
The Global Alliance for TB Drug Development (http://www.tballiance.org) The Stop TB Partnership (http://www.stoptb.org/) The World Health Organisation TB Resource (http://www.who.int/gtb/) The TubercuList World-Wide Web Server (Pasteur Institute) (http://genolist.pasteur.fr/TubercuList/) The TIGR Comprehensive Microbial Resource (CMR) (http://www.tigr.org/tigr-scripts/CMR2/CMRHomePage.spl) NIAID Pathogen Functional Genomics Resource Centre (http://www.niaid.nih.gov/dmid/genomes/pfgrc/ http://pfgrc.tigr.org/) The Welcome Trust Sanger Institute, Pathogen Sequencing Unit (http://www.sanger.ac.uk/Projects/Pathogens/)
Zhang, Y. and Anzel, L.M. (2002) Tuberculosis drug targets. Curr.Drug Targets3, 131–154 Duncan, K. (2003) Progress in TB drug development and what is still needed. Tuberculosis 83, 201–207 Gomez, J.E. and McKinney, J.D. (2004) M. tuberculosis persistence, latency and drug tolerance. Tuberculosis 84, 29–44 Smith, C.V. et al.(2004) TB drug discovery: addressing issues of persistence and resistance. Tuberculosis 84, 45–55 Sharma, K. et al.(2004) Recent advances towards identification of new drug targets for Mycobacterium tuberculosis. Expert Opin. Ther.Targets8, 79–93 Mitchison, D. (2004) The search for new sterilising anti-tuberculosis drugs. Front. Biosci. 9, 1059–1072
the mutant can be applied in complementation analysis, which is central to the target validation process. Random mutagenesis enables the generation and analysis of a large pool of mutants in a single experiment, effectively combining the identification and validation steps. No prior knowledge of individual gene function is required, so bias inherent in target selection is avoided. However, the coverage of mutant libraries is dependent on the randomness of the insertion element and might be restricted by peculiarities in DNA sequence (e.g. G + C content). In addition, the transposon insertion site must be determined and there is a risk that false conclusions might be drawn from disruptions that are non-inactivating or confer polar effects on neighbouring genes. Finally, mutants identified by certain randomised approaches (such as TraSH) cannot be isolated from the pool for further analysis in pure culture, necessitating follow-up by a targeted approach. Both targeted and random mutagenesis approaches are limited by the fact that the disruption, and therefore, abrogation of gene function, is permanent. A solution to this problem lies in the development of systems for conditional mutagenesis. However, the science is new and unproven in mycobacteria, and currently, is limited by the paucity of tightly regulated promoters, which are in urgent need of identification and/or development. Comparative genomics offers an attractive approach to the analysis of gene traits both within and between species, the main advantage being that the bulk of the analysis is performed in silico. Insights gained from genomic analyses can be used to identify and validate surrogate models/organisms. However, the technology platforms are expensive and many analyses are dependent on the ability to source clinical material. Bacterial mRNA analysis offers a real-time glimpse into the transcriptional response of the organism to the imposed environmental or physiological condition, and so is not reliant on the ability to detect a mutant phenotype. However, RNA work is technically demanding and the results are exquisitely sensitive to sample handling and preparation. Major technical hurdles remain to be overcome, particularly the
current inability to extract sufficient high-quality RNA from M. tuberculosis in situ. In addition, the use of an appropriate standard in many situations is still to be satisfactorily resolved.
Conclusions The last decade has witnessed a rapid expansion in the complement and flexibility of available molecular genetic tools for TB drug discovery, and this review has highlighted the application of these techniques to the process of target validation. However, key technical and conceptual limitations exist which remain to be overcome if molecular genetics is to fulfil its crucial role in expediting the development of improved anti-tubercular drugs. M. tuberculosis presents an undeniably daunting and intractable adversary, yet sufficient impetus is surely to be found in the profound and inescapable impact this pathogen continues to exert on global health.
Acknowledgements We thank Ken Duncan for critically reviewing the manuscript.
Outstanding issues Development or identification of tightly regulated mycobacterial promoters Development of methods for generating conditional knockouts in mycobacteria Technical advances in the isolation and detection of mRNA from M. tuberculosis in situ, particularly from paucibacillary lesions in individuals with latent TB infection
References 1 Cole, S.T. et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544 2 Fleischmann, R.D. et al. (2002) Whole-genome comparison of Mycobacterium tuberculosis clinical and laboratory strains. J. Bacteriol. 184, 5479–5490 3 Hingley-Wilson, S.M. et al. (2003) Survival perspectives from the world’s most successful pathogen, Mycobacterium tuberculosis. Nat. Immunol. 4, 949–955 www.drugdiscoverytoday.com
97
Drug Discovery Today: Technologies | Target validation
4 McKinney, J.D. et al. (2000) Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase. Nature 406, 735–738 5 Sambandamurthy, V.K. et al. (2002) A pantothenate auxotroph of Mycobacterium tuberculosis is highly attenuated and protects mice against tuberculosis. Nat. Med. 8, 1171–1174 6 Tullius, M.V. et al. (2003) Glutamine synthetase GlnA1 is essential for growth of Mycobacterium tuberculosis in human THP-1 macrophages and guinea pigs. Infect. Immunol. 71, 3927–3936 7 Parish, T. and 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 8 Bardarov, S. et al. (2002) Specialized transduction: an efficient method for generating marked and unmarked targeted gene disruptions in Mycobacterium tuberculosis, M. bovis BCG and M. smegmatis. Microbiology 148, 3007–3017 9 Hensel, M. et al. (1995) Simultaneous identification of bacterial virulence genes by negative selection. Science 269, 400–403 10 Cox, J.S. et al. (1999) Complex lipid determines tissue-specific replication of Mycobacterium tuberculosis in mice. Nature 402, 79–83 11 Sassetti, C.M. et al. (2003) Genes required for mycobacterial growth defined by high-density mutagenesis. Mol. Microbiol. 48, 77–84 12 Sassetti, C.M. and Rubin, E.J. (2003) Genetic requirements for mycobacterial survival during infection. Proc. Natl. Acad. Sci. USA 100, 12989– 12994
98
www.drugdiscoverytoday.com
Vol. 1, No. 2 2004
13 Parish, T. and Stoker, N.G. (1997) Development and use of a conditional antisense mutagenesis system in mycobacteria. FEMS Microbiol. Lett. 154, 151–157 14 Marmiesse, M. et al. (2004) Macro-array and bioinformatic analyses reveal mycobacterial ‘core’ genes, variation in the ESAT-6 gene family and new phylogenetic markers for the Mycobacterium tuberculosis complex. Microbiology 150, 483–496 15 Tsolaki, A.G. et al. (2004) Functional and evolutionary genomics of Mycobacterium tuberculosis: Insights from genomic deletions in 100 strains. Proc. Natl. Acad. Sci. USA 101, 4865–4870 16 Schnappinger, D. et al. (2003) Transcriptional adaptation of Mycobacterium tuberculosis within macrophages: insights into the phagosomal environment. J. Exp. Med. 198, 693–704 17 Fenhalls, G. et al. (2002) In situ detection of Mycobacterium tuberculosis transcripts in human lung granulomas reveals differential gene expression in necrotic lesions. Infect. Immunol. 70, 6330–6338 18 Shi, L. et al. (2003) Expression of Th1-mediated immunity in mouse lungs induces a Mycobacterium tuberculosis transcription pattern characteristic of nonreplicating persistence. Proc. Natl. Acad. Sci. USA 100, 241–246 19 Timm, J. et al. (2003) Differential expression of iron-, carbon-, and oxygenresponsive mycobacterial genes in the lungs of chronically infected mice and tuberculosis patients. Proc. Natl. Acad. Sci. USA 100, 14321–14326 20 Talaat, A.M. et al. (2004) The temporal expression profile of Mycobacterium tuberculosis infection in mice. Proc. Natl. Acad. Sci. USA 101, 4602– 4607
Drug Discovery Today: Technologies Editors-in-Chief Kelvin Lam – Pfizer, Inc., USA Henk Timmerman – Vrije Universiteit, The Netherlands DRUG DISCOVERY
TODAY
TECHNOLOGIES
Target validation
Mycobacterial genetics in target validation Digby F. Warner, Valerie Mizrahi* MRC/NHLS/WITS Molecular Mycobacteriology Research Unit, School of Pathology of the University of the Witwatersrand and the National Health Laboratory Service, PO Box 1038, Johannesburg 2000, South Africa
One-third of the world’s population is infected with Mycobacterium tuberculosis and TB claims 2 million lives every year, yet no new first-line anti-TB drugs have been introduced since 1970. The urgent need to find new agents to reduce the duration and complexity of TB chemotherapy and to eradicate persistent infection
Section Editors Luis Menandez-Arias, Pierre Chatelain, Bernard Masareel Hundreds of lives are lost each day to Mycobacterium tuberculosis, the world’s most deadly bacterial pathogen, yet more than three decades have elapsed since a new TB drug was last developed. In this article, the authors relate how the sequenced genomes of two strains of M. tuberculosis are being used for target validation.
is reinforced by the emergence of multi-drug resistant strains and the HIV pandemic. In this review, we highlight the application of recent advances in mycobacterial genetics in drug target validation.
Introduction Although Mycobacterium tuberculosis presents an enormously challenging therapeutic target, the erosion of the efficacy of TB therapy through non-compliance, the emergence of multidrug resistant (MDR) TB and the HIV pandemic have added impetus to the call for improved drugs. The current anti-tubercular drugs were selected over 30 years ago for their well-defined bactericidal activity against actively replicating tubercle bacilli in the human host. However, the dire need for a sterilising drug to shorten the duration of chemotherapy and ultimately eradicate the massive reservoir of bacteria in the estimated 2 billion asymptomatically infected individuals worldwide has never been addressed. Detailed guidelines for the entire development process have been proposed by the Global Alliance for TB Drug Development (http://www.tballiance.org). In this review, we focus exclusively on the application of molecular genetics in drug target validation. *Corresponding author: (V. Mizrahi) [email protected] 1740-6749/$ ß 2004 Elsevier Ltd. All rights reserved.
DOI: 10.1016/j.ddtec.2004.07.001
Five key technologies The availability of the complete genome sequence of two strains of M. tuberculosis [1,2] underlies all current approaches to target validation and the genetic technologies highlighted below should be viewed as complementary. However, to place these technologies in context, it is necessary to consider briefly the properties required of a TB drug target. The attractiveness of essential genes is obvious – abrogation of function is lethal – and those proteins shown to be essential for growth in vitro have always been considered high-priority targets for the rational design of bactericidal compounds. However, the identification and classification of mutants of M. tuberculosis with phenotypes dependent on the nature and duration of host–bacterial interactions in vivo [3] has demonstrated convincingly that lack of essentiality in vitro should not automatically disqualify a gene for consideration as a TB drug target. On the contrary, the conditional essentiality exemplified by the failure of mutants in genes such as icl (encoding isocitrate lyase) to persist in vivo in the chronic mouse infection model [4] might be precisely the property required of targets for sterilising drugs.
Targeted gene knockout The development of robust methods for targeted gene disruption in M. tuberculosis has been driven by the desire to www.drugdiscoverytoday.com
93
Drug Discovery Today: Technologies | Target validation
create rationally attenuated mutants as potential vaccine candidates [5] and as a means of investigating the function of specific genes, including those encoding potential drug targets [4,6]. Using suicide or conditionally replicating phage or plasmid vectors to deliver mutant alleles, knockout mutants in non-essential genes can now be generated within a matter of weeks [7,8]. The availability of integrating and replicating vectors not only enables mutant strains to be genetically complemented, but also allows the essentiality of gene function to be established for cases in which knockout mutants cannot be recovered.
Random (transposon) mutagenesis Although targeted gene knockout is the method of choice for investigating individual gene function, large-scale approaches to target identification and validation required the development of techniques based on the random insertion of transposable elements throughout the genome. The immense power of transposon mutagenesis rests in the ability to generate and screen large numbers of mutants simultaneously. Although initially applied in the screening of mutant libraries in vitro, the potential of this technique for target validation was fully realised when adapted for use in vivo by the addition of DNA sequence tags [9,10]. The application of signature-tagged mutagenesis (STM; [9]) in the mouse infection model provides a powerful means of identifying and classifying mutants based on phenotypes revealed throughout the infection process [10] and revealing and validating persistence targets which would otherwise be impossible to predict. A particularly innovative tool for target identification and validation in M. tuberculosis is the transposon site hybridisation (TraSH) method, in which mutant pools can be mapped by a DNA microarray-based approach [11,12]. In the first of two studies demonstrating its applicability, the TraSH method was used to the identify genes required for optimal growth of M. tuberculosis in vitro [11]. Many of the genes identified as essential are conserved in the M. leprae genome, which might be considered the minimal mycobacterial gene set, and a large proportion of these have no identifiable bacterial orthologs. Moreover, many of the essential genes have no assigned function. The utility of the TraSH technique in vivo was confirmed by its successful application in a mouse model of infection [12], underscoring the enormous potential of TraSH and related methods for functional genomic screening during various stages of the infection process. However, the most exciting finding of the latter study was that a significant subset of the genes required for growth in vivo are unique to mycobacteria and closely related species and, as such, hold considerable promise for new drug development. In recognition of its value as an information resource, the TraSHbased classifications revealed by these studies have been 94
www.drugdiscoverytoday.com
Vol. 1, No. 2 2004
added to the individual gene descriptions on the TubercuList World Wide Web Server of the Pasteur Institute (http:// genolist.pasteur.fr/TubercuList/). These advances notwithstanding, there are, however, limitations inherent in each of the high-density mutagenesis methods such as potential ‘pooling’ effects, which can accentuate mutant phenotypes owing to competition, the possible complementation of certain mutants in trans and problems associated with the animal model (e.g. mode of infection, relevance to the disease in humans). Unlike STM, TraSH is further limited by the inability to recover individual mutants, necessitating a reverse genetic approach (e.g. targeted knockout) for further analysis. Nonetheless, the essential genes and targets revealed by these methods have already provided a solid foundation for drug discovery efforts. Moreover, modification of experimental conditions, models and application of TraSH in specific mutant backgrounds can be expected to yield many more persistence targets as well as other classes of conditionally essential genes, genes of complementary function, and genes in related pathways.
Conditional mutagenesis Although useful, the single-gene and high-density mutagenesis approaches described above are limited by the permanent abrogation of gene function. Although this might not always affect the ability of a mutant to compete (or infect), there is the possibility that disruption ab initio is unfairly prejudicial, immediately invalidating the observed phenotype. Validation of alternative persistence targets will require the development of techniques to allow genes to be switched off at a particular stage of infection. In addition, the availability of methods for conditional mutagenesis would facilitate the study of essential genes by allowing, for example, the identification of those genes whose functional inactivation leads to cell lysis. Inducible expression of anti-sense RNA provides one possible approach, which has been validated to some extent in preliminary studies [13]. However, this and other methods for generating conditional knockouts are critically limited by the current lack of tightly regulatable promoters for use in mycobacteria, the identification of which is considered an imperative outstanding issue (see Box).
Comparative genomics Mycobacterial genetics has been revolutionised by the advent of the post-genomic era, which was ushered in by the completion of the genome sequence of the M. tuberculosis strain H37Rv in 1998 [1]. Several other mycobacterial genomesequencing projects have subsequently been completed or are underway (M. bovis, M. leprae, M. avium, M. marinum, M. microti and M. smegmatis – http://www.sanger.ac.uk; http:// www.tigr.org). The availability of complete sequences enables whole-genome comparisons within and between
Vol. 1, No. 2 2004
closely related species [14]. The ability to identify patterns of evolutionary inheritance through the analysis of homologous or paralogous gene content, or to isolate those genes which have been inherited/lost together, can be used to infer essentiality for a host of metabolic and virulence functions, or simply to cluster genes based on participation in the same or related biological pathways. Beyond the identification of conserved genes and the assigning of putative function, inter-strain genome comparisons can provide a more dynamic sense of genome evolution than inter-species analyses. Two clinical isolates were recently selected for sequencing (M. tuberculosis strains CDC1551 [2] and 210 [http://www.tigr.org]). Together with the genome sequence of laboratory strain, H37Rv, these additional sequences are providing important insights into inter-strain variation [2], which are of significant value in drug discovery and vaccine development. Furthermore, the application of array-based technology has now made genome-wide comparisons possible in the absence of wholegenome sequence data. A recent study [15] comparing 100 clinical isolates of M. tuberculosis with the H37Rv reference strain uncovered a significant extent of genome plasticity and revealed those genes of H37Rv that have been lost in this group of clinical isolates. Of particular relevance to target validation is the potential to apply the knowledge gained from genomic studies of clinical isolates to further refine the list of genes comprising the essential virulence gene set, which could be used to restrict targets to those unique mycobacterial genes whose inactivation will not adversely affect normal intestinal flora – a problem that has plagued broad-spectrum antibiotics and is especially relevant in an HIV-prevalent context.
Bacterial mRNA analysis Treatment duration is determined by the sterilising activities of the drugs. The principal benefit of the current four-drug combination therapy accrues during the first two months when the compounds are bactericidal; the four to six month continuation phase is required to eliminate persisting bacilli and decrease the risk of relapse. A significant reduction in the duration of treatment will require the development of potent sterilising compounds. However, sterilising activity is difficult to measure and requires an appropriate model of persistence. In addition, conventional microbiological methods for resistance determination might be rendered inapplicable by the fact that persistence targets are likely not be essential for normal growth of the bacteria. All of the above considerations place an extremely high premium on the rapid development and validation of models for persistence in which sterilising compounds might be analysed. Recent advances in RNA-based technologies have shifted the focus of mycobacterial genetics to the analysis of gene expression profiles [16]. A fundamental assumption under-
Drug Discovery Today: Technologies | Target validation
lying the use of expression analysis for target validation is that the expression of a gene suggests that its encoded product is necessary under the given condition. Based on this assumption, sentinel genes can be used as surrogate markers to probe prevailing physiological or environmental conditions. Such markers have already been employed to provide some measure of the appropriateness of particular model systems [17] and to investigate the environments encountered by M. tuberculosis in vitro, in mice [18,19] and in humans [19] using samples containing relatively large numbers of bacilli. Sentinel genes could be of particular use in the rapid evaluation of the effectiveness of new treatment regimens for persistent or latent TB infection. This has the further potential of shortening the duration of clinical trials and thus removing a significant disincentive to new drug development. Realisation of the full potential of mRNA analysis tools is completely reliant on the ability to isolate RNA from bacteria grown in vivo. Studies reported to date have been limited to RNA analysis by RT-PCR-based methods because of the current technical limitations in extracting RNA of sufficient quality and quantity for whole-genome expression profiling by microarray [19] (see Table 1). A novel attempt to overcome this limitation involved the pooling of RNA extracted from some 50 isogenic mice [20]. However, a technological leap in RNA isolation methodologies and in detection sensitivity will be required before expression profiling of samples obtained from the paucibacillary lesions of humans with latent TB infection could be considered, even at the single-gene level.
Strategy comparison These technologies should not be considered as alternatives and it is likely that different approaches will be applied to aspects of the same problem. However, mycobacterial molecular genetics is currently performed almost exclusively in academic laboratories. As such, the technologies employed are dictated primarily by the availability of resources, a sizeable portion of which is allocated to the operation and maintenance of specialised, Biosafety Level 3 (BSL-3) containment facilities. Target validation requires the demonstration of gene essentiality at some stage of the infection process. Three methods for generating mutants and validating phenotypes were highlighted. Targeted gene knockout is a reverse genetic approach to the elucidation of gene function. Pre-selection of the target implies that this approach is hypothesis-driven, and therefore, potentially limited by prevailing dogma or preconception. Although techniques for targeted mutagenesis have advanced considerably in the past few years, the process remains time-consuming and inapplicable to large gene sets. However, the advantages are that the precise nature of the mutation introduced assures the abrogation of gene function and that the same techniques employed to generate www.drugdiscoverytoday.com
95
www.drugdiscoverytoday.com
Name of specific type of technology
Targeted gene knockout
Random (transposon) mutagenesis
Conditional mutagenesis
Comparative genomics
Bacterial mRNA analysis
Names of specific technologies with associated companies and company websites
Allelic exchange mutagenesis using suicide or conditionally replicating delivery vectors; anti-sense. Not commercialised
Signature-tagged mutagenesis (STM); transposon-site hybridisation (TraSH). Not commercialised
Regulatable gene expression; anti-sense. Not commercialised
Bioinformatics, DNA microarray-based technologies. Tuberculist WWW Server http:// genolist.pasteur.fr/ TubercuList/
Transcriptomics (whole-genome expression profiling); real-time PCR. Qiagen tuberculosis Genome array-ready oligo set http://oligos.qiagen.com/ arrays/oligosets_tb.php; real-time PCR machines see for example http://www.appliedbiosystems.com; molecular beacon design – http:// www.molecular-beacons.org
Pros
Direct; target specific; low-tech
Direct; rapid; genome-wide
Direct; target specific; regulatable
In silico
Real-time; dynamic
Cons
Laborious; time-consuming; hypothesis-driven; requires specialist skills and BSL-3 containment facility
Undirected; limited by insertion-site specificity of transposon; requires validation/follow-up; requires specialist skills and BLS-3 containment facility
Paucity of tightly regulatable mycobacterial promoters; requires development before assessment/implementation
Technologically intensive; requires specialist skills
Indirect; technologically intensive; technically demanding; limited by quality and quantity of RNA; requires access to clinical material, specialist skills and BSL-3 containment facility
Costs
Financial – low; time/labour – high
Financial – intermediate; time/labour – intermediate
Not yet established
Financial – high; time/labour – high
Financial – high; time/labour – intermediate
References
[7,8]
[9–12]
[13]
[1,2,14,15]
[17–20]
Drug Discovery Today: Technologies | Target validation
96
Table 1. Comparison of key mycobacterial target validation technologies
Vol. 1, No. 2 2004
Vol. 1, No. 2 2004
Drug Discovery Today: Technologies | Target validation
Links
Related articles
The Global Alliance for TB Drug Development (http://www.tballiance.org) The Stop TB Partnership (http://www.stoptb.org/) The World Health Organisation TB Resource (http://www.who.int/gtb/) The TubercuList World-Wide Web Server (Pasteur Institute) (http://genolist.pasteur.fr/TubercuList/) The TIGR Comprehensive Microbial Resource (CMR) (http://www.tigr.org/tigr-scripts/CMR2/CMRHomePage.spl) NIAID Pathogen Functional Genomics Resource Centre (http://www.niaid.nih.gov/dmid/genomes/pfgrc/ http://pfgrc.tigr.org/) The Welcome Trust Sanger Institute, Pathogen Sequencing Unit (http://www.sanger.ac.uk/Projects/Pathogens/)
Zhang, Y. and Anzel, L.M. (2002) Tuberculosis drug targets. Curr.Drug Targets3, 131–154 Duncan, K. (2003) Progress in TB drug development and what is still needed. Tuberculosis 83, 201–207 Gomez, J.E. and McKinney, J.D. (2004) M. tuberculosis persistence, latency and drug tolerance. Tuberculosis 84, 29–44 Smith, C.V. et al.(2004) TB drug discovery: addressing issues of persistence and resistance. Tuberculosis 84, 45–55 Sharma, K. et al.(2004) Recent advances towards identification of new drug targets for Mycobacterium tuberculosis. Expert Opin. Ther.Targets8, 79–93 Mitchison, D. (2004) The search for new sterilising anti-tuberculosis drugs. Front. Biosci. 9, 1059–1072
the mutant can be applied in complementation analysis, which is central to the target validation process. Random mutagenesis enables the generation and analysis of a large pool of mutants in a single experiment, effectively combining the identification and validation steps. No prior knowledge of individual gene function is required, so bias inherent in target selection is avoided. However, the coverage of mutant libraries is dependent on the randomness of the insertion element and might be restricted by peculiarities in DNA sequence (e.g. G + C content). In addition, the transposon insertion site must be determined and there is a risk that false conclusions might be drawn from disruptions that are non-inactivating or confer polar effects on neighbouring genes. Finally, mutants identified by certain randomised approaches (such as TraSH) cannot be isolated from the pool for further analysis in pure culture, necessitating follow-up by a targeted approach. Both targeted and random mutagenesis approaches are limited by the fact that the disruption, and therefore, abrogation of gene function, is permanent. A solution to this problem lies in the development of systems for conditional mutagenesis. However, the science is new and unproven in mycobacteria, and currently, is limited by the paucity of tightly regulated promoters, which are in urgent need of identification and/or development. Comparative genomics offers an attractive approach to the analysis of gene traits both within and between species, the main advantage being that the bulk of the analysis is performed in silico. Insights gained from genomic analyses can be used to identify and validate surrogate models/organisms. However, the technology platforms are expensive and many analyses are dependent on the ability to source clinical material. Bacterial mRNA analysis offers a real-time glimpse into the transcriptional response of the organism to the imposed environmental or physiological condition, and so is not reliant on the ability to detect a mutant phenotype. However, RNA work is technically demanding and the results are exquisitely sensitive to sample handling and preparation. Major technical hurdles remain to be overcome, particularly the
current inability to extract sufficient high-quality RNA from M. tuberculosis in situ. In addition, the use of an appropriate standard in many situations is still to be satisfactorily resolved.
Conclusions The last decade has witnessed a rapid expansion in the complement and flexibility of available molecular genetic tools for TB drug discovery, and this review has highlighted the application of these techniques to the process of target validation. However, key technical and conceptual limitations exist which remain to be overcome if molecular genetics is to fulfil its crucial role in expediting the development of improved anti-tubercular drugs. M. tuberculosis presents an undeniably daunting and intractable adversary, yet sufficient impetus is surely to be found in the profound and inescapable impact this pathogen continues to exert on global health.
Acknowledgements We thank Ken Duncan for critically reviewing the manuscript.
Outstanding issues Development or identification of tightly regulated mycobacterial promoters Development of methods for generating conditional knockouts in mycobacteria Technical advances in the isolation and detection of mRNA from M. tuberculosis in situ, particularly from paucibacillary lesions in individuals with latent TB infection
References 1 Cole, S.T. et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544 2 Fleischmann, R.D. et al. (2002) Whole-genome comparison of Mycobacterium tuberculosis clinical and laboratory strains. J. Bacteriol. 184, 5479–5490 3 Hingley-Wilson, S.M. et al. (2003) Survival perspectives from the world’s most successful pathogen, Mycobacterium tuberculosis. Nat. Immunol. 4, 949–955 www.drugdiscoverytoday.com
97
Drug Discovery Today: Technologies | Target validation
4 McKinney, J.D. et al. (2000) Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase. Nature 406, 735–738 5 Sambandamurthy, V.K. et al. (2002) A pantothenate auxotroph of Mycobacterium tuberculosis is highly attenuated and protects mice against tuberculosis. Nat. Med. 8, 1171–1174 6 Tullius, M.V. et al. (2003) Glutamine synthetase GlnA1 is essential for growth of Mycobacterium tuberculosis in human THP-1 macrophages and guinea pigs. Infect. Immunol. 71, 3927–3936 7 Parish, T. and 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 8 Bardarov, S. et al. (2002) Specialized transduction: an efficient method for generating marked and unmarked targeted gene disruptions in Mycobacterium tuberculosis, M. bovis BCG and M. smegmatis. Microbiology 148, 3007–3017 9 Hensel, M. et al. (1995) Simultaneous identification of bacterial virulence genes by negative selection. Science 269, 400–403 10 Cox, J.S. et al. (1999) Complex lipid determines tissue-specific replication of Mycobacterium tuberculosis in mice. Nature 402, 79–83 11 Sassetti, C.M. et al. (2003) Genes required for mycobacterial growth defined by high-density mutagenesis. Mol. Microbiol. 48, 77–84 12 Sassetti, C.M. and Rubin, E.J. (2003) Genetic requirements for mycobacterial survival during infection. Proc. Natl. Acad. Sci. USA 100, 12989– 12994
98
www.drugdiscoverytoday.com
Vol. 1, No. 2 2004
13 Parish, T. and Stoker, N.G. (1997) Development and use of a conditional antisense mutagenesis system in mycobacteria. FEMS Microbiol. Lett. 154, 151–157 14 Marmiesse, M. et al. (2004) Macro-array and bioinformatic analyses reveal mycobacterial ‘core’ genes, variation in the ESAT-6 gene family and new phylogenetic markers for the Mycobacterium tuberculosis complex. Microbiology 150, 483–496 15 Tsolaki, A.G. et al. (2004) Functional and evolutionary genomics of Mycobacterium tuberculosis: Insights from genomic deletions in 100 strains. Proc. Natl. Acad. Sci. USA 101, 4865–4870 16 Schnappinger, D. et al. (2003) Transcriptional adaptation of Mycobacterium tuberculosis within macrophages: insights into the phagosomal environment. J. Exp. Med. 198, 693–704 17 Fenhalls, G. et al. (2002) In situ detection of Mycobacterium tuberculosis transcripts in human lung granulomas reveals differential gene expression in necrotic lesions. Infect. Immunol. 70, 6330–6338 18 Shi, L. et al. (2003) Expression of Th1-mediated immunity in mouse lungs induces a Mycobacterium tuberculosis transcription pattern characteristic of nonreplicating persistence. Proc. Natl. Acad. Sci. USA 100, 241–246 19 Timm, J. et al. (2003) Differential expression of iron-, carbon-, and oxygenresponsive mycobacterial genes in the lungs of chronically infected mice and tuberculosis patients. Proc. Natl. Acad. Sci. USA 100, 14321–14326 20 Talaat, A.M. et al. (2004) The temporal expression profile of Mycobacterium tuberculosis infection in mice. Proc. Natl. Acad. Sci. USA 101, 4602– 4607
