Jacqueline E. Shea is Group Head, Research and Development, Microscience Ltd, Imperial College of Science, Technology, and Medicine, Hammersmith Campus, London, United Kingdom, and David W. Holden is Professor, Department of Infectious Diseases, Imperial College of Science, Technology, and Medicine, Hammersmith Campus, London, United Kingdom.
Links to Other ASM Pages:
Signature-Tagged Mutagenesis Helps Identify Virulence Genes
This genetic method offers researchers an efficient means for identifying virulence genes in bacterial pathogens
Jacqueline E. Shea and David W. Holden
Signature-tagged mutagenesis (STM) was developed as a strategy to enable researchers to screen large numbers of mutants of bacterial pathogens simultaneously to determine whether a specific genetic change results in attenuated virulence in a single host. It is one of a number of ``functional'' genomic techniques available to complement current bacterial pathogen genome sequencing efforts.
The genes identified from the analysis of genome sequences of bacterial pathogens can be classified into three groups. One group consists of those genes whose functions can be predicted with some certainty directly from their nucleotide sequences. This group includes genes encoding certain highly conserved enzymes with well defined biochemical properties, such as DNA polymerases and tRNA synthetases.
A second group includes gene products whose specific biochemical functions are not known, but whose general roles may be inferred from overall amino acid sequence similarities or conserved motifs that are indicative of particular activities. Members of this group include genes encoding two-component regulatory systems and ABC transporters. A third class is made up of genes whose sequences reveal no recognizable information about their function. In the genomes sequenced to date, genes belonging to this third class usually represent in excess of 30% of the genome total.
Genome Sequencing Efforts Drive Need for Rapid Functional Analyses
For sequences in the second and third groups further genetic analysis is required to determine their functions and to evaluate their contributions to virulence. Gene function is often investigated through careful phenotypic analysis of mutant strains (classical genetics). This analysis may involve creating temperature-sensitive strains following UV- or chemical-induced mutagenesis (particularly useful for the analysis of essential genes) or by insertional mutagenesis. Convenient methods for producing insertional mutations include the use of transposons, insertion-duplication mutagenesis, and in vitro transposition followed by transformation of a naturally competent bacterial host.
Once individual mutants are available, they can then be assessed for phenotypic defects by using a variety of screening techniques. Some of these techniques are designed to reflect the effects of the host environment on the infection process, such as limited iron availability, oxidative stress, or invasion and replication within host cells. Although such screening techniques have proved successful in identifying bacterial virulence determinants, they tend to be useful only for identifying single steps of the infective process.
The definitive screen for a virulence defect is to test a mutant strain for its effects on a living host, which requires that a suitable model of infection be available. However, screening large numbers of bacterial mutants individually for attenuation of virulence in an animal model can be costly and often is inefficient.
STM Technique Offers an Efficient Virulence Factor-Screening Alternative
STM combines the power of conventional mutagenesis methods with the ability to trace the fate of individual mutant strains within complex pools. It is an alternative method to in vivo expression technology (IVET) pioneered by John Mekalanos and his collaborators at Harvard Medical School, Boston, Mass., which identifies those genes that are expressed during bacterial growth in the host. Similarly, differential fluorescence induction (DFI), developed by Stanley Falkow and colleagues at Stanford University, also identifies genes whose expression is altered during growth in host cells.
Instead of screening a single mutant for attenuation of virulence, STM enables a researcher to screen many strains simultaneously in a single animal. STM relies on two key elements. First, it exploits the fact that the host will select against growth of mutant strains that carry mutations in virulence genes, and these mutants will fail to be recovered from the host following inoculation of a mixed pool of mutants (negative selection).
Second, STM distinguishes different mutants by incorporating different identifying DNA ``signature-tags'' into the mutant strains. This is accomplished by including signature-tags on a transposon or insertion-duplication plasmid (depending on the method of mutagenesis), such that the tags mark the resulting mutations.
In their original design, these polynucleotide tags consisted of a central 40-base-pair (bp) variable region that permits differentiation between the tags, flanked by two 20-bp invariant arms, to which oligonucleotide primers bind for PCR amplification. Different tagged mutants are then assembled into pools of up to 96 separate mutants, and each pool is screened after inoculation into and recovery from an animal host. The presence of each of the tagged strains in the inoculum and their presence or absence after infection of the host is detected by PCR amplification and labelling of the tags followed by hybridization analysis with DNA dot or colony blots of the individual mutant strains. Attenuated mutants are identified by their failure to be recovered from the host.
Challenges in Defining Virulence Factors, Limitations of STM
Although there appears to be little consensus on the precise definition of virulence factors, Stanley Falkow at Stanford University, Stanford, Calif., Richard Moxon at John Radcliffe Hospital in Oxford, United Kingdom, and their respective colleagues divide the molecular basis of virulence into three main areas.
First, virulence consists of bacterial factors required for host and tissue tropism. Second, it includes factors required for multiplication within the host, including those required for evasion of host defences and nutrient acquisition within the host. And, third, virulence includes factors involved in aspects of host toxicity.
STM theoretically can be used to identify genes encoding the first two classes of virulence factors. However, as a result of mixed infections with different strains, it is unlikely to detect mutations in virulence determinants whose functions can be transcomplemented by the presence of other strains (for example, certain toxins). STM screens are also unlikely to identify mutations in essential genes, unless the insertion mutation results in an altered protein with reduced function, but which is still sufficient for viability and growth of mutant cells in vitro. Furthermore, if there is genetic redundancy for a virulence function, then a single gene mutation may lead to a subtle virulence defect below the level of detection by signature-tag hybridization analysis.
STM Revealed Expected and Surprise Virulence Factors from S. typhimurium
STM was first applied to Salmonella typhimurium in 1995. This pathogen was chosen because so much is known about its molecular genetics, there is a well-established model for studying typhoid disease in mice, and many of its virulence genes had already been identified, some of which we expected to isolate through STM analysis.
In this early application, signature-tags were incorporated into a mini-Tn5 transposon cassette, and a pool of tagged transposons was used to generate an S. typhimurium mutant library. The library was then prescreened for those mutants that contained tags that amplified well by PCR and gave strong hybridization signals when hybridized back to colony blots of their genomic DNA. (Approximately 25% of mutants contained tags that did not amplify or hybridize efficiently and were discarded.) The selected tagged mutants were then reassembled into 12 pools of 96 mutants in microtiter dishes and screened for attenuated virulence in the mouse model of typhoid (Fig. 1).
From this screen of 1,152 mutants using a total of 24 mice, 40 mutants were identified that were attenuated in virulence. Nucleotide sequencing of the DNA immediately adjacent to the transposon insertion of 28 mutants revealed that 13 mutants contained transposon insertions into known virulence genes, thereby validating the method. Other mutants contained transposon insertions into genes with similarity to other known genes, including genes with similarity to other type III secretion system genes of gram-negative bacteria. LD50 and mixed infection experiments with the wild-type strain confirmed that these mutants were strongly attenuated in virulence.
When we characterized the genes with similarity to type III secretion genes, we found that they form part of a large pathogenicity island (Salmonella pathogenicity island 2, SPI2) encoding a type III secretion system apparatus and some of its probable secreted targets. Type III secretion systems are important virulence determinants in some gram-negative bacterial pathogens of plants and animals.
Nevertheless, this result was surprising, as most other pathogens only seem to contain one type III secretion system, and a type III secretion system involved in invasion of epithelial cells encoded by the inv/spa genes and located on another pathogenicity island (SPI-1) had already been identified in S. typhimurium. In contrast, the type III secretion system encoded by SPI-2 is required for systemic infection in mice, and work from several laboratories indicates that it is required for bacterial replication within macrophages.
How STM Fares when Applied to Salmonella and Other Bacterial Pathogens
The STM method has been successfully used to analyze virulence factors from other bacterial pathogens, including Staphylococcus aureus (by two separate research groups), Streptococcus pneumoniae, Vibrio cholerae, Yersinia enterocolitica, and Legionella pneumophila (Table 1). Other STM screens have also been performed in Pseudomonas aeruginosa, Proteus mirabilis, and, most recently, in Mycobacterium tuberculosis. Moreover, in these studies, STM successfully identified genes from two of the three main classes of virulence determinants.
First, STM helped to identify genes involved in host and tissue tropism, such as the tcp operon of V. cholerae. Second, it also identified genes required for bacterial multiplication within the host, including the many factors involved in nutrient acquisition (such as purD, lysA, and trpA), and others required for the evasion or suppression of host defenses (such as SPI-2 of S. typhimurium).
As might be expected, the STM screens did not identify secreted cytotoxins, suggesting that if these mutants were present they may have been rescued by transcomplementation. However, as many other genes were identified whose role in virulence is not known, it may be too early to conclude that this class of virulence genes will not be recovered in an STM screen.
The animal and tissue model chosen for the STM screen can also significantly influence which mutations are identified as resulting in attenuation. For instance, Alessandra Pollissi and colleagues at the Glaxo Wellcome research facility in Verona, Italy, conducted an STM screen of S. pneumoniae in mice that develop pneumonia when exposed to this pathogen and found attenuating mutations in 126 different genes. On rescreening these mutant strains individually in a mouse model of septicemia, only 71% are attenuated. In other words, some genes which are required for virulence in the pneumonia model do not appear to be important in the septicemia model.
Taking this approach still further, Silvija Coulter and colleagues at Pathogenesis, in Seattle, Wash., screened mutants of S. aureus in three separate mouse infection models--abscesses, wounds, and systemic bacteremia. A total of 237 attenuated mutants were identified, of which 54% are attenuated in the wound model, 47% in the systemic bacteremia model, and 46% in the abscess model. There is also considerable overlap among the infection models, with 27% of the mutants being attenuated in two types of infections and 9.7% being attenuated in all three types. Mutations that result in attenuation in several types of infections are more likely to represent systemic virulence factors than those that affect only one type. This approach to testing mutants provides valuable insights into the requirements for a pathogen to cause infections in different tissue types of the same species.
The STM Technique Has Undergone Several Technical Refinements
Table 1 - What you should know to view this information
The original STM method used DNA colony blots for hybridization between target DNAs and labelled tags. However, colony blot hybridizations are relatively insensitive and can be difficult to perform with some pathogens. Therefore the STM screens of S. aureus and S. pneumoniae were refined by employing plasmid and signature-tag DNA dot blots as target DNAs for hybridizations. These modifications enhance sensitivity of the detection process and allow nonradioactive detection methods to be employed.
Perhaps the most useful refinement to date has been to make an initial selection of 96 tags that amplify and hybridize well but that do not cross-hybridize with the other tags. These tags can then be used to generate 96 individually tagged transposons or suicide plasmids that are then used repeatedly to generate as many mutants as are required. This modification removes the need to prescreen mutants for tags that amplify and hybridize efficiently and thus simplifies the overall process.
Most of the STM screens published to date have used transposons as mutagens. However, Pollissi and colleagues used insertion-duplication mutagenesis in their STM screen of Streptococcus pneumoniae. Specifically, they used a library of signature-tagged suicide plasmids containing small genomic fragments of S. pneumoniae to generate mutations by targeted integration of the plasmid into the genome, thereby disrupting targeted genes. Importantly, when using this method of mutagenesis, the genomic fragment should be small enough to avoid reconstituting the targeted gene on integration of the plasmid.
Several Issues To Consider Before Setting Up an STM Screen
Before setting up an STM screen procedure, investigators need to consider a number of issues. First, they will need an efficient mutagenesis method, allowing incorporation of signature-tags. Both transposon (mini-Tn5 and Tn917) and insertion-duplication mutagenesis have been used for this purpose. In vitro transposon mutagenesis of genomic DNA may also be an option for those organisms that are naturally competent for the uptake and integration of their own DNA.
Second, there must be a suitable animal model of infection. The animal model chosen is vital to the success of the screen and must support simultaneous replication of multiple bacterial strains. The other variables in an STM screen, such as the number of mutants in each pool (pool size), inoculum level, and site and time of recovery, are specific to the pathogen and model of infection employed.
The pool size is critical and is limited by the animal model and minimum concentration of each tag required to generate an adequate and reproducible hybridization signal. As the pool size increases, there is a decreased likelihood that each tagged virulent strain will be recovered in sufficient numbers for detection. The largest pool size used to date has been 96 tagged strains. The inoculum level should be such that the infective dose is capable of establishing an infection for each tagged strain. However, if the inoculum level is too high, the host's immune response may be overwhelmed and attenuated mutants will not be identified.
Pool size and level of inoculum are also dictated by the route of inoculation and the presence of any steps in the infective process likely to limit the passage of bacteria from one organ to another. Models of infection that contain limiting steps--for instance, passage of pathogens through the M cells of Peyer's patches in the gut epithelium or through the blood-brain barrier-may give rise to clonal infections and therefore would not be suitable for use in an STM screen. To set up the parameters for a successful screen it is useful to have a known attenuated mutant to use as a control and to repeat the screening of an initial pool in several animals to detect random deletion of tagged strains.
Future Applications
Once genes are identified by STM, they can be used to investigate the processes they control by biochemical and physiological analysis of the gene products and mutant phenotypes. Perhaps the most interesting genes identified by STM are those for which either their nucleotide or predicted protein sequence provides no clue as to their specific function. However, as the strains possess an attenuated virulence phenotype, they can be subjected to a range of in vitro tests based on knowledge of the environment they encounter during infection to help pinpoint their precise role in pathogenesis.
It should also be possible to perform STM screens in other haploid pathogens that have a suitable system of mutagenesis and infection model. For instance, STM screens should be feasible for fungal pathogens, such as Aspergillus fumigatus, and possibly for protozoan pathogens such as Toxoplasma spp. Indeed, Brendan Cormack at Johns Hopkins University School of Medicine together with Stanley Falkow and colleagues at Stanford University, have successfully screened a signature-tagged mutant bank of Candida glabrata in a human epithelial cell assay and identified a novel adhesin.
On a practical note, the attenuated pathogens that STM identifies may prove helpful for those researchers who are trying to develop live, attenuated vaccines. Moreover, some of the proteins encoded by virulence genes may be suitable targets for future antimicrobial drug development.
SUGGESTED READING
Chiang, S. L., and J. J. Mekalanos. 1998. Use of signature-tagged transposon mutagenesis to identify Vibrio cholerae genes critical for colonisation. Mol. Microbiol. 27:797-805.
Coulter, S. N., R. W. Schwan, E. Y. W. Ng, M. H. Langhorne, H. D. Richie, S. Westbrock-Wadman, W. O. Hufnagle, K. R. Folger, A. S. Bayer, and C. K. Stover. Staphylococcus aureus genetic loci impacting growth and survival in multiple infection environments. Mol. Microbiol. 30:393-404.
Falkow, S. What is a pathogen? ASM News 63:359-365.[PDF]
Hensel, M., J. E. Shea, C. Gleeson, M. D. Jones, E. Dalton, and D. W. Holden. 1995. Simultaneous identification of bacterial virulence genes by negative selection. Science 269:400-403.
Mei, J.-M., F. Nourbakhsh, C. W. Ford, and D. W. Holden. 1997. Identification of Staphylococcus aureus virulence genes in a murine model of bacteraemia using signature-tagged mutagenesis. Mol. Microbiol. 26:399-407.
Polissi, A., A. Pontiggia, G. Feger, M. Altieri, H. Mottl, L. Ferrari, and D. Simon. 1998. Large-scale identification of virulence genes from Streptococcus pneumoniae. Infect. Immun. 66:5620-5629.
Email: webmaster@asmusa.org