Transposon Mutagenesis

Most of our current understanding of the mechanisms of microbial pathogenesis comes from studies where scores of mutants, most often randomly generated by transposon mutagenesis, were monitored in suitable models of infection in search of those presenting reduced virulence [3]. For example, the individual screening of 9516 transposon mutants of Salmonella typhimurium in a macrophage culture assay led to the identification of 115 candidates with a diminished capacity for intracellular survival, of which 83 were also less virulent in vivo [9]. For example, the loss of virulence of one of the mutants, which was mutated in the phoP gene, was subsequently shown to be the result of its reduced resistance to defensins, an important microbicidal mechanism of the phagocyte [10]. The main advantages of this approach over directed mutagenesis are that (a) no a priori assumptions (which are often misleading) have to be made about the identity of the genes important for virulence, and (b) large number of mutants are generated and analyzed, which provides plenty of functional information, at a much faster pace. This throughput can even be improved by the use of transposons engineered to contain unique DNA tags, a procedure known as signature-tagged mutagenesis [11], which allowed large number of mutants to be screened simultaneously in vivo for pathogenic microbes as diverse as Staphylococcus aureus, Legionella pneu-

mophila, Aspergillus fumigatus, or M. tuberculosis [12]. This was previously unfeasible due to the excessively high numbers of animals that would have been necessary and therefore represents an important improvement. For example, 100 infant mice were sufficient to screen 9600 mutants of the intestinal pathogen Vibrio cho-lerae for defects in the colonization of the small intestine [13]. This led to the identification of 251 attenuated mutants that contained mutations in known colonization factors, as well as in genes whose role in colonization was not previously appreciated. Another study performed in M. tuberculosis led to the identification of mutants presenting impaired replication in mouse lungs that harbored several transposon insertions in a cluster of genes [14]. Interestingly, these genes were subsequently shown to be involved in the synthesis and transport of a complex cell-wall-associated lipid, which is solely required for growth in the lungs. Recently, even more processive screening methods were described, such as TraSH (transposon site hybridization), where the mutants within a pool are identified by hybridizing RNA probes corresponding to the chromosomal sequences immediately adjacent to each transposon insertion site, to a microarray containing DNA fragments specific for every predicted ORF [15]. TraSH was used to monitor simultaneously the behavior of as many as 100 000 transposon mutants of M. tuberculosis in a mouse model of infection [16], which led to the identification of 194 genes important for mycobacterial growth in vivo.

In general, this approach has contributed immensely to our understanding of the mechanisms of microbial pathogenesis, and its potential has been further increased by the availability of entire genomes, even though this is less apparent than for reverse genetics. For example, it is now much easier to identify the interrupted genes in most of the selected mutants, which was a laborious and impractical task in the pregenome era. Moreover, it is possible to estimate, following a Poisson distribution, how saturating a library of mutants might be, i.e., how many genes are likely to have been mutated and analyzed, which helps to give an idea of the exhaustiveness of the functional analyses that are performed.

0 0

Post a comment