||Mapping, using pulse-field gel electrophoresis, of the entire genomes of various Salmonella strains has shown that the typhoid fever bacterium, Salmonella typhi, contains several "pathogenicity islands", defined as segments of DNA (the largest being 118kb in size) which may have no homologous counterparts in the genome of Salmonella typhimurium (which is not normally invasive for humans) (Liu and Sanderson, 1995). While the viaB gene cluster, necessary for synthesis of the capsular Vi antigen, resides in the 118kb "pathogenicity island" (Liu and Sanderson, 1995), no other genes with functions in the pathogenic process have yet been mapped in these regions. Earlier (Zhang et al., 1997), a clone bank (500 cosmids) of S. typhi DNA was screened by hybridization with total S. typhimurium DNA. Some 5 cosmids, pUST95, pUST88, pUST89, pUST90 and pUST96 did not hybridize with the probe (Zhang, 1997; Zhang et al., 1997). Previous work analysed two overlapping cosmids (pUST95 and pUST90) that encoded a pil operon and a site-specific recombinase gene (rci) (Zhang et al., 1997), and showed that these cosmids contained DNA from the largest pathogenicity island of S. typhi (see summary in Fig. 3.12 in Chapter 3). In this work, a third, non-overlapping cosmid (pUST96) was initially studied. The cosmid was mapped. The intact viaB region was found to be located in this cosmid. A subclone of pUST96 was used to probe total S. typhi DNA. It was shown that the insert of pUST96 did not contain naturally contiguous DNA. A fragment from the pUST96 insert which contained the "break point" was found. This DNA fragment contained DNA naturally contiguous with insert DNA to the left, but also contained (to the right) DNA not naturally contiguous, and was therefore used as a probe to detect another cosmid, pUST97, in the cosmid bank. pUST97 was mapped. It also contains the intact viaB region. The insert DNA of pUST97 was shown to be naturally contiguous. Analysis of pUST97 showed that much of the insert DNA was not probed by total DNA of S. typhimurium, but that DNA on the right end of the insert did, indeed, show homology with the S. typhimurium probe. A fragment containing the rightward boundary of the pathogenicity island was subcloned and part-sequenced, and the boundary point was localized. The switch from S. typhi-unique DNA to DNA expected to be homologous to DNA of S. typhimurium occurred in a purA sequence which, however, appears incapable of contributing to the encoding of a protein with PurA function, as no start codon is evident. In order to find the relationship and distance between pUST96 and pUST95, polymerase chain reaction was used. A ~20kb PCR product was obtained to "bridge" pUST96 and pUST95 with primers from the left end of pUST96 and right end of pUST95. On the left end of the pathogenicity island, the boundary region (defined as a transition between S. typhimurium homology and non-homology) has already been located (Zhang et al., 1997) in a 1.7kb SacI fragment. To define the boundary more precisely, this DNA fragment was largely sequenced, and a particular region of interest was double-strand sequenced. This was a 0.37kb NheI-SphI fragment, which contained the boundary between DNA with database homologues, and S. typhi-unique DNA. There is perfect homology between S. typhimurium DNA upstream of phoN, and S. typhi DNA in the NheI-SphI fragment. There is also some similarity in this region between S. typhi DNA and the phenylalanine tRNA genes of Escherichia coli K-12. A 50nt duplication which is a feature of the ends of the Salmonella senftenberg conjugational transposon CTnscr94 (Hochhut et al., 1997) was also found. Finally, the total size of the largest pathogenicity island in S. typhi was calculated to be approximately 120kb.