Burkholderia pseudomallei is a Gram-negative saprophytic soil bacterium.  It is the aetiological agent of melioidosis, a potentially fatal infectious disease occurring in man and animal.  This disease is endemic in Southeast Asia and Northern Australia, and in inter-tropical zones of Africa, the Indian subcontinent and South-America (Leelarasamee and Bovornkitti, 1989).  In Singapore, melioidosis is a disease of increasing incidence, with a total of 114 cases reported in 1998 alone and an overall incidence rate of 3.6 per 100,000 population (Mi, 1999).  The infection is mostly likely to occur via inhalation or subcutaneous inoculation of contaminated soil or surface water (Dance, 2000a; 2000b).  The disease may be manifested in acute, sub-acute and chronic forms and its incubation period ranges from two days to several decades.

     The pathogenesis of melioidosis is poorly understood.  B. pseudomallei has a broad range of virulence factors that are likely to influence pathogenesis and to be responsible for the various clinical presentations of melioidosis.  It can secret substances that cause tissue necrosis, haemolytic cytolysis and death (Ismail et al., 1987; Ashdown and Koehler, 1990; Sexton et al., 1994; Haase et al., 1997).  The bacterium is inherently resistant to a wide range of antibiotics.  It is also resistant to the bactericidal action of normal human serum despite of strong antibody responses elicited against some bacterial antigens.  Several lines of study, as well as clinical evidences indicate that B. pseudomallei is a facultative intracellular pathogen that is able to survive and grow in eukaryotic cells and causes relapses of the disease, and that invasion, serum resistance and intracellular survival are central to the pathogenesis of diseases due to this organism.

     Most of the recent research has been focused on the identification of virulence factors of B. pseudomallei and the development of diagnostic methods to melioidosis.  These virulence factors are often encoded within clusters of multiple genes that are involved in the process of pathogenesis.  Such gene clusters are called pathogenicity islands (PAIs).  PAIs are present in the genomes of pathogenic organisms but absent from the genomes of nonpathogenic organisms of the same or closely related species, and they contain genes involved in diseases, such as genes encoding invasions, adhesions and secretion factors, and are often sources of toxins.  They often consist of DNA regions that differ from the whole genome in G+C content and in codon usage, which may reflect the generation of PAIs by horizontal gene transfer.  PAIs are often flanked by small directly repeated (DR) sequences.  These sequences may be generated after integration of PAI-specific DNA regions into the host genome via recombination.  PAIs are often associated with transfer RNA (tRNA) genes.  tRNA loci often act as integration sites for foreign DNA.  The association of PAIs and tRNA loci may therefore reflect the generation of PAIs by horizontal gene transfer.  PAIs often carry cryptic or functional genes encoding mobility factors such as integrases, transposases, and insertion sequence (IS) elements or parts of these elements.  PAIs often do not represent homogeneous pieces of DNA but rather are made up of mosaic-like structures which have been generated by a multistep process.  They often represent unstable DNA regions, whose deletions may occur via the direct repeats (DRs) at their ends or via IS elements or other homologous sequences located on PAIs (Hacker and Haper, 2000).

In B. pseudomallei and many other Gram-negative bacteria, PAIs have been associated with type III secretion systems (Winstanley et al., 1999; Attree and Attree, 2002; Mecsas and Strauss, 1996; Hueck, 1998), which are made up of a number of homologous proteins with export functions and are involved in delivering virulence factors directly to host cells (Figure 1a).  The two type III secretion gene clusters of B. pseudomallei were both discovered by sequence homology to known type III secretion apparatuses of other pathogens (Figure 1b).

Figure 1 Type III secretion systems.  a. Type III secretion system of Yersinia (Ysc).  Ysc consists of YopD, YopF, YopJ, YopL, YopN, YopQ, YopR, YopS, YopU, LcrD and YscC (not shown), connected to YopB (B) allowing the export of effector proteins YopE (E), YopH (H), YopM (M), YopO (O), YopP (P) and YopT (T). The genes encoding proteins for the Ysc are conserved among several animal and plant pathogens and in Yersinia are clustered on the 70 kb virulence plasmid.  (Adopted from Wren, 2000)  b. The currently identified two type III secretion gene clusters of B. pseudomallei and the comparison with their homologues, based on which they were discovered.  ORFs are represented as arrows according to the pattern code of predicted proteins as indicated.  (Adopted from Attree and Attree, 2002)

     However, as the genome of B. pseudomallei is being sequenced, an efficient computational tool for the systemic analysis of the genome sequence data and the identification of virulence factors of B. pseudomallei is lacking.  Karlin (2001) recently proposed five criteria for detecting PAIs in bacterial genomes, which are based on G+C frequency, genome signature profile, codon biases from the complete genome and extremes of amino acid usage in the proteome.  To capitalise on the sequence information with regards to our understanding of virulence factors and PAIs, we in silico generated all the putative open reading frames (ORFs) from currently assembled genome sequence of B. pseudomallei and assessed their potential of pathogenicity by analysing the five parameters proposed by Karlin of each putative ORF.  This will provide us with useful information on the more likely candidate genes involved in the pathogenesis of B. pseudomallei and will facilitate the research on its virulence and the development of vaccines against melioidosis.  Furthermore, this project will also help to evaluate the feasibility of using computational methods to predict putative protein functions and properties from nucleotide and amino acid sequences.