In Vivo-Induced Genes in Pseudomonas aeruginosa

Generation of chromosomal cointegrated P. aeruginosaPAO909 library.A library of random genomic DNA fragments fromP. aeruginosa were cloned to the promoterlesspurA-lacZY into pIVET1. Genomic DNA fragments from P. aeruginosa strain PAO1 from 1 to 4 kb were size selected, purified, ligated with pIVET1, and electroporated intoEscherichia coli DH5αλpir (strains and plasmids are listed in Table 1). Analysis of 48 recombinant plasmids confirmed that 99% had different P. aeruginosa DNA fragments ranging between 1 and 4 kb (data not shown). This random pool of plasmids was transformed into E. coli SM10λpir and transferred by conjugation into thepurA mutant P. aeruginosa strain PAO909. The resulting chromosomal cointegrated library was represented by at least 2 × 10⁵ colonies of P. aeruginosatransformants.

Selection of P. aeruginosa in vivo-induced genes.The cointegrated PAO909 library was used to infect BALB/c mice weighing 18 to 20 g (a septicemia model) intraperitoneally with 10⁶ to 10⁷ bacteria/mouse and to infect Sprague-Dawley rats intratracheally with 10⁵ bacteria enmeshed into agar beads per lung (a chronic lung infection model [1]). After incubation, bacteria recovered from mouse livers and rat lungs were plated on rich selective medium containing the sensitive chromogenic substrate, 5-bromo-4-chloro-3-indoyl-β-d-galactopyranoside (X-Gal). A collection of 100 ivi fusions were recovered from infected mouse livers and infected rat lungs (Fig.1).

• Open in new tab
• Download powerpoint

Fig. 1.

Induction of ivi genes is required for survival in the animal. The vertical axis represents the number of CFU recovered from the organ of interest after inoculation. The inoculum bar represents the number of CFU of P. aeruginosa injected into each animal. Results are from the BALB/c mouse model of septicemia induced by intraperitoneal injection (10⁴ CFU/mouse; 3 days) (A) and the Sprague-Dawley rat model of chronic lung infection induced via intratracheal instillation of bacterial cells enmeshed in agar beads (5 × 10⁵ CFU/rat; 5 days) (B). Cells were grown overnight at 37°C in rich (adenine-supplemented) laboratory medium. Strains 100 and 101 (Lac⁻) and strain 102 (Lac⁺) were preselected purA-lacfusion strains. PAO909 is a P. aeruginosa auxotroph for adenine. Data are presented as averages of two to five independent assays ± standard deviations.

Characterization of ivi genes.Plasmid preparations from in vivo-selected PAO909 clones were electroporated into E. coli DH5αλpir to allow plasmid rescue. Next, rescued plasmids which had unique restriction patterns and which gave a Lac⁻ phenotype in vitro were selected for further analysis by DNA sequencing. ivi junctions were sequenced using primers homologous to the 5′ region of the purA gene. Similarity searches with the P. aeruginosa genome were performed at the National Center for Biotechnology Information using the uncompleted P. aeruginosa sequence genome database (http://www.pseudomonas.com ). Bioinformatics analysis was done using GeneMark and software in the University of Wisconsin Genetics Computer Group package (version 10.0). We identified three ivi genes with homology to known sequences:pvdD, ftsY, and hag2. The remaining six ivi genes were open reading frames (ORFs) found to have no DNA or protein similarity (Table2).

Strain 131-17, identified by IVET (henceforth IVET 131-17), contains an unidentified ORF of 15,450 nucleotides (named pvdI) coding for a 5,150-amino-acid synthetase having 43% identity with PvdD ofP. aeruginosa. We refer to this synthetase gene located upstream and in the same orientation as pvdD. The PvdD pyoverdine synthetase is involved in the synthesis of the fluorescent siderophore pyoverdine that is essential for iron uptake (6, 7). The independent isolation of IVET 131-17 from both animal models reflects the relative importance of iron acquisition in the establishment and/or maintenance of P. aeruginosa mucosal and systemic infections. Large-scale isolation of candidate virulence genes of P. aeruginosa strain PAK identified the pyochelin receptor (fptA), known to be inducible under iron-deprived conditions, providing further evidence that animal host tissues are deficient in free iron due to the presence of high-affinity iron binding proteins like transferrin (12, 13).

IVET 134-21 carries an ORF that encodes a protein sharing 65% identity with E. coli FtsY, a docking protein that interacts with the prokaryotic signal recognition particle-like complex involved in protein targeting and membrane biogenesis (10). Severalivi genes are involved in bacterial membrane modifications, presumably in response to overcoming environmental stresses imposed on the pathogen during infection (4).

IVET 131-19 carries a predicted peptide which has 43% identity with hemagglutinin Hag2 of Eikenella corrodens, an oral bacterium found in dental plaque (9). Similarly, the adherence ofP. aeruginosa to the mucosa of the oropharynx is believed to be the initial step in colonization of the lower respiratory tract (14). The ivi of IVET 131-19 in both infection models suggests a mucosal and systemic requirement for P. aeruginosa adhesins, as is the case for other mucosal and systemic infection models (4). Cross talk of virulence factors between different in vivo pathogenesis models has been described previously using plants as hosts to identify P. aeruginosavirulence factors (8). The remaining six ivigenes code for proteins having no significant homology to reported proteins found in databases.

Induction of fusions is required for in vivo survival.All eight ivi clones showed no or weak β-galactosidase activity when in vitro promoter activity was tested as described by Slauch et al. (11) (data not shown). Results shown in Fig. 1indicate that the mutant P. aeruginosa purAstrain PAO909 could not be recovered from mouse liver and rat lung tissues, confirming the efficacy of the selection in both animal models. Moreover, the eight ivi fusions showed a 10³- to 10⁵-fold growth advantage in both infection models. Thus, induction of all eight ivi fusions is required for survival in both animal models under conditions of the IVET selection.

These eight ivi genes were shown to be required for survival under the conditions of IVET selection in both animal models, suggesting that at least some host signals present during mouse systemic infection are also present in the rat respiratory mucosa. The propensity to isolate ivi genes coding for proteins related to the expression of surface proteins such as FtsY, PvdI, and Hag2 may suggest that they play a role in virulence by some unknown mechanisms. IVET selects bacterial ivi genes that presumably contribute to the in vivo fitness of the pathogen host tissues. Many of theivi genes that have been recovered from several pathogens infecting a wide variety of animal models are unknown (4). The high possibility of recovering ivi genes of unknown function may reflect our limited knowledge of the bacterial functions required to survive during infection. Many of these presumably reflect the unique lifestyle of each individual pathogen during growth in the host and may not be shared by other pathogens. Thus, further studies on both known and unknown P. aeruginosa ivi gene products will contribute to a better understanding of the pathobiology of P. aeruginosa as an opportunistic pathogen.