Identification and characterization of hemolysin-like proteins similar to RTX toxin in Pasteurella pneumotropica

Abstract

Pasteurella pneumotropica is an opportunistic pathogen that causes lethal pneumonia in immunodeficient rodents. The virulence factors of this bacterium remain unknown. In this study, we identified the genes encoding two RTX toxins, designated as pnxI and pnxII, from the genomic DNA of P. pneumotropica ATCC 35149 and characterized with respect to hemolysis. The pnxI operon was organized according to the manner in which the genes encoded the structural RTX toxin (pnxIA), the type I secretion systems (pnxIB and pnxID), and the unknown orf. The pnxII gene was involved only with the pnxIIA that coded for a structural RTX toxin. Both the structural RTX toxins of deduced PnxIA and PnxIIA were involved in seven of the RTX repeat and repeat-like sequences. By quantitative PCR analysis of the structural RTX toxin-encoding genes in P. pneumotropica ATCC 35149, the gene expression of pnxIA was found to have increased from the early log phase, while that of pnxIIA increased from the late log to the early stationary phase. As expressed in Escherichia coli, both the recombinant proteins of PnxIA and PnxIIA showed weak hemolytic activity in both sheep and murine erythrocytes. On the basis of the results of the Southern blotting analysis, the pnxIA gene was detected in 82% of the isolates, while the pnxIIA gene was detected in 39%. These results indicate that the products of both pnxIA and pnxIIA were putative associations of virulence factors in the rodent pathogen P. pneumotropica.

FIG. 1.

Predicted genetic map of Pasteurella pneumotropica ATCC 35149 pnxI operon and pnxIIA (a) and structure of PnxIA and PnxIIA (b). In approximately the 6-kb region of the pnxI operon, there are four putative open reading frames. Putative hemolysin-type Ca²⁺-binding regions are indicated by white boxes. In the lower section, the RTX repeats of the glycine-rich regions and the repeat-like regions that are similar but not completely identical to the RTX repeats are indicated by solid and dotted lines, respectively. A white box filled with hatched lines contains the Zn²⁺-binding domain, and the serralysin-like sequence is indicated by double lines in the lower section. A gray and black box indicates predicted hydrophobic and hydrophilic regions.

FIG. 2.

Changes in growth rate (a), mean relative quantification of the expression of the pnxIA and pnxIIA genes (b), and the rate of hemoglobin release from the sheep and murine erythrocytes (c) in the case of Pasteurella pneumotropica ATCC 35149 cultured in BHI broth at 37°C. The growth rate was measured at OD₆₀₀. Gene expression was determined by SYBR green I-based quantitative PCR. Hemoglobin release was monitored at OD₄₁₅, followed by an additional incubation in the sheep or murine erythrocytes at 37°C for 2 h; the rate of hemoglobin release was determined by dividing the OD₄₁₅ values by the OD₆₀₀ values.

FIG. 3.

Induced expression analysis of the Pasteurella pneumotropica pnxIA and pnxIIA genes in E. coli (a) and purified recombinant fusion proteins of PnxIA and PnxIIA (b). Coomassie blue-stained SDS-polyacrylamide gel electrophoresis (5 to 20%) analysis of the following proteins. Lane 1, total protein from E. coli transformed with pPNX-IA before induction; lane 2, total protein from E. coli transformed with pPNX-IA induced by 0.1% l-arabinose; lane 3, total protein from E. coli transformed with pPNX-IIA before induction; lane 4, total protein from E. coli transformed with pPNX-IIA induced by 0.1% l-arabinose; lane 5, purified fusion protein of PnxIA; lane 6, purified fusion protein of PnxIIA. Arrowheads indicate the positions of the recombinant fusion proteins.

FIG. 4.

Hemoglobin release from the sheep and murine erythrocytes cultured with the crude lysates prepared from the TMU0812 strain, TMU0812 harboring pPNX-IA, and pPNX-IIA supplemented with 50 mM CaCl₂.