Identification of a novel adhesion molecule involved in the virulence of Legionella pneumophila

Abstract

Legionella pneumophila is an intracellular bacterium, and its successful parasitism in host cells involves two reciprocal phases: transmission and intracellular replication. In this study, we sought genes that are involved in virulence by screening a genomic DNA library of an L. pneumophila strain, 80-045, with convalescent-phase sera of Legionnaires' disease patients. Three antigens that reacted exclusively with the convalescent-phase sera were isolated. One of them, which shared homology with an integrin analogue of Saccharomyces cerevisiae, was named L. pneumophila adhesion molecule homologous with integrin analogue of S. cerevisiae (LaiA). The laiA gene product was involved in L. pneumophila adhesion to and invasion of the human lung alveolar epithelial cell line A549 during in vitro coculture. However, its presence did not affect multiplication of L. pneumophila within a U937 human macrophage cell line. Furthermore, after intranasal infection of A/J mice, the laiA mutant was eliminated from lungs and caused reduced mortality compared to the wild isolate. Thus, we conclude that the laiA gene encodes a virulence factor that is involved in transmission of L. pneumophila 80-045 and may play a role in Legionnaires' disease in humans.

FIG. 1.

Schematic diagrams of laiA in the chromosome of L. pneumophila 80-045 and of plasmids constructed in this study. The chromosome of 80-045 is presented as a solid line, and the vectors are presented as broken lines. The region corresponding to the laiA gene is shown as heavy lines. The putative translational start codon of the LaiA protein is indicated as 1. The arrow indicates the direction of transcription of the laiA gene. Sites for primers f3 and r8 used in PCR analysis are shown. Triangles show the insertion of the Km^r cassette. Abbreviations for the restriction enzymes: Ac, AccII; Ao, Aor51HI; Bam, BamHI; Ban, BanII; Ps, PstI; Pv, PvuII; Sp, SphI; Xb, XbaI; Xh, XhoI.

FIG. 2.

(A) Schematic representation of the chromosomal region containing laiA and its five paralogs in L. pneumophila 80-045. The corresponding genes of Philadelphia-1 (3) and Lp02 (27) are presented under the predicted ORFs. (B) Domain structure of LaiA. The putative divalent-cation-binding site is shown as a hatched box, the transmembrane domain is shown as a gray box, and the DXSX motifs are shown as black boxes. (C and D) Comparison of divalent-cation-binding motifs. (C) The consensus amino acid sequences for the 13 amino acid residues of EF hand divalent-cation-binding motifs (1). Parentheses, acceptable amino acids; brackets, unacceptable amino acids; X, any amino acid. Cation-coordinating sites are shown in boldface. (D) Alignment of the putative cation-binding site of LaiA, two cation-binding sites in C. albicans αInt1p, three cation-binding sites in α_M, and one cation-binding site in β₃ (14, 15, 24). The standard single-letter code is used. A dash indicates a gap. The figure was designed according to the work of Hostetter (24).

FIG. 3.

Immunoblot of the total proteins of L. pneumophila with mouse monospecific anti-LaiA convalescent-phase sera. After suspension in SDS-PAGE sample buffer, samples were electrophoresed on an 8% running gel, blotted, and reacted as described in Materials and Methods. An open arrowhead indicates the full-length LaiA protein. Bands shown by closed arrowheads seem to be the degraded fractions of LaiA. Positions of molecular mass markers are indicated on the right. Lane 1, total cellular protein of 80-045; lane 2, total cellular protein of PK-treated 80-045; lane 3, total cellular protein of LAM0101.

FIG. 4.

Ability of L. pneumophila to adhere to (A) and invade (B) A549 alveolar epithelial cells. Data points and error bars represent the means and standard errors. Shown here are the results with the wild L. pneumophila strain 80-045, the laiA mutant strain LAM0101, the laiA-complemented (pMMBLG0503) strain LAM0102, and LAM0103, the strain carrying the empty pMMB207C vector. All experiments were performed more than three times with triplicate cultures in each experiment.

FIG. 5.

Growth of L. pneumophila in U937 human monocytic cells. Differentiated U937 cells (1 × 10⁶ cells/well) were infected with approximately 1 × 10⁶ bacteria/well for 1 h. The data shown are representative of those from three experiments, which showed similar results. Error bars indicate the standard errors of the mean. The growth rates of 80-045 (squares) and LAM0101 (circles) at 0, 24, and 48 h after infection showed no significant difference.

FIG. 6.

Elimination of L. pneumophila from the lungs of intranasally inoculated mice. A/J mice were inoculated with 10⁶ cells of L. pneumophila. In each experiment, three mice were used for the test at the 1-h infection point and four mice each were used for the 24- and 48-h infection points. At the time points indicated, the mice were sacrificed and the number of L. pneumophila in the lungs was determined as described in Materials and Methods. The number of bacteria at 1 h postinoculation was set at 1, and the numbers at the 24- and 48-h time points are presented as the relative ratio. Shown here are the results for the wild isolate 80-045 (squares), the laiA mutant LAM0101 (circles), and the laiA-complemented strain LAM0102 (triangles). Results indicate the means ± standard deviations of the bacterial counts at the indicated time point. The numbers of LAM0101 bacteria at the 24- and 48-h time points are significantly less than those of 80-045 and LAM0102. Three independent experiments showed similar results.