The type II secretion system of Legionella pneumophila elaborates two aminopeptidases, as well as a metalloprotease that contributes to differential infection among protozoan hosts

Abstract

Legionella pneumophila, the agent of Legionnaires' disease, is an intracellular parasite of aquatic amoebae and human macrophages. A key factor for L. pneumophila in intracellular infection is its type II protein secretion system (Lsp). In order to more completely define Lsp output, we recently performed a proteomic analysis of culture supernatants. Based upon the predictions of that analysis, we found that L. pneumophila secretes two distinct aminopeptidase activities encoded by the genes lapA and lapB. Whereas lapA conferred activity against leucine, phenylalanine, and tyrosine aminopeptides, lapB was linked to the cleavage of lysine- and arginine-containing substrates. To assess the role of secreted aminopeptidases in intracellular infection, we examined the relative abilities of lapA and lapB mutants to infect human U937 cell macrophages as well as Hartmannella vermiformis and Acanthamoeba castellanii amoebae. Although these experiments identified a dispensable role for LapA and LapB, they uncovered a previously unrecognized role for the type II-dependent ProA (MspA) metalloprotease. Whereas proA mutants were not defective for macrophage or A. castellanii infection, they (but not their complemented derivatives) were impaired for growth upon coculture with H. vermiformis. Thus, ProA represents the first type II effector implicated in an intracellular infection event. Furthermore, proA represents an L. pneumophila gene that shows differential importance among protozoan infection models, suggesting that the legionellae might have evolved some of its factors to especially target certain of their protozoan hosts.

FIG. 1.

Secreted aminopeptidase activities of the wild type versus lapA and lapB mutants of L. pneumophila. Cell-free supernatants obtained from late log cultures of wild-type 130b (black bars), lapA mutant NU320 (white bars), lapB mutant NU322 (gray bars), and lapA lapB double mutant NU324 (hatched bars) were tested for their ability to cleave Leu-pNI (A), Phe-βNA (B), Tyr-βNA (C), Arg-βNA (D), Lys-pNI (E), and Ala-βNA (F). The values presented are the means and standard deviations obtained from triplicate samples and are representative of at least four independent experiments. The reduced levels of activity seen for the lapA mutants (A to C) and the lapB mutants (D and E) were statistically significant (P < 0.05; Student's t test).

FIG. 2.

Secreted aminopeptidase activities of the wild type versus an lspF mutant of L. pneumophila. Cell-free supernatants obtained from late log cultures of wild-type 130b (black boxes) and lspF mutant NU275 (white bars) were tested, as indicated, for their ability to cleave Leu-pNI, Lys-pNI, and Ala-βNA. The values presented are the means and standard deviations obtained from triplicate samples and are representative of at least four independent experiments. The lspF mutant's reduced levels of activity against Leu-pNI and Lys-pNI were statistically significant (P < 0.05; Student's t test).

FIG. 3.

Intracellular infection of amoebae and macrophages by wild-type and aminopeptidase mutants of L. pneumophila. H. vermiformis (A), U937 cells (B), and A. castellanii (C) were infected with wild-type 130b (⧫), lapA mutant NU320 (○), lapB mutant NU322 (•), or lapA lapB double mutant NU324 (□) and then, at various times postinoculation, the numbers of bacteria per well were determined. The values presented are the means and standard deviations obtained from four (A) or three (B and C) infected wells and are representative of at least two independent experiments.

FIG. 4.

Secreted aminopeptidase activities of the wild type versus a proA mutant of L. pneumophila. Cell-free supernatants obtained from late log cultures of wild-type 130b (black boxes) and proA mutant AA200 (white bars) were tested, as indicated, for their ability to cleave Phe-βNA, Tyr-βNA, Ala-βNA, and Leu-pNI (A) or Arg-βNA and Lys-pNI (B). The values presented are the means and standard deviations obtained from triplicate samples and are representative of at least three (A) or four (B) independent experiments. Mutant AA200's reduced levels of activity against Arg-βNA and Lys-pNI (B) were statistically significant (P < 0.05; Student's t test).

FIG. 5.

Intracellular infection of A. castellanii by the wild type, an lspF mutant, and a proA mutant of L. pneumophila. Acanthamoebae, cultured in standard proteose peptone-yeast extract-glucose medium (A) or 1034 medium (B), were infected with wild-type 130b (⧫), lspF mutant NU275 (•), and proA mutant AA200 (○) and then, at various times postinoculation, the numbers of bacteria per well were determined. The values presented are the means and standard deviations obtained from three infected wells and are representative of at least two independent experiments. At 48 to 96 h postinoculation, and under both growth conditions, the recovery of the lspF mutant was significantly less than that of the wild type and the other mutant (P < 0.05; Student's t test). In contrast, the slightly reduced recoveries of the proA mutant relative to the wild type seen at some points were not significant.

FIG. 6.

Intracellular infection of H. vermiformis by the wild type and metalloprotease mutants of L. pneumophila. Hartmannellae in 1034 medium were infected and then, at various times postinoculation, the numbers of bacteria per well were determined. The values presented are the means and standard deviations obtained from four infected wells. (A) Infection comparison between wild-type 130b (⧫), lspF mutant NU275 (•), and proA mutant AA200 (○). At 48 to 96 h postinfection, significant differences in bacterial recovery were obtained between the wild type and both of the mutants (P < 0.01; Student's t test). (B) Infection comparison between 130b(pMMB2002) (⧫), 130b(pMproA) (⋄), AA200(pMMB2002) (○), and AA200(pMproA) (•). pMMB2002 is the vector into which proA was cloned for the purposes of this complementation experiment. At 72 h, significant differences were obtained between AA200(pMMB2002) and all other strains (P < 0.01). The experiments presented here are representative of at least six (A) or three (B) other independent experiments. The proA mutant was similarly defective in an additional experiment that used a multiplicity of infection equal to 1.0 (versus 0.1).