Legionella pneumophila catalase-peroxidases are required for proper trafficking and growth in primary macrophages

Abstract

Legionella pneumophila, a parasite of aquatic amoebae and pathogen of pulmonary macrophages, replicates intracellularly, utilizing a type IV secretion system to subvert the trafficking of Legionella-containing phagosomes. Defense against host-derived reactive oxygen species has been proposed as critical for intracellular replication. Virulence traits of null mutants in katA and katB, encoding the two Legionella catalase-peroxidases, were analyzed to evaluate the hypothesis that L. pneumophila must decompose hydrogen peroxide to establish a replication niche in macrophages. Phagosomes containing katA or katB mutant Legionella colocalize with LAMP-1, a late endosomal-lysosomal marker, at twice the frequency of those of wild-type strain JR32 and show a decreased frequency of bacterial replication, in similarity to phenotypes of mutants with mutations in dotA and dotB, encoding components of the Type IV secretion system. Quantitative similarity of the katA/B phenotypes indicates that each contributes to virulence traits largely independently of intracellular compartmentalization (KatA in the periplasm and KatB in the cytosol). These data support a model in which KatA and KatB maintain a critically low level of H(2)O(2) compatible with proper phagosome trafficking mediated by the type IV secretion apparatus. During these studies, we observed that dotA and dotB mutations in wild-type strain Lp02 had no effect on intracellular multiplication in the amoeba Acanthamoeba castellanii, indicating that certain dotA/B functions in Lp02 are dispensable in that experimental model. We also observed that wild-type JR32, unlike Lp02, shows minimal contact-dependent cytotoxicity, suggesting that cytotoxicity of JR32 is not a prerequisite for formation of replication-competent Legionella phagosomes in macrophages.

FIG. 1.

Sodium resistance. Titers of PE cultures in AYE (25D, JR32, katA, and katB) or in AYET (Lp02, dotA, and dotB) broth were determined on CYE or CYET, respectively, without and with 0.1 M NaCl. The percentages of resistance (± standard deviation) were calculated by comparing the plating efficiency with 0.1 M NaCl to that without 0.1 M NaCl. The P value for a one-sided t test comparing each mutant strain with the corresponding wild-type strain is less than 3 × 10⁻⁵.

FIG. 2.

Intracellular growth and infection efficiency in bone marrow-derived macrophages. (A) Intracellular growth. Adherent macrophages were infected with PE cultures at an MOI of 1.0 for 2 h and then washed to remove extracellular bacteria. At the indicated times, macrophages were lysed and titers were determined (± standard deviation) on CYE. (B) Infection efficiency. Macrophages were infected with PE (open bars) or E (filled bars) cultures of the indicated strains at an MOI of 1.0 for 2 h. External bacteria were removed by washing the monolayers. The percentages of internalization (± standard deviation) were calculated from the titers of bacteria as determined after lysis of the macrophages compared to the number of bacteria in the infection medium. Above the error bar is a symbol representing a P value from a one-sided t test comparing PE strains (open bars) with wild-type strain JR32, and below the x axis are symbols representing P values for comparisons of E and PE cultures of each strain. ∗, P < 0.05; ++, P ≤ 0.02; ∗∗, P ≤ 0.003.

FIG. 3.

Frequency of replication of individual L. pneumophila cells in bone marrow-derived macrophages. Macrophages were infected with PE cultures of L. pneumophila. After 18 h, the number of bacteria per vacuole in infected macrophages was counted and scored (± standard deviation) as belonging to one of the four categories indicated below the x axis. P values from a one-sided t test comparing mutant strains with the wild-type strain are indicated as follows: ∗, P ≤ 0.05; ++, P < 0.01. For comparisons of katA and katB strains at >20 bacteria per vacuole, P = 0.04.

FIG. 4.

Colocalization of L. pneumophila with endosomal markers. Macrophages on coverslips were incubated with PE bacteria of the indicated strains. At 2 h after infection, Legionella-containing phagosomes were identified by immunofluorescence and the percentages (± standard deviation) of those which were also positive for LAMP-1 (A) or TRov (B) were determined. For strain Lp02 in panel B, results for PE cultures are indicated by the rightmost open bar and results for E (Exp) cultures are indicated by the rightmost filled bar. P values for a one-sided t test comparing mutant strains with the corresponding wild-type strain or (in panel B only) comparing E and PE cultures are indicated as follows: ∗, P ≤ 0.05; ++, P < 0.02; +++, P < 0.00015.

FIG. 5.

Intracellular growth in A. castellanii. A. castellanii was infected at an MOI of 1.5 to 3.0 with the cultures of indicated PE L. pneumophila strains, and titers (± standard deviation) of bacteria released into the medium were determined on CYE (strains JR32, katA, and 25D [A] and strains JR32 and katB [B]) or CYET (strains Lp02, dotA, and dotB [C]).

FIG. 5.

Intracellular growth in A. castellanii. A. castellanii was infected at an MOI of 1.5 to 3.0 with the cultures of indicated PE L. pneumophila strains, and titers (± standard deviation) of bacteria released into the medium were determined on CYE (strains JR32, katA, and 25D [A] and strains JR32 and katB [B]) or CYET (strains Lp02, dotA, and dotB [C]).

FIG. 5.

Intracellular growth in A. castellanii. A. castellanii was infected at an MOI of 1.5 to 3.0 with the cultures of indicated PE L. pneumophila strains, and titers (± standard deviation) of bacteria released into the medium were determined on CYE (strains JR32, katA, and 25D [A] and strains JR32 and katB [B]) or CYET (strains Lp02, dotA, and dotB [C]).

FIG. 6.

Cytotoxicity of L. pneumophila to bone marrow-derived macrophages. Macrophages were incubated for 1 h with PE bacteria of the indicated Legionella strain at each of the MOI values shown. The percentages of viable macrophages (± standard deviation) were determined by the Alamar Blue method.