MyD88-dependent responses involving toll-like receptor 2 are important for protection and clearance of Legionella pneumophila in a mouse model of Legionnaires' disease

Abstract

Legionella pneumophila is a gram-negative facultative intracellular parasite of macrophages. Although L. pneumophila is the causative agent of a severe pneumonia known as Legionnaires' disease, it is likely that most infections caused by this organism are cleared by the host innate immune system. It is predicted that host pattern recognition proteins belonging to the Toll-like receptor (TLR) family are involved in the protective innate immune responses. We examined the role of TLR-mediated responses in L. pneumophila detection and clearance using genetically altered mouse hosts in which the macrophages are permissive for L. pneumophila intracellular replication. Our data demonstrate that cytokine production by bone marrow-derived macrophages (BMMs) in response to L. pneumophila infection requires the TLR adapter protein MyD88 and is reduced in the absence of TLR2 but not in the absence of TLR4. Bacterial growth ex vivo in BMMs from MyD88-deficient mice was not enhanced compared to bacterial growth ex vivo in BMMs from heterozygous littermate controls. Wild-type mice were able to clear L. pneumophila from the lung, whereas respiratory infection of MyD88-deficient mice caused death that resulted from robust bacterial replication and dissemination. In contrast to an infection with virulent L. pneumophila, MyD88-deficient mice were able to clear infections with L. pneumophila dotA mutants, indicating that MyD88-independent responses in the lung are sufficient to clear bacteria that are unable to replicate intracellularly. In vivo growth of L. pneumophila was enhanced in the lungs of TLR2-deficient mice, which resulted in a delay in bacterial clearance. No significant differences were observed in the growth and clearance of L. pneumophila in the lungs of TLR4-deficient mice and heterozygous littermate control mice. Our data indicate that MyD88 is crucial for eliciting a protective innate immune response against virulent L. pneumophila and that TLR2 is one of the pattern recognition receptors involved in initiating this MyD88-dependent response.

FIG. 1.

MyD88 is required for cytokine production by macrophages in response to L. pneumophila. BMMs from A/J mice, MyD88-deficient mice (MyD88^−/−), and heterozygous littermates (MyD88^+/−) were incubated with L. pneumophila. Supernatants were collected 24 h postinfection, and IL-12 p40 (A), IL-6 (B), and TNF-α (C) levels were determined by ELISA. Each point represents data for BMMs derived from a different mouse. All data points represent the average cytokine concentration determined from three wells infected independently. Each line indicates the mean calculated from the data for the three different mice.

FIG. 2.

MyD88 signaling does not restrict L. pneumophila replication in macrophages ex vivo. BMMs from A/J mice, MyD88-deficient mice (MyD88^−/−), and heterozygous littermates (MyD88^+/−) were infected with L. pneumophila. The numbers of bacterial CFU were determined at 1-h postinfection and 72 h postinfection. Intracellular growth is expressed as the fold increase in the number of CFU detected over this period. Each point represents data for BMMs derived from a different mouse. All data points represent the average increase in the number of bacterial CFU determined from three wells infected independently. Each line indicates the mean calculated from the data for the three different mice.

FIG. 3.

MyD88 is required for clearance of L. pneumophila in vivo. MyD88-deficient mice (MyD88^−/−) and heterozygous littermates (MyD88^+/−) were infected with L. pneumophila intranasally. (A) Numbers of bacterial CFU in the lungs of infected mice at different times. Each data point represents the number of CFU for the lungs of a single mouse. The average numbers of CFU detected for the MyD88-deficient mice (dotted line) and heterozygous control mice (solid line) are also indicated. (B) Numbers of bacterial CFU in the spleens (solid bars) and livers (open bars) on day 9 postinfection for three MyD88-deficient mice. (C) IFN-γ levels in BAL fluid on day 2 postinfection for MyD88-deficient mice and heterozygous littermates. Each point represents data from a different mouse. The lines indicate the means calculated from the data for the two groups of mice.

FIG. 4.

L. pneumophila dotA mutants are cleared from the lungs by MyD88-independent responses. MyD88-deficient mice (MyD88^−/−) and heterozygous littermates (MyD88^+/−) were infected with an L. pneumophila dotA mutant intranasally. The numbers of bacterial CFU in the lungs of infected mice were determined at different times. Each data point represents the number of CFU in the lungs of a single mouse. The average numbers of CFU detected for the MyD88-deficient mice (dotted line) and heterozygous control mice (solid line) are also indicated.

FIG. 5.

TLR4-deficient BMMs have no apparent defect in the response to L. pneumophila. BMMs from A/J mice, TLR4-deficient mice (TLR4^−/−), and heterozygous littermates (TLR4^+/−) were incubated with L. pneumophila. Supernatants were collected 24 h postinfection, and the levels of IL-12 p40 (A), IL-6 (B), and TNF-α (C) were determined by ELISA. Each point represents data for BMMs derived from a different mouse. All data points represent the average cytokine concentration determined from three wells infected independently. Each line indicates the mean calculated from the data for the three different mice.

FIG. 6.

TLR4 is not important for pulmonary clearance of L. pneumophila. TLR4-deficient mice (TLR4^−/−) and heterozygous littermates (TLR4^+/−) were infected with L. pneumophila intranasally. The numbers of bacterial CFU in the lungs of infected mice were determined at different times. Each data point represents the number of CFU in the lungs of a single mouse. The average numbers of CFU detected for the TLR4-deficient mice (dotted line) and heterozygous control mice (solid line) are also indicated.

FIG. 7.

Macrophage responses to L. pneumophila involve TLR2. BMMs from A/J mice, TLR2-deficient mice (TLR2^−/−), and heterozygous littermates (TLR2^+/−) were incubated with L. pneumophila. Supernatants were collected 24 h postinfection, and the levels of IL-12 p40 (A), IL-6 (B), and TNF-α (C) were determined by ELISA. Each point represents data for BMMs derived from a different mouse. All data points represent the average cytokine concentration determined from three wells infected independently. Each line indicates the mean calculated from the data for the three different mice. For all three cytokines, the defect in cytokine production by the TLR2-deficient macrophages compared to the cytokine production by the heterozygous control macrophages was determined to be highly significant (P < 0.005).

FIG. 8.

TLR2-mediated responses restrict pulmonary replication of L. pneumophila. TLR2-deficient mice (TLR2^−/−) and heterozygous littermates (TLR2^+/−) were infected with L. pneumophila intranasally. The numbers of bacterial CFU in the lungs of infected mice were determined at different times. Each data point represents the number of CFU in the lungs of a single mouse. The average numbers of CFU detected for the MyD88-deficient mice (dotted line) and heterozygous control mice (solid line) are also indicated. (A) Mice were sacrificed in groups of 3 after infection, and the numbers of CFU in the lungs were determined. (B) Mice were sacrificed in groups of 6 on days 2 and 3 postinfection, and the numbers of CFU in the lungs were determined. For the TLR2-deficient mice statistically significant increases in the numbers of CFU were observed on day 2 postinfection (one asterisk; P < 0.05) and on day 3 postinfection (two asterisks; P < 0.005) compared to the numbers of CFU in the heterozygous control mice. (C) IFN-γ levels in BAL fluid on day 2 postinfection for TLR2-deficient mice and heterozygous littermates. Each point represents data from a different mouse. The lines indicate the means calculated from the data for the two groups of mice. The difference between the two groups of mice was not statistically significant (P > 0.05).