Enhanced lung injury and delayed clearance of Pneumocystis carinii in surfactant protein A-deficient mice: attenuation of cytokine responses and reactive oxygen-nitrogen species

Abstract

Surfactant protein A (SP-A), a member of the collectin family, selectively binds to Pneumocystis carinii and mediates interactions between pathogen and host alveolar macrophages in vitro. To test the hypothesis that mice lacking SP-A have delayed clearance of Pneumocystis organisms and enhanced lung injury, wild-type C57BL/6 (WT) and SP-A-deficient mice (SP-A(-/-)) with or without selective CD4(+)-T-cell depletion were intratracheally inoculated with Pneumocystis organisms. Four weeks later, CD4-depleted SP-A-deficient mice had developed a more severe Pneumocystis infection than CD4-depleted WT (P. carinii pneumonia [PCP] scores of 3 versus 2, respectively). Whereas all non-CD4-depleted WT mice were free of PCP, intact SP-A(-/-) mice also had evidence of increased organism burden. Pneumocystis infection in SP-A-deficient mice was associated histologically with enhanced peribronchial and/or perivascular cellularity (score of 4 versus 2, SP-A(-/-) versus C57BL/6 mice, respectively) and a corresponding increase in bronchoalveolar lavage (BAL) cell counts. Increases in SP-D content, gamma interferon, interleukin-4, interleukin-5, and tumor necrosis factor alpha in BAL fluid occurred but were attenuated in PCP-infected SP-A(-/-) mice compared to WT mice. There were increases in total BAL NO levels in both infected groups, but nitrite levels were higher in SP-A(-/-) mice, indicating a reduction in production of higher oxides of nitrogen that was also reflected in lower levels of 3-nitrotyrosine staining in the SP-A(-/-) group. We conclude that despite increases in inflammatory cells, SP-A-deficient mice infected with P. carinii exhibit an enhanced susceptibility to the organism and attenuated production of proinflammatory cytokines and reactive oxygen-nitrogen species. These data support the concept that SP-A is a local effector molecule in the lung host defense against P. carinii in vivo.

FIG. 1.

CD4-depleted SP-A-deficient mice demonstrate delayed clearance of Pneumocystis. The intensity of the infection was determined after intratracheal inoculation of P. carinii in CD4 T-cell-depleted WT and SP-A-deficient mice. Histological sections of lung from each mouse were scored blindly for intensity of infection by using a scale previously described and validated for Pneumocystis. Bars: □, uninfected mice; ▪, P. carinii-infected mice. Bars represent median values, with six samples in each group. ✽, Significant difference from the corresponding uninfected level (P < 0.05); #, significant difference from the corresponding WT level (P < 0.05 [Kruskal-Wallis test]).

FIG. 2.

P. carinii-infected SP-A-deficient mice demonstrate increased lung tissue inflammation. (A) Representative morphological changes in formalin-fixed, paraffin-embedded, hematoxylin-and-eosin-stained left lung sections of CD4-depleted WT and SP-A-deficient mice. Uninfected (upper panels) or P. carinii-infected (lower panels) WT and SP-A-deficient mice previously treated with GK1.5 were inoculated with sterile lung homogenate or 2 × 10⁵ Pneumocystis organisms and sacrificed 4 weeks postinfection as labeled. Original magnification: × 200. (B) Intensity of inflammation in the lung tissue CD 4 depleted WT and SP-A-deficient mice inoculated with either uninfected lung homogenate or 2 × 10⁵ Pneumocystis organisms. Histologic sections of lungs from each of the mice obtained 4 weeks postinoculation were scored blindly for the intensity of inflammation by using a scale previously described and validated (53). Bars: □, uninfected mice; ▪, P. carinii-infected mice. Bars represent median values, with six samples in each group. ✽, Significant difference from the corresponding uninfected level (P < 0.05); #, significant difference from the corresponding WT level (P < 0.05 [Kruskal-Wallis test]).

FIG. 3.

P. carinii-infected SP-A-deficient mice develop increased BAL cellularity. WT and SP-A-deficient mice previously depleted of CD4 T cells by using GK1.5 were inoculated with sterile lung homogenate or 2 × 10⁵ Pneumocystis organisms and then sacrificed 4 weeks postinfection. (A) Total numbers of BAL cells were derived from counts of stained cytospin preparations as described in Materials and Methods. Bars: □, uninfected mice; ▪, P. carinii-infected mice. The data are expressed as means ± the SEM absolute numbers of BAL cells per mouse lung (n = 6 in each group). ✽, Significant difference from the corresponding uninfected level (P < 0.05); #, significant difference from the corresponding WT level (P < 0.05). (B) Differential cell counts (MP, macrophage; EP, eosinophil; NP, neutrophil; LC, lymphocyte) in each BAL sample was determined by Coulter counting as described in Materials and Methods. Bars: □, WT mice; ▪, P. carinii-infected WT mice; ▩, SP-A-deficient mice; ▨, P. carinii-infected SP-A-deficient mice. The data expressed as the means ± the SEM absolute numbers of 10⁵ BAL cells per mouse lung (n = 6 in each group). ✽, Significant difference from the corresponding uninfected level (P < 0.05); #, significant difference from the corresponding WT level (P < 0.05).

FIG. 4.

Lung injury by Pneumocystis is enhanced in SP-A-deficient mice. (A) Total protein content of BAL fractions of uninfected or P. carinii-infected CD4-depleted WT and SP-A-deficient mice were determined by the Bradford method as described in the text. The data are expressed as mean amounts ± the SEM (in micrograms) of total protein per mouse lung (n = 3 to 6 in each group). Bars: □, uninfected mice; ▪, P. carinii-infected mice. ✽, Significant difference from corresponding uninfected level (P < 0.05); #, significant difference from corresponding WT level (P < 0.05). (B) Wet weight of the right lung of each mouse was determined and normalized to total body weight with data expressed as mean ± the SEM of right lung wet weight to body weight ratio × 1,000 (n = 5 in each group). Bars: □, uninfected mice; ▪, P. carinii-infected mice. ✽, Significant difference from the corresponding uninfected level (P < 0.05); #, significant difference from the corresponding WT level (P < 0.05).

FIG. 5.

SP-D protein levels are attenuated in P. carinii-infected SP-A-deficient mice. WT and SP-A-deficient mice previously depleted of CD4 T cells by using GK1.5 were inoculated with sterile lung homogenate or 2 × 10⁵ Pneumocystis organisms and sacrificed at 4 weeks postinfection. Western blots normalized for total protein (10 μg of total protein/sample) of SA surfactants prepared from BAL of WT and SP-A-deficient mice were prepared with polyclonal antisera to SP-D and visualized by using enhanced chemiluminescence. Quantification of total SA SP-D content was performed by densitometric scanning of multiple blots as described in Materials and Methods. The data are expressed as a percentage of uninfected WT levels (n = 3 to 5 in each group). Bars: □, uninfected mice; ▪, P. carinii-infected mice. ✽, Significant difference from the corresponding uninfected level (P < 0.05); #, significant difference from the corresponding WT level (P < 0.05).

FIG. 6.

Cytokine levels in BAL fluid are attenuated in P. carinii-infected SP-A-deficient mice. WT and SP-A-deficient mice previously depleted of CD4 T cells by using GK1.5 were inoculated with sterile lung homogenate or 2 × 10⁵ Pneumocystis organisms and sacrificed 4 weeks postinfection. Cytokine levels in BAL fluid were determined by ELISA as described in Materials and Methods. The data are expressed as total picograms per mouse lung. Values are represented as means ± the SEM (n = 6 in each group). Bars: □, uninfected mice; ▪, P. carinii-infected mice. ✽, Significant difference from corresponding uninfected level (P < 0.05); #, significant difference from corresponding WT level (P < 0.05).

FIG. 7.

The production of nitric oxide metabolites is increased in P. carinii-infected SP-A-deficient mice. WT and SP-A-deficient mice previously depleted of CD4 T cells by using GK1.5 were inoculated with sterile lung homogenate or 2 × 10⁵ Pneumocystis organisms and sacrificed 4 weeks postinfection. BAL samples were analyzed by chemical reduction chemiluminescence for total nitrogen oxide (A) or nitrite (B) as described in Materials and Methods. The data are expressed as total nanomoles in BAL (n = 5 animals in each group). Values are means ± the SEM. Bars:, uninfected mice; ▪, P. carinii-infected mice. ✽, Significant difference from the corresponding uninfected level (P < 0.05); #, significant difference from the corresponding WT level (P < 0.05).

FIG. 8.

Nitrotyrosine production is attenuated in the lung tissue of P. carinii-infected SP-A-deficient mice. WT and SP-A-deficient mice previously depleted of CD4 T cells by using GK1.5 were inoculated with sterile lung homogenate or 2 × 10⁵ Pneumocystis organisms and sacrificed 4 weeks postinfection. Paraffin-embedded sections of lung tissue from uninfected (upper panels) and P. carinii-infected (lower panels) mice are shown. Staining with monoclonal anti-nitrotyrosine antibody was performed as described in Materials and Methods. The specificity of staining was confirmed by using parallel sections treated with dithionite to remove nitrotyrosine (not shown). Images shown are representative of five WT and five SP-A-deficient mice in each group. Magnification, ×200.