Cryptococcal heat shock protein 70 homolog Ssa1 contributes to pulmonary expansion of Cryptococcus neoformans during the afferent phase of the immune response by promoting macrophage M2 polarization

Abstract

Numerous virulence factors expressed by Cryptococcus neoformans modulate host defenses by promoting nonprotective Th2-biased adaptive immune responses. Prior studies demonstrate that the heat shock protein 70 homolog, Ssa1, significantly contributes to serotype D C. neoformans virulence through the induction of laccase, a Th2-skewing and CNS tropic factor. In the present study, we sought to determine whether Ssa1 modulates host defenses in mice infected with a highly virulent serotype A strain of C. neoformans (H99). To investigate this, we assessed pulmonary fungal growth, CNS dissemination, and survival in mice infected with either H99, an SSA1-deleted H99 strain (Δssa1), and a complement strain with restored SSA1 expression (Δssa1::SSA1). Mice infected with the Δssa1 strain displayed substantial reductions in lung fungal burden during the innate phase (days 3 and 7) of the host response, whereas less pronounced reductions were observed during the adaptive phase (day 14) and mouse survival increased only by 5 d. Surprisingly, laccase activity assays revealed that Δssa1 was not laccase deficient, demonstrating that H99 does not require Ssa1 for laccase expression, which explains the CNS tropism we still observed in the Ssa1-deficient strain. Lastly, our immunophenotyping studies showed that Ssa1 directly promotes early M2 skewing of lung mononuclear phagocytes during the innate phase, but not the adaptive phase, of the immune response. We conclude that Ssa1's virulence mechanism in H99 is distinct and laccase-independent. Ssa1 directly interferes with early macrophage polarization, limiting innate control of C. neoformans, but ultimately has no effect on cryptococcal control by adaptive immunity.

Figure 1

Cryptococcal SSA1 gene deletion delays fungal growth in the lungs and the CNS dissemination, and improves survival of infected mice. (A) Mice were infected with 10⁴ cells of wild type (WT) C. neoformans H99 (serotype A), SSA1-gene deleted mutant Δssa1, or complement strain Δssa1::SSA1 on an H99 background. Pulmonary fungal burdens were quantified on 3, 7, or 14 days post infection (dpi). Mice infected with Δssa1 had significantly reduced pulmonary fungal burden relative to H99 at 3, 7, and 14 dpi and had significantly reduced pulmonary fungal burden relative to Δssa1::SSA1 at 3 and 7dpi. Note the faster growth rates of the WT and complement strains at 0–7 dpi, and parallel growth rates for all 3 strains beyond day 7. Data represent pooled average CFU per lung or brain from separate matched experiments; n=13 mice per group per time point for H99 and Δssa1 and n=4 for Δssa1::SSA1. (B) Survival study was performed on 7–8 mice per group. Infection with Δssa1 conferred a modest but significant survival advantage compared to H99 and Δssa1::SSA1. (C, D) Post-mortem CFU assay was performed on homogenized lung (C) and brain (D) from mice in the survival study. Pulmonary fungal burden did not differ significantly between any strains at the time of death, while fungal burden in the brain was significantly decreased in Δssa1-infected mice. († p < 0.01 relative to H99, * p < 0.05 relative to Δssa1::SSA1, ** p < 0.01 relative to Δssa1::SSA1).

Figure 2

Cryptococcal SSA1 deletion suppresses laccase activity in a serotype D JEC21-Δssa1 mutant but not in a serotype A Δssa1 mutant. Laccase activity was determined by melanization of C. neoformans colonies cultured for 48 h in YPD media at 37°C, then plated on asparagine salts agar with norepinephrine, and incubated at room temperature for 5 days. Note strong melanin pigmentation displayed by WT serotype A strain H99, the Δssa1 mutant in an H99 background, and the corresponding Δssa1::SSA1 complemented serotype A, and the absence of pigmentation in the laccase-deficient mutant (Δlac) in an H99 background. Also note the pigmentation in the serotype D WT strain JEC21 and corresponding Δssa1::SSA1 but the minimal pigmentation displayed by the Δssa1 strain on JEC21 background.

Figure 3

Deletion of cryptococcal SSA1 results in delayed but similar type lung and CNS pathology. Lungs and brains of mice infected with H99 (A, C, respectively) and Δssa1 (B, D, respectively) strains were isolated at week 3. Tissues were preserved in formalin, processed for histology, and cut and stained with mucicarmine to visualize C. neoformans. Lungs were counterstained with H&E. Representative images taken at 20x magnification with 100x magnification insets for lung histology images. A and B) Inflammatory infiltrates (black arrows) are present in both H99 and Δssa1 strains. Note the cryptococci (red arrows, 100x) residing in the alveolar space and the presence of inflammatory cell infiltrates in H99- and Δssa1-infected lungs. C and D) Cryptococcal mucoid cysts infiltrating and displacing brain tissue. Note the infiltrating inflammatory cells at the periphery and cocci and cell debris in the middle of the cysts (blue arrows), where normal brain tissue has eroded (yellow arrows) in both H99 and Δssa1 strains.

Figure 4

Deletion of cryptococcal SSA1 resulted in a similar magnitude but distinct polarization of early but not late inflammatory responses in the infected lungs. Lung leukocytes were isolated and characterized by microscopy at 3, 7, and 14 dpi from C. neoformans-infected lungs. Total leukocyte numbers in the lungs were equivalent in H99 versus Δssa1-infected lungs with the exception of a small but significant decrease on day 7 in the Δssa1-infected lungs (A). Monocytes (B), eosinophils (C), neutrophils (D) and lymphocytes (E) were identified by morphology and staining characteristics. Monocyte frequency was significantly increased at 3 and 7 dpi in Δssa1-infected mice, while eosinophil and neutrophil frequency was significantly decreased at 7 dpi in Δssa1-infected mice relative to H99-infected mice, and lymphocyte frequency never differed significantly between the two strains at any time points. N=5–11 mice per time point per strain, representing 3 separate matched experiments. *p < 0.05 compared to H99.

Figure 5

Cryptococcal SSA1 expression promotes the induction of crucial M2 proteins ARG1 and CD206 while preventing induction of M1 protein iNOS in the lungs. Defrosted lung sections taken at 8 dpi were stained with DAPI (blue) to show nuclei, and antibodies for M1 activation marker iNOS (left panel), M2 activation marker Arginase (Arg1, middle panel) and M2 activation marker mannose receptor (CD206, right panel). Secondary FITC conjugated antibodies (green) were used to visualize immunoreactive proteins. Note the significant induction of Arg1 and CD206 and the absence of iNOS staining in H99-infected lung sections. In contrast, iNOS protein can be detected in Δssa1-infected sections, while the M2 markers Arg1 and CD206 show strikingly less immunoreactivity in Δssa1-infected lungs. Sections were imaged at 40X magnification.

Figure 6

Cryptococcal Ssa1 increases surface expression of M2 activation markers and decreases surface expression of MHCII during the innate but not adaptive phase of the immune response in the infected lungs. Flow cytometric analysis of leukocyte populations from infected mouse lungs used the gating scheme in (A) to sequentially gate out debris, non-immune cells, lymphocytes, and granulocytes and then we selected the CD11c⁺ subset of this population. M1 activation marker MHC class II (B) and M2 activation markers Gal3 (C) and CD206 (D) surface expression was assessed at 3 and 14 dpi. Note the significant increase in MHCII and decrease in CD206 and Gal3 surface staining in Δssa1-infected mice at 3 dpi, but a resolution of these differences by 14 dpi. Representative histograms were selected from two separate, matched experiments. N = 5 per treatment per time point. Combined mean fluorescence intensity plots represent high-expressing cells (MHC II) or positively staining cells (CD206 and Gal3), and reflect the cumulative phenotype. * p < 0.05 using Student’s t-test.

Figure 7

Cryptococcal Ssa1 increases induction of macrophage M2 activation markers and decreases induction of M1 activation marker iNOS during the early but not the late phase of lung infection in mice. Macrophages were isolated from infected mouse lung leukocyte preparations by adherence selection. iNOS (A), Arg1 (B), the iNOS/Arg1 expression ratio (C), CD206 (D) and Gal3 (E) mRNA transcript levels were assessed by qPCR. Note the increase in iNOS in Δssa1-infected mouse macrophages relative to H99-infected mouse macrophages at day 3 but not at day 14. Arg1 is not significantly different between the macrophages at day 3. However, a paired iNOS/Arg1 ratio (C) where an individual mouse’s macrophage iNOS mRNA fold change was divided by its own Arg1 mRNA fold change yielded highly significant M1-skewed signature in macrophages from Δssa-infected mice but a far lower ratio in H99-infected mouse macrophages at day 3, but not at day 14, where both Δssa1- and H99-infected mice had M2-skewed macrophages. Also note the decrease in expression of both M2 activation markers CD206 and Gal3 in Δssa1- infected mice relative to H99-infected mice at day 3, but not at day 14. N = 8 from 2 separate experiments. * p < 0.05, ** p < 0.01 using Student’s t-test.

Figure 8

Cryptococcal SSA1 expression results in increased pro-Th2 cytokine induction and increased IL4:IFNγ regulation ratio in pulmonary leukocytes from infected mice during innate phase of the immune response to C. neoformans. Leukocyte preparations from H99 and Δssa1- infected mouse lungs were processed for mRNA quantification of IL-13 (A), IL-4 (B), IFNγ (C) and for the IL-4:IFNγ induction ratio (D) at 3 and 14 dpi. Note the decrease in non-protective cytokines IL-4 and IL-13, along with the decreased IL-4:IFNγ induction ratio in the leukocytes obtained from Δssa1-infected mice relative to the H99-infected mice at day 3, followed by progressive but no longer differential cytokine induction in the lungs of Δssa1- and H99-infected mice on day 14. N = 5/time point/group from 2 separate experiments. * p < 0.05 using Student’s t-test.

Figure 9

In vitro stimulation with an SSA1-expressing cryptococcus is sufficient to decrease M1 and increase M2 activation markers in bone marrow derived macrophages. mRNA was extracted from bone marrow-derived macrophages stimulated with live H99 or Δssa1 C. neo for 24 hours and qPCR was performed for M1 activation marker iNOS (A) and M2 activation markers Arg1 (A), CD206 (B) and Gal3 (B). iNOS/Arg1 expression ratio (A) was also prepared. Note the increase in iNOS expression and iNOS/Arg1 ratio and the decreases in Arg1, CD206, and Gal3 expression in the Δssa1-infected macrophages relative to the H99-infected macrophages. N=6–9 from three separate experiments. * p < 0.05, ** p < 0.01 using Student’s t-test.