Loss of allergen 1 confers a hypervirulent phenotype that resembles mucoid switch variants of Cryptococcus neoformans

Abstract

Microbial survival in a host is usually dependent on the ability of a pathogen to undergo changes that promote escape from host defense mechanisms. The human-pathogenic fungus Cryptococcus neoformans undergoes phenotypic switching in vivo that promotes persistence in tissue. By microarray and real-time PCR analyses, the allergen 1 gene (ALL1) was found to be downregulated in the hypervirulent mucoid switch variant, both during logarithmic growth and during intracellular growth in macrophages. The ALL1 gene encodes a small cytoplasmic protein that is involved in capsule formation. Growth of an all1Delta gene deletion mutant was normal. Similar to cells of the mucoid switch variant, all1Delta cells produced a larger polysaccharide capsule than cells of the smooth parent and the complemented strain produced, and the enlarged capsule inhibited macrophage phagocytosis. The mutant exhibited a modest defect in capsule induction compared to all of the other variants. In animal models the phenotype of the all1Delta mutant mimicked the hypervirulent phenotype of the mucoid switch variant, which is characterized by decreased host survival and elevated intracranial pressure. Decreased survival is likely the result of both an ineffective cell-mediated immune response and impaired phagocytosis by macrophages. Consequently, we concluded that, unlike loss of most virulence-associated genes, where loss of gene function results in attenuated virulence, loss of the ALL1 gene enhances virulence by altering the host-pathogen interaction and thereby impairing clearance. Our data identified the first cryptococcal gene associated with elevated intracranial pressure and support the hypothesis that an environmental opportunistic pathogen has modified its virulence in vivo by epigenetic downregulation of gene function.

FIG. 1.

ALL1 expression in SM and MC variants and high levels of homology of the ALL1 protein in C. neoformans serotypes. (a) Comparison of ALL1 in SM and MC variants during intracellular growth in macrophages by using RT-PCR. In two different macrophage experiments (black bar and gray bar) ALL1 was more highly expressed at 2 and 8 h. Expression was measured relative to actin expression, and similar results were obtained with glyceraldehyde-3-phosphate dehydrogenase. Standard deviations were determined by using triplicate measurements for RT-PCR and are indicated by the error bars. (b) High levels of homology of the putative protein sequences in C. neoformans serotype D, A, and B strains.

FIG. 2.

Confirmation and characterization of the all1Δ mutant. (a) Plasmid construct used in this work to generate homologous recombinants. Van91I restriction sites on the individual primers permitted rapid directional cloning. (b) PCR amplification and Southern blotting demonstrated that there was correct homologous recombination in the mutant compared to the SM parent variant (Wt). (c) India ink preparations (magnification, ×1,000) demonstrating differences in capsule size in vitro (upper panel) and in vivo after induction (cells derived from lungs of mice on day 7 after i.t. infection) (lower panel) between the SM, MC, all1Δ, and all1Δ+ALL1 cells.

FIG. 3.

Cytoplasmic location of All1 protein. The Western blot is a Western blot of cytoplasmic protein (CP) extracts from SM and all1Δ cells (WP, whole-cell proteins). Membranes were stained with polyclonal serum to All1. Note that there is a smaller 25-kDa cross-reacting band. This is consistent with detection of the second homologous protein encoded by CNM02200. Examination of single optical sections by confocal microscopy showed that red staining and green staining occurred in different compartments of the cell. Green, intracellular staining of All1; red, nuclear staining with DAPI.

FIG. 4.

Virulence of the all1 mutant is enhanced. (a and b) Differences in survival of the SM variant-, all1Δ mutant-, all1Δ+ALL1 mutant-, and MC variant-infected mice in an i.p. and i.t. infection model. (c) Increased ICP in all1Δ mutant-infected rats compared to sham- and SM variant-infected rats. (d) Different ICP results are consistent with premature mortality in all1Δ mutant-infected rats.

FIG. 5.

Host response is modified in the absence of ALL1. (a and b) Hematoxylin- and eosin-stained and mucicarmine-stained lung tissue of SM variant-infected (a) and all1Δ mutant-infected (b) mice on day 10 after the mice were given 10⁶ cells i.t. Note the beginning of granuloma formation in the SM variant-infected mouse (a) and the extracellular accumulation of yeast cells in the alveolar spaces of the all1Δ mutant-infected mouse (b). At day 46 hematoxylin and eosin staining of lung tissue revealed an excessive but ineffective host response in the all1Δ mutant-infected mice (d), whereas the SM variant-infected mice (c) successfully cleared the infection. (e and f) Immunohistochemistry with a macrophage-specific MAb, MAC3, revealed significantly more macrophages in all1Δ mutant-infected lungs (f) than in SM variant-infected lungs (e). (g) The inflammatory response in an SM variant-infected mouse whose strain switched to a MC phenotype is similar to the inflammatory response in all1Δ mutant-infected mice. (h) Accumulation of yeast cells in the subarachnoid space in an all1Δ mutant-infected rat.

FIG. 6.

Phagocytosis and T-cell proliferation are affected by loss of ALL1. (a) The phagocytosis index is significantly (P < 0.05) lower in macrophages incubated with the all1Δ mutant and the MC variant than in macrophages incubated with the SM variant and the all1Δ+ALL1 mutant. The standard deviations for phagocytosis in three wells were determined by Student's t test and are indicated by the error bars. (b) Dose-dependent T-cell proliferation in lymphocytes incubated with All1. The test was done in triplicate, and the error bars indicate the standard deviations. The lymphocytes were isolated from popliteal lymph nodes of mice on day 7 after immunization with CFA and All1. The T-cell proliferation was comparable to concanavalin A (conA)-induced proliferation and was significantly greater (P < 0.05, as determined by Student's t test) than the proliferation of lymphocytes of control antigen (Ag) (protein from empty vector bacterial lysate)-immunized mice.