The fatal fungal outbreak on Vancouver Island is characterized by enhanced intracellular parasitism driven by mitochondrial regulation

Abstract

In 1999, the population of Vancouver Island, Canada, began to experience an outbreak of a fatal fungal disease caused by a highly virulent lineage of Cryptococcus gattii. This organism has recently spread to the Canadian mainland and Pacific Northwest, but the molecular cause of the outbreak remains unknown. Here we show that the Vancouver Island outbreak (VIO) isolates have dramatically increased their ability to replicate within macrophages of the mammalian immune system in comparison with other C. gattii strains. We further demonstrate that such enhanced intracellular parasitism is directly linked to virulence in a murine model of cryptococcosis, suggesting that this phenotype may be the cause of the outbreak. Finally, microarray studies on 24 C. gattii strains reveals that the hypervirulence of the VIO isolates is characterized by the up-regulation of a large group of genes, many of which are encoded by mitochondrial genome or associated with mitochondrial activities. This expression profile correlates with an unusual mitochondrial morphology exhibited by the VIO strains after phagocytosis. Our data thus demonstrate that the intracellular parasitism of macrophages is a key driver of a human disease outbreak, a finding that has significant implications for a wide range of other human pathogens.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.

(A) Significant inter-strain variation in IPR occurs within the C. gattii species (n = 39). Black bars represent VIO isolates, which proliferate far better than other C. gattii strains (gray bars). The only exception was A1M-R272, which belongs to the minor (AFLP6B) group of the outbreak and has previously been shown to be less virulent than other VIO strains (12). An asterisk (*) denotes strains used for the microarray study. (B) IPR values of 8 strains in human primary blood-derived macrophage cells (HPBMCs) correlate significantly with those observed in J774 (P = 0.00044, Spearman's test, n = 8). The IPR values obtained with HPBMC are generally lower than those seen in J774 macrophages, which may be a result of the higher cryptococcal expulsion rates observed in primary cells, as previously reported (36).

Fig. 2.

A significant correlation between mouse survival data [both previously published in ref. (■) and newly generated (♦)] and intracellular proliferation rate (IPR). Mice survived longer when infected with strains with low IPR values as compared with animals infected with high IPR strains (P = 0.00017, linear regression, n = 18). For the unpublished mouse survival assays, experiments were conducted as described in the Materials and Methods section. Published mouse survival data are taken from Fraser et al. (12). Note that strain WM276 has been omitted from the graph, since published ST50 values on this strain vary widely for reasons that are not currently understood (12, 45). For strains that did not cause mortality of 50% or more within the 45-day timeframe of the experiment, we arbitrarily assigned an ST50 of 55 days.

Fig. 3.

(A) A1M-R265 Supercontig 25 is highly over-represented in the transcriptional analysis (P = 10−24, χ2 test). Probes (1,367) whose expression showed significant correlation with IPR were mapped onto the 28 supercontigs of the A1M-R265 genome and compared to the expected hit rate (assuming a random probe distribution) using the χ2 test. Supercontig 25 is 10-fold over-represented. (B) The predicted genome structure of A1M-R265 supercontig 25. The genome is 34.7 kb in size and, as with other cryptococcal mitochondria, encodes 17 genes important for mitochondrial function and protein synthesis. The ORFs were predicted using Open Reading Frame Finder at NCBI (http://www.ncbi.nlm.nih.gov/gorf/gorf.html) in combination with alignment to the H99 mitochondrial sequence (available at http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db = nucleotide&val = AY101381). The supercontig represents the whole mitochondrial genome of A1M-R265. Sections with light blue color are either introns or intergenic spaces, both of which are substantially expanded relative to C. neoformans H99.

Fig. 4.

Vancouver Island Outbreak isolates acquire a characteristic mitochondrial morphology after intracellular growth. (A) Representative images of the 3 different mitochondrial morphologies observed (diffuse, globular, and tubular) and a table showing the percentage of intracellular yeast cells with tubular mitochondria in 6 AFLP6 strains (3 VIO and 3 non-VIO isolates). For each strain, a random selection of cryptococcal cells were scored blindly for the 3 different mitochondrial morphologies. Mitochondria with a tubular morphology were found only rarely in strains with low IPR values (non-VIO strains) or in strains (both VIO and non-VIO) that had been grown extracellularly. (B) The percentage of intracellular yeast exhibiting a tubular mitochondrial morphology correlates significantly with IPR value (n = 6, P = 0.00021, linear regression). (C) A Z-projection confocal image showing the tubular mitochondrial morphology of a VIO strain. C. gattii strain ENV152 (IPR = 2.28), isolated 24 h of growth within J774 macrophages and labeled with MitoTracker.