Distinguishing Features of C. gattii

C. gattii was easily and reliably differentiated from C. neoformans on creatinine dextrose bromthymol blue (CDB) medium. This work built on the discovery that C. neoformans can assimilate creatinine as sole source of carbon and nitrogen. Further modification in CDB medium led to development of canavanine-glycine-bromthymol blue agar, which has since become the differential medium of choice (1,11,12). Unfortunately, the medium is still not widely used in diagnostic laboratories, most likely because of limited availability from commercial suppliers.

Figure

Figure. Worldwide isolations of Cryptococcus gattii from human clinical, veterinary, and environmental sources. Circles indicate serotype information, diamonds indicate genotype information, and rectangles indicate hybrids between C. gattii and C. neoformans. Existing...

C. gattii populations can be distinguished by the pairing of unknown isolates with compatible tester strains to distinguish MATα from MATa strains. MATα is most prevalent clinically and environmentally, and MATa is recovered less frequently (7,13,14). Four distinct C. gattii molecular subtypes (VGI, VGII, VGIII, and VGIV) have been recognized by PCR amplification of genomic DNA by using bacteriophage M13 single-stranded primers. Genotypes VGI and VGII are prevalent worldwide, and VGIII and VGIV are less common (Figure). Serotypes B and C are randomly dispersed among the M13 molecular types. Further molecular subtypes are now known to exist within the 4 molecular types (15).

The VGI molecular genotype has been reported from many areas and from 2 C. neoformans–C. gattii hybrids reported from Canada and the Netherlands (14,16). VGII has been reported from the Western Hemisphere, Australasia, Asia, and Africa and is reportedly the most fertile and virulent of the strains responsible for infection in healthy and immunocompromised humans and animals. This strain also is associated with an ongoing outbreak of C. gattii cryptococcosis from Vancouver Island, British Columbia, Canada (14). VGIII, which contains both serotypes B and C, is most commonly reported from South America but also is reported from North America, Central America, Australasia, and Southern Asia. VGIV, frequently associated with serotype C, has been reported from Africa and South and Central America. Further experimental studies on VG genotypes are needed to explain possible connections between distribution of various genotypes and propensity to cause cryptococcosis among exposed hosts.

C. gattii in Clinical and Environmental Specimens

Initially, reports of C. gattii originated from human clinical samples in tropical and subtropical regions, including portions of Africa, Europe, Australia, the United States, and South America. This accounted for the long-held impression that C. gattii is a tropical or subtropical pathogen (4,17,18). More recent clinical isolations from temperate regions in the United States, Canada, Europe, and Asia have greatly expanded the incidence areas of C. gattii (4). Accordingly, C. gattii has been reported from such diverse countries as Argentina, Austria, Canada, China, Congo, India, Italy, Japan, South Korea, the Netherlands, Spain, South Africa, United Kingdom, United States, and Democratic Republic of Congo, and this list expands every year. Infections in domestic animals, such as goats, dogs, cats, and horses, are common in Australia, New Zealand, Canada, and Brazil (4,19,20). Additionally, infections are reported in migratory, water-dwelling animals, including porpoises and dolphins (21). C. gattii also was associated with native animals in Australia (koala, echidna), New Zealand (kiwi), Africa (cheetah), and Canada (squirrel) (14,19,22). Non-native zoo animals (koala, ferret, tapir, cheetah, llama) and exotic birds (cockatoo and parrots) have been affected with cryptococcosis caused by C. gattii in Australia, Canada, the United States, and Cuba (23). Thus, C. gattii affects a wide variety of native and domestic animals in regions of known clinical presence. Because, in comparison with human clinical samples, veterinary samples are less frequently analyzed for and diagnosed with C. gattii infection and subsequently reported, we suggest that the actual infection rates of C. gattii in animal populations are possibly much higher than presently known.

Published literature on the environmental isolation of C. gattii is patchy, sporadic, and centered in geographic regions reporting a high clinical incidence of C. gattii cryptococcosis. This is true for Canada, South America, and Australia. However, in India the environmental prevalence of C. gattii appears more pervasive than the reported prevalence of the fungus in clinical specimens (Figure). The first environmental isolation of C. gattii (serotype B) was reported by Ellis and Pfeiffer from Eucalyptus trees in 1990 (24) after unsuccessful attempts at environmental isolations from the same tree species in Oklahoma (18) and California (25). The first environmental isolation of serotype C was reported in 1998 from almond trees in Colombia (26). Environmental sampling is much more limited than clinical sampling because clinical isolates are a public health priority.

C. gattii serotype B is the most prevalent serotype in clinical and environmental samples (17,18). Curiously, C. gattii serotype C is a less common constituent of clinical and environmental isolations even though it is associated with AIDS patients and immunocompetent persons (27). C. gattii serotype C has been isolated from humans in clinical samples from India, the Western Hemisphere, and Africa (18,27). Although serotype C is rarely isolated from the environment, its most notable isolation occurred in association with detritus around nonnative almond trees in Colombia (26). More recent and extensive clinical surveys combining serotyping and molecular typing suggest that serotype C is less rare or restricted than previously thought (27). Future studies are unlikely to provide serotype information because the commercial typing reagents are no longer readily available, and thus genotypes will be the primary means to correlate strain characteristics with their environmental and clinical prevalence.

Existing reports and surveys of C. gattii from human clinical, veterinary, and environmental sources are patchy (Figure). Several areas between positive regions would most likely harbor the pathogen. Thus, C. gattii is likely to have a wider geographic distribution than documented. The environment has not been systematically explored to identify the source of C. gattii in the Congo River basin, where the first definitive report of C. gattii emerged. Such environmental surveys are imperative in view of reports of C. gattii cryptococcosis from a number of African countries (3,17,27,28).

C. gattii and Trees

Ellis and Pfeifer reported the first environmental isolation of C. gattii in 1990 in Australia from wood, bark, leaves, and plant debris of Eucalyptus trees (24). Although Eucalyptus is present in many of the areas known to have C. gattii cryptococcosis, the actual isolation of C. gattii from Eucalyptus trees is rare outside Australia, despite extensive sampling. Imported Eucalyptus has not been associated with the environmental presence of C. gattii in Spain, central Africa, or Canada, and most Eucalyptus trees tested in Papua New Guinea, Egypt, and Italy were negative for C. gattii. Furthermore, early environmental surveys for C. gattii in imported Eucalyptus spp. rarely included other local tree species for testing (4). Although understandable, this was unfortunate because C. gattii is now known to have extensive associations with other tree species.

Evidently, C. gattii is established ecologically in trees other than Eucalyptus in many parts of the world, as supported by C. gattii association with native trees in Canada, Brazil, Colombia, India, and Argentina (Figure). C. gattii has been reported from 54 tree species; most (77%) are angiosperms; gymnosperms account for 23% of positive species (Table 2). Gymnosperms and angiosperms can develop decayed hollows, which differ in biochemical composition, available nutrients, presence of water, microbial communities, and fungal associations (29). C. gattii exhibits associations with the gymnosperms Abies spp., Arbutus menziesii var. menziesii, Cedrus spp., Abies grandis, Picea spp., Pinus spp., Pseudotsuga menziesii var. menziesii, and Thuja plicata in Canada; Pinus radiate (Monterey pine) and Cupressus lusitanica in Colombia; and Cedrus deodara and Cupressus sempervirens in Argentina. Angiosperms other than Eucalyptus spp. have been reported positive for C. gattii from North America, South America, Africa, and India. Like Eucalyptus spp., other angiosperm tree species reported as hosts for C. gattii have been extensively exported from their native areas (Table 2). Two prominent examples are Ficus spp. and Terminalia spp. (almond) trees. Ficus spp. are widely distributed in the tropics and subtropics, and many are exported as ornamentals. Ficus spp. have been recorded as C. gattii hosts in Brazil and Colombia but not in other regions (30).

C. gattii vis-à-vis C. neoformans

Cryptococcosis due to C. gattii is unlikely to be recognized in the laboratory without heightened awareness and sustained effort to differentiate these 2 closely related pathogens. Given the much more recent recognition of C. gattii, historical reports are likely to mention only C. neoformans; this is a major consideration in evaluating historical publications on cryptococcosis due to C. gattii. Pigeon droppings are a known ecologic niche for C. neoformans because the pathogen is predominantly isolated from avian environments or areas contaminated with avian feces (1). Thus, urban dwellings frequented by pigeons and containing accumulated pigeon droppings are an important reservoir for human and animal infections.

Pigeons are not known to acquire symptomatic disease but can carry yeast on feathers, skin, crops, or cloaca (1). Other animals reported positive for C. neoformans include macaw, swan, parakeet, Guenon monkey, fox, potoroo, and sheep (1). Thus, exotic and migratory birds and domestic and wild animals can be carriers or susceptible hosts for C. neoformans. The overwhelming association with avian droppings and environment, especially pigeons, sets C. neoformans apart from C. gattii. The ecologic niches for C. neoformans and C. gattii appear to be distinct.

Few reports exist of isolations of C. neoformans and C. gattii from the same habitats with the recognitions of natural hybrids between the 2 species. For instance, C. neoformans and C. gattii have been isolated from same sources, such as Eucalyptus spp. or Syzygium cumini trees or bird feces (1,30,31). C. grubii association with trees might represent fecal contamination by birds inhabiting these trees. Hybrid strains have been isolated from samples of bird feces in urban areas of South America and from patient samples obtained from the Netherlands and Canada (16). The existence of these hybrid strains suggests that at least in some parts of the world C. neoformans and C. gattii occupy either the same ecologic niche or closely overlapping areas.