The ability of fungi to grow well at mammalian body temperatures is a fundamental characteristic of invasive human fungal pathogens. Cryptococcus neoformans, with its genetics, molecular biology, robust animal models and clinical importance, has become a premier fungal model system for molecular fungal pathogenesis studies. There has been over a half century of study into how C. neoformans grows at high temperatures. However, recently the understanding of high-temperature growth at a molecular level has dramatically accelerated. Many strategies have been used to identify genes and over a dozen genes have already been identified to be necessary for high-temperature growth. It is likely that there are many more to discover. It is clear that, as further studies understand how this encapsulated yeast is able to withstand the stresses of high temperature at a genetic and molecular basis, we will also know more about how it and other fungal pathogens have evolved into well-established human pathogens.
In an elegant discussion of the potential ecological and evolutionary importance of high-temperature growth by certain members of the fungal kingdom and their ability to cause human disease, Casadevall (2005) proposed the hypothesis that during the cretaceous–tertiary boundary period of history, as the fungal biomass and its spores were dramatically increasing due to environmental temperatures changes, the avian and mammalian presence of endothermy became a selective evolutionary advantage. The use of high body temperatures by these species to act as a physical barrier to the vast inocula of fungal spores in the environment allowed a competitive advantage of birds and mammals over the ectothermic species such as the dinosaurs. Of course, this hypothesis cannot be proven but it is provocative to consider that the high energy costs of high body temperatures in mammals were rewarded by a profound, innate resistance to the ubiquitous environmental fungi present within the Earth's biomass.
There are at least 1.5 million known fungal species and they are responsible for most of the biomass degradation on the planet. However, the vast majority of fungal species grow optimally between 25 and 35°C and there are only a few fungal species that appear relatively thermotolerant (growth at 35–40°C) and yet this physical characteristic becomes a pre-requisite phenotype for invasive mycoses to produce disease within a mammalian host. In fact, there have been only c. 270 fungal species reported to cause human disease. Although this number of species has been increasing slightly as we further attend to severely immunocompromised hosts, it is clear that the human body temperature is a primary protective factor for even the severely immunosuppressed mammalian host. It is interesting to note that the most common human fungal pathogens, the dermatophytes, grow poorly at high environmental temperatures and their primary site of infection remains in the lower body temperature areas of the skin and nails, even in immunosuppressed hosts. However, it is with this simple but profound biological characteristic, the ability to grow at high environmental temperatures, that begins to frame how Cryptococcus neoformans has developed into a major fungal pathogen in humans.
Cryptococcal species variability to high-temperature growth
Cryptococcus neoformans and Cryptococcus gattii are the only Tremallelles, which include at least 38 different cryptococcal species, that can grow optimally above 30°C. For instance, there are cryptococcal species, like Cryptococcus podzolicus, which possess a capsule and/or produce melanin, but do not possess the ability to grow at high environmental temperatures (35–40°C) and do not produce mammalian disease (Petter, 2001). On the other hand, strains of C. neoformans have been shown to have optimum and maximum growth temperatures of 32 and 40°C, respectively. However, within the various strains and varieties of C. neoformans the tolerance for high temperature can actually vary. For example, when studying the growth of a large number of strains, it appears that strains of C. neoformans var. grubii (serotype A) generally have a better tolerance for high temperature than C. neoformans var. neoformans (Martinez, 2001). It is interesting to speculate that C. neoformans var. neoformans is primarily found in northern Europe because it is in general less fit for the high environmental temperatures of the tropics and areas closer to the equator. However, there can be overlap between individual strains and varieties in their thermotolerance for growth and thus this physical characteristic does not precisely distinguish between the varieties.
The history of host ecology on high-temperature Cryptococcus neoformans growth
The ability for this encapsulated yeast to be a pathogen and its survival at high temperatures has fascinated scientists for many years and a series of experiments have been performed on this yeast's ability to survive elevated environmental temperatures. For instance, Cryptococcus neoformans yeast cells can survive in the gastrointestinal tract of pigeons (∼40°C; Swinne-Desgain, 1976) and this potential colonization vehicle may be functional for the environmental spread of the yeast from its ecological niche to areas inhabited by birds and humans. Furthermore, in some of the first antifungal strategy experiments, the survival of chicken embryos infected with Cryptococcus were found to be greater in those embryos incubated at temperatures above 39°C (Kligman, 1951), to show that the host temperature at the site of infection has impact on the outcome. In normal rabbits, C. neoformans can survive in the testes at lower body temperatures (<37.0°C) (Bergman, 1966) but the yeast is rapidly eliminated in the central nervous system of the rabbit when inoculated directly into the subarachnoid space (>39.5°C) (Perfect, 1980). Further experiments have emphasized that the rabbit's protective immunity was necessary to have high body temperatures present, but temperature was not wholly sufficient for complete protection, since rabbits immunosuppressed with corticosteroids, despite high body temperatures, were still susceptible to fatal meningoencephalitis with C. neoformans when inoculated directly into the subarachnoid space (Perfect, 1980). It was also shown that the environmental temperature might impact a mammalian response to cryptococcal infection. In mice, it was shown that animals with cryptococcosis could survive longer when they were placed in an environment with higher ambient temperatures (Kuhn, 1949). These early experimental observations on the temperature effect on yeast growth in vivo, prior to the use of antifungal agents, actually led to reported case studies of individuals with cryptococcal meningoencephalistis being treated with hyperthermia. Although one case report suggested improvement of cryptococcal meningitis during the pyrexia of a malarial attack, the clinical experience with hyperthermia therapy was generally unfavorable (Littman & Zimmerman, 1956). It is clinically apparent today that individuals with cryptococcal meningitis can at times present with high fevers of 39–40°C and a large burden of viable yeasts present in the cerebrospinal fluid. It is clear that there are many cryptococcal strains that can adapt well to high-temperature environments and produce invasive mycoses.
Two other pathobiological observations have been noted in relationship to high-temperature growth in mammals. Rhinotrophic strains in both mice and dogs have been identified (Dixon & Polak, 1986; Chung, 2003). It is interesting that these natural strains possessed a temperature-sensitive (ts) phenotype of growth at 24–30°C but not at 37°C. In one of these pathogenic strains, the actual gene defect for ts growth has been identified and it is apparent that the nasal site of infection represented a lower temperature environment for this rhinotrophic strain. In at least two clinical reviews, there have been reports of cryptococcal strains primarily infecting the human skin (Dromer, 1996; Singh, 1997). In both reviews, there was speculation that the presentation was caused by ts phenotypic strains. For instance, in one study, the majority of isolates were C. neoformans var. neoformans which tend to be less thermotolerant (Dromer, 1996). In the second study, these cryptococcal infections of the skin occurred in solid organ transplant recipients receiving calcineurin-inhibiting agents such as cyclosporine or tacrolimus (Singh, 1997). Since it is known that an intact calcineurin pathway is necessary for high-temperature growth of C. neoformans, it is postulated that these immunosuppressive agents might protect against internal invasive disease, compared to skin disease where environmental temperatures are lower. Although this clinical situation occurs in occasional patients with some strains of C. neoformans, high body temperature is surely not protective for all transplant recipients, since there are many individuals who still develop cryptococcal meningitis while receiving calcineurin-inhibiting agents as part of their immunosuppressive regimens.
The genetic and molecular biology of Cryptococcus neoformans high-temperature growth
In the early history of the genetic and molecular biology experiments for understanding the Cryptococuus neoformans virulence composite, the phenotype of high-temperature yeast growth has played a special part. Although there are now multiple virulence phenotypes associated with this encapsulated yeast, the classical virulence phenotypes have been: (1) capsule formation; (2) melanin production; and (3) the ability to grow well at 37°C. The early genetic studies with this yeast have shown that these three phenotypes to be under classical genetic controls (Kwon-Chung & Rhodes, 1986) and with these early studies as a substrate, it was now possible to find the genes that controlled high-temperature growth once the molecular-biology foundation of this yeast was established during the early 1990s. The first gene replaced through biolistic transformation in C. neoformans was the n-myristyl transferase (NMT1) gene (Lodge, 1994). As predicted from Saccharomyces cerevisiae, the change of an amino acid in the NMT1 gene of C. neoformans created a mutant which possessed a ts auxotrophy and the mutant was found to be avirulent in an animal model. The second C. neoformans gene which produced a ts phenotype was not predicted by studies in S. cerevisiae and it was the calcineurin A gene (CNA1) (Odom, 1997). However, the gene's phenotype was predicted from some discrepant results in mice and rabbits with cryptococcosis treated with cyclosporine (Perfect & Durack, 1985; Mody, 1989). When in vitro susceptibility testing showed that cyclosporine had no anticryptococcal activity at 30°C but potent anticryptococal activity at 37°C, the prediction was that a null mutant of calcineurin A would have a ts growth phenotype and be avirulent. The site-directed cna1 mutant was created and the hypothesis was proven to be correct, as the cna1 mutant was ts and completely avirulent in two animal models (Odom, 1997).
We have known that C. neoformans strains that have been frequently passed in the laboratory can microevolve under nonselecting environmental conditions and a fully virulent yeast strain can under frequent in vitro passage become attenuated for virulence in the same animal model (Perfect & Cox, 2000). Furthermore, it has been postulated that C. neoformans, which does not include a mammalian host in its normal life-cycle, has developed and maintained many of its pathogenic features in nature as it encountered environmental predators such as ameba and worms (Steenbergen & Casadevall, 2003). These environmental exposures /encounters might cause the yeast to select certain characteristics that are essential for microbial invasion of the mammalian host. However, direct exposure to high temperatures may also be essential for maintaining optimal fitness of the yeast for production of disease. For example, in elegant but simple experiments, Xu (2004) examined patterns of genotype–environmental interactions from multiple clinical and environmental strains. He showed that spontaneous mutations frequently occur during vegetative growth of strains, but clones maintained at 37°C compared to those at lower temperatures showed less decline in vegetative in vitro growth. Furthermore, clinical isolates with their exposure to mammalian host temperatures had greater vegetative growth compared to those from the natural environment. A hypothesis created by this work is that the total C. neoformans virulence composite can rapidly change for a single C. neoformans strain and that fitness of the strain benefits from periodic high-temperature growth. This high-temperature exposure may be a critical component of maintaining the C. neoformans virulence composite of individual strains.
High-temperature growth has become a very productive focus for the study of molecular pathogenesis in Cryptococcus neoformans. In Table 1 there is a list of 20 genes which have been validated to be necessary for high-temperature growth and pathogenesis in C. neoformans. These genes have all been confirmed by creation of site-directed mutants and their pathobiological importance has been studied in animal models. These over a dozen identified genes are likely just the beginning of our understanding of the genetic controls for high-temperature growth, which are predicted to include over 100 genes. However, these early studies are already yielding a variety of insights into how C. neoformans uses certain genes or pathways to survive the stress of high-temperature exposure and in particular the mammalian host environment.
Cryptococcus neoformans genes shown to be necessary for high-temperature growth (references in parentheses)
CNA1 (Odom, 1997)
CNB1 (Fox, 2001)
TPS1 (Wills, 2003)
TPS2 (Wills, 2003)
SOD2 (Giles, 2005)
TSA1 (Missall, 2004)
CCN1 (Chung, 2003)
VPH1 (Erickson, 2001)
ILV2 (Kingsbury, 2004a)
SPE3/LYS9 (Kingsbury, 2004b)
MPK1 (Kraus, 2003)
CTS1 (Fox, 2003)
STE20 (Wang, 2002)
RAS1 (Alspaugh, 2000)
MGA1 (Kraus, 2004)
Strategies for identification of high-temperature growth genes in Cryptococcus neoformans
The study and identification of the Cryptococcus neoformans genes associated with high-temperature growth has followed two specific pathways and both have been productive investigative strategies. The first strategy has been directed to gene studies in known pathways or to genes with specific functions that were identified and characterized to be ts. The second strategy is a direct attempt to discover genes through complementation of ts strains, prediction of conservation of gene function from studies with other fungal species, use of transcriptional profiling at different temperatures, and screening of signature-tagged mutant libraries for a ts phenotype (Idnurm, 2004).
With the first strategy, a series of genes and the principles regarding their impact on high-temperature growth have been elucidated. High-temperature growth genes are associated with structural components of the encapsulated yeast, including energy for its intracellular vacuoles (VPH1), or specifically involved in the creation of its protective capsule (UGD1) (Griffith, 2004; Moyrand & Janbon, 2004). In fact, in its basic functions such as amino acid metabolism there are several genes essential for high-temperature growth (ILV2; SPE3/LYS9). (Kingsbury, 2004a, b). In the study of the signaling pathways of C. neoformans, it has now been confirmed that several of these important networks possess genes which are necessary for optimal growth at 37°C (RAS1, CNA1, CNB1, MPK1, CTS1) (Odom, 1997; Alspaugh, 2000; Fox, 2003; Kraus, 2003). Another important gene concept in high-temperature growth of C. neoformans is that there is a continuum for a gene's impact on growth, depending on the specific temperature. For instance, there are several genes (STE2O and CPA1) that are required for very-high-temperature growth (39–40°C) but are dispensable for efficient growth at 37°C (Wang, 2001, 2002).
In the second investigative strategy for gene discovery of temperature-related genes, the direct study of ts genes included complementation of a natural ts strain isolated from a dog's nasal cavity and then the isolation and characterization of the gene, CCN1 (Chung, 2003). This strategy can continue to be followed by the use of constructed libraries of signature-tagged mutants and the screening for ts mutants (Idnurm, 2004). A further direct strategy was to examine stress genes or genes/pathways known to be conserved among other fungal species. With this strategy, several genes involved in oxidative stress have been shown to be linked with a ts phenotype (Missall, 2004; Giles, 2005). For instance, the SOD2, a mitochondrial manganese-superoxide dismutase, merges the focus on mitochondrial function, which is predicted to be important during high-temperature exposure in C. neoformans, and this enzyme's function to detoxify this important cellular organelles from its oxidative wastes. C. neoformans appears to produce more wastes as the environmental temperature rises. For example, it has been shown that, with elevated temperature exposure increasing in a continuous fashion, the need for a functioning mitochondrial SOD2 becomes more critical in C. neoformans. The sod2 mutant displays various levels of decreased growth as the ambient temperature rises, until it has very poor viability at 37°C (Giles, 2005). Also, in the area of conserved fungal pathways, the stress-protectant sugar, trehalose, and its pathway have been studied. It is clear that the synthesis of this sugar is essential for high-temperature growth in several fungi (Singer & Lindquist, 1998). The ts phenotype of reduced high-temperature growth was also found in site-directed mutants unable to synthesize trehalose in C. neoformans (Wills, 2003). Trehalose and the expression of genes for its synthesis have been identified at the site of infection (Himmelreich, 2002; Steen, 2003) and mutants in the trehalose synthesis pathway are avirulent in animal models (Wills, 2003).
Insights regarding cryptococcal high-temperature genes and virulence composite
A productive area of temperature-related research has been to examine the Cryptococcus neoformans transcriptome at various temperatures and with several different methods: (differential display RT-PCR (Rude, 2002), cDNA subtraction libraries (Cox, 2003), SAGE (Steen, 2002), and microarray analysis (Kraus, 2004). There are several principles noted from these transcriptional profiling studies. First, at high temperatures compared to lower temperatures, there appear to be a series of both up-regulated and down-regulated genes and the critical regulation depends on the yeast growth stage at the time of temperature exposure and also on the time of exposure to the environmental temperature (Kraus, 2004). Second, temperature-regulated genes may (MGA1) (Kraus, 2004) or may not (AOX1, SOD1) (Akhter, 2003; Cox, 2003) be correlated with a ts phenotype. In fact, the first ts mutant (cna1) does not have regulation of its gene (CNA1) by temperature. However, temperature-regulated genes do appear to be a part of a very enriched gene pool associated with the virulence composite. For instance, virulence genes such as AOX1 (Akhter, 2003), SOD1 (Cox, 2003), SKN7 (Wormley, 2004) and MGA1(Kraus, 2004) have been identified as transcriptionally regulated by temperature. However, temperature-regulated genes may not have a virulence phenotype (CBP1; Gorlach, 2000). Third, there are several genes, down-regulated when yeasts are exposed to high temperatures, that if over-expressed at high temperatures would interfere with the normal growth of C. neoformans (G. Cox, unpublished data). Fourth, in the study of global transcriptional profiles in relation to environmental temperature changes it has been observed that different cryptococcal varieties (serotypes) can have different transcriptional profiles and even can show differences in gene splicing at different temperatures, such as in studies on COX1, a mitochondrial gene (Toffaletti, 2003). Finally, transcriptional profiles in relation to temperature have also led to establishing gene linkages between transcriptional factors such as MGA1 and specific biochemical pathways such as fatty acid synthesis.
There are several observations regarding ts genes of C. neoformans on pathobiology. First, all null ts mutants for growth at 37°C appear to be attenuated in virulence studies within mammalian models. While this observation appears to be an obvious consequence, it is still likely that all ts mutants will not be killed within the host at the same rate. Furthermore, some of these genes will have an effect on the virulence composite through mechanisms unrelated to their ts phenotype. Second, known ts mutants have been used to determine their impact on the host immune response. For instance, the cna1 mutant was able to live and establish infection in the host long enough to produce an inflammatory reaction at the site of infection. However, this immune reaction was not protective for a secondary yeast challenge (Wormley, 2005). Therefore, the use of ts mutants as live vaccines will need further study to create strains that produce a protective immune response but are then eliminated completely from an immunosuppressed host.
The ability of certain fungi to develop the molecular machinery to survive mammalian host temperatures has been a pre-requisite for the virulence composite. High-temperature growth of fungi is necessary but not sufficient to be a mammalian pathogen. However, in studying high-temperature growth and its stresses, scientists will get closer to understanding fundamental aspects of fungal pathogenicity. For Cryptococcus neoformans, the road to this understanding of high-temperature growth has begun and genetic insights have now been made in the roles of specific genes and the pathways they control, but present studies are only the beginning. This road is paved with both knowns and unknowns, but this focus on high-temperature growth has the potential to lead to identification of new antifungal targets and/or protective vaccine candidates.