Updated view of Cryptococcus neoformans mating type and virulence.

Abstract

A paper by Nielsen et al. in this issue of Infection and Immunity (5) will have significant impact on the study of Cryptococcus neoformans, an opportunistic fungal pathogen that causes meningitis in immunocompromised and immunocompetent hosts (1). It changes the practice of C. neoformans genetics by opening up simple genetic manipulation for the most frequently isolated variety of the organism. It changes our thinking by reinvestigating the role of mating type in the virulence of this organism. Mating type has a well-documented but poorly understood relationship with C. neoformans infection. The vast majority of infections are caused by isolates of mating type α, not mating type a (1). This extreme bias is also reflected among environmental isolates, where again mating type α is predominant. In a seminal report some time ago, Kwon-Chung and colleagues created congenic mating type α and mating type a strains in a strain of C. neoformans serotype D and found that the mating type α strain was substantially more virulent than the mating type a strain in a mouse infection model (3). This correlation was extended through analysis of several progeny of the two strains. Indeed, sequence analysis has revealed that the mating type locus has at least 20 genes, several of which might reasonably contribute to virulence (4). These observations have focused much attention on the mating type locus and specifically on its role in virulence. There are two limitations in the analysis of serotype D strains. First, the vast majority of infections in the Western hemisphere are caused by serotype A strains, not serotype D (1). Second, the serotype A strain commonly used at the bench, H99, is substantially more virulent than the congenic pair of serotype D strains in typical animal models. C. neoformans serotypes A and D are quite distant relatives; they can mate only to form sterile hybrids (2). In addition, there are already a few genes known to govern virulence in one serotype and not the other. There is thus the reasonable concern that conclusions from an elegant analysis of serotype D strains might have to be verified with serotype A strains before they can be useful medically. Thus, many research groups have elected to work directly with serotype A strains despite the lack of Mendelian genetics. The new report by Nielsen et al. eliminates this compromise (5). They have used a rare mating type a isolate of serotype A to create congenic mating type a and α strains in the H99 serotype A background. The strategy that they employed is interesting in its own right, because the limited fertility of the mating type a isolate required the use of a surrogate mating type α partner for the first mating. Success of a second cross depended upon the use of a hyperresponsive mating partner. Ultimately, the congenic pair was created through 10 sequential backcrosses to the H99 genetic background, and genetic homogeneity was demonstrated by pulsed-field whole chromosome gel, restriction fragment length polymorphism and amplified fragment length polymorphism analyses. The analyses provide confidence that this congenic pair will be of broad utility to the community. How can Mendelian genetics help to understand a pathogen? Transformation for gene disruption is potentially problematic for C. neoformans, because DNA can be incorporated into the genome through homologous and nonhomologous integration. Mendelian segregation can provide simple proof that a defined mutation is the cause of a phenotypic alteration. In addition, the limited spectrum of selectable markers for this organism creates difficulties for the construction of double mutants in an analysis of regulatory pathways or gene families. Mendelian genetics circumvents these problems. Finally, it will simplify characterization of unmarked suppressor mutations and forward genetic studies of the kind that have been of immeasurable value in serotype D strains. Does mating type govern virulence in serotype A? Nielsen et al. assessed virulence in two frequently used animal models. They saw no difference in the levels of virulence of the mating type a and mating type α strains. Thus, the situation with serotype A seems quite different from that with serotype D. There is the caveat that H99 represents only one particular serotype A isolate, and it remains possible that other serotype A genetic backgrounds may show mating type-responsive virulence. However, the equivalent levels of virulence of the mating types brings to the foreground an equally compelling question that is experimentally quite challenging: How are mating type α strains able to populate the environment more effectively than mating type a strains? Nielsen et al. have contributed valuable starting materials with which to address this question.