On the relationship between serotype, chemotype and genotype for Cryptococcus spp. including a method for including polysaccharide structure in strain characterization

Abstract

Over the past eight decades the classification of cryptococcal strains has relied on the technologies available to discriminate among isolates, such as serology, exopolysaccharide (EPS) NMR, and most recently multi-locus sequencing, which yielded serotypes, chemotypes, and genotypes, respectively. However, as one method superseded the other, the relationship between classification schemes became uncertain, resulting in assumptions that have not been rigorously validated. Here we compared the serotype, chemotype, and genotype as defined by multi-locus sequence typing (MLST) of 63 strains for which both serotype and chemotype characterization was available. None of the three strain typing methods were correlative with each other, although for about 50% of the strains there was correlation between chemotype and genotype. To address this, we updated the methodology of GXM motif categorization for cryptococcal strains through analyzing filter-isolated EPS by 1D [1H] NMR and assigning GXM motifs along with the O-acetylation level. The result is a facile method using minimally processed EPS material coupled with simple cryptococcal strain classification by GXM motif expression. While this method increases the correlation between MLST genotype and GXM motif expression, further studies to establish the polysaccharide regulatory genes in cryptococcus will be necessary to understand the polysaccharide differences. In summary, none of the current classification methods correlated with each other, indicating a dissociation between genotype and phenotype, which poses a challenge to the cryptococcal field for explaining how phenotypic characteristics arise and are maintained. Author SummaryThis work examines the relationship between methods of cryptococcal strain classification and the primary structure of the predominant polysaccharide glucuronoxylomannan (GXM). We find no strong correlation between the expression of trimannose repeat motifs of GXM by a strain and serotype, MLST genotype, or the previous polysaccharide classification system, chemotyping. With an understanding that none of these classification systems accurately represent the GXM sequences in cryptococcal polysaccharides, we describe a new, facile, rapid method of polysaccharide isolation and GXM motif expression characterization. Application of this method reveals a strong association between high levels of GXM O-acetylation, necessary for antibody binding, and the expression of a single motif of GXM. Additionally, the association between MLST genotype and GXM motif expression is higher when utilizing this new method. While much remains to be understood about this complex system, this work provides a new method to include GXM motif expression in cryptococcal strain expression allowing for more accurate strain descriptions in the future.