Cryptosporidium hominis gene catalog: a resource for the selection of novel Cryptosporidium vaccine candidates

Abstract

Human cryptosporidiosis, caused primarily by Cryptosporidium hominis and a subset of Cryptosporidium parvum, is a major cause of moderate-to-severe diarrhea in children under 5 years of age in developing countries and can lead to nutritional stunting and death. Cryptosporidiosis is particularly severe and potentially lethal in immunocompromised hosts. Biological and technical challenges have impeded traditional vaccinology approaches to identify novel targets for the development of vaccines against C. hominis, the predominant species associated with human disease. We deemed that the existence of genomic resources for multiple species in the genus, including a much-improved genome assembly and annotation for C. hominis, makes a reverse vaccinology approach feasible. To this end, we sought to generate a searchable online resource, termed C. hominis gene catalog, which registers all C. hominis genes and their properties relevant for the identification and prioritization of candidate vaccine antigens, including physical attributes, properties related to antigenic potential and expression data. Using bioinformatic approaches, we identified ∼400 C. hominis genes containing properties typical of surface-exposed antigens, such as predicted glycosylphosphatidylinositol (GPI)-anchor motifs, multiple transmembrane motifs and/or signal peptides targeting the encoded protein to the secretory pathway. This set can be narrowed further, e.g. by focusing on potential GPI-anchored proteins lacking homologs in the human genome, but with homologs in the other Cryptosporidium species for which genomic data are available, and with low amino acid polymorphism. Additional selection criteria related to recombinant expression and purification include minimizing predicted post-translation modifications and potential disulfide bonds. Forty proteins satisfying these criteria were selected from 3745 proteins in the updated C. hominis annotation. The immunogenic potential of a few of these is currently being tested.Database URL: http://cryptogc.igs.umaryland.edu.

Figure 1.

Cryptosporidium hominis gene catalog (ChGC). The landing page includes an overview of ChGC and links to related information and resources. Several data subsets are readily available for download (right hand bar), and the full dataset can be further queried with user-selected criteria (bottom button). Direct links to the definition of each criterion, as well as related publications, are also available (top right).

Figure 2.

Approaches used for antigen identification. (A) Genes homologous to previously proposed C. hominis (Ch) or C. parvum antigens (purple) were identified among the gene set from the new C. hominis TU502_2012 genome assembly. The structural annotation of these genes was then manually curated, and targeted analyses were conducted to identify genes encoding proteins with the desired properties. (B) The structural annotation of C. hominis TU502_2012 was improved using information from related species and several of gene finders. The resulting gene set was assigned functional annotation. This gene set was then screened from desired properties. The gene structure of antigen candidates was manually curated.

Figure 3.

Selection of potential of Cryptosporidium vaccine candidates. (A) Overlap between set of potential antigens, one collected from the literature (purple) and the other generated using a bioinformatic screen for genes with predicted GPI-anchor motifs, secretion signals or at least five transmembrane motifs (orange). Of the total 407 potential antigens, roughly one-half were identified with both approaches. (B) Down-selection of genes to be used in immunogenicity experiments. The complete gene complement was first reduced by 90% to 407 candidates from (A), and a further 90% reduction resulted from the use of stricter criteria.

Figure 4.

Properties stored in the C. hominis Gene Catalog (ChGC). The database contains a variety of searchable properties for each gene, including physicochemical properties, gene expression data, presence of potential T-cell epitopes and distribution of detectable homologs across the Cryptosporidium genus and in the human genome.

Figure 5.

The ChGC interface. Key elements: (a) ‘Help’ button; (b) click on a column header to sort by that column; (c) ‘columns’ menu available in the drop-down menu on any column header is used to add hidden, or remove visible, columns; (d) ‘Sort/Filter’: multiple columns can be filtered to generate customized datasets of interest; (e) filtered datasets can be downloaded as an Excel or a CSV file, using these buttons.