Stage-specific expression and divergent functions of two insulinase-like proteases associated with host infectivity in Cryptosporidium

Abstract

Background: The determinants of differences in host infectivity among Cryptosporidium species and subtypes are poorly understood. Results from recent comparative genomic studies suggest that gains and losses of multicopy subtelomeric genes encoding insulinase-like proteases (INS-19 and INS-20 in Cryptosporidium parvum and their orthologs in closely related species) may potentially contribute to these differences.

Methodology/principal findings: In this study, we investigated the expression and biological function of the INS-19 and INS-20 of C. parvum. CRISPR/Cas9 was used to endogenously tag both genes with the hemagglutinin epitope. Immunofluorescence analysis revealed that INS-19 and INS-20 are expressed at different developmental stages of the pathogen. Although knockout of either had no detectable effect on the in vitro growth of C. parvum, knockout of INS-20, deletion of its multiple domains, or mutation of the active motif in the functional domain reduced the intensity of C. parvum infection in IFN-γ knockout mice. Consistent with this, mice infected with the INS-20-deleted mutant had reduced intestinal damage and parasite burden.

Conclusions/significance: These results suggest that INS-19 and INS-20 have stage-specific expression with distinct biological functions, and that the presence of the INS-20 in zoonotic C. parvum contributes to its infectivity and fitness in mice.

Fig 1. Sequence characteristics of INS-19 and INS-20 from several Cryptosporidium species.

(A) Domains of INS-19 and INS-20 sequences predicted by InterPro software. INS-19 and INS-20 possess a signal peptide, an active M16 domain containing a zinc-binding motif, an inactive M16 domain at the N-terminus, and an inactive M16 domain at the C-terminus. (B) 3D structures of INS-19 and INS-20 predicted by AlphaFold. INS-19 and INS-20 each consist of four distinct structural components. (C) Phylogenetic relationship of INS-19 and INS-20 from several Cryptosporidium spp. based on neighbor-joining analysis of amino acid sequences, with Cryptosporidium tyzzeri as the outgroup. INS-19 and INS-20 sequences from multiple Cryptosporidium spp. form several clusters that are not completely segregated by species. Sequences in parentheses represent the predicted active motif within the active domain. Numbers on nodes are percent bootstrap values from 1000 replicate analyses.

Fig 2. Expression profiles of INS-19 and INS-20 in Cryptosporidium parvum based on endogenous tagging of the genes.

(A) Illustrations of endogenous tagging of the INS-19 and INS-20 genes with a 3× HA tag and the Nluc-P2A-Neo sequence at the C-terminus. (B) Expression patterns of INS-19 and INS-20 in transgenic lines as indicated by immunofluorescence microscopy of developmental stages using a mAb against the HA tag, with the Sporo-glo (a polyclonal antibody against developmental stages of C. parvum) and the Hoechst (a nuclear stain) as controls. INS-19 and INS-20 have different expression patterns during the developmental stages in C. parvum. Scale bars = 5 μm.

Fig 3. Effect of INS-19 and INS-20 gene knockout on the growth of Cryptosporidium parvum.

(A) Growth patterns of the IId-GFP, Δins-19 and Δins-20 lines in HCT-8 cell cultures. Differences between groups are not significant (p = 0.1995, p = 0.4194, p = 0.1534 for Δins-19 at 3, 24, and 48 HPI, respectively; and p = 0.7912, p = 0.0688, p = 0.1049 for Δins-20 at 3, 24, and 48 HPI, respectively). N = 4. Bars are standard deviations. (B) Oocyst shedding patterns in GKO mice infected with IId-GFP, Δins-19 and Δins-20 lines as indicated by fecal luciferase activity. At the peak of infection, mice infected with the Δins-19 line showed a threefold decrease in fecal luciferase activity compared to those infected with the IId-GFP line, while those infected with the Δins-20 line showed a 6-27-fold decrease in fecal luciferase activity. N = 6. Bars are standard deviations. (C) Differences in weight gain of GKO mice after infection with Δins-19, Δins-20 and IId-GFP lines. Compared to the mice infected with the IId-GFP line, mice infected with the Δins-20 line had higher body weights. N = 6. Bars are standard deviations. (D) Survival rates of the GKO mice infected with IId-GFP, Δins-19 and Δins-20 lines. No significant differences in survival rates were observed between the groups (p > 0.9999 and p = 0.1201 for Δins-19 and Δins-20, respectively).

Fig 4. Differences in histological changes and parasite load of intestinal tissues among mice infected with Δins-19, Δins-20, and IId-GFP lines of Cryptosporidium parvum.

(A) Hematoxylin and eosin microscopy images of the ileum from GKO mice infected with IId-GFP, Δins-19 and Δins-20 lines taken at high magnification. Scale bars = 10 μm. (B) Parasite load in the intestine from GKO mice infected with IId-GFP, Δins-19 and Δins-20 lines. The INS-20 gene knockout significantly reduces the Cryptosporidium load (p = 0.2294 and p < 0.0001 for Δins-19 and Δins-20). N = 25. Bar indicates standard deviation. (C) Hematoxylin and eosin microscopy images of the ileum of GKO mice infected with IId-GFP, Δins-19 and Δins-20 lines taken at low magnification. Scale bars = 50 μm. (D) The ratio of villus to crypt depth of GKO mice infected with IId-GFP, Δins-19 and Δins-20 lines. Knockout of INS-19 or INS-20 genes attenuates the pathological damage in the mouse intestine (p = 0.0004 and p = 0.0054 for Δins-19 and Δins-20). N = 25. Bar indicates standard deviation. (E) Cryptosporidium load in the intestines of GKO mice infected with IId-GFP, Δins-19 and Δins-20 lines using scanning electron microscopy. The absence of each reduces the Cryptosporidium load. For the top three images, scale bar = 30 μm, for the bottom three images, scale bar = 10 μm.

Fig 5. Effect of deletion of the functional domains of INS-20 on Cryptosporidium parvum.

(A) Illustration of the native and modified loci within the INS-20 gene for WT, INS-20ΔNter, INS-20ΔCter, and INS-20^HLLEQ/5A lines. (B) Illustration of the mutation of the functional motif ’HLLEQ’ within the active domain of the INS-20 to ’AAAAA’. (C) Sequence analysis of PCR products from native INS-20 (left) and mutant INS-20^HLLEQ/5A (right). (D) Nucleic acid electrophoresis analysis of the mutation of the motif ’HLLEQ’ within the active domain of the INS-20 to ’AAAAA’. The PCR products "5’ Insertion" and "3’ Insertion" confirm correct integration. The "Control" product is specific for the INS3 locus. (E) Oocyst shedding patterns in GKO mice infected with INS-20ΔNter, INS-20ΔCter, and INS-20^HLLEQ/5A and IId-GFP lines as indicated by fecal luciferase activity. Mice infected with the transgenic lines of INS-20ΔNter, INS-20ΔCter, and INS-20^HLLEQ/5A showed a 3-fold decrease in fecal luciferase activity compared to those infected with the IId-GFP line. N = 3, bar indicates standard deviation.