Modelling Cryptosporidium infection in human small intestinal and lung organoids

Abstract

Stem-cell-derived organoids recapitulate in vivo physiology of their original tissues, representing valuable systems to model medical disorders such as infectious diseases. Cryptosporidium, a protozoan parasite, is a leading cause of diarrhoea and a major cause of child mortality worldwide. Drug development requires detailed knowledge of the pathophysiology of Cryptosporidium, but experimental approaches have been hindered by the lack of an optimal in vitro culture system. Here, we show that Cryptosporidium can infect epithelial organoids derived from human small intestine and lung. The parasite propagates within the organoids and completes its complex life cycle. Temporal analysis of the Cryptosporidium transcriptome during organoid infection reveals dynamic regulation of transcripts related to its life cycle. Our study presents organoids as a physiologically relevant in vitro model system to study Cryptosporidium infection.

Figure 1. Development of asexual and sexual stages of C. parvum in human SI organoids.

(a) Schematic representation of C. parvum life cycle. (b) Scheme and bright-field images of microinjection. (c) C. parvum 18S rRNA was measured at each time point after injection in differentiated and expanding SI organoids by qRT-PCR (n=2 biologically independent experiments). Mean value at each time point was used for connecting line. (d) Immunofluorescence of C. parvum epicellular stages in expanding organoids. (top) At 24 hr post-injection, meront I (arrow) and possibly meront II (arrowhead) were observed. (bottom) At 72 hr, a microgamont with 16 nuclei was detected. Sporo-Glo marks epicellular stages. DAPI mark nuclei. Scale bars indicate 2μm. More than 40 meronts I, 10 meronts II and 3 microgamonts were observed independently. (e) TEM of distinct stages of C. parvum life cycle after injection. Invading sporozoite and meront II were observed in differentiated organoids at 1 day. Trophozoite was observed in expanding organoid at 1 day. PV: parasitophorous vacuole, FO: feeder organelle, AP: amylopectin granule, WB: wall-forming body, LB: lipid body, DG: dense granule, N: nucleus, RB: residual bod, DB: dense band, EDC: electron dense collar, AI: anterior invagination. Scale bars indicate 2μm. Macrogamont, zygote and developing oocyst were detected in expanding organoids at 5 day. More than 2 sporozoites, 4 trophozoites, 5 meronts II, 5 macrogamonts, 1 zygote and 3 oocysts were observed independently.

Figure 2. C. parvum completes its life cycle inside organoids.

(a) Schematic representation showing microinjection of in vitro excysting sporozoites. (b) Immunofluorescence of C. parvum oocysts isolated from sporozoite-injected organoids at 4 day post-injection. Scale bars indicate 5μm. Isolated oocysts were observed from two independent experiments. (c) (left) Schematic representation showing inoculation of sporozoite-injected organoids into neonatal mice (right). Scatter dot plot showing the level of infection in mice by sporozoite-injected organoids. At 94 hr post-inoculation of organoids, the level of C. parvum HSP70 gene in mice intestine was measure by qPCR (n=biologically 3 independent mice). Numbers of C. parvum stages were calculated by comparison to standards of known quantities of C. parvum. Organoids that were infected with sporozoites and incubated for 5 days (S_5d) were infectious to mice whereas organoids that were infected and incubated for 1 day (S_1d) were not (one-way ANOVA). The lines in the scatter dot plot depict the medians with error bars (± standard deviation). (d) Histological section of the ileum-cecum junction of the inoculated mice. Scale bars indicate 1mm. Similar results were observed from two independent experiments.

Figure 3. In vitro culture of C. parvum in human lung organoids

(a) (left) Schematic representation showing microinjection. (right) C. parvum 18S rRNA was measured at each time point after injection in lung organoids by qRT-PCR (n=2 biologically independent samples). Mean value at each time point was used for connecting line. (b) Immunofluorescence of meront I and microgamont stages in lung organoids. Anti-gp-15 antibody (yellow) marks both merozoites in meront and microgametes in microgamont. Scale bars indicate 5μm. More than 20 meronts I and 5 microgamonts were observed. (c) Immunofluorescence of newly formed oocysts inside lung organoids at 168 hr post-injection. Scale bars indicate 20μm. The oocysts were observed from more than 6 independent organoids. Media-injected organoids were used as a control.

Figure 4. Transcriptome analysis of host epithelia and C. parvum.

(a) Volcano plot showing differential expression of host genes between oocyst- and media (control)-injected organoids at 24 and 72 hr post-injection (n=3 biologically independent samples). Each dot represents a gene. The red and green dots represent differentially expressed genes with p-value < 0.05 and with p-value <0.1 (Wald test), respectively upon injection. The vertical lines represent log2 fold change values as indicated in x axis. List of differentially expressed genes are available in Supplementary Table 1. (b) Heatmap showing the expression of top 30 induced genes in lung organoids at 24 hr post-injection (n=3 biologically independent samples). Expression values are expressed as Z-score transformed transcript count. (c) (top) Heatmap showing the expression of 30 induced genes in lung organoids at 72 hr post-inject (n=3 biologically independent samples). (bottom) Gene set enrichment assay (GSEA) plot showing strong enrichment of type I interferon regulation genes in oocyst-injected lung organoids. NES: normalized enrichment score, p-value is nominal p-value. Black bars underneath the graph present the rank position of genes from the gene set. GSEA plots of genes in differentiated SI organoids are available in Supplementary Fig. 5. (d) Volcano plot showing differential expression of C. parvum genes between 24 and 72 hr post-injection into differentiated SI and lung organoids (n=3 biologically independent samples). The magenta and blue dots indicate enriched genes with p-value <0.05 and with p-value <0.1 (Wald test), respectively. List and GO-term analysis of differentially expressed C. parvum genes are available in Supplementary Table 2 and Fig 6, respectively. (e) Heatmap showing the expression of ribosomal protein coding genes (top) and oocyst wall protein genes (bottom) of C. parvum in lung organoids at 24 and 72 hr post-injection (n=3 biologically independent samples). Heatmap of C. parvum genes in SI organoids are available in Supplementary Fig. 7.