A single-cell RNAseq atlas of Schistosoma mansoni identifies a key regulator of blood feeding

Abstract

Schistosomiasis is a neglected tropical disease that infects 240 million people. With no vaccines and only one drug available, new therapeutic targets are needed. The causative agents, schistosomes, are intravascular flatworm parasites that feed on blood and lay eggs, causing pathology. The function of the parasite’s various tissues in successful parasitism are poorly understood, hindering identification of novel therapeutic targets. Using single cell RNAseq we characterize 43,642 cells from the adult schistosome, identifying 68 distinct cell populations including specialized stem cells that maintain the parasite’s blood-digesting gut. These stem cells express the gene hnf4, which is required for gut maintenance, blood feeding, and pathology in vivo. Together, these data provide molecular insights into the organ systems of this important pathogen and identify potential therapeutic targets.

One Sentence Summary:

Single-cell RNAseq is used to study the basic biology of and identify a novel therapeutic target against a deadly parasite

Schistosomes dwell inside the host’s circulation, often for decades, where they feed on blood and lay eggs, which become trapped in host tissues and cause disease pathology. As a metazoan comprised of multiple tissue types, understanding the schistosome’s biology on a molecular level during parasitism could suggest novel therapeutic strategies. Single-cell RNAseq (scRNAseq) has been used to comprehensively describe tissue types and physiology of diverse metazoans (1) including larval schistosomes (2) but we lack a comprehensive description of the cell types present in egg-laying adults as specific molecular markers are known for only a small number of cell types (3–8).

To define the molecular signature of adult schistosome cell types, we dissociated adult Schistosoma mansoni, isolated cells by Fluorescence-Activated Cell Sorting (FACS), and generated scRNAseq libraries using a 10x genomics chromium controller (Fig. S1A). Schistosomes are dioecious and sexual maturation of the female worm’s reproductive organs, including the ovary and vitellaria, requires sustained physical contact with the male worm (9). Accordingly, we generated scRNAseq libraries from adult male parasites, adult sexually mature female parasites, and age-matched virgin female parasites. We then performed clustering, identifying 68 molecularly distinct clusters composed of 43,642 cells (Figs. 1A, S1B, Table S1). These included: three clusters of cells expressing somatic stem cell (i.e., neoblast) markers such as the RNA binding protein nanos2, the cell surface receptor notch, and the receptor tyrosine kinase fgfra (3) (Figs. 1B, S2A); eight clusters expressing markers of tegument (“skin”-like surface) progenitors (4, 5) (Fig. S2B); two clusters of parenchymal cells (Figs. 1C, S2C); one cluster of ciliated flame cells that are part of the worm’s protonephridial (excretory) system (Figs. 1D, S2D); eight clusters of muscles (Fig. 1E); and a cluster of esophageal gland cells (Figs. 1F, S2E). Despite being composed of thousands of nuclei, our analysis also identified clusters corresponding to syncytial tissues: the tegument (4) (Figs. 1G, S2F) and gut (Figs. 1H, S2G). We failed to identify cells from the female ootype (an organ involved in eggshell formation) (9) and the protonephridial ducts (10), possibly because of their multinucleate nature. Gene ontology (GO) analyses of these clusters (Table S2) confirmed expected findings (enrichment of “DNA replication” in “neoblast 1”) and revealed novel biology such as the enrichment of “extracellular matrix structural components” in muscle clusters suggesting muscles are the source of extracellular matrix in schistosomes, similar to planarians (11).

Fig. 1. Schistosoma mansoni single cell atlas.

(A) Uniform Manifold Approximation and Projection (UMAP) plot of the 68 scRNAseq clusters. (B-I), (left) UMAP plot and whole-mount in situ hybridization (WISH) of the indicated gene and its expression in the noted tissue in the head (middle, top) and body (middle, bottom) of a male and the ovary (right, top) and vitellaria (right, bottom) of a mature female parasite. Scale bars, 100μm. UMAP plots colored by gene expression (blue = low, red = high).

We uncovered unexpected molecular complexity within the schistosome nervous system, identifying 30 clusters expressing the neuroendocrine protein 7b2 (Figs. 1I) and one apparent neuronal cluster that did not express high levels of 7b2 but expressed several of synaptic molecules (e.g. synapsin) (Figs. S3A, Table S1). Examination of genes from these neuronal cell clusters uncovered unique molecular fingerprints for several populations (Figs. S3A, S4, Table S1) and highly-ordered structural and regional specialization in the central and peripheral nervous systems, including left-right asymmetry (Fig. S3B) and nine types of apparently ciliated neurons (Fig. S3C,D). This complexity is surprising given the relatively “sedentary” lifestyle of adult parasites in the portal vasculature (9).

Schistosome muscle is also very heterogeneous, with eight muscle clusters that possess unique expression patterns (Fig. S5A–C). Some populations occur diffusely throughout the animal (“muscle 1” and “muscle 2”), whereas others are anatomically restricted such as “muscle 7” cells that reside next to the gut, suggesting they are enteric muscles.

Similar to planarians (12), many morphogens that regulate wnt (Fig. S6A–D) and tgfb signaling (Fig. S6E–H) are expressed in muscle and neuronal cells. Homologs of many of these genes are expressed specifically in planarian muscles (1) and have been implicated in regeneration in planarians (12). Though schistosomes survive amputation (13), there is no evidence of whole-body regeneration. This expression pattern in a non-regenerative animal suggests these genes may regulate schistosome neoblasts during homeostasis.

The pathology of schistosome infection is driven by the host’s inflammatory responses to parasite eggs(14). Thus, we examined the differences between male, sexually mature female, and age-matched virgin females at the cellular level (Fig. 2A). All adult parasites have germline stem cells (GSCs) marked by expression of nanos1(6). Our scRNAseq data revealed that GSCs have very similar gene expression regardless of sex or maturity (Figs. 2B, S7A). Like GSCs, GSC progeny fall into the same clusters in both male and female parasites, suggesting no major sex- or maturation-dependent differences in early gametogenesis (Figs. 2C, S7B). However, later germ cells cluster according to sex, with expression of “late female germ cells” markers found predominantly in mature females (Figs. 2D and S7C) and “late male germ cells” markers only in males (Fig. S7D).

Fig. 2. The germ lineage in schistosome ovaries.

(A) UMAP plots of all clusters split by parasite sex. Sexual tissues are labeled. (B-D) (top) WISH and UMAP plot of gene expression of indicated gene in sexually mature females (m♀) (top) and in virgin females (v♀) (bottom) for the “GSCs” marker nanos1 (B), the “GSC progeny” marker meiob (C), and the “late female germ cells” marker bmpg (D). Dashed line indicates boundary of ovary. Scale bars, 100μm. UMAP plots are colored by gene expression (blue = low, red = high).

The sexually mature schistosome ovary is structured such that GSCs reside at the anterior and mature oocytes at the posterior end (6, 15). The “GSCs” marker nanos1 is expressed in the proliferative anterior compartment (Figs. 2B, top, S8A–D) whereas the “late female germ cells” marker bmpg is expressed most highly in the posterior ovary (Figs. 2D, top, S8C). Our single-cell RNAseq data shows that the “GSC progeny” cluster sits between “GSCs” and “late female germ cells” on the UMAP plot, (Fig. 2A), with the “GSC progeny” marker meiob expressed most highly between the anterior and posterior ovary (Figs. 2C, S7B). Concurrent visualization of these clusters reveals an organized linear architecture (Fig. S8E). Interestingly, both mature and virgin females express the marker meiob (Fig. 2C), suggesting that virgin female GSCs express differentiation markers without male stimulus. Thus, male parasites may regulate this developmental checkpoint by promoting survival of differentiating GSCs rather than inducing commitment, consistent with studies suggesting that male-female pairing can suppress apoptosis in the vitellaria of virgin female worms (16).

We also examined the vitellaria, another male-sensitive, stem-cell dependent tissue that produces the yolk cells of the parasite’s eggs. Despite a different function and organization, we observed parallels between ovary and vitellaria maturation, such as an apparent lineage from stem cell to mature tissue (Fig. S9A–D). We also found a low frequency of vitellocyte-like cells in males (17) (Fig. S9A). Finally, we identified pairing-independent sexual tissues such as the flatworm-specific Mehlis’ gland that plays an enigmatic role in egg production (9) (Fig. S9E).

In addition to sexual tissues, we observed sexual dimorphism in non-reproductive tissues as well including 3 muscle clusters (muscle 5, 6, and 8) that appear to be largely restricted to female parasites (Table S3), with “muscle 8” representing muscle cells that surround the ovary (Fig. S10). In some cases, we observed unexpected numbers of male cells in clusters of female sexual tissues which we attribute to neoblasts expressing low levels of differentiated tissue markers, like what has been observed in planarians (1) (Fig. S11, Supplementary Text).

Egg production is the primary driver of pathology, but this pathology is exacerbated by the parasite’s stem cell-mediated longevity (3). Previous work suggests adult neoblasts are molecularly homogeneous and predominantly give rise to cells involved in tegument production (4, 5) but free-living flatworms are known to possess functionally distinct neoblasts that produce specific tissues (18). We identified a subpopulation of neoblasts (“eled⁺ neoblasts”) that formed a putative non-tegument lineage as suggested by a linear “path” of cells from eled⁺ neoblasts to the gut (Figs. 3A, S12A–F). These eled⁺ neoblasts expressed hnf4 (Figs. 3A and S12B,C), a marker of gut neoblasts in planarians (18). Given the importance of gut-mediated blood digestion for egg production (19), we sought to perturb this lineage by RNAi of genes expressed in this lineage (Fig. S13A,B). We found that knocking down hnf4 resulted in a ~3.8-fold increase in eled⁺ neoblasts (Figs. 3B and S13C–F) and a concomitant decrease in the expression of several gut markers (Fig. S14A,B). Indeed, RNAseq on hnf4(RNAi) animals demonstrated that over 70% of transcripts expressed in the “gut” cluster were downregulated (Fig. S14C,D, Table S4).

Fig. 3. An hnf4 homolog regulates a novel gut lineage.

(A) UMAP plots of the expression pattern of the indicated gene on the (top) original dataset or the (bottom) re-clustered dataset, and (right) a colorimetric WISH of a male parasite’s trunk for eled, hnf4, prom2, and ctsb. Insets: magnifications of dashed boxes. (B) Fluorescence in situ hybridization (FISH) and EdU labeling showing the expression of eled (green) and EdU⁺ proliferative cells (yellow) in control or hnf4(RNAi) animals. n = ≥ 18 parasites, two biological replicates. (C) FISH of ctsb (cyan) and fluorescent dextran (red) in the gut lumen in control(RNAi) and hnf4(RNAi) animals. n = 15 parasites, three biological replicates. (D) TEM micrographs showing gut of control(RNAi) and hnf4(RNAi) animals. ‘mv’ microvilli, ‘ga’ gastrodermis, ‘L’ lumen, ‘em’ enteric muscle. n = 4 parasites, two biological replicates. Nuclei: blue (B) or grey (C). The number of parasites similar to the representative image is in the upper right of each panel. Scale bars: A, 100 μm, B, 50 μm, C, 20 μm, D, 1 μm. UMAP plots are colored by gene expression (blue = low, red = high).

To understand whether stem cells functioned normally in hnf4(RNAi) animals, we first looked at apoptosis using TUNEL and found no difference in hnf4(RNAi) animals, ruling out increased cell death (Fig. S15A). Next we looked at tegument production using EdU pulse-chase approaches. We found a significant increase in tegument production compared to controls (Fig. S15B,C), ruling out a broad stem cell differentiation defect. Our ability to monitor new gut production by EdU pulse-chase approaches was complicated by the fact that gut marker expression was largely absent in most hnf4(RNAi) parasites (Fig. S14A,B). In cases where we could detect gut marker expression in EdU pulse-chase experiments, we found gut-like tissue was being produced in hnf4(RNAi) parasites but was morphologically abnormal (Fig. S15D). Examination of the expression of eled and the gut marker ctsb revealed that locations where eled⁺ neoblasts were abundant lacked normal gut tissue (Fig. S15E). This suggests that the impairment of gut production is at least partially responsible for the gut defects following hnf4 RNAi.

To assess gut structure, we next supplemented the culture media of hnf4(RNAi) parasites with fluorescently-labeled dextran (which labels the gut lumen (20)). After 12 hours of culture, all control(RNAi) parasites but only 1 out of 15 hnf4(RNAi) parasites had dextran in the lumen (Fig. 3C). The dextran failed to enter the digestive tract of the hnf4(RNAi) parasites (Fig. S16A), suggesting either a complete loss of patency or a defect in the parasite’s ability to coordinate dextran ingestion. We then examined hnf4(RNAi) animals by transmission electron microscopy (TEM). The schistosome gut is a syncytial blind tube-like structure with a microvilli-filled lumen (21). Though gut tissue was still present, we found a significant decrease in luminal microvilli (Figs. 3D, S16B) and 2 out of 4 of hnf4(RNAi) animals had dilated lumens compared to controls (Fig. S16C).

To assess the digestive capability of hnf4(RNAi) parasites, we added red blood cells (RBCs) to the media and observed the parasites’ ability to uptake and digest the cells. hnf4(RNAi) parasites either failed to ingest (15/69) or digest RBCs (54/69) (Figs. 4A, S17A). Because we observed a decrease in the expression of proteolytic enzymes by RNAseq (Table S4), we studied whether hnf4 RNAi resulted in loss of cysteine (cathepsin) protease activity (which contributes to hemoglobin digestion (22, 23)). Measuring cathepsin activity of lysates in hnf4(RNAi) parasites using a fluorogenic peptidyl substrate, we found cathepsin B activity was decreased 8.2-fold relative to control parasites (Fig. 4B) consistent with gene expression analyses (Table S4). In contrast, aspartyl protease activity was similar in control and hnf4(RNAi) parasites (Fig. S17B), which could reflect expression of aspartic proteases in non-gut tissues that were unaffected following hnf4 RNAi (Table S1, S4). Together, these data suggest hnf4 is at least indirectly required for the digestion of hemoglobin, in part by regulating the expression of cathepsin B, a key contributor to the digestion of blood proteins including hemoglobin (22, 23) in S. mansoni.

Fig. 4. hnf4 is required for blood feeding and pathology.

(A) Brightfield images of control(RNAi) or hnf4(RNAi) animals cultured with red blood cells. Inset: magnification of boxed area. (B) Cathepsin activity of lysates from control(RNAi) or hnf4(RNAi) animals determined by cleavage of Z-FR-AMC with no inhibitor (DMSO), a broad cysteine protease inhibitor (E-64), or a cathepsin B-selective inhibitor (CA-074). n = 3, three biological replicates. (C), H&E-stained mouse liver sections 22 days post-transplant with RNAi-treated parasites. Arrows: granulomata. Sections from n = 3 recipients. (D) Parasites recovered from transplant recipients. n > 15 from three recipients. Nuclei: white. The number of parasites/sections similar to the representative micrograph is in the upper left of each panel. Scale bars: A,C, 100μm, D, 1 mm. ****, p < 0.0001 (Welch’s t-test).

We examined whether hnf4 was required to cause disease in the host by transplanting control and hnf4(RNAi) parasites into uninfected mice and then perfusing the mice 22–30 days post-transplant. Worm recovery was statistically indistinguishable (control(RNAi) = 72% vs. hnf4(RNAi) = 49%, p = 0.136, Welch’s t-test) (Fig. S17C). This observation is not entirely unexpected as schistosomes can acquire nutrients though their tegument (19). Nonetheless, mice receiving hnf4(RNAi) parasites had morphologically normal livers in contrast to abundant egg-induced granulomata in control parasite recipients (Figs. 4C, S17D). Additionally, recovered male hnf4(RNAi) parasites were significantly shorter than controls (2.87mm vs. 5.21mm, respectively, p < 0.0001, Welch’s t-test) (Figs. 4D, S17E). These results show hnf4 is at least indirectly required for parasite growth and egg-induced pathology in vivo. Together, these data suggest hnf4 specifically and gut homeostasis generally are potential therapeutic targets to blunt the pathology caused by adult parasites.

Here we describe a comprehensive single-cell atlas of the adult schistosome, identify regulators of gut biology, and leverage this knowledge to experimentally block schistosome-induced pathology in the host. We envision these data serving as a catalyst towards understanding other aspects of schistosome biology (e.g., reproductive biology) and serving as a foundation for understanding the development of various cellular lineages during the parasite lifecycle. Indeed, our approach serves as a template for the investigation of other understudied and experimentally challenging parasitic metazoans, improving our understanding of their biology and enabling the discovery of novel therapies for these pathogens.