Fasciola hepatica possess the core components of functional Wnt signalling pathways

Mining genomic datasets facilitated the identification of putative orthologues of all major canonical and non-canonical Wnt pathway components in F. hepatica, including receptors and antagonists. Previously, McVeigh et al. [21] reported a greatly reduced and dispersed complement of Wnt and Frizzled orthologues in F. hepatica, comprising three putative Wnt protein ligands and five Frizzled class GPCRs. Here, this was corroborated by leveraging the most recent F. hepatica genome assemblies (PRJEB58756 and PRJNA179522), with sequence annotations revealing FhWnts to be representative of the WNT1, WNT4 and WNT9 subfamilies. Two additional Wnt ligands belonging to the WNT2 and WNT5 subfamilies were also identified. All gene annotations were assigned based upon the top hit generated from reciprocal BLASTs against H. sapiens using the NCBI non redundant protein (nr) database. While this represents a greatly reduced complement of both protein families when compared to mammalian genomes which typically harbour 19 Wnts and 10 FZDs, similar numbers have been reported in other members of the phylum. For example, the genome of S. mediterranea encodes 9 Wnts and 13 FZDs, S. mansoni 5 Wnts and 9 FZDs, while Echinococcus spp have 6 Wnts and 8 FZDs [19]. Both F. hepatica and S. mansoni were found to lack orthologues of low-density lipoprotein receptor-related proteins 5 and 6. This constitutes a noteworthy absence given the key roles of LRP5/6 as co-receptors in the Wnt/β-catenin pathway. Further, while the related S. mediterranea and E. multilocularis lack LRP5, they do possess LRP6 orthologues. In the absence of LRP5/6, it may be that F. hepatica and S. mansoni instead opt to employ receptor tyrosine kinase like orphan receptors 1 and 2 (ROR1/ROR2) as co-receptors in their respective Wnt/β-catenin signal cascades and as such, these proteins warrant future functional interrogation. Additionally, F. hepatica possess a LRP1B orthologue (FhHiC23_g15024). LRP1B is not recognised as a Wnt/β-catenin pathway component in higher organisms, however, there exist reports of interactions with Wnt and FZDs [33,34]. While the results of such interactions repress Wnt signalling, it suggests that the docking site of LRP1B is structurally and functionally compatible with Wnt protein ligands and as such, FhLRP1B may have displaced FhLRP5/6 over the course of evolution. While the core molecular machinery required for functional Wnt pathways is clearly evident in F. hepatica, it is noteworthy that much like other parasitic flatworm species, they lack homologues of the antagonists Dikkopf and Cerberus. A third antagonist, Wnt inhibitory factor (WIF), was not identified in liver fluke datasets, despite its presence in the related H. microstoma, Echinococcus spp, Schistosoma spp and S. mediterranea [19]. While two FhSFRPs were identified, there is a general paucity of pathway antagonists within the F. hepatica genome, a pattern which appears to be unanimous across several basal metazoan lineages, including Drosophila and C. elegans [35,36].

With few exceptions, a full complement of orthologues unique to both non-canonical pathways were also identified (S1 Table), with the conserved domains of each gene family intact, indicating that both canonical and non-canonical Wnt signalling is functional in F. hepatica. The sole planar cell polarity (PCP) pathway component lacking any discernible orthologue in F. hepatica was dishevelled-associated activator of morphogenesis 1 (DAAM1), while nuclear factor of activated T-cells (NFAT) and cell division control protein 42 homolog (CDC42) were absent from the FhWnt/Ca²⁺ pathway.

FhWnt pathway orthologues possess key domains associated with gene function in higher organisms

Of the three characterised Wnt pathways, the Wnt/β-catenin signal cascade is the most widely studied and its significance in platyhelminth development has been underscored in regenerating planarians. As such, putative FhWnt/β-catenin pathway orthologues were subjected to further in silico analyses to validate their integrity as true orthologues of those annotated in both higher and related organisms. The predicted protein sequences of all genes tentatively selected for RNAi experiments were aligned with corresponding sequences from H. sapiens, M. musculus, D. melanogaster, S. mediterranea, E. multilocularis and S. mansoni. Alignments of FhWnts revealed full conservation of the Wnt family signature motif (CKCHGVSGSC) within the frizzled receptor binding site [37], while the conserved C-terminal cytoplasmic motif of frizzled receptors (KTXXXW, where X denotes any amino acid), required for both the activation of the Wnt/β-catenin pathway, and for membrane localisation and phosphorylation of dishevelled [38], is present in FhFZDs (S1A and S1B Fig). FhCTNNB1 possesses N-terminal residues essential for protein-level regulation including Ser33/Ser37/Thr41 and Ser45 which are phosphorylated by GSK3β and CK1 of the destruction complex, respectively. Also in consensus with higher and related organisms, was a conserved lysine residue at position 49 and the DSG motif (DSG□XS where □ denotes a hydrophobic amino acid, and X any amino acid), in which both serines are phosphorylated and recognised by the E3 ubiquitin ligase complex, resulting in the ubiquitination and degradation of β-catenin [39] (S1D Fig). A second FhCTNNB was identified (FhHiC23_g16384), however, as it lacks the aforementioned conserved domains, it was not investigated further in this study. FhDSHs were found to possess the characteristic dishevelled PDZ binding domain [40,41] in addition to the class III PDZ-binding motif [42] (S1C Fig). The highly conserved motifs CDFGSAK and SYICSR [43], present only within members of the GSK-3 subfamily of serine/threonine protein kinases were identified in both FhGSK3β isoforms (S1F Fig). FhAPC possesses two SAMP repeats, believed to mediate interactions with axin of the destruction complex [44] (S1E Fig). Finally, alignments of the antagonists, FhSFRPs revealed the presence of a fully conserved cysteine rich domain homologous to that of FZDs, containing 10 invariantly spaced cysteine residues [45] (S1G Fig).

Transcriptomic data support a role for Wnt/β-catenin signalling in juvenile Fasciola hepatica growth and development

Prior to the functional interrogation of individual pathway components, two distinct transcriptomic datasets were exploited in attempts to elucidate the relative importance of Wnt signalling in juvenile F. hepatica, the key life stage for control. Analyses of life stage transcriptome data [46] revealed that genes involved in active Wnt signalling were most highly transcribed in the early juvenile stages, notably 24 h old NEJs, while the expression of members of the Wnt/β-catenin destruction complex demonstrated bias towards adult stage fluke (S2 Fig) – both of these observations are consistent with elevated Wnt/β-catenin signalling in rapidly growing juveniles and reduced Wnt/β-catenin activity in adults. Few components involved in active Wnt/β-catenin signalling diverge from this trend, with the exception of dishevelled (DSH) which is most highly transcribed in the adult stage. The key effector of the pathway, β-catenin is also highly expressed in adult stage fluke, however, remains biased towards early juvenile stages.

Previously, through comparative transcriptomic analyses, we demonstrated the downregulation of FhWnts and FhFZDs in ex vivo three-week-old juveniles relative to developmentally delayed, time matched in vitro-maintained specimens [26], suggestive of key roles in early development. Following excystment, newly excysted juveniles (NEJs) enter a period of remarkably rapid growth and development, as evidenced by a twofold increase in size every two weeks, with the concurrent upregulation of neoblast-like cell (cell-cycle) associated transcripts [47] and elaboration of the gut and reproductive structures [48]. As such, the upregulation of Wnt pathway components in these early developmental stages is likely reflective of the influential role of Wnt signalling in organogenesis and the formation of distinct cell and tissue types [49–53]. We would hypothesise that a key role for FhWnts during juvenile development is the establishment and maintenance of anterior/posterior identity for various tissue types. While four FhWnts were differentially expressed in in vitro and in vivo juvenile fluke, just two FhFZDs were downregulated in in vivo maintained juveniles [26]. This may represent interchangeability among FhWnt-FZD interactions, highlighting the potential for FhWnt signalling cascades to be highly complex and dynamic systems.

Localisation of FhWnt pathway components using fluorescence in situ hybridisation (FISH) reveals the partitioning of expression sites

This study localised eight FhWnt pathway component transcripts, revealing visually striking polychotomy among expression patterns throughout the F. hepatica body plan (Fig 1). With the exception of the gut branches, FhWNT1 appears to be absent from the posterior third of the worm and was predominantly detected within sub-tegumental parenchyma. FhWNT2B transcripts were detected in a total of eight, relatively large cells comprising two distinct clusters in the anterior region of the worm. Despite their close proximity to one another, disparate morphologies indicate two unique cell types. Much like FhWNT1, a strong signal was also observed in the guts of FhWNT2B-labelled worms. While autofluorescence of the F. hepatica gut is a common occurrence following staining procedures, the guts of worms exposed to FhWNT1 and FhWNT2B sense probes were not labelled, supporting the specificity of antisense probe binding and thus, the expression of FhWNT1 and FhWNT2B in the F. hepatica gut (S3 Fig). The expression loci of FhWNT4 and FhWNT5A transcripts both exhibit mid-anterior bias, being totally absent from the posterior region. FhWNT4, localised to a network of cells scattered throughout the midbody region of the worm, while FhWNT5A was detected within a population of cells across the mid-anterior region, in addition to parenchymal tissue surrounding the oral sucker. This curiously distinct band of cells highlighted by the FhWNT5A antisense probe is perhaps reflective of a role in the patterning of the mediolateral axis, as is the case with a planarian WNT5 sub-family orthologue [15]. As such, juvenile F. hepatica exhibit a unique mid-anterior bias in Wnt transcript loci, distinct from the posterior bias commonly reported in related species. FhWNT9A was the only Wnt transcript to diverge from this trend, localising to cells displaying a highly regimented pattern around the lateral edges of the posterior half of the worm. Widespread staining of the midbody was also detected in addition to a strong signal emanating from the parenchyma surrounding the oral sucker. While it is impossible to unequivocally identify and differentiate between specific cell types based on FISH alone, disparate morphologies and loci suggest that no two pathway components localised to the same cell population. Similarly polychotamous expression patterns have been observed in Hymenolepis microstoma [54]. Leveraging the recently generated single cell RNAseq atlas of S. mansoni [55] revealed that homologues of all FhWnt pathway components localised using FISH were particularly enriched in various neuronal and muscle cell clusters, with sporadic expression in neoblasts, S1 progeny and flame cells (S. mansoni Wnt pathway gene IDs orthologous to those silenced in F. hepatica can be found in S3 Table). FhCTNNB1 and FhWNT1 S. mansoni homologues are ubiquitously expressed, corroborating the staining patterns observed in this study. In planarians, muscle cells are the source of developmental and positional control genes, including Wnts and SFRPs [56] and serve as landmarks during differentiation by relaying positional information to neoblasts. This, coupled with the S. mansoni homology data suggest that the majority of cell types expressing Wnt pathway transcripts in F. hepatica are likely muscle.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Juvenile Fasciola hepatica Wnt/β-catenin pathway components display distinct tissue expression patterns.

Fluorescence in situ hybridisation (FISH)-mediated localisation of FhWnt pathway transcripts in four-week old juvenile F. hepatica. Target localisation is denoted by red (TAMRA) fluorescence, green fluorescence denotes EdU+ (neoblast-like) cells. DAPI (blue) served as a counterstain. Scale = 50 µm.

https://doi.org/10.1371/journal.ppat.1012562.g001

There exist multiple reports of Wnt pathway in situ hybridisation in planarians, the majority of which detail a posterior bias in Wnt protein ligand expression [15,17,51], while antagonists localise to the anterior pole [13,15,51]. The only SmedWNTs to diverge from this trend are a WNT2 subfamily member which is expressed laterally in the head and a WNT5 subfamily member demonstrating lateral expression along the dorsoventral axis [57,58]. Wnt pathway transcript localisation in the E. multilocularis and H. microstoma metacestode stage mirrors this trend [22]. The highly dispersed spatial patterns of FhWnt transcript loci (Fig 2) are, therefore, at odds with the strict anterior or posterior classification of Wnt antagonists and ligands reported in other flatworm species. In vivo, F. hepatica inhabit extremely caustic environments where the rapidity of growth and development is uncompromisingly vital and tissue damage commonplace. The polychotomy of expression patterns observed here is, therefore, unsurprising and may also be reflective of the involvement of Wnt signalling in the maintenance of stem cell niches [59–61].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Schematic summarising the diverse expression sites of all Wnt/β-catenin pathway components successfully localised in juvenile Fasciola hepatica using fluorescence in situ hybridisation.

https://doi.org/10.1371/journal.ppat.1012562.g002

As anticipated due to its pleiotropic role, the expression of FhCTNNB1 appears ubiquitous, localising throughout the dorso-ventral axis. FhDSH3.1 localisation revealed expression within a population of small cells, uniform in size along the anterior/posterior (A/P) axis of the worm and, accordingly, this transcript may be involved in the establishment/ maintenance of the A/P axis during development. The antagonistic FhSFRP2 transcript localisation uncovered a remarkably symmetrical pattern, localising to discrete cells, likely muscle cell bodies, around the lateral edges of the worm in addition to a single cell above and below the ventral sucker. Although in many cases, Wnt targets localised in close proximity to EdU+ cells, the co-localisation of target Wnt pathway component transcripts and EdU within the same cell was not observed.

As their expression profiles are akin to FhWnts, the failure to localise FhFZDs remains somewhat of an enigma. This discrepancy may be the result of secondary structures within FhFZD mRNAs which may hinder RNA probe binding. However, as Wnt protein ligands exhibit limited diffusion in the extracellular environment and tend to act on neighbouring cells [62], it would seem logical to hypothesise that FhFZD receptor loci and expression patterns would largely mirror those of their cognate FhWnt ligands.

RNA interference of Fasciola hepatica Wnt pathway components

In order to probe the role of Wnt signalling in juvenile F. hepatica, 18 canonical pathway components were selected for functional characterisation using RNAi-mediated gene silencing. Targets were prioritised based upon reports of differential expression in select transcriptomic datasets, plausibility as druggable targets and the existence of reported aberrant phenotypes in other organisms (see S3 Table for full list of RNAi targets and orthologues). All targets investigated were amenable to RNAi, as evidenced by significant target transcript knockdown relative to no dsRNA controls (S4 Fig). It should be noted that due to consistent, significant die-off in FhCTNNB1-silenced juveniles during the fourth week of RNAi trials, extractions were instead performed following 3 weeks of dsRNA exposures prior to the onset of mass death. Of the 18 targets silenced, 13 produced measurable growth phenotypes. With regards to FhWnts, knockdown led to significant reductions in worm growth for all five genes, evident from week one in the experimental setup (Fig 3A, n = 3; FhWNT1, p = 0.0253; FhWNT2B, p = 0.0359; FhWNT4, p = < 0.0001; FhWNT5A, p = 0.048; FhWNT9A, p = 0.0274). This inhibition of growth was sustained for the full duration of the trial, with FhWNT4 yielding the most dramatic growth phenotype following four weeks of silencing (Fig 3A, p = <0.0001). Interestingly, FhWNT4 was also the most significantly differentially expressed WNT in the in vitro vs in vivo transcriptomes [26]. In regenerating planaria, the silencing of Wnt protein ligands produces a plethora of aberrant phenotypes, including two-headed organisms, those void of tails, ectopic pharynges and the disappearance of both the anterior-posterior (A/P) and mediolateral axes [15,58], illustrating functional diversity in platyhelminths. It is, therefore, unsurprising that the silencing of FhWnts profoundly inhibited the growth trajectories of juvenile F. hepatica.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Silencing juvenile Fasciola hepatica Wnt/β-catenin pathway component gene transcripts dysregulates growth.

Growth of juvenile F. hepatica following four weeks of RNAi-mediated silencing of A) FhWnts B) FhFrizzleds in addition to light microscope images, C) FhDishevelled and D) FhAdenomatous polyposis coli. Worm area measured in μm², data presented as μm²±SEM (n = 3). Scale = 100 µm. Statistical analyses were performed using Kruskal Wallis with Dunn’s post hoc tests. *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

https://doi.org/10.1371/journal.ppat.1012562.g003

Wnt protein ligands promote the progression of myogenic progenitors throughout proliferative expansion and determine their terminal differentiation as both muscle cells and myogenic stem cells [63,64]. They do so by inducing the transcriptional activation of myogenic regulatory factors including Myf5 and MyoD [63] and as such, are heavily involved in myogenesis. F. hepatica possess three distinct types of muscle fibre (longitudinal, circular and diagonal) of which the major organ systems including the sub-tegumental musculature, gut, reproductive and attachment apparatus are comprised [65]. Indeed, the highly organised musculature constitutes the bulk of the worm’s mass. As such, the growth inhibition observed in FhWnt-silenced juveniles may principally be a consequence of dysregulated myogenesis.

As with FhWnts, a significant inhibition of growth was observed following the silencing of each FhFZD. Unlike FhWnts, only one FhFZD (FhFZD5) exhibited a significant reduction in growth following one week of dsRNA exposures, however at week two, the significant inhibition of growth was observed in all FhFZD targets (Fig 3B). While the level of significance fluctuated in the cases of FhFZD5 and FhFZD5.1, reduced growth was still evident for all targets at week four of the experimental setup (n = 3, FhFZD1: p = < 0.0001; FhFZD4: p = < 0.0001; FhFZD5: p = 0.0247; FhFZD5.1: p = 0.0463; FhFZD8, p = < 0.0008). As the activation of FZD GPCRs is contingent upon their association with a Wnt protein ligand, it is probable that the silencing of FhFZDs phenocopied that of FhWnts and vice versa. The apparent absence of functional redundancy in FhWnts and FhFZDs is an observation corroborated by the stark partitioning of FhWnt transcript expression observed in FISH.

Of two Dishevelled (DSH) orthologues, a reduced growth phenotype was only evident in FhDSH3.1-RNAi worms (Fig 3C). DSH regulates and channels Wnt signals into the various pathway branches and as such, the silencing of FhDSH.3 likely inhibited growth via the disorganised funnelling of FhWnts into the canonical, or indeed, non-canonical branches. The localisation of FhDSH3.1 transcripts along the anterior/posterior axis is also suggestive of a role in the maintenance of A/P identity in F. hepatica, as is the case in S. mediterranea [58]. FhAPC was the only target silenced to yield an increased growth phenotype following four weeks of silencing (Fig 3D, n = 3, p = 0.0190). As APC ordinarily serves as a tumour suppressor and negative regulator of Wnt/β-catenin signalling [9], its knockdown would allow cytoplasmic β-catenin to accumulate, aberrantly activating Wnt signalling and, ultimately, juvenile F. hepatica growth. The importance of APC has also been demonstrated in regenerating planarians where silencing results in the formation of a tail in place of a head [13].

Taken alongside the observed reductions in growth exhibited by FhWnt/ β-catenin pathway component-silenced worms, it was hypothesised that neoblast-like cell activity would also be significantly reduced. To investigate this hypothesis, silenced juveniles were incubated in EdU for 24 h alongside those exposed to control dsRNA (Fig 4A). This revealed that FhWNT1- and FhWNT2B-silenced juveniles displayed significantly reduced EdU+ cell numbers compared to control dsRNA-treated juveniles (Fig 4D, p = 0.0009 and p = 0.0085, respectively). In other systems, Wnt protein ligands serve as niche factors [66], stimulating both the proliferation and self-renewal potential of tissue-specific stem cells [61]. Further, the silencing of β-catenin inhibits the proliferation of cancer stem cells by inducing cell cycle arrest [67–69]. Such functions appear to be conserved in F. hepatica. Of FhFZD-silenced juveniles (Fig 4B), only FhFZD1 and FhFZD4 caused significant reductions in neoblast-like cell proliferation (Fig 4D, p = 0.0478 and p = 0.0152, respectively). As the results of this experiment mirrored those of FhWnts, in that silencing of two of five family members led to reduced cell proliferation, it may be hypothesised that FhFZD1 and FhFZD4 constitute the receptors of FhWNT1 and FhWNT2B protein ligands. Despite silencing yielding growth phenotypes, neither FhDSH3.1 or FhAPC affected neoblast-like cell proliferation (S6 Fig). This appears somewhat surprising given that neoblast-like cells are proposed to drive growth and development in F. hepatica and related species [29,48,70]. In addition to its role in cell proliferation, Wnt signalling also mediates cell differentiation and fate determination, therefore, silencing the remaining FhWnts, FhFZDs and FhDSH3, may result in fewer cells of a specific cluster, manifesting as reduced growth in juvenile worms.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. The silencing of selected juvenile Fasciola hepatica Wnt/β-catenin pathway component gene transcripts dysregulates neoblast-like stem cell proliferation.

Maximally projected confocal z-stack images of EdU+ nuclei (green) in juvenile F. hepatica showing reduced cell proliferation following four weeks of silencing A) FhWnts, B) FhFrizzleds and C) FhCTNNB1. DAPI (blue) served as a counterstain. Scale = 50 µm. D) Neoblast-like cell proliferation in FhWnt pathway component silenced juveniles yielding significant growth phenotypes relative to dsRNA control groups. Data presented as mean EdU+ nuclei ± SEM (n = 3). Statistical analyses were performed using a one-way ANOVA with Dunnett’s post hoc tests. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

https://doi.org/10.1371/journal.ppat.1012562.g004

As Wnt/β-catenin signalling is negatively regulated by the presence of a destruction complex, it was hypothesised that silencing components of this complex would remove regulatory pressure, thereby leading to increased growth. Despite significant levels of transcript knockdown, no growth phenotypes were observed in any of the destruction complex targets, with the exception of FhAPC (Fig 3D and S5 Fig). The most common mutation preceding the development of colon cancer is the inactivation of APC which serves as a tumour suppressor via the modulation of cytoplasmic β-catenin levels. This, coupled with the fact that APC orchestrates the Wnt/β-catenin destruction complex assembly and function, may explain why it was the sole destruction complex component yielding a phenotype. With regards to FhGSK3B, the lack of phenotypes arising from silencing may be explained relatively simply in that the generation of life stage expression profiles for FhWnt/β-catenin pathway components revealed that members of the destruction complex exhibit their lowest levels of expression in early juvenile stage fluke. As the RNAi screen was performed on early juvenile stage fluke, knocking down the transcripts of FhGSK3Bs further had no discernible effect. It is also plausible that the lack of phenotypic response observed in FhGSK3B-silenced juveniles was a result of functional redundancy.

Following FISH, a juxtaposition was identified between the expression patterns of FhSFRP2 and FhWNT9A, perhaps suggestive of a functional relationship. While neither FhSFRP agonists yielded growth phenotypes, FhSFRP2-silenced juveniles exhibited a significant increase in motility (S5 Fig, n = 3, 199.1±22.65%, p = < 0.0001). However, it has been suggested that SFRPs may not serve as Wnt antagonists in parasitic flatworms due to the absence of a netrin-like domain required for their anchorage to the extracellular matrix [19]. As such, the observed increase in motility may be the result of an unknown function executed by FhSFRP2 in F. hepatica.

FhCTNNB1 is essential for the development of the juvenile Fasciola hepatica neuromuscular system

The phenotypic consequences of silencing FhCTNNB1 were undoubtedly the most profound observed in this study. This is supported by the ubiquitous expression of the FhCTNNB1 transcript, an observation also reported in planarians [71,72]. Significant reductions in worm growth were evident in two-week-old juveniles, an effect that was further exaggerated at weeks three and four (Fig 5A, n = 3, p = 0.0009, p = 0.0009, p = 0.0003, respectively), accompanied by concomitant reductions in neoblast-like cell proliferation (Fig 4C, p = 0.0020). Interestingly, the silencing of FhCTNNB1 also resulted in significant fluke die off, with average survival being just 19% at week four (Fig 5B, n = 3, 19±5.85%, p = < 0.0001). Worm death was defined by a darkened appearance, gut oedema and total loss of tissue integrity within 24 h of the most recent dsRNA exposure (Fig 5C). As the key effector of the canonical signalling pathway, β-catenin regulates gene transcription and triggers conformational changes in chromatin [73,74], thereby acting as a master regulator, fulfilling a myriad of context-dependent roles during development, regeneration and homeostasis. In vertebrates, the most prominent role of β-catenin during development is in the anteroposterior patterning of the central nervous system [75].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. FhCTNNB1 silencing in juvenile Fasciola hepatica dysregulates muscle/nervous system structure/development, inhibits growth and reduces survival

A) Growth following four weeks of RNAi-mediated FhCTNNB1-silencing. Worm area measured in μm², data presented as μm²±SEM (n = 3). B) Survival. Data show mean survival (±SEM) relative to untreated control (RPMI) juveniles. Statistical analyses were performed using unpaired t-tests and Kruskal Wallis with Dunn’s post hoc tests. ***, p < 0.001; ****, p < 0.0001. C) Light microscope images demonstrating the morphological features of FhCTNNB1-silencing induced death in a three-week old FhCTNNB1-silenced juvenile compared to a time-matched control dsRNA-treated juvenile. D) Confocal scanning laser micrographs of wholemount dsRNA control and FhCTNNB1 silenced three-week old F. hepatica subjected to ICC. Green staining denotes NPF immunoreactivity, highlighting the nervous system, while TRITC-labelled phalloidin (red) highlights filamentous-actin. Staining patterns shown in D and E were consistent across all 10 juveniles stained. CG, cerebral ganglia; VNC, ventral nerve cord; OS, oral sucker; VS, ventral sucker. E) Phalloidin-labelling of the filamentous-actin structure of the musculature comprising the ventral suckers (red) of dsRNA control and FhCTNNB1 silenced three-week old F. hepatica. Scale = 50 µm.

https://doi.org/10.1371/journal.ppat.1012562.g005

To investigate the phenotypic consequences of FhCTNNB1 knockdown in greater detail, silenced and control specimens were subjected to neuropeptide F (NPF) immunocytochemistry (ICC) and fluorophore-phalloidin staining in order to visualise a significant component of the peptidergic nervous system and musculature arrangements, respectively. FhCTNNB1-silenced worms were prepared for ICC at three weeks old prior to the onset of mass die off. Indeed, FhCTNNB1-silenced juveniles exhibited aberrant, dysregulated development of the central nervous system, with unusually dense NPF-immunoreactivity in both the cerebral ganglia and ventral nerve cords, and the formation of spherical, ectopic neuronal cell bodies (Fig 5D). Such observations could be deemed consistent with hypercephalisation, a phenomenon also reported in S. mediterranea, where the silencing of CTNNB1 produces organisms exhibiting ectopic eyes, radial-like hypercephalisation and a loss of A/P identity [71].

Phalloidin staining of F-actin revealed an equally striking visual phenotype, demonstrating a loss of filamentous-actin patterning and structural integrity within the musculature of FhCTNNB1 silenced juveniles (Fig 5D). This was particularly evident when examining the highly muscular oral and ventral suckers (Fig 5E). In addition to its role in signal transduction, β-catenin also serves as a scaffold molecule and is a key constituent of cadherin-based adherens junctions in higher organisms [76]. The adhesive competence of adherens junctions is dependent upon members of the cadherin superfamily which bind homotypically to complementary cadherin molecules on neighbouring cells [77]. The direct interaction of β-catenin with the cytoplasmic domain of cadherins prompts the establishment of a complex with α-catenin and the subsequent linkage of actin filaments to adherens junctions, forming a structural continuum [78]. In this respect, these specialised junctions assist in the formation of the polarised epithelial tissues required for morphogenesis and organism integrity [79,80]. As such the silencing of FhCTNNB1 may preclude the formation of cellular junctions, leading to an imbalance in the structural properties of the cytoskeleton and thereby, loss of tissue integrity and death. It should also be mentioned that β-catenin specialisation has been reported in S. mediterranea whereby Smed-β-catenin-1 mediates Wnt signalling while Smed-β-catenin-2 functions in cell adhesion [39], however, evidence for such functional segregation in parasitic species is lacking.

While this study presents both transcriptomic and phenotypic evidence lending support to the hypothesis that Wnt signalling is a key molecular driver of growth and development in juvenile F. hepatica, similar studies in related species oppose such findings. A large-scale RNAi screen performed in S. mansoni silenced four FZD orthologues (Smp_118970, Smp_139180, Smp_155340, Smp_174350), in addition to a β-catenin-like protein (Smp_089400) with no measurable phenotypes following 30 days of intermittent dsRNA exposures [81]. This appears somewhat peculiar given the close phylogenetic relationship between F. hepatica and S. mansoni. Quite simply, this apparent disparity may be a consequence of the RNAi screen of Wang et al employing fully developed, adult parasites. Further, target expression analyses were not conducted, therefore, the absence of phenotypic readouts could be due to the absence of, or modest transcript and/or protein knockdown. As schistosomulae are amenable to RNAi [82,83], silencing Wnt pathway components in this earlier life stage may yield tangible phenotypes.

Pharmacological inhibition of FhWnt/β-catenin pathway components

Due to the identification of robust links between aberrant Wnt signalling and tumorigenesis, recent years have observed a concerted effort within the field of oncology to exploit small-molecule inhibitors targeting various Wnt pathway components as novel anti-cancer treatments. Pyrvinium pamoate (PP) is an FDA-approved anthelmintic, originally employed in the treatment of Enterobius vermicularis (pinworms). More recently, PP was found to promote the degradation of β-catenin by selectively potentiating casein kinase 1α activity in various cancer cell lines, inhibiting tumour growth via the downregulation of Wnt signalling [84,85]. Similarly, biweekly exposures to PP significantly impeded growth and neoblast-like cell proliferation in juvenile F. hepatica (Fig 6B and 6C, n = 3, p < 0.0001, Fig 6E, n = 3, p = 0.0114, respectively). After just one week, the degree of growth inhibition observed in PP-treated juveniles paralleled that of TCBZ. Moreover, concentrations greater than 0.1 µM resulted in a loss of tissue integrity and proved lethal. Due to significant die off in 0.1 µM PP-treated juveniles during week three, the experiment had to be terminated. PP-induced lethalities have also been reported in S. mansoni schistosomula [86].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Selected Wnt/β-catenin pathway antagonists dysregulate neoblast-like cell proliferation and growth in juvenile Fasciola hepatica.

A) Growth of juvenile F. hepatica following biweekly exposures to A) FZM1 and B) Pyrvinium pamoate (PP). TCBZ served as a positive control. Worm area measured in μm², data presented as μm²±SEM (n = 3). Statistical analyses were performed using Ordinary One-way ANOVA or Kruskal Wallis with Dunn’s post hoc tests. *, p < 0.05; **, p < 0.01; ****, p < 0.0001. C) Light microscope images showing size and aberrant morphology of two-week old 0.1 µM PP-treated juveniles relative to DMSO controls. D) Neoblast-like cell proliferation in juvenile Fasciola hepatica following treatment with PP and FZM1 relative to DMSO controls. Data presented as mean EdU+ nuclei ± SEM. Statistical analyses were performed using a one-way ANOVA with Dunnett’s post hoc tests. *, p < 0.05. E) Maximally projected confocal z-stack images of EdU+ nuclei (green) in DMSO control and PP-treated two-week old juvenile F. hepatica. DAPI served as a counterstain (blue). Scale = 50 µm.

https://doi.org/10.1371/journal.ppat.1012562.g006

The inhibition of Wnt signalling is known to interfere with the recruitment of stem cells to wound sites following mechanical injury [87,88] and disrupts their proliferative potential during the healing process [89]. Moreover, it has been demonstrated that TCBZ enters F. hepatica tissues via trans-tegumental diffusion [90], with tegumental damage proving fatal [91]. If PP does indeed inhibit Wnt signalling in F. hepatica, the dysregulation of neoblast-like cells coupled with compromised musculature could render worms less equipped to defend themselves from drug action.

While we observed reductions in both growth and cell proliferation following PP treatment, two-week-old PP-treated worms did not phenocopy those of FhCTNNB1-silenced juveniles with regards to aberrant nervous system development (S7 Fig). Although not unequivocally determined, PP is postulated to inhibit glucose uptake and mitochondrial fumarate reductase in pinworms [92]. Since Wnt activation has been shown to modulate mitochondrial function which, in a feedforward loop, helps to drive Wnt signalling in human cells [93], the possible dysregulation of multiple pathways and processes contributing to the observed PP-induced phenotype in F. hepatica cannot be ruled out. Biweekly exposures to FZM1, a FZD4 inhibitor also significantly reduced juvenile fluke growth (Fig 6A, n = 3, p = 0.0066), again, phenocopying the effects of RNAi-mediated silencing of FhFZD4. FZM1 binds to an allosteric binding site within intracellular loop 3 of FZD4, resulting in conformational changes within the receptor which ultimately inhibit Wnt/β-catenin signalling [94].

In silico characterisation of Fasciola hepatica Wnt signalling pathways

Putative homologues of all major components of both canonical and non-canonical Wnt signalling pathways were identified in F. hepatica using a reciprocal BLAST (basic local alignment search tool) approach. Sequences from Homo sapiens, Mus musculus, Drosophila melanogaster and Caenorhabditis elegans were obtained from Uniprot for each gene of interest (see S1 Data for query sequence gene accessions) and queried against the F. hepatica genomes (PRJEB25283/PRJEB58756 and PRJNA179522) available on WormBase ParaSite v16 using BLASTp and tBLASTn with default parameters. Returned hits (sorted by E-value) were subsequently screened against the NCBI non-redundant (nr) protein sequence database using BLASTp to validate sequence identity. The organism field was set to exclude Platyhelminthes. Where the top hit was not the gene of interest, that sequence was omitted from the dataset.

As an additional measure of confidence, putative F. hepatica Wnt, frizzled (FZD), β-catenin, dishevelled (DSH) and glycogen synthase kinase 3 (GSK3) protein sequences were aligned with those of H. sapiens, M. musculus, D. melanogaster, S. mediterranea and S. mansoni orthologues in Clustal Omega using ClustalW with character counts. The integrity and classification of homologues was further examined through the identification of conserved domains with the aid of Pfam HMMScan (https://www.ebi.ac.uk/Tools/hmmer/search/hmmscan) and Interpro (www.ebi.ac.uk/interpro/search/sequence-search). The presence of functional domains and characteristic motifs were considered to be evidence of gene conservation and functional viability.

To gain insight into the degree of expression of various pathway components and thus, the relative importance of Wnt signalling throughout the F. hepatica life cycle, the life-stage transcriptomic data supporting the F. hepatica genome project was mined (available on WormBase Parasite v16; egg (n = 1), metacercariae (n = 3), 1-hour newly excysted juvenile (NEJ, n = 2), 3-hour NEJ (n = 2), 24-hour NEJ (n = 2), 21-day old juvenile (n = 1) and adult-stage (n = 1) F. hepatica) to obtain median transcripts per million-mapped reads (TPM) values for all F. hepatica canonical and non-canonical Wnt pathway components. TPM values were subsequently transformed to a Log2 scale and uploaded to heatmapper (http://www.heatmapper.ca/) [95] set for Average Linkage, and Pearson Distance Measurement to generate a life stage expression heatmap.

Excystment and in vitro maintenance of juvenile Fasciola hepatica

Experiments were carried out using the Italian (TCBZ-susceptible) F. hepatica isolate (Ridgeway Research Ltd, UK). Encysted metacercariae were excysted as described by McVeigh et al. [96]. Our excystment protocol is available, in full at https://www.protocols.io/view/in-vitro-excystment-of-juvenile-fasciola-hepatica-14egn212qg5d/v1. In short, encased metacercariae were popped from their outer cyst walls and treated with 10% sodium hypochlorite for 2-3 min prior to the addition of excystment solution. Metacercariae were incubated in excystment solution at 37°C for one hour. Newly excysted juveniles (NEJs) were washed in warmed RPMI 1640 (#11835105, ThermoFisher Scientific), before being transferred to either 40 µl of dsRNA (if destined for RNAi experiments) or 200 µl of 50% chicken serum (#16110082, Thermo Fisher Scientific) in RPMI (v/v) (CS50) in round bottomed, 96 well plates (Sarstedt). Juvenile fluke were maintained for the duration of experiments in groups of ~20 worms per well in a humidified incubator set to 37°C with 5% CO₂ atmosphere. Culture medium was replaced every three days.

Whole mount fluorescence in situ hybridisation (FISH)

Juvenile worms were excysted as above and maintained under standard culture conditions for four weeks. On day 27 of culture, worms were incubated in 500 µM 5-ethynyl-2′-deoxyuridine (EdU) in CS50 for 24 h to enable the co-localisation of FISH targets with EdU⁺ (neoblast-like) cells. At four weeks, worms were washed in RPMI and anaesthetised via a three-minute incubation in 0.25% tricaine (ethyl 3-aminobenzoate methanesulfonate, #A5040, Sigma-Aldrich). From this point, RNase-free pipette tips were used, and reagents were made using diethyl pyrocarbonate (DEPC)-treated RNase-free water (this is applicable to all steps until post Riboprobe incubation). Worms were transferred to a 20 µl droplet of 4% formaldehyde (36% methanol formaldehyde [Sigma Aldrich], diluted in PBSTx (1x phosphate buffered saline, PBS) with 0.3% Triton X-100 [Sigma-Aldrich]) in a petri dish and flat fixed under a weighted coverslip for 10 min. Following removal of the coverslip, worms were collected into NoStick hydrophobic microtubes (#1210-S0, SSIbio) containing 1 ml 4% formaldehyde, and free fixed for a further 10 min. After fixation, worms were washed in PBSTx before being dehydrated via a 10-min incubation in 1:1 PBSTx:Methanol (MeOH). Worms were stored in 100% MeOH at -20°C until use. All incubation and wash steps were carried out under rotation at room temperature (RT).

The synthesis of DIG-labelled RNA probes was achieved through in vitro transcription reactions. Amplicons labelled with the T7 promoter sequence; 5’-TAATACGACTCACTAT AGGGT-3’ were generated with end-point PCRs using FastStart Taq Polymerase (#4738357001, Sigma Aldrich) and purified using the ChargeSwitch PCR Clean-Up Kit (#CS12000, ThermoFisher Scientific). Primer sequences can be found in S2 Table. Transcription reactions consisted of 2 µl transcription buffer, 1 µl T7 polymerase (T7 RNA polymerase kit, ThermoFisher Scientific), 1 µl digoxigenin-labelled ribonucleotide mix (Roche) and 6 µl DNA template. Probes were then incubated overnight at 27°C before being DNase treated for 15 min at 37°C (0.25 µl DNase I, 0.25 µl Transcription Buffer and 2 µl RNase-free H₂O). Probes were precipitated overnight at -80°C with the addition of 1.25 µl 4 M LiCl and 37.6 µl chilled molecular grade ethanol [97]. Post-precipitation, RNA was pelleted via a 20-min centrifugation (x16,000 g at 4°C). The pellet was washed in 70% chilled molecular grade ethanol and centrifuged for a further 5 min before discarding the supernatant and allowing to air dry for 15 min. The pellet was then resuspended in 20 µl RNase-free H₂O and probe concentrations and purities were measured using a DeNovix DS-11 FX spectrophotometer/fluorometer. Concentrations were adjusted to 50 ng/µl via the addition of hybridisation solution (25 ml de-ionized formamide, 12.5 ml 20X SSC, 100 µl of 50 mg/ml yeast RNA, 5 ml 10% (v/v) Tween-20, 5 ml 50% Dextran Sulfate [Sigma Aldrich] and 2.4 ml DEPC-treated H₂O) and probes were stored at -20°C until use. Yeast RNA was generated as described by Jing [98].

The core FISH protocol employed here (including the preparation of all stock solutions) was that detailed by King & Newmark [97]. The only modifications made were in the treatment of parasites pre-hybridisation and are as follows; worms were rehydrated in 50:50 PBSTx:MeOH for 10 min at RT, followed by a further 10 min incubation in PBSTx. Worms were then incubated in 1 ml bleaching solution (1.77 ml H₂O, 100 µl formamide, 50 µl 20X SSC, 80 µl 30% hydrogen peroxide; all Sigma Aldrich) for 1 h under bright light [99]. The remainder of the protocol was carried out as described by King & Newmark [98] and is outlined below. A full protocol is available at https://www.protocols.io/view/fluorescent-in-situ-hybridisation-for-juvenile-fas-3byl4q7xjvo5/v1. Following the removal of bleaching solution, worms were rinsed in 1x SSC (20x SSC diluted to 1x in DEPC treated H₂O) for 10 min before being rinsed in PBSTx for 10 min on rotation. To permeabilise the worms, they were subjected to a 15 min incubation in 1 ml of proteinase K solution (50 µl 10% SDS and 2.75 µl of 20 mg/ml proteinase K [Sigma Aldrich] in 4.95 ml PBSTx) prior to being postfixed in 4% formaldehyde for 10 min on rotation. Worms were then washed for a further 10 min in PBSTx, on rotation, before being transferred to 1:1 PBSTx:PreHybridisation solution (Prehyb: 25 ml de-ionized formamide, 12.5 ml 20X SSC, 100 µl of 50 mg/ml yeast RNA, 5 ml 10% (v/v) Tween-20 [Sigma Aldrich], and 7.4 ml DEPC-treated H₂O) in 35 µm incubation baskets (Intavis) within an 18 well plate and incubated on a shaker at RT for 10 min. Baskets containing worms were then transferred to wells containing PreHybridisation solution before being placed in a hybridisation oven where they were incubated at 52^oC for 2 h. Following this 2 h incubation, worms were hybridised by transferring baskets to wells containing Riboprobe mix (20 µl of 50 ng/µl target riboprobe was heated to 80°C for 5 min and placed on ice prior to the addition of 980 µl of hybridisation solution) and incubating overnight at 52^oC in a Boekel Scientific Shake’N Bake hybridisation oven.

The following morning, 500 µl of Riboprobe mix was removed from wells and replaced with 500 µl of preheated 2X SSCx (20X SSC stock diluted to 2X in ddH₂O + 0.1% Triton X-100) in which worms were incubated for 20 min. Baskets were then transferred to new wells containing 2X SSCx and incubated for 20 min. This was repeated an additional 2x. Worms then underwent 4x 20 min washes in 0.2X SSCx (20X SSC stock diluted to 0.2X in ddH₂O + 0.1% Triton X-100). All SSC washes were carried out at 52^oC. Following the final 0.2X SSCx wash, baskets were transferred to wells containing TNTx (1 L ddH₂O, 12.11 g Tris Base, 8.77 g NaCl, 3 ml Triton X-100, pH 7.5) and incubated for 10 min x2 on a shaker at RT before being transferred to wells containing blocking solution (500 µl horse serum and 500 µl Roche Western Blocking Reagent [both Sigma Aldrich] diluted in 9 ml TNTx), in which they were incubated at RT for 1 h. Baskets were then transferred to wells containing antibody solution (anti-DIG-POD [Sigma Aldrich] diluted in blocking solution, 1:2000), and incubated overnight at 4°C.

Worms were washed in TNTx for 5 min, 10 min and 6x for 20 min at RT via the transfer of baskets to new wells prior to a 10 min incubation in freshly made tyramide solution (1 ml tyramide signal amplification (TSA) buffer [2 M NaCl; 0.1 M Boric acid, pH 8.5], 6 µl 5% H₂O₂, 1 µl 4 IPBA [20 mg 4-iodophenylboronic acid in 1 ml dimethylformamide (DMF)], 2 µl TAMRA). The fluor-conjugated NHS ester employed in the preparation of tyramide conjugates was TAMRA (5-(and-6)-carboxytetramethylrhodamine, Sigma Aldrich).

During tyramide incubation and for the remainder of the protocol, plates were protected from light. Post tyramide incubation, EdU detection and DAPI staining were performed as detailed below. During all wash/incubation steps post SSCx washes, plates containing baskets of worms were placed on a shaker at 100 rpm. Worms were mounted in 10 µl Vectashield (Vector Laboratories) on standard microscope slides and coverslips were sealed with clear nail polish. Only those RNAi targets yielding phenotypes were subjected to FISH.

RNA interference (RNAi)

Double-stranded (ds)RNA templates specific for each target gene were amplified from juvenile F. hepatica cDNA using primers labelled with the T7 promoter sequence; 5’-TAATACGACTCACTATAGGGT-3’. Primers were designed using Primer3Plus and amplicons were later sequence confirmed by Eurofins Genomics (https://eurofinsgenomics.eu/en/). Primer sequences used to generate dsRNA templates can be found in S2 Table. Amplicons were generated using end-point PCRs (as above) and sizes confirmed on a 2% agarose gel. dsRNA templates were purified using the ChargeSwitch PCR Clean-Up Kit (#CS12000, ThermoFisher Scientific) and dsRNA synthesis performed using the T7 RiboMAX Express RNA System (#P1700, Promega) according to manufacturer’s instructions. The resulting dsRNA constructs were re-suspended in nuclease free H₂O and their RNA concentrations and purities measured using a DeNovix DS-11 FX spectrophotometer/fluorometer. All dsRNAs were stored at -20 °C as single use aliquots.

All RNAi experiments were performed in triplicate. Worms were soaked in 100 ng/µl of target (or control) dsRNA in 50 µl RPMI reactions for 24 h under standard culture conditions. dsRNA exposures were repeated biweekly for four weeks from the date of excystment. Between exposures, worms were returned to CS50 and maintained as standard. All RNAi experimental setups included an untreated control group (no dsRNA), while a bacterial dsRNA template (neomycin phosphotransferase [U55762]) served as a negative control. Upon completion of trials, juveniles were transferred to 2 ml round-bottomed Eppendorf tubes and snap frozen in liquid nitrogen prior to messenger (m)RNA extraction and reverse transcription.

Quantification of RNAi target transcript knockdown

Frozen samples from each treatment group were lysed using a Qiagen TissueLyser LT at 50 oscillations/minute for one minute. Poly-adenylated mRNA was then extracted from each sample using the Dynabeads mRNA Direct Kit (#61012, ThermoFisher Scientific) before being treated with DNase (Turbo DNA-free Kit, #AM1907, ThermoFisher Scientific) and reverse transcribed to cDNA (High-Capacity RNA-to-cDNA Kit, #4387406, ThermoFisher Scientific). All cDNAs were diluted 1:1 in nuclease-free H₂O prior to use. Target transcripts were amplified via qPCR on a Qiagen RotorGene Q 5-plex HRM instrument using the following cycling parameters: 95^oC for 10 min, followed by 40 cycles of 95^oC 10 s, 60^oC 15 s and 72^oC for 30 s. qPCRs were performed using 10 µl reactions consisting of 5 μl SensiFAST SBYR No-ROX Kit (#BIO-98005, Bioline), 0.2 μM of each primer and 2 μl of relevant cDNA (or H₂O in the case of no template controls). Primers used in qPCR reactions were tested prior to use to ensure efficiencies of ≥1.7. All reactions were performed in triplicate, including no-template controls. F. hepatica glyceraldehyde 3-phosphate dehydrogenase (FhGAPDH, [AY005475]) served as a reference gene. Melt-curve analyses were employed as standard.

Pfaffl’s augmented ΔΔCt method [100] was employed to calculate relative gene expression. Ratios of target transcript abundance relative to the untreated control were then converted to percentage transcript expression where 100% represents no change. Transcript expression was plotted for both target and control-dsRNA-treated groups relative to no dsRNA control.

Compound exposures

Pyrvinium pamoate (#HY-A0293, MedChem Express), FZM1 (1-(3-Hydroxy-5-(thiophen-2-yl)phenyl)-3-(naphthalen-2-yl)urea, #5343580001, Sigma Aldrich) and triclabendazole (TCBZ; #1681611, Sigma Aldrich) were dissolved to the desired concentrations using dimethyl sulphoxide (DMSO, #D2650, Sigma Aldrich) within NoStick hydrophobic microtubes. All compound exposures were carried out in 3 ml reactions within 35x10 mm Petri dishes (#82.1135.500, Sarstedt), with compounds being added at 1:1000, such that the final concentration of DMSO was 0.1% (v/v). All assays were performed alongside TCBZ-treated and vehicle (DMSO)-treated positive and negative control groups, respectively. Parasites were exposed to compounds for 18-hours biweekly from the date of excystment for three weeks under standard culture conditions. Prior to exposures, worms were washed in RPMI to remove excess chicken serum. Phenotypic observations were recorded immediately after the second exposure each week. Post compound exposures, worms were washed in RPMI to remove any residual drug before being returned to CS50 and cultured as standard. Following the final exposure, worms from treatment groups exhibiting growth phenotypes were incubated in EdU for 24 h to allow for the visualisation of any effects on neoblast-like cell proliferation.

Labelling neoblast-like cells using 5-ethynyl-2′-deoxyuridine (EdU)

Following the final dsRNA exposure on day 28 of RNAi experiments, or after final compound exposures, a cohort of worms from control groups and each treatment group producing a growth phenotype were incubated in 500 µM EdU (dissolved in 1x PBS) in CS50 under standard culture conditions for 24 h. Worms were then washed in RPMI and transferred to a 20 µl droplet of 4% paraformaldehyde (PFA) in PBS within a Petri dish and flat fixed under a weighted coverslip for 10 min. Following removal of the coverslip, worms were collected into NoStick hydrophobic microtubes containing 1 ml 4% PFA, and free fixed O/N at 4°C, while under constant rotation.

EdU detection was performed using a modified version of an S. mansoni EdU staining protocol [101]. Fixed, EdU-labelled worms were permeabilised for 30 min in PBSTx (PBS containing 0.5% Triton-X-100). EdU detection was performed via a 30-min incubation in a 6-carboxyfluorescein (6-FAM) azide conjugate (Metabion) solution while protected from light. Worms were then washed 2x in PBSTx and counterstained via a 20-min incubation in 1:1000 4′,6-diamidino-2-phenylindole (DAPI, 1 mg/ml, Sigma Aldrich) in PBS, again, while protected from light. Two final 5-min washes were performed in PBSTx prior to mounting (as previously described). All steps were performed under constant rotation at RT. A full version of the protocol is available at https://www.protocols.io/view/edu-staining-protocol-in-juvenile-f-hepatica-eq2lyjnrrlx9/v1.

Immunocytochemistry (ICC)

Following the final dsRNA exposure on day 28 of FhCTNNB1 RNAi experiments, 10 worms from each treatment group were processed for fixation (as above). ICC was performed using neuropeptide F (NPF) antiserum (raised against the C-terminal decapeptide-amide of NPF from Moniezia expansa [102]) and tetramethylrhodamine-isothiocyanate (TRITC)-conjugated phalloidin (#P1951, Sigma Aldrich) to enable the visualisation of a significant portion of the peptidergic nervous system and musculature, respectively. Worms were incubated in rabbit anti-NPF primary antiserum (1:1000) for 72 h, washed in antibody diluent (AbD; 0.1 M PBS, 0.1% (v/v) Triton X-100, 0.1% (w/v) bovine serum albumin) and incubated for a further 48 h in 1:100 fluorescein isothiocyanate (FITC)-labelled anti-rabbit secondary antiserum (Sigma Aldrich). A final overnight incubation in 1:100 TRITC-labelled phalloidin (200 ng/ml, Sigma Aldrich) was performed before three 15-min washes in AbD prior to mounting (as above). The dilution of antisera and phalloidin was achieved using AbD and all incubation periods were carried out under constant rotation at 4°C.

Phenotypic analyses

Parasite motility and size were recorded following the second dsRNA/compound exposure each week for the duration of RNAi and compound assays, respectively. Darkfield videos, each one minute in length, were captured using an Olympus SZX10 microscope with attached Olympus SC50 camera. Videos were analysed using the wrMTrck plugin for ImageJ (https://www.phage.dk/plugins/wrmtrck.html). Changes in worm length between frames provided values for individual length changes (µm/minute) and served as a measure of motility. The motility of individual worms was then normalised to the average movement of untreated (no-dsRNA) and vehicle (DMSO) controls for RNAi and compound assays, respectively. Values were presented as percentage motility relative to the relevant control. Area values (µm²) for individual worms were also obtained using wrMTrck.

At the conclusion of experiments, brightfield images of each treatment group were captured using a Leica M205 C stereomicroscope with attached Leica MC190 HD camera to document the presence of any aberrant, morphological changes following RNAi and compound exposures. Neoblast-like cell activity was quantified via the enumeration of EdU⁺ nuclei using the cell counter plugin for ImageJ (https://imagej.nih.gov/ij/plugins/cell-counter.html) which were then normalised to worm area, giving values per 1000 µm².

Confocal microscopy

All FISH, EdU-labelled and ICC specimens were imaged using a Leica TCS SP8 inverted confocal scanning laser microscope. Images were captured as maximally projected z-stacks, each generated from 30-40 optical sections between the dorsal and ventral surfaces of whole worms.

Statistical analyses

All statistical analyses were performed and graphs produced using GraphPad Prism 9 (La Jolla California USA, www.graphpad.com). Each dataset was first tested for normality. Where data were normally distributed, parametric tests including ANOVAs and t-tests were performed, while the nonparametric Kruskal-Wallis and Mann-Whitney tests were employed where data were not normally distributed. Post-hoc analyses were carried out using Dunn’s, Dunnett’s and Tukey’s multiple comparisons tests to identify differences between several groups.