The life cycle of a Liberian strain of S. mansoni (Bayer AG, Monheim, Germany) was maintained by infecting Biomphalaria glabrata snails as intermediate hosts and Syrian hamsters (Mesocricetus auratus) as final hosts in accordance with the European Convention for the protection of vertebrate animals used for Experimental and other Scientific Purposes (ETS, No. 123; revised Appendix A). The experiments have been approved by the Regional Council Giessen, Germany (V54-19c 20 15h 02 GI18/10). Single-sex (ss) and bisex (bs) worms were generated by monomiracidial and polymiracidial infection of the intermediate host, respectively (Grevelding, 1999). Adult bs and ss worms were recovered by hepatoportal perfusion of the final host 46 days (d; for bs) and 67 d (for ss) post infection, respectively. Worms were transferred to petri dishes containing M199 (3+) medium (Gibco, Germany), supplemented with 10% (v/v) newborn calf serum (NCS; Sigma-Aldrich, Germany), 1% (v/v) HEPES (Carl Roth, Germany, pH 7.4), and 1% (v/v) antibiotic-antimycotic solution (ABAM; CCPro, Germany) consisting of 10,000 U penicillin, 10 mg streptomycin and 25 µg amphotericin B per ml. The parasites were incubated at 37°C and 5% CO₂.

Three eIF4A isoforms of S. mansoni were identified by Basic Local Alignment Search Tool for proteins (BLASTp) using the three known human eIF4A isoform sequences. A multiple sequence alignment of human and S. mansoni eIF4A sequences was generated with Clustal Ω (Madeira et al., 2024) and visualized using ESPript 3.0 (Robert and Gouet, 2014) to analyze the parasite eIF4A sequences for the presence of conserved motifs described for DEAD-box RNA helicases (Li et al., 1999; Cordin et al., 2006). The phylogenetic tree was created for grouping three eIF4A isoforms of S. mansoni with their potential orthologs of other platyhelminths and selected model organisms. Appropriate sequences were obtained from NCBI protein Blast (Altschul et al., 1990), the ortholog finder of WormBase ParaSite (Howe et al., 2016, Howe et al., 2017), InterProScan (Blum et al., 2025), and the UniProt database (Consortium et al., 2025). The DEAD-box RNA helicases p68 of Mus musculus and Saccharomyces cerevisiae served as outgroup. Identified orthologs and paralogs with their respective accession numbers are listed in Supplementary Table S1. Multiple sequence alignments were performed using MAFFT v7.520 (Katoh and Standley, 2013) with the L-INS-i parameters for less than 1,000 sequences. The alignment was further refined with MUSCLE v5.1.0 (MUltiple Sequence Comparison by Log-Expectation; Edgar, 2004). Bayesian inference analysis was performed using MrBayes v3.2.7a (Ronquist et al., 2012) with the LG+I+G (Le and Gascuel + invariable sites + gamma distributed rate heterogeneity) model and fixed state frequencies. Twenty-five percent of the 1,000,000 generated trees were excluded as burn-in, whereas the remaining 75% were utilized to construct the majority-rule consensus tree. The generated phylogenetic tree was visualized by the interactive tool Tree Of Life (iTOL; Letunic and Bork, 2024).

For RNA isolation, S. mansoni couples were separated by incubation with 0.26% (w/v) ethyl 3-aminobenzoate methanesulfonate (Sigma-Aldrich, Germany) dissolved in M199 (3+) medium, as described elsewhere (Collins et al., 2013). Males and females were collected separately in 1.5 ml reaction tubes, and then washed twice with PBS (pH 7.4). Supernatants were replaced by DNA/RNA protection reagent (New England Biolabs, NEB; UK) and the worms immediately frozen in liquid nitrogen and stored at -80°C until further use. After thawing on ice, tissue was mechanically homogenized and RNA isolated using the Monarch^® Total RNA Miniprep Kit (NEB, UK) according to the manufacturer’s instructions. Concentration and integrity of isolated RNA were determined by electropherograms (Bioanalyzer 2100; Agilent Technologies, USA) using the Agilent 6000 Nano or Pico kit. Synthesis of cDNA was performed with 200 ng and 100 ng RNA for males and females, respectively, with the QuantiTect^® Reverse Transcription Kit (Qiagen, Germany). Transcript levels were defined by RT-qPCR and amplified by gene-specific and exon-spanning primers. The primers were analyzed for the absence of primer dimers and only used if their efficiencies ranged between 1.8 – 2.2 (Supplementary Table S2). The final reaction consisted of 10 µl 2x KAPA SYBR^® Fast qPCR Master Mix (Kapa Biosystems, USA), 0.8 µl forward and reverse primer (10 µM, each), 3.4 µl PCR-grade water (Carl Roth, Germany), and 5 µl cDNA (1:20 diluted for male, 1:10 dilution for female samples). Cycling conditions were as follows: initial denaturation at 95°C for 3 min, followed by 45 cycles of denaturation at 95°C for 10 s, annealing at 60°C for 15 s, and elongation at 72°C for 20 s. Final elongation was performed at 60°C for 3 min. Amplification was followed by a melt curve analysis ranging from 60°C to 95°C with a stepwise increase of 1°C for 20 s to confirm amplification of a single product and the absence of primer dimers (Rotor Gene Q with Q-rex software, Qiagen, Germany). Relative transcript levels after RNAi were calculated by the Pfaffl method (Pfaffl, 2001) and are shown as log(2) fold changes. Smletm1 is a housekeeping gene of S. mansoni and was used as reference gene for all RT-qPCR analyses, as described before (Haeberlein et al., 2019). Transcript profiles of SmeIF4A-a and SmeIF4A-b were analyzed in bM (males obtained from couples after separation), sM (males obtained from ss infections, without any previous pairing experience), bF (females obtained from couples after separation), and sF (females obtained from ss infections, without any previous pairing experience), and were calculated using the Pfaffl method (Pfaffl, 2001). Smletm1 (Haeberlein et al., 2019) was used for normalization, and transcript profiles are shown as relative transcript levels.

Cloning of gene-specific fragments in the T7, T3, and SP6 promotors-containing pJC53.2 plasmid (Collins et al., 2010) was done as follows. By polymerase chain reaction (PCR), gene-specific amplicons of Smeif4a-a and Smeif4a-b of 400–500 bp (Supplementary Table S3) were generated. The Accu Prime Taq DNA Polymerase High Fidelity kit (Invitrogen, USA) was used to amplify PCR products with 3’ A overhangs. The reaction was set up according to the manufacturer’s instructions with 10 ng of plasmids carrying the coding sequences of Smeif4a-a and Smeif4a-b, respectively, 1 µl of corresponding primers (10 µM, each), and 1.5 µl ethylene glycol (100%; Merck, Germany) in a 50 µl reaction. The initial denaturation at 94°C for 2 min was followed by 34 cycles of denaturation at 94°C for 30 s, annealing at 58°C for 30 s, and elongation at 68°C for 1 min. A final elongation was performed at 68°C for 5 min (C1000™ Thermal Cycler, Bio-Rad, USA). PCR products were analyzed by agarose gel electrophoresis and extracted from the gel using the Monarch^® DNA Gel Extraction Kit (NEB, UK). Amplicons were recovered in 15 µl elution buffer. The pJC53.2 plasmid was linearized by the restriction enzyme AhdI (NEB, UK), which resulted in T overhangs. Inserts were ligated into pJC53.2 in 3:1 ratio using T4 ligase (NEB, UK) at 4°C overnight. Transformation of chemically competent Escherichia coli of the NEB 5α strain (NEB, UK) was performed by incubating 10 µl of the ligation reaction with 100 µl E. coli cells for 30 min on ice. A heat shock was applied for 80 s at 42°C followed by 5 min on ice before 900 µl of lysogeny broth medium (LB; Carl Roth, Germany) were added. Cells were incubated for 1 h at 37°C shaking and centrifuged for 1 min at 900 g. Approximately 900 µl of the supernatant were discarded, and cells were resuspended in the remaining volume of 100 µl. Transformed E. coli were selected on kanamycin (50 µg/ml; Sigma-Aldrich, Germany) containing LB agar plates and incubated at 37°C overnight. The sequence of integrated inserts into pJC53.2 was verified by Sanger sequencing (Microsynth SeqLab, Germany). The created constructs were used for riboprobe and dsRNA synthesis.

Gene-specific PCR amplicons of approximately 500 bp (Supplementary Table S3), which derived from the pJC53.2 plasmid constructs, were used to synthesize dsRNA for RNA interference (RNAi) experiments or for synthesizing riboprobes for whole mount in situ hybridization (WISH). Additionally, a non-schistosomal dsRNA encoding the E. coli ampicillin resistance gene (ampR) was synthesized as a control for RNAi as described elsewhere (Moescheid et al., 2023). Inserts of the pJC53.2 construct were amplified with a primer specific for the T7 promotor sequence (T7_extended 5’-CCT AAT ACG ACT CAC TAT AGG GAG-3’) in a 50 µl reaction consisting of: 8 ng plasmid, 2 µl dNTPs (10 mM each; Solis BioDyne, Estonia), 6 µl T7_extended primer (10 µM), 1 µl Betaine (5 M; Sigma-Aldrich, Germany), 10 µl Q5 reaction buffer (NEB, UK), and 0.5 µl Q5 High-Fidelity Polymerase (NEB, UK). Samples were first denatured at 98°C for 2 min, followed by 35 cycles of denaturation at 95°C for 15 s, annealing at 64°C for 20 s, and elongation at 72°C for 40 s. The final elongation occurred at 72°C for 3 min. PCR success was confirmed by agarose gel electrophoresis. For riboprobe synthesis, amplicons were purified with the Monarch^® Spin PCR & DNA Cleanup Kit (NEB, UK). In vitro transcription was modified after Collins et al., 2010. In brief, approximately 0.5 µg PCR product was used in a reaction with 20 µl rNTP mix (25 mM each; NEB, UK), 10 µl reaction buffer (10x), 1 µl inorganic pyrophosphatase (NEB, UK), 10 µg T7 RNA polymerase (expression vector kindly provided by Prof. Collins, UT Southwestern Medical Center), and Diethylpyrocarbonate (DEPC)-treated water to a final volume of 100 µl. The reaction was incubated overnight at 37°C. Remaining template DNA was removed by incubation with 10 U DNase I (NEB, UK) at 37°C for 30 min. Synthesized dsRNA was precipitated with 100 µl lithium chloride (7.5 M; Merck, Germany) at -20°C for at least 1 h. Samples were centrifuged for 30 min at 4°C, top speed. The supernatant was removed, and the pellet washed with 500 µl ice-cold ethanol (70%). Samples were centrifuged for 3 min at 4°C, top speed. Ethanol was cleaned from the reaction, and the dsRNA pellet resuspended in DEPC-treated water. For riboprobe synthesis, in vitro transcription of 0.5 µg PCR product comprising the T3 and SP6 promotors was performed to synthesize single-stranded RNA (ssRNA) probes. To this end, a Digoxigenin-11-UTP (DIG-UTP) rNTP mix was prepared by mixing 5 µl ATP, CTP, GTP (100 mM each; NEB, UK) with 3.5 µl UTP (100 mM; NEB, UK), and 17.5 µl DIG-UTP (10 mM; Roche, Switzerland) and filled up to 50 µl with DEPC-treated water. The in vitro transcription reaction consisted of: 0.5 µg PCR product, 1 µl T3 or SP6 RNA polymerase (Roche, Switzerland), 2 µl transcription buffer (10x; Roche, Switzerland), 2 µl DIG-rNTP mix, 0.6 µl murine RNase inhibitor (NEB, UK) in a final volume of 20 µl. Samples were incubated at 28°C for approximately 16 h. Template DNA was digested by 10 U RNase-free DNase I (NEB, UK) for 20 min at 37°C. An equal volume of 7.5 M lithium chloride was added, and the mixture incubated at -80°C for at least 30 min. Samples were centrifuged at 16,000 g for 30 min at 4°C. The supernatant was replaced by ice-cold ethanol (70%), and the sample centrifuged again at 16,000 g for 3 min at 4°C. Ethanol was removed and the pellet resuspended in 20 µl DEPC-treated water. The concentrations of dsRNA and ssRNA-probes were photometrically determined and their integrity confirmed by agarose gel electrophoresis.

WISH was performed for transcript localization of Smeif4a-a and Smeif4a-b, using a previously established protocol (Cogswell et al., 2011; Collins and Collins, 2017; Li et al., 2024; Moescheid et al., 2025) with the following modifications. Bs male tissue was permeabilized with 45 µg/ml proteinase K (Ambion, UK) in 1x PBSTx (PBS in DEPC-treated water + 0.3% TritonX-100; Sigma-Aldrich, Germany), and tissue of bs females with 20 µg/ml for 45 min at room temperature (RT) with moderate agitation. For hybridization, the worms were incubated with 200 ng/ml (Smeif4a-a, Smvlg1, Smmyst4) or 20 ng/ml (Smeif4a-b, Smtsp-2) of the appropriate riboprobe in a final volume of 300 µl at 55 °C and 135 rpm (GFL, Germany) for at least 16 h. The probe concentration of Smeif4a-b and Smtsp-2 was adjusted to reduce background signals. Worm samples were incubated with an anti-DIG-AP antibody (11093274910; Roche, Switzerland) diluted 1:2,000 in colorimetric block solution [7.5% heat-inactivated horse serum (H1138; Sigma-Aldrich, Germany) in Tris-NaCl-Tween-20 buffer solution] at 4°C overnight to detect target-bound riboprobes. Signals developed after the addition of the substrates nitro blue tetrazolium (11383213001; Roche, Switzerland) and 5-bromo-4-chloro-3-indolylphosphate (11383221001; Roche, Switzerland). Development was directly performed at RT for Smeif4a-a, Smvlg1, and Smmyst4, while Smeif4a-b and Smtsp-2 samples were incubated in AP buffer with added substrates at 4°C overnight and signals continued to be developed the next day at RT. Signal development was stopped at different time points by transferring worms in PBSTx. Worms were finally embedded on slides in 80% glycerol and imaged using an inverted laboratory microscope (DM IL LED; Leica Microsytems, Germany) with the WaveImage software (VWR, USA).

Transcriptional knockdowns (KD) of Smeif4a-a and Smeif4a-b were performed by RNAi. To this end, gene-specific dsRNAs were synthesized and the worms incubated in a modified version of the ABC169 medium (Wang et al., 2019) for soaking, as described before (Li et al., 2024; Moescheid et al., 2025). In brief, Basch medium (Basch, 1981) was supplemented with 10% (v/v) NCS, 1% (v/v) ABAM, 0.2% (v/v) red-blood cells (10% suspension; Biochrome, Germany), 200 µM ascorbic acid (Sigma-Aldrich, Germany), and 0.25% (v/v) low-density lipoprotein (LDL; Trina Bioreactives, Switzerland). Worms were incubated in 6-well plates (Greiner Bio-One, Germany) with 5 ml prewarmed ABC169 medium and 15 S. mansoni couples per well and kept in an incubator (nunc™ Galaxy S+; RS Biotech, Germany) at 37°C and 5% CO₂. Each approach was performed as technical duplicate, and worms deriving from different hamsters served as biological replicates. In total, six biological replicates were analyzed. DsRNAs against the target genes or the control were added one day after perfusion at a concentration of 15 µg/ml (Smeif4a-a) or 10 µg/ml (Smeif4a-b), respectively. Pilot experiments were performed with varying concentrations of dsRNA for each gene to determine the optimal concentration to reach sufficient knockdown efficiencies. However, these experiments were conducted in a different medium (M199 3+), and knockdown efficiencies might vary depending on the medium composition (Štefanic et al., 2010). Controls consisted of irrelevant dsRNA (ampR) of the same concentration and a control without dsRNA but adding an equal volume of DEPC-treated water to the parasite culture. Worms were transferred into new 6-well plates filled with prewarmed ABC169 medium 24 h after the start of the experiment using feather-weight tweezers alongside with the addition of dsRNA or DEPC-treated water. This procedure was repeated every 2–3 d until d 21. Phenotypes were investigated by an inverted laboratory microscope (DM IL LED; Leica Microsytems, Germany) considering the pairing status (paired or separated), worm attachment to the petri dish, worm motility, egg production, and morphological changes. Worm motility was classified as 0 = total absence of movement; 1 = minimal movement restricted to the gut, head or tail region; 2 = reduced motility; 3 = normal motility; 4 = hyperactivity according to Ramirez et al., 2007; Keiser, 2010; Manneck et al., 2010. In vitro laid eggs were counted and grouped as normal or abnormal eggs with respect to their morphology. Abnormal eggs were classified based on alterations in shape (e.g. absence of the spine) or a reduced egg size.

The vitellarium of S. mansoni bs females consists of fully differentiated, lipid-rich vitellocytes (Kunz, 2001; Lu et al., 2015). In order to investigate RNAi effects on mature vitellocytes, lipid droplets were stained by a modified version of the Oil-Red O staining method (Anthony et al., 2010; Wang et al., 2019). After 21 d of RNAi, couples were separated by 0.26% (w/v) ethyl 3-aminobenzoate methanesulfonate (Sigma-Aldrich, Germany) dissolved in M199 (3+), as described before (Collins et al., 2013). Then, worms were fixed and stored in 2% paraformaldehyde (Carl Roth, Germany) in 1x PBSTx at 4°C until further use. Next, the worms were washed twice in 1x PBSTx and transferred into netwell inserts (Science Service) fitting into individual wells of a 12-well plate. Worms were incubated in 99% propane-1,2-diol (Merck, Germany) for 5 min at RT, while shaking at 135 rpm (10 worms/ml). The solution was replaced by 0.5% (w/v) Oil-Red O (Sigma-Aldrich, Germany) dissolved in propane-1,2-diol, and staining occurred for 45 min at RT on a shaker. Afterwards, worms were washed twice with 85% propane-1,2-diol for 5 min each at RT with agitation. A final washing step was performed with 1x PBS until worms were embedded with ROTI Mount FluorCare (Carl Roth, Germany). The worms were immediately analyzed by phase contrast microscopy using an inverted laboratory microscope (DM IL LED; Leica Microsytems, Germany). The size of the vitellarium (mm²) was determined with ImageJ using the selection brush tool (Schindelin et al., 2012; Schneider et al., 2012; Rueden et al., 2017).

To investigate stem-cell proliferation after 21 d of RNAi, 5-ethynyl-2’-deoxyuridine (EdU) was added to the worms in vitro at a final concentration of 10 µM (Thermo Fisher Scientific, USA) for 24 h at 37°C and 5% CO₂. Treated and control couples were separated by 0.26% (w/v) ethyl 3-aminobenzoate methanesulfonate (Sigma-Aldrich, Germany) dissolved in M199 (3+) as described before (Collins et al., 2013). Subsequently, worms were collected in 2 ml reaction tubes and fixed in 4% paraformaldehyde in 1x PBSTx at 4°C overnight. For storage, worms were rinsed in PBSTx for 3 min shaking at RT, followed by dehydration in 50% methanol (MeOH; Carl Roth, Germany) in PBSTx and 100% MeOH for 10 min each on a shaker. Samples were stored in fresh 100% MeOH at -20°C. The worms were rehydrated in 50% MeOH in PBSTx at RT on a rocker for 10 min and subsequently washed in PBSTx for 10 min. Thereafter, worms were transferred to small sized baskets (Intavis Bioanalytical instruments, Germany) fitting into single wells of a 48-well plate. Bleaching occurred for 30 min under bright light in 1.2% (v/v) H₂O₂ (Carl Roth, Germany), 5% (v/v) formamide (Carl Roth, Germany), 0.5x saline-sodium citrate (SSC) pH 7.0 in DEPC-treated water. After bleaching, the worms were permeabilized with 6 µg/ml proteinase K (Ambion, UK) in PBSTx for 45 min. Post fixation was carried out with 4% paraformaldehyde in 1x PBSTx for 10 min at RT. Two washing steps followed with 1x PBS for 5 min each. Staining was performed using the Click-iT Plus EdU Alexa Fluor 488 imaging kit (Thermo Fisher Scientific, USA) for 30 min with shaking. From this point on, samples were protected from light. The worms were washed twice with PBS for 5 min each at RT on a shaker before counterstaining the parasites overnight with 10 µM Hoechst 33342 (Sigma-Aldrich, Germany) in PBSTx at 4°C with shaking. Next, the worms were washed three times with PBSTx at RT for 10 min each with shaking followed by mounting on slides, as described previously (Hahnel et al., 2014). In order to quantify cell proliferation in the immature part of the ovary or the testicular lobes, where gonadal stem cells mostly occur in schistosomes (Collins and Collins, 2017; Li et al., 2021), the surface areas were determined using the selection brush tool of ImageJ (Schindelin et al., 2012; Schneider et al., 2012; Rueden et al., 2017). Proliferating EdU-positive (EdU⁺) oogonia and spermatogonia were counted, and the ratio of proliferating cells per mm² was calculated, as described elsewhere (Moescheid et al., 2025).

After 21 d of RNAi, morphological changes in the reproductive organs of males and females were investigated by confocal laser scanning microscopy (CLSM) as described before (Machado-Silva et al., 1994; Neves et al., 2005; Beckmann et al., 2010). For this, couples were separated by 0.26% (w/v) ethyl 3-aminobenzoate methanesulfonate (Sigma-Aldrich, Germany) dissolved in M199 (3+). Males and females were fixed in alcohol formalin glacial acetic fixative (AFA) consisting of 2% acetic acid (Carl Roth, Germany), 1.1% paraformaldehyde, and 66.7% ethanol for 24 h at RT. The parasites were transferred to netwell inserts fitting into single wells of a 12-well plate, and staining was performed on a shaker at RT with Certistain carmine red (Merck, Germany) for 1 h. The worms were destained up to five times in acidic ethanol [70% (v/v) ethanol, 2.5% (v/v) hydrochloric acid (Carl Roth, Germany)] for 5 min each. Thereafter, worms were dehydrated by increasing ethanol concentrations (80%, 90%, 100%) for 5 min each, and fixed on slides in Euparal (Carl Roth, Germany). The software package “IMARIS for cell biologists” (v 8.4.2; Bitplane, Switzerland) was used to determine the volume of ovaries and testes, as described before (Tonk et al., 2020). Z-stacks of these reproductive organs, generated by CLSM, served as input data. For microscopic evaluation, the confocal microscope TCS SP5 VIS with the LAS AF software (Leica Microsystems, Germany) was used. Alexa 488 and carmine red were excited with an argon-ion laser at 488 nm, whereas Hoechst 33342 was excited at 405 nm. Background signals and optical section thickness were defined by setting the pinhole size to airy unit 1 (Kellershohn et al., 2018; Mughal et al., 2021). Z-stacks of ovaries and testes were acquired with a step size of 0.5 µm.

Statistical analyses were performed using the GraphPad Prism V.8 software (GraphPad Software; San Diego, USA) applying the paired or unpaired two-tailed t-test for normally distributed data, determined by the Shapiro-Wilk test, or the two-tailed Mann-Whitney test for non-normally distributed data, as previously described (Moescheid et al., 2023, Moescheid et al., 2025). Volumes of ovaries and testicular lobes were statistically evaluated by the unpaired Welch’s t-test. Statistical differences with p < 0.05 were considered significant.

Three isoforms of eIF4A have been identified in S. mansoni by BLASTp using human orthologs as templates (accession numbers, see Supplementary Table S1). These SmeIF4A isoforms were analyzed for the presence of conserved motifs (Li et al., 1999; Cordin et al., 2006), as shown by multiple sequence alignment with human eIF4A sequences (Figure 1). All three isoforms have the Q motif and motifs I-VI, which are important for the function of the protein. The Q motif and motifs I, II, and VI are involved in ATP binding and hydrolysis, while motifs Ia, Ib, and IV are required for RNA binding (Andreou and Klostermeier, 2013). eIF4AIII can be distinguished from eIF4AI and eIF4AII due to additional conserved motifs specific for this isoform (Li et al., 1999). These motifs were found for Smp_034190 (Figure 1, red boxes) having a sequence identity greater than 80% to HseIF4AIII (Figure 1, Supplementary Table S4). Therefore, it was classified as S. mansoni eIF4AIII (SmeIF4AIII). HseIF4AIII has an amino acid (aa) sequence identity of 67% to HseIF4AI and HseIF4AII. A similar range was observed for SmeIF4A with Smp_097660 and Smp_166400 being 53% or 61% identical to SmeIF4AIII, respectively. In contrast to the high sequence identity of ~ 90% between HseIF4AI and HseIF4AII, the sequence of Smp_097660 is only to 59% identical to Smp_166400. Each of these two S. mansoni eIF4A isoforms exhibits a similar sequence identity to both HseIF4AI and HseIF4AII, with Smp_097660 being 57% and 58% identical to HseIF4AI and HseIF4AII, respectively, while the sequence identity of Smp_166400 to HseIF4AI and HseIF4AII was 70% in both cases (Supplementary Table S4). Consequently, the two SmeIF4A isoforms could not be clearly assigned as orthologs to either HseIF4AI or HseIF4AII. Since the SmeIF4A isoforms appear to be no one-to-one orthologs to the human eIF4A isoforms, they have been designated SmeIF4A-a (Smp_097660) and SmeIF4A-b (Smp_166400).

Based on their aa sequences, we conducted a phylogenetic analysis for the SmeIF4A isoforms, which were compared to orthologs of organisms with different evolutionary distance, including model organisms and other platyhelminths (Figure 2). The DEAD-box RNA helicase p68, which is closely related to Vasa proteins, served as outgroup (Skinner et al., 2012). The resulting phylogenetic tree showed clustering of eIF4AI and eIF4AII isoforms of model organisms, consistent with the fact that they share approximately 90% sequence identity (Nielsen and Trachsel, 1988). In platyhelminths, two (or more) eIF4A genes (eIF4A-a, eIF4A-b) were identified that were not classified as eIF4AIII, and they clustered in a subgroup. The eIF4AIII isoform contains additional conserved aa motifs, which distinguish it from eIF4AI and eIF4AII isoforms (Li et al., 1999). eIF4AIII of model organisms was found to be distant to eIF4AI and eIF4AII, and it grouped together with eIF4AIII of platyhelminths. As expected, eIF4A-a, eIF4A-b, and eIF4AIII sequences from trematodes were evolutionary closely related, but distinct from the cestode Echinococcus. Our results confirm previous phylogenetic analyses that identified S. rodhaini as the closest relative of S. mansoni and S. japonicum, a member of the Asian clade, as the most distant relative (Lawton et al., 2011). The corresponding branch lengths of this phylogenetic tree are illustrated in Supplementary Figure S1. Additionally, we performed individual multiple sequence alignments for all eIF4A isoforms of each species (Table 1). The results showed a sequence identity of approximately 90% between eIF4AI and eIF4AII of vertebrates, while eIF4AIII is different with about 60% sequence identity to eIF4AI (Nielsen and Trachsel, 1988; Li et al., 1999). In platyhelminths, the sequence identity between the eIF4A-a and eIF4A-b isoforms was lower (~ 60%), and it is only slightly higher than the similarity between eIF4AIII and eIF4A-a of platyhelminths (~ 55%; Table 1). This indicates that the platyhelminth eIF4A-a and eIF4A-b are no one-to-one orthologs of the vertebrate isoforms eIF4AI and eIF4AII. Thus, the gene duplication of eIF4A in platyhelminths appears to be independent from that of vertebrates.

Previous bulk RNA-seq data showed transcript levels of genes of interest in males and females before and after pairing and in their corresponding isolated gonads (testis, ovary; Lu et al., 2016, Lu et al., 2018). This data set indicated a higher transcript level of Smeif4a-a in females compared to males, and slightly more transcripts in bF after pairing compared to sF as well as higher transcript levels in ovaries of bF compared to ovaries of sF (Supplementary Figure S2, Lu et al., 2016, Lu et al., 2018). RT-qPCRs with cDNA obtained from bM, bF as well as sM, sF, respectively, supported previous RNA-seq data with a similar tendency of slightly, but not significantly higher expression levels in females after pairing. In contrast to previous RNA-seq data, we observed significantly higher transcript levels in sF compared to sM using RT-qPCR (Figure 3A). Next, we performed WISH to localize Smeif4a-a transcripts using a gene-specific single-stranded riboprobe. The sense probe served as negative control. As positive controls, we used previously published riboprobes of Smvlg1 (Skinner et al., 2012), Smtsp-2 (Cogswell et al., 2011), and Smmyst4 (Li et al., 2024; Supplementary Figure S3). Already after short signal development (~ 30 min), Smeif4a-a transcripts were detected in the middle and posterior parts of the bF ovary, where intermediate and mature oocytes are located (Figure 3B). After longer signal development (> 2 h), additional signals were found in the anterior part of the bF ovary, the location of oogonia, and in the vitellarium of females. Moreover, at this later time point we found signals of Smeif4a-a transcripts in the testis and in striping patterns along the male body, which could be indicative for muscle and neuronal cells (Wendt et al., 2020; Li et al., 2024). These results confirmed single-cell (sc) RNA-seq data from adults, which revealed high transcript levels in the germline of bF and bM (Supplementary Figure S4, Wendt et al., 2020). In a recent scRNA-seq study of oocytes, Smeif4a-a transcripts were found to be highly expressed in germ cells/germ cell progeny, intermediate, and late germ cells in ovaries of bF, but to a lesser extent in the somatic cell cluster (Supplementary Figure S5, Moescheid et al., 2025).

Bulk RNA-seq data indicated that Smeif4a-b may not be gender- or pairing-dependently expressed in adults (Supplementary Figure S2; Lu et al., 2016; Lu et al., 2018). In contrast, our RT-qPCR data revealed significantly higher transcript levels in sF compared to bF, whereas no difference was detected between males with and without pairing experience (Figure 4A). WISH localized transcripts of Smeif4a-b in both parts of the ovary, along the uterus, and in the vitellarium of bF after short and long signal developments. Longer signal developments localized Smeif4a-b transcripts in the testis and in muscles and neurons, but weaker compared to Smeif4a-a (Figure 4B). ScRNA-seq analysis identified Smeif4a-b transcripts in the GSC progeny, neurons, and muscles. However, there was no significant enrichment of Smeif4a-b in any cluster (Supplementary Figure S4, Wendt et al., 2020). ScRNA-seq of oocytes indicated that Smeif4a-b transcripts are present in the GSC/GSC progeny and intermediate stage oocytes (Supplementary Figure S5, Moescheid et al., 2025). Signals detected along the uterus might represent the muscles surrounding the uterus (Mair et al., 2000).

To gain first insights into the potential functions of Smeif4a-a, gene-specific KD experiments were performed by RNAi. The used dsRNAs had sizes of ~ 500 bp and were not covered by the RT-qPCR primers used for downstream analysis. The designed dsRNA was analyzed using the small interfering RNA-finder (si-fi) software to predict potential off-targets (Lück et al., 2019) as previously described (Moescheid et al., 2023). None of the listed off-targets with > 3 siRNA hits were annotated as RNA helicases, showing that the used dsRNA had a certain specificity for Smeif4a-a. RNAi experiments were performed with 30 couples each and performed with six biological replicates. Couples incubated with a similar volume of DEPC-treated water, the solvent for dsRNA, or with non-schistosomal dsRNA (ampR) of the same concentration as in the RNAi group, served as controls. Previous studies showed no effects of ampR dsRNA on pairing stability, motility, egg production, and morphology of S. mansoni couples at concentrations between 30 – 60 µg/ml (Moescheid et al., 2023; Li et al., 2024). In this study, lower concentrations of dsRNA (15 µg/ml; 10 µg/ml) were applied, therefore a comparison between ampR dsRNA (15 µg/ml; 10 µg/ml) and a corresponding water control was performed to evaluate ampR as suitable dsRNA control at the concentrations used in this study. Although egg production was significantly higher in the ampR group compared to the water control (no dsRNA) at d 13 and d 15 of treatment, ampR dsRNA (15 µg/ml) showed no further effects on worm viability, morphology, or transcript levels of the genes examined in this study (Supplementary Figure S6). Therefore, the ampR dsRNA control was used in the following to interpret Smeif4a-a RNAi effects. After 21 d of dsRNA treatment, the KD efficiency was determined. For this, couples were separated, and RNA from each gender was isolated and used for RT-qPCR. The results confirmed an efficient Smeif4a-a KD of 88% in males and females, respectively (88.0 ± 1.0%, 88.0 ± 1.4% reduction compared to the water control; Figure 5A). The remaining protein level of SmeIF4A-a after RNAi could not be analyzed due to the lack of an appropriate antibody. The transcript level of isoform Smeif4a-b was also analyzed, but it was unaffected upon Smeif4a-a RNAi (Supplementary Figure S7). During the treatment period, control and Smeif4a-a RNAi worms were monitored every 2–3 d and evaluated for physiological or morphological changes based on a standardized scoring system (Moescheid et al., 2023; Li et al., 2024). KD of Smeif4a-a caused no pairing instability (Figure 5B), while worm attachment and motility were significantly reduced after 20 and 21 d of treatment, respectively (Figures 5C, D). Furthermore, eggs laid in vitro were quantified and categorized as abnormal or normal eggs depending on their size and shape. Examples for abnormal eggs are shown in Supplementary Figure S8. The overall egg production was evaluated by calculating abnormal plus normal eggs per couple. After 10 d of treatment, egg production was significantly reduced (Figure 5E) in the Smeif4a-a RNAi group and further declined until no eggs were laid after 21 d (data not shown). The percentage of abnormal eggs did not significantly differ between Smeif4a-a dsRNA-treated and control worms (Figure 5F), although a tendency of more abnormal eggs was found from d 10 on. Next, we investigated the morphology of the reproductive organs of all worm groups after 21 d of treatment by CLSM (Figure 6). Z-stacks of these organs were acquired to compare volumes (Figure 6B). This analysis revealed reduced volumes of the testicular lobes and the ovary in the Smeif4a-a RNAi group (Figures 6A, B) compared to the control group. In contrast to control males, which possess densely filled testes, empty spaces were observed within the testicular lobes of Smeif4a-a RNAi males. In ovaries of this group, no clear separation between both parts of the ovary was found, and low numbers of immature and mature oocytes appeared to be mixed within the comparatively small ovary. In addition, the total number of vitellocytes and the amount of stage 3/4 vitellocytes (Kunz, 2001; Lu et al., 2015) seemed to be reduced in the vitellarium of Smeif4a-a RNAi females (Figure 6A). To substantiate this RNAi effect on vitellocytes, worms were incubated in Oil-red O to stain lipid droplets that are characteristic for fully differentiated S3/S4 cells (Kunz, 2001; Lu et al., 2015). Following Smeif4a-a RNAi, the vitellarium exhibited lower staining intensities compared to the controls. Furthermore, in contrast to the control (ampR dsRNA), in which intensively stained S4 vitellocytes accumulated in the vitelloduct (which connects the vitellarium with the ootype) and thus made this structure evident, in females of the Smeif4a-a RNAi group the vitelloduct was hardly visible (Figure 7A). This can be explained by the absence of Oil-red O-stained S3/S4 vitellocytes, which corresponds to the CLSM data. Next, we quantified the area of the vitellarium and found a significantly smaller size of this organ following Smeif4a-a RNAi (Figure 7B). Additionally, we determined the relative transcript levels of the eggshell precursor protein p14 (Smp14) and tyrosinase 1 (Smtyrosinase1), as vitellarium marker genes (Ebersberger et al., 2005; Fitzpatrick et al., 2007). RT-qPCRs of control and Smeif4a-a RNAi worms showed significantly lower transcript levels for both marker genes in females of the RNAi group (Figure 7C).

EdU assays were performed to comparatively investigate stem-cell proliferation in testes and ovaries of the treatment groups. No EdU⁺ cells were detected in males and females after Smeif4a-a RNAi, whereas EdU⁺ cells occurred in the control with irrelevant dsRNA (ampR; Figures 8A, B). The absence of EdU⁺ cells was not restricted to the reproductive organs where GSCs are located (Wang et al., 2018; Lee, 2023). In addition, we observed no EdU⁺ cells in SSCs (neoblasts; Collins et al., 2013) around the gut, in the parenchyma, and next to the tegument. A similar phenotype was found before in a large-scale, unbiased RNAi study that aimed to identify therapeutic targets in S. mansoni. However, no causal link between SmeIF4A-a and its role in stem-cell proliferation was described (Wang et al., 2020). The results of EdU assays were substantiated by RT-qPCR results obtained for the stem cell marker genes Smnanos-1 and Smnanos-2, which were shown before to regulate proliferation processes in GSCs and SSCs (Collins et al., 2013; Wang et al., 2013, Wang et al., 2018). Upon Smeif4a-a RNAi, the transcript levels of both genes were significantly downregulated in males and females, which corresponds to the results of the EdU assays (Figure 8C).

As for Smeif4a-a, RNAi experiments were performed to functionally characterize Smeif4a-b in male and female worms. To this end, Smeif4a-b dsRNA was administered to worm couples every 2–3 d, which resulted in KD efficiencies of 69.3 ± 13.6% and 67.8 ± 3.9% in males and females, respectively, after 21 d of treatment (Figure 9A). Analysis of potential off-targets by si-fi (Lück et al., 2019) predicted four genes with > 3 siRNA hits. These genes were annotated as RNA helicases including Smeif4a-a (Smp_097660). RT-qPCR showed lower transcript levels for three of these genes, Smp_333940, Smp_034190, and Smp_031010, but only in males and not in females (Supplementary Figure S9). Transcript levels of Smeif4a-a were unaffected by the KD of Smeif4a-b in males and females, indicating no reciprocal influence of RNAi on the transcript levels of each of these genes. During the treatment period of 21 d, the worms were examined for morphological or physiological phenotypes – as for Smeif4a-a. While pairing stability, motility, total egg production, and the percentage of abnormal eggs were similar between the Smeif4a-b RNAi group and the control (ampR dsRNA), only worm attachment was significantly reduced after 17 d of Smeif4a-b RNAi (Figure 9). After 21 d, only about 60% of the worms were still attached (Figure 9C). The morphology of the reproductive organs of both genders and stem-cell proliferation of GSCs and SSCs were evaluated in control and Smeif4a-b RNAi worms by CLSM, lipid staining, and EdU assays. CLSM analysis of the reproductive organs of testis, ovary, and vitellarium, revealed no differences between the control and the Smeif4a-b RNAi group (Figure 10). The vitellarium was studied in more detail by staining the lipid droplets of mature vitellocytes and quantifying the relative transcript levels of Smp14 and Smtyrosinase1 as marker genes. Lipid staining of control and Smeif4a-b RNAi worms resulted in similar staining intensities between both treatment groups (Figure 11A). Furthermore, the relative transcript levels of Smp14 and Smtyrosinase1 were not significantly different from those of the control group (Figure 11B). Next, EdU assays to investigate stem-cell proliferation of GSCs and SSCs revealed no differences of the numbers of EdU⁺ cells in ovaries and testes upon Smeif4a-b RNAi (Figures 12A, B). In males and females, finally, the transcript levels of Smnanos1 and Smnanos2 as stem-cell marker genes were similar to those of the control (Figure 12C).

A comparison between the water control and 10 µg/ml ampR dsRNA was performed for each type of post-RNAi analysis to confirm the suitability of ampR dsRNA as irrelevant dsRNA control. The only difference between the water and ampR group was found on d 13 with a significantly higher egg production of ampR worms. Otherwise, ampR dsRNA (10 µg/ml) had no effects on worm viability, morphology, stem-cell proliferation or transcript levels of the genes examined in this study (Supplementary Figure S10).