Schistosoma mansoni α-N-acetylgalactosaminidase (SmNAGAL) regulates coordinated parasite movement and egg production

α-galactosidase (α-GAL) and α-N-acetylgalactosaminidase (α-NAGAL) are two glycosyl hydrolases responsible for maintaining cellular homeostasis by regulating glycan substrates on proteins and lipids. Mutations in the human genes encoding either enzyme lead to neurological and neuromuscular impairments seen in both Fabry- and Schindler/Kanzaki- diseases. Here, we investigate whether the parasitic blood fluke Schistosoma mansoni, responsible for the neglected tropical disease schistosomiasis, also contains functionally important α-GAL and α-NAGAL proteins. As infection, parasite maturation and host interactions are all governed by carefully-regulated glycosylation processes, inhibiting S. mansoni’s α-GAL and α-NAGAL activities could lead to the development of novel chemotherapeutics. Sequence and phylogenetic analyses of putative α-GAL/α-NAGAL protein types showed Smp_089290 to be the only S. mansoni protein to contain the functional amino acid residues necessary for α-GAL/α-NAGAL substrate cleavage. Both α-GAL and α-NAGAL enzymatic activities were higher in females compared to males (p<0.05; α-NAGAL > α-GAL), which was consistent with smp_089290’s female biased expression. Spatial localisation of smp_089290 revealed accumulation in parenchymal cells, neuronal cells, and the vitellaria and mature vitellocytes of the adult schistosome. siRNA-mediated knockdown (>90%) of smp_089290 in adult worms significantly inhibited α-NAGAL activity when compared to control worms (siLuc treated males, p<0.01; siLuc treated females, p<0.05). No significant reductions in α-GAL activities were observed in the same extracts. Despite this, decreases in α-NAGAL activities correlated with a significant inhibition in adult worm motility as well as in egg production. Programmed CRISPR/Cas9 editing of smp_089290 in adult worms confirmed the egg reduction phenotype. Based on these results, Smp_089290 was determined to act predominantly as an α-NAGAL (hereafter termed SmNAGAL) in schistosome parasites where it participates in coordinating movement and oviposition processes. Further characterisation of SmNAGAL and other functionally important glycosyl hydrolases may lead to the development of a novel anthelmintic class of compounds.


Introduction
Schistosomiasis is a parasitic disease caused by infection with blood flukes of the genus Schistosoma and is estimated to affect more than 250 million people worldwide [1]. With up to 200,000 deaths per annum, schistosomiasis is considered one of the most devastating neglected tropical diseases (NTDs) of resource poor communities within developing countries [2]. Currently, the primary method used to control schistosomiasis is mediated by praziquantel chemotherapy [3] with recent data suggesting that this drug targets transient receptor potential (TRP) channels expressed in adult schistosomes [4,5]. The continued use of praziquantel in mass drug administration (MDA) programmes has given rise to the fear of possible drug resistant blood flukes developing. Indeed, multiple studies have demonstrated that praziquantel insensitive schistosomes can be generated in the laboratory [6][7][8]. These findings, thus, highlight the need for the development of novel anti-schistosomal drugs, which can act as praziquantel replacements if praziquantel-resistant schistosomes eventually arise. Identifying and inhibiting processes critical to schistosome developmental biology and/or host interactions presents a strategy that may contribute to this objective.
It is well established that schistosomes carefully regulate their production of specific glycans and glycan elements as they transition between diverse environmental niches including water, mammals and snails [9][10][11][12]. Some parasite glycans, as moieties of glycoproteins or glycolipids, are immunomodulatory in nature and contribute to host (snail/mammal) as well as dioecious (male/female schistosome) interactions [12][13][14][15]. Regardless of their significance in orchestrating these heterospecific, conspecific or developmental activities, the gene products responsible for schistosome glycoprotein and glycolipid anabolism and degradation are only slowly being characterised [16,17]. One such glycogene family, critical to glycosylation homeostasis in other eukaryotes, is the glycosyl hydrolases (or glycoside hydrolases, GH). These enzymes hydrolyse glycosidic bonds between monosaccharide residues in a carbohydrate chain or bonds between and 100 μg/ml penicillin/streptomycin in an atmosphere of 5% CO 2 with a 70% media exchange performed after 24 hr and 48 hr for adult male/adult pairs and female worms respectively. Cultured worms were subsequently used for RNA interference, CRISPR/Cas9 genome editing, total RNA isolation/complementary DNA (cDNA) generation, dual RNA and genomic DNA (gDNA) extraction and enzyme assays.

Homology modelling of Smp_089290
The three-dimensional structure of Smp_089290 was derived by homology modelling using MODELLER [40]. The α-GAL/α-NAGAL template selected for Smp_089290 modelling was the three-dimensional structure of H. sapiens α-NAGAL (Protein Databank identification code 4DO4) [32]. The sequence identity between Smp_089290 and the human α-NAGAL was 40% with a sequence coverage of 97%, hence well within the acceptable range for comparative modelling [41]. The stereochemical quality of the Smp_089290 model was assessed by RAM-PAGE (Ramachandran Plot Analysis) [ Total RNA isolation and cDNA generation S. mansoni total RNA was isolated using the Direct-zol RNA MiniPrep Kit (Zymo Research) with slight modifications. Briefly, worms were removed from culture, transferred to 2 ml Eppendorf tubes, and homogenised in QIAZOL reagent (Qiagen) with a 5 mm diameter stainless steel bead (Qiagen) for a total of 6 min (2 x 3 min with 1 min on ice in between homogenisations) at 50 Hz using a TissueLyser LT (Qiagen). Thereafter, the RNA was treated with DNase I to remove contaminating gDNA. RNA was eluted into collection tubes by adding 30 μl of DNase/RNase-free H 2 O directly to the Direct-zol column matrix and subsequently centrifuged at 17,000 x g for 60 sec. Yields of total RNA samples were assessed using a Nano-Drop ND-1000 UV-Vis spectrophotometer; RNAs were concentrated when necessary in a Concentrator plus (Eppendorf) and re-quantified. Schistosome cDNAs were reverse transcribed from the total RNAs using the SensiFAST cDNA Synthesis Kit (Bioline) according to the manufacturer's instructions.

cDNA cloning and sequencing of smp_089290
PCR primers used to amplify the full coding sequence of Smp_089290 included forward (5´-ATG GCT ACC GTA CCA CCG-3´) and reverse (5´-CTA TAA CAA TGT CTG AAA CAG TCC ATC-3´) pairs. PCR amplification utilised cDNA prepared from total RNA obtained from mixed-sex adult worms [45]. The amplification was performed in 25 μl containing Biomix reaction mix (Bioline) and ultra-stable Taq DNA polymerase (Bioline). Amplicons were subjected to electrophoresis through 1% w/v agarose, stained and visualised with SYBR Safe dye (Thermo Fisher Scientific) under UV light. Products of the predicted size of 1,463 bp were ligated into pGEM-T Easy (Promega) after which α-Select Bronze Efficiency Competent Cells (Bioline) were transformed with the ligation products. Blue/white screening of E. coli colonies was accomplished on LB agar plates supplemented with 5-bromo-4-chloro-3-indolyl-β-Dgalactopyranoside (X-Gal) and Isopropyl β-D-1-thiogalactopyranoside (IPTG). Recombinant (white) clones were isolated from a 5 ml High Salt (HS) Luria Bertani (LB) culture containing ampicillin (50 μg/ml) using the QIAprep Spin Miniprep Kit (Qiagen). Inserts in pGEM-T Easy were sequenced at Aberystwyth University's Translational Genomics Facility and the sequence trace was analysed using FinchTV 1.4.0 software [46]. The full coding sequence of Smp_089290 is deposited under the GenBank accession number MZ508282.

DNA microarray analysis
The S. mansoni long oligonucleotide DNA microarray was designed and constructed by Fitzpatrick and colleagues [45]. The DNA microarray consists of 35,437 S. mansoni oligonucleotide 50-mers as well as 2,195 controls and is deposited in the ArrayExpress database under the accession number A-MEXP-830. S. mansoni lifecycle stages profiled by the DNA microarray included egg, miracidia, mother (2 day) sporocysts, daughter sporocysts, cercariae, 3 hr schistosomula, 24 hr schistosomula, 3 day schistosomula, 6 day schistosomula, 3 wk worms, 5 wk  worms, 7 wk worms, male 7 wk worms and female 7 wk worms. Normalised fluorescence  intensity values (available via ArrayExpress under the experimental accession number  E-MEXP-2094), corresponding to 50-mers representing exons 2 and 3 of smp_089290, enabled the quantification of gene expression profiles across these developmental stages of the schistosome.

RNA-Seq meta data analysis
The RNA-Seq meta database was created by Lu

RNA interference (RNAi)
Small interfering RNA (siRNA) duplexes were designed (siSmp_089290: sense strand 5'-CUA AUG AAA UCG UUG CAG A-3'; anti-sense strand 5'-UCU GCA ACG AUU UCA UUA G-3') based on the cDNA sequence verified smp_089290 amplicon and possessed no significant homology to other gene products (i.e. less than 15 out of 19 nucleotides conserved) in the S. mansoni genome assembly (v7.0) [54,55]. An siRNA duplex designed for firefly luciferase (siLuc) without significant homology to gene products in the S. mansoni genome assembly (v7.0) served as a negative control [56]. Briefly, sets of 10 worms (sex-separated) or five adult pairs were transferred to 0.4 cm pathway electroporation cuvettes (Invitrogen) containing DMEM supplemented with 2 mM L-glutamine and 100 μg/ml penicillin/streptomycin. siRNA duplexes (5 μg) were subsequently added and worms were electroporated with a single pulse at 125 V (LV mode) for 20 ms using an ECM 830 Square Wave Electroporation System (BTX, Harvard Apparatus, Holliston, MA). Worms were subsequently transferred to a 48 well tissue culture plate and cultured for up to seven days.

Preparation of CRISPR/Cas9 plasmid constructs
Two CRISPR target sites (single guide RNA; sgRNA sequence) for Cas9-catalysed gene editing for smp_089290 were designed by Breaking-Cas [57] with default parameters compatible for the protospacer adjacent motif (PAM) of Cas9 from Streptococcus pyogenes (NGG) [58,59]. Two sgRNAs targeting the coding regions exons 1 (SmNAGALX1: 5´-CUA CCG UAC CAC CGA UGG GU-3´) and 2 (SmNAGALX2: 5´-UUG UAA UCU AUG GCG UAU GC-3´) were used in this study. The sgRNAs contained >40% GC-content, no self-complementarity and no off-target sites against the S. mansoni genome assembly (v7.0) as predicted by Breaking-Cas software. A 20 nucleotide (nt) 'Scramble' sgRNA designed with low homology to the S. mansoni genome assembly (v7.0) and lack of an adjacent PAM site (necessary for Cas9 function) served as a non-targeting control (5´-GCA CUA CCA GAG CUA ACU CA-3´). CRISPR/Cas9 plasmid constructs were assembled using the pLenti-Cas-Guide construction Kit (GE100010, OriGene, Maryland, USA) and each sgRNA was ligated into the pLenti-Cas-sgRNA backbone as per the manufacturer's instructions. Expression of the sgRNA (SmNA-GALX1, SmNAGALX2 or Scramble containing plasmids) was driven by the mammalian U6 promoter and expression of Cas9 from S. pyogenes with nuclear localisation signals was driven by the human cytomegalovirus (CMV) promoter. Each CRISPR/Cas9 plasmid construct was independently transformed into DH5-α Chemically Competent Cells (GoldBio) with chloramphenicol (25 μg/ml) as drug selection. The chloramphenicol resistant transformants were confirmed for corrected orientation of sgRNA in the vector backbone by Sanger direct sequencing. To amplify the plasmid DNA, the transformants were first cultured in a 10 ml HSLB culture containing chloramphenicol (25 μg/ml) in a shaking incubator at 37˚C, 225 rpm for approximately 7 hr. Thereafter, 10 ml of the culture was added to 240 ml of fresh HSLB culture containing chloramphenicol (25 μg/ml) and cultured overnight, as above. Plasmid DNA was recovered from 150 ml of this overnight culture using the GenElute HP Plasmid Midiprep Kit (Sigma-Aldrich) as per the manufacturer's instructions except for a modified elution step; here, accomplished with 800 μl of nuclease-free H 2 O. Plasmid DNAs were quantified using the NanoDrop ND-1000 UV-Vis spectrophotometer (Thermo-Fisher Scientific), after which the DNA was precipitated and subsequently dissolved in Opti-MEM Reduced Serum Medium (Thermo-Fisher Scientific) and stored at -20˚C.

CRISPR/Cas9 mediated genome editing
A mixture of 24 μg of CRISPR/Cas9 plasmid DNA (either SmNAGALX1, SmNAGALX2 or Scramble) reconstituted in a total volume of 400 μl Opti-MEM Reduced Serum Medium was dispensed into a 0.4 cm pathway electroporation cuvette (Invitrogen). Five pairs of worms were transferred to the cuvettes, which had been chilled on wet ice, and subjected to square wave electroporation with a single pulse of 125 V (LV mode) for 20 ms (BTX, Harvard Apparatus). Worms were subsequently transferred to a 48 well tissue culture plate and cultured for up to seven days at 37˚C in an atmosphere of 5% CO 2 . Upon completion of the seven day experiment, RNA and gDNA was extracted from adult male and female worms using both RNAzol reagent (Sigma-Aldrich) and DNAzol reagent (Invitrogen) based on the manufacturer's instructions. One ml of RNAzol reagent was added to the parasite material, then subsequently homogenised four times with a 5 mm diameter stainless steel bead (Qiagen) for 3 min at 50 Hz using a TissueLyser LT (Qiagen) with 1 min incubation on ice in between homogenisations. Once the tissues of the schistosomes were completely disrupted, the protocol steps were followed until completion. RNA samples were immediately used for cDNA synthesis, as above. gDNA samples were amplified by PCR using primers encompassing the programmed doublestrand break (DSB) site at exon 1 (Forward: 5´-CTT ATA GGT GTG CCA TAT TAA CGA T-3´, Reverse: 5´-ATG CAC TAC ATT CGA AAG ACA-3´) or exon 2 (Forward: 5´-AGT GTT CTC ATG CAG TTA TCC T-3´, Reverse: 5´-TCC ATG TCA GCT GAG ATC A-3´) of smnagal. Correct amplification was verified by 1% w/v agarose gel electrophoresis. Thereafter, amplicons were subjected to the QIAseq 1-Step Amplicon Library Kit (Qiagen) for Illumina compatible next generation sequencing (NGS) library construction with GeneRead DNAseq Targeted Panels V2 (Qiagen) as per the manufacturer's instructions. Amplicon size of each NGS library was verified using a 2100 Bioanalyzer (Agilent, Santa Clara, CA). The NGS libraries were quantified using the GeneRead Library Quant Kit (Qiagen). NGS libraries were pooled at the Genewiz NGS facility (Genewiz, NJ) and processed with a MiSEQ configuration of 2x250 bp. The demultiplexing data generated by the Genewiz NGS facility was exported as Fastq files (.qz format). The mutations around the cut sites were analysed by CRIS-PResso2 software using default parameters [60,61]. Background mutations (i.e. mutations not attributable to genome editing) were inferred by identifying mutations present in the control samples and the treatment samples. The unique mutations around the programmed DSB sites only reported in the high quality sequence reads of the target sample were designated as CRISPR/Cas9-induced mutations. The software analysed and reported the percentage of indels (insertions and deletions) and substitutions in the target sample (inferred by the number of reads that are modified in comparison to the total reads), the percentages of each mutation type and a list of each specific mutation showing which nts have been altered. Sequence reads from the NGS libraries are available at the Sequence Read Archive (SRA) under BioProject ID PRJNA743897.

Quantitative reverse transcription (RT)-PCR (qRT-PCR) analysis
cDNAs synthesised from freshly perfused, siRNA treated or CRISPR/Cas9 plasmid treated schistosomes were used as templates for qRT-PCR to analyse transcript abundance. smp_089290 transcript levels in freshly perfused schistosomes were quantified relative to Ras-Related GTP-binding protein D (smrgbd) and Tubulin-specific chaperone D (smtscd); these two genes were identified by Normfinder [62] as being constitutively expressed across nine S. mansoni lifecycle stages, including the adult [63]. smp_089290 transcript levels in siRNA treated and CRISPR/Cas9 plasmid treated schistosomes were quantified relative to α-tubulin (smat1). qRT-PCR was performed using a StepOnePlus Real-Time PCR System (Applied Biosystems) and SensiFAST SYBR Hi-ROX mix (Bioline). Total reaction volume was 10 μl with 150 nM of each primer, 5 μl of SensiFAST SYBR Hi-ROX mix, 2 μl of cDNA template and 2.7 μl of nuclease-free H 2 O. qRT-PCR primers for smp_089290 were designed based on the sequence verified smp_089290 amplicon and included forward (5'-CAC GAC TGA TGG TGG TGG-3') and reverse (5'-CTC GAT ACA TCA TTA TCC CGC T-3') pairs. smrgbd primers included forward (5'-CGG CTT TAA CTC GCC TAC AC-3') and reverse (5'-CAT TCG ACG GTT GTT CAC AC-3') pairs. smtscd primers included forward (5'-GCC AAC AAA TTT CGT GGT CT-3') and reverse (5'-TTC ACC ATT TCG GTC GTA CA-3') pairs. smat1 primers included forward (5'-CTT CGA ACC AGC AAA TCA GA-3') and reverse (5'-GAC ACC AAT CCA CAA ACT GG-3') pairs. To determine smp_089290 knockdown (KD) in adult worms following RNAi and CRISPR/Cas9 programmed knockout, smp_089290 transcript levels in siRNA treated samples (at 48 hr post electroporation) and CRISPR/Cas9 plasmid treated samples (at seven days post electroporation) were quantified relative to smat1 as described previously [64]. The Geometric mean method was used to quantify smp_089290 transcript levels in freshly perfused male and female schistosomes as described previously [65,66]. Melting curves (dissociation curves) were generated for each qRT-PCR analysis to verify the amplification of one product only. A two-tailed student's t-test was used to test for significance between siRNA treatments. A Kruskal-Wallis ANOVA with Dunn post-hoc comparisons was used to test for significance between CRISPR/Cas9 plasmid treatments. A two-tailed student's t-test was used to test for significance between freshly perfused adult male and female worms.

Whole mount in situ hybridisation (WISH)
The full coding sequence of Smp_089290 was amplified using forward (5´-ATG GCT ACC GTA CCA CCG-3´) and reverse (5´-CTA TAA CAA TGT CTG AAA CAG TCC ATC-3´) primer pairs. The PCR products were subsequently cloned into the pJC53.2 vector [67] using standard cloning methods as mentioned previously. These recombinant plasmids were subsequently used to generate digoxigenin-labelled riboprobes using the Riboprobe System (Promega) with SP6 or T3 RNA polymerases and digoxigenin-labelled Uracil triphosphate (Roche) [68]. Antisense treatment riboprobes (generated by SP6 polymerase) and sense control riboprobes (generated by T3 polymerase) were processed as described [68] and stored at -20˚C until needed for the WISH staining protocol. Riboprobes were used within 2 wk of their initial storage at -20˚C.
Adult male and female worms were relaxed and separated by incubation (15 min) in a 0.25% solution of the anaesthetic ethyl 3-aminobenzoate methanesulphonate (Sigma-Aldrich) dissolved in DMEM [69]. Relaxed parasites were subsequently killed in a 0.6 M solution of MgCl 2 (1 min) and fixed for 4 hr in 4% formaldehyde in PBSTx (PBS + 0.3% Triton X-100). Fixed parasites were dehydrated in MeOH and stored at -20˚C until needed. Parasite samples were rehydrated in 1:1 MeOH:PBSTx (5-10 min) followed by incubation in PBSTx (5-10 min). Rehydrated parasite samples were bleached in formamide bleaching solution (0.5% v/v de-ionised formamide, 0.5% v/v SSC and 1.2% w/w H 2 O 2 , brought to a final volume of 10 ml with diethyl pyrocarbonate H 2 O) for 90 min, rinsed twice in PBSTx, treated with proteinase K (10 μg/ml, Invitrogen) for 45 min at room temperature and post-fixed for 10 min in 4% formaldehyde in PBSTx. Parasite samples were processed as previously described [67,68] with several modifications. Antisense treatment riboprobes and sense control riboprobes were mixed with hybridisation solution (50% v/v de-ionised formamide, 10% w/v dextran sulphate, 1.25% v/v SSC, 1 mg/ml yeast RNA, 1% v/v Tween-20, brought to a final volume of 40 ml with diethyl pyrocarbonate H 2 O) and hybridised overnight at 52˚C. Parasite samples were transferred to fresh colorimetric developing solution consisting of alkaline phosphatase buffer (100 mM Tris pH 9.5, 100 mM NaCl, 50 mM MgCl 2 , 0.1% v/v Tween-20, brought to a final volume of 10 ml with 10% polyvinylalcohol solution) supplemented with 4.5 μl/ml NBT (Roche) and 3.5 μl/ml BCIP (Roche). All parasite samples treated with the antisense treatment riboprobe were developed at room temperature in the dark until the desired level of purple signal was reached (male samples = 2 hr, female samples = 45 min). In parallel, all parasite samples treated with the sense control riboprobe were developed for the same length of time. Once the desired level of signal was reached, the colorimetric developing solution was removed and worms were rinsed twice in PBSTx to stop any further development from occurring. Worms were dehydrated in 100% ethanol for 5 min and subsequently submerged in 80% glycerol in 1x PBS and incubated overnight at 4˚C. Thereafter, stained worms were mounted onto microscope slides and examined with a light microscope (Leica LMD6000 Laser Microdissection Microscope).

α-GAL/α-NAGAL enzymatic activity measurements
Soluble worm antigen preparation (SWAP) was derived from worms (freshly perfused or electroporated with siRNAs) removed from culture after 1 wk. Worms were homogenised with a 5 mm diameter stainless steel bead (Qiagen) for 4 min at 50 Hz in 100 μl of 0.15 M McIlvaine buffer pH 4.6 [72] with EDTA-free protease inhibitors (Sigma-Aldrich) using a TissueLyser LT (Qiagen). After homogenisation, tubes were centrifuged at 21,100 x g for 30 min at 4˚C and SWAP was collected and quantified by the Bradford method (Sigma-Aldrich).
The enzymatic activity of SWAP derived from freshly perfused adult male and female worms, siLuc treated and siSmp_089290 treated adult male and female worms was measured using 4-Nitrophenyl α-D-galactopyranoside (α-GAL colorimetric substrate) and 4-Nitrophenyl N-acetyl-α-D-galactosaminide (α-NAGAL colorimetric substrate) in separate reactions. Differing concentrations of human α-GAL (Fabrazyme, kindly provided by the Leiden University Medical Centre, LUMC) and commercially sourced α-NAGAL cloned from Chryseobacterium meningosepticum and expressed in E. coli (New England Biolabs, Ipswich, MA) were measured with corresponding substrates.
Enzyme assays were performed in standard flat-bottomed 96 well plates (STARLAB). α-GAL/α-NAGAL assays consisted of 100 μl of α-GAL or α-NAGAL substrate dissolved in 0.15 M McIlvaine buffer (pH 4.6) with the addition of differing concentrations of α-GAL or α-NAGAL and 0.15 M McIlvaine buffer pH 4.6 to give a final volume of 125 μl per well. Concentrations of α-GAL used to generate a standard curve of activity were as follows: 0.95, 0.475, 0.19, 0.0475, 0.0095, 0.00475 and 0.0019 μg/ml. Concentrations of α-NAGAL used to generate a standard curve of activity were as follows: 20, 15, 10, 5, 2 and 1 μg/ml. SWAP reactions were set-up in the same way but α-GAL/α-NAGAL enzymes were replaced with different quantities of sample specific SWAP and incubated with both α-GAL and α-NAGAL substrates. For untreated adult male and female worms, 5 μg of SWAP was used per well to enable comparison between gender. For siLuc treated and siSmp_089290 treated adult male and female worms, 6.45 μg and 1.08 to 2.44 μg of SWAP was used per well to enable comparison between siRNA treatments, respectively. Reaction proceeded for 60 min at 37˚C and terminated by the addition of 70 μl of 0.4 M glycine (pH 10.4) [72]. Final absorbances were quantified at 410 nm using a POLARstar Omega microplate reader (BMG Labtech). Absorbance values produced from different SWAP treatments were compared to α-GAL/α-NAGAL standard curves to calculate α-GAL and α-NAGAL activities (μg/ml). A two-tailed student's t-test was used to test for significance between siRNA treated samples or between genders.

Motility analysis quantified by WormAssayGP2 and adult worm scoring matrix
The digital image processing-based system known as WormAssayGP2 was derived from Marcellino et al. [73] and implemented by us as previously described [74]. Individual wells containing up to 10 adult worms were recorded for 60 sec each day and analysed by the Lucas-Kanade algorithm. Once recording was completed, the data were quantified and stored as a .csv file which was further processed to calculate the mean motility for the control and treatment groups. In parallel, adult worm motility was also scored using the WHO-TDR scoring matrix [75]. All worms were scored daily from two days after electroporation until seven days after electroporation. Worms were ranked between 4-0 based on their motility; 4 = normal active/ paired up, 3 = full body movement but slowed activity, 2 = minimal activity, occasional movement of head and tail only, 1 = movement in the suckers only or slight contraction of the body and 0 = total absence of motility. For both motility analyses, a General Linear Mixed-Effects Model was fitted to each dataset ('NLME' package) and statistical differences were determined by performing pairwise comparisons of the estimated marginal means of each group per time point ('EMMEANS' package) in R. Video footage of adult worms was captured using a Nexius-Zoom stereo microscope (Euromex) and edited with ImageFocus 4 software (Euromex).

Enumeration of vitellocytes and egg volume measurements
The eggs released by adult female worms from RNAi and CRISPR/Cas9 genome editing experiments were collected and fixed in 1 ml 10% neutral buffered formalin solution (Sigma-Aldrich) for 24 hr at 4˚C. Thereafter, eggs were enumerated using a Sedgewick Rafter Counting Chamber [76]. Prior to visualisation by laser scanning confocal microscopy (LSCM), stored eggs from RNAi experiments only were immersed in PBS supplemented with DAPI (4',6-diamidino-2-phenylindole, 2 μg/ml). Fluorescence images (10 eggs per treatment) were captured on a Leica TCS SP8 super resolution laser confocal microscope fitted with a 63x (water immersion) objective using the Leica Application Suite X (LAS X). Green (egg

PLOS PATHOGENS
Characterisation of SmNAGAL autofluorescence) fluorescence was visualised with an argon or diode-pumped, solid state (DPSS) laser at 488 nm. DAPI was visualised using a 405 nm blue diode laser. The number of vitellocytes (DAPI + cells) and overall volume (mapped by the green autofluorescence) for individual eggs were calculated using IMARIS 7.3 software (Bitplane). A two-tailed student's t-test was used to test for significance between siRNA treated samples with regard to the total number of eggs produced and individual egg volume. A Kruskal-Wallis ANOVA with Dunn posthoc comparisons was used to test for significance between CRISPR/Cas9 plasmid treated samples with regard to the total number of eggs produced. IMARIS 7.3 software (Bitplane) was also used to create a video showing the 360˚horizontal rotation of a representative egg from each siRNA treatment.
Among these five S. mansoni gene products, only one (Smp_089290, highlighted yellow, Fig 1) contained all 13 ligand binding residues (green shaded amino acids, Fig 1) as well as both catalytic aspartic acid residues (light blue Ds, Fig 1) critical for α-NAGAL activity [32]. MS3_11280 (S. haematobium homolog) and EWB00_005284 (S. japonicum homolog) also shared these diagnostic characteristics. Furthermore, while two additional S. japonicum homologs (EWB00_005283 and EWB00_005285) contained both catalytic aspartic acid residues, they did not possess all 13 ligand binding residues. These five schistosome proteins, containing the essential catalytic aspartic acid residues, clustered into a separate clade discrete from other putative schistosome GH27 family members (Fig 2).
The predicted three-dimensional structure of Smp_089290 was derived by homology modelling and passed all stereochemical quality assessments (S1 Fig). Homology modelling of Smp_089290 as a monomer (HsNAGAL functions as a homodimer [32]) revealed the positioning of these 13 ligand binding residues and two catalytic residues around the putative active site (purple, Fig 3). Despite the melibiase-2 C-terminal domain (red, Fig 3) not possessing any residues involved in ligand binding and substrate cleavage mechanisms, the predicted Smp_089290 model and H. sapiens α-NAGAL structure are consistent across this region. Both

PLOS PATHOGENS
Characterisation of SmNAGAL the predicted Smp_089290 model and H. sapiens α-NAGAL structure possess eight anti-parallel β-strands (Fig 3) [33]. Overall, the homology modelling analysis predicted that Smp_089290 likely possesses folding topology and spatial arrangements that closely resemble typical α-GAL and α-NAGAL proteins. Phylogenetic analyses were conducted using a concatenated multiple sequence alignment of Schistosoma proteins that contained >70% of the residues within the

Characterisation of SmNAGAL
Together, these data suggest that the S. mansoni genome encodes a single gene product (Smp_089290) containing all essential catalytic amino acid residues for hydrolysis and release of α-N-acetylgalactosamine from glycosylated substrates. Therefore, this putative schistosome α-N-acetylgalactosaminidase (SmNAGAL, Smp_089290) was taken forward for further transcriptional, enzymatic, and functional genomics studies.

Smnagal is developmentally regulated, female-enriched and localised to vitellaria, mature vitellocytes and parenchymal cells
To begin deciphering SmNAGAL function, both transcriptomic-and enzymatic-based approaches were initiated (Fig 4). Meta-analysis of historical DNA microarray data across the S. mansoni lifecycle [45] was facilitated by two 50-mer oligonucleotide probes that retained melibiase-2 domain and α-GAL/α-NAGAL protein types across phyla. Proteins were analysed using both Maximum Likelihood and Bayesian inference approaches. Branch lengths (indicated by scale bar) represent distance among different taxa as predicted by the Bayesian inference approach. Node labels outside parentheses represent Bayesian posterior probability support values whilst those within parentheses represent percentage bootstrap support values from Maximum Likelihood analysis. Schistosoma proteins conserving catalytic aspartic acid residues are highlighted in the blue dashed box. The phylogram includes protein sequences from S. mansoni (

Characterisation of SmNAGAL
100% base-pair complementarity to exon 2 (CONTIG7235) and exon 3 (CONTIG6265), respectively, of smnagal (Fig 4A). For each of these two oligonucleotides, similar patterns of smnagal abundance were deduced. While smnagal expression was low in eggs, it increased in miracidia only to wane as schistosome development (sporocysts-cercariae) continued in the molluscan host (Fig 4B). Upon early intra-mammalian schistosome maturation (3 hr-3 day schistosomula), smnagal transcription remained invariably low, until day six post schistosomula transformation. At this point and extending into more developmentally mature lifecycle forms (3 wk-7 wk schistosomes), smnagal expression increased, reaching peak abundance in 7 wk old schistosomes. Here, female-enriched smnagal expression was clearly observed in adult schistosomes, which was independently confirmed by qRT-PCR (Fig 4C). Additionally, female-biased expression of smnagal was observed when interrogating S. mansoni RNA-Seq meta data (S2 Fig) [47]. Further support for female biased expression of smnagal was obtained from the analysis of α-N-acetylgalactosaminidase activity in adult worm protein extracts (Fig  4D). In a direct comparison, female protein extracts contained significantly higher levels of α-NAGAL activity when compared to males. This female biased trend was also observed with α-GAL activity measured in the same extracts (Fig 4D).
Expanding these temporal investigations of smnagal abundance and enzymatic activities to spatial localisation, by WISH and scRNA-Seq, in adult worms revealed additional gender-specific traits. An antisense RNA probe spanning the full length coding sequence of smp_089290 was used for each WISH localisation experiment (Fig 5) [68]. A negative control was prepared using a sense smp_089290 probe (S3 Fig); no specific staining was observed. While the ovary (containing mature and immature oocytes) was not a rich source of smnagal expression, the vitellarium (and mature vitellocytes passing through the vitello-oviduct) was highly enriched for this putative α-NAGAL gene product (Fig 5A). These observations were supported by scRNA-Seq approaches, which showed prominent smnagal abundance within the vitellaria and mature vitellocytes and no expression within female gametes (S4 Fig). In addition, low levels of smnagal expression were found within the parenchyma, which was more clearly observed in the female scRNA-Seq plots (S4 Fig). Moderate smnagal expression was also observed in some clusters of neuronal cells, tegument lineage cells and muscle cells when inspecting the female scRNA-Seq plots. No appreciable expression was found in any other female tissue examined. The WISH analysis of males revealed that smnagal expression was predominantly localised to parenchymal cells widely distributed throughout the body (Fig 5B), which was confirmed by the male scRNA-Seq expression profile plots (S5 Fig). A lack of staining was additionally observed throughout the primary reproductive organs (testes), which was supported by scRNA-Seq profiles. WISH analyses also showed that cells lining the tegument and intestine lacked intense smnagal expression in both genders, although this was easier to deduce in males. Similar to females, small yet noticeable levels of smnagal expression were evident in some neuronal cell clusters when inspecting male scRNA-Seq plots (S5 Fig). Appreciable expression was not observed elsewhere in any other male tissue examined.

α-N-acetylgalactosaminidase activity is required for worm motility, egg production and development
The localisation of smnagal to adult parenchymal cells, neuronal clusters and mature vitellocytes as well as the well-documented neuromuscular defects characteristic of Schindler/ well) were used for both sexes and measured for α-NAGAL and α-GAL activity on diagnostic α-NAGAL and α-GAL substrates. Statistical significance is indicated (Student's t-test, two tailed, unequal variance, � = p<0.05 and �� = p<0.01). https://doi.org/10.1371/journal.ppat.1009828.g004

Characterisation of SmNAGAL
Kanzaki disease (due to α-nagal deficiencies, [23,24]) implicated key roles for this gene product in schistosome motility, oviposition and development. Therefore, to assess whether these processes were dependent upon α-NAGAL activity, functional genomics investigations of smnagal/smp_089290 were conducted in adult schistosomes (Fig 6). RNA interference (RNAi) of smnagal/smp_089290, using small interfering RNAs (siRNAs), led to a highly significant knockdown (92%) of smnagal in adult male worms when compared to controls (Fig 6A). In parallel, genome editing approaches were implemented with CRISPR/Cas9 plasmid constructs and used to complement RNAi experiments. The CRISPResso2 pipeline was utilised to

PLOS PATHOGENS
Characterisation of SmNAGAL quantify NHEJ-associated mutations within aligned MiSEQ library sequencing reads from genome-edited worms (S1 Table). Oligonucleotide primers designed for MiSEQ library generation were designed to encompass the programmed DSB sites for each sgRNA used (S6 Fig). All genome edited samples showed detectable levels of genome editing (0.25-0.31%) with substitutions being the most frequently observed mutation (S2 Table). Further analyses of modified sequence reads revealed all CRISPR/Cas9 plasmid treatment samples displayed NHEJassociated mutations predicted to introduce frameshifts at the smnagal locus (likely resulting in the translation of substantially truncated proteins) or to ablate smnagal transcription (S7 Fig). Similar to RNAi, CRISPR/Cas9 genome editing led to significant knockdowns (46-60%) of smnagal in mixed-sex adult worms when compared to controls (Fig 6B). Reassuringly, smnagal depletion in siRNA-treated worms significantly reduced α-NAGAL activity in SWAP derived from both males and females when compared to siLuc controls (Fig 6C). However, RNAi-mediated smnagal knockdown did not result in a significant reduction in SWAPderived α-GAL activity when compared to siLuc treated schistosome samples (S8 Fig).
Having established that smnagal encodes a functional α-NAGAL and given that RNAi and CRISPR/Cas9 depleted this transcript from intracellular RNA pools, motility and egg-laying phenotypes of siSmnagal treated schistosome pairs were subsequently examined and quantified. Regardless of the quantification metric used (WormAssayGP2 [73,74] or WHO-TDR standards [75]), a clear motility defect was observed in both male and female schistosomes when smnagal was depleted by RNAi (Figs 7A and S9 and S1 Video). This motility defect was apparent by day two post RNAi and maintained until day 7, when the assay was terminated. smp_089290 (siSmp_089290). After 48 hr, total RNA was isolated and used to generate cDNA, which was subjected to qRT-PCR. Percent knockdown (KD) is indicated. Statistical significance is indicated (Student's t-test, two tailed, unequal variance, �� = p<0.01). (B) 7 wk old adult male and female schistosomes were electroporated with either lentiviral CRISPR/Cas9 plasmid constructs targeting exon 1 (SmNAGALX1) or exon 2 (SmNAGALX2) of smnagal. Additionally, both exons were targeted by electroporating a mixture of both plasmid constructs (SmNAGALX1/X2). Electroporations with CRISPR/Cas9 plasmid DNA containing a Scramble sgRNA were used as a control. After seven days, total RNA was isolated and used to synthesise cDNA for qRT-PCR analysis. Percent knockdown (KD) is indicated. Statistical significance is indicated (Kruskal-Wallis ANOVA with Dunn post-hoc comparisons, � = p<0.05). (C) α-NAGAL activity was measured in SWAP derived from siLuc treated and siSmp_089290 treated adult male and female worms (6.45 μg for males and 2.44 μg for females) as described above. Statistical significance is indicated (Student's t-test, two tailed, unequal variance, � = p<0.05 and �� = p<0.01). https://doi.org/10.1371/journal.ppat.1009828.g006

Characterisation of SmNAGAL
Upon completion of the RNAi assays (day seven), eggs were collected from in vitro cultures and quantified for number, volume and retention of mature vitellocytes. Here, a significant reduction in the quantity of deposited eggs was associated with smnagal deficiency (Fig 7B). The CRISPR/Cas9 genome editing approach further supported these observations with all smnagal-edited worms displaying a significantly reduced number of deposited eggs when compared to controls (Fig 7C). Notably, eggs derived from siRNA-treated female worms were also significantly smaller than those collected from wells of siLuc treated worms (Fig 7D) and contained abnormally spaced vitellocytes (Figs 7E and S10 and S2 Video).
Mature vitellocytes contain two types of large cytoplasmic inclusions known as vitelline droplets and lipid droplets, which both coalesce to assist in eggshell formation [79,80]. Furthermore, eggshell formation also requires phenol oxidase-mediated protein cross-linking (quinone tanning) facilitated by tyrosinase activity originating from mature vitellocytes [77,79,[81][82][83]. To further investigate the potential role of SmNAGAL in vitellaria biology and egg-laying, siLuc treated and siSmnagal treated adult female worms were either stained with Fast Blue BB (to visualise vitelline droplets in mature vitellocytes) or used for diphenol oxidase localisation (to assess tyrosinase activity in mature vitellocytes) assays and imaged by bright-field microscopy. In both assays, no differences were observed between siRNA treatments (S11 Fig).

Discussion
The identification and inhibition of gene products responsible for essential developmental or gender-associated processes provide a pathway by which schistosome drug discovery can progress rationally. To fast track such investigations, an extensive collection of putative S. mansoni drug candidates is currently available within the TDR Targets database [84]. Alongside this resource, multiple reports have recently described how genome sequencing outputs can be effectively leveraged by both cheminformatics and functional genomics for characterising next-generation schistosome drug targets and chemotherapeutics [74,[85][86][87][88][89][90]. Complementary evidence is provided here to support an essential role for SmNAGAL in the regulation of worm movement and reproductive processes.
SmNAGAL (Smp_089290) encoded by S. mansoni, MS3_11280 encoded by S. haematobium and EWB00_005284 encoded by S. japonicum were the only melibiase-2 domain-containing Schistosoma proteins conserving all functionally important amino acid residues necessary for the hydrolysis of α-galactose and α-N-acetylgalactosamine residues from glycolipid and glycoproteins (Fig 1). Closer inspection of these residues showed these three proteins possessed the exact same ligand binding residues to H. sapiens α-NAGAL, G. gallus α-NAGAL and C. elegans gana-1, which are all α-NAGAL enzymes with hydrolytic activity against terminal α-N-acetylgalactosamine and α-galactose moieties [25,32,33]. Accordingly, it is likely that Smp_089290, MS3_11280 and EWB00_005284 would exhibit enzymatic activity towards both terminal α-N-acetylgalactosamine and α-galactose residues unlike α-GAL enzymes, which can only cleave terminal α-galactose residues. Studies characterising human α-GAL/α-NAGAL activity provide evidence that H. sapiens α-GAL cannot use α-N-acetylgalactosamine as a substrate due to steric hindrance mediated by Glu 203 and Leu 206 [31,32]. In contrast, H. sapiens α-NAGAL, Smp_089290, MS3_11280 and EWB00_005284 all possess Ser and Ala at homologous positions (Ser 188,168,168,197 and Ala 191,171,171,200 ), which are required/essential for using α-N-acetylgalactosamine as a substrate [32]. Additionally, EWB00_005283 and EWB00_005285 encoded by S. japonicum contain both catalytic aspartic acid residues (Fig 1), which suggests that S. japonicum possesses three melibiase-2 domain-containing proteins capable of enzymatic activity. However, neither of these S. japonicum homologs possess all 13 ligand binding residues, which may influence affinity to target substrates. Although mutagenesis studies

PLOS PATHOGENS
Characterisation of SmNAGAL performed on Pichia pastoris α-GAL and α-NAGAL showed Trp 16 to be essential for enzymatic activity [91], little is known on the essentiality of the other ligand binding residues. Nevertheless, phylogenetic analyses reinforced that Smp_089290, MS3_11280, EWB00_005284, EWB00_005283 and EWB00_005285 are schistosome α-NAGALs as all demonstrated stronger relations to representative α-NAGAL proteins compared to representative α-GAL proteins (Fig 2). Whether the other GH27 family members identified here contribute to the α-GAL activities measured within adult worm extracts (Fig 4D) has yet to be determined.
Interrogating DNA microarray (Fig 4B) and RNA-Seq meta-analysis (S2 Fig) databases provided the first insight to smnagal's temporal expression profile across the developmental stages of the schistosome. In both cases, expression of smnagal increased throughout intramammalian schistosome development until full adult worm maturation, suggesting a potential role for SmNAGAL in adult schistosome development within the definitive human host. Confirmation of smnagal expression in adult male and female worm stages by qRT-PCR analyses (Fig 4C) and dominant α-NAGAL activity as shown by enzymatic assays (Fig 4D) suggests SmNAGAL preferentially cleaves off α-N-acetylgalactosamine (and not α-galactose) residues from glycan substrates (functionally confirmed by RNAi , Figs 6 and S8). Homology modelling suggests SmNAGAL possesses a pattern of α-helices and β-strands throughout its N-terminal domain (Fig 3) comparable to the (β/α) 8 barrel structure commonly observed in GH27 family members [31-33, [92][93][94]. Furthermore, all enzymatically important residues were shown to be arranged in an exposed catalytic active pocket, which suggests SmNAGAL utilises a double displacement mechanism for substrate binding and cleavage (Fig 3). This type of cleavage mechanism is commonly utilised by retaining GH enzymes (yielding a product that possesses the same anomeric configuration as the cleaved substrate) and involves two nucleophilic attacks on the 1-carbon of the substrate [33, [95][96][97][98].
Amongst a variety of glycans and glycoconjugates in adult schistosomes is the O-glycopeptide Galβ1-3GalNAcα1-Ser/Thr (also known as the oncofetal Thomsen-Friedenreich antigen or TF antigen), which is an abundant α-N-acetylgalactosamine-containing structure [99]. The TF antigen is present on the surface syncytium and may be involved in protecting tegumental structures that are essential for schistosome survival within the vasculature of the human host [99][100][101]. Furthermore, the TF antigen has been suggested to interfere with the functions of host Kupffer cells and hepatocytes [99,102]. Therefore, SmNAGAL activity required for hydrolysis of the GalNAcα1-Ser/Thr linkage during O-glycopeptide degradation may have an impact on adult worm tegument metabolism and host interactions. Expression of smnagal in adult female and male tegument lineage cells identified by scRNA-Seq analyses (S4 and S5 Figs) further supports a potential role for SmNAGAL in tegument metabolism and host interactions. Parenchymal expression of smnagal (Fig 5) and noticeable abundance in other cell types might be explained by the reported localisation of α-NAGAL enzymatic activity in lysosomes [103]. Lysosomes are found throughout many different S. mansoni tissues [104], which suggests other functional roles for SmNAGAL in addition to those identified in this study. Clearly, these observations require further exploration.
Knockdown of smnagal correlated with striking motility defects (Figs 7A and S9) as early as day two post siRNA treatment. Smaller reductions in motility observed in siSmNAGAL treated female worms may be due to the higher expression of smnagal in female worms compared to male worms. Higher mRNA turnover rates and transcript abundances have been shown to be a limiting factor of siRNA efficiency when targeting DGKE and ARHGAP27 kinases in HeLa and HepG2 cells [105]. Regardless of this potential discrepancy of RNAi efficiency between the sexes, the motility defect became more severe at day three (Figs 7A and S9 and S1 Video) and was maintained for the seven day experiment. The observed abnormal motility defects were consistent with the neurological and neuromuscular impairments associated with the human lysosomal storage disorder known as Schindler/Kanzaki disease (human α-NAGAL deficiency) [23,24,32]. Symptoms of Schindler/Kanzaki disease include a wide range of clinical neurological/neuromuscular deficits due to the accumulation of substrates possessing α-N-acetylgalactosamine residues, which are grouped into three distinct types [32]. The most severe form, type I, is characterised by stiff movements (spasticity) caused from involuntary muscle spasms, developmental retrogression, decorticate posturing, profound psychomotor retardation and muscular hypotonia, which begins in infancy [24, [106][107][108]. Worms substantially depleted of smnagal/SmNAGAL activity display motility defects consistent with the spasticity associated with type I Schindler/Kanzaki disease. Due to the neurological nature of Schindler/Kanzaki disease, the onset of impaired motility phenotypes may be due to smnagal depletion within neuronal cell clusters in both adult female and male worms (scRNA-Seq expression profiles ; S4 and S5 Figs). Similarly, female muscle cells also possess moderate smnagal levels, which suggests the onset of RNAi-mediated motility phenotypes could be driven by depletion in muscle cell activity and, therefore, collectively characterised as a neuromuscular impairment. However, minimal smnagal expression observed in male muscle cells suggests the abnormal motility phenotype is exclusively associated with neurological impairments in this sex. Regardless of the differential molecular mechanisms involved, it is clear that SmNAGAL also contributes to coordinated movement in adult schistosomes.
A fundamental key difference between the male and female schistosome centres on production of eggs, which leads to the pathology and transmission of schistosomiasis [109,110]. An important organ necessary for egg production is the vitellarium, which extends throughout the majority of the female worm and is involved in the production of mature vitellocytes [80,111]. Vitellocyte maturation within the vitellarium progresses through four stages involving cell division and differentiation to ultimately produce mature vitellocytes (also referred to as stage 4 vitellocytes) [112]. Mature vitellocytes are transported through the vitello-oviduct and surround the ovum, which is initially produced in the ovary [113][114][115]. The rigid insoluble eggshell is subsequently synthesised by phenol oxidase-mediated protein cross-linking (quinone tanning) as a result of increased tyrosinase activity originating from late/mature vitellocytes [77,79,[81][82][83]. Subsequently, the egg enters the uterus and is expelled through the gonopore of the female into the blood [83]. The elevated expression of smnagal within the vitellaria and mature vitellocytes traversing the vitello-oviduct (Figs 5A and S4) suggest SmNAGAL may be involved in aspects of egg production. The detection of smnagal expression within the egg (Figs 4B and S2) may also be explained by the presence of mature vitellocytes and strongly supports a role for SmNAGAL in vitellogenesis and oviposition, which was subsequently confirmed by both RNAi and programmed gene knockout (Fig 7B and 7C). Here, diminished numbers of eggs produced by smnagal depleted adults exhibited two predominant abnormalities; lack of typical vitellocyte structuring and spacing (Figs 7E and S10 and S2 Video) and reductions in egg volume (Fig 7D). smnagal deficiency, however, does not influence vitellocyte incorporation as DAPI + cells are observed in all eggs obtained from siSmNAGAL treatment groups. Reductions in SmNAGAL enzymatic activity (S11A Fig) also does not appear to influence the presence of vitelline droplets (S11B Fig) or the activity of tyrosinase (S11C Fig) in mature vitellocytes (within the vitellaria). While both of these features (along with lipid droplets) contribute to the formation of hardened/tanned eggs [77,[79][80][81][82][83], our data would suggest that their presence is unaffected by smnagal knockdown. Instead, SmNAGAL activity seems to affect an area of egg-laying biology (involving the vitellaria and vitellogenesis) that is currently unknown, but is associated with normal spacing of mature vitellocytes within the in vitro laid eggs (IVLEs) and total egg volume (i.e. siSmNAGAL treatment groups-approximately 15,000 μm 3 per egg; siLuc treatment groups-approximately 35,000 μm 3 per egg; Fig 7D). In some cases, smnagal deficiency also contributed to additional observed phenotypes such as abnormal shaped eggs and incomplete development of the lateral spine (S10 Fig). Therefore, in addition to adult worm motility, SmNAGAL clearly participates in vitelline droplet/tyrosinase-independent mechanisms of oviposition. In light of this evidence, S. japonicum may need three enzymatically active α-NAGAL proteins (EWB00_005284, EWB00_005283 and EWB00_005285, Fig 1) due to markedly higher rate of oviposition (>2000 eggs per day per worm pair) compared to S. mansoni and S. haematobium (>300 and >200 eggs per day per worm pair, respectively) [116,117]. Further characterisation of these other schistosome homologs could provide further insight into the differential rate and absolute numbers of eggs produced by the three main, human-infecting schistosome species.
The use of RNAi to characterise S. mansoni genes of interest [74], such as smnagal, has been established for several years and is currently considered the functional genomics gold standard for this parasite. However, CRISPR/Cas9 genome editing approaches to characterise schistosome gene function are increasingly being explored as recently exemplified by studies of S. mansoni omega-1 [118], ache [119], and sult-or [120]. Our use of this technology to edit smnagal contains some broadly-overlapping similarities to these previous studies. For example, the overall percentages of modified sequence reads observed in smnagal-edited worms (0.25-0.31%) were comparable to those observed in omega-1-edited eggs (approximately 4.5%) [118], ache-edited eggs (0.0295-0.12%) [119], sult-or-edited worms (0.3-2.0%) and sult-oredited sporocysts (0.1-0.2%) [120]. Furthermore, the predominant types of mutation in smnagal-edited worms were substitutions (S2 Table), which is consistent with omega-1-edited eggs [118] and ache-edited eggs [119]. The CRISPR/Cas9 investigation targeting the liver fluke granulin (ov-grn-1) locus in Opisthorchis viverrini also showed substitutions (98.7%) to be introduced at a higher rate than insertions (0.6%) or deletions (0.7%) [121]. However, this observation was not consistent for the sult-or investigation, which only showed deletions attributable to genome editing [120]. Additional investigations are necessary in S. mansoni and other platyhelminths to confirm if predominant substitution rates resulting from programmed genome editing are a conserved feature. It is notable that worms treated with CRISPR/Cas9 plasmids targeting smnagal exon 1 (SmNAGALX1 and dual SmNAGALX1/X2) exhibited modified sequence reads with complex rearrangements consisting of insertions with deletions (S2 Table), which was not previously reported in the other S. mansoni genome editing studies to date. This may be due to the infrequency of this type of indel being introduced when DSBs are repaired after CRISPR/Cas9 genome editing [122]. Nonetheless, low percentages of S. mansoni genome editing quantified within omega-1-edited eggs and ache-edited eggs led to pronounced phenotypes [118,119]. This observation was also consistent for smnagaledited worms, which showed significant reductions in IVLE production (Fig 7C) similar to siSmNAGAL treated worms (Fig 7B). Further comparisons can be made between smnagaledited worms and omega-1-edited eggs in which both manipulated groups showed significant reductions in target transcript abundance. However, this is in contrast to the CRISPR/Cas9 study targeting sult-or, which showed no mRNA knockdown or expected phenotypes despite NHEJ-associated deletions predicted to cause frameshifts that ablate sult-or transcription [120]. Collectively, these results suggest that phenotypic effects/reductions in mRNA abundance associated with RNAi-treated S. mansoni parasites targeting a particular transcript may not always be equivalent to those found in CRISPR/Cas9-edited S. mansoni parasites that target the associated gene locus. Furthermore, the overall percentages of genome editing may be underestimated in smnagal-edited worms due to the presence of large deletions completely removing the primer regions over exon 1 and 2 and, thus, some mutations remain undetected by amplicon sequencing and bioinformatics analysis of the alleles. This situation has been reported with S. stercoralis [123] and C. elegans [124] and speculated to occur in the reports on omega-1 [118], sult-or [120] and ache [119].
In addition to the presence of large deletions, the omega-1 investigation suggested that several non-synonymous substitutions may have disrupted the ribonuclease catalytic site and contributed to the mutant phenotypes. Similarly, the smnagal genome edited-associated phenotypes observed (smnagal knockdown and reduced egg production) may also be due to NHEJ-associated indels/substitutions producing frameshifts (leading to the translation of substantially truncated proteins), ablating smnagal transcription or leading to translated proteins that lack all the functional amino acid residues (S7 Fig) [33,125]. The continued refinement of CRISPR/Cas9 technology in S. mansoni will help resolve some of these outstanding queries.
While the specific SmNAGAL targets within adult schistosomes have yet to be identified, our functional characterisation of this glycogene product suggests that glycoproteins/glycolipids containing α-N-acetylgalactosamine residues are critical for coordinated worm movement and egg production. These traits suggest that SmNAGAL is an essential schistosome gene product, representing a novel parasite vulnerability for exploiting further as a next-generation anthelminthic target for controlling schistosomiasis. As an important component of these investigations, we additionally confirm that smnagal is susceptible to somatic genome editing and contribute to the growing literature on utilising the CRISPR/Cas9 system in S. mansoni as a tool for functional genomics in parasitic platyhelminths [126]. An enhanced understanding of SmNAGAL or other S. mansoni glycan machinery components in lifecycle functions or host interactions will aid the search for urgently-needed, next-generation interventions.

Characterisation of SmNAGAL
for SmNAGALX1 (red), SmNAGALX2 (dark blue) and dual SmNAGALX1/X2 plasmid treated worms (purple). Nts that are modified by each indel/substitution within the alignment are coloured green. Insertions appear as additional nts not found in the original sequence, deletions appear as lines (-) and substitutions appear as replaced nts located in the same positions as the original sequence. The nt sequences of SmNAGALX1_sgRNA and SmNA-GALX2_sgRNA are highlighted in yellow within the original sequence. DSB sites for SmNAGALX1_sgRNA and SmNAGALX2_sgRNA are indicated by yellow arrows, which are located three nts upstream of the PAM sequence highlighted in light blue (note: the PAM sequence is outside and downstream of the sgRNA sequence). An additional colour code is used to distinguish between start (orange nts) and termination (red nts) codons found in the sequences, which have been highlighted in black. The motility of individual worms (five adult pairs per well) were scored between 4-0 based on the WHO-TDR scoring matrix guidelines; 4 = normal active/paired up, 3 = slowed activity, 2 = minimal activity and occasional movement of head and tail, 1 = absence of motility apart from gut movements and 0 = total absence of motility. The total occurrences of each score was plotted per day for siLuc treated and siSmNAGAL treated adult female (A) and male (B) worms starting from day two after electroporation up until day seven (day of electroporation considered as day zero). Statistical significance is indicated (General Linear Mixed-Effects Model, NLME and EMMEANS R packages, �� = p<0.01). Day one was not included in the statistical analysis due to worms appearing stunned and immobile, likely due to electroporation manipulation. Six wells/biological replicates per siRNA treatment were used for this analysis.  Table. Initial processing of MiSEQ deep-coverage sequence reads for CRISPResso2 analysis. The number of sequence reads at each of the three initial processing stages before indel characterisations can be made by further CRISPResso2 analysis is presented. These three initial processing stages are "Reads in inputs" (highlighted in red, first stage), "Reads after preprocessing" (highlighted in blue, second stage) and "Reads aligned" (highlighted in yellow, third stage). "Reads in inputs" refers to the total number of sequence reads from raw MiSEQ sequencing data. "Reads after pre-processing" refers to the number of sequence reads after PCR amplification or trimming artefacts are removed. "Reads aligned" refers to the number of sequence reads that are of high quality (>60% homology to reference amplicon sequence), which are used for indel characterisations. The table also lists the primer pair set and sample (samples amplified by SmNAGALX1_MiSEQ and SmNAGALX2_MiSEQ primers are highlighted in orange and green, respectively) used for each barcoded MiSEQ amplicon library constructed. (DOCX) S2 Table. Detectable frequencies of insertions, deletions, insertions with deletions and substitutions in smnagal-edited worms as quantified by CRISPResso2 analysis. The mutation frequencies attributable to genome editing (i.e. insertions, deletions, insertions with deletions, and substitutions) in smnagal-edited worms as quantified by CRISPResso2 analysis is indicated. The percentage of unmodified sequence reads is included. The primer pair set and sample (samples amplified by SmNAGALX1_MiSEQ and SmNAGALX2_MiSEQ primers are highlighted in orange and green, respectively) used for each barcoded MiSEQ amplicon library constructed are indicated.

PLOS PATHOGENS
Characterisation of SmNAGAL