Diminished adherence of Biomphalaria glabrata embryonic cell line to sporocysts of Schistosoma mansoni following programmed knockout of the allograft inflammatory factor

Background Larval development in an intermediate host gastropod snail of the genus Biomphalaria is an obligatory component of the life-cycle of Schistosoma mansoni. Understanding of the mechanism(s) of host defense may hasten the development of tools that block transmission of schistosomiasis. The allograft inflammatory factor 1, AIF, which is evolutionarily conserved and expressed in phagocytes, is a marker of macrophage activation in both mammals and invertebrates. AIF enhances cell proliferation and migration. The embryonic cell line, termed Bge, from Biomphalaria glabrata is a versatile resource for investigation of the snail-schistosome relationship since Bge exhibits a hemocyte-like phenotype. Hemocytes perform central roles in innate and cellular immunity in gastropods and in some cases can kill the parasite. However, the Bge cells do not kill the parasite in vitro. Methods Bge cells were transfected by electroporation with plasmid pCas-BgAIFx4, encoding the Cas9 nuclease and a guide RNA specific for exon 4 of the B. glabrata AIF (BgAIF) gene. Transcript levels for Cas9 and for BgAIF were monitored by reverse-transcription-PCR and, in parallel, adhesion of gene-edited Bge cells during co-culture with of schistosome sporocysts was assessed. Results Gene knockout manipulation induced gene-disrupting indels, frequently 1–2 bp insertions and/or 8–30 bp deletions, at the programmed target site; a range from 9 to 17% of the copies of the BgAIF gene in the Bge population of cells were mutated. Transcript levels for BgAIF were reduced by up to 73% (49.5 ± 20.2% SD, P ≤ 0.05, n = 12). Adherence by BgAIF gene-edited (ΔBgAIF) Bge to sporocysts diminished in comparison to wild type cells, although cell morphology did not change. Specifically, as scored by a semi-quantitative cell adherence index (CAI), fewer ΔBgAIF than control wild type cells adhered to sporocysts; control CAI, 2.66 ± 0.10, ΔBgAIF, 2.30 ± 0.22 (P ≤ 0.01). Conclusions The findings supported the hypothesis that BgAIF plays a role in the adherence of B. glabrata hemocytes to sporocysts during schistosome infection in vitro. This demonstration of the activity of programmed gene editing will enable functional genomics approaches using CRISPR/Cas9 to investigate additional components of the snail-schistosome host-parasite relationship.


Background
Evolution endowed the schistosomes with a complex life-cycle that includes both a freshwater gastropod intermediate host and a definitive mammalian host. Several species of the freshwater snail genus Biomphalaria are the intermediate host for Schistosoma mansoni. The Neotropical species, Biomphalaria glabrata has been studied extensively with respect to host-parasite relationship and coevolution with S. mansoni, especially on mechanisms of susceptibility and/or resistance to the compatible parasites [1,2]. Genetic variation is evident among isolates and strains of B. glabrata, both in the laboratory and in the field, resulting in a spectrum of the susceptibility of infection with S. mansoni [3]. Considerable advances have been made in the exploration and characterization of mechanisms of the internal defenses system (IDS) of the snail that determine susceptibility and resistance to schistosome [4][5][6][7][8][9][10][11]. The resistance phenotype is underpinned by a complex genetic trait, where the schistosome larva fails to develop as the consequence of innate and cellular immune responses. Hemocytes of resistant snails encapsulate and destroy the sporocyst [11][12][13][14][15][16][17][18].
Biomphalaria glabrata embryonic cell line (Bge) [19] remains to date the only established cell line from any mollusk. The cell line originates from 5-day-old embryos of B. glabrata susceptible to infection with S. mansoni. The Bge cell line has been studied extensively to interrogate the host-parasite relationship because the Bge cell exhibits a hemocyte-like behavior that includes encapsulation of the larval parasite, but does not kill the parasites [20][21][22][23][24][25][26][27][28].
The genome sequence of B. glabrata has been reported [29], along with ongoing transcriptome and proteome catalogues that include factors participating in immunological surveillance, phagocytosis, cytokine responses, and pathogen recognition receptor elements including Toll-like receptors and fibrinogen-related proteins [30][31][32][33][34][35][36]. An orthologue of the evolutionary conserved allograft inflammatory factor (AIF) is an evolutionary conserved protein typically expressed in phagocytes and granular leukocytes in both vertebrate and invertebrate. Functions demonstrated for AIF include macrophage activation, enhancement of cellular proliferation and of migration in mammalian and invertebrate cells; protostomes and deuterostomes [37][38][39][40][41]. AIF also plays a key role in the protective response by B. glabrata to invasion by schistosomes [8,9]. BgAIF, the orthologue in B. glabrata is expressed in hemocytes, which participate in phagocytosis, cellular proliferation, and cellular migration. Elevated expression of BgAIF is a characteristic of the resistance of B. glabrata to schistosome infection and has been considered as a marker of hemocyte activation [8,9].
Expression of AIF is also seen during hemocyte activation in oysters [36,38,42,43] and during hepatic inflammation during murine schistosomiasis [44,45]. We hypothesized that BgAIF was involved in cell mediated immune response(s) by B. glabrata through activation of hemocyte cell adhesion and/or migration after the schistosome miracidium has penetrated into the tissues of the snail. We addressed this hypothesis by using CRISPR/Cas9-based programmed genome editing to interrupt the BgAIF gene of B. glabrata in the Bge cell line, following reports that indicated the utility of using CRISPR-based programmed gene knockout approach in other mollusks including the Pacific oyster, Crassostrea gigas and the slipper limpet, Crepidula fornicata and the gastropod, Lymnaea stagnalis [46][47][48]. As detailed below, we demonstrated the activity of programmed genome editing in Bge cells, with gene knockout at the BgAIF locus.

Gene editing construct
The gene encoding the allograft inflammatory factor of B. glabrata, BgAIF (2226 bp; accession number BGLB005061, https ://www.vecto rbase .org/) includes five exons interrupted by four introns (Fig. 1a). A guide RNA (gRNA) for Cas9-catalyzed gene editing specific for the target B. glabrata gene locus, BgAIF, was identified in the BGLB005061 sequence using the 'CHOPCHOP' v3 tool, https ://chopc hop.cbu.uib.no/, with default parameters compatible for the protospacer adjacent motif, NGG, of Cas9 from Streptococcus pyogenes [49][50][51] and screened for off-target sites against the B. glabrata genome [29]. Based on the guidance from the CHOPCHOP analysis, we chose the top ranked guide RNA (gRNA), AGA CTT TGT TAG GAT GAT GC, specific for exon 4 of the AIF gene, with predicted high CRISPR/Cas9 efficiency for double-stranded cleavage in tandem with an absence of off-target activity in the genome of B. glabrata (Fig. 1a). A CRISPR/Cas9 vector encoding the gRNA targeting exon 4 of BgAIF under the control of the mammalian U6 promoter and encoding Cas 9, with nuclear localization signals 1 and 2, driven by the human cytomegalovirus (CMV) immediate early enhancer and promoter was assembled using the GeneArt CRISPR nuclease vector system (Thermo Fisher Scientific, Waltham, MA, USA), according to the manufacturer's protocol. Briefly, the 20 nt of either target (including 'GTTTT' on the 3′ end) or complementary to target (including 'CGGTG' on the 3′ end) sequences were synthesized commercially (Integrated DNA Technology, IA, USA), and annealed according to manufacturer's protocol. The annealed double-strand DNA (dsDNA) was ligated into the linearized GeneArt ® CRISPR Nuclease vector via BamHI and BsmBI restriction sites, respectively, and the construct was termed pCas-BgAIFx4 (Fig. 1b). (The sequence of GeneArt CRISPR nuclease vector backbone is available at https ://www.therm ofish er.com/order /catal og/produ ct/ A2117 4#/A2117 4). Chemically competent TOP10, E. coli cells (Invitrogen, Thermo Fisher Scientific) were transformed with pCas-BgAIFx4 by the heat shock method and cultured on LB-agar supplemented with ampicillin at 100 µg/ml. Subsequently, the integrity of the recombinant plasmids from several single colonies of ampicillin-resistant E. coli transformants was confirmed by amplicon PCR-based Sanger direct nucleotide sequence analysis using a U6 gene-specific primer for gRNA ligation and orientation (Fig. 1b).

Biomphalaria glabrata embryonic (Bge) cell line culture
The Bge cell line was provided by the Schistosomiasis Resource Center (SRC), Biomedical Research Institute (BRI), Rockville, MD, USA. Historically, the Bge cell line was sourced by the SRC from the American Type Culture Collection (Manassas, VA, USA), catalog no. ATCC CRL 1494, and thereafter maintained at BRI for > 10 years. Bge cells were maintained at 26 °C in air in 'Bge medium' , which is comprised of 22% (v/v) Schneider's Drosophila medium, 0.13% galactose, 0.45% lactalbumin hydrolysate, 0.5% (v/v) phenol red solution, 20 µg/ml gentamycin, and supplemented with 10% heat-inactivated fetal bovine serum [24,52]. Bge cells were grown to 80% confluence before transfection by electroporation with

Transfection of Bge cells by square wave electroporation
Bge cells were harvested using a cell scraper, washed twice in Bge medium, counted, and resuspended at 20,000 cell/µl in Opti-MEM medium (Sigma-Aldrich). Two million cells were transferred into 0.2 mm path length electroporation cuvettes (BTX Harvard Apparatus, Hollister, MA, USA) containing 6 µg pCas-BgAIFx4 in ~ 100 µl Opti-MEM. The cells were subjected to electroporation using one pulse at 125 V for 20 ms, using a square wave pulse generator (ECM 830, BTX Harvard Apparatus). Immediately thereafter, the Bge cells were maintained in 12-well plates (Greiner Bio-One, NC, USA) at 26 °C. The mock control included Opti-MEM only for electroporation. The presence of transcripts encoding the B. glabrata actin and the Cas9 was monitored daily for 9 days following transfection by electroporation (Fig. 1c).

Sequential isolation of total RNA and genomic DNA
To monitor the transfection of Bge cell by pCas9-BgAIFx4, we investigated the expression of Cas9 in Bge cells by reverse transcription PCR (RT-PCR). Both total RNA and genomic DNA were extracted sequentially from cell pellets, as described [53,54]. In brief, each sample of total RNA sample was extracted using RNAzol ® RT reagent (Molecular Research Center, Inc., Cincinnati, OH, USA) according to the manufacturer's protocol. Subsequently, the DNA/protein pellet retained after recovery of RNA was resuspended in DNAzol ® solution (Molecular Research Center, Inc), from which total DNA was recovered. The RNAs and DNAs were dissolved in nuclease-free water and their concentration and purity established by spectrophotometry (Nanodrop 1000, Thermo Fisher Scientific).

Expression of Cas9 in Bge cells
To investigate transcription from the pCas-BgAIFx4 vector following transfection of Bge cells, levels of transcribed Cas9 were investigated by semi-quantitative RT-PCR. The Cas9-specific primers were Cas9-F (5′-AGC ATC GGC CTT GAT ATC GG-3′) and Cas9-R (5′-AGA AGC TGT CGT CCA CCT TG-3′) (Fig. 1b). Total RNA from the non-transfected m ock (Opti-MEM electroporated-), and pCas-BgAIFx4 DNA electroporated-Bge cells were treated with DNase I (Ambion, Thermo Fisher Scientific) to digest any residual vector pCas-BgAIFx4 DNA and contaminating genomic DNAs. The RNAs were reverse transcribed to cDNA using ProtoScript II reverse transcriptase with oligo dT and random primers (First Strand cDNA Synthesis Kit, New England Biolabs, Ipswich, MA, USA). RT-PCRs specific for the Cas9 or actin gene of B. glabrata, BgActin (GenBank: U53348.1) were undertaken, with BgActin serving as the positive control for RNA integrity. The primer pairs used for the BgActin coding sequences were termed actin-F (5′-AAG CGA CGT TTT CTT GGT GC-3′) and actin-R (5′-ACC CAT ACC AAC CAT CAC ACC-3′). Amplicons and molecular size standards were separated by electrophoresis through Tris-acetate-EDTA-buffered agarose 1% stained with ethidium bromide (Fig. 1c).

Analysis of programmed mutation of the allograft inflammatory factor gene of B. glabrata
Genomic DNA samples from the mock-transfected and pCas-BgAIFx4-transfected cells were amplified by PCR using the primers AIF-F (5′-GCA GAT TTG CAA TTC AAC ACT TA-3′) and AIF-R (5′-TGC CAG CTA GCT TAC TGC AT-3′) that flank the CRISPR/Cas9 programmed double-stranded break (DSB) site (Fig. 1a). Amplicons of 568 nt in length (from residues 489 to 1056 of the BgAIF_BGLB0055061 gene) were obtained using the AIF-F and -R primer pair. Amplicons were isolated from the agarose gel using the PCR cleanup and gel extraction kit (Takara Bio, Mountain View, CA, USA) and the nucleotide sequence of amplicons determined by Sanger direct sequencing (GENEWIZ, South Plainfield, NJ, USA). Chromatograms of the sequence reads from the control and experimental groups in each replicate experiment were subjected to online analysis using the TIDE algorithm, https ://tide.deskg en.com/ [55,56] and also using the Inference of CRISPR v2 Edits analysis (ICE) software, https ://ice.synth ego.com/#/ (Synthego Corporation, Redwood City, CA, USA) [57]. Estimates of CRISPR efficiency, insertion-deletion (INDEL)-substitution percentages, and the nucleotide sequence of mutant alleles were obtained using both the TIDE and the ICE platforms [55,56] (Fig. 2a, b).

Quantitative real time PCR analysis of transcription of BgAIF
To evaluate the differential levels of the BgAIF transcript among the control and experimental groups, total RNAs were extracted and treated with DNase I, as above. DNase I treated-RNA (200 ng) was reverse transcribed to cDNA, followed by quantitative RT-PCR, using the ViiA7 real time PCR system (Applied Biosystems, MA, USA), and the SSoAdvanced Universal SYBR Green Supermix reagents (Bio-Rad, CA, USA), according to the manufacturer's recommendations. The following nucleotide primers used BgAIF gene-specific primers amplify 119-257 nt of BgAIF GenBank accession number EX001601.1: BgAIF-rt-F (5′-CCT GCT TTT AAC CCG ACA GA-3′) and BgAIF-rt-R (5′-TGA ATG AAA GCT CCT CGT CA-3′). Differential BgAIF gene expression was calculated after normalizing with BgActin (primers as above) and comparison with the non-treated (control) cells. The ΔΔCt method was used to calculate the differential gene expression [58], with assistance of the GraphPad Prism 8 software (San Diego, CA, USA) (Fig. 2c).

Schistosome sporocysts
Miracidia of the NMRI strain of S. mansoni were hatched from eggs recovered from livers of schistosome infected mice (SRC, Biomedical Research Institute, Rockville, MD, USA) under axenic conditions [28], primary sporocysts were transformed from the miracidia in vitro, as previously described [26]. Briefly, miracidia were immobilized by chilling on ice for 25 min, followed by pelleting using centrifugation, 500×g at 4 °C, 60 s. The miracidia were washed with ice cold Chernin's balanced salt solution (28 mM NaCl, 0.5 mM Na 2 HPO 4 , 2 mM KCl, 1.8 mM MgSO 4 .7H 2 O, 0.6 mM NaHCO 3 and 3.6 mM CaCl 2 .2H 2 O) with 1 mg/ml glucose, trehalose, and antibiotic, 10 µl/ml of 100 × penicillin/streptomycin (Thermo Fisher Scientific), termed CBSS + . Approximately 5000 miracidia per well of a 24-well plate were cultured in CBSS + at 26 °C for 24 h, after which the sporocysts were washed to remove shed ciliated epidermal plates and other debris, followed by transfer to a 1.5 ml microcentrifuge tube [26].

Sporocyst-Bge cell binding assay and cell adhesion index (CAI)
To investigate the if BgAIF would affect the ability of cell adhesion to S. mansoni sporocysts, we co-cultured the non-transfected Bge cell or non-selected-, transfected-pCas-BgAIFx4 cells (BgAIF depleted-cells named 'ΔBgAIF-Bge') with in vitro transformed sporocysts, then the cell adhesion index (CAI) was calculated as previously described [26,59]. A limitation of this study was that we were not be able to select or enrich for BgAIF edited-cells, and hence the ΔBgAIF-Bge cell populations can be considered to be a population of gene mutated mixed with non-modified (wild type) cells. CAI is a semiquantitative method of cell adhesion to primary sporocysts using four categories of scores ranging from one to four-lower to higher numbers of cells adherent to the parasite's surface. In brief, we mixed single cell suspensions of 500,000 Bge cells with 200 freshly-prepared sporocysts (total volume 200 µl of CBSS + ) in sterile, siliconized tubes (Bio Plas, Thomas Scientific, Swedesboro, NJ, USA). The Bge cell-sporocyst co-culture was maintained at 26 °C for 24 h. Cellular morphology and adhesion of the cells to the surface of the sporocysts was monitored and recorded using an inverted microscope, at 20× magnification (Zeiss Axio Observer A1, Carl Zeiss LLC, White Plains, NY, USA) after gently transferring the parasite-cell suspension to a tissue culture plate (Greiner Bio-One). Scoring of the adherence index was carried out in a blinded fashion to the investigator reading the score; ≥ 50 sporocysts from each experimental group were counted each time, and triplicates of each treatment group were scored. Seven independent biological replicates of this CAI-based sporocyst-Bge cell binding assay were carried out. In total, ≥ 400 sporocysts were examined from each treatment and control group. Averages for the CAI values were calculated from the cell adhesion scores ranging from 1 to 4 (examples presented in Fig. 3a) according to the formula, CAI = total binding value per number of sporocysts [26].

Cas9 nuclease transcribed in transfected Bge cells
Total RNA was extracted from non-transfected cell (wild type; WT), mock control and pCas-BgAIFx4-transfected Bge cells to assess the expression of Cas9 (Fig. 1b). The cDNAs from either controls or pCas-BgAIFx4-transfected Bge cells were employed as the template in PCRs using two primer pairs, one specific for Cas9 and the other for BgActin, the actin gene of B. glabrata that served as the reference gene (Fig. 1b, c). Transcripts encoding Cas9 in transient pCas-BgAIFx4 transfected-Bge cells were detected at 24 h after transfection and expression was maintained for the 9 days of the assay.
The specific amplicon of Cas9 mRNA (277 bp) was observed from pCas-BgAIFx4 transfected cells, but was absent from the non-transfected cells (Fig. 1c). Our findings supported previous findings that revealed a CMV promoter driven by luciferase in Bge cells [60]. Expression of the control reference BgActin was observed at a 214 bp amplicon in both the controls and experimental samples (Fig. 1c).

Programmed mutation of BgAIF confirmed functional CRISPR/Cas9 activity in Bge cells
Genomic DNAs from wild type Bge, medium-transfected (mock) and pCas-BgAIFx4-transfected cells were used as the template for PCRs with the primer pair, AIF-F and AIF-R, flanking the programmed Cas9 cleavage site on BgAIF, exon 4 ( Fig. 1a, green arrows; amplicon size, ~ 200 nt). The red arrow indicates the predicted site of the Cas9-catalyzed double-strand break (DSB) within the BgAIF locus (Fig. 1a). The nucleotide sequence of the amplicons was determined by Sanger direct sequencing using the same primers. Both forward and reverse Sanger direct sequencing reads from the same amplicon were estimated for insertion-deletion (INDELs) by the ICE and the TIDE algorithms [55,56]. The reads from the Bge cells transfected with the pCas9-BgAIFx4 contained INDELs at or around the programmed CRISPR/ Cas9 cleavage site. The percentage of reads that included INDELs ranged from 8.9% to 17.1%, in the 12 biological replicates that were carried out (Fig. 2a, b). Notably, the mutation profile in the vicinity of the predicted DSB in BgAIF was similar among these 12 replicates, which were undertaken independently. Commonly observed INDELs at the DSBs site as revealed by the ICE analysis included deletions of 8 to 30 bp and insertions of 1 or 2 bp (Fig. 2a). These mutations were predicted to result in frameshift mutations, the consequent loss of the open reading frame, and hence and permanent knockout of BgAIF in the gene-edited Bge cell. The profile of the frequency of mutations observed in each biological replicate was used to plot the curve (Prism 8 software; https ://www.graph pad.com/) presented in Fig. 2b. These findings demonstrated that programmed genome editing using CRISPR/Cas9 was active in Bge cells, and that the non-homologous end-joining (NHEJ) pathway [61] was active in B. glabrata for the repair of programmed double-stranded breaks, leading to targeted gene knockout.

Programmed mutation interrupted expression of BgAIF
The aims of the study included the investigation of the activity of CRISPR/Cas9 gene editing in the Bge cell line and addressing the hypothesis that AIF functions in the activation of a macrophage like phenotype by the Bge cell. Accordingly, Bge cells were transfected with pCas9-BgAIFx4 plasmid DNA. The experimental approach did not include drug resistance and/or reporter gene markers in order to enrich for transfected Bge cells. However, even without enrichment for transfected cells, there was a highly statistically significant reduction in levels of BgAIF transcripts in the transfected Bge cell population. Expression of BgAIF transcripts were assessed using RNAs from the cells at 9 days post-transfection. Comparison of the experimental and control groups revealed significantly reduced levels of the BgAIF in the pCas-BgAIFx4-transfected cells, mean 49.55 ± 20.22%, range 28.1 to 86.3% (n = 12) compared to the wild type Bge (normalized sample, 100% expression), mock control cells (unpaired t-test and F-test to compare variances; F (11,11) = 294.7, P < 0.0001) (Fig. 1b). An inverse correlation between the percentage of INDELs and reduction in transcript levels was not apparent (not shown). c Bar chart to present the CAI values from control (mock-transfected) ∆BgAIF-Bge cells during co-culture with primary sporocysts at a co-culture ratio of one sporocyst to 100 Bge cells; CAI = 2.66 ± 0.10 (mean ± SD; 476 sporocysts in total scored) for the mock-transfected Bge and 2.31 ± 0.23 for the ∆BgAIF-Bge cells (424 sporocysts in total scored); P = 0.0033, unpaired Student's t-test; n = 7 biological replicates. d The cell adhesion score ranging from 1 to 4 of individual sporocysts. Cell adhesion scores 1, 2, 3 and 4 from mock cells were 2.5 ± 1.58 (SE), 29.28 ± 4.06, 20.85 ± 2.15 and 13.14 ± 2.55, respectively, whereas in the ∆BgAIF-Bge cells cell adhesion scores 1, 2, 3 and 4 were 8.86 ± 1.57, 31.42 ± 4.07, 13.28 ± 2.51 and 13.13 ± 2.55, respectively. There was a higher amount of sporocyst from the BgAIF edited-mixed cell population scored as '1′ than control cells with statistic significant by t-test (t ratio = 3.98, P = 0.001). Also, there was a lower amount of sporocysts scored as '3′ in ∆BgAIF-Bge cells compared with the control cells (t ratio = 2.40, P = 0.004)

Programmed knockout of BgAIF interfered with adherence of Bge cells to schistosome sporocysts
Single-cell suspensions of Bge cells in the mock-treated and ΔBgAIF groups were co-cultured for 24 h in siliconized tubes with primary S. mansoni sporocysts. At that point, the numbers of cells that had adhered to each sporocyst were scored. This was accomplished by examination of at least 5 discrete sites of the well of the 24-well plate with ≥ 50 sporocysts of each group. The cell adhesion index (CAI) were scored from 1 to 4, with a score of 1 indicating few or no adherent cells and a score of 4 indicating that cells or clumps of cells covered more than half the tegumental surface of the sporocyst, as defined in earlier reports [26] (Fig. 3a). Cells from the mocktransfected control group mostly adhered in clumps or singly to the surface of the parasite (representative images in the upper panels of Fig. 3b), with CAI values that ranged from 2 to 4. By contrast, fewer cells adhered to the surface of the sporocysts in the ΔBgAIF-Bge group (representative images, lower panels in Fig. 3b), with CAI values ranging from 2 to 3. Only ~ 20% of the ΔBgAIF-Bge cells adhered to the surface of the sporocysts and most of the cells retained remained spread singly on the surface of the well of tissue culture plate (Fig. 3b).
The CAI values ascertained from the seven biological replicates (> 50 parasites in each replicate) with more than 400 parasites in total were scored). The average CAI were 2.66 ± 0.10 (range: 2.53-2.78) and 2.25 ± 0.22 (range, 2.08 to 2.55) in the mock-treated and transfection control group, respectively. The average of CAI value from control group was significantly higher than the ΔBgAIF group (Fig. 3c) (unpaired t-test: t = 3.661 df = 12, P = 0.0033). More specifically, the CAI categoryspecific CAI values for mock-treated cells averaged from the cell adherence to single sporocysts, were 2.5 ± 1.58 (SE), 29.28 ± 4.06, 20.85 ± 2.15, and 13.14 ± 2.55 for categories 1, 2, 3 and 4, respectively. For the ∆BgAIF-Bge cells, the CAI values were 8.86 ± 1.57, 31.42 ± 4.07, 13.28 ± 2.51, and 13.13 ± 2.55 for categories 1, 2, 3 and 4, respectively. Although we observed CAI scores of 1 to 4 in both mock control cells and ΔBgAIF cells, nonetheless there were statistically significant higher numbers of sporocysts with the lowest adherence, scored as '1′, in the ΔBgAIF group compared with the mock control group as confirmed using a multiple t-test (t = 3.98, df = 12, P = 0.001). By contrast, there were significantly higher numbers of sporocysts scored as '3′ in the control compared with the ΔBgAIF group (t = 3.52032, df = 12, P = 0.004) (Fig. 3d). Finally, morphological changes were not apparent between the ΔBgAIF and the control group Bge cells.

Discussion
This report describes a novel use of programmed genome editing by the CRISPR/Cas9 approach in the embryonic cell line from the g astropod snail, B. glabrata, an intermediate host snail of the human blood fluke, S. mansoni. The Bge cell line is an informative tool in investigation of snail-schistosome, host-parasite interactions. A key attribute of the Bge cell is its hemocyte-like phenotype, given the central role of the snail hemocyte in innate and cellular immunity. However, even though Bge cells adhere to the schistosome, the parasite is not killed by these cells in vitro. The allograft inflammatory factor 1 (AIF) is a conserved calcium-binding protein typically expressed in phagocytic and granular leukocytes and is a marker of macrophage activation [38,41,45,[62][63][64][65]. An orthologue, termed BgAIF, is highly expressed in isolates of B. glabrata that are resistant to infection with S. mansoni and this gene may be linked to hemocyte activation [8,9]. Here, we targeted the AIF gene of B. glabrata embryonic cell line using programmed gene knockout to further interrogate its role in the intermediate host-schistosome interaction. We constructed a plasmid vector encoding the CRISPR/Cas9 nuclease and a guide RNA targeting exon 4 of BgAIF gene and the Cas9 nuclease from Streptococcus pyogenes. Bge cells were transfected with the gene-editing construct by square wave electroporation. Transcript levels of BgAIF were significantly reduced by up to 71.9% following transformation. In parallel, sequence reads of amplicons spanning the locus targeted for programmed gene knock-out revealed on-target mutation on the BgAIF gene, that had been repaired by non-homologous end joining leading to gene-inactivating insertions and deletions. In addition, the adherence of gene-edited Bge cells to sporocysts was significantly impeded in comparison to control cells, as ascertained using a semi-quantitative, cell adherence index. In our study, the % INDELs (8.9-17.1%) resulting from NHEJ after CRISPR/Cas9 gene editing on BgAIF exon 4 locus did not correlate with its transcript reduction (~ 50%) in all experimental samples. Nonetheless, alternative mechanisms could be used as the fusion of suppressors with a 'dead' Cas9 which enables gene regulation and increase the level of repression of the target gene [66].
The B. glabrata IDS comprises hemocytes and soluble proteins found in the hemolymph, among them the BgAIF [67][68][69]. The response of resistant mollusks is given by the adherence and encapsulation of sporocysts by hemocytes, leading to the parasite destruction [70].
The AIF-1 was demonstrated to be a pro-inflammatory cytokine that regulates immune-related genes of the oyster, Crassostrea ariakensis [36]. An orthologue in the leech Hirudo medicinalis promotes macrophagelike migration by a chemotactic activity, in addition to being involved in the innate immune responses as also seen in other species [41]. The adherence of the mixed populations of BgAIF gene-edited/non-edited-Bge cells to sporocysts was significantly impeded in comparison to control cells, as ascertained using a semi-quantitative cell adherence index. These cells, albeit in a low percentage, are less responsive to the S. mansoni parasite. These data suggested that, in the presence of S. mansoni, the Bge cells need to secrete BgAIF for activating the recruitment of more adherent Bge cells. Thus, the BgAIF protein appears to play a role in cell recognition, migration, and/or adhesion, and to participate in the early immune response to the parasite. The AIF gene is conserved broadly among protostomes and deuterostomes, including vertebrates, and also in prebilaterian including sponges, where it likely performs similar functional roles in macrophage activation and migration [71]. In humans, the HmAIF1 is an NF-κB pathway regulator, a pathway that comprises a family of evolutionarily conserved proteins, important to the immune system by participating in the expression of other proteins related to the immune system [72,73]. Although more studies will be required to decipher the regulation of these pathways in B. glabrata, after the pathogen invasion, the BgAIF possibly acts throughout the activation of the NF-κB pathway, leading to the recruitment of hemocytes and consequent pathogen elimination [72,74].
These findings confirmed the tractability of transfection of Bge cells by electroporation with the genomeediting construct, pCas-BgAIFx4, and that the CMV promoter drove transcription of Cas9 in this snail species. Whereas transformation by plasmid DNA of Bge cells by square wave electroporation appears to be novel, Bge cells have been transformed using DNA complexed with cationic lipid-based transfection reagents and with polyethyleneimine [23]. Nevertheless, our study has some limitations. Thus far we have yet to enrich the transfected cells from wild type cells. Future studies using a drug selectable marker can be designed to address this issue. Other approaches to deliver the CRISPR/Cas geneediting cargo can be tried including repeated inoculation with ribonuclear protein complexes [75], titration of the transfection chemicals [76], titration of electroporation parameters [77], and/or transduction by lentiviral virions encoding the gRNA and S. pyogenes Cas9 nuclease as we have demonstrated with eggs of S. mansoni [53,54]. Moreover, CIRCLE-Seq and similar approaches can be employed to investigate the off-target mutations [78].

Conclusions
Here we demonstrated CRISPR/Cas-based gene editing in a cell line from a medically important taxon of freshwater gastropods that are vectors for the transmission of schistosomiasis. We showed the functional role of a B. glabrata allograft inflammatory factor in the recognition/ attachment of S. mansoni sporocysts in vitro. The demonstration of the activity of CRISPR/Cas9 gene editing in Bge cells suggests that genome editing in the germline and somatic tissues of intact B. glabrata snails will also be functional. Whereas improvements can be anticipated in these approaches, an obvious next step will be to gene edit the intact snail B. glabrata. Transfection of germline cells within the snail using microinjection can be considered [79]. These findings, together with the first application of the CRISPR/Cas technique in the genetic edition of the snail Lymnaea stagnalis [48] are a step-change since they can favor the creation of a genetically modified Biomphalaria line to study the biology and physiology of the snail as well the schistosome-intermediate host relationship. Functional genomics using CRISPR/Cas-based genome editing in schistosomes and other trematodes responsible for major neglected tropical diseases has been reported [53,54]. The establishment of a functional genomic protocols involving programmed gene editing to address fundamental questions in this host-parasite relationship using genetically modified snails and schistosomes now seems to be feasible.