in vitro egg production by the human parasite Schistosoma mansoni

Schistosomes infect over 200 million people. The prodigious egg output of these parasites is the sole driver of pathology due to infection, yet our understanding of their sexual reproduction is limited because egg production is not sustained for more than a few days in vitro. Here, we describe culture conditions that support schistosome sexual development and sustained egg production in vitro. Female schistosomes rely on continuous pairing with male worms to fuel the maturation of their reproductive organs. Exploiting these new culture conditions we explore the process of male-stimulated female maturation and demonstrate that physical contact with a male worm, and not insemination, is sufficient to induce female development and the production of viable parthenogenetic haploid embryos. Furthermore, we show that RNAi can be used to robustly perturb the maintenance of the female reproductive system and blunt egg production in vitro. Taken together, these results provide a platform to study the fascinating sexual biology of these parasites on a molecular level, perhaps illuminating new strategies to control schistosome egg production.


ABSTRACT 17
Schistosomes infect over 200 million people. The prodigious egg output of these parasites is the 18 sole driver of pathology due to infection, yet our understanding of their sexual reproduction is 19 limited because egg production is not sustained for more than a few days in vitro. Here, we 20 describe culture conditions that support schistosome sexual development and sustained egg 21 production in vitro. Female schistosomes rely on continuous pairing with male worms to fuel the 22 maturation of their reproductive organs. Exploiting these new culture conditions we explore the 23 process of male-stimulated female maturation and demonstrate that physical contact with a male 24 worm, and not insemination, is sufficient to induce female development and the production of 25 viable parthenogenetic haploid embryos. Furthermore, we show that RNAi can be used to 26 robustly perturb the maintenance of the female reproductive system and blunt egg production in 27 vitro. Taken together, these results provide a platform to study the fascinating sexual biology of 28 these parasites on a molecular level, perhaps illuminating new strategies to control schistosome 29 egg production. Schistosomes are blood-dwelling parasitic flatworms that cause serious disease in 52 millions of people in the developing world (1). The pathology caused by these parasites is 53 entirely due to the parasite's prodigious egg output (2). Although the goal of the parasite is to 54 pass these eggs from the host to ensure the continuity of the parasite's complex life cycle, 55 approximately half of these eggs become trapped in host tissues inducing inflammation that 56 represents the primary driver of disease (3). Since parasites incapable of producing eggs 57 produced little pathology in infected hosts, understanding the biology of schistosome egg 58 production could suggest new therapeutic strategies against these devastating parasites. 59 Schistosomes are unique among flatworms as they do not sexually reproduce as 60 hermaphrodites instead they have evolved separate male and female sexes (2,4,5). This 61 transition from hermaphroditism to dioecism has led to some intriguing biological phenomena, in 62 particular the observation that female schistosomes rely on continuous pairing with a male worm 63 to become sexually mature and produce eggs (2, 6-8). Indeed, females grown in the absence of 64 male worms are developmentally stunted and their reproductive organs are undeveloped. Upon 65 pairing with a male worm the female's sexual organs become mature and egg production 66 commences. Interestingly, this process is reversible since females deprived of male contact will 67 regress to an immature state (9). Although the process of male-induced female maturation was 68 described almost a century ago little is known about the nature of the molecular signals that 69 induce female maturation upon pairing with a male worm. 70 A major bottleneck for understanding the biology of egg production and female sexual 71 development is that normal egg production ceases within days of removal of the parasite from 72 the host (10-13). While work by numerous investigators has established robust conditions for the 73 maintenance (14-17) and growth of adult-staged parasites (12,18,19), no in vitro conditions that 74 sustain continuous egg production have been described. Here, we report conditions that support 75 long-term schistosome egg production in vitro. As a proof of principle, we use these culture 76 conditions to explore the process by which male worms stimulate female maturation. We find 77 that direct contact with a male worm along the female worm's entire body is essential for sexual 78 maturation. Furthermore, we demonstrate that in the absence of sperm transfer that contact with a 79 male worm is sufficient for female worms to produce viable parthenogenetic haploid embryos. 80 These studies provide new insights in the biology of schistosome egg production and lay the 81 groundwork for the application of a growing molecular tool kit to understanding the fascinating 82 sexual biology of these important pathogens. 83 84

RESULTS AND DISCUSSION 85
Media containing ascorbic acid, red blood cells, and cholesterol supports egg production in 86 vitro. The most successful systematic effort for culturing schistosomes in vitro are those of 87 Basch (12,18,19). While Basch's "medium 169" (BM169) was able to support the in vitro 88 growth of larval parasites to adulthood (18), it was insufficient for maintaining sexually mature 89 egg-laying female parasites (12,19). Nevertheless, given the success of BM169 to support 90 parasite growth, we reasoned BM169 was the ideal place to begin optimizing conditions for egg 91 production. As previously reported (12), adult schistosomes recovered from mice and cultured in 92 BM169 progressively lost the ability to lay eggs with the morphological characteristics of those 93 laid in vivo or immediately ex vivo (Fig. 1A). 94 A schistosome egg is constructed from cells derived from two organs: the ovary, that 95 contributes an oocyte, and the vitellaria, that provide 20-30 vitellocytes (2,20). Although female 96 worms paired with male worms retained the ability to produce oocytes during in vitro culture 97 (Fig. S1A,B), we noted a rapid loss in the ability of cultured parasites to generate vitellocytes. 98 Vitellocytes contain two types of large cytoplasmic inclusions: lipid droplets and vitelline 99 droplets that coalesce to form the eggshell (21, 22). Using Fast Blue BB to label vitelline 100 droplets we found that vitellaria progressively ceased production of large numbers of vitellocytes 101 in BM169 (Fig. 1B,C). Similarly, BODIPY 493/503 labeling found that the vitellaria of females 102 in BM169 possessed few lipid droplets at D20 (Fig. 1C). Examination of a panel of genes 103 expressed in mature vitellocytes (23) by whole mount in situ hybridization (Fig. 1D) and 104 quantitative PCR (Fig. 1E) found a significant decrease in the expression of vitellaria-specific 105 transcripts during culture, similar to previous studies (13). Thus, the capacity for vitellogenesis is 106 rapidly lost in vitro. 107 To improve the rate of vitellogenesis and egg production we examined supplements that 108 could potentially satisfy either known metabolic requirements for egg production (e.g., lipids (24))  109 or documented auxotropies of the worm (e.g., polyamines, fatty acids, sterols(25-27)). From 110 these analyses we observed no qualitative effects on egg production from supplements, including: 111 albumin (e.g., Lactalbumin or Linoleic Acid-Oleic Acid-Albumin), spermidine, commercially 112 available lipid supplements, commercially available antioxidant supplements, red blood cells 113 (RBCs), low density lipoprotein, L-carnitine, N-acetyl-cysteine, or sera from various species 114 (chicken, bovine, or horse). During this process, however, we came across a report detailing the 115 formation of abnormal eggs in schistosome-infected guinea pigs fed an Vitamin C (L-ascorbic 116 acid) deficient diet (28). Strikingly, addition of ascorbic acid to BM169 led to a marked increase 117 in vitelline development (Fig. 1C, Fig. S1C) and the production of eggs morphologically similar 118 to those laid by parasites immediately ex vivo (Fig. 1C). 119 Although L-ascorbic acid had profound effects on the quality of eggs generated in vitro, 120 the rate of egg production and development of the vitellaria remained inferior to that of fresh ex 121 vivo parasites (Fig. 1C). Thus, we re-examined some of the previous assessed media 122 supplements. Given the critical role for lipid metabolism in egg production (24), we reasoned 123 that adding complex sources of lipids (and other nutrients) that the parasite encounters in vivo 124 might act synergistically with ascorbic acid. We found that supplementation with either red 125 blood cells or a commercial "cholesterol" concentrate containing purified Low Density 126 Lipoprotein (LDL) increased lipid stores along the intestine but had little effect on vitelline 127 development or the production of normal eggs (Fig. 1C). However, combination of red blood 128 cells, the cholesterol/LDL concentrate, and ascorbic acid produced a dramatic increase in the rate 129 of vitellogenesis (Fig. 1C), the rate of egg production (Fig. 1C), and a marked increase in the 130 expression of vitellaria-specific transcripts (Fig. 1D,E). From here on we refer to this formulation 131 as ABC169 (Ascorbic Acid, Blood Cells, Cholesterol and Basch Media 169). 132 In vertebrate cells, L-ascorbic acid acts not only as an antioxidant, but an essential co-133 factor for a variety of enzymes (29). Mammals deficient for L-ascorbic acid cannot perform key 134 enzymatic reactions (e.g., collagen hydroxylation) leading to symptoms commonly known as 135 scurvy (29). Interestingly, the effects of vitamin C to prevent scurvy are stereoselective as D-136 isoascorbic acid cannot replace L-ascorbic acid at equimolar concentrations (30). Similarly, we 137 observed that D-isoascorbic acid could not replace L-ascorbic acid in egg production (Fig. 1F), 138 suggesting that either L-ascorbic acid is selectively transported into cells or that it acts in a 139 stereoselective fashion to facilitate one or more enzymatic reactions within the cell. Future 140 studies aimed at determining the function of vitamin C in schistosome cells could suggest new 141 approaches to blunt egg production. 142 Parasites cultured in ABC169 produce eggs capable of developing to miracidia. After an 143 initial peak, egg production in BM169 dropped precipitously and by D7 of culture parasites laid 144 ~13 egg-like masses per day ( Fig. 2A,B). We noted a similar peak in egg production using 145 ABC169, however, after D7 these parasites sustained production of ~44 morphologically normal 146 eggs per day ( Fig. 2A). Indeed, eggs laid in ABC169 possessed a lateral spine typical of S. 147 mansoni eggs, had smooth shells, and contained a "germinal disc" corresponding to the early 148 embryo (Fig. 2B). Eggs freshly laid in ABC169 media were larger on average than those laid in 149 BM169 (Fig. 2C) and contained similar numbers of nuclei as eggs laid by parasites freshly 150 recovered from mice (Fig. 2D). Due to a high concentration of phenolic proteins that originate in 151 the vitelline droplets that form the eggshell, schistosome eggshells are highly autofluorescent 152 (20). In contrast to the egg-like masses from BM169, where parasites produce few vitelline 153 droplets ( Fig. 1C), eggs from parasites in ABC169 possessed autofluorescence comparable to 154 those from parasites freshly ex vivo (Fig. 2E). Furthermore, nearly half of the eggs laid by 155 parasites at D18-20 of culture in ABC169 contained clusters of proliferative embryonic cells 156 visualized by labeling with thymidine analog EdU (Fig. 2F, G). S. mansoni eggs are passed from 157 the host to release larvae called miracidia (2). Approximately 10-20% of eggs laid on the first 158 day cultured either in BM169 or ABC169 produced miracidia (Fig. S2). However, this rate 159 dropped during time in culture (Fig. S2). While eggs laid in BM169 after D7 were incapable of 160 liberating miracidia, about 1-2% of eggs laid in ABC169 produced viable and morphologically 161 normal miracidia (Fig. 2H, Movie S1-2). Given the disparity between the rate of entering 162 embryonic development (Fig. 2G) and the capacity for these embryos to mature to miracidia ( Physical contact with a male worm is sufficient for female worms to produce viable 179 parthenogenetic haploid embryos. Several theories have been put forward to explain the 180 mechanism by which male worms stimulate female development (2, 33-35). However, the 181 experiments supporting (or refuting) these hypotheses were conducted using sub-optimal culture 182 conditions (34) and in many cases have not been subject to extensive reproduction in the modern 183 literature. Thus, we were compelled to revisit key observations using ABC169. The prevailing 184 thought is that female development requires direct contact with a male worm (7, 34, 35). The 185 most intriguing study supporting this hypothesis are those of Popiel and Basch (32). These 186 authors observed that small segments of male worms could stimulate vitelline development. 187 Interestingly, vitelline development was confined to regions in direct contact with the male 188 segment (32). Consistent with these observations, we found a large fraction of small posterior 189 fragments could pair with immature females (Fig. 4A,B, Movie S3). These posterior segments 190 often paired with posterior regions of female worms, and consistent with observations of Popiel  191 and Basch, vitelline maturation occurred only in regions in direct contact the male segment (Fig. 192 4B). Thus, pairing with a male induces a signal that induces localized female vitelline maturation. 193 While culturing male posterior segments with female worms we observed that these 194 female worms laid morphologically normal eggs (Fig. 4C). However, examination of these eggs 195 found that they contained no embryos capable of incorporating EdU (Fig. 4C). To explore this 196 observation in more detail we examined the ovaries of the female worms paired with male 197 segments and found no evidence of mature oocyte production ( Fig. 4D). Since the schistosome 198 ovary is located anterior to the vitellaria, and this region was not in contact with male posterior 199 segments, we reasoned that ovaries, like the vitellaria, might also require local contact with a 200 male worm to mature and begin oocyte production. To test this model, we amputated males 201 behind the testes (Fig. 4E) and paired the decapitated and castrated fragments with immature 202 female worms. Since these large posterior fragments typically ensheathed the entire female 203 worm ( Fig. 4E), we reasoned that this pairing might be sufficient to stimulate oogenesis. 204 Consistent with this model, we found that ovaries of females paired with decapitated males 205 produced oocytes (Fig. 4F). Furthermore, these parasites had fully developed vitellaria along 206 their entire length ( Fig. 4G) and laid morphologically normal eggs (Fig. 4H). 207 Despite the fact that females paired with castrated male segments had no chance of being 208 inseminated, we observed that eggs laid by these parasites possessed the ability to initiate 209 embryogenesis ( Fig. 4H) and could even give rise to miracidia (0.23%, n=4704 eggs). Previous 210 studies have suggested that female schistosomes can produce parthenogenetic offspring 211 containing a haploid set of maternal chromosomes when mated with males of distantly related 212 schistosome species (36, 37). From these experiments it is not clear if simply coming into 213 contact with a male schistosome of another species is sufficient to induce the production of 214 parthenogenetic offspring or if parthenogenesis occurs only following the transfer of sperm (36, 215 37). Therefore, we examined the karyotypes of eggs laid by females paired with castrated males. 216 Unlike the diploid karyotypes of embryos from eggs laid by freshly ex vivo parasites (2n=16), 217 mitotic cells from eggs laid by females paired with castrated males were haploid containing only 218 8 chromosomes (Fig. 4I, Fig. S3). These results suggest that only contact with a male worm 219 along the entire length of the female body, and not insemination, is sufficient for the production 220 of viable haploid embryos. These data highlight the value of ABC169 media to enhance our 221 understanding of schistosome reproductive biology. 222 RNA interference can be used to study schistosome reproduction in vitro. Given the rapid 223 rate at which the vitellaria degenerate in vitro, the use of RNA interference to study sexual 224 reproduction is restricted to the first few days of culture (24,38). This approach is not ideal since 225 the functional consequences of certain RNAi perturbations (39) may not manifest in the narrow 226 window before control parasites cease normal egg production. Therefore, we examined the 227 efficacy of RNAi using ABC169. The production of both oocytes and vitellocytes is dependent 228 on dedicated populations of stem cells: oogonia in the ovaries (40) and S1 cells in the vitellaria 229 (21). Despite the fact that oocyte and vitellocyte production depends on these stem cells there 230 are no direct studies of the function of either of these stem cell populations. Therefore, we 231 examined the effect of RNAi against Histone H2B that is expressed in both S1 cells (23) and  232 oogonia (41) and was previously reported to be essential for proliferative somatic stem cell 233 maintenance due presumably to its role in cell cycle progression (42). For these studies we 234 treated parasites with dsRNA and monitored egg production and the status of the reproductive 235 organs after 14 days of culture in ABC169. Consistent with studies in male worms, we observed 236 a complete ablation of somatic stem cells (54/54 female parasites). Similarly, we observed a 237 massive depletion of the EdU incorporation in the stem cell compartments of the vitellaria (50/54 238 parasites) and ovaries (48/54 parasites) (Fig. 5A). As anticipated, stem cell depletion led to a 239 loss of mature vitellocytes (Fig. 5B) and a substantial regression in the size of the ovary (Fig.  240 5A). Consistent with the effects of Histone H2B RNAi treatment on the reproductive organs we 241 noted a significant decline in the rate of egg production in Histone H2B (RNAi) worm pairs (Fig. 242 5C). Given these observations, we suggest the combination of ABC169 with approaches such as 243 RNAi will provide a platform to dissect schistosome reproductive biology on a molecular level. 244

Conclusions 245
During his extensive studies to develop in vitro culture conditions for S. mansoni, Paul 246 Basch lamented that the production of viable eggs in vitro "remains a formidable challenge" (19). 247 We find three additions to Basch's base media formulation (ascorbic acid, blood cells, and LDL) 248 can sustain the production of viable eggs for several weeks in vitro. infections were obtained by infecting mice with male or female cercariae recovered from NMRI 266 Biomphalaria glabrata snails infected with single miracidia. Following recovery from the host, 267 worms were rinsed several times in DMEM + 5% serum before placing into culture media. The 268 base medium for these studies was BM169 (18)  Antimycotic (Gibco/Life Technologies, Carlsbad, CA) for 7 days at 37 o C in 5% CO 2 . After 313 culture, eggs were collected on 10 µm cell strainer and rinsed with artificial pond water 1-2 times 314 and placed in artificial pond water under light. A small aliquot of eggs was counted to determine 315 the total number of eggs examined. Active miracidia were observed and counted under light 316 microscopy every 30 mins for up to 4hrs. Miracidia fixation and labeling was performed as 317 previously described (47). 318

Karyotype analysis 319
Karyotypes from schistosome embryos were determined using a modified version of a previously 320 published method (48). Briefly, eggs were incubated in M199 for two days after deposition and 321 then incubated in 5 µM Nocodazole for 1 hour at 37 o C. Eggs were pelleted at 20,000 g for 1 min, 322 rinsed in DI water, pelleted, resuspended in 1ml of water, and incubated for 20 min at RT. 323 Following centrifugation, the pelleted eggs were resuspended in 1 ml of fixative (3:1 324 methanol:acetic acid) for 15 minutes. Eggs were again pelleted and resuspended in 100 µl of 325 fixative and the eggshells were disrupted using a Kontes pellet pestle (DWK Life Sciences, 326 Rockwood, TN). The disrupted eggshells were allowed to settle for 10 min, the supernatant was 327 collected, and centrifuged for 8 min at 240 g. All but ~20 µl of supernatant was removed and the 328 remaining liquid was pipetted dropwise on to a glass slide (Superfrost Plus, ThermoFisher, 329 Carlsbad, CA) freshly dipped in PBS. After air-drying chromosomes were labeled with 330 Vectashield containing DAPI (Vector Laboratories, Burlingame, CA) and imaged on a Nikon 331 A1+ laser scanning confocal microscope with a 60x/1.4 NA objective. 332 333 RNA Interference 334 Double stranded RNA was generated as previously described (41,49). For dsRNA treatment 335 worm pairs were electroporated with a square-wave pulse (125 V for 20 ms (50)) on a BTX 336 Gemini in 4 mm cuvettes with 160 µg of dsRNA in 400 µl of DMEM. The worms were then 337 supplemented with 60 µg/ml of dsRNA on day 0, 1, 2, 6 and 10. At D14, parasites were pulsed 338 for 4 hours with EdU and processed as previously described (41)

D1 Fresh egg
Movie S1. Miracidium inside egg laid on D20 of culture in ABC169.
Movie S2. Miracidia after hatching from eggs laid on D20 of culture in ABC169.
Movie S3. Previously immature female paired with a male segment in ABC169.
Additional Dataset S1 (separate file). Oligonucleotide sequences for cDNA cloning and qRT-PCR analyses.