Somatic regulation of female germ cell regeneration and development in planarians

SUMMARY

Female germ cells develop into oocytes, with the capacity for totipotency. In most animals, these remarkable cells are specified during development and cannot be regenerated. By contrast, planarians, known for their regenerative prowess, can regenerate germ cells. To uncover mechanisms required for female germ cell development and regeneration, we generated gonad-specific transcriptomes and identified genes whose expression defines progressive stages of female germ cell development. Strikingly, early female germ cells share molecular signatures with the pluripotent stem cells driving planarian regeneration. We observe spatial heterogeneity within somatic ovarian cells and find that a regionally enriched foxL homolog is required for oocyte differentiation, but not specification, suggestive of functionally distinct somatic compartments. Unexpectedly, a neurotransmitter-biosynthetic enzyme, aromatic L-amino acid decarboxylase (AADC), is also expressed in somatic gonadal cells, and plays opposing roles in female and male germ cell development. Thus, somatic gonadal cells deploy conserved factors to regulate germ cell development and regeneration in planarians.

Graphical abstract

In brief

Unlike most animals, planarians can regenerate germ cells. Here, Khan and Newmark characterize gene expression in the planarian ovary, identifying genes expressed at progressive stages of female germ cell development and in somatic ovarian cells. Functional characterization revealed somatically expressed genes required for specification or differentiation of female germ cells.

INTRODUCTION

Germ cells serve as a link between generations, producing gametes that propagate the genetic material from parent to offspring. During their development, germ cells undergo dramatic, sex-specific differentiation to generate highly specialized cell types, the sperm and egg, which ultimately yield a totipotent zygote. As the female germline differentiates into oocytes, it acquires the capacity for totipotency (Reik and Surani, 2015; Seydoux and Braun, 2006). This capacity is exemplified by somatic cell nuclear transfer, in which oocyte cytoplasm reprograms differentiated cell nuclei to produce viable clones (Campbell et al., 1996; Gurdon, 1962; Wakayama et al., 1998). Parthenogenesis, in which an unfertilized egg generates an entire new organism (Simon et al., 2003), provides another striking example of the oocyte’s capacity for totipotency. Given the oocyte’s critical roles in embryonic development, understanding the mechanisms underlying female germ cell development has enormous significance.

Somatic support cells within the gonads play critical roles in regulating germ cell development across the animal kingdom. Soma-germline communication is necessary throughout a germ cell’s life, from regulating fate choices and survival, to proliferation and differentiation (Kiger et al., 2000; Kimble and White, 1981; Korta and Hubbard, 2010; Li and Albertini, 2013; Murray et al., 2010). The in vitro derivation of functional oocytes and spermatid-like cells from pluripotent stem cells requires co-culture with somatic gonadal cells (Hikabe et al., 2016; Zhou et al., 2016), further emphasizing the importance of soma-germ cell interactions in facilitating proper germ cell development.

Most animals, including model organisms typically used to study development, set aside germ cells early during embryonic development and cannot replace them if lost (Nieuwkoop and Sutasurya, 1979, 1981). By contrast, freshwater planarians demonstrate the striking ability to regenerate both male and female germ cells from tissues lacking any reproductive structures (Morgan, 1901; Sato et al., 2006; Wang et al., 2007). Planarians are best known for their remarkable whole-body regeneration, driven by stem cells called neoblasts (Baguñà et al., 1989), a subset of which are pluripotent (Wagner et al., 2011; Zeng et al., 2018). Planarian germ cells are specified inductively and are derived from neoblasts (Baguñà et al., 1989; Newmark et al., 2008; Sato et al., 2006; Wang et al., 2007). The vast majority of research on planarian regeneration focuses on asexual strains that reproduce by fission; however, planarians can also reproduce sexually as simultaneous hermaphrodites. Some planarians can even reproduce parthenogenetically: sperm serve to activate development, without contributing genetic information (a process known as gynogenesis), demonstrating the potency of the female germline (Benazzi and Benazzi, 1976). Germ cell development in sexual planarians is responsive to physiological cues: following prolonged starvation, gonads regress and accessory reproductive organs are resorbed (Berninger, 1911; Morgan, 1901; Schultz, 1904); when feeding is resumed, gonads again produce gametes and accessory reproductive organs are re-formed. Planarians’ remarkable developmental plasticity provides a unique opportunity to investigate mechanisms regulating sex-specific germ cell specification and differentiation in the context of an adult organism.

In sexual planarians, ovarian tissue is scarce relative to testes; as such, previous transcriptomic analyses predominantly identified genes with enriched expression in the testes rather than ovaries (Rouhana et al., 2017; Wang et al., 2010; Zayas et al., 2005). Whole-animal single-cell sequencing approaches produced similar results (Fincher et al., 2018). Thus, little is known about the gene expression changes driving female germ cell development in planarians. An orphan G-protein-coupled receptor (GPCR), Ophis, is expressed in somatic gonadal cells of ovaries and testes and is required for differentiation of both oocytes and sperm (Saberi et al., 2016); however, gene expression differences distinguishing female from male somatic gonadal cells have yet to be identified, and the roles of somatic ovarian cells in various stages of female germ cell development remain to be determined. Here, to circumvent the limited quantity of ovarian tissue and identify genes with enriched expression in planarian ovaries, we generated gonad-specific transcriptomes using laser-capture microdissection followed by RNA sequencing (RNA-seq). Analysis and validation of these transcriptomic data enabled us to characterize the developmental progression of female germ cells, define somatic ovarian cell types, and identify somatically expressed genes required for proper regeneration and development of female germ cells.

RESULTS

LCM sequencing identifies transcripts that mark and distinguish ovaries from testes

The reproductive system of the sexual strain of the planarian Schmidtea mediterranea consists of two ovaries located ventrally under the posterior lobes of the brain, numerous dorsolaterally distributed testes, and accessory reproductive organs (yolk glands [vitellaria], oviducts, sperm ducts, copulatory apparatus, and gonopore) (Figure 1A) (Hyman, 1951; Issigonis and Newmark, 2019). To precisely excise ovarian, testis, and surrounding non-gonadal tissues for transcriptomic comparisons, we used laser-capture microdissection (LCM; Figure 1B) (Emmert-Buck et al., 1996; Espina et al., 2006; Forsthoefel et al., 2020). RNA-seq analysis (Mortazavi et al., 2008; Nagalakshmi et al., 2008) of the laser-captured tissues identified 36,036 non-redundant transcripts, of which 7,577 (21%) were upregulated (≥2-fold, false discovery rate p value ≤0.01) in either or both gonads: 1,491 (4%) in ovaries; 4,880 (14%) in testes; and 1,206 (3%) in both (Figures 1B, 1C, and S1A–S1D; Table S1). To validate the LCM-generated gonadal transcriptomes, we examined genes with previously reported gonadal expression (Wang et al., 2010). As expected, we found significant upregulation of the majority of these genes (82 out of 89) in the testis transcriptome (Figure S2A), including nanos, a conserved regulator of germ cell development (Handberg-Thorsager and Saló, 2007; Sato et al., 2006; Wang et al., 2007), which was also upregulated in the ovary transcriptome (Figure S2A).

Figure 1. Ovary transcriptome reveals progressive stages of female germ cell development.

(A) Reproductive system of the sexual strain of S. mediterranea.

(B) LCM-RNA-seq approach to generate gonadal transcriptomes.

(C) Hierarchical clustering of genes with significantly enriched expression in ovaries, testes, or both.

(D) Representative whole-mount RNA in situ hybridizations (WISH) to detect candidate ovary-enriched transcripts.

(E) Double fluorescent RNA in situ hybridization (FISH) to detect klf4 and lecg reveals a field of klf4⁺ cells anterior to the ovary.

(F) FISH to detect lecg and ubp8, revealing expression in oocytes. Smaller, immature oocytes (white arrows) are observed at the periphery of the ovary; larger, mature oocytes (yellow arrows) reside internally, proximal to the tuba (fertilization duct; asterisks).

(G) FISH to detect gwîn and klf4 or lecg shows gwîn expression in klf4⁺ cells anterior to the ovary (i) and at the margin of the ovary (ii). Co-expression of gwîn with lecg is observed in oocytes (iii). klf4⁺gwîn⁺ cells are presumptive female germ cell progenitors (FGPs) that differentiate into gwîn⁺lecg⁺ oocytes within the ovary. Nuclei are labeled with DAPI. Scale bars: (D) 500 μm, insets 200 μm; (E) 100 μm; (F) 50 μm; (G) 100 μm, insets 10 μm. See also Figures S1 and S2.

To validate candidates and identify new genes expressed in the ovary, we performed whole-mount colorimetric RNA in situ hybridization (WISH) for ovary-enriched transcripts and detected diverse expression patterns in and around the ovaries (Figures 1D and S2B). Approximately 90% (204 out of 227) of genes tested showed enriched expression in the ovary or associated reproductive structures. These results demonstrate the successful application of LCM-RNA-seq for generating gonad-specific transcriptomes, providing a resource for identifying regulators of germ cell development and regeneration.

Female germ cell progenitors are specified outside the ovaries and express markers of pluripotency

Histological and ultrastructural studies have shown that planarian oocytes grow and mature in a progression from the margins of the ovary toward the tuba (fertilization duct) (Gremigni and Nigro, 1983; Harrath et al., 2011). Markers of the earliest stages of germ cell development (nanos and krüppel-like factor 4 [klf4]) label oogonia at the ovary periphery (Issigonis et al., 2021; Wang et al., 2007); nanos and klf4 are also expressed in two fields of cells, anterior to the ovary at the ventro-medial portion of each brain lobe (Issigonis et al., 2021). To help elucidate the origins and differentiation of the female germ cell lineage, we sought to identify ovary-enriched transcripts expressed at various stages of oogenesis. Fluorescent RNA in situ hybridization (FISH) analysis detected galactose-binding lectin (lecg) expression in differentiating cells throughout the ovary, but not in the anterior field of klf4-expressing cells (Figure 1E). Expression of ubiquitin carboxyl-terminal hydrolase 8 (ubp8) was detected in larger oocytes residing within the ovary, proximal to the tuba, but not within smaller lecg⁺ cells at the ovary periphery (Figure 1F). Thus, ubp8 is upregulated later in oocyte differentiation than lecg (Figure 1F). The expression patterns of these oocyte markers and the corresponding growth of oocytes reveal progressive stages of female germ cell differentiation, from the periphery of the ovary toward its interior.

Like the previously described patterns of nanos and klf4 (Issigonis et al., 2021), we detected expression of a gene originally annotated as polyketide synthase-1 (pks1) (Zeng et al., 2018) in fields of cells anterior to the ovary; its expression was also detected throughout the ovary (Figures 1D and 1G). This gene was identified via single-cell sequencing as a marker of a pluripotent subpopulation of neoblasts in asexual S. mediterranea (Zeng et al., 2018); because its predicted product shares no conserved domains with any reported proteins, including polyketide synthases, and based upon its expression dynamics (described below), we propose renaming it germline-expressed, wound-induced in neoblasts (gwîn; Sindarin for “youth,” corresponding to its expression beginning in early stages of female germ cell development). gwîn is expressed in the klf4⁺ fields anterior to (Figures 1G and 1Gi) and around the margins of the ovaries (Figures 1G and 1Gii). The gwîin⁺ cells in the anterior fields and around the margins of the ovary were labeled by anti-phospho-Histone H3 (pHH3) antibodies (Figure S3Ai and ii), indicating that they are proliferative. gwîn is also expressed in lecg⁺ oocytes within the ovary (Figures 1G and 1Giii), and, thus, serves as a marker spanning all stages of female germ cell development. These results suggest that the nanos⁺ klf4⁺ gwîn⁺ cells located at the margin of the ovary are oogonia that give rise to gwîn⁺lecg⁺ oocytes within the ovary (Figure 1G). Finding proliferating cells expressing the earliest markers of germ cell development (nanos⁺ klf4⁺gwîn⁺) anterior to the ovaries suggests that female germ cell progenitors (FGPs) can also be specified outside of the ovaries (Issigonis et al., 2021). Identifying markers for progressive stages of female germ cell development allows us to study early steps in female germ cell specification and differentiation.

A marker of pluripotent neoblasts is expressed in germ cells and only upregulated in neoblasts after wounding

In addition to gwîn, single-cell analysis of a pluripotent neoblast subpopulation identified other cluster-defining transcripts, including tspan group-specific gene-1 (tgs-1) and tetraspanin-1 (tspan-1) (Zeng et al., 2018). The expression of gwîn in FGPs of sexual planarians suggested affinities between pluripotent neoblasts and early germ cells; thus, we sought to further explore this relationship. First, we examined the expression of the neoblast marker piwi-1 (Reddien et al., 2005), a planarian PIWI homolog that is also expressed in germ cells (Davies et al., 2017). We detected relatively high levels of piwi-1 mRNA in gwîn⁺ FGPs (Figures 2A and 2Ai and ii). Consistent with previous studies (Davies et al., 2017), we detected piwi-1 expression in oocytes, suggesting that piwi-1 expression is maintained through oocyte differentiation (Figures 2A and 2Aiii). Similar to gwîn, tgs-1 is enriched in the ovarian transcriptome and expressed abundantly in FGPs (Figures 2B and S3C); however, tgs-1 was also detected in a subset of neoblasts (Figure S3D). In the sexual strain, gwîn expression was most pronounced in germ cells in both ovaries and testes (Figures 1G, 2A, 2B, and S3B) and, aside from co-expression in the germ cells, its expression did not overlap with that of piwi-1.

Figure 2. FGPs express markers associated with a pluripotent neoblast subpopulation.

(A) FISH to detect gwîn and the neoblast marker, piwi-1. Expression of piwi-1 is maintained from neoblasts (i) into FGPs (ii) and oocytes (iii).

(B) FISH to detect tgs-1 with klf4 and gwîn. tgs-1 marks a pluripotent subset of neoblasts and is enriched in FGPs.

(C and D) WISH to detect gwîn sexuals (C) and asexuals (D) during homeostasis (uncut) and 72 h post-amputation (hpa) or-wounding (hpw). gwîn is upregulated at wound sites in both sexuals and asexuals.

(E and F) FISH to detect gwîn and piwi-1 in asexuals during homeostasis (E) and 72 hpa (F), shows upregulation of gwîn in neoblasts at wound sites after amputation. Scale bars: (A and B) 100 μm, insets 10 μm; (C and D) 500 μm; (E and F) 10 μm. See also Figures S3 and S4.

To validate the expression of neoblast markers in germ cells, we first assessed gwîn expression in asexual S. mediterranea. Although these animals reproduce by fission, they develop gonad primordia containing germ cells that fail to differentiate (Handberg-Thorsager and Saló, 2007; Wang et al., 2007). In small asexuals (worm length: 2.3 ± 0.2mm) we did not detect gwîn expression anywhere in the animal; larger asexuals (4.2 ± 0.5mm), however, displayed a small group of gwîn⁺ cells ventral to the posterior brain lobes (Figures 2D and 2E), reminiscent of FGPs in the sexual strain and consistent with the increase of nanos⁺ klf4⁺ female germ cells in this region of larger asexual planarians (Issigonis et al., 2021). We did not detect gwîn expression in testis primordia of asexuals, suggesting that gwîn is upregulated during a later stage of male germ cell development (Figure 2D).

Considering the scarcity of FGPs in asexuals, it seems unlikely that these cells were captured during single-cell sequencing by Zeng et al.; rather, gwîn expression in pluripotent neoblasts could result from cell dissociation and the subsequent induction of wound-response transcriptional programs (Lacar et al., 2016; Wu et al., 2017). Indeed, wound responsiveness was one of the criteria used to define this neoblast subcluster (Zeng et al., 2018). To test this idea, we examined gwîn expression by WISH after amputating or injuring the worms and found that gwîn mRNA was upregulated at wound sites in both sexual and asexual strains (Figures 2C–2F). This dramatic upregulation at wound sites may account for the expression of gwîn in additional clusters beyond that expressing tgs-1, as well as its broad expression in neoblast colonies observed after partial irradiation (Zeng et al., 2018). We corroborated our ISH analysis by examining previous regeneration transcriptomes (Kao et al., 2013; Rozanski et al., 2019), which revealed that gwîn is upregulated as early as 4 to 6 h after injury, a time frame consistent with the possibility of cell-dissociation-induced upregulation in neoblasts (Figure S4A). In contrast, other early germ cell markers, such as klf4 and nanos, are not upregulated at wound sites (Figures S4B and S4C). Next, we confirmed that gwîn was upregulated in neoblasts after amputation by performing double FISH with the neoblast marker piwi-1. In uninjured asexual planarians, gwîin⁺piwi-1⁺ cells were only detected in the FGP region (Figure 2E). However, after amputation, gwîn⁺piwi-1⁺ cells were detected at both anterior and posterior wound sites (Figure 2F). These data indicate that, in intact planarians, gwîn marks germ cells rather than neoblasts. Wound-inducible expression of this gene in neoblasts provides another link between germ cells and the planarian’s pluripotent stem cells.

Gene-expression profiling reveals spatial heterogeneity in ovarian somatic cells

How are FGPs maintained and what regulates their differentiation within the ovary? Somatic gonadal cells play critical roles in regulating germ cell development throughout the animal kingdom. To date, studies of soma-germline interactions in the planarian ovary have been limited by the availability of appropriate markers. The first such marker, ophis, encodes an orphan GPCR expressed in somatic support cells of ovaries and testes and is required for both female and male germ cell differentiation (Saberi et al., 2016). Our ISH screen of ovary-enriched transcripts identified three genes with expression patterns distinct from those observed for germ-cell-enriched transcripts, and resembling a scaffold surrounding the germ cells: delta3; the forkhead-family transcription factor, foxL; and endothelin converting enzyme 1 (ece1) (Figures 1C, 1D, and S5A). We confirmed that delta3, foxL, and ece1 transcripts were not detected in female germ cells using double FISH with pooled probes for the germ cell markers klf4 and lecg (Figures 3A, 3B, and 3C). Additionally, unlike ophis (Saberi et al., 2016), delta3, foxL, and ece1 transcripts were not detectable in the testes, indicating sexually dimorphic expression (Figure S5A). delta3⁺ cells are closely associated with oocytes, and co-express ophis within the ovary, confirming delta3 as a marker of ovarian somatic cells (Figure 3D). Double FISH to detect foxL and delta3 revealed higher expression of foxL in delta3⁺ somatic cells situated proximal to the tuba, compared with delta3⁺ cells at the periphery (Figure 3E). In contrast, ece1 is expressed in the peripheral delta3⁺ cells (Figure 3F). Double FISH to detect both ece1 and foxL confirmed the presence of ece1^highfoxL^low cells at the periphery of the ovary (Figures 3G, 3Gi, and 3H) and ece1^low foxL^high cells proximal to the tuba (Figures 3G, 3Gii, and 3H), indicating heterogeneity among ovarian somatic support cells. This heterogeneity could reflect distinct functional roles at different stages of oogenesis.

Figure 3. Characterization of somatic ovarian cells.

(A–C) FISH to detect delta3 (A), foxL (B), ece1 (C), and pooled probes for klf4 (FGPs) and lecg (oocytes) shows expression of these transcripts in non-germ-cell populations of the ovary.

(D) ophis⁺ ovarian somatic cells co-express delta3.

(E and F) delta3⁺ somatic support cells co-express foxL (E) and ece1 (F).

(G) FISH to detect ece1 and foxL reveals heterogeneity in gene expression within somatic ovarian cells: (i) ece1^low foxL^high cells are located proximal to the tuba and (ii) ece1^high foxL^low cells are located distal to the tuba. Arrowheads indicate ectopic ovaries often found in large sexual worms.

(H) Schematic depicting marker gene expression in somatic ovarian cells and associated reproductive structures (tuba and oviduct). Asterisks indicate tuba. Scale bars: 50 μm. See also Figure S5.

Ovarian somatic cells regulate female germ cell development via conserved factors

foxL encodes a planarian homolog of the forkhead-family transcription factor FoxL2 (Figure S5B), which is a critical regulator of granulosa cell differentiation in the mammalian ovary. Ovaries fail to develop properly in FoxL2 knockout mice: the absence of functional granulosa cells in these mutants leads to oocyte atresia and infertility due to progressive follicular depletion (Schmidt et al., 2004; Uda et al., 2004). Human mutations in FoxL2 can lead to premature ovarian failure or ovarian tumors (Crisponi et al., 2001; Georges et al., 2014; Schmidt et al., 2004; Shah et al., 2009; Uda et al., 2004; Uhlenhaut and Treier, 2011). FoxL2 is also required to actively maintain female fate in the adult mouse ovary (Uhlenhaut et al., 2009). Although FoxL2 genes have been identified in a range of invertebrates, its expression in ovarian somatic cells, but not germ cells, has been proposed to be a vertebrate innovation (Bertho et al., 2016). Furthermore, functional roles for FoxL2 homologs in ovarian development have yet to be demonstrated in any invertebrate.

Thus, it was noteworthy to find planarian foxL similarly expressed in ovarian somatic cells, enriched in the tuba-proximal population surrounding late-stage oocytes (Figures 3B, 3E, 3G, and S5A). To examine whether foxL also plays a functional role in oogenesis, we used RNAi to knock down its expression in the context of decapitated worms that will regenerate new ovaries (Figure 4A). foxL(RNAi) worms regenerated their heads normally and displayed normal feeding behavior (Figure S5C). The distribution and numbers of klf4⁺ FGPs in the anterior fields were not affected significantly in these animals (Figures 4B and 4C). However, foxL RNAi resulted in a dramatic and significant reduction in the number of lecg⁺ oocytes (Figures 4B and 4C). These data suggest that foxL is required cell non-autonomously for oocyte differentiation and maintenance in planarians.

Figure 4. Ovarian somatic cells are essential for female germ cell development.

(A) Regeneration assay for RNAi-mediated gene knockdown: sexually mature planarians were fed double-stranded RNA (dsRNA) corresponding to the target gene four times over 2 weeks. Then, the heads were amputated posterior to the ovaries and discarded; after head regeneration in tail fragments, eight dsRNA feedings were performed for 4 weeks before fixing the animals for further analysis.

(B) FISH to detect klf4 and lecg in control(RNAi) and foxL(RNAi) worms.

(C) Quantification of klf4⁺ FGPs and lecg⁺ oocytes in control(RNAi) and foxL(RNAi) worms (mean with 95% confidence interval; ns, not significant).

(D) FISH to detect klf4 and lecg in control(RNAi), delta3(RNAi), notch2(RNAi), and notch4(RNAi) worms.

(E) Quantification of klf4⁺ FGPs and lecg⁺ oocytes in control(RNAi), delta3(RNAi), notch2(RNAi), and notch4(RNAi) worms (mean with 95% confidence interval; ns, not significant).

(F) Density plot of klf4⁺ FGPs in control(RNAi), delta3(RNAi), notch2(RNAi), and notch4(RNAi) worms. The brain lobe and tuba were used to define the relative position of cells.

(G) Double FISH for notch2 or notch4 with the somatic marker delta3. Asterisks indicate tuba. Scale bars: 100 μm. See also Figure S5.

As reported above, ovarian somatic cells also express a delta3 homolog (Figures 3A and 3B), which encodes a transmembrane Notch-signaling ligand (Figure S5D) (Bray, 2006). Notch signaling plays critical roles in various aspects of soma-germ cell interactions across species. For example, the Caenorhabditis elegans niche uses Notch signaling to regulate germline stem cell maintenance and proliferation (Byrd and Kimble, 2009), whereas, in the Drosophila ovary, Notch signaling controls the formation and maintenance of the niche (Xie et al., 2008). To explore the role of Notch signaling in planarian oogenesis, we performed delta3 RNAi experiments in the context of the ovarian regeneration paradigm (Figure 4A). delta3(RNAi) planarians displayed a significant increase of klf4⁺ FGPs, the distribution of which was skewed toward the midline (Figures 4D, 4E, and 4F). Additionally, the ovaries of delta3(RNAi) planarians appeared disorganized, with klf4⁺ FGPs intermingled with lecg⁺ oocytes; oocyte numbers were significantly reduced (Figures 4D and 4E).

Because direct cell-cell interaction is necessary for Notch signaling (Bray, 2006), we asked whether Notch receptors were also expressed in the ovary. We found enriched expression of transcripts encoding both planarian Notch receptors in the ovarian transcriptome (Figures S5D and S5E). FISH revealed mRNAs of both notch2 and notch4 in the tuba and oviduct; notch4 mRNA was also detected in the delta3⁺ ovarian somatic cells (Figure 4G). Knockdown of notch2 resulted in a significant expansion, midline-skewed distribution of klf4⁺ FGPs, and disorganized ovaries with depleted oocytes, similar to delta3(RNAi) worms. By contrast, klf4⁺ FGPs appeared unaffected after notch4 RNAi, but oocyte numbers were reduced (Figures 4D, 4E, and 4F). Although notch2(RNAi) worms regenerated normally, they were smaller at the end of our assay, suggesting roles in growth or feeding, which could also influence oocyte development (Figure S5F). Altogether, Notch signaling plays several roles in female germ cell development in planarians, regulating the number and distribution of FGPs, as well as the spatial organization of the ovary, with effects upon oocyte production. The expression patterns of the Notch-signaling components reported here suggest that interactions between accessory reproductive organs (oviduct, tuba) and ovarian somatic cells may also help establish proper structure of the ovary.

A female-specific regulator of germ cell regeneration: bidirectional soma-germline communication

In addition to markers of ovarian somatic cells, the ovary transcriptome enabled us to identify zfs1, a female-specific early germ cell marker. This gene encodes a predicted RNA-binding protein (Figure S6A) and is expressed in ovaries as well as in cells anterior to the ovaries (Figures 5A and 5B). Double FISH with gwîn indicated that zfs1 is co-expressed in FGPs and oocytes (Figure 5C); however, unlike other FGP markers (klf4 and nanos) that are also expressed in early stages of male germ cell development, zfs1 expression was restricted to female germ cells (Figures 5A and 5B). RNAi knockdown of zfs1 using the previously described ovary-regeneration paradigm (Figure 4A) resulted in a dramatic reduction in the number of klf4⁺ FGPs and ablation of lecg⁺ oocytes (Figures 5D and 5E). By contrast, male germ cell development was unaffected by zfs1 RNAi: klf4⁺ early male germ cells at the testis periphery, and DAPI-stained sperm in the lumen (Issigonis et al., 2021) appeared normal (Figure 5F). Accessory reproductive structures and egg laying were also unaffected. However, zfs1(RNAi) animals were sterile: they failed to produce hatchlings (Figures S6B–S6D). These results indicate a germ-cell-intrinsic, female-specific role for zfs1, which belongs to a small, yet ancient, family of RNA-binding proteins that regulate mRNA turnover (Cuthbertson et al., 2008; Wells et al., 2017). The yeast ortholog is a regulator of sexual differentiation (Kanoh et al., 1995; Navarro et al., 2017), and the mammalian homolog, ZFP36L2, is critical for oocyte maturation, targeting transcriptional regulators and chromatin modifiers for degradation (Dumdie et al.,2018). Understanding how zfs1 expression is restricted to female germ cells will provide insight into the modulation of germ cell sex in a simultaneous hermaphrodite.

Figure 5. Female-specific role of zfs1 in germ cell regeneration.

(A) zfs1 expression is enriched in the ovary versus testis and non-gonadal transcriptomes (mean ± SD).

(B) WISH to detect zfs1 reveals expression in the ovary and cells anterior to the ovary (ventral view). Dorsal view shows no detectable zfs1 expression in the testes.

(C) Double FISH to detect zfs1 and gwîn shows zfs1 expression in gwîn⁺ FGPs anterior to the ovary (i) and oocytes within the ovary (ii).

(D) FISH to detect klf4⁺ FGPs and lecg⁺ oocytes in control(RNAi) and zfs1(RNAi) worms.

(E) Quantification of klf4⁺ FGPs and lecg⁺ oocytes in control(RNAi) and zfs1(RNAi) worms.

(F) FISH to detect klf4 (early male germ cells) in testes of control(RNAi) and zfs1(RNAi) worms.

(G) FISH to detect delta3 (somatic ovarian cells) and gwîn (female germline) in control(RNAi), zfs1(RNAi), and klf4(RNAi) worms.

(H) Quantification of delta3⁺ support cells in control(RNAi), zfs1(RNAi), and klf4(RNAi) worms. (E and H): means with 95% confidence intervals. Scale bars: (B) 500 μm; (C–F) 100 μm. See also Figure S6.

Although zfs1(RNAi) animals lack female germ cells, DAPI staining suggested that ovaries were still present (Figure 5D). To examine whether the somatic ovary was affected, we used delta3 as a marker, revealing increased somatic cells in zfs1(RNAi) worms (Figures 5G and 5H). We recently reported that klf4 RNAi resulted in loss of female germ cells and concomitant expansion of the somatic ovary; however, klf4 RNAi also leads to loss of testes and vitellaria (Issigonis et al., 2021). The ovary-specific effects of zfs1 RNAi suggest that the absence of female germ cells, rather than altered systemic cues resulting from lack of testes and vitellaria, triggers expansion of ovarian somatic cells in klf4(RNAi) worms (Figures 5G and 5H). Thus, germ cells and somatic support cells communicate bidirectionally during ovary regeneration.

Unexpected roles of AADC in germ cell development and regeneration

The laser-capture transcriptomes also revealed unanticipated gonadal expression of conserved genes not previously implicated in gonadal function. Aromatic L-amino acid decarboxylase (AADC) is a broadly conserved enzyme (Figure S7A) that catalyzes a key reaction in the production of monoamine neurotransmitters, such as serotonin and dopamine. Previous studies of asexual planarians showed that aadc is expressed in serotonergic neurons within the central nervous system, as well as in pigment cups of the photoreceptors and secretory cells around the pharynx (Currie and Pearson, 2013; März et al., 2013). Thus, it was surprising to find enriched expression of aadc mRNA in the planarian ovary (Figure S7B). We confirmed ovary-enriched aadc expression using in situ hybridization (Figure 6Ai), and also detected aadc expression in the testes and vitellaria (Figure 6Aii, iii). Double FISH to detect aadc and somatic support cell markers in ovaries (delta3, foxL, and ece1) and testes (ophis) revealed that aadc is expressed in both female and male somatic gonadal cells (Figures 6B and 6C). In ovaries, aadc expression was enriched in the ece1⁺foxL^low cells at the periphery (Figure 6B). To confirm this expression in somatic gonadal cells, we generated and validated polyclonal antibodies against AADC (Figures S7C, S7D; STAR Methods). Immunofluorescence labeling with anti-AADC antibodies detected AADC within the somatic support cells, which form scaffolds surrounding the germ cells in ovaries and testes (Figures 6D and 6E, and S7D). These data suggest that AADC could act locally within somatic gonadal cells to produce monoamines. In support of this idea, we find differential expression of various serotonin synthetic enzymes and other pathway components in the ovary and testis transcriptomes (Figure S7E).

Figure 6. Opposing roles of AADC in female and male germ cell regeneration.

(A) Expression of aadc (aromatic L-amino acid decarboxylase) is detected in the planarian reproductive system: (i) ovaries, (ii) testes, and (iii) vitellaria.

(B) FISH to detect aadc and ovarian somatic markers (delta3, ece1, and foxL): aadc transcripts are enriched in the distal ece1^high foxL^low ovarian somatic cells.

(C) FISH to detect aadc and testis somatic marker (ophis) indicates aadc expression in testis somatic support cells.

(D and E) Immunostaining for AADC protein in testes (D) and ovaries (E).

(F) FISH to detect klf4⁺ FGPs and lecg⁺ oocytes in control(RNAi) and aadc(RNAi) worms.

(G) FISH for delta3 to mark somatic ovarian cells in control and aadc(RNAi) worms.

(H and I) FISH to detect klf4 (H) and nanos (I), labeling early male germ cells in testes of control(RNAi) and aadc(RNAi) worms.

(J) FISH for ophis to mark testis somatic support cells in control and aadc(RNAi) worms.

(K–N) Quantification of klf4⁺ FGPs and lecg⁺ oocytes (K); delta3⁺ ovarian support cells (L); klf4⁺ early male germ cells (M); and ophis⁺ male somatic cells (N), in control(RNAi) and aadc(RNAi) worms. Means with 95% confidence intervals; ns, not significant. Asterisks in (B) and (E) indicate tuba. Scale bars: (A and H) 500 μm; (B–E, J) 50 μm; (F, G, and I) 100 μm. See also Figure S7.

To determine whether monoamines play a role in female germ cell development, we disrupted AADC function using the RNAi paradigm described above (Figure 4A). Knockdown of aadc resulted in the failure to regenerate ovaries after amputation: we observed a dramatic reduction of klf4⁺ FGPs and loss of oocytes (Figures 6F and 6K). Furthermore, aadc(RNAi) worms failed to regenerate the somatic compartment of the ovary (Figures 6G and 6L). These results suggest that AADC activity is necessary for the specification of FGPs and implicates ovarian somatic cells in this process. More broadly, accessory reproductive organs (vitellaria, oviducts, sperm ducts, and the gonopore) failed to regenerate or be maintained, and egg laying was abolished in aadc(RNAi) worms (Figures S7F–S7H). No obvious aadc expression was detected in oviducts or sperm ducts (Figure 6A), suggesting that AADC function within neurons, somatic gonadal cells, and/or vitellaria could act extrinsically in regeneration of these organs.

In contrast to the lack of ovaries after aadc knockdown, aadc(RNAi)animals displayed hyperplastic testes, full of early male germ cells (Figure 6H). Testes of control worms consist of peripherally located, immature (klf4⁺nanos⁺ and klf4⁻nanos⁺) germ cells, with luminal differentiating germ cells (spermatids and sperm) (Issigonis et al., 2021; Wang et al., 2007). Testes of aadc(RNAi) worms, however, consisted mainly of klf4⁺nanos⁺ early male germ cells and lacked differentiated cells (Figures 6H, 6I, and 6M). Unlike the ovaries, somatic cells in the testes appeared unaffected after aadc knockdown (Figures 6J and 6N). Taken together, these results implicate somatic gonadal cells as a potential source of monoamines, which may be acting differentially upon male and female germ cells in planarians.

DISCUSSION

Planarians exhibit extraordinary plasticity in reproductive development, including the ability to regenerate germ cells de novo; the mechanisms underlying this plasticity remain poorly understood. To uncover mechanisms underlying female germ cell development and regeneration, we generated gonad-specific transcriptomes. These studies identified genes defining progressive stages of female germ cell development and revealed heterogeneity of somatic ovarian cells. We uncovered intriguing similarities between female germ cells and neoblasts, the pluripotent stem cells that drive regeneration. We identified conserved, somatically expressed regulators, which have sex-specific functions in germ cell development and regeneration. Our findings underscore the key role played by somatic support cells in both development and regeneration of germ cells.

Germ cell sex in a simultaneous hermaphrodite and the relationship of FGPs to pluripotent stem cells

Planarians are simultaneous hermaphrodites, developing ovaries and testes in different regions of the body. Because the genes implicated thus far in the earliest stages of germ cell development (nanos and klf4) are expressed in both male and female germ cells (Handberg-Thorsager and Saló, 2007; Issigonis et al., 2021; Sato et al., 2006; Wang et al., 2007), it was unclear whether the sex of these early germ cells was already determined. The identification of zfs1 as a female-specific, germcell-intrinsic factor indicates that planarian germ cells “know” their sex early in their development. Presumably, inputs from global-patterning signals that define the animal’s major body axes (Reddien, 2018) are integrated such that germ cells born antero-ventrally toward the base of the brain adopt female fates, whereas those born dorso-laterally adopt male fates. Exploring the impact of global-patterning signals on sex-specific germ cell specification will be an important future direction.

The distribution of FGPs in fields anterior to the ovaries suggests two plausible scenarios. These fields could represent streams of migratory female germ cells that will later be incorporated into the ovary, similar to the example of planarian eye progenitors that are found in streams outside of the photoreceptors (Atabay et al., 2018; Lapan and Reddien, 2012). Alternatively, early female germ cells may be specified in a relatively broad antero-ventral region, but only those cells interacting with somatic gonadal cells (or other anatomical landmarks) will be incorporated into the ovary and produce oocytes. Whichever of these models best reflects reality, the gene expression changes we identified throughout female germ cell development suggest that FGPs/oogonia found around the margin of the ovary differentiate peripherally into oocytes that mature and move internally toward the tuba, where they exit from the ovary for fertilization (Figure 7).

Figure 7. Schematic of planarian female germ cell differentiation and the roles of somatic gonadal cells.

(i) Neoblasts specialize into FGPs antero-ventrally toward the base of the brain. FGPs around the margin of the ovary differentiate peripherally into oocytes that mature and move internally toward the tuba. Gene expression at distinct stages of female germ cell differentiation is indicated to the left. Expression of female-specific zfs1 reveals early and spatially restricted specification of germ cell sex in a simultaneous hermaphrodite. (ii) Somatic ovarian cells are closely associated with germ cells and display spatial heterogeneity in gene expression. Tuba-distal aadc⁺ somatic cells are required for regeneration of the female germ cells, whereas foxL⁺ tuba-proximal cells regulate oocyte development, suggesting distinct functional domains within the ovarian somatic compartment. (iii) Loss of female germ cells triggers an expansion of ovarian somatic cells, suggesting that feedback from female germ cells regulates support cell numbers.

Previous studies have noted the similarities between neoblasts and germ cells, from the presence of ribonucleoprotein granules called chromatoid bodies, to the expression of several genes associated with germline development in other animals, including vasa, piwi, tudor, pumilio, and bruno (Guo et al., 2006; Reddien et al., 2005; Salvetti et al., 2005; Shibata et al., 1999; Solana et al., 2009). Single-cell transplantation into lethally irradiated hosts revealed a subpopulation of neoblasts capable of giving rise to all cell types in the animal; these pluripotent neoblasts are functionally defined as clonogenic neoblasts (cNeoblasts) (Wagner et al., 2011). Single-cell sequencing efforts to characterize a pluripotent neoblast subpopulation identified TSPAN-1 as a cell-surface marker that can be used to enrich for cNeoblasts (Zeng et al., 2018). These authors identified tgs-1 and pks1 (here renamed gwîn) as signature transcripts associated with this tspan-1⁺ cluster in the asexual strain of S. mediterranea. We have shown that, in the sexual strain, these two genes are expressed abundantly in FGPs. Although gwîn is expressed at high levels constitutively in germ cells during homeostasis, it is only upregulated in neoblasts in response to wounding. This wound-responsive, neoblast-specific upregulation of genes associated with early stages of female germ cell development is particularly intriguing given the recent demonstration that neoblast fates are far more plastic than generally appreciated: specialized neoblast subclasses that express transcription factors associated with specific cell types are capable of generating daughters that adopt other fates (Raz et al., 2021). These authors also found that tgs-1 was not expressed in all early neoblast colonies; rather, their single-cell sequencing analysis detected tgs-1 expression in clusters displaying features of neural specification (Raz et al., 2021). Whether tgs-1 expression in female germ cells reflects additional similarities between FGPs and neural fates (e.g., reliance on post-transcriptional control of gene expression; Kulkarni et al., 2020) will be a fascinating avenue for future study. Nonetheless, our results suggest that neoblast plasticity may involve transient activation of some germ-cell-associated genes. Similarities in gene expression between germ cells and pluripotent neoblasts during regeneration are consistent with previous observations that planarian germ cells are capable of contributing to the regeneration of somatic tissues (Gremigni et al., 1980a; 1980b).

Ovarian somatic support cells: Key players in germ cell regeneration

Somatic support cells of planarian ovaries were observed ultrastructurally decades ago (Fischlschweiger, 1991, 1994; Gremigni and Falleni, 1998; Gremigni and Nigro, 1983; Harrath et al., 2011), but their functional characterization began with the discovery of ophis, a GPCR-encoding gene that is expressed in somatic gonadal cells and required for germ cell differentiation (Saberi et al., 2016). The paucity of these cells had thus far hindered their discovery by bulk and single-cell sequencing approaches (Fincher et al., 2018; Zayas et al., 2005). We used LCM to overcome this limitation, and the resulting gonad-specific transcriptomes enabled us to identify a set of sexually dimorphic ovarian support cell markers (delta3, foxL, and ece1). Their expression patterns revealed heterogeneity within ovarian somatic cells: enriched expression of ece1 and aadc was observed in cells distal to the tuba, whereas enriched expression of foxL was detected in cells proximal to the tuba, associated with later-stage oocytes. We found that aadc was required for FGP specification and maintenance, whereas foxL was required for oocyte differentiation and maintenance, without affecting FGPs (Figure 7). These disparate effects suggest that the tuba-distal and tuba-proximal somatic cell populations may control distinct stages of female germ cell development (Figure 7). Such distinct functional domains would be consistent with the recently described functional compartmentalization of the somatic niche in the germarium of the Drosophila ovary (Shi et al., 2021; Tu et al., 2021) and the distinct populations of granulosa cells regulating reproductive onset versus duration in the mouse ovary (Niu and Spradling, 2020).

The expression patterns and knockdown phenotypes of the Notch-signaling components reported here suggest that interactions between ovarian somatic cells (which express delta3 and notch4) and accessory reproductive organs, the oviduct and tuba (which express notch2 and notch4), help establish (and/or maintain) proper ovarian structure. Notch signaling also affects the germ cells: RNAi knockdown of delta3 or notch2 resulted in both the skewed distribution of FGPs toward the midline and an increase in the number of FGPs. The midline skewing could reflect the role of delta3 in midline patterning: it is expressed at the midline and knockdown in the asexual strain leads to cyclopia (Sasidharan et al., 2017). We did not observe cyclopic worms in our RNAi regeneration assays, so it remains to be determined if FGPs are more sensitive to the disruption of midline cues than the photoreceptors, or if the altered distribution reflects altered signaling from the somatic ovary. Nonetheless, potential alterations in midline patterning seem unlikely to affect the number of FGPs; thus, the somatic ovary may be capable of communicating with FGPs and regulating their specification and/or proliferation at a distance.

FoxL2 is a critical regulator of granulosa cell fate in mammals; homologs of this gene had yet to be characterized functionally in any invertebrate. Our observations that a planarian homolog (foxL) is expressed robustly in a subpopulation of somatic ovarian cells and is required for oocyte differentiation and maintenance suggest that the female-specific role of FoxL family members in the somatic gonad may well predate the emergence of vertebrates. The sexually dimorphic expression of planarian foxL is particularly intriguing in the context of this simultaneous hermaphrodite. In mice, female-specific FoxL2 and male-specific Dmrt1 transcription factors act in a mutually antagonistic manner to maintain gonadal sex: post-natal conditional knockout of foxL2 results in granulosa cell transdifferentiation into Sertoli cells (Uhlenhaut et al., 2009); by contrast, post-natal conditional knockout of dmrt1 results in Sertoli cell transdifferentiation into granulosa cells (Matson et al., 2011). In addition to foxL, planarians also have a dmrt1 homolog (dmd-1), which is expressed specifically in male reproductive organs (including somatic cells of the testes) and is required for the specification, differentiation, and maintenance of male germ cells as well as accessory reproductive organs (Chong et al., 2013). Whether functional antagonism, like that observed in mice, between female-specific foxL and male-specific dmd-1 also plays a role in maintaining gonadal sex in this simultaneous hermaphrodite will be an important avenue for future studies.

Potential non-neuronal roles of monoamines in germ cell development

Biogenic monoamines, like serotonin and dopamine, act as neurotransmitters across the animal kingdom (Weiger, 1997; Yamamoto and Vernier, 2011). It is perhaps less widely appreciated that many of these molecules predate the evolution of nervous systems (Roshchina, 2016) and play other, non-neuronal roles (Bellono et al., 2017; Lv and Liu, 2017; Matsuda et al., 2004). Our finding that AADC, an enzyme required for monoamine synthesis, plays sex-specific roles in planarian germ cell regeneration implicates monoamines in the regulation of crucial stages of germ cell development, from specification to differentiation. AADC is involved in the synthesis of many different monoamines (including serotonin, tyramine, tryptamine, histamine, dopamine, and other catecholamines); thus, it is unclear whether the observed sex-specific effects reflect sex-specific activities of different monoamines or whether male and female tissues respond differentially to the same monoamine.

Although we do not yet know the identity of the monoamine(s) mediating the aadc(RNAi) phenotype, planarian gonads possess the machinery to respond differentially to the same monoamine. For example, we queried the gonad transcriptomes and found evidence for serotonergic pathways in both testes and ovaries (Figure S7E). Tryptophan hydroxylase (TPH) catalyzes the conversion of L-tryptophan to 5-hydroxy-L-tryptophan, which is decarboxylated by AADC to produce serotonin. Vesicular monoamine transporters (VMAT) then package serotonin into vesicles for release, where it can bind serotonin receptors (5HTR). Serotonin transporters (SERTs) terminate signaling by taking up extracellular serotonin. A homolog of the Drosophila peripheral tryptophan hydroxylase (TPH), henna, is highly expressed in both testes and ovaries, relative to the neuronally expressed tph1 (Currie and Pearson, 2013; März et al., 2013; Sarkar et al., 2019). Planarian gonads also express vmat and sert homologs, suggesting the ability to release and recycle serotonin. Notably, ovaries and testes appear to differentially express different vmats, serts, and serotonin receptors. Thus, gonadal cells are capable of producing and responding locally to monoamines such as serotonin; the expression of distinct receptors in ovaries and testes suggests the possibility of differential responses. The expression of AADC by somatic support cells and the presence of specific monoamine receptors in both ovaries and testes supports the idea that these molecules can be produced and sensed locally within the gonads. Moreover, a recent report showed that serotonin could induce ovary development in another planarian species (Dugesia ryukyuensis), further supporting the role of monoamines in planarian reproductive system development (Sekii et al., 2019).

Might monoamines act in germ cell development in other animals? Genes encoding an entire serotonergic network, consisting of synthetic enzymes, receptors, and transporters, are expressed in mammalian ovaries (Dubé and Amireault, 2007). In mice and humans, two paralogous genes encode tryptophan hydroxylase, the rate-limiting enzyme in serotonin synthesis. Expression of one paralog (Tph1) is found in a wide range of non-neuronal tissues and is responsible for synthesizing ~95% of the body’s serotonin, whereas expression of the other (Tph2) is detected largely in neural cells (reviewed in Amireault et al., 2013). The embryonic defects observed in offspring of Tph1^−/− knockout females (Côté et al., 2007) are confounded by potential indirect effects from the compromised physiological states of these mice, which have diabetes and anemia (Amireault et al., 2011; Paulmann et al., 2009). These complications of interpreting Tph1^−/− phenotypes and the finding that the placenta serves as the source of serotonin that acts in the fetal forebrain (Bonnin et al., 2011) seem to have quelled investigations into potential roles of serotonin in oogenesis and early embryonic development. Furthermore, because Tph2 is among the genes showing the greatest degree of upregulation in the female somatic gonad at the time of sex determination (Beverdam and Koopman, 2006), it seems quite likely that it may act redundantly with Tph1 in the ovaries. Similar expression of multiple serotonin receptors in the gonads would also likely compensate for loss of any single gonadally enriched gene. Thus, exploring potential roles of these deeply conserved molecules in the reproductive organs of a simpler animal, like the planarian, may overcome such concerns about redundancy and guide future tissue-specific knockout studies.

Regardless of whether or not the role of monoamines uncovered here reflects deeply conserved mechanisms for regulating germ cell development across animal phylogeny, the observed effects upon the planarian’s reproductive system have other important implications. Planarians are free-living representatives of the phylum Platyhelminthes (the flatworms) and their reproductive system shares several important features with those of parasitic flatworms (flukes and tapeworms), which have major impacts on global public health (Collins and Newmark, 2013). All of these flatworms have ectolecithal eggs (i.e., yolk on the outside); yolk cells are produced by specialized accessory reproductive organs known as vitellaria and they are essential for embryonic development. Furthermore, transmission of parasitic flatworms requires prolific egg and yolk cell production. Because inhibition of aadc results in the loss of ovaries and other female accessory reproductive organs (including vitellaria), unraveling the role of monoamines in the female flatworm reproductive system may lead to new approaches for preventing parasite transmission.

Limitations of the study

This work represents a first step in characterizing global gene expression in the planarian ovary and identifying genes that play critical roles in various stages of female germ cell development in this animal. The use of LCM to identify genes with enriched expression in ovarian tissues was quite successful, as indicated by the ~90% validation using in situ hybridization as well as by our ability to define successive stages of female germ cell development and distinct somatic ovarian cell types. However, this initial expression-based screen likely missed genes that are also expressed abundantly in non-gonadal tissues. Future single-cell sequencing efforts will provide a more comprehensive picture of gene expression in all ovarian cell types. Moreover, the functional analysis presented here, using RNAi in the context of an ovary-regeneration assay, only allowed us to examine germ cell phenotypes in knockdown animals that could properly regenerate their heads (including cephalic ganglia) and their feeding circuitry. Thus, any genes required for proper female germ cell development that are also required more broadly for proper neoblast function, neural activity, or viability would be missed by our assay, because knockdown planarians would fail to regenerate, feed, or survive. Despite such limitations, the genes identified here will facilitate future studies in which subtler experimental perturbations (e.g., localized ovarian ablation) are used to dissect more precisely the processes by which planarian neoblasts give rise to functional female germ cells.

STAR★METHODS

RESOURCE AVAILABILITY

Lead contact

• Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Phillip A. Newmark (pnewmark@morgridge.org).

Materials availability

• Newly generated materials generated in this study will be made available upon request.

Data and code availability

• All next-generation sequencing data have been deposited at the GEO repository under the accession number GSE191229 and are publicly available as of the date of publication.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Planarian maintenance and care

Hermaphroditic, sexual S. mediterranea were maintained at 18°C in 0.75X Montjuïc salts (Cebrià and Newmark, 2005) supplemented with 50 μg/mL Gentamicin (Gemini Bio-Products) and fed beef liver paste. Asexual S. mediterranea (clone CIW4; Sánchez Alvarado et al., 2002) were maintained at 22°C in 1X Montjuïc salts and fed beef liver paste. Planarians were maintained in Ziploc containers and kept in Petri dishes for RNAi experiments. Planarians were starved at least one week before initiating RNAi experiments, in situ hybridizations, or immunostainings.

METHOD DETAILS

Laser-capture microdissection and RNA extraction

Planarians were killed in chilled 2% HCl in PBS (RNAse-Free) and then fixed in 100% Acetone for 1 hour at −20°C. Planarians were then incubated in 10, 20, 30% sucrose in PBS (RNAse-Free) for 20 minutes each before embedding in Tissue Freezing Medium (TFM) blocks (Ted Pella). The samples were then cryosectioned at 16 mm thickness onto PEN membrane slides (Thermo Fisher). The slides were stained with 1% cresyl violet and 1% eosin Y (Sigma) in 75% ethanol (Clément-Ziza et al., 2008). The tissue samples were dissected using Veritas Laser Capture Microdissection System (Arcturus) and captured on to Capsure HS LCM Caps (Thermo Fisher). After capture, RNA was isolated using the Arcturus PicoPure RNA isolation kit (Thermo Fisher). Ovarian tissue samples were collected from 14 worms (10 worms used for cross-sections and 4 worms used for sagittal sections) for each replicate (3 replicates total). Testis- and non-gonadal-tissue samples were collected from 6 worms for each replicate (4 worms used for cross-sections and 2 worms used for sagittal sections). RNA concentrations were determined using Qubit Fluorometer (Thermo Fisher), and RNA integrity was analyzed using 2100 Bioanalyzer Instrument (Agilent).

RNA sequencing and gene expression analysis

Libraries were generated using Trueseq RNA stranded kit (Illumina), and 100 nt single-end sequencing reads were generated on Hi-Seq 2500 (Illumina). Trimming of adapters and low-quality reads as well as subsequent read mapping and differential gene expression analyses were performed using CLC Genomics Workbench (Qiagen). Reads were mapped to a de novo assembled S. mediterranea transcriptome (smed_20140614) (Robb et al., 2015) with the Smed-zfs1 transcript added. The transcriptome was annotated in OmicsBox (BioBam) using the SwissProt Protein database. Heatmaps were generated in R using the ComplexHeatmaps package (Gu et al., 2016).

Molecular biology methods

200–1000 bp fragments of the gene of interest (Table S2) were amplified using Platinum Taq DNA Polymerase (Invitrogen) from cDNA. Amplified fragments were cloned into pJC53.2 (Collins et al., 2010) via TA cloning. Riboprobes and dsRNA were synthesized as previously described (King and Newmark, 2013; Rouhana et al., 2013).

In situ hybridization and immunohistochemistry

Whole-mount ISH was performed as previously described (King and Newmark, 2013) with modifications for larger sexual worms: formaldehyde fixation was increased to 30 min, proteinase K treatment was increased to 20-25 min, and post-proteinase K fixation was increased to 30 min. pHH3 was labeled using anti-phospho-Histone H3 (Ser10) (Millipore Sigma). AADC polyclonal antibody was generated by injecting a synthetic peptide (DVYTPKMDAEEFRKRGKE) into rabbits (Pierce Biotechnology, Rockford, IL). The serum was affinity purified with the peptide antigen and used at a dilution of 1:2000 for immunofluorescence. Colorimetric and FISH/immunofluorescence samples were imaged on Axio Zoom.V16 (Carl Zeiss) and LSM 710 or 880 confocal microscope (Carl Zeiss), respectively. Cell counts were performed using the spot detection tool in Imaris (Bitplane).

RNA interference

Knockdowns were performed by feeding in vitro-transcribed dsRNA, as previously described (Rouhana et al., 2013). 6-8 mature sexual animals were fed ~10-20 ug dsRNA mixed with 90 uL liver puree: water mixture 5:1 (with food coloring) in a petri dish. For all RNAi experiments, dsRNA corresponding to bacterial ccdB gene was used for negative controls. Animals were fed dsRNA 4 times, cut posterior to the ovaries, and tail fragments were allowed to regenerate for two weeks, followed by 8 dsRNA feedings.

Protein domain \ analysis

Conserved protein domains were analyzed using InterProScan, SMART and Phobius protein domain analysis tools. Predicted protein sequences were aligned using MUSCLE.

QUANTIFICATION AND STATISTICAL ANALYSIS

All two-sample and three-sample comparisons were analyzed using Welch’s t-test or one-way ANOVA (Dunnett’s test), respectively, in Prism (GraphPad). Differential expression for RNAseq was analyzed in CLC Genomics Workbench (Qiagen) using the Wald test.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-Digoxigenin-POD, Fab fragments (sheep polyclonal)	Roche	Cat: 11207733910 RRID: AB_514500
Anti-Dinitrophenyl-HRP (rabbit polyclonal)	Vector Laboratories	Custom made
Anti-Fluorescein-POD, Fab fragments (sheep polyclonal)	Roche	Cat: 11426346910 RRID: AB_840257
Anti-phospho-Histone H3 (Ser10) (rabbit monoclonal)	Millipore Sigma	Cat: 05-817R-I; RRID: AB_11215621)
Anti-SmedAADC (rabbit polyclonal)	This paper	N/A
Peroxidase AffiniPure Goat Anti-Rabbit IgG (H+L)	Jackson ImmunoResearch Laboratories	Cat: 111-035-003 RRID: AB_2313567
Alkaline Phosphatase AffiniPure Goat Anti-Rabbit IgG (H+L)	Jackson ImmunoResearch Laboratories	Cat: 111-055-144 RRID: AB_2337953
Bacterial and virus strains
DH5-alpha	Thermo Fisher	Cat: EC0112
Chemicals, peptides, and recombinant proteins
Alcoholic Eosin Y	Sigma Aldrich	Cat: HT110116
Cresyl Violet Acetate	Sigma Aldrich	Cat: C5042
Critical commercial assays
PEN membrane glass slides	Thermo Fisher	Cat: LCM0522
Capsure HS LCM caps	Thermo Fisher	Cat: LCM0214
PicoPure RNA Isolation Kit	Thermo Fisher	Cat: KIT0204
Deposited data
Raw and analyzed data	This paper	GEO: GSE191229
Experimental models: Organisms/strains
Schmidtea mediterranea, Sexual strain	Zayas et al., 2005	N/A
Schmidtea mediterranea, Asexual clonal strain CIW4	Sánchez Alvarado et al., 2002	N/A
Oligonucleotides
Sequences used for probes and dsRNA	This Paper	Table S2
Recombinant DNA
Plasmid: pJC53.2	Collins et al., 2010	Cat: 26536 RRID: Addgene_26536
Software and algorithms
CLC Genomics Workbench	Qiagen	RRID: SCR_011853
R Studio	RStudio	RRID: SCR_000432
Complex Heatmaps	Gu et al., 2016	N/A
Zen	ZEISS	RRID: SCR_013672
GraphPad Prism	GraphPad Software	RRID: SCR_002798
Fiji	Schindelin et al., 2012	RRID: SCR_002285
Imaris	Oxford Instruments	RRID: SCR_007370
Other
Differential gene expression analysis	This paper	Table S1
Transcriptome (smed_20140614)	Robb et al., 2015	https://planosphere.stowers.org
PlanMine	Rozanski et al., 2019	https://planmine.mpibpc.mpg.de

Highlights.

• Laser-capture microdissection generated planarian gonad-specific transcriptomes

• Identified transcripts defining female germ cells and somatic ovarian cells

• Defined functional roles for conserved somatic gonadal genes in ovary regeneration

• Implicated somatically derived monoamines in male and female germ cell development