βcatenin-1 localization and transcriptomic profiling provide insights into the early development of planarians Authors: P. Coronel-Córdoba, M. Sureda-Gómez, J. M. Martín-Durán, F. Monjo, E. Saló, J. F. Abril, T. Adell Published: 2025 ABSTRACT BackgroundThe development of planarians is unique among Spiralians. Instead of the stereotypical spiral cleavage, planarians exhibit a dispersed cleavage. There is no apparent gastrulation, and the morphogenesis of the yolk-feeding embryo remains a mystery. In this study, we examine the subcellular localization of βcatenin-1 and the transcriptomic profile during the early embryonic development of Schmidtea polychroa to shed light on these early events.ResultsThe first localization of βcatenin-1 in the nucleus occurs in yolk cells surrounding the embryonic syncytium. By 24 h post-deposition, βcatenin-1 starts to be nuclear in blastomeres, coinciding with the activation of signaling and cell motility genes. During morphogenesis of the yolk-feeding embryo, βcatenin-1 is first localized in the nucleus at one pole (gut and pharynx progenitors) and in epidermal progenitors, and afterward in the embryonic pharynx. At this stage, genes involved in a first morphogenetic event are turned on. Following the yolk ingestion by the embryo, a dramatic transcriptomic shift occurs that coincides with the activation of genes related to cell proliferation. Finally, between 5 and 7 days post-deposition, βcatenin-1 is massively located in the nucleus, and genes involved in the morphogenesis and patterning of the adult tissues get activated.ConclusionsOur findings provide new insights into the early developmental events of Schmidtea polychroa, including cleavage, the involvement of βcatenin-1 in forming the embryonic tissues, and the morphogenesis of two distinct body plans. These findings are significant to understanding the evolution of the peculiar mode of planarian development.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12915-025-02451-0. Background Within the core of developmental biology lies the understanding of the processes by which a single fertilized egg gives rise to the final outline of an organism. This question has been historically tackled using a few model organisms established as animal models [1]. However, the study of non-model organisms is essential to elucidate the evolutionary history of certain lineages and to unravel the developmental diversity across animals. Here, we study the early events during planarian embryogenesis. Planarians, the common name that receive the order of Tricladida among Platyhelminthes, are a known model for studying regeneration thanks to their plasticity, which relies on a population of adult pluripotent stem cells called neoblasts [2, 3]. Even though the regeneration process in planarians, especially in freshwater species, has been extensively documented [4, 5], just a few studies have focused on understanding planarian embryonic development, mainly due to the difficulties of experimentally handling planarian embryos [6]. However, a detailed study of the early embryogenesis of the planarian embryo is key to understand the peculiar mode of Tricladida development and their unique plasticity. Planarians or triclads belong to the Spiralia (Lophotrocozoa) [7]. However, planarians have lost the ancestral mode of spiral cleavage that characterizes the clade and present a poorly understood mode of development. Unlike the synchronous and stereotypical cleavage that characterize spiralian embryos, planarian blastomeres do not maintain the direct contact with each other and are found dispersed among yolk cells [8–10]. Planarians develop from polyembryonic ectolecithal eggs, in which the yolk is stored in the vitellocytes surrounding oocytes [11]. Each egg, or cocoon, may contain from 5 to 20 fertilized zygotes, depending on the species. Le Moigne characterized planarian embryogenesis and proposed a staging that was later adopted and modified by Cardona et al. [9, 12], which comprises 8 stages. During stage 1 (0–32 h post-deposition, hpd) each zygote divides into daughter blastomeres that remain unattached, while the closest vitellocytes/yolk cells fuse to form a syncytium that encloses the blastomeres [13]. This dispersed cleavage was described by German embryologists as “Blastomerenanarchie” [14–16] since no apparent organization comparable to other embryonic cleavage programs was noticed. At stage 2 (32–72 hpd), the yolk-feeding embryo develops. Some blastomeres migrate to the periphery of the syncytium and differentiate into primary epidermis cells that will end up enclosing the whole syncytium. At the same time, a group of blastomeres that remain together form a primordium that will originate two structures: the embryonic pharynx and the embryonic gut [10]. Some undifferentiated blastomeres remain in the periphery underneath the epidermis, composing the germ band, which will give rise to the adult structures at later stages. The apparent lack of coordinated cell movements during the formation of the yolk-feeding embryo has raised the question whether, in planarians, there is a process of gastrulation itself that gives rise to the three germ layers [17]. At stage 3 (72 hpd–4 days post-deposition, dpd), the embryo starts to feed the surrounding yolk cells and grows in size. The first signs of cellular proliferation are reported at this stage [17]. During stages 4 and 5 (5–7 dpd), yolk engulfment stops, and the embryo starts to flatten dorso-ventrally, while the cells in the germ band start to differentiate into the adult cell types, such as epidermal cells that will intercalate into the embryonic epidermis [10] and the neuromuscular tissue [18]. Furthermore, at this stage, the neoblasts acquire the capacity to repopulate irradiated host after transplantation, suggesting that neoblast differentiation could be also occurring at this stage [19]. During this period, the embryonic structures degenerate and become replaced by the adult tissues, while the bilaterian body axes of the adult are established [12]. Finally, at stages 6 to 8 (7–15 dpd), differentiation and morphogenesis of the adult definitive tissues take place, giving rise to the juveniles. The canonical Wnt/βcatenin pathway is a key intercellular signaling pathway of multicellular animals and is a main player in constructing the metazoan body plan [20, 21]. It is a universal symmetry-breaking mechanism in animals, leading to pattern formation and self-organization. The central role of the Wnt/βcatenin pathway during development relies on the transduction of both chemical and mechanical signals to polarize cells, orient the axis of cell division, and simultaneously program daughter cells to adopt diverging fates according to the environment [22]. The activation of the pathway through the binding of the extracellular Wnt ligands to its coreceptors Frizzled/Lrp6 triggers the inhibition of the βcatenin destruction complex, allowing the nuclear localization of βcatenin. In the nucleus, βcatenin interacts with the transcription factor TCF to modulate chromatin accessibility and gene expression [23]. During early development, the Wnt/βcatenin pathway spatially organizes embryonic cells and specifies the primary body axis, referred to as the animal-vegetal, oral-aboral, or anteroposterior in bilaterians [21]. βcatenin nuclear localization at the vegetal pole is also an evolutionarily conserved mechanism to define the site of gastrulation and endomesoderm specification [24–27]. In adult planarians, the Wnt/βcatenin pathway specifies the anteroposterior axis during regeneration and homeostasis [28–34]. Accordingly, its inhibition leads to a complete anteriorization of the animals [28–30]. The immunostaining analysis using an antibody against the canonical βcatenin ortholog in planarians, βcatenin-1 (βCAT-1), demonstrated the existence of a gradient of nuclear βCAT-1 from posterior to anterior, and additional roles of the pathway, for instance, during neural regeneration and gonad development [35–37]. To further understand planarian early development, we took advantage of the anti-βCAT-1 antibody to analyze the activation of the Wnt/βcatenin during the early stages of embryogenesis of the planarian Schmidtea polychroa (Spol). We combined this information with the transcriptomic profile of seven time points corresponding to the early stages of planarian embryogenesis, from the early cleavage to the onset of the adult body plan (from 4 hpd to 7 dpd). Our results show that the first event of βCAT-1 nuclear localization occurs in yolk cells, suggesting their signaling role during cleavage. The first embryonic nuclear localization of βCAT-1 occurs at 24 hpd in blastomeres with an elongated morphology, coinciding with the expression of genes involved in cell signaling and migration. At 32 hpd, βCAT-1 is located in the nucleus of progenitors of the three embryonic tissues: the primary epidermis, the embryonic gut, and the pharynx. At 48 hpd, βCAT-1 is located in the nucleus of cells of the embryonic pharynx primordium, the first noticeable asymmetry of the planarian embryo. At 72 hpd, we detect activation of the genes related to the first event of tissue morphogenesis. A dramatic transcriptional shift occurs between 72 hpd and 4 dpd accompanied by the activation of genes related to cell proliferation. Finally, between 5 and 7 dpd, βCAT-1 becomes extensively located in the nucleus, coinciding with the activation of genes related to the morphogenesis and patterning of the adult planarian tissues. Results The first event of βcatenin-1 nuclear localization occurs in yolk cells Planarians, or triclads, belong to the Spiralia (Lophotrocozoa) clade (Fig. 1A). They are hermaphrodites, and fertilization takes place internally during mating. Fertilized eggs, or cocoons, are laid and attached to the substrate to avoid being swept away by the current (Fig. 1B, C). Cocoons contain several embryos immersed in yolk cells, which can be identified as type I, filling most of the cocoon, and type II, which directly surrounds the zygote and will form a syncytium during cleavage (Fig. 1D). To study the early stages of planarian development, two strategies were carried out: (1) immunostaining to analyze the subcellular dynamics of βcatenin-1 (βCAT-1), with a specific anti-βCAT-1 antibody [36], and (2) bulk RNA-seq of the whole cocoons (Fig. 1E). Both analyses were performed at eight time points spanning from the zygote cleavage to the establishment of the adult body plan: 4, 24, 32, and 48 hpd and 3, 4, 5, and 7 dpd (Fig. 1E). We analyzed this data using paired comparisons between time samples to find differentially expressed genes (DEGs) at each stage (Table 1) [38]. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of the DEGs were performed (Tables 2 and 3) [38]. To check temporal patterns of the DEGs throughout the developmental stages, time clustering along the all-time series was calculated using temporal clustering analysis (TCA) (Additional file 1: Fig. S1). The multidimensional scaling (MDS) plot of the RNA-seq shows a temporal split in between the samples as expected in a developmental transcriptomic profile (Additional file 2: Fig. S2 A, B).Fig. 1Experimental design. βCAT-1 nuclear localization and RNA-seq analysis indicate that the first embryonic asymmetry is at 24 hpd. A Representative metazoan phylogenetic tree. Red asterisk: Platyhelminthes. B Representative image of S. polychroa species collected from Tordera river (Sant Celoni, Barcelona). Red asterisk: mating adults. Blue asterisks: cocoons attached to the substrate. C Coocon corresponding to 32 hpd. D Schematic representation of a cocoon right after fertilization. E Experimental strategy. Immunostaining with anti-βCAT-1 and anti-TYR-TUB combined with bulk RNA-seq of embryos corresponding to 4–24–32–48 hpd and 3–4–5–7 dpd was performed. F Schematic representation of a planarian embryo at 4 hpd. G TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 4 hpd. Nucleus are stained with DAPI (blue). Red arrows indicate βCAT-1 + yolk cells (n = 7 sections corresponding to 3 embryos). H Magnification of a blastomere showing cytoplasmic βCAT-1. I Magnification of a yolk cell showing nuclear βCAT-1. J Volcano plot representing DEGs at 24 hpd (red) vs. 4 hpd (blue); green dashed lines correspond to adjusted p values > 0.05 (horizontal) and |log2FC|> 1 (vertical). K Schematic representation of a planarian embryo at 24 hpd. L TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 24 hpd. Nucleus are stained with DAPI (blue). (n = 3 sections corresponding to 3 embryos). M Magnification of a blastomere showing nuclear βCAT-1. Red arrow indicates the trail of Tyr-tub (n = 2/3). N Magnification of a blastomere showing cytoplasmic βCAT-1. O Magnification of a yolk cell with nuclear βCAT-1. P Heatmap of the DEGs at 24 hpd vs. 4 hpd involved in cell signaling. The heatmap color gradient represents log2(CPM). hpd hours post deposition, dpd days post deposition, DEG differentially expressed gene, CPM counts per million. Scale bars: (B): 2 mm (C): 200 µm; (G, L): 40 µm, (H, I; M, O): 10 µm Experimental design. βCAT-1 nuclear localization and RNA-seq analysis indicate that the first embryonic asymmetry is at 24 hpd. A Representative metazoan phylogenetic tree. Red asterisk: Platyhelminthes. B Representative image of S. polychroa species collected from Tordera river (Sant Celoni, Barcelona). Red asterisk: mating adults. Blue asterisks: cocoons attached to the substrate. C Coocon corresponding to 32 hpd. D Schematic representation of a cocoon right after fertilization. E Experimental strategy. Immunostaining with anti-βCAT-1 and anti-TYR-TUB combined with bulk RNA-seq of embryos corresponding to 4–24–32–48 hpd and 3–4–5–7 dpd was performed. F Schematic representation of a planarian embryo at 4 hpd. G TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 4 hpd. Nucleus are stained with DAPI (blue). Red arrows indicate βCAT-1 + yolk cells (n = 7 sections corresponding to 3 embryos). H Magnification of a blastomere showing cytoplasmic βCAT-1. I Magnification of a yolk cell showing nuclear βCAT-1. J Volcano plot representing DEGs at 24 hpd (red) vs. 4 hpd (blue); green dashed lines correspond to adjusted p values > 0.05 (horizontal) and |log2FC|> 1 (vertical). K Schematic representation of a planarian embryo at 24 hpd. L TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 24 hpd. Nucleus are stained with DAPI (blue). (n = 3 sections corresponding to 3 embryos). M Magnification of a blastomere showing nuclear βCAT-1. Red arrow indicates the trail of Tyr-tub (n = 2/3). N Magnification of a blastomere showing cytoplasmic βCAT-1. O Magnification of a yolk cell with nuclear βCAT-1. P Heatmap of the DEGs at 24 hpd vs. 4 hpd involved in cell signaling. The heatmap color gradient represents log2(CPM). hpd hours post deposition, dpd days post deposition, DEG differentially expressed gene, CPM counts per million. Scale bars: (B): 2 mm (C): 200 µm; (G, L): 40 µm, (H, I; M, O): 10 µm Double immunostaining analyses at the earliest time point analyzed (4 hpd) using anti-βCAT-1 and anti-tyrosine-tubulin (TYR-TUB), which specifically labels blastomeres, shows that βCAT-1 is not localized in the nucleus of the cleaving blastomeres but in the nucleus of yolk cells around them (Fig. 1F–I, Table 4 [38]). Immunostainings performed in embryos before 4 hpd did not show nuclear localization of βCAT-1 in the surrounding yolk cells, suggesting that this occurs at 4 hpd and onwards (Additional file 3: Fig. S3 A–D). The medium distance for βCAT-1 + yolk nucleus to the closer blastomere was 65,32 µm (n = 4) (Table 4) [38]. GO enrichment analysis of the RNA-seq DEGs revealed that the genes downregulated at 24 hpd vs. 4 hpd are related to mitochondrial activity (mitochondrial 2-oxoglutareate7malate carrier), ERAD pathway (E3 ubiquitin-protein ligase RNF185), Golgi transport (Golgi complex subunit 6), and secretory function (phosphopentomutase) (Fig. 1. J; Additional file 1: Fig. S1 E). Moreover, genes downregulated in several comparisons against 4 hpd appear overrepresented in KEGG pathways related to processing in the endoplasmic reticulum (ER) (Table 3) [38]. These results are in concordance with the accumulation of mitochondria and ER in the yolk cells surrounding the dividing blastomeres previously described by electron microscopy studies [10], indicating active biogenesis and proteins traffic during the early cleavage (4 hpd). Our results suggest that yolk cells could have a signaling role during the early cleavage since we observed nuclear βCAT-1 and transcriptomic signs of protein trafficking. β-Catenin-1 nuclear localization in blastomeres and the expression of signaling and motility genes occurs at 24 hpd Immunostaining analysis of 24 hpd embryos with anti-βCAT-1 and anti-TYR-TUB shows that βCAT-1 is still nuclear in the surrounding yolk cells and starts to be expressed in the embryo (Fig. 1K–O). Interestingly, while βCAT-1 is localized in the cytoplasm of some blastomeres, others with an elongated morphology show nuclear βCAT-1 (Fig. 1M, N). Both the morphology and the trail of TYR-TUB observed in blastomeres with nuclear βCAT-1 suggest that they could be migrating through the forming syncytium. Further supporting this possibility, GOs of the DEGs up-regulated at 24 hpd vs. 4 hpd are related to cell migration and cell motility (actin-binding LIM protein 2 and activin receptor-1) (Fig. 1J; Additional file 3: Fig. S3 F). Interestingly, genes involved in syncytium formation, like hu-li tai shao, which is involved in the ring canal formation during Drosophila oogenesis [39], are also upregulated at this stage, while the syncytium is developing (Fig. 1J; Table 1 [38]). Signaling genes (egr-2, activin receptor 1, lox5b) were also specifically turned on from 24 hpd onwards (Fig. 1J, P). These results show that the earliest nuclear localization of βCAT-1 in blastomeres occurs in elongated cells that could be migrating through the new syncytium. At this point, the transcriptomic activation of signaling genes suggests that the early patterning of the zygote is starting. The polarization of blastomeres with nuclear βcatenin-1 coincides with the specification of the embryonic tissues During stage 2 (32–72 hpd), some blastomeres remain together, giving rise to the primordium of the embryonic pharynx and gut, while others migrate through the syncytium to give rise to the primary epidermis [10, 13]. By 48 hpd, the primary epidermis has fully enclosed the syncytium, and the primordium differentiates into the embryonic gut and pharynx. At 72 hpd, the embryo begins to ingest the surrounding yolk cells into the inner gut cavity and pushes the rest of the blastomeres to the periphery, forming the germ band (Fig. 2A, E, I).Fig. 2βCAT-1 is located in the nucleus in blastomeres of the pharynx and gut primordia, in the embryonic pharynx and epidermal precursors during stage 2. A Schematic representation of a planarian embryo at 32 hpd and the section shown. B TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 32 hpd. Nucleus are stained with DAPI (blue). pp pharynx primordia, pe primary epidermis. (n = 10 sections corresponding to 7 embryos). C Magnification of the pharynx primordium showing some blastomeres with nuclear βCAT-1. D Magnification of a blastomere, probably a precursor of primary epidermal cells, showing nuclear βCAT-1. E Schematic representation of a planarian embryo at 48 hpd and the section shown. F TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 48 hpd. Nucleus are stained with DAPI (blue). eg embryonic gut (n = 13 sections corresponding to 5 embryos). G Magnification of the developing embryonic gut (eg) showing two cells with nuclear βCAT-1 (red arrowhead). H Developing embryonic pharynx (ep) with two cells showing nuclear βCAT-1 (red arrowhead). I Schematic representation of a planarian embryo at 72 hpd and the section shown. J TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 72 hpd. Nucleus are stained with DAPI (blue). (n = 11 sections corresponding to 9 embryos). K Magnification of the cells in the germ band (gb), showing cytoplasmic βCAT-1 shared equally (red arrowhead). hpd hours post deposition. Scale bar: (B; F; J): 40 µm, (C, D; G, H; K): 10 µm βCAT-1 is located in the nucleus in blastomeres of the pharynx and gut primordia, in the embryonic pharynx and epidermal precursors during stage 2. A Schematic representation of a planarian embryo at 32 hpd and the section shown. B TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 32 hpd. Nucleus are stained with DAPI (blue). pp pharynx primordia, pe primary epidermis. (n = 10 sections corresponding to 7 embryos). C Magnification of the pharynx primordium showing some blastomeres with nuclear βCAT-1. D Magnification of a blastomere, probably a precursor of primary epidermal cells, showing nuclear βCAT-1. E Schematic representation of a planarian embryo at 48 hpd and the section shown. F TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 48 hpd. Nucleus are stained with DAPI (blue). eg embryonic gut (n = 13 sections corresponding to 5 embryos). G Magnification of the developing embryonic gut (eg) showing two cells with nuclear βCAT-1 (red arrowhead). H Developing embryonic pharynx (ep) with two cells showing nuclear βCAT-1 (red arrowhead). I Schematic representation of a planarian embryo at 72 hpd and the section shown. J TYR-TUB (magenta) and βCAT-1 (green) immunostaining on paraffin sections of a planarian embryo at 72 hpd. Nucleus are stained with DAPI (blue). (n = 11 sections corresponding to 9 embryos). K Magnification of the cells in the germ band (gb), showing cytoplasmic βCAT-1 shared equally (red arrowhead). hpd hours post deposition. Scale bar: (B; F; J): 40 µm, (C, D; G, H; K): 10 µm Immunostaining shows that at 32 hpd βCAT-1 is found in the nucleus of cells of the pharynx and gut primordium (Fig. 2A–C), located at one pole of the embryo, and in blastomeres surrounding the syncytium, where the primary epidermal cells are being specified (Fig. 2D). At 48 hpd, βCAT-1 is exclusively located in the nucleus of cells at one pole of the embryo, specifically in the cells of the primordium that give rise both to the embryonic pharynx and the embryonic gut (Fig. 2E, H; Additional file 4: Fig. S4 C, D). Nuclear βCAT-1 is not maintained in the primary epidermis (Additional file 4: Fig. S4 A-B), while some cells of the embryonic pharynx maintain nuclear βCAT-1 during the yolk ingestion at 72 hpd (Fig. 2I, J). At this stage, βCAT-1 is also detected in blastomeres of the developing germ band, although it remains in the cytoplasm (Fig. 2. I-K; Additional file 4: Fig. S4 E, F). Cell divisions were observed in these blastomeres of 32 hpd embryos (Table 4) [38], in which βCAT-1 was always cytoplasmatic and evenly distributed between daughter cells (Fig. 2K; Additional file 4: Fig. S4 E, F). The TCA analysis reveals a group of genes that specifically gets shut down at 32 hpd (cluster 1) (Fig. 3A, B; Additional file 1: Fig. S1 A) while another two clusters correspond to genes specifically upregulated at 48–72 hpd (cluster 4 and cluster 5) (Fig. 3C, D; Additional file 1: Fig. S1 M; O). On the one hand, the GOs of genes in cluster 1 show their involvement in protein transport (Fig. 3E). Further supporting this finding, we observed that DEGs upregulated at 32 vs. 24 hpd include genes involved in ubiquitination (marh6 and ube2o), sphingolipids metabolism and negative regulation of autophagy (degs1) which are known to regulate protein transport [40, 41] (Fig. 3F; Additional file 5: Fig. S5 A). The transcriptome also shows four autophagy-related genes specifically downregulated at 32 hpd (Additional file 6: Fig. S6 A). On the other hand, the GOs of cluster 4 correspond to genes related to protein translation (Fig. 3G), indicating that these functions are upregulated at 48–72 hpd. Considering that protein transport and autophagy are functions associated with the stress produced by protein biosynthesis [42, 43], its downregulation at 32 hpd indicates relatively low levels of biosynthesis occurring compared to the increase detected at 48–72 hpd.Fig. 3A transition in biosynthesis levels occurs between the 32 hpd and the 48–72 hpd, followed by the first morphogenetic event of the yolk-feeding embryo. A Schematic representation of a planarian embryo at 32 hpd. B Cluster 1 from TCA analysis showing genes specifically shut down at 32 hpd; each line corresponds to the expression pattern for one gene in the cluster, while color gradient is proportional to the membership score. C Schematic representation of planarian embryos at 48 hpd and 72 hpd. D Cluster 4 from TCA analysis, gathering genes specifically up-regulated at 24 hpd and between 48 and 72 hpd. E BP over-represented GO terms for the genes belonging to cluster 1, pinpointing functions related to protein transport. Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in up-regulated genes. F Volcano plot representing DEGs at 32 hpd (red) vs. 24 hpd (blue); green dashed lines correspond to adjusted p values > 0.05 (horizontal) and |log2FC|> 1 (vertical). G BP over-represented GO terms for cluster 4 genes, listing functions related to protein translation. H Volcano plot representing DEGs at 72 hpd (red) vs. 32 hpd (blue); green dashed lines correspond to adjusted p values > 0.05 (horizontal) and |log2FC|> 1 (vertical). hpd hours post deposition, DEG differentially expressed gene, TCA temporal clustering analysis, GO gene ontology, BP biological process, CPM counts per million A transition in biosynthesis levels occurs between the 32 hpd and the 48–72 hpd, followed by the first morphogenetic event of the yolk-feeding embryo. A Schematic representation of a planarian embryo at 32 hpd. B Cluster 1 from TCA analysis showing genes specifically shut down at 32 hpd; each line corresponds to the expression pattern for one gene in the cluster, while color gradient is proportional to the membership score. C Schematic representation of planarian embryos at 48 hpd and 72 hpd. D Cluster 4 from TCA analysis, gathering genes specifically up-regulated at 24 hpd and between 48 and 72 hpd. E BP over-represented GO terms for the genes belonging to cluster 1, pinpointing functions related to protein transport. Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in up-regulated genes. F Volcano plot representing DEGs at 32 hpd (red) vs. 24 hpd (blue); green dashed lines correspond to adjusted p values > 0.05 (horizontal) and |log2FC|> 1 (vertical). G BP over-represented GO terms for cluster 4 genes, listing functions related to protein translation. H Volcano plot representing DEGs at 72 hpd (red) vs. 32 hpd (blue); green dashed lines correspond to adjusted p values > 0.05 (horizontal) and |log2FC|> 1 (vertical). hpd hours post deposition, DEG differentially expressed gene, TCA temporal clustering analysis, GO gene ontology, BP biological process, CPM counts per million If we focus on the changes occurring specifically at 72 hpd, the comparison of the DEGs between 72 and 32 hpd shows an upregulation at 72 hpd of genes related to cell, tissue, and anatomical structure development (e.g., cathepsin, procathepsin, basement membrane heparan sulfate proteogycan, lmnl). Particularly, genes related to muscle and pharynx morphogenesis (e.g., troponin-1, tropomyosin alpha chain 1, unc52/perlecan) and involved in neural adhesion and recognition (e.g., ncam1, ncam2, zig4) can be identified. Genes activated in response to bacteria (macrophage expressed gene 1) were also identified, suggesting the possibility that microbiota is present inside the cocoon (Fig. 3H; Additional file 6: Fig. S6 B). A previous study also reported bulk RNA-seq data at this stage for the sister species Schmidtea mediterranea [19]. In order to assess the comparability of both datasets we integrated DEGs mapped between the two sets into a Sankey visualization that represents the expression changes of each gene pair across the different developmental stages. After comparing the DEG upregulated at 72 vs. 32 hpd in our dataset with the genes peaking at S3 in [19], we have found shared GO reporting functions related to tissue and animal organ morphogenesis (including mesoderm, muscle, pharyngeal, gut, and neural), cell adhesion, and basement membrane organization (Additional file 7: Fig. S7 A–C), which is in concordance with our report for the 72 hpd. Overall, our results demonstrate the polarization of βCAT-1 in the embryo during the formation of the embryonic pharynx primordia. Additionally, we observe that the onset of tissue specification coincides with the activation of the βCAT-depending Wnt signaling, both in differentiating primary epidermal cells (32 hpd) and embryonic pharynx and gut cells (32–48 hpd). The transcriptomic profile indicates a transition in biosynthesis levels between 32 hpd (low levels) and 48–72 hpd (high levels). This event is followed by the specific activation at 72 hpd of genes related to the morphogenesis of the embryonic tissues that coincides with the first mitosis detected in the germ band. At stage 3 (3–4 dpd), extensive cell proliferation coincides with a shift in gene expression and the activation of adult patterning genes During 4 dpd (end of stage 3), the embryo finishes ingesting yolk cells and has grown in size. Immunostaining shows an increase in the number of cells in the germ band with respect to the previous stage (Fig. 4A, B). βCAT-1 is consistently found in the nucleus in cells of the embryonic pharynx at this stage, but in contrast to the findings at 72 hpd, it starts to localize in the nucleus of blastomeres in the germ band (Fig. 4C, D, G; Additional file 4: Fig. S4 G, H; Table 4 [38]). Several mitoses are detected thanks to the anti-TYR-TUB labeling. Interestingly, βCAT-1 was found to be shared both symmetric and asymmetrical between daughter cells (Fig. 4E, F, H; Table 4 [38]).Fig. 4At 4 dpd massive cell proliferation coincides with the turning on of adult patterning genes. A Schematic representation of a planarian embryo at 4 dpd. B Whole mount immunostaining of a planarian embryo at 4 dpd with anti-TYR-TUB (magenta) and anti βCAT-1 (green). Nucleus are stained with DAPI (blue). White dotted line highlights the embryonic mouth (n = 8 embryos). C TYR-TUB (magenta) and βCAT-1 (green) immunostaining on a paraffin section of a planarian embryo at 4 dpd, showing the cells in the germ band (gb) close to the embryonic pharynx (ep). Nucleus are stained with DAPI (blue) (D). TYR-TUB (magenta) and βCAT-1 (green) immunostaining on whole-mount planarian embryo at 4 dpd, showing germ band blastomeres with nuclear βCAT-1. Nucleus are stained with DAPI (blue). E Magnification of a mitotic cell in the germ band showing asymmetric distribution of βCAT-1. F Mitotic cell in the germ band showing symmetric distribution of βCAT-1. G Percentage of cells in the germ band that present nuclear βCAT-1 staining at 4 dpd (n = 4) and 5 dpd (n = 5). Unpaired t test p value 0.0002. H Percentage of dividing cells in the germ band (stained with Tyr-Tub) at 4 dpd (n = 4) and 5 dpd (n = 5). Unpaired t test p value 0.013. I Cluster 6 from TCA analysis, grouping genes specifically up-regulated at 4 dpd; each line corresponds to the expression pattern for one gene in the cluster, while color gradient is proportional to the membership score. J BP over-represented GO terms for cluster 6 genes highlighting processes related to DNA replication, DNA compaction, cell division, and DNA repair. Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in up-regulated genes. K Whole mount immunostaining of planarian embryo at 4 dpd with DAPI (gray) which shows the chromosome compaction at the level of the germ band (red arrowheads). L Enriched KEGG pathways for upregulated DEGs at 4 dpd vs. 32 hpd, accounting for functions related to cell cycle, and molecular pathways involved in patterning the adult planarian. dpd: days post deposition. DEG differentially expressed gene, TCA temporal clustering analysis, GO gene ontology, BP biological process, KEGG Kyoto Encyclopedia of Genes and Genomes. Scale bars: (B): 80 µm, (C): 40 µm, (D): 30 µm, (E, F; I): 10 µm At 4 dpd massive cell proliferation coincides with the turning on of adult patterning genes. A Schematic representation of a planarian embryo at 4 dpd. B Whole mount immunostaining of a planarian embryo at 4 dpd with anti-TYR-TUB (magenta) and anti βCAT-1 (green). Nucleus are stained with DAPI (blue). White dotted line highlights the embryonic mouth (n = 8 embryos). C TYR-TUB (magenta) and βCAT-1 (green) immunostaining on a paraffin section of a planarian embryo at 4 dpd, showing the cells in the germ band (gb) close to the embryonic pharynx (ep). Nucleus are stained with DAPI (blue) (D). TYR-TUB (magenta) and βCAT-1 (green) immunostaining on whole-mount planarian embryo at 4 dpd, showing germ band blastomeres with nuclear βCAT-1. Nucleus are stained with DAPI (blue). E Magnification of a mitotic cell in the germ band showing asymmetric distribution of βCAT-1. F Mitotic cell in the germ band showing symmetric distribution of βCAT-1. G Percentage of cells in the germ band that present nuclear βCAT-1 staining at 4 dpd (n = 4) and 5 dpd (n = 5). Unpaired t test p value 0.0002. H Percentage of dividing cells in the germ band (stained with Tyr-Tub) at 4 dpd (n = 4) and 5 dpd (n = 5). Unpaired t test p value 0.013. I Cluster 6 from TCA analysis, grouping genes specifically up-regulated at 4 dpd; each line corresponds to the expression pattern for one gene in the cluster, while color gradient is proportional to the membership score. J BP over-represented GO terms for cluster 6 genes highlighting processes related to DNA replication, DNA compaction, cell division, and DNA repair. Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in up-regulated genes. K Whole mount immunostaining of planarian embryo at 4 dpd with DAPI (gray) which shows the chromosome compaction at the level of the germ band (red arrowheads). L Enriched KEGG pathways for upregulated DEGs at 4 dpd vs. 32 hpd, accounting for functions related to cell cycle, and molecular pathways involved in patterning the adult planarian. dpd: days post deposition. DEG differentially expressed gene, TCA temporal clustering analysis, GO gene ontology, BP biological process, KEGG Kyoto Encyclopedia of Genes and Genomes. Scale bars: (B): 80 µm, (C): 40 µm, (D): 30 µm, (E, F; I): 10 µm The heatmap of all the DEGs comparing all samples vs. 4 hpd shows a dramatic transcriptional change in the transition from 72 hpd to 4 dpd (Additional file 6: Fig. S6 B). The multidimensional scaling (MDS) plot of the RNA-seq also supports a temporal split between 72 hpd and 4 dpd (Additional file 2: Fig. S2 A, B). Supporting these transcriptional changes, two TCA clusters, clusters 6 and 3, corresponding to genes increasing or decreasing, respectively, from 4 dpd onwards, were found (Additional file 1: Fig. S1 G; O). Thus, the transition from 72 hpd to 4 dpd appears as a turning point regarding gene expression. GO analysis of the genes from cluster 6 corresponds to functions related to DNA replication, DNA repair, chromosome compaction, and G2/M transition (Fig. 4I, J), which agrees with the high mitotic activity observed. These results are supported by the nuclear compaction with visible chromosomes observed after DNA staining in blastomeres of the germ band (Fig. 4K). Shared GO terms from the comparison between the up-regulated DEGs at 4 dpd vs. 72 hpd with the genes peaking at S4 in [19] dataset, also reported functions related to DNA replication and chromatin compaction (Additional file 7: Fig. S7 D, E), further reinforcing this message. Cluster 3, grouping the genes decreasing from 4 dpd onwards, corresponds to functions related to intracellular protein transport, endocytosis, and exocytosis, which were found upregulated at stage 1 (Additional file 1: Fig. S1 G, I). GO and KEGG analyses of the upregulated DEGs at 4 dpd versus 72 hpd also show gene functions related to cell proliferation and DNA replication (Fig. 4L; Additional file 5: Fig. S5 C). Interestingly, the KEGG analysis shows the upregulation of intercellular communication pathways at this stage, such as the Hedgehog, Wnt, and Notch (Fig. 4L). In agreement, genes involved in anteroposterior (sfrp-1 and ndk) and dorsoventral (ngl-1 and ngl-7) patterning of adult planarians are found among the DEGs at 4 dpd vs. 72 hpd in the RNA-seq (Table 1) [38]. Those results indicate that at 4 dpd the program for patterning the definitive adult tissues is turning on. Our data shows that during germ band expansion at 4 dpd (end of stage 3), a massive cell proliferation concomitant to a shift in gene expression occurs. This result, together with the dynamic expression of βCAT-1 revealing asymmetric cell divisions and the onset of adult patterning gene expression, indicates that a process of cell specification of the adult tissues begins before the dorsoventral flattening of the embryo. A second event of tissue morphogenesis takes place at stage 5 (5–7 dpd) At 5 dpd, the embryo undergoes dorsoventral flattening, losing the radial symmetry (Fig. 5A, B). Immunostaining shows that βCAT-1 is extensively found in the nucleus of the germ band blastomeres in comparison with the previous stage (Figs. 4G; 5B, C; Table 4 [38]). Symmetric and asymmetric cell divisions in terms of βCAT-1 expression are also found at this stage, although the ratio of mitotic cells is lower than in 4 dpd embryos (Fig. 4H; Table 4 [38]). The chromatin decompaction observed with DAPI in cells that also show high levels of nuclear βCAT-1 (Fig. 5C) indicates an active transcriptional activity of blastomeres, which is βCAT-1-dependent, as described in adult planarians [36]. Interestingly, coinciding with the dorsoventral flattening, blastomeres presenting high levels of nuclear βCAT-1 accumulate on one side of the embryo (Additional file 4: Fig. S4 I, J).Fig. 5A second event of tissue morphogenesis takes place at stage 5. A Schematic representation of a planarian embryo at 5 dpd. B Whole mount immunostaining of a planarian embryo at 5 dpd with anti-TYR-TUB (magenta) and anti-βCAT-1 (green). Nucleus are stained with DAPI (blue) (n = 7 embryos). C Magnification of the germ band in (C) showing the DNA decompaction (weak DNA staining) in cells with high levels of nuclear βCAT-1 (green). D Schematic representation of a planarian embryo at 7 dpd. E Whole mount immunostaining of a planarian embryo at 7 dpd with anti-TYR-TUB (magenta) and anti-βCAT-1 (green). Nucleus are stained with DAPI (blue). ep, embryonic pharynx. F Magnification of the germ band in (H) showing differentiating cells with nuclear βCAT-1 staining; red arrow: probably a muscle cell; yellow arrow: probably a neuron. G Cluster 2 for TCA analysis, grouping genes specifically up-regulated at 7 dpd; each line corresponds to the expression pattern for one gene in the cluster, while color gradient is proportional to the membership score. H BP over-represented GO terms for cluster 2 genes include processes related to cell adhesion, neuromuscular differentiation, and eye morphogenesis. Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in upregulated genes. I Heatmap of genes involved in planarian neural and muscle regeneration, which are specifically turned on between 4 and 7 dpd. The heatmap color gradient represents log2(CPM). J Heatmap of genes involved in dorso-ventral, medio-lateral and anterio-posterior axis specification during planarian regeneration, which are specifically turned on between 4 and 7 dpd. The heatmap color gradient represents log2(CPM). K In situ hybridization to detect Spol-slit at 5, 7, and 14 dpd. L In situ hybridization to detect Spol-slit at 5, 7, and 14 dpd. dpd: days post deposition. DEG differentially expressed gene, TCA temporal clustering analysis, CPM counts per million, GO gene ontology, BP biological process, ep embryonic pharynx. Scale bars: (B; E): 15 µm, (C; F): 10 µm, (K, L): 250 µm A second event of tissue morphogenesis takes place at stage 5. A Schematic representation of a planarian embryo at 5 dpd. B Whole mount immunostaining of a planarian embryo at 5 dpd with anti-TYR-TUB (magenta) and anti-βCAT-1 (green). Nucleus are stained with DAPI (blue) (n = 7 embryos). C Magnification of the germ band in (C) showing the DNA decompaction (weak DNA staining) in cells with high levels of nuclear βCAT-1 (green). D Schematic representation of a planarian embryo at 7 dpd. E Whole mount immunostaining of a planarian embryo at 7 dpd with anti-TYR-TUB (magenta) and anti-βCAT-1 (green). Nucleus are stained with DAPI (blue). ep, embryonic pharynx. F Magnification of the germ band in (H) showing differentiating cells with nuclear βCAT-1 staining; red arrow: probably a muscle cell; yellow arrow: probably a neuron. G Cluster 2 for TCA analysis, grouping genes specifically up-regulated at 7 dpd; each line corresponds to the expression pattern for one gene in the cluster, while color gradient is proportional to the membership score. H BP over-represented GO terms for cluster 2 genes include processes related to cell adhesion, neuromuscular differentiation, and eye morphogenesis. Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in upregulated genes. I Heatmap of genes involved in planarian neural and muscle regeneration, which are specifically turned on between 4 and 7 dpd. The heatmap color gradient represents log2(CPM). J Heatmap of genes involved in dorso-ventral, medio-lateral and anterio-posterior axis specification during planarian regeneration, which are specifically turned on between 4 and 7 dpd. The heatmap color gradient represents log2(CPM). K In situ hybridization to detect Spol-slit at 5, 7, and 14 dpd. L In situ hybridization to detect Spol-slit at 5, 7, and 14 dpd. dpd: days post deposition. DEG differentially expressed gene, TCA temporal clustering analysis, CPM counts per million, GO gene ontology, BP biological process, ep embryonic pharynx. Scale bars: (B; E): 15 µm, (C; F): 10 µm, (K, L): 250 µm At 7 dpd, the embryo finally shows a recognizable bilaterian body plan (Fig. 5D, E). Immunostaining reveals a highly dynamic expression of βCAT-1, which is not cytoplasmic anymore but nuclear, as found in adult planarians (Fig. 5F) [36]. Although double staining is required to assign a cell identity, we can identify elongated cells with a fusiform nucleus that are positive for βCAT-1, which could correspond to muscular cells, as well as cells with TYR-TUB + projections and nuclear βCAT-1 that could correspond to newly formed neurons (Fig. 5F). This observation agrees with the onset of neuromuscular differentiation at this stage reported in [9, 18, 44] and with the neuromuscular-related genes found in the DEGs at 7 dpd (Table 1) [38]. Thus, TCA analysis shows a cluster of genes (cluster 2) that specifically increase their expression at 7 dpd (Fig. 5G; Additional file 1: Fig. S1 D). GO analysis of this cluster indicates a role of those genes in cell adhesion (cadherins and protocadherins) and differentiation of neuromuscular tissues (Fig. 5H). Consistently, GO and KEGG analyses of upregulated genes at 7 dpd versus 4–5 dpd showed their relationship with cell adhesion and differentiation of neuromuscular tissues (Additional file 5: Fig. S5 F; Table 3 [38]). Through a heatmap analysis, we show how DEGs at 7 dpd involved in neural and muscle differentiation are turned on (Fig. 5I), some of them already being active at 4 dpd: tgs-1, pc-1, neuroD1, and hesl3 among neural genes and troponin, foxF-1 and myoD among muscular genes (Fig. 5I). Additional categories found over-expressed at 7 dpd include genes related to neoblasts maintenance, epidermal, excretory, and digestive systems differentiation, as well as eye morphogenesis (Additional file 5: Fig. S5 D; Additional file 6: Fig. S6 C). Our transcriptomic data also reveals the transcriptional up-regulation of genes required for anteroposterior, dorsoventral, and mediolateral axial patterning of adult planarians at 7 dpd (Additional file 5: Fig. S5 D). Heatmap analysis illustrates the turn-on of polarity genes at 7 dpd (Fig. 5J). Wnt signaling elements are activated at 7 dpd, although, as previously mentioned, some of them are already turned on at 4 dpd (lrp5/6, evi/wntless, sFRP1, frizzled 5/8–4, axinA, foxD, zicA, and APC) (Fig. 5J). Notably, most genes activated at 4 dpd are related to the inhibition of the Wnt pathway. The heatmap focusing on BMP and Slit/Wnt5 related elements shows its upregulation as well at 7 dpd (Fig. 5J). As observed with the Wnt signaling pathway, inhibitory elements such as ngl-7and ngl-1, are detected already at earlier stages. GO functions related to cellular signaling pathways, like the Wnt pathway, and several GOs related with neural differentiation are also found in common when comparing up-regulated DEGs at 7 dpd vs. 4dpd and genes peaking at S5 in [19], further emphasizing that a second event of tissue morphogenesis is occurring in the embryo at this time point (Additional file 7: Fig. S7F–G). Previous works demonstrated by in situ hybridization that most of the genes involved in axial patterning (Wnt and BMP pathway elements) are detected at late stage 5 (7dpd) [17]. However, anteroposterior axis elements, such as sFRP1 and wnt1, as well as the dorsal determinant bmp, are already expressed at 5 dpd in a polarized manner [17]. Here, we demonstrate that the mediolateral determinants slit and wnt5 are detected by in situ hybridization as soon as 7 dpd and show an expression pattern similar to the adult planarians (Fig. 5K, L). To note, slit, which specifies the midline [45], is detected already at 5 dpd, close to the embryonic pharynx (Fig. 5K). The ventral determinant ADMP is not detected at 7 dpd, although it shows the adult expression pattern at 14 dpd (Additional file 6: Fig. S6 D). Altogether, this data demonstrates that from 5 to 7 dpd a second event of tissue morphogenesis takes place, giving rise to the definitive body pattern and tissue types of the adult planarian. Discussion This study illuminates key aspects of planarian embryonic development that have remained poorly understood, largely due to the challenges in handling planarian cocoons. Below, we discuss how our findings enhance our understanding of planarian development. Cross signaling between yolk cells and blastomeres during cleavage Cleavage is one of the most intriguing aspects of planarian development. Spiralian development relies on maternal determinants and cell–cell interactions to specify the cell lineages and axial organizers during cleavage [46]. In planarians, the fertilized zygote divides into daughter blastomeres that do not retain cell–cell contacts (“Blastomerenanarchie”). It was proposed that maternal determinants present in yolk cells could signal the first steps of planarian development since vitellocytes of abortive cocoons do not fuse to form the syncytium [9]. Furthermore, it has been shown that planarian yolk cells are transcriptionally active [10]. Our data supports the signaling role of yolk cells during cleavage since (1) we observe βCAT-1 in the nucleus of yolk cells surrounding the syncytium from the earlier stage analyzed (4 hpd), and (2) the transcriptomic profiling reveals ER and Golgi active cell secretion, which has a key role connecting yolk components and the developing embryo [47, 48]. This last point is in agreement with the electron microscope images reported in [10] at the level of the yolk cells, which show an accumulation of mitochondria and ER surrounding the nuclei of yolk cells. Few studies focus on the role of the canonical Wnt/βcatenin pathway in yolk cells. In zebrafish, a role of maternal βcatenin in the yolk cells during the formation of the early organizer was demonstrated [49]. Our results favor the signaling role of the yolk cells, which could present maternal components, such as βCAT-1, that organize the early cleavage in planarians. However, planarian embryogenesis occurs within a polyembryonic cocoon. Thus, a gradient of maternally inherited mRNAs or proteins with respect to a zygote is not probable. An alternative is that βCAT-1 + yolk cells could be signaling other yolk cells to form the syncytium, which is a requirement for a successful cleavage. Simultaneously, a differential distribution of mRNAs could cause the first asymmetry, as proposed by [17]. βCAT-1 nuclear localization in yolk cells is maintained at 24 hpd, when it starts to be expressed in the cytoplasm of blastomeres, which agrees with in situ hybridization results demonstrating βcat-1 mRNA expression in the cleaving blastomeres at 24 hpd [17]. In the same study, a qPCR analysis showed high levels of βcat-1 mRNA expression at 0 hpd, drastically decreasing at 24 hpd. This result supports the hypothesis that nuclear βCAT-1 in yolk cells could be translated from maternal mRNA, while βCAT-1 from the blastomeres could be zygotic. The kinetics of protein synthesis and gene expression, which specifically drops at 32 hpd (cluster 1 of the TCA), agrees with the possibility that maternal mRNAs present in the yolk are translated during the early stages. Symmetry breaking and germ layer specification in planarian embryos A central question in planarian embryology is how the first asymmetry, essential for pattern formation of the yolk-feeding embryo, arises in a system where blastomeres remain unattached during cleavage. In body patterning, symmetry breaking is understood as the initial event where a significant difference emerges between cells or regions of the embryo [50, 51]. According to our findings, the symmetry-breaking event in the planarian embryo occurs at 24 hpd, when the first blastomeres with nuclear βCAT-1 are observed, showing an elongated shape and located away from the remaining ones. This observation, together with the up-regulation at 24 hpd of genes related to cell migration and motility and the expression of snail in blastomeres in the periphery of the syncytium [17, 52], suggest that these blastomeres are migrating from the cleavage area. Although differentiated primary epidermal cells do not express βCAT-1, some elongated cells underneath the hull membrane show nuclear βCAT-1, supporting the hypothesis that these first migrating blastomeres could give rise to the embryonic epidermis. Nuclear localization of βcatenin in a few cells of the cleaving embryo is a conserved sign of gastrulation, which specifies the blastopore and the site of endomesoderm specification [27, 53–56]. During the stereotypical spiral cleavage of most Lophotrocozoans, descendants of the D quadrant blastomere act as the embryonic organizer, controlling axial identities [46]. However, planarian blastomeres lack cell–cell contacts during cleavage, and quadrants cannot be identified. Morphologically, the first observed asymmetry occurs during the formation of the primary tissues of the embryo with the appearance of the pharynx primordia at one pole, which establishes radial symmetry in the embryo and is the first to show cell–cell contacts. βCAT-1 nuclear localization, an evolutionarily conserved mechanism to specify the primary axis, suggests that it also provides the signals to confer polarity in planarians. Not only nuclear βCAT-1 but also βcat-1 mRNA expression and endomesodermal markers (foxA, twist) have been reported surrounding the pharynx cavitation between 32 and 48 hpd [17]. This data suggests that the primordium of the pharynx at 32 hpd could correspond to the specification of the endomesoderm of the planarian embryo and a particular mode of gastrulation. Thus, in planarians, βCAT-1 nuclear localization could represent a conserved mechanism for endomesoderm specification and to define the primary axis [53, 54]. However, the earliest βCAT-1 nuclear localization is observed in migrating blastomeres that likely give rise to the transient epidermis. This suggests that the initial βCAT-1 localization in the nucleus may play a role in promoting cell polarity and migration, potentially as an adaptation to the unique dispersed cleavage of planarians. Although our data support the presence of a mode of gastrulation in planarian embryos, ectodermal derivatives have not been identified. Some authors have described the primary embryonic epidermis as the ectodermal layer in planarians [10]. However, we observed βCAT-1 nuclear localization in precursors of the embryonic epidermal cells, and data from several model organisms indicate that β-catenin is not involved in, and may be incompatible with, ectodermal specification [55, 57, 58]. Some evidence suggests that the formation of derivatives of the three germ layers occurs during the formation of the pharynx primordium at 32 hpd. Thus, the embryonic pharynx is surrounded by epidermal cells [10]; and some neural activity is required for muscle contraction [18]. In support of the presence of the ectoderm, we detect neural differentiation genes upregulated at 48–72 hpd, while the embryonic pharynx is being developed. Moreover, the upregulation at 32 hpd of transcripts related to sphingolipids, which are essential for gastrulation in mice [59] and for gut phagocyte differentiation in planarians [60], further supports a process of gastrulation and germ layer segregation at this stage. Nuclear localization of βCAT-1 in 32 hpd planarian embryos is not likely to depend on a Wnt signal, since no Wnt ligand has been detected in the pharynx primordia. Particularly, Wnt1 (the canonical Wnt ligand in adult planarians) is not expressed in the pharynx primordia but in migrating blastomeres [17]. Therefore, we speculate that the βCAT-1 nuclear localization may result from the new cell–cell contacts at this stage. In several models, the mechanical forces exerted in the blastopore maintain the Wnt/βcatenin pathway activated at the site of blastopore formation [27]. It is possible that the proximity between these blastomeres is enough to self-organize themselves, promoting a gastrulation event and endomesoderm progeny, as described in in-vivo and in-vitro models [61–63]. Patterning during planarian embryogenesis: two morphogenetic events giving rise to two body plans The shift in the transcriptomic profile observed between 72 hpd and 4 dpd coincides with the two different events of tissue morphogenesis that must occur during the formation of the yolk-feeding embryo and the adult body plan. During this interval, the embryo feeds the yolk, and right after, embryonic tissues start their disassembly, which occurs concomitantly to the differentiation of the germ band into the definitive tissues [10]. While the mechanisms of planarian adult patterning and morphogenesis have been extensively studied [64–67], the one of the yolk-feeding embryo is largely unknown. The Wnt/βcatenin pathway is activated during the formation of the pharynx primordia (32 hpd), suggesting its conserved role in endomesoderm specification as discussed in the previous section, and remains activated in the differentiated embryonic pharynx and gut. The transcriptomic data shows the activation of neural and mesodermal development at 48–72 hpd, when the embryo is formed and feeds on the yolk. Our results also show that during and after ingesting the yolk (72 hpd–4 dpd), the cells from the germ band start to proliferate massively, as previously reported through pH3 immunodetection [17]. At 72 hpd, βCAT-1 is in the cytoplasm and is evenly distributed between daughter cells. However, at 4 dpd, βCAT-1 is found in the nucleus in some blastomeres, and its asymmetric distribution is observed. This result could indicate that blastomeres divide to make the embryo grow during feeding, and βCAT-1 accumulates. Later on, after yolk-ingestion, the asymmetric nuclear localization of βCAT-1 coincides with the triggering of the differentiation program corresponding to the adult. The complex regulation of βCAT-1 expression, coupled with the activation of cell differentiation genes, demonstrates the essential role that βCAT-1 has during the specification of the adult cellular lineages, in agreement with the results found in adults [36]. At 4 dpd, genes related to adult axial patterning appear upregulated in the RNA-seq (sFRP1, ngl-1, ngl-7). In previous studies, the earliest expression of patterning genes was detected at 5 dpd by in situ hybridization. Thus, our transcriptomic analysis has provided a more sensitive description of the gene expression occurring during planarian embryogenesis. At 5–7 dpd, the dorsoventral flattening of the embryo starts, concomitant to the overexpression of cell differentiation and patterning genes, as previously reported [17, 19]. A massive βCAT-1 nuclear localization and chromatin decompaction is also observed, suggesting a general cellular reprogramming event. Interestingly, during the dorsoventral flattening, there is an accumulation of blastomeres, most showing high levels of nuclear βCAT-1 at one-half of the embryo, indicating the ongoing establishment of the adult body axis. We speculate that the site where blastomeres accumulate corresponds to the ventral side of the developing juvenile, where neuromuscular progenitors are starting to differentiate [18] since we identify the differentiation of putative muscle fibers and neurons with nuclear βCAT-1 at 7 dpd. Furthermore, this accumulation of βCAT-1 positive cells in the ventral part is also observed in adults [36]. At 7 dpd, a strong up-regulation of genes related to tissue morphogenesis and cell adhesion indicates that the outline of the adult is being established [17, 19]. Importantly, our observations detect βCAT-1 nuclear localization in the embryonic pharynx from the 48 hpd until the 4 dpd when the yolk engulfment finishes. This could indicate that the embryonic pharynx can serve, between 4 and 7 dpd, as an anatomical reference to pattern the adult tissues. The first detection of the morphogen slit surrounding the embryonic pharynx at 5 dpd further supports this hypothesis. Our transcriptomic results, showing a shift in gene expression between 72 hpd and 4 dpd, coinciding with the transition from the yolk-feeding embryo to the adult body plan, together with the bi-phasic expression of signaling genes and tissue-determinants during both morphogenetic events, favor the hypothesis of an indirect mode of development. Previous histological studies provided direct evidence of the degeneration of the muscle embryonic pharynx, which is being replaced by the definitive during stage 5 (5–6 dpd) and the replacement of the primary epidermis by the definitive one at the same stage [44]. Additional evidence supporting a process of metamorphosis can be found in the analysis of the embryonic origin of neoblasts. It has been demonstrated that PIWI-1 + cells found before and after stage 5 have different morphological and molecular signatures, and that only the second ones are able to give rise to adult tissues after transplantation in irradiated hosts [19]. Our results, together with previous publications, support the indirect development of planarians with two different body plans sustained by different stem cells. Conclusions Planarian embryonic development has been previously classified into eight developmental stages. Due to the difficulties associated with their manipulation, most studies have focused on stages 5 and beyond, when the adult body plan is being established. In this study, we have examined and dissected the cellular and genetic events occurring from stage 1 to 5, corresponding to the cleavage and early patterning of planarian embryos, which is crucial for understanding the evolution of their unique developmental mode among Spiralians. Our data reveals a possible signaling role for the yolk cells during the dispersed cleavage of planarians. It also suggests that the βCAT-1-dependent Wnt signaling is involved in specifying the three embryonic tissues: the primary epidermis, the embryonic gut, and the pharynx. This challenges the traditional view of the primary epidermis as an ectodermal derivative and supports the hypothesis that the embryonic primordium is the site of endomesoderm specification, where a gastrulation-like event occurs. Furthermore, our observations uncover two distinct morphogenetic events, one giving rise to the yolk-feeding embryo and the other forming the adult planarian (Fig. 6).Fig. 6Schematic representation of the main events during planarian early embryogenesis with the contribution of the findings presented in this work. Yolk cells could have a role in signaling during cleavage (4 hpd). The first asymmetry is observed at 24 hpd, when migrating blastomeres show nuclear βCAT-1. They could be the precursors of the primary epidermis. At 32 hpd, the first cell–cell contacts and βCAT-1 nuclear localization to give rise to the pharynx and gut embryonic tissues, together with the decrease in biosynthesis suggests that gastrulation (endomesoderm specification) is occurring. From 48 hpd to 4 dpd, the first event of tissue morphogenesis takes place, when the embryonic structures are being differentiated and the embryo gets ready to engulf the yolk. From 4 to 5 dpd, high cell proliferation coincides with the disassembling of the embryonic tissues and the starting of the adult cell differentiation. From 5 dpd onwards, germ band cells differentiate, adult patterning genes are expressed, and the second event of morphogenesis to give rise to the adult body plan takes place Schematic representation of the main events during planarian early embryogenesis with the contribution of the findings presented in this work. Yolk cells could have a role in signaling during cleavage (4 hpd). The first asymmetry is observed at 24 hpd, when migrating blastomeres show nuclear βCAT-1. They could be the precursors of the primary epidermis. At 32 hpd, the first cell–cell contacts and βCAT-1 nuclear localization to give rise to the pharynx and gut embryonic tissues, together with the decrease in biosynthesis suggests that gastrulation (endomesoderm specification) is occurring. From 48 hpd to 4 dpd, the first event of tissue morphogenesis takes place, when the embryonic structures are being differentiated and the embryo gets ready to engulf the yolk. From 4 to 5 dpd, high cell proliferation coincides with the disassembling of the embryonic tissues and the starting of the adult cell differentiation. From 5 dpd onwards, germ band cells differentiate, adult patterning genes are expressed, and the second event of morphogenesis to give rise to the adult body plan takes place Our results are significant for understanding the evolutionary history of Tricladida and the unique plasticity of the adult worm. Recent studies indicate that a feature of planarian regeneration is that neoblasts and progenitor cells remain intermingled during their specification [68]. The intermingled manner in which the blastomeres of the germ band are specified during stage 4 may have a relationship with the pluripotency of planarian adult stem cells. Methods Animals maintenance A sexual population of Schmidtea polychroa was collected from Sant Celoni, Barcelona, Spain, and maintained in the lab as described by [69]. Planarians were kept at 20 °C under a day–night light cycle (16 h of light followed by 8 h of darkness) in glass aquariums in a 1:1 (v:v) mixture of distilled water and tap water. The tanks had a surface of 1800 cm2, and a water depth of 11 cm. The animals were fed twice a week with sliced bovine liver. Egg capsules collection and fixation Egg collection and fixation was conducted as described in [9, 69]. During the breeding period, egg capsules were collected and placed in Petri dishes, where they hatched after approximately 14–20 days. If no visible signs of development (the boundaries between the different embryos when the egg capsule is observed under transmitted light) were present after 3 weeks, the egg was considered to be non-viable. Eggs were collected right after deposition (orange color) and were placed in 24-well plates in order to be able to process each cocoon at a specific hour or day post-deposition. Plates were maintained at 20 °C. Immunohistochemistry Embryos at different developmental stages were dissected as previously described by [70]. Briefly, planarian egg capsules were laid on a glass slide, perforated using a 0.5-mm needle and immediately fixed in 4% paraformaldehyde at 4 °C for 4 h. After an overnight wash in PBS at 4 °C, egg capsules were stored at 4 °C in 70% alcohol. When needed, they were rehydrated with PBT (0.1% Triton X-100 in PBS) and placed in an agar-coated petri dish to be dissected over an agar plate under the scope with the help of tweezers. At this point, embryos can either be stored, after proper dehydration, in 100% ethanol at – 20 °C, embedded in paraffin, or used directly for whole mount in immunostaining or situ hybridization. Paraffin embedding and sectioning were conducted as previously described by [70]. Immunohistochemistry of whole mount samples and paraffin sections were conducted as previously described by [36]. Primary antibodies were used at the following dilutions: anti-βcatenin-1, 1:200 [36], anti-tyr-tubulin 1:400 (Sigma). Secondary antibodies were used at the following dilutions: anti-rabbit HRP-conjugated antibody 1:500 (Pierce); anti-mouse Alexa Fluor 568-conjugated antibody, 1:1000 (Molecular probes). In anti-βcatenin-1 immunostainings, HRP signal was developed with a tyramide signal amplification kit following manufacturer’s recommendations (Perkin Elmer). Nuclei were stained with DAPI (Sigma). Embryos were mounted and stored in 80% glycerol. Whole mount in situ hybridization Embryos were dissected described in the above section and the whole mount in situ hybridization protocol was performed as described by [70]. After dissection, embryos were permeabilized with proteinase K (Sigma; 10 µg/ml in PBT for 5 min at room temperature) and treated with triethanolamine (TEA, 0.1 M pH 7.6, Fluka; 2 × 5 min and 2 × 5 min following addition of 2.5 µl/ml of acetic anhydride without changing the TEA solution) to block positive charges. Following a PBT wash, embryos were post-fixed in 4% formaldehyde in PBT for 20 min at room temperature and washed again in PBT (at least five changes of 5 min each). Pre-hybridization and hybridization steps were carried out at 56 °C in a dry oven. Antisense and sense digoxigenin-labeled riboprobes were synthesized using an RNA in vitro transcription kit (Takara), and column purified (Roche). Probes were then diluted to a final 1:100–1:200 ratio in the hybridization solution. After hybridization, probes were recovered, and samples were washed at 56 °C (3 × 20 min with 2 × SSC; 2 × 20 min with 0.2 × SSC). After two washes in maleic buffer (MAB; 100 mM maleic acid, 150 mM NaCl, 0.1% Tween 20, pH to 7.5 with NaOH) embryos were blocked for 2 h in blocking solution (1% Blocking reagent (Roche) with 5% heat inactivated fetal calf serum (Invitrogen)). Anti-digoxigenin AP conjugated antibody (Roche) at a final 1:2000 dilution in blocking solution was incubated for 4 h at room temperature. After overnight washes, samples were transferred to AP buffer and developed with BM Purple (Roche) until a signal was visible. The reaction was stopped in MAB and embryos were dehydrated and kept in 100% ethanol until the desired signal-to-noise ratio was reached. After rehydration, embryos were fixed in 4% formaldehyde in PBT and stored in glycerol:PBS, 1:1, at − 20 °C. Embryos were mounted and stored in 70% glycerol. Image analysis Brightfield colorimetric in situ hybridization images were observed under a Leica MZ16F stereomicroscope (Leica Microsystems, Mannheim, BW, Germany) and images were captured with a ProgRes C3 camera from Jenoptik (Jena, TH, Germany). A Leica TCS-SP5 confocal laser-scanning microscope (Leica Lasertchnik, Heidelberg, BW, Germany) or a Zeiss LSM 880 confocal microscope (Zeiss, Oberkochen, Germany) was used to obtain confocal images of paraffin sections and whole-mount immunostainings. Fiji/ImageJ was used to show representative confocal stacks for each experimental condition [71]. RNA-sequencing sample preparation and analysis RNA-sequencing samples were obtained with the following protocol. At each corresponding time point (4 h, 24 h, 32 h, 48 h, 72 h, 96 h, 120 h, and 168 h), egg capsules were placed in a microscope concave glass slide and were treated with 1% diluted commercial NaOCl, for no more than 3 min in order to remove the mucus and parasites. Then the cocoons were moved to an Eppendorf tube, washed with RNAlater (Invitrogen) and 50% RNAlater/Trizol. Finally, liquids were removed and 100 μl of Trizol reagent (Invitrogen) was added. The whole cocoons were triturated with a tissueLyser (Kontes Microtube Pellet Pestle Rods with Motor), and total mRNA extraction was performed as described in [72]. Three biological replicates were analyzed per time point. Each biological replicate contained 4 whole cocoons. Planarian early embryos are not synchronous, and each sample was identified by the hour or days post deposition (hpd or dpd) in which they were collected. Library preparation and sequencing were carried out by the Centre Nacional d’Anàlisi Genòmica (CNAG). TruSeq stranded mRNA-seq libraries were sequenced on HiSeq 2000 (Illumina) with a read length of 2 × 100 bp. Sequencing run produced 1,763,225,906 raw reads totaling 178,085,816,506 bp; those reads were cleaned with Trimmomatic version 0.32 [73], resulting in 1,679,606,938 clean reads longer than 30 bp summing up 166,255,109,697 bp. Basic quality control on raw and clean reads was done by FastQC version 0.11.2 [74]. RNA reads from all sample replicates were pooled to be de novo assembled into contigs using Trinity version 2.5.1 [75]; this yielded 1,554,770 contigs longer than 100 bp (mean length 264.6 bp, longest 19,880 bp). Other assembly approaches were considered, yet this one recovered more genes when assessing assembly completeness with BUSCO (version 3.0.2) [76, 77] [910 complete, 49 fragmented, and 19 missing from 978 candidate BUSCOs; see Additional File 2: Fig. S2 C]. Nucleotide NCBI-BLAST (megablast, e value < 10e-25 [78] against RefSeq genomes database [79] returned 42,271 contigs mapping to other species than Tricladida (i.e., human, bacteria, etc.); those alignments had more than 90% coverage and > 90% identity, therefore those contigs were discarded as contaminants. Moreover, only contigs longer than 1000 bp were considered for further downstream analyses, totaling 62,218 transcript sequences (mean length 1843 bp) assigned by Trinity to 23,300 genes. After contamination screen when uploading to NCBI-TSA repository with their internal pipeline [80], 1297 sequences were discarded and 60,921 transcripts were kept (mean length 1847 bp, summing up 112,526,258 bp). From now on, we will refer to this assembly as “T1kcln,” submitted as NCBI-TSA GLGS01000000 (https://identifiers.org/nucleotide:GLGS01000000) [81]. Isoform expression levels were estimated using RSEM algorithm [82] following the integrated Trinity procedures [83]; such procedure implied a Trimmed Mean of M-values (TMM) normalization by library size to produce the counts' matrix for expression analysis [84]. All downstream analyses were performed using R version 4.2.2 packages. Differential gene expression analysis Only those transcript isoforms having at least five counts per million (CPMs) on at least three replicates were kept undergoing differential gene expression analyses; this left 21,060 transcripts out of 60,921 from previous filtering steps. DGEList object was created from counts’ matrix using edgeR package (version 3.40.2) [85] yet differentially expressed genes were identified employing the lima-voom pipeline (version 3.54.2) [86], considering an FDR adjusted p-value cut-off of 0.05 and a log fold change cut-off of ± 1. All 28-time pair-wise distinct combinations were considered in the contrast design, focusing on the comparisons of each time point samples relative to initial time (ti versus t0, where t0 = 4hpd) and the pairs for a given time related to previous time (ti versus ti-1). Differential expression results over all transcript isoforms for those comparisons were provided as (Table 1) [38] and is summarized in the UpSetR (version 1.4.0) [87] plots on (Additional file 2: Fig. S2 D-E), for up and down differentially expressed genes. Full matrices for RSEM-raw and TMM-normalized counts are available from NCBI-GEO repository [88] under the series GSE296729 (https://identifiers.org/geo:GSE296729) [89]. Gene Ontology analysis MMseqs2 (release 12-113e3 + ds-3 + b1, [90] compared T1kcln transcript isoforms against all UniProt Swi + sProt and TREMBL protein sequence databases (local mirror last update on May 2025) [91], NCBI GenBank nucleotide sequences database nt [92], as well as the full set of transcriptomic sequences for dd_Spol_v3, dd_Smed_v6, and smesGENES_v2 (downloaded from PlanMine, [93]); 55,550 out of 60,921 transcripts had at least a match to a sequence from those datasets. Gene Ontology annotations from the aforementioned datasets were transferred based on those alignments for 37,173 transcripts. Analysis of over-represented Gene Ontology (GO) terms on sets of differentially expressed genes was performed using topGO (version 2.50.0) [94] functions; visualization was done with extended versions of this package plotting functions. For the functional comparison of T1k with [19], the same procedure was applied on those sequences from Schmidtea mediterranea, which were downloaded from GEO GSE72389 (https://identifiers.org/geo:GSE72389) [95, 96]. They were aligned via MMseqs2 to T1kcln transcript isoforms: 2063 were found only in Davies’ set, 11,659 only in T1kcln, and 49,262 T1kcln had a crossmatch. After mapping to the same reference sets, 14,437 transcripts of the Davies’ set were assigned to one or more GO term. The number of DGE transcripts at the different stages of both sets that were sharing same GO annotations for the DGE transcripts at the different stages, were then compared by means of circle-plots. Sankey plots visualization shown on (Additional file 7: Fig. S7) were generated with ggsankey R package [97], taking into account the correspondence between the two transcriptomes and the reported log fold-change to define the different expression categories (UP, DOWN, etc.), as well as the closest time-points in both experimental designs (Additional file 7: Fig. S7 A). Time course clustering analysis TCseq (version 1.22.6) [98] clustering method was used to characterize the expression patterns through time of total genes. The optimal number of clusters was set to k = 6 after exploring the different clusters obtained from k = 2 to k = 20; the choice was made considering the robustness of each new cluster based on membership score calculated by TCseq, for all clusters in the set having a membership score greater than 0.25. GO over-representation analysis was undertaken on each cluster with total genes, as well as selecting differentially expressed genes on those clusters. clusterProfiler (version 4.6.2) [99] DOSE (version 3.24.2), and enrichPlot (version 1.18.4) [100] R packages were considered for this purpose. KEGG pathway analysis As there is no organism-level pathway set predefined for planarian in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database [101, 102], mapping transcript isoforms to KEGG pathways was done through the KEGG Orthology (KO) database. Longest open reading frames (ORFs) were translated from transcripts; but due to frameshifts from sequencing errors, nucleotide and indel variants, or assembly mismatches, they may not define the complete CDS region [103]. To overcome this limitation, a set of complete Platyhelminthes mRNAs was downloaded from NCBI Genbank (7187 sequences), in order to train coding and non-coding models for estscan (version 3.0.3) [104]. estscan produced a CDS prediction for 54,819 of 62,218 transcripts (before NCBI-TSA contamination screen), increasing average CDS length from 1199.84 bp (Additional file 2: Fig. S2 F) to 1422.75 bp (Additional file 2: Fig. S2 G) with respect to the full transcript average length of 1843.53 bp. The estscan translated CDSs were analyzed with KofamScan (version 1.3.0) [105] using the kofam HMM profiles downloaded from KEGG repository (https://www.genome.jp/ftp/db/kofam/); as a result, KOs were assigned to 15,326 transcripts out of 60,921 of the T1kcln set. KEGG pathways (ko) annotation was downloaded with download_KEGG function of clusterProfiler package and enrichKEGG function was employed to calculate enriched pathways with an FDR adjusted p value threshold of 0.05. Finally, significant pathways were fed to pathview (version 1.38.0) [106] to visualize enriched KEGG pathways identified by clusterProfiler. Those pathways pinpoint the KOs linked to Trinity transcripts that were overlaid on the corresponding diagram nodes, color scales being mapped to KO abundance or to the transcript log fold-change expression levels. Supplementary Information Additional file 1. Figure S1. Temporal Clustering Analysis (TCA) of the 6 optimal clusters from the RNA-seq data. (A; D; G, J; M; O) Cluster 1 to cluster 6 expression patterns (vertical axis) over time points (horizontal axis). Each line corresponds to the expression pattern for one gene in the cluster, its color being proportional to the cluster “membership” score (from magenta for genes leading to the cluster formation, higher membership score, to yellow for genes loosely related to the cluster, lower membership score). (B; E; H; K; N; P) CC over-represented GO terms for the genes from each cluster. (C; F; I; L; Q). MF over-represented GO terms for the genes from each cluster. TCA: Temporal Clustering Analysis. GO: Gene Ontology. CC: Cellular Compartment. MF: Molecular Function. Red arrows indicate the functions related to the biological processes highlighted in the different sections of the manuscript. No overrepresented MF GO terms were found for cluster 5 (M).Additional file 2. Figure S2. Complementary analyses of the RNA-seq data for the Material and Methods section. (A) Multidimensional scaling (MDS) plot of RNA-seq samples clustering across the two principal components, accumulating 26.75% and 10.81% of variance respectively. hpd: hours post deposition. dpd: days post deposition. Scale bar: 200μm. (B) Multidimensional scaling (MDS) plot of RNA-seq samples clustering after removing E048h_1 and E120h_3 replicates. (C) Comparison of BUSCO results for the transcriptome assembly completeness assessment. Ratio of complete (blue), partial (yellow) and missing (red) genes from the BUSCO orthologs set for the different assemblies considered prior to the differential gene expression analysis. Trinity assembly achieves the larger number of total genes found from BUSCO set, both for raw (all predicted contigs) and clean (cln, filtered contigs longer than 1kbp). (D-E) Summary of up (green) and down (red) regulated DEGs, respectively, for all the 28 time point pairwise comparisons. Numbers on top of the bars correspond to the total up/down DEGs found for each pairwise comparison. (F-G) Each Trinity assembled transcript contig (or isoform) was translated into the six frames to obtain the longest ORF (E), and it was also analyzed by estscan to predict the more likely CDS (F). Each transcript is shown as a dot, where X-axis has the total transcript length (bp) and Y-axis has that transcript longest ORF (E) or its CDS (F) lengths, both at nucleotide level (bp). Dashed diagonal lines define coverage thresholds; the longer the coding region encoded in the transcript is, the closer to the 100% diagonal should be. Density curves highlight what coverage ratio has more transcripts on it, where it is evident that the predicted estscan CDSs approach returns longer protein coding segments than the longest ORF method. DEG: Differentially Expressed Gene. ORF: Open Reading Frame. CDS: Coding DNA Sequence.Additional file 3. Figure S3. TYR-TUB and βCAT-1 staining of 0-24 hpd embryos and over-represented GO terms for up- and down-regulated DEGs at 24 vs 4 hpd. (A) Schematic representation of a planarian cocoon at 0-4 hpd. (B) Immunostaining on paraffin sections of a planarian whole cocoon at 0-4 hpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing the cleavage after deposition. Nucleus are stained with DAPI (Blue). (C) Schematic representation of a planarian embryo at 0-4 hpd. (D) Magnification of the cleaving blastomeres immunostained with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing that during the first cleavage division the blastomeres did not show βCAT-1 staining. Nucleus are stained with DAPI (Blue). (E) CC overrepresented GO terms for down-regulated DEGs at 24 hpd vs 4 hpd, showing genes involved in mitochondrial function, ERAD pathway (derlin-1 and fibronectin), Golgi trafficking and secretory vesicle production (purple arrows). Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in down-regulated genes. (F) BP over-represented GO terms for up-regulated DEGs at 24 hpd vs 4 hpd showing up-regulation in functions involved in cell migration among others (red arrows). Barplot color gradient as in Supp3.E, but relative to up-regulated genes. hpd: hours post deposition. GO: Gene Ontology. CC: Cellular Compartment. BP: Biological Process. Scale bar: (B): 200 μm (D): 10 μm.Additional file 4. Figure S4. βCAT-1 staining of stage 2, 3, 4 and 5 embryos (A) Schematic representation of a planarian embryo at 32 hpd. (B) Immunostaining on paraffin sections of a planarian embryo at 32 hpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing how primary epidermal differentiated cells do not present nuclear βCAT-1 (red arrowhead). Nucleus are stained with DAPI (Blue). (C) Schematic representation of a planarian embryo at 48 hpd. (D) Immunostaining on paraffin sections of a planarian embryo at 48 hpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing nuclear staining of βCATENIN-1 in the embryonic pharynx (red arrowheads). Nucleus are stained with DAPI (Blue). (E) Schematic representation of a planarian embryo at 72 hpd. (F). Immunostaining on paraffin sections of a planarian embryo at 72 hpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing cells in the germ band sharing equally βCAT-1. Nucleus are stained with DAPI (Blue). (G) Schematic representation of a planarian embryo at 4 dpd. (H) Whole mount Immunostaining of a planarian embryo at 4 dpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), in the cells of the germ band surrounding the embryonic pharynx showing nuclear βCAT-1. Nucleus are stained with DAPI (Blue). Red arrowhead points to an embryonic pharynx cell positive for nuclear βCAT-1 staining. (I) Schematic representation of a planarian embryo at 4 dpd. (J) TYR-TUB (magenta) and βCAT-1 (green) immunostaining on a paraffin section of a planarian embryo at 5 dpd, showing nuclear βCAT-1 mostly at one side of the embryo.Additional file 5. Figure S5. Over-represented GO terms for 32 vs 24 hpd, 72 vs 32 hpd, 4 dpd vs 72 hpd and 7 dpd vs 4 dpd. (A) BP over-represented GO terms for up-regulated DEGs at 32 hpd vs 24 hpd showing an up regulation in functions related to ubiquitination, lipid metabolism and negative regulation of autophagy (red arrows). Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in up-regulated genes. (B) BP over-represented GO terms for up-regulated DEGs at 72 hpd vs 32 hpd showing an up-regulation in functions related to tissue morphogenesis as cell adhesion, muscle, neural pharynx, and gut differentiation (red arrows). Barplot color gradient as in Figure S5 A. (C) BP over-represented GO terms for up-regulated DEGs at 4 dpd vs 72 hpd, which shows an up-regulation of processes related to DNA replication, cell division and patterning of the adult (dorso-ventral and anterio-posterior) (red arrows). Barplot color gradient as in Figure S5 A. (D) BP over-represented GO terms for up-regulated DEGs at 7 dpd vs 4 dpd, which shows an up-regulation of processes related to a second event of tissue morphogenesis: cell adhesion and differentiation of several adult tissues. Barplot color gradient as in Figure S5 A. hpd: hours post deposition. dpd: days post deposition. DEG: Differentially Expressed Gene. CPM: Counts Per Million. GO: Gene Ontology. BP: Biological Process.Additional File 6. Figure S6. Heatmap of autophagy related genes, consecutive time-points comparisons and adult tissue morphogenesis genes. admp expression pattern. (A) Heatmap of atg genes specifically shut down at 32 hpd. The heatmap color gradient represents log2 (CPM). (B) Heatmap of the consecutive timepoints versus 4 hpd comparisons showing clustering of two groups: samples comparisons between 24-72 hpd vs 4 hpd and 4-5-7 dpd vs 4 hpd. Arrow indicates the transition from 72h to 4 dpd (C) Heatmap of genes involved in maintenance of neoblasts, differentiation of the eyes, the epidermis, the excretory system, and the pharynx and the digestive system in adult animals, which are specifically turned on between 4 and 7 dpd. The heatmap color gradient represents log2(CPM). (D) In situ hybridization to detect Spol-ADMP in embryos at 5, 7 and 14 dpd. dpd: days post deposition. hpd: hours post deposition. dpd: days post deposition. Scale bar: (D) 15 μm.Additional file 7. Figure S7. Comparison to Davies et al 2017. (A) Sankey plot showing the distribution of genes on each of the 7 DEGs categories, which define the nodes, on every developmental stage being compared across time, which where interspersed according to the closest time point, between the two experiments (Y, S2, S3, S4, S5, S6, S7, and S8 for Davies et al; 24x4hpd, 32x24hpd, 48x32hpd, 72x48hpd, 96x72hpd, 120x96hpd, and 168x120hpd for Coronel-Córdoba et al). 4,001 unique DEGs of 4,918 from all stages reported in Davies et al (from total transcripts), were calculated relative to all time points average expression, whilst for Coronel-Córdoba they came from the previous time pair-wise comparison DEGs (1,310 transcripts of 21,006); this higher resolution for the later may explain the observed switch in upregulated genes at 72hpd, which was described at S2 for the Davies’ set. A total of 27,259 transcript contigs merged from both experiments’ assemblies were classified into the 7 expression categories, then those categories were pair-wise linked to define the internode flows between each developmental stage time point pairs; transcripts not mapped in both sets, or those for which no pre-computed expression data was available from Davies et al, were assigned to “NA” category. Significant DEGs are classified as “UP” or “DOWN”; those genes just passing one of the significance criteria are labeled as “up” or “down” (adjusted p-value = 0.05 or|log(fold-change)| = 1); non-significant genes are classified as non-significant up (“nsu”) or down (“nsd”). Number of genes assigned to each node/category in function of their adjusted p-value and log fold-change are shown below all node labels. (B) (D) (F) Zoom on three pair-wise time point comparisons between Davies and Coronel-Córdoba sets. Sankey-plots description as in (A). (C) (E) (G) Bubble-plots show the number of genes (proportional to circle size), classified as “up” or “UP” in both sets, that are annotated with the same BP GO, for each pair-wise comparison ((B) (D) (F) respectively). n accounts for the number of GOs shown in the plots; only GOs having at least one gene annotated in both sets were considered. Red arrow-heads indicate the functions related to the BP highlighted in the different sections of the manuscript. hpd: hours post deposition. DEG: Differentially Expressed Gene. GO: Gene Ontology. BP: Biological Process. Additional file 1. Figure S1. Temporal Clustering Analysis (TCA) of the 6 optimal clusters from the RNA-seq data. (A; D; G, J; M; O) Cluster 1 to cluster 6 expression patterns (vertical axis) over time points (horizontal axis). Each line corresponds to the expression pattern for one gene in the cluster, its color being proportional to the cluster “membership” score (from magenta for genes leading to the cluster formation, higher membership score, to yellow for genes loosely related to the cluster, lower membership score). (B; E; H; K; N; P) CC over-represented GO terms for the genes from each cluster. (C; F; I; L; Q). MF over-represented GO terms for the genes from each cluster. TCA: Temporal Clustering Analysis. GO: Gene Ontology. CC: Cellular Compartment. MF: Molecular Function. Red arrows indicate the functions related to the biological processes highlighted in the different sections of the manuscript. No overrepresented MF GO terms were found for cluster 5 (M). Additional file 2. Figure S2. Complementary analyses of the RNA-seq data for the Material and Methods section. (A) Multidimensional scaling (MDS) plot of RNA-seq samples clustering across the two principal components, accumulating 26.75% and 10.81% of variance respectively. hpd: hours post deposition. dpd: days post deposition. Scale bar: 200μm. (B) Multidimensional scaling (MDS) plot of RNA-seq samples clustering after removing E048h_1 and E120h_3 replicates. (C) Comparison of BUSCO results for the transcriptome assembly completeness assessment. Ratio of complete (blue), partial (yellow) and missing (red) genes from the BUSCO orthologs set for the different assemblies considered prior to the differential gene expression analysis. Trinity assembly achieves the larger number of total genes found from BUSCO set, both for raw (all predicted contigs) and clean (cln, filtered contigs longer than 1kbp). (D-E) Summary of up (green) and down (red) regulated DEGs, respectively, for all the 28 time point pairwise comparisons. Numbers on top of the bars correspond to the total up/down DEGs found for each pairwise comparison. (F-G) Each Trinity assembled transcript contig (or isoform) was translated into the six frames to obtain the longest ORF (E), and it was also analyzed by estscan to predict the more likely CDS (F). Each transcript is shown as a dot, where X-axis has the total transcript length (bp) and Y-axis has that transcript longest ORF (E) or its CDS (F) lengths, both at nucleotide level (bp). Dashed diagonal lines define coverage thresholds; the longer the coding region encoded in the transcript is, the closer to the 100% diagonal should be. Density curves highlight what coverage ratio has more transcripts on it, where it is evident that the predicted estscan CDSs approach returns longer protein coding segments than the longest ORF method. DEG: Differentially Expressed Gene. ORF: Open Reading Frame. CDS: Coding DNA Sequence. Additional file 3. Figure S3. TYR-TUB and βCAT-1 staining of 0-24 hpd embryos and over-represented GO terms for up- and down-regulated DEGs at 24 vs 4 hpd. (A) Schematic representation of a planarian cocoon at 0-4 hpd. (B) Immunostaining on paraffin sections of a planarian whole cocoon at 0-4 hpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing the cleavage after deposition. Nucleus are stained with DAPI (Blue). (C) Schematic representation of a planarian embryo at 0-4 hpd. (D) Magnification of the cleaving blastomeres immunostained with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing that during the first cleavage division the blastomeres did not show βCAT-1 staining. Nucleus are stained with DAPI (Blue). (E) CC overrepresented GO terms for down-regulated DEGs at 24 hpd vs 4 hpd, showing genes involved in mitochondrial function, ERAD pathway (derlin-1 and fibronectin), Golgi trafficking and secretory vesicle production (purple arrows). Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in down-regulated genes. (F) BP over-represented GO terms for up-regulated DEGs at 24 hpd vs 4 hpd showing up-regulation in functions involved in cell migration among others (red arrows). Barplot color gradient as in Supp3.E, but relative to up-regulated genes. hpd: hours post deposition. GO: Gene Ontology. CC: Cellular Compartment. BP: Biological Process. Scale bar: (B): 200 μm (D): 10 μm. Additional file 4. Figure S4. βCAT-1 staining of stage 2, 3, 4 and 5 embryos (A) Schematic representation of a planarian embryo at 32 hpd. (B) Immunostaining on paraffin sections of a planarian embryo at 32 hpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing how primary epidermal differentiated cells do not present nuclear βCAT-1 (red arrowhead). Nucleus are stained with DAPI (Blue). (C) Schematic representation of a planarian embryo at 48 hpd. (D) Immunostaining on paraffin sections of a planarian embryo at 48 hpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing nuclear staining of βCATENIN-1 in the embryonic pharynx (red arrowheads). Nucleus are stained with DAPI (Blue). (E) Schematic representation of a planarian embryo at 72 hpd. (F). Immunostaining on paraffin sections of a planarian embryo at 72 hpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), showing cells in the germ band sharing equally βCAT-1. Nucleus are stained with DAPI (Blue). (G) Schematic representation of a planarian embryo at 4 dpd. (H) Whole mount Immunostaining of a planarian embryo at 4 dpd with anti TYR-TUB (magenta) and anti βCAT-1 (green), in the cells of the germ band surrounding the embryonic pharynx showing nuclear βCAT-1. Nucleus are stained with DAPI (Blue). Red arrowhead points to an embryonic pharynx cell positive for nuclear βCAT-1 staining. (I) Schematic representation of a planarian embryo at 4 dpd. (J) TYR-TUB (magenta) and βCAT-1 (green) immunostaining on a paraffin section of a planarian embryo at 5 dpd, showing nuclear βCAT-1 mostly at one side of the embryo. Additional file 5. Figure S5. Over-represented GO terms for 32 vs 24 hpd, 72 vs 32 hpd, 4 dpd vs 72 hpd and 7 dpd vs 4 dpd. (A) BP over-represented GO terms for up-regulated DEGs at 32 hpd vs 24 hpd showing an up regulation in functions related to ubiquitination, lipid metabolism and negative regulation of autophagy (red arrows). Barplot color gradient is defined by z-score that is proportional to the relative abundance of the term in up-regulated genes. (B) BP over-represented GO terms for up-regulated DEGs at 72 hpd vs 32 hpd showing an up-regulation in functions related to tissue morphogenesis as cell adhesion, muscle, neural pharynx, and gut differentiation (red arrows). Barplot color gradient as in Figure S5 A. (C) BP over-represented GO terms for up-regulated DEGs at 4 dpd vs 72 hpd, which shows an up-regulation of processes related to DNA replication, cell division and patterning of the adult (dorso-ventral and anterio-posterior) (red arrows). Barplot color gradient as in Figure S5 A. (D) BP over-represented GO terms for up-regulated DEGs at 7 dpd vs 4 dpd, which shows an up-regulation of processes related to a second event of tissue morphogenesis: cell adhesion and differentiation of several adult tissues. Barplot color gradient as in Figure S5 A. hpd: hours post deposition. dpd: days post deposition. DEG: Differentially Expressed Gene. CPM: Counts Per Million. GO: Gene Ontology. BP: Biological Process. Additional File 6. Figure S6. Heatmap of autophagy related genes, consecutive time-points comparisons and adult tissue morphogenesis genes. admp expression pattern. (A) Heatmap of atg genes specifically shut down at 32 hpd. The heatmap color gradient represents log2 (CPM). (B) Heatmap of the consecutive timepoints versus 4 hpd comparisons showing clustering of two groups: samples comparisons between 24-72 hpd vs 4 hpd and 4-5-7 dpd vs 4 hpd. Arrow indicates the transition from 72h to 4 dpd (C) Heatmap of genes involved in maintenance of neoblasts, differentiation of the eyes, the epidermis, the excretory system, and the pharynx and the digestive system in adult animals, which are specifically turned on between 4 and 7 dpd. The heatmap color gradient represents log2(CPM). (D) In situ hybridization to detect Spol-ADMP in embryos at 5, 7 and 14 dpd. dpd: days post deposition. hpd: hours post deposition. dpd: days post deposition. Scale bar: (D) 15 μm. Additional file 7. Figure S7. Comparison to Davies et al 2017. (A) Sankey plot showing the distribution of genes on each of the 7 DEGs categories, which define the nodes, on every developmental stage being compared across time, which where interspersed according to the closest time point, between the two experiments (Y, S2, S3, S4, S5, S6, S7, and S8 for Davies et al; 24x4hpd, 32x24hpd, 48x32hpd, 72x48hpd, 96x72hpd, 120x96hpd, and 168x120hpd for Coronel-Córdoba et al). 4,001 unique DEGs of 4,918 from all stages reported in Davies et al (from total transcripts), were calculated relative to all time points average expression, whilst for Coronel-Córdoba they came from the previous time pair-wise comparison DEGs (1,310 transcripts of 21,006); this higher resolution for the later may explain the observed switch in upregulated genes at 72hpd, which was described at S2 for the Davies’ set. A total of 27,259 transcript contigs merged from both experiments’ assemblies were classified into the 7 expression categories, then those categories were pair-wise linked to define the internode flows between each developmental stage time point pairs; transcripts not mapped in both sets, or those for which no pre-computed expression data was available from Davies et al, were assigned to “NA” category. Significant DEGs are classified as “UP” or “DOWN”; those genes just passing one of the significance criteria are labeled as “up” or “down” (adjusted p-value = 0.05 or|log(fold-change)| = 1); non-significant genes are classified as non-significant up (“nsu”) or down (“nsd”). Number of genes assigned to each node/category in function of their adjusted p-value and log fold-change are shown below all node labels. (B) (D) (F) Zoom on three pair-wise time point comparisons between Davies and Coronel-Córdoba sets. Sankey-plots description as in (A). (C) (E) (G) Bubble-plots show the number of genes (proportional to circle size), classified as “up” or “UP” in both sets, that are annotated with the same BP GO, for each pair-wise comparison ((B) (D) (F) respectively). n accounts for the number of GOs shown in the plots; only GOs having at least one gene annotated in both sets were considered. Red arrow-heads indicate the functions related to the BP highlighted in the different sections of the manuscript. hpd: hours post deposition. DEG: Differentially Expressed Gene. GO: Gene Ontology. BP: Biological Process.