Zic4(RNAi) leads to an up-regulation of genes normally expressed in epithelial cells
Next, we sought to characterize the cellular behavior that underlies deficient tentacle maintenance and tentacle formation in Zic4(RNAi) animals. The characteristic cells that provide the functionality of the tentacles are the tentacle battery cells (TBCs) in the epidermis, which terminally differentiate in the tentacle roots from the epidermal ESCs that reach this zone when displaced from the body column (Fig. 4A). To monitor the cellular composition of tentacles, we used transgenic Hydra constitutively expressing a tandem cell cycle sensor (TCCS) in epidermal epithelial cells (Fig. 4B). We noted a twofold decrease in the TBC number after Zic4(RNAi) counted either after maceration or on whole mounts (Fig. 4C and fig. S7). Nevertheless, we found that the level of epithelial apoptosis in the apical region and the level of epithelial proliferation in the body column were unchanged after Zic4 knockdown (fig. S8). Similarly, we did not record any obvious change in the displacement behavior of epidermal epithelial cells toward the head region (fig. S9). Thus, the altered size of the tentacles and the lower number of TBCs cannot be explained by decreased proliferation, increased apoptosis, or reduction in the number of cells allocated to tentacles. Ultimately, we postulated that the reduced size of the tentacles could be the result of an overall shape change on the part of the cells that compose them.
Genes specifically expressed in the basal disc get up-regulated in tentacles upon Zic4 knockdown
To evidence molecular markers downstream of Zic4, we performed comparative RNA-seq analysis on apical regions of scramble and Zic4(RNAi) animals. One day after EP4, the comparison reveals a notable decrease in the transcript level of four tentacle markers ks1, HyAlx, Nematocilin A, and Nematocilin B (Fig. 4D and dataset S2) (12, 15, 36, 37), consistent with the Zic4(RNAi)–induced loss of TBCs. While nematocilins are expressed in the nematocytes anchored in the TBCs, ks1 and HyAlx have been identified as genes exclusively expressed in TBCs, responsible for tentacle development in case of ks1 (12) and tentacle patterning in case of HyAlx (15). As expected, hybridization chain reaction (HCR) RNA–fluorescence in situ hybridization (FISH) experiments showed a striking decrease in ks1 expression upon down-regulation of Zic4 and Sp5 (Fig. 4E). Furthermore, we noted an up-regulation of genes normally expressed in the basal disc, including the peroxidases PPOD1, PPOD2, PPOD2l, and PPOD11, the homeogenes Dlx1 and NK2, the transmembrane bone morphogenetic protein (BMP) regulator Crim-1, and the BMP antagonist NBL1 (Fig. 4D and table S2). These findings suggest that Zic4 controls tentacle identity by repressing basal gene expression.
To further investigate gene expression changes that occur in tentacles of animals knocked down for Zic4 and/or Sp5, we used proximal and distal tentacle tissue for RNA-seq analysis (fig. S10, A and B). When projected into a Principal Component Analysis (PCA) space, generated from the positional sequencing of different body parts, the overall identity of the RNAi tentacles samples seems to shift only marginally from the location of the intact tentacles (fig. S10 and dataset S3). However, when the identity map is constructed with epithelia-specific genes, we noted a clear shift from tentacle toward basal identity in all RNAi samples (Fig. 4F). All investigated conditions show an increase in the expression of basal markers, combined with a decrease of tentacle marker genes (Fig. 4G and dataset S4), with the strongest modulations not only in the tentacles of Zic4/Sp5(RNAi) animals but also in ectopic structures of Sp5(RNAi) animals. These results, which confirm the basal transformation of TBCs, suggest a key role for Zic4 as we recorded the highest level of Zic4 silencing after Zic4/Sp5(RNAi) when compared to Sp5(RNAi) or Zic4(RNAi) (fig. S11).
Tentacles transform into basal discs upon Zic4/Sp5(RNAi)
Using in situ hybridization (ISH) approaches, we could monitor the spatial distribution of changes in gene expression. In Zic4(RNAi), Sp5(RNAi), and Zic4/Sp5(RNAi) animals, we found 3 days after EP3 tentacle regions that express Crim-1, a gene exclusively expressed in epidermal BDCs according to the single-cell analysis (38), but no longer Nematocilin A, as well as ectopic structures along the body column that do not express Wnt3 but express Crim-1 (Fig. 5A and figs. S12 and S13). In the same conditions, we observed MPS droplets typical of BDCs in tentacle cells as well as in the structures ectopically formed along the body column when Zic4 and/or Sp5 are knocked down (Fig. 5B). This phenotype is enhanced in Zic4/Sp5(RNAi) animals that develop complete ectopic basal discs, which are morphologically and functionally indistinguishable from those of control animals (100%, n = 70) (movies S1 and S2). We could monitor the BDC identity by detecting peroxidase activity (Fig. 5C). Just 5 days after EP3, the tentacles of Zic4/Sp5(RNAi) animals display similar localized peroxidase patterns and include cells almost indistinguishable from BDCs (Fig. 5D). We observed a similar tentacle to basal disc transformation upon β-catenin silencing (fig. S14), likely because of Zic4 down-regulation. We were able to monitor this tissue transformation in two distinct Hydra strains, Hv_AEP and Hv_Basel, although with a stronger penetrance in the former (see figs. S12 and S13). For this reason, we decided to perform all subsequent experiments in Hv_AEP2 animals.
Formation of ectopic basal discs upon Zic4/Sp5(RNAi) also occurs in neuron-depleted animals
Upon examination of Zic4 expression in single-cell RNA-seq data (38), we could detect, on top of the epithelial expression, some nerve cell populations that express relatively high amounts of Zic4 (Fig. 6A). To assess the role of these Zic4-expressing cells in the observed phenotype, we eliminated them by submitting the animals to a three-course hydroxyurea (HU) treatment, a procedure that eliminates in a few days all interstitial cycling cells and progressively their progeny (Fig. 6B) (39, 40). Forty hours after the last HU exposure, animals were bisected to regenerate their head in the absence of interstitial cells. We could see a marked decrease in the number of RFamide+ nerve cells, one of the interstitial cell lineage progeny, in HU-treated animals having regenerated their head after 1 week (Fig. 6C). However, the down-regulation of Zic4 and Sp5 in these nerve-depleted animals resulted 3 days after EP3 in the transformation of tentacles in basal disc structures as evidenced with the ectopic expression of Crim-1 (Fig. 6D). Thus, the tentacle phenotype induced by a low Zic4 expression is identical in animals with and without interstitial cells, implying that the epithelial cells are primarily involved in the observed transformation.
TBCs transdifferentiate into BDCs upon Zic4/Sp5(RNAi)
As a consequence, we hypothesized that the formation of ectopic basal discs in tentacles relies on the transdifferentiation of TBCs. As supporting evidence, we found TBCs in Zic4/Sp5(RNAi) animals, which exhibit a mixed identity, half tentacle, half basal disc already 2 days after EP2 with the typical granular cytoplasm of basal mucus aspect in the close vicinity of degenerating nematocytes (Fig. 7, A and B; fig. S15; and movies S3 to S7). The cytoplasm of TBCs actually contains lipid droplets with peroxidase activity, never observed in apical regions of scramble (RNAi) animals (Fig. 7C and fig. S16, A to D). Upon maceration of the Zic4/Sp5(RNAi) animals, these intermediate cells that transition from a TBC to a BDC fate are clearly visible, harboring simultaneously multiple nematocytes while expressing peroxidases normally found in BDCs (Fig. 7D and fig. S16E). Furthermore, isolated TBCs of apical tissues macerated 2 days after EP2 express the nematocyte marker Nematocilin A together with the BDC marker Crim-1 (Fig. 7E and fig. S17). These results confirm the transdifferentiation process, i.e., TBCs that still contain embedded nematocytes, while starting to express basal disc markers.
The cells in the tentacle epidermis originate from the epidermal ESCs located all along the gastric part of the body, which are being passively displaced toward the extremities. To exclude the possibility that in Zic4/Sp5(RNAi) animals the new epithelial cells that reach the tentacles directly differentiate into BDCs without first undergoing differentiation into TBCs, we performed a tracing experiment with GFP-labeled TBCs transplanted onto a nonlabeled host. We excised tentacles from animals expressing GFP in epidermal epithelial cells and transplanted them in the body column of Hydra not expressing GFP (Fig. 7F and movie S8). These chimeric animals were subsequently knocked down for Zic4 and Sp5, and 3 days after EP3, we identified several GFP+ cells, originally TBC cells from the graft, now expressing Crim-1 (Fig. 7G, fig. S18, and movies S9 to S12). This result confirms the transdifferentiation of a formerly TBC into a BDC.
TBCs undergoing transdifferentiation reenter the cell cycle
Next, we investigated whether cells that show signs of transdifferentiation also reenter the cell cycle. We first electroporated the animals once with small interfering RNAs (siRNAs) and exposed them for 2 hours to 5-bromo-2′-deoxyuridine (BrdU) before fixation. In control animals, BrdU+ nuclei are not detected in the tentacles, while in Zic4/Sp5(RNAi) animals, BrdU+ nuclei are detected in the proximal part of the tentacles (Fig. 8A). To confirm cell cycle reentry of TBCs, we performed a 5-hour BrdU labeling of TCCS transgenic Hydra knocked down for Zic4/Sp5. Two days after EP2, we found the proximal part of tentacles populated with BrdU+ nuclei among TBCs (Fig. 8, B to D), which contain peroxidase+ droplets (Fig. 8E and fig. S19A). The typical organization of TBCs becomes disrupted in the proximal part of tentacles of Zic4/Sp5 animals, with a sharp boundary with the distal part that remains well organized (Fig. 8E). We also found BrdU+ TBCs that strongly express Crim-1 (Fig. 8F and fig. S19B). These results suggest that TBCs that undergo transdifferentiation reenter the cell cycle by activating DNA synthesis. However, we did not find any evidence of increase in mitotic activity when we analyzed the spatial distribution and the density of phosphohistone-positive nuclei (fig. S20).
To collect evidence on the molecular mechanism that possibly drives cell cycle reentry, we analyzed the Gene Ontology (GO) term of the genes consistently up-regulated when Zic4 and/or Sp5 are knocked down, and we identified biological functions linked to cell cycle, i.e., G1-S transition, positive regulation of the cell cycle, and chromosome segregation (Fig. 8G and dataset S5), confirming that epithelial cell cycling is enhanced upon Zic4 and/or Sp5 silencing. To test whether cell cycling is necessary for tentacle transdifferentiation, we inhibited DNA synthesis with two pulses of HU. We observed that ectopic Crim-1 spots no longer form in tentacles of HU-treated Zic4/Sp5(RNAi) animals, at least transiently (Fig. 8H and fig. S21) , in agreement with the transient HU-induced blockade of DNA synthesis (40). Together, these data indicate that upon Zic4/Sp5 silencing, TBCs do not maintain their typical tentacle organization, reenter the cell cycle but do not divide, and fully transform within less than 7 days into BDCs.
DISCUSSION
The Hydra apical organizer relies on Wnt/β-catenin signaling as well as Sp5 and Zic4 activities
This study places the Zic4 function under the regulation of the head organizer. The results obtained in intact and regenerating animals indicate that distinct levels of Wnt/β-catenin signaling as well as Sp5 and Zic4 activity define three regions in the apical region and trigger the formation of two distinct structures, the region of the mouth opening at the tip of the hypostome and the tentacle zone at the base of the hypostome. At the tip of the hypostome, Wnt/β-catenin signaling is high, while Zic4 and Sp5 are low. In the intermediate region (the dome), Wnt/β-catenin signaling is expected to be moderate given the lower level of Wnt3 expression, the moderate level of Sp5 expression, and the low level of Zic4. Last, the tentacle zone is characterized by low Wnt/β-catenin signaling, high Sp5, and high Zic4 expression, a condition that promotes tentacle formation (Fig. 9A). This molecular definition of the head organizer actually fits the anatomical description of the Hydra head based on cellular analyses, with the inner hypostome (region surrounding the mouth opening), the outer hypostome (intermediate region), and the tentacle zone.
The Zic4 expression in the tentacle zone, the tentacle roots, and the proximal region of tentacles together with the short tentacle phenotype obtained in intact Zic4(RNAi) Hydra and the lack of tentacles in ALP-treated or regenerating Zic4(RNAi) Hydra indicate that Zic4 acts at three distinct levels: (i) in the tentacle zone, where Zic4 contributes to commit epidermal epithelial cells to the TBC fate as deduced from the loss of ks1-expressing cells upon Zic4/Sp5(RNAi); (ii) in the tentacle roots, where Zic4 is necessary to differentiate TBCs as deduced from the down-regulation of ks1 and HyAlx after Zic4(RNAi); and (iii) in the proximal region of tentacles where Zic4 is required to maintain the differentiated status of TBCs, as deduced from the loss of TBCs after Zic4/Sp5(RNAi). These spatially distinct activities of Zic4 correspond to the established three distinct phases of tentacle formation—TBC commitment in the tentacle zone, TBC differentiation in tentacle roots, and tentacle elongation (15). These Zic4 activities are supported by RNA-seq and whole-mount ISH analyses of two epithelial markers, ks1 and HyAlx, down-regulated in Zic4(RNAi) and/or Zic4/Sp5(RNAi) animals
Comments
Post a Comment