Loss of plac8 expression rapidly leads pluripotent stem cells to enter active state during planarian regeneration

Planarian regeneration, which can produce morphologically and functionally complete individuals from tiny body fragments, has fascinated many scientists for a long time. In the past two decades, several kinds of cells involved in planarian regeneration have been identified at the molecular level by employing newly established biological methodologies. Adult pluripotent stem cells (aPSCs) in planarians called neoblasts are central to planarian regeneration as they give rise to all of the differentiated cells that compose the planarian body (Agata and Watanabe, 1999; Shibata et al., 2010). In addition to this ability to produce functional somatic cells, aPSCs can produce germline cells during sexualization from the asexual state in the planarian Dugesia japonica (Sato et al., 2006). The first report about the expression of a germ-cell-related gene described a vasa-related gene in aPSCs in D. japonica (Shibata et al., 1999). Following that finding, many reports showed that in D. japonica and a related species, Schmidtea mediterranea, most representative genes of aPSCs are germ cell-specific or germ cell-related genes, including piwi homologs (Guo et al., 2006; Hayashi et al., 2010; Reddien et al., 2005; Rossi et al., 2006; Salvetti et al., 2005; Shibata et al., 2012, 2010; Solana et al., 2009; Yoshida-Kashikawa et al., 2007). Among planarian piwi homolog genes, the Djpiwi and smedwi families in D. japonica and S. mediterranea, respectively, are expressed specifically or predominantly in aPSCs, and DjpiwiA (and its ortholog smedwi-1) is a specific marker gene for aPSCs (Hayashi et al., 2010; Palakodeti et al., 2008; Reddien et al., 2005; Shibata et al., 2016; Yoshida-Kashikawa et al., 2007). Specific expression of DjpiwiA in all aPSCs has been confirmed by morphological analysis and single cell-based real-time (RT)-PCR in D. japonica (Yoshida-Kashikawa et al., 2007; Hayashi et al., 2010).

Previously, DjpiwiA (or smedwi-1)⁺ aPSCs were considered to be a homogeneous population that constituted ∼25% of total planarian cells (Yoshida-Kashikawa et al., 2007; Reddien et al., 2005). However, recently, more detailed work has shown molecular heterogeneity of the DjpiwiA (or smedwi-1)⁺ aPSC population. There are subpopulations that specifically express genes related to differentiated cells together with DjpiwiA (or smedwi-1) (Hayashi et al., 2010; Zeng et al., 2018). These subpopulations of stem cells appear to be specified stem cells which will differentiate into a specific lineage of progeny cells (Zeng et al., 2018). Although we still do not know how each cell subtype in the aPSC population performs its own unique behavior in intact animals, piwi⁺ aPSCs appear to behave uniformly like a homogeneous single population in certain situations. Injury is a biologically urgent situation in which homeostasis collapses and it is necessary to restore it by reconstructing missing body parts and/or rearranging remaining body parts. In planarians, amputation of their body [loss of body part(s)] is a cue to start their regeneration, which is completed within 1 week (Agata et al., 2003). Therefore, aPSCs have to become fully prepared to produce differentiated cells after amputation more quickly than aPSCs in the steady state (tissue homeostasis).

The aPSCs are the only cell population showing continuous proliferation even in normal conditions (Newmark and Sánchez Alvarado, 2000; Sakurai et al., 2012; Tasaki et al., 2011b; Wenemoser and Reddien, 2010). Interestingly, acceleration of cell proliferation after amputation or feeding has been reported in several planarian species (Baguñà, 1976; Saló and Baguñà, 1984). A recent report showed that the accelerated mitotic activity consisted of two peaks after amputation in S. mediterranea (Wenemoser and Reddien, 2010). Accelerated proliferation is also observed in specified stem cells. Specified stem cells increase in number and undergo differentiation during accelerated proliferation. For example, ovo⁺ stem cells, which are eye-specified, are increased and produce progenitors after amputation (Lapan and Reddien, 2012). Thus, it appears that accelerated proliferation might occur in all of the aPSCs that play an important role in regeneration.

Although we reported that an aPSC-specific gene, P2X-A, which encodes an ATP-dependent ion channel, modulates the accelerated proliferation after feeding in D. japonica (Sakurai et al., 2012), the molecular mechanism responsible for accelerated proliferation after amputation is still unknown. However, two intracellular signaling pathways are predicted to be involved in the transient acceleration of aPSC proliferation after amputation. During regeneration, a blastema, composed of differentiating and differentiated cells derived from the aPSCs, is formed at a wound (Tasaki et al., 2011b). The signaling pathway of c-JUN N-terminal kinase (JNK), a pathway required for normal proliferation of aPSCs, is strongly activated in the post-blastema region, in which accelerated proliferation takes place during the early stage of regeneration (Saló and Baguñà, 1984; Tasaki et al., 2011b). Another signaling pathway, extracellular signal-regulated kinase (ERK) signaling, which is involved in the differentiation of aPSCs in D. japonica (Tasaki et al., 2011a), might also regulate aPSC proliferation, because ERK signaling is involved in cell proliferation in many adult stem cell (ASC) systems in other animals (Hasegawa et al., 2013). However, there is no evidence about whether either or both of these signaling pathways are involved in the transient acceleration of global aPSC proliferation.

To understand this transient acceleration of global aPSC proliferation at the molecular level, we first tried to identify molecules involved in accelerated proliferation after amputation and feeding. We screened aPSC-specific genes identified by comprehensive gene expression analysis (HiCEP) and focused on placenta specific gene 8 homolog-A (plac8-A) (Shibata et al., 2012), as plac8 has functions in the regulation of proliferation and differentiation in diverse phyla, from plants to animals (Cabreira-Cagliari et al., 2018). Here, we show that, after amputation, expression of plac8-A disappeared at the post-blastema region as a result of activated JNK signaling, followed by accelerated proliferation of the aPSCs. When JNK signaling was attenuated during regeneration by an appropriate dose of specific inhibitor, the proliferation of the aPSCs failed to accelerate, resulting in death and regenerative defects of the planarians. However, simultaneous knockdown of plac8-A restored the accelerated aPSC proliferation in these planarians, resulting in survival and increase of regenerative ability. From these results, we conclude that accelerated aPSC proliferation induced by loss of plac8-A expression is indispensable for appropriately producing differentiated cells required for regeneration. We propose that plac8-A is a switch for transition of the cellular state from steady state to regenerative state in a broad aPSC subpopulation in response to amputation.

The proliferation status of aPSCs after feeding can be assessed by counting the number of fissions of planarians subjected to long-term feeding RNAi. For example, long-term P2X-A(RNAi) animals showed enhanced accelerated proliferation followed by higher fission frequency, suggesting that P2X-A is involved in aPSC proliferation after feeding (Sakurai et al., 2012). We employed this unique method to identify aPSC-specific genes involved in the accelerated proliferation of aPSCs. Long-term RNAi was introduced by feeding planarians liver extract containing double-stranded RNA (dsRNA) targeting the gene of interest once per week for a desired period in addition to the three successive feedings that are used for standard planarian RNAi experiments (Rouhana et al., 2013; Sakurai et al., 2012). First, we screened the aPSC-specific genes identified by HiCEP, based on expectable conserved gene functions known in other organisms (Shibata et al., 2012). Among them, we focused on HiCEP clone number 37 gene, which has homology to plac8, because plac8 has various known functions in cellular regulation, including regulation of proliferation and differentiation (Guo et al., 2010; Jimenez-Preitner et al., 2011). In particular, many previous reports have shown that the major function of plac8 is regulation of proliferation (Guo and Simmons, 2011; Libault et al., 2010). Thus, we named this gene plac8-A and performed long-term RNAi targeting plac8-A.

Interestingly, plac8-A knockdown caused increased fission frequency (Fig. 1A). plac8-A(RNAi) animals reproducibly showed three times higher fission frequency compared with the control animals (n=10×3), which was comparable with the increase of fission rate observed in P2X-A(RNAi) planarians (Fig. 1A; Sakurai et al., 2012). A decrease of the expression level of plac8-A in plac8-A(RNAi) animals was confirmed by RT-PCR (Fig. S1) and plac8-A(RNAi) animals regenerated normally after fission as observed in P2X-A knockdown animals (Sakurai et al., 2012).

In situ hybridization of plac8-A in whole-mount planarians revealed a typical expression pattern of this aPSC-specific gene, in which plac8-A-expressing cells were located throughout all body regions posterior to the head region except in the pharyngeal region (Fig. 1B). X-ray irradiation is known to specifically eliminate the aPSC population (Wolff and Dubois, 1948; Shibata et al., 1999; Hayashi et al., 2010), and indeed the in situ hybridization signal of plac8-A disappeared in X-ray-irradiated planarians (Fig. 1B). Thus, we confirmed the reported aPSC-specific expression of plac8-A (Shibata et al., 2012). This expression pattern is identical to the expression pattern of piwiA (one of the most reliable aPSC marker genes). Among the aPSC-specific genes that showed a similar expression pattern to piwiA, P2X-A was reported to be expressed heterogeneously in the piwiA⁺ population (in ∼50% of piwiA⁺ aPSCs) (Sakurai et al., 2012; Shibata et al., 2012). Next, we examined the expression of plac8-A at the single cell level in the piwiA⁺ population using fluorescence-activated cell sorting (FACS)-based single cell PCR (FBSC-PCR). We performed gene expression analysis on the Index Sorting profile obtained by FBSC-PCR in intact animals. Nearly all (97%) of plac8-A-expressing cells were in the X1 fraction plus X2 fraction (231/238), which were sorted into aPSC-enriched fractions. Also, 84% of the expression of plac8-A was detected in piwiA⁺ aPSCs (190/227) (Fig. 1C) (Hayashi et al., 2010, 2006). Thus, plac8-A was expressed throughout the piwiA⁺ aPSC population.

Next, we analyzed the subcellular localization of Plac8-A protein in the aPSCs by immunostaining using an anti-Plac8-A antibody that we raised. This antibody specifically recognized a 16-kDa protein in western blotting (Fig. S2A), which was coincident with the predicted size of Plac8-A protein. To check the specificity of the antibody, we conducted immunostaining of plac8-A(RNAi) planarians, and found that the signal disappeared in the RNAi animals (Fig. S2B), confirming the specificity of the antibody. Immunostaining using this antibody revealed that the signal was generally observed in aPSCs that were also positive for PiwiA protein, in accord with the result of FBSC-PCR (Fig. 1C; see also Yoshida-Kashikawa et al., 2007). Plac8-A was localized in the outermost region in the cytoplasm of the aPSCs, whereas PiwiA protein was widely observed throughout the cytoplasm of the aPSCs (Fig. 1D, upper panel). We also performed co-staining with a planarian membrane protein, P2X-A, as a previous report showed that P2X-A is localized on the cell membrane of about half of the neoblasts (Sakurai et al., 2012). Co-localization of Plac8-A and P2X-A was observed (Fig. 1D, lower panel), strongly suggesting that Plac8-A is a cell membrane protein or associated with a cell membrane protein(s). This suggested that subcellular localization of Plac8-A protein is in accord with the reported localization of Plac8 protein in other organisms (Guo et al., 2010). Although most PiwiA⁺ cells were also positive for Plac8-A, it was noteworthy that two subpopulations of aPSCs were negative for Plac8-A: nanos⁺ aPSCs localized in the dorsolateral side of the body (Sato et al., 2006) and the piwi-1⁺ subpopulation localized on the dorsal longitudinal midline (Rossi et al., 2006) (Fig. S3).

Plac8 family genes have a conserved domain called the Plac8 superfamily domain (Fig. S4A). In our D. japonica EST database (Nishimura et al., 2015; An et al., 2018), we identified six other genes possessing the Plac8 superfamily domain in addition to plac8-A (Fig. S4B). One of them was reported to be highly similar to plac8 and to be related to immune response and development in planarian (Pang et al., 2017). We named the five newly found homologs plac8-B to -F, and determined their expression patterns: the plac8-B to -F genes were not expressed in the aPSCs (Fig. S4C), and therefore we focused on only plac8-A in this study.

In the past, the accelerated proliferation of the aPSCs has been called a ‘mitotic burst’ after amputation or feeding, or a ‘local response to loss of tissue’ after amputation in various reports (Sakurai et al., 2012; Saló and Baguñà, 1984; Wenemoser and Reddien, 2010). Hereafter, we call this phenomenon ‘induced hyperproliferation’ in order to unify the definition of this proliferation mode across all of these types of its induction.

First, we performed RT-PCR to examine the expression dynamics of plac8-A after feeding. The significance of differences of expression levels was evaluated by comparing the expression at each time point to that in the starved state. The expression levels of cell-proliferation marker genes such as pcna and mcm2 increased soon after feeding. The expression level of pcna and mcm2 increased significantly at 12 h after feeding, and then decreased to the steady-state level 1 week after feeding, as previously reported (Fig. 2A; see also Sakurai et al., 2012), indicating that hyperproliferation was induced after feeding. The expression level of plac8-A was decreased at 12 h and 24 h after feeding, and returned to the initial steady-state level 5 days after feeding (Fig. 2B). A negative correlation between induced hyperproliferation and gene expression after feeding was similarly observed for P2X-A (Fig. 2B; Sakurai et al., 2012), suggesting that plac8-A and P2X-A might have similar roles in induced hyperproliferation after feeding.

As mentioned above, induced hyperproliferation was reported at various times after feeding in P2X-A knockdown planarians (Sakurai et al., 2012). We therefore investigated whether increased proliferation was induced in plac8-A knockdown animals after feeding. plac8-A knockdown animals showed higher expression levels of pcna than control animals from 12 h to 7 days after feeding, indicating that the feeding-induced hyperproliferation of aPSCs was enhanced by plac8-A knockdown, as it was by P2X-A knockdown (Fig. 2C). Also, after feeding, P2X-A(RNAi) animals showed an increase of M-phase aPSCs [aPSCs positive for anti-phosphohistone H3 (pH3) immunostaining] cells at the region directly anterior to the pharynx in a previous report (Sakurai et al., 2012). We employed the same method to count pH3⁺ cells as in that previous report (Sakurai et al., 2012), and found that the number of pH3⁺ cells increased in plac8-A(RNAi) animals, as it did in P2X-A(RNAi) animals (Fig. 2D,D′). These results indicate that plac8-A expression is negatively correlated with induced hyperproliferation after feeding, and thus may modulate aPSC proliferation, as does P2X-A expression.

Although the previously reported findings and those obtained here show that plac8-A and P2X-A might modulate the induced hyperproliferation of aPSCs after feeding, whether these genes are also involved in induced hyperproliferation after amputation still remained unknown. To answer this question, first we examined whether aPSC proliferation was induced after amputation in D. japonica, because there have been no reports regarding this question. The significance of differences of expression levels was evaluated by comparing the expression at each time point to that at 0 h after amputation. The expression levels of proliferative markers were significantly increased rapidly (within 12 h) after amputation compared with the levels at 0 h after amputation, as they were in the case of induced hyperproliferation after feeding, and returned to the steady-state level within 1 week (Fig. 3A). Although the patterns of the time course and level of increases or decreases of proliferative marker expression were slightly different depending on the particular experiments and proliferative markers examined, we confirmed that their expression levels certainly increased after amputation (and also after feeding), and then returned to the steady-state expression levels within 1 week (Figs 2 and 3). Thus, we concluded that proliferation is accelerated after amputation in D. japonica, as it is in other planarian species. In S. mediterranea, two types of induced hyperproliferation after amputation were reported (Wenemoser and Reddien, 2010): an early response to lesions caused by piercing or lesioning with/without amputation, and a late regeneration-dependent response caused by loss of body part(s) by amputation (Wenemoser and Reddien, 2010). However, we could not detect any obvious induced hyperproliferation in simply pierced or injured planarians without amputation in D. japonica (Fig. S5).

Next, we examined the expression dynamics of P2X-A and plac8-A during regeneration after amputation. Interestingly, the expression level of plac8-A was significantly reduced (by ∼50%) within 12 h after amputation. In contrast, P2X-A expression did not significantly change after amputation (Fig. 3B). Downregulation of plac8-A expression was detected during an early stage of regeneration (until 3 days after amputation), and then plac8-A expression gradually recovered to the steady-state level by 7 days after amputation. The expression patterns of proliferation markers and plac8-A showed opposite trends, as observed in the case of induced hyperproliferation after feeding, suggesting that plac8-A, but not P2X-A, is negatively correlated with the induced hyperproliferation after amputation.

To confirm this, we monitored the induced hyperproliferation in plac8-A(RNAi) animals. For this, we introduced dsRNA of plac8-A directly into the planarian digestive tract by injection (to avoid the induction of proliferation by feeding). As expected, plac8-A(RNAi) animals showed higher expression of pcna than control animals at all time points examined within 1 week after amputation (Fig. 3C). The number of M-phase cells, as indicated by anti-pH3 immunostaining, was also significantly increased in these plac8-A knockdown animals (Fig. 3D,D′). These results indicate that plac8-A exerts a negative effect on induced hyperproliferation after amputation. We also examined whether aPSC proliferation was induced after amputation in P2X-A(RNAi) planarians, but no acceleration of induced hyperproliferation was observed after amputation in these planarians (Fig. 3E,F,F′), indicating that P2X-A is dispensable for induced hyperproliferation after amputation. Thus, plac8-A is the only gene demonstrated thus far to be involved in induced hyperproliferation after both feeding and amputation, and P2X-A is only involved in the induced hyperproliferation after feeding.

We then performed in situ hybridization to examine the spatial expression pattern of plac8-A expression during regeneration after amputation. It is known that during regeneration, a blastema is formed at the wounding site and is composed of differentiating and/or differentiated cells derived from the aPSCs (Tasaki et al., 2011b). Interestingly, we found that expression of plac8-A was drastically reduced in the post-blastema region (i.e. the regions posterior to the head blastema and anterior to the tail blastema) during regeneration at 12 h and 24 h after amputation (Fig. 4A), although aPSCs (as indicated by PiwiA immunostaining) were present in the post-blastema region (Fig. 4B). As regeneration proceeded, the expression of plac8-A recovered in the anterior region (except in the head) and in the posterior region, with restoration of the normal expression pattern at 7 days after amputation (Fig. S6A). Immunostaining and western blotting also indicated that Plac8-A protein decreased in the post-blastema region during regeneration, whereas the Plac8-A level remained unchanged in the rest of the body (Fig. 4C,C′; Fig. S6B). Therefore, we concluded that the expression of plac8-A and its protein were transiently reduced in aPSCs located in the post-blastema region during the early stage of regeneration.

Previous studies have shown that the post-blastema region after amputation contains many M-phase aPSCs (Saló and Baguñà, 1984; Tasaki et al., 2011b). Also, it has been reported that JNK is activated in the post-blastema region (Tasaki et al., 2011b). Accordingly, next we examined the possible role of JNK signaling in plac8-A expression and induced hyperproliferation. First, we visualized the activation of JNK using anti-phosphorylated JNK (pJNK) antibody in head-regenerating planarians, and confirmed that JNK was activated in the post-blastema region (where plac8-A mRNA was decreased), as previously reported (Fig. 5A; Tasaki et al., 2011b). Then, to test whether JNK signaling can indeed regulate the expression of plac8-A in this region during regeneration, we examined the expression of plac8-A in planarians treated with 25 µM SP600125, JNK inhibitor, after amputation. A previous report showed that treatment with 25 µM SP600125 caused loss of almost all mitotic aPSCs by blocking the entry into M-phase of the cell cycle (Tasaki et al., 2011b). In control animals, expression of plac8-A decreased in the post-blastema region at 24 h after amputation (Fig. 5B, upper panel), but in SP600125-treated planarians, plac8-A was still expressed in aPSCs there (Fig. 5B, lower panel). We then measured the expression level of plac8 after amputation in SP00125-treated planarians by qRT-PCR. In the inhibitor-treated animals, the level of expression of pcna was not increased. Furthermore, the level of expression of plac8-A was not changed in SP600125-treated planarians after amputation, in contrast to the decrease of the expression level of plac8-A in control amputated animals (Fig. 5C). Taken together, these findings led us to conclude that activation of JNK signaling contributes to the induced hyperproliferation via repression of plac8-A expression in aPSCs located in the post-blastema region after amputation.

Next, we tested the possibility that ERK signaling affects the expression of plac8-A and the induced hyperproliferation of aPSCs, as the ERK signaling pathway is also known to be one of the major signaling pathways for promoting the proliferation of stem cells, such as spermatogonial stem cells (Hasegawa et al., 2013), although the role of ERK signaling was reported to be promotion of the differentiation of the aPSCs in D. japonica (Tasaki et al., 2011b). We checked the induced hyperproliferation of aPSCs and the expression of plac8-A in regenerating animals treated with U0126, an MEK-specific inhibitor that specifically blocks ERK signaling (Favata et al., 1998; Tasaki et al., 2011a). Planarians were amputated and treated with 25 µM U0126, which blocks the differentiation of aPSCs (Tasaki et al., 2011a; Umesono et al., 2013). First, we monitored induced hyperproliferation by detecting the expression level of pcna, and found relatively weaker elevation of pcna expression in the U0126-treated animals compared with the increase of pcna expression in the amputated control animals (Fig. 5D). The expression level of plac8-A was not changed by U0126 treatment after amputation (Fig. 5D). These results raised the question of whether ERK signaling can directly affect the expression of plac8-A, or whether ERK signaling affected plac8-A expression via a signaling cascade through JNK phosphorylation. To answer this, we monitored the activation of JNK in U0126-treated planarians by immunostaining with anti-pJNK antibody, which revealed that the signal of pJNK was greatly reduced in the post-blastema region in U0126-treated planarians (Fig. 5E). Taken together, these findings suggest that U0126 induced failure of the activation of JNK signaling, resulting in a low level of induced hyperproliferation in planarians after amputation. Thus, induced hyperproliferation after amputation might require both JNK signaling and ERK signaling.

Next, to elucidate the biological consequences of the induced hyperproliferation of aPSCs after amputation, we investigated whether inhibition of this induced hyperproliferation affected regeneration. First, we examined the effect of SP600125 by treating planarians with 25 µM SP600125 after amputation and allowing them to regenerate. Planarians treated with 25 µM SP600125 throughout regeneration showed severe regenerative failure. At 1 week after amputation, planarians treated continuously with 25 µM SP600125 after amputation showed head and tail regression or died. This regenerative failure was in accord with but more severe than the reported defects when the inhibitor treatment was delayed, namely, incomplete head or tail structure regeneration in planarians treated with 25 µM SP600125 during only a limited time (from after 12 h after amputation until 3 days after amputation) (Tasaki et al., 2011b). That previous report showed that treatment with 25 µM SP600125 after amputation caused failure of regeneration by blocking wound healing in addition to blocking progression of aPSC cell cycle from S- to M-phase in almost all aPSCs, as mentioned above (Tasaki et al., 2011b). Therefore, we tested lower concentrations of SP600125, and found that the optimal concentration that enabled normal proliferation by allowing entry into M-phase to some extent, but did not induce hyperproliferation, was 5 µM. At this optimal concentration, planarians regenerated partially, showing regenerative defects as indicated by the incidence of cyclopia and eye-absent planarians. Then we tested whether this partial regenerative defect could be abrogated by RNAi of plac8-A. For this, we amputated and treated planarians with 5 µM SP600125 after a series of injection RNAi treatments against plac8-A. Three groups of biological replicates were performed and the regeneration of these planarians was observed at 1 week after amputation. Most of the control animals (injection of dsRNA of EGFP) with inhibitor treatment (70.8% of planarians) failed to regenerate normally: 13.4% of control animals were dead and 57.4% of them showed cyclopia or no eyes (Fig. 6A,B). In contrast, there was no death of plac8-A(RNAi) animals treated with the SP600125: they all survived and underwent regeneration, and 40% showed normal eye regeneration, suggesting that they had higher regenerative ability than control animals (Fig. 6B). The ratios of eye-absent animals and cyclops were also decreased compared with those in control animals (Fig. 6B). We also examined the regeneration of visual neurons by performing immunostaining of Arrestin. Most plac8-A knockdown plus inhibitor-treated animals (86% of planarians; 18/21) completely regenerated the visual neurons, whereas EGFP(RNAi) plus inhibitor-treated animals failed to regenerate these neurons (Fig. 6C). Thus, attenuation of plac8-A expression by RNAi could abrogate the regenerative defects caused by inhibition of JNK signaling.

We then tested whether these abrogations by RNAi against plac8-A in SP600125-treated animals are associated with enhanced cell cycling of the aPSCs, like the induced hyperproliferation caused by reduction of plac8-A expression in normal planarians. Control SP600125-treated animals did not show induced hyperproliferation after amputation, although they maintained their basal rate of mitosis (Fig. 6D). In contrast, SP600125-treated plac8-A(RNAi) animals showed an increase of mitotic cell number relative to that in control animals, comparable with the amputation-induced increase of mitotic cell number in normal planarians. That is, RNAi of plac8-A could restore the induced hyperproliferation in SP600125-treated animals after amputation. These results suggested that reduction of plac8-A expression alone is sufficient to cause the induced hyperproliferation after amputation, and this induced hyperproliferation might play an important role in regeneration by promoting the progression of aPSC differentiation.

Then, to test the importance of induced hyperproliferation in the cell differentiation of aPSCs, we checked the expression of runt-1 in SP600125-treated planarians. It was reported that runt-1 is expressed in a certain population in the aPSC pool located in the post-blastema region, and is involved in the commitment of neuronal and eye cells during head regeneration in S. mediterranea (Wenemoser et al., 2012). First, we confirmed the expression pattern of runt-1 in D. japonica. Expression of runt-1 was rapidly increased at 3 h after amputation, and continued to increase during early regeneration (Fig. 6E). This elevation of runt-1 expression after amputation was attenuated by X-ray irradiation (Fig. 6E). Also, in situ hybridization showed that runt-1 was expressed in the post-blastema region in early regeneration, and this expression was weakened by X-ray irradiation (Fig. 6E). Thus, we confirmed that runt-1 was expressed in the aPSCs localized in the post-blastema region during regeneration, as reported for runt-1 in S. mediterranea (Wenemoser et al., 2012). Finally, we tested the relationships among activation of JNK, reduction of plac8-A expression and expression of runt-1 by a combinatory experiment using SP600125 and plac8-A(RNAi). When planarians were treated with 5 µM SP600125 after amputation, runt-1 expression was slightly increased, but not to its normal expression level at 12 h after amputation (Fig. 6F). However, RNAi of plac8-A in the inhibitor-treated animals could significantly restore runt-1 expression during early regeneration (Fig. 6F). Taken together, these results suggest that the induced hyperproliferation caused by inhibition of plac8-A expression as a result of activation of JNK signaling is crucial for commitment of the aPSCs after amputation.

Induced hyperproliferation after amputation or feeding has been known for decades in various species of planarians, for example Dugesia tigrina and S. mediterranea (Baguñà et al., 1989; Wenemoser and Reddien, 2010). During the induced hyperproliferation, mitotic cells (M-phase cells) are rapidly increased, and thereafter gradually return to the steady-state level within about 1 week (Saló and Baguñà, 1984). Induced hyperproliferation in response to amputation or feeding, in which aPSCs supply differentiated cells during regeneration or growth, appears to be a common event across planarian species. In the case of S. mediterranea, induced hyperproliferation after amputation occurs with two peaks. The first peak occurs as a result of simple wounding such as stabbing, and the second peak occurs as a result of complete amputation of the bodies (Wenemoser and Reddien, 2010). However, in D. japonica, we found here that amputation could accelerate aPSC proliferation, but simple wounding without regeneration, such as stabbing, could not evoke such acceleration (which is relevant to the first mitotic peak in S. mediterranea) (Fig. 3A). Thus, the regulation of stem cell responses to wounding appears to differ among planarian species.

A recent report showed that one of the plac8 homologs in planarians is related to immune response and development (Pang et al., 2017). Here, we identified six additional plac8 homologs and showed that, among them, plac8-A is the only aPSC-specific gene that is involved in aPSC proliferation. plac8 has multiple functions in diverse phyla, from plants to animals, as mentioned above. Thus, plac8 has been called several names, including cell number regulator, onzin, depending on its function and known related genes. New consolidated nomenclature for Plac8 was proposed recently (Cabreira-Cagliari et al., 2018). In plants, most studies showed that plac8 homologs regulate cell proliferation negatively and control the weight and size of fruits in tomato, maize, etc. (Guo et al., 2010; Guo and Simmons, 2011; Libault et al., 2010). On the other hand, diverse functions of plac8 in animals, including regulation of cell differentiation, proliferation, apoptosis and migration, have been reported (Bedell et al., 2012; Li et al., 2006; Mao et al., 2019). Here, we revealed that plac8-A is involved in induced hyperproliferation of aPSCs via a decrease of its expression level after both amputation and feeding. This supports the notion that plac8 and its function in regulating cell proliferation and/or differentiation are conserved in many organisms across diverse phyla.

How does Plac8-A regulate proliferation of aPSCs in planarian? As common features of Plac proteins between plants and planarians are that Plac8 is a negative regulator of mitotic activity and is a membrane-associated protein (Frary et al., 2000; Fig. 1D), it is possible to speculate about the detailed molecular function of Plac8-A by referring to Plac8 of plants. Because most of the Plac8 in plants is localized as a transmembrane protein, it is considered to affect mitotic activity indirectly (Cong and Tanksley, 2006; Li and He, 2015; Libault et al., 2010). For example, Plac8 regulates cell proliferation negatively by interacting with casein kinase II (CKII)β subunit in tomato and soybean. Modified CKIIβ after interacting with Plac8 translocates from the cytosol to the nucleus and activates the CKIIα subunit, which represses the cell cycle (Cong and Tanksley, 2006; Libault et al., 2010). In addition, another report showed a negative correlation between organ size and plac8 expression level (Li and He, 2015). In this case, Plac8 interacts with AG2, which is an agamous-like MADS domain protein localized at the cell membrane. Then, modified AG2 translocates to the nucleus and binds to the CArG-box in the Cyclin promotor, resulting in enhanced repression of the cell cycle. However, after a decrease of the Plac8 protein level, AG2 is released without modification from the membrane, resulting in weakened repression of the cell cycle (Li and He, 2015). These results obtained in plants suggest that absence of Plac8 leads to loss of a cell cycle repressor, and consequently promotes cell proliferation. Therefore, investigating the molecules that interact with planarian Plac8-A in future studies will be important for deeply understanding the cell cycle control by Plac8-A.

Translocation of the subcellular localization of Plac8 protein from the cell membrane is also commonly observed in vertebrates. For example, in mice, Plac8 is known to translocate from the cell membrane to the nucleus, and to bind there to the promoter of a transcription factor, C/EBPβ, to induce its transcription, and consequently promote the differentiation of brown adipocytes (Jimenez-Preitner et al., 2011). We could not observe obvious translocation of Plac8-A during regeneration in this study. However, producing Plac8-A-GFP transgenic animals might enable us to show more clearly the Plac8-A localization pattern in the aPSCs during regeneration in the future. Our studies revealed that induced hyperproliferation of aPSCs is important not only for aPSC expansion but also for supplying differentiated cells for normal regeneration. The biological role of induced hyperproliferation in cellular differentiation during planarian regeneration is discussed below.

We showed that activated JNK signaling downregulates the expression of Djplac8-A in the aPSCs located in the post-blastema region after amputation. Treatment of animals with an appropriate concentration of SP600125 attenuated the expression of plac8-A and resulted in failure of induced hyperproliferation, but not of proliferation in control animals, indicating that JNK signaling elicited induced hyperproliferation in the aPSCs pool after amputation (Fig. 7). In addition, our results suggest that ERK signaling also impacts the expression of plac8-A after amputation by cooperating with JNK signaling. ERK signaling is known to be important for differentiation of the aPSCs (Tasaki et al., 2011a). At the high concentration of ERK inhibitor, U0126, de novo differentiation of somatic cells from the aPSCs was disrupted (Tasaki et al., 2011b). During regeneration in D. japonica, aPSCs supply all types of differentiated cells, including cells that secrete growth factors (Hayashi et al., 2011; Yazawa et al., 2009). This leads us to hypothesize that, after amputation, newly differentiated cells are also required for maintaining activated JNK signaling during regeneration, and this activated JNK signaling might regulate the expression level of plac8-A during regeneration (Fig. 7). It appears likely that U0126 treatment blocked the supply of the cells needed to maintain the activation of JNK signaling after amputation, and consequently the reduction of plac8-A expression and induced hyperproliferation were prevented. Thus, we propose that both JNK and ERK signaling might be required to maintain the induced hyperproliferation during regeneration.

So far, induced hyperproliferation has been thought to be a unique feature of aPSCs for enlarging the aPSC population at the onset of the regeneration process (Wenemoser and Reddien, 2010). However, our results strongly suggest that induced hyperproliferation is an indispensable event for rapid and proper supply of differentiated cells during regeneration. Losing body part(s) leads immediately to various kinds of reactions, including wound healing, rearrangement of body polarity and so on in the remaining tissues (Gurley et al., 2010; Kato et al., 2001). Induced hyperproliferation is regarded as the initial reaction of aPSCs during regeneration. Here, we revealed a molecular mechanism involving transition of the cellular state through induced hyperproliferation. We propose that plac8-A acts as a switch to shift the state of aPSCs from steady state to active state in early regeneration (Fig. 7).

plac8-A is expressed in almost all piwiA-expressing aPSCs, except in the nanos⁺ aPSC subpopulation and the piwi-1⁺ aPSC subpopulation (Fig. 1C; Fig. S2). Thus, decreasing plac8-A expression can induce a qualitative state change in almost the whole stem cell pool. Though the above two Plac8-A⁻ populations should be further characterized, it is considered that nanos⁺ aPSCs are germline-specified stem cells, not somatic cell-specified stem cells, based on previous reports showing a relationship between nanos and germline differentiation (Sato et al., 2006; Wang et al., 2007). It is interesting that a simple mechanism of regulation by one gene can activate almost the whole aPSC population, which consists of several subsets with molecular heterogeneity. Considering this, we can expect that plac8-A is upstream in the molecular pathway of aPSC regulation and thus can influence a variety of consequences, including proliferation and differentiation. Therefore, suppression of plac8-A might be a rapid strategy to regulate the participation of as many and as varied aPSCs as possible for regeneration. A recent report showed a close relationship between cell cycle regulation by factors such as Cyclin and CDKs and proliferation or differentiation of stem cells. In particular, it was reported that specification or differentiation of stem cells occurred during unusual cell cycling such as increasing cell division or expanded G₁ phase (Liu et al., 2019). Therefore, the results of this study strongly suggest a qualitative change of the aPSC state through cell cycle control by the plac8-A-pathway.

We revealed that a decrease of plac8-A expression is required to induce expression of a transcription factor for differentiation (Fig. 6F). This suggested that a qualitative state change to an active state of aPSCs by decreasing plac8-A expression is indispensable for normal regeneration. In hematopoietic stem cells (HSCs), an activated HSC population is increased by upregulating the CD34 mRNA level to produce differentiated hematopoietic cells after injury (Wilson et al., 2008). Likewise, the active state of plac8-A⁻ aPSCs is essential for supplying a sufficient differentiated cell supply by promoting differentiation.

In future studies, it will also be important to examine the function of plac8-A in the steady state or return to the steady state from the active state after regeneration through recovery of the expression level of plac8-A. As plac8-A is expressed in almost all aPSCs in intact animals, it is possible that there is a relationship between plac8-A expression and the constant differentiated cell supply for homeostasis. Also, our results suggested that plac8-A represses acceleration of the cell cycle in the steady state. Thus, recovery of the expression of plac8-A at the end of regeneration is important for precisely balancing the differentiated cell supply and demand in order to avoid abnormal regeneration. Understanding this kind of mechanism will be helpful in the medical field, including applications of stem cell therapy. Transplantation of stem cells in the activated state would appear to be useful for increasing the efficiency of stem cell-based therapy, because activated stem cells would proliferate rapidly in vivo and respond faster to environmental stimuli without needing to undergo the process of in vivo activation. Therefore, in-depth studies of the molecular mechanisms regulating heterogeneous stem cell populations will be important for not only stem cell research but also the medical field.

A clonal strain of the planarian D. japonica [sexualizing special planarian (SSP) (2n=16); Shibata et al., 2012] was maintained at 24°C in 0.005% artificial sea water (Instant Ocean). Chicken liver was fed every 1 or 2 weeks to the planarians to maintain them. Planarians had been starved for at least 1 week before all experiments. Regenerating planarians used for experiments were obtained by amputation anterior or posterior to the pharynx. Three or five groups of biological replicates were tested for statistical analyses.

Animals placed on wet filter paper on ice were irradiated with 120 roentgens of X-rays using an X-ray generator (SOFTEX B-5). Planarians were used for experiments 5 days after irradiation.

Double-stranded RNA was synthesized as previously described (Rouhana et al., 2013). The primers for PCR amplification were as follows: SP6+ T7 forward primer (for plac8) 5′-GATCACTAATACGACTCACTATAGGGCAAGCTATTTAGGTGACACTATAG-3′; Zap Linker+ T7 forward primer (for P2X-A) 5′-GATCACTAATACGACTCACTATAGGGCTGCAGAATTCGGCACGAGG-3′; M13 reverse primer 5′-GTTTTCCCAGTCACGACGTTGTAA-3′.

RNA interference (RNAi) with dsRNA was performed to knock down the target genes as follows: 25 μl of chicken liver solution (liver homogenate:culture water=1:1), 7 μl of 2% agarose and 7 μl of 2.0 μg/μl dsRNA were mixed and fed to 15 planarians. The mixture was frozen at −30°C for at least 30 min before use. Three successive feedings were similarly conducted at 3-day intervals after the first feeding. Control animals were fed with dsRNA of EGFP.

dsRNA synthesized as described above (2 µg/µl) was injected into the intestine. Three successive injections were performed daily after the first injection. Control animals were injected with dsRNA of EGFP.

For production of anti-Plac8-A antibody, peptides corresponding to three parts of Plac8-A were synthesized and injected into rabbits. Affinity-purified polyclonal antibody was obtained from the rabbit sera. All procedures were conducted by MBL. The amino acid sequences of the peptides were MNENKRYSNKLDYSQEC, AAEPILQQPPEYPGFPKC and IQQPKSNTGSAREWSSGC. For antibody validation see Fig. S2.

Ten planarians were dissolved in sample buffer [100 mM Tris-HCl (pH 6.8), 4% sodium dodecyl sulfate (SDS), 12% β-mercaptoethanol, 20% glycerol, 0.024% Bromophenol blue] and boiled for 5 min. After SDS-PAGE, samples were transferred to a Hybond P membrane, and stained with 1/800 diluted anti-Plac8-A antibody or 1/5000 diluted anti-α-tubulin (Sigma-Aldrich, T9026) as the primary antibody. Then the membrane was incubated with 1/5000 diluted secondary antibody (Cytiva, RPN420). Signals were detected using SuperSignal West Dura Extended Duration Substrate (Pierce).

FBSC-PCR was performed as previously described (Hayashi et al., 2010). The forward (FW) and reverse (RV) primer sets used were (5′-3′): G3PDH (internal control) FW ACCACCAACTGTTTAGCTCCCTTAG, RV GATGGTCCATCAACAGTCTTTTGC; plac8-A FW AAGAGCAACACAGGTAGTGCTAGGGAGTG, RV AGAAGCACAACAACATTCACCATATCGTG; piwiA FW CGAATCCGGGAACTGTCGTAG, RV GGAGCCATAGGTGAAATCTCATTTG.

Planarians were fixed with 4% paraformaldehyde/5% methanol in 5/8 Holtfreter's solution for 30 min at room temperature after removing mucus using 2% HCl in 5/8 Holtfreter's solution. Then, samples were bleached with 6% H₂O₂ overnight at room temperature under fluorescent light. Bleached samples were treated with 50% xylene/methanol for 30 min at 4°C and rinsed with 100% ethanol for 30 min at 4°C. Then, the samples were rehydrated through a graded ethanol series (75%, 50% and 25%) in 5/8 Holtfreter's solution each for 30 min at 4°C. The rehydrated samples were rinsed with Triton-PBS (TPBS; 2.7 mM KCl, 8.1 mM Na₂HPO₄・12H₂O, 136.9 mM NaCl, 1.5 mM KH₂PO₄, 0.1% Triton X-200) for 30 min at 4°C. Permeabilization was performed with 5 μg/ml Proteinase K in TPBS for 12 min at 37°C. Next, samples were post-fixed in 4% paraformaldehyde/5% methanol in 5/8 Holtfreter's solution for 30 min and rinsed with TPBS at 4°C. After blocking using 10% goat serum in TPBS, the samples were incubated in 10% goat serum in TPBS containing primary antibody overnight at 4°C. The dilution of primary antibody used was 1/800 for anti-Plac8-A, 1/500 for anti-PiwiA (Yoshida-Kashikawa et al., 2007) and 1/200 for anti-pH3 (Upstate, 06-570). The samples were washed with TPBS several times at room temperature and incubated in 10% goat serum in TPBS containing 1/1000 fluorescent-labeled secondary antibody (Alexa Fluor 594 or Alexa Fluor 488; Molecular Probes) and 1 μg/ml Hoechst 33342 (Calbiochem) overnight at 4°C. The samples were observed using a confocal microscope (Fluoview FV10i; Olympus) or a fluorescence stereoscopic microscope (M205FA T-RC 1; Leica).

For RNA probe synthesis, the plasmid pCR^®II-TOPO containing the gene plac8-A or the plasmid pBluescript SK containing the gene runt-1 was used. The DNA linearized with Not1 at the 5′ end of the target gene was used as a template. Before the transcription, linearized DNA was purified using phenol/chloroform extraction and ethanol precipitation. Then the product was used as a template for antisense RNA transcription by Sp6 RNA polymerase (Promega) or T7 RNA polymerase (Fermentas). After transcription, the probe was purified using ethanol precipitation and stored at −80°C. For fixation, bleaching, permeabilization and post-fixation steps, the same procedures as used for whole-mount immunohistochemistry were employed. After post-fixation, the samples were soaked in hybridization solution for 1 h at 55°C and hybridized with Dig-labeled antisense RNA probe (which had been denatured previously for 20 min at 55°C) in hybridization solution for 36 h at 55°C. Then, the samples were washed with wash solution three times for 30 min and three times for 1 h at 55°C and rinsed with Buffer I [0.1 M maleic acid, 5× SSC, 0.1% Tween-20 (pH 7.5)] twice for 5 min each at room temperature. Then samples were treated with Buffer II [1% blocking reagent (Roche Diagnostics) in Buffer I] for blocking for 30 min at room temperature and treated with 1/2000 alkaline phosphatase-conjugated anti-Digoxigenin antibody (Roche Diagnostics, 1093274) in Buffer II overnight at 4°C. Then, the samples were rinsed with Buffer I six times for 30 min each at room temperature and washed with TMN [0.1 M Tris-HCl, 0.1 M NaCl, 50 mM MgCl₂ (pH 9.5)] solution twice for 5 min each at room temperature. A mixture of 3.5 μg/ml 5-bromo-4-chloro-3-indolyl phosphatase (Roche Diagnostics) and 1.8 μg/ml 4-nitro blue tetrazolium chloride (Roche Diagnostics) in TMN solution was used for detecting signals. After the detection, TE buffer [10 mM Tris-HCl, 1 mM EDTA (pH 8)] was used to stop the reaction and the samples were kept in TE buffer at 4°C.

All steps before detection of the signal were performed in the same way as for whole-mount in situ hybridization. To detect signals, a Tyramide Signaling Amplification kit (Molecular Probes TSA™ Kit #12, with HRP-goat anti-rabbit IgG and Alexa Fluor 488^® tyramide, Eugene) was used for color development.

Total RNA was extracted using ISOGEN-LS (Wako) as follows: 50 μl of ISOGEN-LS was added to samples (all fragments obtained by amputation were used for regenerating samples). The samples were homogenized and then 700 μl of ISOGEN-LS was added. Total RNA was extracted according to the manufacturer's instructions and cDNA was synthesized using a QuantiTect Reverse Transcription Kit^® according to the manufacturer's instructions (Qiagen). The synthesized cDNA was diluted (×20) and used for gene expression analysis by quantitative (q)RT-PCR. Ten microliters of RT-PCR mixture containing 1× QuantiTect SYBR green PCR master mix (Qiagen), 0.3 μM gene-specific FW/RV primers and 1 μl of diluted cDNA template was analyzed using an ABI PRISM 7900 HT (Applied Biosystems). The reactions were carried out as follows: 50°C for 2 min, 95°C for 15 min, 50 cycles of 95°C for 15 s, 60°C for 30 s, 72°C for 1 min. The FW and reverse RV primer sets used were (5′-3′): G3PDH (internal control) FW ACCACCAACTGTTTAGCTCCCTTAG, RV GATGGTCCATCAACAGTCTTTTGC; pcna FW ACCTATCGTGTCACTGTCTTTGACCGAAAA, RV TTCATCATCTTCGATTTTCGGAGCCAGATA; mcm2 FW CGCTGTTGGACAAGGTCAGAAGAATGAACA, RV CCAGAAACACAAATCTACATCTTCCAAAGG; plac8 FW AAGAGCAACACAGGTAGTGCTAGGGAGTG, RV AGAAGCACAACAACATTCACCATATCGTG; P2X-A FW GATTTCAACAATGGAATGAATTTTAGATA, RV AAAATGTGAAACAAGTAGCAGGATCA; runt-1 FW CGGCCATCGAGTATGGTTAT, RV ACGGCAACAATGTTTGGATT.

JNK inhibitor SP600125 (Sigma-Aldrich) and MAPK⁄ERK kinase (MEK) inhibitor U0126 (Cell Signaling Technology) were dissolved in dimethylsulfoxide (DMSO) at 100 mM and 10 mM, respectively. Amputated planarians were treated with 5 μM SP600125 from 12 h after amputation until the indicated period of regeneration for optimal JNK inhibition, or with 25 μM SP600125 or U0126 at 4 h after amputation for strong inhibition of JNK or ERK, respectively, for the indicated period of regeneration for each experiment.

Standard deviations and paired two-tailed Student’s t-tests were calculated using Microsoft Excel.

We thank Shigehiro Kuraku for helping us to perform FACS-based single cell RT-PCR. We are grateful to Elizabeth Nakajima for critical reading of the manuscript. We also thank all of our laboratory members for their help and encouragement.

The peer review history is available online at https://journals.biologists.com/dev/article-lookup/doi/10.1242/dev.199449.