Caenorhabditis elegans PTR/PTCHD PTR-18 promotes the clearance of extracellular hedgehog-related protein via endocytosis

Spatiotemporal restriction of signaling plays a critical role in animal development and tissue homeostasis. All stem and progenitor cells in newly hatched C. elegans larvae are quiescent and capable of suspending their development until sufficient food is supplied. Here, we show that ptr-18, which encodes the evolutionarily conserved patched-related (PTR)/patched domain-containing (PTCHD) protein, temporally restricts the availability of extracellular hedgehog-related protein to establish the capacity of progenitor cells to maintain quiescence. We found that neural progenitor cells exit from quiescence in ptr-18 mutant larvae even when hatched under starved conditions. This unwanted reactivation depended on the activity of a specific set of hedgehog-related grl genes including grl-7. Unexpectedly, neither PTR-18 nor GRL-7 were expressed in newly hatched wild-type larvae. Instead, at the late embryonic stage, both PTR-18 and GRL-7 proteins were first localized around the apical membrane of hypodermal and neural progenitor cells and subsequently targeted for lysosomal degradation before hatching. Loss of ptr-18 caused a significant delay in GRL-7 clearance, causing this protein to be retained in the extracellular space in newly hatched ptr-18 mutant larvae. Furthermore, the putative transporter activity of PTR-18 was shown to be required for the appropriate function of the protein. These findings not only uncover a previously undescribed role of PTR/PTCHD in the clearance of extracellular hedgehog-related proteins via endocytosis-mediated degradation but also illustrate that failure to temporally restrict intercellular signaling during embryogenesis can subsequently compromise post-embryonic progenitor cell function.

Introduction C. elegans L1 larvae hatch out of their eggshells with quiescent stem and progenitor cells. Sufficient food supply initiates L1 development by coordinating the release of stem and progenitor cells from quiescence [1][2][3][4]. Conversely, under nutritionally unfavorable conditions, the newly hatched larvae enter developmental dormancy, called L1 arrest or L1 diapause, and survive for weeks until ample food becomes available [5]. Because of the ease of manipulating and tracking their quiescence and reactivation, stem and progenitor cells in C. elegans L1 larvae have served as excellent models to study the nutritional regulation of stem cells in vivo.
Previous studies have shown that the insulin/insulin-like growth factor signaling (IIS) pathway plays a critical role in the developmental decision in response to nutrient availability. For example, loss of daf-16/foxo, which results in the constitutive activation of the IIS pathway, causes unwanted reactivation of many types of somatic progenitor cells, such as the P neuronal and M mesodermal progenitor cells [2,6]. MicroRNA (miR)-235 acts partly downstream of the IIS pathway to regulate a subset of L1 developmental events during L1 arrest [7]. Expression analysis and rescue experiments suggested that the miRNA primarily acts in the hypodermis to suppress the reactivation of P and M progenitor cells [7]. These findings led us to identify grl-5 and grl-7 as target genes of miR-235 that mediate the non-autonomous regulation of P cell quiescence by the hypodermis [8]. Both grl-5 and grl-7 encode putative secreted proteins and belong to the hedgehog-related (hh-r) gene family, which has been proposed to have evolved from the same ancestral gene as hedgehog (Hh) in other animals [9]. Inhibition of grl-5 and grl-7 can partially suppress the inappropriate reactivation of P cells in starved mir-235 larvae, suggesting that these grl genes promote cellular events.
Several genes in the Hh signaling pathway, such as Hh, Smoothened, Cos2, Fused, and Suppressor of Fused are absent in the C. elegans genome [10]. However, the nematode possesses two patched orthologs, ptc-1 and ptc-3 [11]. Knockdown of ptc-3 and multiple hh-r genes results in molting defects, raising the possibility that these genes act together in the same genetic pathway [12,13]. In addition to hh-r and ptc genes, some of the ptr (patched-related) genes were also shown to result in similar molting defects [12,14]. PTR, also called patched domain-containing (PTCHD) proteins, are found in other species such as Drosophila, mouse, and human [11,15]. PTR/PTCHD contains a region that is conserved among the Niemann-Pick Type C proteins, which are involved in cholesterol transport, the Dispatched protein, which promotes the secretion of Hh protein, and the Hh receptor Patched [16][17][18][19]. This membrane-spanning region contains the "sterol-sensing domain," which is involved in binding to and sensing cholesterol [19,20]. In addition, PTR/PTCHD, NPC1, Dispatched, and Patched proteins belong to the resistance, nodulation, and division (RND) transporter superfamily [21]. Members of the superfamily generally contain a 12-transmembrane (TM) domain, which is thought to have arisen from an intragenic duplication of a six-transmembrane domain [22], and a conserved GXXXD motif, which has been shown to play a critical role in the bacterial RND transporter activity [23,24]. Furthermore, Patched, Dispatched, and PTR/PTCHD proteins contain not only an expanded GXXXDD motif within TM4 but also a GXXXD/E motif within TM10 [13]. Introduction of mutations in Patched GXXXDD motif and in Dispatched GXXXDD and GXXXD/E motifs impairs their activities [13,[25][26][27].
Although the ptr/ptchd genes are conserved among several animal species, little is known about their cellular functions. Mutations in the human ptchd1 gene are found in patients with autistic spectrum disorders and learning disabilities [28][29][30][31][32][33][34]. Furthermore, ptchd1 knockout mice show attention-deficit hyperactivity disorder (ADHD)-like phenotypes [35]. Loss of the Drosophila ptr gene results in embryonic lethality [36]. In addition to their involvement in molting, the functions of few C. elegans ptr genes have been reported in detail. For instance, daf-6 is involved in the formation of the glial channel that surrounds the receptive endings of the sensory neurons and likely regulates vesicular transport [37][38][39][40]. ptr-24 has been proposed to act downstream of the hh-r gene, grl-21, to regulate mitochondrial fragmentation and lipid accumulation [41]. Additionally, wrt-10, which belongs to another subfamily of hh-r genes, reportedly promotes oocyte quality maintenance and delays reproductive decline via ptc-1 and ptr-2 [42].
Here, we show that C. elegans PTR-18 promotes the clearance of extracellular Hh-related proteins via endocytosis-mediated degradation, potentially acting as its decoy receptor. Under nutrient-deficient conditions that force the wild-type larvae to enter L1 arrest, newly hatched ptr-18 mutant L1 larvae show reactivation of P progenitor cells. This arrest-defective phenotype is suppressed by the inhibition of a particular set of grl genes, including grl-5, grl-7, and grl-27. Unexpectedly, analysis using reporter genes showed that neither PTR-18 nor GRL-7 were expressed in newly hatched larvae. Instead, these proteins are temporally localized along the periphery of the apical membranes of hypodermal and P neuronal progenitor cells during late embryogenesis and are subsequently targeted to lysosomal degradation before hatching. This temporally controlled clearance of GRL-7 requires activity of PTR-18, so that newly hatched ptr-18 mutant larvae still exhibit extracellular GRL-7 accumulation. Furthermore, the GXXXDD motif within TM4, the GXXXD/E motif within TM10, and the cytoplasmic C-terminal portion of PTR-18 protein are indispensable for its appropriate function. These findings reveal a previously undescribed function of PTR/PTCHD as a sink for extracellular Hh-related proteins and illuminate the importance of the temporal regulation of extracellular signaling in maintaining progenitor cell function.

ptr-18 is required to maintain the quiescence of progenitor cells during L1 arrest
As shown in Fig 1A, six pairs of P neural progenitor cells first reside along the ventrolateral sides in newly hatched larvae. When the larvae are supplied with ample food, the most anterior pair of quiescent P cells migrate into the ventral nerve cord during the mid-L1 stage, followed successively by the more posterior pairs. This reactivation of quiescent P cells is easily detected under a differential interference contrast microscope [43]. Our previous studies showed that forced expression of the hh-r gene, grl-7, in starved L1 larvae can reactivate P neuroblasts [8]. In addition, grl-7 and another hh-r gene, grl-5, partially mediate reactivation of P neuroblasts Drawings were made based on https://www.wormatlas.org/hermaphrodite/hypodermis/Hypframeset. html. Top: A newly hatched L1 larvae with quiescent P cells (indicated in magenta). Anterior is to the left, and dorsal is up. Bottom: A part of the body of developing L1 larvae. P cell reactivation is detected by their ventral migration. (B) Percentage of starved L1 larvae showing �1 reactivated P cell(s). P cells that had migrated to the ventral nerve cord after 5 days of L1 starvation were scored as "reactivated." Experiments were repeated �4 times, and n�35 animals were scored for each experiment. Data are presented as mean ± SD. ���� : p <0.0001 (Fisher's exact test). (C) Differential interference contrast photographs of wild-type and ptr-18(ok3532) L1 larvae after 5 days of L1 starvation. Only the cell bodies of motor neurons (arrows) were observed along the in starved mir-235 mutant L1 larvae [8]. RNAi targeting some of the hh-r and ptr genes results in similar developmental defects, such as failure to complete molting and small body size [12][13][14], suggesting that these genes act in the same genetic pathway. Similarly, previous studies have proposed that grl-21 negatively regulates ptr-23 to regulate mitochondrial fragmentation and lipid accumulation [41]. Thus, we hypothesized that the ptr gene may be involved in maintaining the quiescence of P cells during L1 arrest by antagonizing the activity of grl-5 and grl-7. We found that among the available ptr mutants, the majority of ptr-18 mutant larvae showed reactivation of P cells when starved after hatching (Fig 1B and 1C). Further, we conducted RNAi against ptr -2, 4, 6, 9, 11, 13, 14, 16, 17, 19, 20, 22, 23, and 24, and none of the starved L1 larvae from RNAi-treated mothers showed abnormal P cell reactivation (50 animals were scored; n = 1). Since ptr-18 mutant animals exhibit a P cell defect at relatively high penetrance, we decided to focus our analysis on the ptr-18 gene. In addition, reactivation of the mesoblast M cell and molting were observed in starved ptr-18 mutant larvae ( Fig 1D). As observed in mir-235 mutant animals [7], animals that had undergone M cell proliferation always show migrated P cell(s), whereas molted animals always harbor reactivated P and M cells (Fig 1E), suggesting that P cells, M cell, and molt become activated in this strict order in starved ptr-18 animals. In contrast, the primordial germ cells, Z2 and Z3, remain quiescent in ptr-18 mutant larvae after 5-day L1 starvation (S1A Fig), similar to the daf-16/foxo and mir-235 mutant animals [3,7]. When starved in cholesterol-and ethanol-free complete S medium after hatching, most of the daf-16/foxo-null mutant animals could not survive beyond 10 days (S1B Fig). In contrast, ptr-18 mutant animals were relatively resistant to starvation stress, similar to the wild-type animals (S1B Fig). Activities of grl-5, grl-7, and grl-27 are required for P cell reactivation in starved ptr-18 mutant animals Given the similarity of GRLs to Hh and PTR-18 to PTC, we examined whether GRL-5 and GRL-7 would act through PTR-18 or independently. To this end, we tested whether the activities of grl-5 and grl-7 contribute to the inappropriate reactivation of P cells in the starved ptr-18 mutant larvae. Deletion mutations of these grl genes were introduced in ptr-18 mutant animals. Unexpectedly, we found that the inhibition of both grl-5 and grl-7 almost completely suppressed the phenotype in the ptr-18 mutant larvae, suggesting that ptr-18 acts upstream of, but not downstream of, these grl genes ( Fig 1F). Previous studies have shown that the inhibition of grl-5 and grl-7 activities only partially suppresses the defect in mir-235 mutant larvae [8]. Thus, these observations suggest that the reactivation of P cells in ptr-18 mutant animals is heavily dependent on the activity of these grl genes. Previous studies using reporter genes suggested that in addition to grl-5 and grl-7, several other grl genes are expressed in the P and hypodermal cells [44]. Strikingly, the inhibition of grl-27 comparably suppressed the phenotype in ptr-18 mutant animals, similarly to grl-5 and grl-7 ( Fig 1F). In contrast, the elimination of grl -2, grl-4, grl-5, grl-10, grl-15, grl-17, and grl-21 activities did not significantly affect the defect (Fig 1F). These findings suggest that ptr-18 antagonizes the activity of a specific subset of grl genes among those expressed in the P and hypodermal cells. Because grl-5, grl-7, and grl-27 are required for P cell activation in starved ptr-18 mutant animals, these grl genes might also play a critical role in the exit of P cells from quiescence in well-fed wild type animals. However, the triple mutant animals of grl-5, grl-7, and grl-27 did not exhibit a delay in the timing of P cell activation under the fed condition (S1C Fig). These findings suggest that these grl genes do not contribute to P cell reactivation under the fed condition. Alternatively, an additional hh-r gene may act together with these grl genes.

Spatiotemporal dynamics of PTR-18::GFP reporter expression
To elucidate the expression pattern of ptr-18, we constructed a GFP translational reporter by inserting the gfp gene into the open reading frame of the ptr-18 gene in the fosmid WRM0613dH03.1. This fosmid ptr-18 reporter gene was introduced as an extrachromosomal array. Expression of PTR-18::GFP was bright enough for live imaging, though it should be noted that genes in the extrachromosomal array generally tend to be overexpressed. PTR-18:: GFP was first detected along the apical side of surface cells that cover the whole body, which consists of hypodermal, seam, and P cells at the 3-fold stage during embryogenesis (Fig 2A; also see S4 Fig below). To determine the localization pattern of PTR-18::GFP in 3-fold embryos in detail, we dissolved the gravid, transgenic animals to harvest early embryos and allowed them to grow synchronously. Transition of the population ratio of 2-fold embryos, 3-fold embryos, and L1 larvae from 9 to 16 h after harvest confirmed the developmental synchronization of the collected embryos (S2A Fig). During the early time point, the majority of the 3-fold embryos initially showed apical localization of PTR-18::GFP (Fig 2A and 2B). As development continued, the percentage of the apical localization decreased, the fraction that exhibited localization of PTR-18::GFP in the vesicular punctate structures increased, and eventually, most of the embryos did not show its expression (Fig 2A and 2B). These observations indicate that PTR-18::GFP first localizes at the apical side of the hypodermal, seam, and P cells, subsequently accumulates in the vesicular structures, and eventually disappears before hatching. Unexpectedly, despite the robust phenotypes of ptr-18(ok3532) animals during L1 arrest, faint expression of PTR-18::GFP was only occasionally observed in the excretory duct and G1 pore cells during L1 starvation ( Fig 2C). Expression of PTR-18::GFP only reappeared along the apical surface of the hypodermal, seam and P cells 11 h after the L1-arrested larvae were fed, which was several hours after P cell reactivation was initiated (S2B Fig). These observations raise the possibility that ptr-18 acts before hatching to regulate the quiescence of P cells. PTR-18::GFP was also expressed in the descendants of P cells (S2B Fig; arrows

ptr-18 acts in P and hypodermal cells
There is an open reading frame, Y38F1A.4, within the third intron of the ptr-18 gene. However, the expression of ptr-18 cDNA::venus fusion gene under the control of its native promoter almost completely suppressed the inappropriate reactivation of P cells in starved ptr-18 mutant larvae, indicating that the observed defects were caused by the reduced activity of ptr-18 and not of Y38F1A.4 ( Fig 2D).
To determine the site of ptr-18 action, ptr-18 cDNA::venus was expressed under the control of promoters whose activities were specific to the cells and tissues where the ptr-18::gfp was expressed (S2D Fig). When ptr-18 cDNA::venus was expressed under the control of the dpy-7 promoter, which is active in hypodermal, part of seam, and P cells [45], it rescues the P cell activation defect of ptr-18 mutant animals as efficiently as ptr-18 cDNA::venus driven by the native promoter ( Fig 2D). In contrast, the phenotype was hardly affected when ptr-18::venus Photographs were taken with fixed long and short exposure times, as indicated by "long" and "short." Note that apical PTR-18::GFP distribution can be visualized by the short exposure, whereas vesicular PTR-18::GFP patterns become only conspicuous by the long exposure. PTR-18::GFP localized in either the apical side of surface cells that cover the whole body, which consists of hypodermal and P cells (left images; see also S4 Fig), or intracellular vesicular structures (middle was expressed under the control of G1 pore-specific lin-48 [46] and excretory duct-specific dct-5 [47]. Furthermore, the expression of ptr-18::venus in either P or hypodermal and some seam cells driven by the hlh-3 promoter [48] and Q system [49], respectively, could efficiently suppress the phenotype (Figs 2E and S2E-S2G), suggesting that ptr-18 acts both autonomously and non-autonomously to maintain the quiescence of P cells. Because PTR-18::GFP was detected in the hypodermal, seam, and P cells only during the 3-fold stage before P cells become reactivated (Fig 2A and 2B; see also S4 Fig below), ptr-18 acts before hatching to regulate the quiescence of P cells.

GRL-5 and GRL-7 reporter proteins show expression patterns similar to that of PTR-18::GFP
Although the loss of grl-5, grl-7, and grl-27 can suppress the defects in ptr-18 mutant animals, whether PTR-18 and these GRL proteins act in spatial and temporal proximity remains undetermined. To assess the interaction between ptr-18 and grl genes, we constructed a grl-7:: mcherry::3xflag gene by inserting the mcherry::3xflag tag into the grl-7 genomic region in the fosmid WRM0615cE01 (see Materials and Methods). Similar to PTR-18::GFP, GRL-7::mCherry::3xFLAG was first detected in the 3-fold embryos ( Fig 3A). Previous studies have shown that grl-7 encodes a protein with a predicted signal sequence at the N-terminus, and its transcriptional reporter genes are expressed in hypodermal, seam and P cells [44]. Consistently, GRL-7::mCherry::3xFLAG localized along the apical side of the surface cells that cover the whole body as well as in the intracellular structures of these cells (Fig 3A; also see S4 Fig for details). As observed for PTR-18::GFP, the majority of the GRL-7::mCherry::3xFLAG embryos initially showed an apical distribution. However, as the embryos neared hatching, internal, vesicular localization became predominant (Figs 3B and S3A). After hatching under the feeding condition, apical localization of GRL-7::mCherry::3xFLAG remained undetectable until 11 h post-feeding of the L1-arrested larvae (S3B Fig). To further define the expression patterns of GRL-7, the mcherry tag was introduced before the stop codon of the grl-7 gene using the CRISPR/Cas9 system, and its expression patterns were analyzed via super-resolution confocal microscopy. We first confirmed that GRL-7::mCherry was removed from the apical side of the cell before hatching by comparing its fluorescence intensity along the apical side of the cell and inside the cell, which is marked by cytoplasmically localized GFP::RAB-7, (Fig 3C and 3D; also see below). As predicted by the presence of N-terminal signal sequence [9], the apically localized GRL-7::mCherry does not overlap with cytoplasmic GFP::RAB-7, suggesting that GRL-7 was secreted ( Fig 3C). During this analysis, we noticed that GRL-7::mCherry-positive vesicular structures were found in hypodermal, seam, and P cells in newly hatched larvae (Fig 3E). To images). The top panel is a magnified view of the image below (indicated by the rectangle), and the apical border of cells is indicated by arrowheads. A fraction of the 3-fold embryos carrying the transgene, marked by mCherry expression driven by the ptr-18 promoter, did not show detectable expression of PTR-18::GFP (right images). Scale Bar: 10 μm (wholeembryo image) and 1 μm (magnified image). (B) Percentage of animals showing the indicated expression patterns were scored at the indicated times after harvesting early embryos. Data are presented as mean ± SD. Experiments were repeated three times, and �35 animals were scored for each time point. (C) Faint expression of PTR-18::GFP was detected in the excretory duct (arrowhead) and G1 pore (arrow) cells in L1 larvae after 24 h L1 starvation. Scale Bar: 10 μm. (D) Effects of PTR-18::VENUS expression on inappropriate reactivation of P cells in starved ptr-18(ok3532) mutant larvae. PTR-18::VENUS was expressed under the control of the promoters indicated (see S2D Fig). Animals that had lost the array expressing PTR-18::VENUS under the control of the lin-48 promoter (line #1) were used as control. (E) Expression of PTR-18::VENUS in P or hypodermal cells restored the defect in maintaining P cell quiescence. PTR-18:: VENUS was expressed in P or hypodermal cells using the hlh-3 promoter or Q system, respectively (S2E- S2G Fig). YB3808 strain expressing mcherry under the control of the dpy-7 promoter was used as control. For D and E, data are presented as mean ± SD. Experiments were repeated three times, and �35 animals were scored for each experiment. � : P<0.05, �� : P <0.01, and ���� : p <0.0001 (Fisher's exact test). https://doi.org/10.1371/journal.pgen.1009457.g002

PLOS GENETICS
Clearance of extracellular hedgehog related protein by C. elegans PTR/PTCHD PTR-18 via endocytosis
In general, GFP and its derivatives are sensitive to acid quenching in lysosomes. In contrast, RFP and its derivatives are relatively acid tolerant and resistant to lysosomal enzymes [55]. Similar to mCherry-fused GRL-7 reporter proteins, GRL-7::VENUS driven by the native promoter was found to exhibit apical localization in wild-type 3-fold embryos ( Fig 7C). However, GRL-7::VENUS was undetectable after hatching ( Fig 7D). These observations are consistent with the above data, suggesting that GRL-7 is internalized via endocytosis before hatching (Figs 5 and 6). Furthermore, most of the newly hatched ptr-18 L1 larvae exhibited detectable levels of GRL-7::VENUS (Fig 7C and 7D), suggesting that the loss of ptr-18 causes a significant delay in GRL-7 endocytosis. Although the difference in these expression patterns between wild-type and ptr-18 mutant animals is obvious, it remains possible that this was caused by the overexpression of the grl-7 reporter from extrachromosomal arrays. To exclude this possibility and obtain more quantitative insights into the regulation of GRL-7 protein levels by ptr-18, we introduced a venus::3xflag tag in front of the stop codon of the grl-7 gene using the CRISPR/ Cas9 system. Although the GRL-7::VENUS::3xFLAG expression was too faint for live cell imaging, immunoblot analysis showed that the level of tagged GRL-7 protein was upregulated in newly hatched ptr-18 L1 animals compared to that in wild-type animals ( Fig 7E). Thus, these observations explain why the loss of ptr-18 causes developmental defects in a grl-7dependent manner under the post-hatch starved condition, although neither PTR-18 nor GRL-7 protein was detected in newly hatched wild-type larvae. The untimely extracellular presence of GRL-7 due to the absence of PTR-18 will lead to the re-activation of the P cell irrespective of the dietary environment.  Although we cannot exclude the possibility that similar to GRL-7, ptr-18 also promotes timely internalization of GRL-5, the above observations suggest that PTR-18 dependent uptake is not the major route, if at all.
Previous studies have shown that both daf-16/foxo and mir-235 mutant larvae fail to maintain the quiescence of multiple progenitor cells [2,6]. The temporally regulated internalization of GRL-7::mCherry::3xFLAG was not affected in newly hatched daf-16/foxo and mir-235 L1 larvae, suggesting that these genes and ptr-18 regulate L1 arrest through distinct mechanisms (S7A Fig). mir-235 downregulates grl-7 via the miR-235 target site on the 3'UTR of grl-7 mRNA [8]. In contrast, our findings suggest that ptr-18 suppresses grl-7 via endocytosis-mediated degradation, raising the possibility that ptr-18 and mir-235 regulate the quiescence of P cells in genetically parallel pathways. Consistent with this, the loss of mir-235 in ptr-18 mutant animals significantly enhanced the quiescent defective phenotype (S7B Fig).
These findings suggest that the internalization and subsequent lysosomal degradation of GRL-7 before hatching play critical roles in establishing the capacity of neural progenitor cells to maintain quiescence under starved conditions.
Altogether, we conclude that ptr-18 temporally restricts the activity of GRL-7 within 3-fold stage of embryogenesis by facilitating its internalization and subsequent lysosomal degradation. Defects in this temporal restriction cause prolonged accumulation of extracellular GRL-7 beyond hatching, which compromises the capacity of neural progenitor cells to respond to nutritional stress (Fig 8D). Data were collected after 24 h L1 starvation and are presented as mean ± SD. Experiments were repeated three times, and n �50 animals were scored for each trial. ���� : p <0.0001 (Fisher's exact test). (E) GRL-7 reporter protein is upregulated in newly hatched ptr-18 mutant animals. Lysates from three biological samples were examined for newly hatched wild-type and ptr-18 mutant larvae. GRL-7::VENUS::3xFLAG was detected by anti-GFP antibody. The specificity of the antibody was confirmed using the lysate from non-transgenic wild-type animals on the right lane. Intensity of GFP bands were normalized by those of the corresponding tubulin bands for quantitation. � : P <0.05 (unpaired two-sided t-test). https://doi.org/10.1371/journal.pgen.1009457.g007

PLOS GENETICS
Clearance of extracellular hedgehog related protein by C. elegans PTR/PTCHD PTR-18 via endocytosis

ptr-18-dependent restriction of GRL activity allows neural progenitor cells to anticipate nutritional stresses before hatching
In this study, we first found that one of the C. elegans ptr/ptchd orthologs, ptr-18, is required to prevent P neuronal progenitor cells from undergoing unwanted reactivation when newly hatched larvae encounter starvation conditions. This reactivation requires the activity of grl-5, grl-7, and grl-27 hh-r genes but not other grl genes known to be expressed in the hypodermal and P cells. While PTR-18 first begin to accumulate along the apical cell membrane of hypodermal, seam and P cells at the late embryonic stage, GRL-7 becomes enriched at the specific regions of the cuticle, which are probably annular furrows. PTR-18 is subsequently internalized slightly earlier than GRL-7. However, both proteins eventually together populate endosomal and lysosomal compartments, resulting in the clearance of their apically localized fractions before hatching. Loss of ptr-18 activity causes significant delay in this endocytosismediated degradation of GRL-7, such that newly hatched ptr-18 mutant larvae still exhibit extracellular GRL-7 accumulation. Furthermore, the potential transporter activity of PTR-18 is required for its appropriate function. These findings suggest that ptr-18 temporally limits the activity of GRL-7 to establish the capacity of neural progenitor cells to maintain quiescence in response to nutritional stresses and also provide unique insights into the cellular role of PTR/PTCHD in promoting the clearance of extracellular Hh-related protein by targeting it to lysosomal degradation.

Potential role of PTR-18 as a decoy receptor for GRL-7
Our studies suggest that PTR-18 temporally restricts the availability of extracellular GRL-7 protein by targeting it to lysosomal degradation via endocytosis. Similarly, previous studies have shown that Hh receptor Patched sequesters Hh through endocytosis to limit the spatiotemporal range of its action in Drosophila and vertebrates [60][61][62][63]. In addition to sequestering Hh ligand, Patched also confers "ligand-independent antagonism" against Hh signaling by inhibiting Smo activity [60]. On the other hand, the reactivation of P cells observed, caused by the loss of ptr-18, was almost completely dependent on the activity of grl genes. Thus, in contrast to Patched, PTR-18 is unlikely to be coupled to the signaling components that act downstream of GRL-7. PTR-18 seems to act as its decoy receptor, whose function is only to sequester extracellular GRL-7 protein. This model is reminiscent of the D6 chemokine receptor, which has been proposed to act as a decoy receptor that is unfit for signaling, but can scavenge chemokines by constitutively delivering it to lysosomes [64,65] and whose loss results in prolonged inflammation due to impaired chemokine clearance [66]. This could explain the observation that the expression of ptr-18 in hypodermal or P cells can restore the quiescent defective phenotype of prt-18 mutant animals (Fig 2E). In these experiments, ptr-18 could be overexpressed under the control of the heterologous promoter from the extrachromosomal arrays, and when sufficient amount of PTR-18 protein is expressed in either type of cells, extracellular GRL-7 protein would be removed before hatching. In addition to PTR-18, other PTR/PTCHD proteins such as Drosophila PTR and the C. elegans DAF-6 reporter protein reportedly localize to unidentified intracellular vesicles [36,37,67]. Conversely, daf-6 mutations result in excessive accumulation of glial-secreted extracellular matrix in a channel formed by glial cells [68], which led to the model where daf-6 antagonizes the secretion of vesicles containing the matrix or promotes their uptake [69]. Although the involvement of hh-r genes in daf-6-dependent processes remains unexplored, reporters of some hh-r genes have been shown to be expressed in the glial socket and sheath cells, in which daf-6 restricts the channel size [9,38,44,70]. In contrast to the hh-r genes [9,44], a comprehensive reporter expression analysis of C. elegans ptr genes has not yet been reported. However, a genome-wide expression analysis showed that most of the ptr genes exhibit oscillatory expression patterns, similarly to hh-r genes [71]. Considering that C. elegans hh-r and ptr genes are extensively diverged [9,11,72], each PTR protein may further fine-tune the oscillatory activity of a particular set of Hh-r proteins via their endocytosis-mediated degradation.

Receptor for GRL-5, GRL-7, and GRL-27
In this study, we showed that ptr-18 acts upstream of the grl-5, grl-7, and grl-27 genes. Little is known about the genetic pathway that the hh-r genes act on. In Drosophila and vertebrates, Hh and sonic hedgehog (Shh) stimulate Hh signaling through their receptor, Patched [73,74]. Although C. elegans possesses two genes, ptc-1 and ptc-3, which encode patched orthologs, their relationship with hh-r genes remains ambiguous. However, RNAi targeting ptc-3 and several ptr and hh-r genes results in molting defects [12], suggesting that ptc-3 and at least some of the ptr and hh-r genes may participate in similar pathways [10]. Our attempt to test whether ptc-3 is involved in maintaining L1 arrest was hampered by the severe embryonic lethality caused by the loss of ptc-3. This technical difficulty will need to be overcome for further elucidation of the ptr-18-dependent regulation of L1 arrest.
Do PTR/PTCHD proteins in other animals also act as decoys for Hh? Overexpression of human PTR/PTCHD, PTCHD1, and PTCH53 (also called PTCHD4) proteins in Shh-responsive C3H10T1/2 cells suppressed Hh signaling [29,75]. In contrast, the expression of PTCHD1 in patched1 (ptch1)-deficient mouse embryonic fibroblasts did not repress the canonical Hh signaling pathway [76]. These observations raise the possibility that PTCHD functions upstream of Patched, possibly by acting as a decoy receptor for Shh. However, the expression of PTCH53 in the DAOY cancer cell line inhibited the upregulation of the Hh pathway via the Smo agonist, purmorphamine, raising the possibility that PTCH53 acts downstream of Smo [75]. In contrast to these in vitro studies, there is insufficient evidence that Drosophila PTR and mammalian PTCHD proteins impinge upon the Hh pathway in vivo. Unlike ptch1-deficient mice, ptchd1-knockout mice did not show overproliferation of neuronal precursors in the brain [77]. In vertebrates, a vertebrate-specific Hh-interacting protein 1 (Hhip1, also called Hip1) also functions together with Patched to antagonize Hh activity through its direct binding [78][79][80][81][82]. Thus, the loss of ptchd1 might be compensated by Hhip1 in vertebrates. In Drosophila, which does not have the Hhip1 ortholog, Hh is still internalized in the absence of Patched [63]. Thus, this patched-independent internalization might be mediated by PTR.

hh-r genes may promote the progression of larval development
We have shown that grl-5, grl-7, and grl-27 contribute to the reactivation of quiescent P cells in starved ptr-18 L1 larvae. Although we could not detect the expression of the grl-27 fosmid reporter gene, we observed the accumulation of extracellular GRL-5 and GRL-7 reporter proteins only prior to hatching and around L1 molting but not around the time when P cells initiate ventral migration. On the other hand, simultaneous loss of grl-5, grl-7, and grl-27 did not cause a significant delay in P cell reactivation in well-fed L1 larvae. These observations raise the possibility that there is an additional Hh-r protein that is induced by feeding earlier than GRL-5, GRL-7, and GRL-27 and activates P cell migration. How can ectopic activation of these grl genes cause P cell reactivation and later developmental events in starved ptr-18 and mir-235 mutant larvae? Starved ptr-18 and mir-235 animals occasionally molt. On the other hand, previous studies have shown that most ptr and hh-r genes show transcriptional oscillations with distinct phases during larval stages and that multiple hh-r and ptr genes as well as ptc-3 are implicated in molting [12,71,83]. Thus, molting observed in starved ptr-18 and mir-235 mutant larvae is likely caused by the ectopic activation of transcriptional oscillations of hhr and ptr genes. Furthermore, the activation of at least two hh-r-dependent processes, P cell reactivation and molting, always occur in this order. This apparent dependency of the latter events on the former can be explained by assuming 1) that each temporal transcriptional upregulation of hh-r genes promotes the subsequent hh-r genes one after another and 2) that each of the waves of upregulation sequentially activates distinct L1 developmental events. If so, the inhibition of grl-7 in starved ptr-18 and mir-235 mutant larvae would block the upregulation of a hh-r gene that promotes P cell reactivation in fed wild-type larvae by preventing the ectopic initiation of the transcriptional oscillations. This idea would also explain why ptr-18 deficiency conspicuously affected only the sequestration of the GRL-7 reporter but GRL-5 reporter. grl-5, potentially as well as grl-27, might act upstream of grl-7 to promote its oscillation of expression. Thus, the loss of grl-5 and potentially grl-27 would significantly suppress the quiescent defective phenotype of ptr-18 mutant animals via the downregulation of grl-7.
Although the molecular mechanism that generates the transcriptional oscillations of hh-r and ptr genes remain to be fully elucidated, it is worth noting that nhr-23, which encodes a nuclear hormone receptor homologous to both mammalian RORα and Drosophila DHR3, has been shown to upregulate multiple hh-r and ptr genes, including grl-5, grl-7, and ptr-18, and regulate molting [84][85][86]. In addition to nhr-23, dozens of genes whose inhibition causes molting defects have already been identified [87]. Instead of just regulating the molting cycle, some of these genes may promote the sequential activation of multiple developmental events by controlling the transcriptional oscillations of hh-r and ptr genes.

Materials and methods
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Masamitsu Fukuyama (mfukuyam@mol.f.u-tokyo.ac.jp).
Construction of grl-7 reporter genes using the CRISPR/Cas9 system venus::3xflag and mCherry tags were inserted into the grl-7 gene using the CRISPR/Cas9 system according to the protocol described in [93]. Briefly, venus::3xflag and mCherry tags, both of which contain grl-7 homologous arm sequences at both the 5' and 3' ends were constructed via PCR using primer pairs MF2984-MF2985 and MF3047-MF3048, respectively. DNA fragments encoding venus and mCherry cDNA were PCR-amplified using primer pairs MF2607-MF2608 and MF3049-MF3050, respectively. These PCR products were subsequently annealed to make "dsDNA hybrid donors", as described in [93]. Each of these dsDNA hybrid donors was co-injected with Alt-R S.p. Cas9 Nuclease V3 (IDT, 1081058), Alt-R CRISPR-Cas9 tracrRNA (IDT, 1072532), gene-specific Alt-R CRISPR-Cas9 crRNA, and pRF4 at certain concentrations as suggested by [93]. The partial sequence of the grl-7 gene CTATATTTACAATT GACGGT was used to custom-order the grl-7-specific Alt-R CRISPR-Cas9 crRNA. On the other hand, we could not insert venus and HA tags using CRISPR-Cas9 crRNAs targeting to the sequences, CTCCGAACGGATTTTTCAAC, AACCAGTTGAAAAATCCGTT, and CCGTTCACTTTGATAACTTT in the ptr-18 locus.

Immunostaining
A drop of the starved L1 larvae was placed on polylysine-coated slides, permeabilized using the freeze-cracking method, and fixed in MeOH/acetone at -20˚C for 5 min, as described previously [94]. The slides were incubated with anti-PGL-1 antibody [95] diluted 1:40,000 in PBS containing 4% BSA at 4˚C overnight. The slides were then washed in PBS three times and incubated with Alexa Fluor 568 goat anti-mouse IgG (H+L) highly cross-adsorbed secondary antibody (Thermo Fisher Scientific) diluted 1:2,000 in PBS containing 4% BSA at 37˚C for 30 min. The slides were washed in PBS three times and subsequently mounted with PermaFluor mounting medium (Thermo Fisher Scientific) containing 1 μg/mL of DAPI (4',6-diamidino-2-phenylindole).

Microscopy
The C. elegans embryos and adults were mounted on 4% agar pads, as described previously [96]. Fluorescent and differential interference contrast (DIC) images were obtained with an Axio Imager M1 equipped with Plan-Apochromat 63x/1.40 Oil DIC and HRm digital camera, and processed with the AxioVision (Carl Zeiss) and Photoshop (Adobe) software.

Super-resolution microscopy
L1 larvae were anesthetized as described above. Super-resolution imaging was performed using a Zeiss LSM980 quipped with Airysccan2 and the Plan-Apochromat 63x/Oil 1.4 DIC objective. Approximately 150 Z-section images/embryo were acquired using the SR-4Y Multiplex mode at 0.17 μm intervals and processed with ZEN pro software and Fiji.

C. elegans culture
The L1 larvae were prepared and either starved in complete S medium or grown synchronously as described previously [7]. The concentration of L1 larvae was adjusted by culturing sterilized embryos at 10 embryos/μl. The starved L1 larvae in polypropylene tubes were continuously rotated at 30-40 rpm. The L1 larvae referred to as 'after 5 days of L1 starvation' indicate larvae that were allowed to hatch and were cultured in complete S medium for 5 days after the alkali/bleach treatment, except for L1 larvae in Figs 6A-6D and S6A-S6D, which were starved in cholesterol-and ethanol-free complete S medium for 24 h. The synchronized embryos shown in Figs 2B and 3B and S2A and S3A were prepared by harvesting the early embryos by bleaching gravid adults as described above, hatched, and cultured in cholesterol-and ethanol-free complete S medium in a 15 mL polypropylene tube with continuous rotation at 30-40 rpm and 20˚C. Embryos were observed 9-17 h after bleach treatment. Embryos in other experiments were harvested in M9 buffer using a spatula and a glass Pasteur pipet (after washing out the well-fed gravid worms) on 100 mm 4x peptone plates, which contained a 4x peptone concentration of the standard nematode growth medium (NGM) agar. The collected embryos were then washed five times with M9 buffer to remove E. coli before observation.

Immunoblot analysis
Wild-type and ptr-18 mutant animals after 24 h L1 starvation were prepared as described in C. elegans culture. Newly hatched larvae were first filtered through an L1 harvest filter (InVivo Biosystems, USA). Filtered animals were collected via centrifugation in 15 mL polypropylene tubes at 3,500 rpm for 1 min at 4˚C, further centrifuged in 1.5 mL microtubes at 15,000 rpm for 1 min at 4˚C, and snap-frozen by liquid nitrogen. Proteins were extracted in the urea lysis buffer (6 M urea, 2 M thiourea, 3% [w/v] CHAPS, 1% [v/v] Triton X-100; [97]) by sonication using an ultrasonic disruptor UD-100 equipped with a TP-120 tip (TOMY, Japan). UD-100 was set to repeat the cycle of 10 s pulses at 99% power and 10 s intervals for 2 min. When unbroken larvae were found under the dissecting microscope, another around of sonication was performed. Protein extracts were resolved in 10% TGX by SDS-PAGE and blotted onto the Immun-blot PVDF membrane by the Transblot Turbo blotting system (Bio-Rad, USA). Blotted membranes were first blocked in the Everyblot blocking buffer for 5 min (Bio-Rad, USA), incubated with primary antibodies at 4˚C overnight, and subsequently incubated with secondary antibodies at room temperature (20-25˚C) for 1 h. Anti-GFP from mouse IgG1κ (clones 7.1 and 13.1) (Roche 11814460001), anti-α-tubulin antibody, mouse monoclonal clone DM1A (Sigma T6199), and Peroxidase-AffiniPure Goat Anti-Mouse IgG (Jackson ImmunoResearch 115-035-071) were used at 1:1,000, 5, 000, and 10,000 dilutions, respectively, in a 19:1 Tris Buffered Saline:Blocking One (Nacalai, Japan) solution containing 1% Tween 20. Signals were detected using Chemi-lumi One (Nacalai, Japan) with ChemiDoc XRS+ (Bio-Rad, USA). The gray intensity of each band circled by a fixed size of ROI was measured using Fiji (https:// imagej.net/Fiji).

Viability assay
The embryos harvested using the alkali/bleach method described above and the resulting hatched larvae were cultured at 10 embryos/μl in 10 mL of cholesterol-, ethanol-free complete S medium in a 15 mL polypropylene tube with rotation at 30-40 rpm at 20˚C for 10 days. The larvae were subsequently transferred onto freshly seeded plates, and the number of transferred and recovered larvae was scored after 3 days.

Statistical analysis
All statistical analyses were conducted using the Microsoft Excel software. . Photographs were taken with the same exposure time. B) Apical localization of GRL-5:: mCherry::3xFLAG was rarely detected in the starved wild-type and ptr-18 mutant L1 larvae. n �50 animals were scored for each genotype. Data were collected after 24 h L1 starvation and are presented as mean ± SD. Experiments were repeated three times, and n �50 animals were scored for each trial. (C) Wild-type and ptr-18 mutant 3-fold embryos and L1 larvae after 24 h L1 starvation are shown. These animals carry transgenes that express GRL-5::VENUS fusion protein under the control of the native grl-5 promoter. Note that the fluorescence of L1 animals under the GFP filter is derived from gut granules. The presence of the transgene is marked by the co-injected plasmid that expresses mCherry in the hypodermis. Photographs were taken with the same exposure time. Scale Bars: 10 μm. (D) Percentage of starved wildtype and ptr-18 mutant L1 larvae showing GRL-5::VENUS expression. Data were collected after 24 h L1 starvation and are presented as mean ± SD. Experiments were repeated three times, and n �50 animals were scored for each trial. (PDF) S7 Fig. mir-235 and ptr-18