Transcriptional and spatiotemporal regulation of the dauer program

ABSTRACT

Caenorhabditis elegans can enter a diapause stage called “dauer” when it senses that the environment is not suitable for development. This implies a detour from the typical developmental trajectory and requires a tight control of the developmental clock and a massive tissue remodeling. In the last decades, core components of the signaling pathways that govern the dauer development decision have been identified, but the tissues where they function for the acquisition of dauer-specific traits are still under intense study. Growing evidence demonstrates that these pathways engage in complex cross-talk and feedback loops. In this review, we summarize the current knowledge regarding the transcriptional regulation of the dauer program and the relevant tissues for its achievement. A better understanding of this process will provide insight on how developmental plasticity is achieved and how development decisions are under a robust regulation to ensure an all-or-nothing response. Furthermore, this developmental decision can also serve as a simplified model for relevant developmental disorders.

Abbreviations: AID Auxin Induced Degron DA dafachronic acid Daf-c dauer formation constitutive Daf-d dauer formation defective DTC Distal Tip Cells ECM modified extracellular matrix GPCRs G protein-coupled receptors IIS insulin/IGF-1 signaling ILPs insulin-like peptides LBD Ligand Binding Domain PDL4 Post Dauer L4 TGF-β transforming growth factor beta WT wild-type

1. The dauer program: Entry, Execution, Maintenance, and Exit

The nematode Caenorhabditis elegans develops through four larval stages (L1-L4) before reaching adulthood. The dauer larva is the outcome of a plastic development in this nematode’s life history, where in response to harsh environments, they alter the L2-L3-L4 continuous development trajectory to the alternative L2d-dauer-L4 path [1,2] Figure 1. Dauers acquire specialized morphological, physiological, and behavioral traits that enhance dispersal and survival until a more suitable environment is found. Namely, they stop feeding and alter their metabolism to gradually make use of the stored energy, change collagen composition of their cuticle to make it more resistant to desiccation and other environmental insults, arrest gonadal and somatic development and perform a hitch-hicking behavior that increases probability of survival [1,3]. The perception of dauer-inducing conditions at the L1 (for L2 vs L2d decision) and L2d (for L3 vs dauer) relies on sensory neurons, mainly through G protein-coupled receptors (GPCRs). Ablation of ASI, ADF, or ASG sensory neurons triggers dauer entry (Daf-c, constitutive dauer formation), while ablation of ASJ or ASK neurons causes Daf-d phenotypes (dauer defective formation) [4,5]. Mutations in GPCRs phenocopy neuronal ablation, with some GPCRs mediating dauer promoting cues, while others signaling reproductive development [6,7]. Importantly, internal metabolic state signals from the gut are also integrated by the nervous system [8]. The neuronal circuits involved are beyond the scope of this review, but they converge on insulin/IGF-1 signaling (IIS), transforming growth factor β (TGF-β) and steroid hormone pathways to modulate entry to the dauer program (reviewed in [9]) Figure 2. Mutations in components of these pathways also have Daf-c and Daf-d phenotypes, and temperature sensitive (ts) Daf-c alleles are often used to induce dauer formation at restrictive temperatures [10].

Figure 1.

Continuous development trajectory (blue arrows) and dauer trajectory (red arrows). the asterisk (*) shows the reproductive development commitment, which can be reached by 3 different paths.

Figure 2.

Overview of the three dauer pathways. Insulin/IGF-1 signaling (DAF-2), TGF-β (DAF-7) and steroid hormone (DAF-12). In bold: transcription factors and co-factors. ILPs: insulin-like peptides.

A further examination of dauer-related phenotypes reveals different “steps” in the dauer program, subject to distinct transcriptional regulation. The Daf-c and Daf-d mutant phenotypes are associated with genes involved in sensing, signal transduction, and transcriptional regulation that mediate the developmental arrest [9]. Therefore, Daf-d mutants are locked in a continuous development trajectory while Daf-c are obligated to arrest (“entry”) and follow a dauer path [10]. The entry is uncoupled from the execution in the partial dauer mutants, with worms that do not proceed to adulthood but fail to showcase the typical dauer features [11,12]. Thus, the execution to acquire dauer traits is secondary, and needs to be actively maintained for the duration of diapause (up to 4+ months) [13,14]. If the environment is sensed as favorable, then dauers will irreversibly commit to exit diapause, losing the specialized features and resuming development by molting into a PDL4 (Post Dauer L4) [1] Figure 1. Post dauer adult morphology resembles that of worms that underwent continuous development, but show significant differences in epigenetic profiles, transcriptome, adult lifespan, and brood size [15–19].

DAF-16/FOXO, DAF-3/co-SMAD, and DAF-12/VDR are well-characterized transcription factors downstream the three dauer pathways (reviewed in [20–23]). They are required for dauer entry and have roles in tissue remodeling, maintenance of the remodeled state and exit/recovery. Among their targets are dauer-specific genes like collagens and GPCRs that need to be induced and reproductive development ones that must be inhibited. For the latter, the dauer program is intertwined with the heterochronic circuit, a transcriptional network of genes that function as temporal regulators of developmental progression (reviewed in [24,25]). Besides said temporal control, each step of the dauer program requires spatial (tissue-specific remodeling) and synchronic (inter-organ coordination) modulation, conditioned to external cues and internal metabolic state [2,8,26]. This review aims to gather the current knowledge on the transcriptional regulation of the dauer program, focusing on how the dauer promoting transcription factors/regulators contribute to the different steps and which are the relevant tissues involved.

2. Entry: to arrest or not to arrest?

Entry can be referred to as the commitment to arrest development upon encountering harsh environmental conditions at the L2d larvae. It relies on two critical periods of sensing and decision-making where food availability, population density, and temperature are assessed and integrated by IIS, TGF-β, and the steroid hormone pathways to determine the developmental path [2]. L1 larvae can choose to commit to reproductive development through molting into an L2 larvae or instead, decide to form the dauer-capable L2d larvae. High food supply, low dauer pheromone (secreted ascarosides, a proxy for population density) or low temperature activate IIS and TGF-β, which feed into the steroid hormone pathway during the L1/L2 molting to signal dauer bypass [2,27,28]. Reduced signaling in response to low food supply, high pheromone and/or high temperature will drive entry to L2d. During the L1 transition to L2d, genes involved in stress response are upregulated, while genes for growth and metabolism, such as the intestinal lipase gene lips-6, are downregulated [29]. While still feeding, L2d larvae start executing dauer fates, including fat accumulation and shrinkage of hypodermis through autophagy [2,30,31]. In addition, they delay reproductive fates, evidenced by the extended intermolt period and slower rate of germline divisions [2,32]. Meanwhile, they continue to assess the environment until mid-L2d, when they commit to dauer or proceed to L3 [2,28]. To facilitate differentiation of specific stages, several genetic markers have been recently developed. L2d stage can be recognized by expression of flp-8 in AVM neurons or ets-10 in the intestine [33,34]. Dauer commitment can be assessed by hypodermal expression of col-183 and intestinal expression of ets-10 and nhr-246. Meanwhile, the expression of F53F1.4 in the hypodermis indicates reproductive fate commitment [34].

Despite the dauer signaling pathways being extensively studied and their role in committing to arrest being well understood, efforts are still being made to understand the relevant tissues for the dauer entry decision [14,35,36]. The transcription factors that drive dauer entry are evident upon their Daf-d mutant phenotypes, while the ones that inhibit it are Daf-c. Table 1 summarizes the transcriptional regulators involved in entry, their expression pattern, and the tissues where their expression has been identified as “sufficient” or “necessary” to drive the transcriptional response.

Table 1.

Transcriptional regulators with mutant dauer Entry phenotypes.

TF	ortholog	mutant phenotype	assigned pathway	larval expression pattern	dauer expression pattern	postembrionic highest expression	sufficient	necessary
DAF-8	R-SMAD	daf-c	DAF-7/TGF-β	subset of head neurons (ASI and ADL) and tail neurons, intestine, DTCs, VNC [37]	n.a	L1 larvae [37,38]	n.a	n.a
DAF-14	R-SMAD	daf-c	DAF-7/TGF-β	neurons in the lateral ganglia, phasmid glia sheet, hypodermis, intestine, excretory cell [39]	n.a	L1 larvae [38]	n.a	n.a
DAF-3	co-SMAD	daf-d (in TGF-β mutant background)	DAF-7/TGF-β	citoplasmatic localization in head and tail neurons, pharinx, hypodermis, intestine, DTCs and VNC [14,40]	nuclear expression in some head neurons [14]	L1 larvae [14,38]	nervous system [40]	cilliated neurons [14]
DAF-5	Sno/Ski	daf-d (in TGF-β mutant background)	DAF-7/TGF-β	nuclear localization in head and tail neurons, pharinx, hypodermis, muscles, intestine and DTCs [41]	n.a	L1 larvae [38]	nervous system [41]	n.a
DAF-16	FOXO	daf-d (in IIS mutant background)	DAF-2/IIS	citoplasmatic localization in sensory head neurons, citoplasmatic [14]	nuclear localization in neurons, hypodermis, intestine, body wall [14,42]	dauer [38]	neurvous system [43], intestine [44]	assayed, not determined [14]
DAF-12	VDR	daf-d (DBD mutant or null allele) or daf-c (LBD mutant)	Steroid hormone	nuclear localization in hypodermis, vulva, somatic gonad, intestine, nervous system, pharynx, pharingeal muscle body wall muscle [20]	nuclear localization in neurons, hypodermis, intestine, body wall [14]	dauer [38] followed by L2: hypodermis, vulva, pharingeal muscle, nervous system [20,38]	n.a	assayed, not determined [14]
DIN-1s	SPEN	daf-d	Steroid hormone	widely expressed, XXX cells [45]	n.a	late L1 and L2 larvae, the XXX cells [38,45]	n.a	n.a
NHR-69	HNF4a	enhanced dauer entry (in TGF-β background)	DAF-7/TGF-β	intestine [46]	n.a	dauer [38]	n.a	n.a
CHD-7	CHD7/8	daf-d (in TGF-β background)	DAF-7/TGF-β	widely expressed [47]	n.a	L2 [38]	n.a	n.a

bold: tissues with highest fluorescent reporter signal.

“sufficient”: tissue-specific expression that rescues mutant LoF phenotype.

“necessary”: tissue-specific deletion that phenocopies mutant LoF phenotype.

n.a: not assayed.

VNC: ventral nerve cord.

DTCs: distal tip cells.

2.1. Entry promoting components in the DAF-7/TGF-β pathway

DAF-3/co-SMAD and DAF-5/Sno/Ski form a transcriptional complex that promotes dauer entry [40,41]. During favorable conditions, upon DAF-7/TGF-β binding to receptors DAF-1/TGFBR1 and DAF-4/BMPR2, active R-SMADs DAF-8/DAF-14 antagonize DAF-3/DAF-5, probably by directly binding to DAF-3/co-SMAD and impeding the formation of the complex [37] Figure 2. Since DAF-7/TGF-β is secreted from ASI neurons and all the components of the TGF-β pathway are widely expressed, early models considered DAF-7/TGF-β as the neuroendocrine signal that traveled toward downstream tissues to promote reproductive development [5,48]. However, this does not appear to be the case for the dauer entry decision: pan-neuronal expression of daf-4 or daf-1 rescues the Daf-c mutant phenotype of daf-4(e1364) or daf-1(m40), respectively [39,49]. In addition, pan-neuronal expression of daf-5 rescues the Daf-d phenotype of daf-5(e1386) mutants [41]. In support of these studies, recent work taking advantage of the auxin-inducible degron (AID) system, demonstrated that daf-7(e1372) mutants depleted of DAF-3/co-SMAD specifically in ciliated sensory neurons bypass dauer arrest at 25°C, arguing that neuronal DAF-3/co-SMAD is necessary for dauer entry [14]. Therefore, the DAF-7/TGF-β pathway acts mainly in the nervous system and functions non-autonomously to promote arrest in the whole organism. How this signal propagates to downstream tissues remains unknown but could be mediated by insulin peptides or other neuropeptides (see Entry: TGF-β and IIS show complex cross-talk). For instance, a secreted neuropeptide that functions non-autonomously to promote reproductive development was described recently [50]. In response to a favorable environment, AIA interneurons secrete the FMRF-like peptide FLP-2 that activates the GPCR NPR-30 in the nervous system and intestine. flp-2(-) and npr-30(-) mutants have an increased dauer formation index and expression of flp-2 in AIA strongly inhibits dauer formation in both WT and npr-30(-) mutant backgrounds [50]. Interestingly, the ASI neurons and AIA interneurons share gap junctions, and could communicate through direct electrical synapses or small molecules that initiate signal transduction and gene expression [51,52]. Whether an ASI mediated signal, such as DAF-7/TGF-β, instructs AIA to secrete FLP-2 still needs to be addressed.

Recently, we found CHD-7/CHD7/8 (a chromodomain helicase) to be necessary for dauer entry in daf-7 mutants, as chd-7(tm6139);daf-7(e1372) mutants bypass arrest at 25°C (Daf-d) [47]. Developmental arrest can be rescued by shifting the worms to a higher temperature (26.5°C), which phenocopies daf-3 and daf-5 mutants [41]. Genetic epistasis puts chd-7 at the level of daf-3 and daf-5, upstream of daf-9 [47]. Interestingly, chd-7 is allelic to scd-3 (suppressor of constitutive dauer-3), which was found more than 20 years ago in a screen for TGF-β suppressors of dauer formation [53]. In double mutants chd-7(tm6139);daf-2(e1370), CHD-7/CHD7/8 was also found to have roles in dauer morphogenesis (see Execution). Like DAF-3/co-SMAD, CHD-7/CHD7/8 is involved in dauer entry decision in the context of reduced TGF-β signaling. Interestingly, unlike DAF-3/co-SMAD, CHD-7/CHD7/8 is also involved in the dauer execution in the context of reduced IIS signaling. At its C-terminus, CHD-7/CHD7/8 has a small motif called BRK (Brahma and Kismet) domain that is critical for its role in both dauer entry and morphogenesis [47]. Since the BRK domain is predicted to function as a protein–protein interaction module [54], CHD-7/CHD7/8 might be interacting with other transcription factors and regulators to modulate the distinct steps of the dauer program.

2.2. Entry promoting components in the DAF-2/IIS pathway

DAF-16/FOXO is antagonized by the IIS pathway Figure 2, upon phosphorylation by AKT-1 and −2 and consequent sequestration in the cytoplasm. C. elegans has only one insulin-like receptor (DAF-2/INSR) and nearly 40 insulin-like genes [55,56]. From these, agonistic INS-4, INS-6 and DAF-28 promote reproductive development while antagonistic INS-1 or INS-18 promote dauer entry [17,44]. In unfavorable conditions and resulting reduced IIS, DAF-16/FOXO translocates to the nucleus to promote dauer entry. In dauers, nuclear DAF-16/FOXO is observed throughout the body [14,42]. An early study showed that DAF-16/FOXO expression in the nervous system was sufficient to rescue the Daf-d phenotype of daf-16(mu86);daf-2(e1370) mutants [43]. Nonetheless, the site of action of DAF-16/FOXO to enter dauer is still under debate, with studies showing a key role for the intestine, while others arguing against it [14,44]. Expression of daf-2 under an intestinal promoter completely rescues the Daf-c phenotype of daf-2(lf) mutants [44]. Consistent with this, intestinal expression of daf-16 completely rescues the Daf-d phenotype of daf-16(mu86);daf-2(lf), while expression with nervous system or muscle-specific promoters had no effect [44]. Together, these experiments suggest that the activity of DAF-2/INSR and DAF-16/FOXO in the intestine is critical for the dauer entry decision. In addition, the Daf-d phenotype of triple mutants for INS-4, INS-6 and DAF-28 genes (redundant agonistic insulins) can be fully rescued by intestinal DAF-16 [44]. This argues that dauer-inhibitory insulin neuropeptides secreted from the nervous system function upstream of DAF-16/FOXO activity in the intestine. Indeed, a model involving intestinal DAF-16/FOXO has been proposed for organismal tissue-coordinated aging, where DAF-16/FOXO inhibits INS-7 insulin secretion from the gut, which in turn reduces IIS and activates DAF-16 transcriptional activity in other tissues [57]. However, by itself, neuronal, intestinal, hypodermal or muscle depletion of DAF-16/FOXO using the AID system failed to prevent dauer entry [14]. Accordingly, depletion of DAF-2/INSR in each of these tissues did not phenocopy ubiquitous depletion [36]. Neither did simultaneously depleting DAF-2/INSR from neurons and intestine or neurons and muscle [35]. This, together with the observation that intestinal DAF-16/FOXO expression alone is sufficient but not necessary for dauer entry, suggests that rather than a single hub for the DAF-16/FOXO activity, this transcription factor is simultaneously required in multiple tissues to coordinately promote dauer entry.

2.3. Steroid hormone pathway: the most downstream regulator for dauer bypass/entry

DAF-12/VDR is the nuclear hormone receptor that integrates both IIS and TGF-β signaling for the dauer entry decision [20,58] Figure 2. It binds to bile acids called Dafachronic Acids (DA), which are cholesterol-derived hormones [59,60]. The enzymes known to be involved in the cholesterol backbone modification are DAF-36, HSD-1 and DHS-16, while DAF-9/CYP450 introduces a carboxyl group, representing the last step of DA synthesis [59,61–65]. Double mutants of these genes have enhanced dauer formation, some having complete Daf-c penetrance and resulting in 100% dauers, like hsd-1(mg433);daf-36(k114) [63,64]. DAF-36 is primarily expressed in the intestine while HSD-1 is restricted to the XXX neuroendocrine cells [59,64]. DAF-9/CYP450 is present in the XXX cells and, together with DHS-16, is also expressed in the hypodermis, suggesting that many tissues are involved in DA synthesis [62,63]. In its liganded state, DAF-12 activates reproductive development and when unliganded, forms a complex with the co-repressor DIN-1S/SPEN for dauer arrest [45,58,60].

The decision to commit to reproductive development earlier (through L2) or later (through L2d) seems to rely ultimately on DA levels in late L1 and mid-L2d, respectively [2,28]. By supplementing daf-9(dh6) animals, which lack endogenous DA, with increasing concentrations of this hormone, it was shown that there is a threshold to be met to bypass dauer entry [28]. This threshold can in fact be shifted by other factors, such as the amount of dauer pheromone. Furthermore, they found daf-9(dh6) become refractory to exogenous DA by mid-L2d, since continued supplementation past that point resulted in 100% dauers [28]. Taking this into account, L2s are presumably worms that met the DA threshold earlier (at late L1) while L2ds, by having a slower rate of production of DA in response to unfavorable conditions, take longer to achieve the threshold. If by mid-L2d, a worm did not meet the DA threshold, it will become dauer Figure 1.

When worms face proper conditions for development in L1, enough DA is produced and DAF-12/DA activates the let-7 family of miRNA, which in turn represses hbl-1 for a rapid L2/L3 transition. A negative feedback loop, where these miRNAs repress DAF-12/VDR expression, inhibits dauer formation, and reinforces the reproductive development commitment [66,67]. In contrast, adverse conditions during L1 and consequent DA levels below the threshold, drive the molt into L2d larva [28]. In this dauer-prone stage, unliganded DAF-12/VDR binds to DIN-1/SPEN and represses the expression of the let-7 family of miRNAs [67]. The L2d delay of L3 fates is achieved by an alternative program relying on lin-46, lin-4, nhl-2 miRNAs to timely repress hbl-1 [68]. Thus, this genetic rewiring in the L2d tunes the developmental clock for a potential diapause entry. While this regulation of hbl-1 during L2 or L2d was studied in hypodermal seam cells, these or other miRNAs could potentially act to synchronize arrest in every tissue. In the past, dauer life history was shown to suppress heterochronic mutant phenotypes (where some tissues fail to follow the stage-specific program), suggesting that genes involved in temporal control became somehow dispensable after dauer life history [69–72]. Recent experiments showed that L2d life history alone is sufficient to suppress heterochronic phenotypes, like the reiteration of seam cell proliferative divisions, which can be attributed to the employment of an alternative genetic circuit for the developmental clock [68].

In which tissues is DAF-12/VDR required for dauer bypass (together with DA) or entry (together with DIN-1/SPEN) has yet not been elucidated. While ubiquitous DAF-12/VDR depletion in daf-7(e1372) mutants prevents dauer entry, tissue-specific depletion was unable to phenocopy this phenotype [14]. Considering that the site of action of DAF-3/co-SMAD for dauer entry is the nervous system, TGF-β in the nervous system functions non-autonomously by feeding into a signal that informs of the dauer entry vs. dauer bypass decision in all relevant tissues. Meanwhile, the apparent broad tissue requirements for DAF-16/FOXO and DAF-12/VDR, argues for a model where IIS and steroid promoting transcription factors/regulators contribute to the different steps and which are the relevant tissues involved. Hormone pathways require inter-organ communication to coordinate dauer entry.

2.4. TGF-β and IIS show complex cross-talk

Mutations in daf-3 are dauer defective in Daf-c mutant backgrounds from the TGF-β pathway and similarly, daf-16 mutations are dauer defective in Daf-c backgrounds from the IIS pathway. However, these TFs are dispensable in a mutant background of the other pathway, namely daf-2(e1370);daf-3(e1376) and daf-16(m26/m27);daf-7(e1372) arrest at non-permissive temperatures [12,73]. These epistasis experiments suggested that both pathways worked in parallel for dauer entry. Arguing against this was the observation that daf-16(mgDf47);daf-7(m62) mutants, when grown at 25°C, showed increased DAF-16:GFP nuclear localization in L2d, when dauer commitment decision is made. Thus, reduced TGF-β signaling activates DAF-16, suggesting a cross-talk between these pathways [42]. However, the mechanism or the level at which these pathways intersect remained unknown. More recently, using RNAi instead of mutants, it was demonstrated that daf-16 RNAi partially suppress dauer entry of daf-7(e1372) worms at comparable levels of daf-3 RNAi [74]. Unexpectedly, daf-8(m85) mutants treated with daf-3 RNAi formed 75% dauers while daf-16 RNAi only formed 4% [74]. This strong suppression by daf-16 RNAi suggests that, when TGF-β is reduced, DAF-16 also functions as a downstream effector in the dauer entry decision, again arguing for communication between these pathways. In addition, active TGF-β changes insulin peptides expression, regulating the IIS pathway and DAF-16 activity [46,74]. One report found a complex between DAF-8/R-SMAD and NHR-69/HNF4a that positively regulates IIS. DAF-8/NHR-69 complex inhibits exp-2 and alleviates EXP-2-mediated repression of DAF-28 in ASI neurons. Active DAF-28 secretion then inhibits DAF-16/FOXO nuclear localization throughout the animal [46]. Another report showed that the phosphatase PDP-1/PDP1/2 antagonizes IIS and promotes DAF-16/FOXO nuclear localization and transcriptional activity. While PDP-1/PDP1/2 genetically interacted with TGF-β at the level of the R-smads DAF-8 and DAF-14, the expression profile of pdp-1 mutants showed misregulation of many insulin peptides and indicates modulation of the IIS pathway, although the mechanism behind this regulation is not fully understood [74]. Both examples show how TGF-β components modify insulin expression in the nervous system and modulate IIS signaling in downstream tissues. Nonetheless, an even greater complexity is revealed when considering that NHR-69/HNF4a, which antagonizes DAF-16/FOXO activity by insulin modulation, can be repressed by DAF-16/FOXO [46]. Furthermore, the insulin agonist ins-7, which is negatively regulated by PDP-1/PDP1/2, can also be inhibited by DAF-16/FOXO [57,74].

The other TGF-β pathway in C. elegans, the Sma/Mab pathway, is involved in body size determination, cuticle formation, and male tail development [22]. Very recently, this pathway has also been implicated in dauer formation [75,76]. Single mutants of this pathway do not exhibit dauer phenotypes but do when put in sensitized genetic backgrounds. This is the case of dbl-1 (encodes for the ligand of the pathway), sma-3 (the R-SMAD activated upon ligand binding) and lon-2 genes (antagonizes ligand binding). The double mutant daf-7(ok3125);dbl-1(ok3749) showed an increased dauer arrest at 20°C when compared to daf-7 single mutants [73], suggesting cross-talk between both TGF-β pathways. What is more, Sma/Mab promoted dauer entry in the context of daf-2(e1370) grown at the permissive temperature of 20°C. daf-2(e1370);sma-3(wk30), which have reduced Sma/Mab signaling, had a smaller dauer fraction when compared to daf-2 single mutants. Meanwhile, daf-2(e1370);lon-2(e678), that have an increased Sma/Mab signaling, showed a greater dauer fraction. Further studies are needed to understand how Sma/Mab can modulate both DAF-7/TGF-β and IIS for dauer entry but interestingly, Sma/Mab targets ins-4, which is known to have inhibitory roles in dauer entry [44,77].

3. Execution

Execution refers to the active remodeling of tissues and changes in physiology and behavior to achieve all the specialized traits of a dauer larva. For this, animals must a. indefinitely delay reproductive fates, b. ensure long-term survival, achieved by secreting a specialized cuticle, shifting the metabolism, and upregulating stress genes and c. promote dispersal, such as the hitch-hiking nictation behavior (for a detailed review on the morphological changes, see [3,78]). The execution is uncoupled from the entry decision in partial dauers, which arrest but fail to acquire all dauer-associated features (i.e., SDS resistance). Since the discovery that dauers survive exposure to this detergent, treatment with 1% SDS has been used in dauer formation assays to differentiate between non-dauers and dauers [1,79]. This can be misleading, since SDS-sensitive worms can be partial dauers that failed to properly form the buccal plug or the proper dauer cuticle [1]. Discriminating between normal and partial dauers is critical to understand the roles of the dauer pathway modulators and effectors, as they could be targeting entry and execution steps differently. For instance, DAF-3/co-SMAD and DAF-16/FOXO show parallel functions in dauer entry, but distinct and unique roles in dauer execution (see next section). Lately, new approaches have been developed to differentiate partial from full dauers, such as edible fluorescent beads combined with shorter 1% SDS incubation periods to find detergent-sensitive non-feeding partial dauers [80]. Other approach makes use of fluorescent-based reporters of dauer commitment genes to track tissue-specific remodeling, since they are often mis-expressed in partial dauers [34].

3.1. DAF-16/FOXO and DAF-3/co-SMAD non-parallel functions

daf-16 and daf-3 mutants are Daf-d in the context of reduced IIS and TGF-β, respectively [12]. To study their function in the execution of the dauer program, specific genetic backgrounds or experimental conditions are needed to force them to arrest. For instance, daf-16(lf) mutants arrest when exposed to dauer pheromone or when grown at 25°C as daf-16(lf);daf-7(e1372) double mutants. Despite being arrested, they form partial dauers that exhibit pharyngeal pumping and SDS sensitivity, and fail to shrink the pharynx, form alae or accumulate gut granules [12,73]. Thus, in daf-16(lf);daf-7(e1372) partial dauers, active DAF-3/DAF-5 repressive complex is able to successfully trigger dauer entry, but fails to accomplish dauer remodeling in these tissues, for which DAF-16/FOXO is required. In contrast, daf-3(mgDf90) mutants exposed to pheromone or as daf-2(e1370);daf-3(mgDf90) double mutants, arrest as full dauers [12,73], suggesting that in the context of reduced IIS signaling, active DAF-16/FOXO promotes dauer entry and tissue morphogenesis independently of DAF-3/co-SMAD. Therefore, DAF-3/co-SMAD is critical for dauer entry in the context of reduced TGF-β but is dispensable for morphogenesis. DAF-16/FOXO, on the other hand, seems to be required for both entry and morphogenesis, having unique and overlapping targets with the steroid hormonal pathway.

3.2. DAF-16/FOXO and DAF-12/VDR shared and unique functions

For dauer entry, daf-16 has been placed upstream of daf-9 and daf-12 by genetic epistasis [61,62,81] Figure 2. For execution, the attribution of a hierarchical role is more complex. Besides targeting a distinct subset of effector genes and tissue remodeling [14,23,82], daf-16 has also been placed downstream of daf-12 [83]. Replacement of cholesterol with the methylated metabolite lophenol in one generation, affects the next generation’s ability to produce DA, driving dauer entry. In these lophenol-induced dauers, DAF-16/FOXO translocates to neuronal nuclei of the pharynx, tail and ventral cord and this translocation is dependent on unliganded DAF-12/VDR activity [83], presumably bound to repressor DIN-1/SPEN. Furthermore, DAF-12/VDR can directly promote daf-16 expression as shown by ChIP-Chip analysis and luciferase-based assays [82].

Recently, overlapping functions have been identified for DAF-16/FOXO and DAF-12/VDR in the dauer program. Pan-neuronal depletion of DAF-16/FOXO or DAF-12/VDR has similar effects: lack of dauer-specific induction of inx-6 in AIB neurons, increased locomotory behavior and pharyngeal pumping [14]. For these functions, daf-16 could be upstream or downstream daf-12, or both converging on the same targets. In addition, they have unique roles; DAF-16/FOXO in the gut and pharyngeal muscles is necessary to inhibit pumping cessation, while DAF-12/VDR is not [14]. The DAF-12/DIN-1 complex is required autonomously in the germline for quiescence, for which DAF-16/FOXO plays no part [32,84]. Moreover, these experiments demonstrate cell autonomous and non-autonomous roles for these transcription factors to coordinate remodeling throughout the body.

3.3. Transcriptional regulation of execution

The regulation of dauer execution seems more complex than that of dauer entry. The latter relies ultimately on not meeting a DA threshold for the whole organismal arrest, while the former depends on each tissue individually remodeling in a coordinated manner. The autonomy of each tissue is especially evidenced in partial dauers failing to express genetic markers in a subset of tissues only. Such is the case of daf-9 partial dauers that show reduced expression of the dauer marker ets-10 in the intestine, but appropriately express it in neurons [34]. Since each tissue must halt reproductive fates and promote dauer ones, the pathways involved often display non-autonomous functions, which ensures coordination through inter-organ communication (see below). For some tissue-specific remodeling events, the pathways and transcription factors modulating dauer execution are yet to be determined. For instance, how the epidermal cuticlins CUT-1, −5 and −6, required for proper alae formation, become induced [85] or how the SCL-12 and −13, involved in cholesterol sequestration in intestinal lumen, are activated is not well understood [86]. For other remodeling events, transcriptional regulators have been identified and we will refer to them in the next sections.

3.3.1. Inhibition of reproductive fates

The systemic inhibition of reproductive fates (i.e., developmental arrest) is regulated by DAF-12/VDR and the heterochronic circuit, which ensures quiescence in hypodermal seam cells (described in Entry). The remaining somatic tissues and the germline require distinct transcriptional regulation to ensure quiescence and are all summarized in Table 2.

Table 2.

Transcriptional regulators that function to inhibit reproductive fates during Execution of dauer program.

TF	required in	function	target tissues	target genes	outcome	ref
DAF-3/DAF-5	somatic gonad (DTCs)	cell autonomously	somatic gonad	lag-2	block Notch signaling, undifferenciated germline	[87,88]
DAF-16	n.a	n.a	hypodermis	lin-41	repress adult collagen expression	[89]
	vulval precursor cells	cell autonomously	vulva	lag-2	block Notch signaling, undifferenciated vulva	[90]
DAF-12/DIN-1	n.a	n.a	hypodermis	lin-46, nhl-2, lin-4, let-7 Fam	hbl-1 expression, delay of L3 fates	[68,91]
	germline, somatic gonad, vulva	cell autonomously	germline, somatic gonad	-	quiescence in germline, somatic gonad	[84,90]
unknown*	-	-	neurons and excretory system	-	PAR-4 and AAK-1 -2 activities promote quiescence in germline, somatic gonad and vulva	[92]
unknown*	-	-	germline, somatic gonad	-	DAF-18 promotes quiescence in germline and somatic gonad	[32,93]

*placed downstream of TGF-β and IIS by genetic epistasis.

n.a not addressed.

• Quiescence of the germline. To ensure post-dauer fecundity, the germline stops proliferating and remains undifferentiated. The study of mutants that display germline hyperplasia allowed the identification of three parallel pathways to maintain germline quiescence: 1) PAR-4/AAK-1 and AAK-2 (LKB1/AMPK): downstream of TGF-β and IIS, these genes function non-autonomously from neurons and the excretory system [32,92]. 2) DAF-18/PTEN: downstream of IIS but independent of DAF-16/FOXO, DAF-18/PTEN functions non-autonomously from the somatic gonad [32,93]. 3) DAF-12/DIN-1 complex: downstream of TGF-β and IIS pathways, this complex acts cell-autonomously in the germline and is linked to the heterochronic circuit [84]. For pathways 1 and 2, the transcriptional regulators that mediate their activity have not been fully elucidated. In addition to promoting quiescence in the germline, dauers must prevent meiotic differentiation and this is achieved by inhibiting the Notch pathway. In the gonadal Distal Tip Cells (DTCs), DAF-3/co-SMAD and DAF-5/SPEN are required to directly repress the expression of the Notch ligand LAG-2 by binding to a 25 bp sequence upstream its promoter [87,88]. Consistent with additive effects involving DAF-12/DIN-1 in germline quiescence and a TGF-β role in germline differentiation, din-1(rr94);daf-7(e1372) show stronger germline hyperplasia compared to din-1(rr94);daf-2(e1372) dauers [84].

• Quiescence of the somatic tissues. Dauers also prevent cellular proliferation in most somatic tissues including the seam cells, gonad and vulva. Both aak-1 and din-1 mutants show vulval and gonad over-proliferation during dauer morphogenesis, suggesting that pathways involved in ensuring germline quiescence are also relevant in somatic tissues. However, they are not required for inhibition of seam cells divisions [84,92], which is achieved by regulation of the HBL-1 transcription factor by miRNAs [68,91]. The LIN-12/Notch pathway is also inhibited in the vulva, where DAF-16/FOXO acts cell autonomously in Vulval Precursor Cells (VPC) to repress lag-2 expression, preventing the expression of vulval specification markers and maintaining multipotency [90]. For this vulval downregulation of Notch, neither DAF-5/Sno/Ski nor DAF-12/VDR are required [90].

4. Promotion of dauer fates: (summarized in Figure 3.)

Figure 3.

DAF-12 and DAF-16 functions in the Execution step of the dauer program. Dashed lines involve inter-organ communication while continuous lines represent tissue autonomous signaling. Created with BioRender.Com.

• Modification of chemoreceptor repertoire: Sensory receptors, generally GPCRs, use cGMP as a second messenger to regulate TGF-β, IIS and steroid hormone pathways (reviewed in [9]). While in dauer, worms change their neuronal receptor repertoire, a feature associated with enhanced sensibility to detect environmental changes [19,94]. For instance, the GPCR sri-9 becomes selectively expressed in 8 neuron classes upon dauer entry and this is largely dependent on DAF-16 activity, since pan-neuronal removal of DAF-16 by the AID system suppresses its induction [14,19]. DAF-12 pan-neuronal removal also affects sri-9 expression in some neuron classes, overlapping with DAF-16 function in NSM and OLL [14]. The precise cell-specific activities of DAF-16 and DAF-12 have been associated with terminal selectors, which are combinatorial transcription factors that are involved in neuronal identity (reviewed in [95]). Losing the homeobox transcription factor and terminal selector unc-86, prevents dauers from expressing sri-9 in NSM neurons. The role of terminal selectors for the dauer-specific expression of certain GPCRs is further supported by the observation that the GPCR srh-71, which is selectively repressed in lateral IL2 neurons during dauer, remains expressed in unc-39(e257) dauers [19,96]. Both SRI-9 and SRH-71 remain orphan receptors, but differential expression in diapause by dauer regulators as well as neuronal identity regulators suggests critical roles in cues sensing to resume a developmental path.

• IL2 neurons arborization: Upon dauer entry, IL2 lateral neurons extend new dendritic trees and branches that retract after dauer exit [97]. This remodeling is necessary for the dauer-specific nictation behavior [97,98] and depends on two transcription factors: UNC-86 and the master regulator of ciliogenesis DAF-19. KPC-1, a proprotein convertase that is required for multi dendritic arborization, is upregulated in IL2 neurons during dauer and acts cell autonomously for the remodeling [97]. The roles of DAF-3, DAF-16 or DAF-12 were not addressed, most likely for being Daf-d, but UNC-86 and DAF-19 could be interacting with these or other TFs to activate KPC-1 downstream the dauer entry decision.

• Supernumerary muscle arms: During embryonic and post-embryonic development, the body wall muscles grow membrane extensions (muscle arms) towards the motor axons to establish synapses [99]. Dauers display supernumerary arm muscles (Sna phenotype) that have been linked to increased neurotransmitter sensitivity and locomotory behavior and persist after resumption of reproductive development [100]. daf-7(e1372) or daf-2(e1370) mutants grown at 20°C form L3 larvae that display Sna phenotype, suggesting that during dauer, reduced signaling from these TGF-β and IIS are responsible for the behavior. Both signaling pathways converge to inhibit daf-9 and accordingly, hypomorphic daf-9(m540) L3s are also Sna, which argues for low levels of DA to be associated with extra arm muscle extension. In agreement, the null allele daf-12(rh61rh411) or mutations in the DAF-12/VDR Ligand Binding Domain (LBD) suppress the Sna phenotype of daf-2(e1370), showing that active DAF-12/DA prevents the Sna phenotype. daf-16;daf-2 grown at 20°C L3s are not Sna and DAF-16/FOXO expression in muscle or intestine induces formation of supernumerary arms [100]. Together, these data argue that in dauers, DAF-16/FOXO functions cell autonomously in muscle and non-autonomously in the intestine to inhibit DAF-12/DA, leading to dauer supernumerary arms. Indeed, it was recently shown that DAF-16 is necessary for the Sna phenotype. Depletion of DAF-16 in muscle leads to daf-2(e1270) dauers that develop fewer muscle arms when compared to control [14]. Whether DAF-3/co-SMAD activity is required for the na phenotype has not been explored.

• Epidermal remodeling: Hypodermal seam cells are responsible for the radial shrinkage, secretion of dauer cuticle and alae formation. Partial dauers raised from dex-1 mutants are specifically defective for epidermal dauer features, missing the radial shrinkage and alae formation and consequently, showing reduced locomotion and increased SDS sensitivity [101]. DEX-1 is a transmembrane protein, which can also be cleaved and secreted [102]. During dauer, DEX-1 is expressed and secreted from seam cells, acting locally as an extracellular matrix component to promote epidermal dauer remodeling. When daf-16(m27) are supplemented with pheromone, they develop into partial dauers with reduced seam cell dex-1 expression [101]. Indeed, dex-1 is a direct transcriptional target of DAF-16/FOXO since deletion of the insulin response sequences (IRS) in dex-1 promoter reduces the expression in dauers at levels comparable to those of daf-16(m27) partial dauers [101].

• Dauer cuticle: The worm’s cuticle is a specialized extracellular matrix, composed predominantly of cross-linked collagens [103]. Dauers achieve a thicker and desiccation-resistant cuticle by changing its collagen composition. Transcriptomics comparing collagen genes in daf-16(mgDf50);daf-7(e1372) partial dauers vs. daf-7(e1372) full dauers identified 20 out of 22 adult-specific collagens mis-expressed in these particular partial dauers [89]. In addition, 23 out of 24 dauer-specific collagens were downregulated. Further experiments with col-19 showed that daf-16;daf-7 mutants precociously express this adult-specific collagen during dauer. Mechanistically, DAF-16/FOXO activates lin-41, which prevents col-19 expression in the hypodermal seam cells [89]. Of importance, this study was conducted in daf-7(e1372) background, where the DAF-3/DAF-5 complex should be active and could be promoting expression of dauer-specific collagens.

• Pumping cessation: Controlled by pharyngeal neurons, this prevention of unnecessary muscle contraction has been associated with reduced energy consumption and protection of muscles. Making use of the AID system, DAF-16/FOXO was shown to be required to inhibit pumping both cell-autonomously within pharyngeal neurons and non-autonomously acting from the nervous system, the intestine or pharyngeal muscle [14]. Similarly, overlapping roles for DAF-12/VDR have been reported in the nervous system and intestine, since its depletion in said tissues results in augmented pumping [14].

• Metabolic switch: As non-feeding, dauers need to survive for long periods from stored energy. For this, they shift from aerobic respiration to glycolysis and gluconeogenesis [1,104,105]. In this context, DAF-16/FOXO is required for the inhibition of catabolism, which allows rationing and avoids rapid depletion of reserves. daf-16(mu86) partial dauers grown in lophenol, exhibit a faster degradation of lipids, sugars and amino acids when compared to wild-type dauers. Consistent with a hyperactive catabolism, AID-mediated intestinal depletion of DAF-16/FOXO in daf-2(e1370) dauers leads to smaller lipid droplets in intestinal cells [14]. In addition, DAF-16/FOXO and DAF-12/VDR are required to inhibit the TCA cycle and promote gluconeogenesis as it was seen by molar quantification of enzymes involved in these metabolic pathways [105]. Interestingly, depletion of either DAF-16/FOXO or DAF-12/VDR in the nervous system reduces the lipid storage in the gut, showing a non-autonomous regulation of metabolic remodeling [14]. Downstream of both DAF-16-mediated metabolic functions is the activity of AAK-2 [105,106].

• Long-term longevity: Dauers can survive for up to 3-4 months. Some mutations can shorten this extended lifespan, as seen in daf-2(e1370);bec-1(RNAi) dauers that are deficient in autophagy and die within a few days at 25°C [30], highlighting how critical this process is during nutrient deprivation. Another example is the daf-2(e1370);aak-2(lf) dauer larvae that die within 12-14 days [32], attributed to increased lipolysis and depletion of reserves [106]. Similarly, din-1(rr94);daf-2(e1370) dauers live up to 15–16 days and the triple mutant aak-2;din-1;daf-2 die within 5-8 days [84]. The latter is consistent with DAF-16 and AAK-2 functioning in parallel with DAF-12 (and probably DIN-1) for metabolic remodeling [105]. Thus, the triple mutant results in a severely mis-regulated metabolism and systemic failure. In addition, other mutations that affect long-term dauer survival are not directly linked to metabolism but rather related to overall regulation of transcription. For instance, xrn-2 dauers die within 10 days, which is associated with defective miRNA machinery that leads to broad mis-expression of miRNAs and their targets [107,108]. Downregulation of the transcription factors blmp-1 and lin-40 by RNAi, caused daf-7(e1372) dauers to die prematurely [109]. These TFs are involved in dauer development and regulation of the heterochronic circuit and have been implicated in histone remodeling [109,110]. As suggested for xrn-2 mutants, their downregulation could be leading to a broad mis-expression of targets and subsequent systemic collapse.

• Nictation: Dauers can stand and wave their bodies to attach onto passing-by animals,increas the chances of being moved to favorable environments [1]. This behavior is specifically mediated by IL2 neurons [98]. Dauers arising from mutations in IIS signaling exhibit enhanced nictation when compared to wild-type dauers [111]. Increased nictation is partially lost in daf-16(mgDf50);daf-2(e1370) dauers carrying extrachromosomal arrays for different DAF-16 isoforms, arguing for it to be a DAF-16-mediated behavior. Mechanistically, the p21-activated kinase regulator max-2, that has critical functions in neuronal remodeling, is placed downstream IIS for nictation behavior [111]. Surprisingly, downregulation of TGF-β by mutations in all Daf-c components (daf-7, −1, −4, −8, −14) inhibits nictation, providing the first report of these two signaling pathways having opposite effects on the same phenotype [111]. The null mutant daf-3(mgDf90) did not enhance nictation, suggesting a DAF-3-independent transcriptional regulation downstream of TGF-β signaling. Rescue experiments to find critical neurons for IIS and TGF-β signaling in the modulation of nictation pointed at ASI/ASJ and RIM/RIC, respectively [111].

It is noteworthy that while DAF-16/FOXO and DAF-12/VDR activities are broadly required for execution, DAF-3/co-SMAD, active in reduced TGF-β signaling, is not Figure 3. So far, DAF-3/co-SMAD has only been reported to play a role in germline quiescence, where it directly inhibits LAG-2 expression [88]. The reduced TGF-β signaling could contribute to the dauer remodeling by interacting with other pathways, like IIS [46,74]. In addition, DAF-3/co-SMAD directly binds to daf-7 and daf-8 promoters to repress their transcription [37,88]. It is possible that when dauering, DAF-3/DAF-5 function to completely shut down the TGF-β signaling.

4.1. Additional transcriptional regulators involved in dauer development

• SWI/SNF complex. This ATP-dependent chromatin remodeling complex regulates gene expression by modifying DNA accessibility to the transcriptional machinery. In daf-2(e1370), inhibition of the swns-1, -4 and -8 complex subunits by RNAi led to arrest as SDS-sensitive partial dauers, arguing for a role in cuticle morphogenesis [112]. Consistent with this observation, RNAi-mediated silencing of swns-1 and -8 in daf-7(e1372) animals also induced partial dauers [47]. Co-immunoprecipitation studies identified a physical interaction between DAF-16/FOXO and the SWI/SNF complex and were found to co-localize in the chromatin [112]. Moreover, the expression of DAF-16/FOXO targets depends on this chromatin remodeler complex [112]. Interestingly, a yeast two-hybrid approach identified that DAF-3/co-SMAD physically interacts with SWSN-1/SMARCC2, suggesting that the SWI/SNF complex could be associating in multiple transcriptional complexes to mediate dauer features [113].

• BLMP-1/PRDM1. B lymphocyte-induced maturation protein-1 is a transcription factor involved in the heterochronic circuit, and mutants display gonad migration and seam cell defects. Wild-type, daf-2(e1368), daf-7(e1372), and daf-9(dh6) dauers treated with blmp-1 RNAi arrest as SDS-sensitive L3/L4 larva that lack alae, displaying a partial dauer phenotype [114]. blmp-1 expression peaks at L2d in daf-7 mutants and then decreases upon dauer entry [109]. During reproductive development, BLMP-1/PRDM1 interacts with DRE-1 to promote gonad migration and seam cell terminal specification [114]. Under stress-inducing conditions, BLMP-1/PRDM1 interacts with LIN-40 to promote dauer development [114]. BLMP-1 can physically bind to a subunit of the SWI/SNF complex called HAM-3 [115], and blmp-1(RNAi) partial dauers share overlapping phenotypes with partial dauers formed by RNAi inhibition of SWI/SNF subunits [112,114]. Target genes of BLMP-1/PRDM1 and HAM-3, including col-124, lin-29 and bed-3, are mainly expressed in the hypodermis, suggesting that BLMP-1/PRDM1 and SWI/SNF could be partnering in the hypodermis for dauer-specific regulatory mechanisms [109].

• CHD-7/CHD7/8. The Chromodomain Helicase DNA binding protein 7 was originally identified in a ChIP-chip assay for the nuclear receptor DAF-12/VDR [82]. chd-7(tm6139);daf-2(e1370) arrest as partial dauers at the non-permissive temperature of 25°C and display dauer alae defects, over-proliferation of the germline and gonad migration defects, all consistent with CHD-7/CHD7/8 having a role in dauer morphogenesis [47]. In contrast, chd-7 mutants bypass the dauer arrest (Daf-d phenotype) in all Daf-c mutant backgrounds belonging to the TGF-β pathway. At higher temperatures like 26.5°C or by daf-9 downregulation, chd-7(tm6139);daf-7(e1372) arrest as SDS-sensitive partial dauers, suggesting that CHD-7/CHD7/8 is still required for some aspect of dauer morphogenesis [47]. One possibility is that CHD-7/CHD7/8, by regulating the BMP/DBL-1 pathway, prevents proper expression of dauer cuticle-specific collagens. This is further supported by the integration of ChIP-seq and RT-qPCR data, showing that CHD-7 binds and regulates dbl-1 expression and downstream components of the Sma/Mab pathway [47]. In addition, chd-7 mutants show aberrant upregulation of daf-9 expression, which might lead to increased DAF-12/DA signaling and defective morphogenesis. The BRK domain is critical for CHD-7 function in dauer entry and morphogenesis [47]. Using the WormBase (version WS286) motif resource, BRK is only found in C. elegans in CHD-7 and the SWI/SNF subunit SWSN-4 [38]. The NMR structure of the BRK domain from BRG1, the human ortholog of SWSN-4, resembles the structure of the glycine-tyrosine-phenylalanine (GYF) domain, which functions as a protein-protein interaction module [54]. Thus, CHD-7’s BRK domain likely mediates interactions with other factors to modulate the transcriptional activity necessary for dauer. Identifying CHD-7/CHD7/8 interacting partners would be instrumental to fully understand its role in dauer development.

5. Maintenance

Maintenance refers to the ability to remain developmentally arrested when dauers are still facing unfavorable environmental conditions. Mutants defective in this step of the dauer program spontaneously exit diapause. Such is the case of worms carrying mutations in the Notch receptor GLP-1 or its ligand LAG-2, revealing how critical Notch signaling is for dauer maintenance. daf-7(e1372);lag-2(q420) or glp-1(e2141);daf-7(e1372) mutant dauers aberrantly resume development within 24 h, despite being kept at high temperatures (25°C) [13]. Laser ablation of IL2 neurons, which express LAG-2, leads to dauer maintenance defects comparable to those observed in daf-7; lag-2. In addition, ablation of AWC neurons in glp-1; daf-7 double mutants suppress maintenance defects, suggesting that these neurons mediate the aberrant recovery [13]. This argues that LAG-2/GLP-1 Notch pathway functions in the nervous system to ensure the arrested state and inhibit exit. The dauer-specific expression of LAG-2 in IL2 neurons is placed downstream IIS, TGF-β and the steroid hormone pathway, since Daf-c mutations in these pathways lead to similar levels of lag-2::GFP [13]. During reproductive development, the forkhead transcription factor UNC-130 inhibits lag-2, and this repression is lost during dauer. During dauer, unc-130 mutants do not mis-express lag-2::GFP, suggesting that this TF only regulates lag-2 in the reproductive path. A dauer conditional transcription factor might be needed to activate lag-2, although DAF-16/FOXO can be ruled out since lag-2::GFP expression was unaffected in daf-16(mgDf50) null mutant dauers [13]. The AMP-activated kinase AAK-2 also contributes to dauer maintenance. Double mutants daf-7(e1372);aak-2(rr48) develop dauers that recover within 5 days at 25°C [32]. The mechanism behind AAK-2 requirement for dauer maintenance is unknown, but since aak-2 mutant dauers display defects in germline and metabolic remodeling, this defective morphogenesis may be the underlying cause [32,92,105].

As shown previously, DAF-3/co-SMAD, DAF-16/FOXO and DAF-12/VDR have many roles in the dauer program, including regulation of entry and execution. Uncoupling dauer maintenance transcriptional regulation from that of entry and execution was achieved by employing the AID system to degrade these TFs [14]: they formed full dauers and then shifted them to auxin containing plates to induce depletion. With this approach, authors found that the ubiquitous depletion of DAF-3/co-SMAD or DAF-12/VDR in daf-7(e1372) full dauers resulted in worms spontaneously exiting diapause at 25°C [14]. Depletion of DAF-16/FOXO in daf-2(e1370) dauers led to the strongest defective maintenance phenotype with all dauers recovering as early as 4 days after depletion. Considering that for dauer entry, DAF-3/co-SMAD and DAF-16/FOXO are only strictly required in daf-7 or daf-2 mutant backgrounds, respectively, it would be interesting to see if similar maintenance loss upon DAF-3/co-SMAD and DAF-16/FOXO depletion is observed in wild-type dauers. It is somewhat unexpected that DAF-3/co-SMAD, which does not play a prominent role in execution (see Execution), is indeed required for maintenance of daf-7(e1372) dauers. As mentioned before, DAF-3/co-SMAD might be ensuring reduced signaling of TGF-β by a negative feedback loop and, like so, contributing to stay developmentally arrested [37,88]. Overall, DAF-3/co-SMAD, DAF-16/FOXO and DAF-12/VDR are actively required to maintain the remodeled state achieved in the execution step and to prevent dauer exit [14]. In addition, aberrant activity of their downstream effectors like AAK-2 and Notch signaling is sufficient to initiate the exit program, resulting in release of the developmental arrest regardless of environmental conditions [13,32].

6. Exit: the resumption of reproductive development

The dauer state is maintained until favorable conditions trigger exit or recovery from diapause, causing the dauer to lose its specialized features necessary for survival and molt into a L4 larva [1]. When dauer start to recover, the first and most noticeable feature is the resuming of pharyngeal pumping [1]. Sensory neurons such as ASJ function in both dauer entry and exit. Others, like the AWC and IL2, have roles in promoting or inhibiting recovery, respectively [4,13]. During dauer, worms go through specific remodeling of sensory neurons [78,116] and express a different subset of GPCRs [19], suggesting that they tune their sensitivity to detect exit cues. Consistent with this, glial GPCR REMO-1 is required in AMsh glia for AMsh and AWC dauer-specific neuronal remodeling [117]. In WT dauers, AMsh glia fuse and AWC receptive endings expand beyond the midline, making them more sensitive to favorable stimuli. In contrast, remodeling-defective remo-1(-) mutants show a significant delay in dauer recovery after exposure to bacteria [117].

Both exit and entry involve the same signaling pathways, while they are reduced to induce entry; they need to be activated for exit. For instance, INS-6 functions to activate IIS in both scenarios: it is expressed in ASI neurons to prevent dauer entry but is expressed in ASJ neurons to promote dauer exit [17]. What is more, entry and exit involve the same transcriptional regulation: DAF-8/NHR-69 complex promotes DAF-28 expression in ASI neurons and activates IIS to inhibit entry or favor exit [46]. DAF-3/co-SMAD, DAF-16/FOXO or DAF-12/VDR are induced to promote entry (see Entry) and are antagonized to promote exit, as evidenced by AID-mediated depletion in which worms exit diapause regardless of environmental conditions (see Maintenance). Likewise, the activation of steroid hormone by liganded DAF-12/VDR is similar in entry and exit; in both cases, it can be modulated by nM concentrations of DA and triggered by daf-9 expression in the hypodermis alone [28,62,118].

Recent studies in daf-9(dh6), which lack endogenous DA, shed light onto the process of recovery. When exposed to unfavorable conditions, such as high temperature or presence of dauer pheromone, daf-9 mutants form normal dauers. By contrast, when faced with a favorable environment, such as low temperature and no pheromone, they form partial dauers that have increased mobility, higher rate of pharyngeal pumping and wider pharyngeal bulbs [62,118]. By performing shifting experiments from one experimental condition to the other and using these phenotypes as a redout, was shown that daf-9 worms are indeed dauer exit defective [118]]. Grown in favorable conditions, worms developed a transient full dauer state and ended as partial dauers that could not be induced to form full dauers, resembling the irreversibility of exit decision [118]. Interestingly, daf-2(e1370);daf-9(dh6) or daf-7(e1372);daf-9(dh6) double mutants formed partial dauers with milder defects and more dauer-like features [118]. Overexpression of either daf-2 or daf-7 exacerbated the partial dauer phenotype, suggesting that active IIS and TGF-β signaling were responsible for the defective exit phenotype [118]. Despite the lack of DA preventing daf-9 mutants to progress to reproductive fates, they still execute a partial exit in response to favorable environments. This suggests that recovery involves two sequential events: 1) activation of IIS and/or TGF-β in response to favorable environments, which will trigger the irreversible commitment to dauer exit (partial exit) and 2) activation of DAF-12/DA, which will trigger commitment to reproductive fates (complete exit) [118]. As there is a DA threshold to be met for dauer bypass during continuous development [28], presumably there is also a DA threshold for post-dauer commitment to reproductive development Figure 1.

While dauer exit involves activity from IIS and TGF-β, it is noteworthy that the IIS has a more predominant role. When daf-9(dh6) facing favorable conditions are induced to overexpress daf-7 or daf-2, the latter one elicits a stronger partial exit, as it results in the locomotory behavior of a L3 larva [118]. In agreement, previous studies showed that reduced IIS signaling impedes recovery, while dauers with reduced TGF-β can still resume development [13,32,119]. The induction of dauer exit by supplementing with exogenous muscarinic neurotransmitters promotes dauer exit in wild-type and daf-7 backgrounds, while daf-2 worms fail to recover [119]. Similarly, mutations in glp-1 or aak-1 promote dauer exit in wild-type and daf-7 but not in a daf-2 background [13,32]. As seen in Maintenance, loss of DAF-16/FOXO results in the most penetrant exit phenotype, suggesting that for dauer recovery the inhibition of DAF-16/FOXO is critical [14]. Since DAF-16/FOXO drives and maintains the remodeling of many tissues during dauer, active IIS and repression of DAF-16/FOXO probably triggers the loss of dauer-specific morphology and behavior. As for TGF-β signaling, its activity is required for sensing favorable conditions, but this contribution is small and perhaps related to the fact that DAF-3/co-SMAD is not broadly required for dauer remodeling. Indeed, TGF-β contributions to dauer exit might rely solely on feeding the IIS by controlling the expression of insulins (see Entry: TGF-β and IIS show complex cross-talk).

Full dauer exit requires DA to resume reproductive development. To meet the hypothetical dauer exit DA threshold, daf-9 needs to be upregulated. The regulation of daf-9 expression by IIS and TGF-β was only demonstrated by genetic epistasis, but whether DAF-3/co-SMAD or DAF-16/FOXO directly regulate daf-9 expression remains unknown. Interestingly, XXX cells are not essential for exit since its ablation in dauers does not prevent the molt into L4 [118]. This is in contrast to XXX cell ablation during L2d, where worms cannot perform the L2d/L3 transition [28]. In the continuous development trajectory, hypodermal daf-9 expression is triggered from anterior to posterior starting in the head near to XXX cells, suggesting that DA secreted from the XXX cells triggers the amplification of daf-9 [28]. This is possible because DAF-12/DA can promote daf-9 expression as a part of a positive feedback loop [28]. In the dauer trajectory, if XXX cells are not essential for recovery, a signal within the hypodermis itself could trigger daf-9 expression and reproductive fate commitment. Alternatively, a signal from the gut could be responsible. Recently, it was shown that upon exit decision, probably downstream IIS and TGF-β (step 1 of Exit), SCL-12 and −13 transport cholesterol stored in the gut lumen to the intestinal cell lysosomes for its release [86]. There, cholesterol seems to activate mTOR pathway, which is required for dauer exit. daf-2(e1370), scl-12(-), scl-13(-) mutants are unable to deliver cholesterol to the lysosomes and therefore cannot exit dauer when shifted to a lower temperature [86]. In these mutants, exit can be fully rescued by constitutive activation of mTOR by nprl-3 RNAi or by supplementation with 200 nM DA [86]. From these isolated observations, one cannot establish a cause-and-effect relationship between mTOR activity and DA production but considering the role of DA in reproductive fate commitment, one could hypothesize that mTOR pathway functions as an energy sensor in response to cholesterol availability which then triggers protein synthesis and growth and ultimately leads to production of cholesterol derived DA and recovery. This coupled to the fact that the expression of the Rieske-like oxygenase DAF-36, that converts cholesterol into 7-dehydrocholesterol for DA biosynthesis, is mainly intestinal [59] argues for the intestine as an important signaling hub for dauer exit.

7. Concluding remarks

The dauer program functions to halt developmental progression (entry) and induce extensive changes for long-term survival (execution). When thinking of relevant tissues for entry and exit, the nervous system is an obvious one as amphid sensory neurons directly sense external cues. Additionally, the intestine senses internal energy states and informs the nervous system by altering DAF-7/TGF-β and insulin expression for entry decision [8] and acts in a yet not fully understood mechanism for exit [86]. The integration of this information leads IIS and TGF-β pathways to converge on daf-9 thereby coupling environmental information to DA synthesis. For dauer entry, the XXX cells are the relevant site of DA production while for exit, the hypodermis produces appropriate levels of DA for adult morphogenesis [28,62]. After the entry decision is made, the dauer promoting transcription factors are broadly required to instruct dauer fates in nearly every tissue (nervous, pharynx, hypodermis, muscle, intestine, etc.) and to repress reproductive development (hypodermis, vulva, gonad, and germ line) thereby contributing to both execution and maintenance.

The transcription factors and regulators reviewed herein function at different steps of the dauer program, have tissue-specific roles, and achieve a coordinated remodeling throughout the organism. How is this regulation achieved? Temporal selectors called heterochronic genes regulate the timing of cellular fates across tissues. The heterochronic gene network, that includes daf-12 and din-1, regulates genes to instruct developmental arrest [24]. In addition, tissue- or cell-specific functions are regulated by partnering with different transcription factors in a combinatorial way. For dauer-promoting transcription factors, their activation is conditional to the very early environmental cues, but they can work together with non-conditional TFs to selectively target genes. In the nervous system, terminal selectors of neuronal identity like UNC-39 or UNC-86 transcription factors are required for stage-specific dauer remodeling [14,96]. For instance, if DAF-16/FOXO is activated in dauer-inducing conditions, a different set of genes will be expressed in neurons that already express UNC-86 [14]. Yet another example of combinatorial interactions between transcription factors involves DAF-8/R-SMAD: it can bind to DAF-14 in the DTCs to antagonize the formation of the pro-quiescence DAF-3/DAF-5 complex [37,87,88], but also can bind to NHR-69/HNF4a in ASI neurons to ultimately promote expression of the pro-reproductive development DAF-28 insulin [46]. Thus, DAF-8/R-SMAD promotes reproductive development through distinct transcriptional complexes in different tissues. We also described epigenetic regulators, including the SWI-SNF complex, CHD-7/CHD7/8 and BLMP-1/PRDM1, which probably function to make chromatin more accessible to dauer-inducing transcription factors thereby allowing expression/repression of relevant target genes.

The dauer larva in C. elegans is considered a great model to study plasticity in an organism with otherwise stereotypical and invariant cell-lineages [9]. Besides, the dauer program is also providing developmental robustness to environmental and genetic perturbations [82]. For instance, DAF-12/DA functions in a positive feedback loop with hypodermal daf-9 [28,62] and in a negative feedback loop with members of the let-7 family of miRNA to robustly lock in the decision between dauer bypass or entry [66,67]. Because DA is a diffusible hormone, it can bind to and modulate DAF-12/VDR throughout the whole animal. There is even a combination of feedback loops and cross-talk involving IIS and TGF-β, where a DAF-8/NHR-69 complex functions to ultimately activate IIS and repress DAF-16/FOXO activity. The expression of the components in this complex is regulated by IIS and TGF-β: DAF-16/FOXO directly represses nhr-69 and DAF-3/co-SMAD directly represses daf-8 [37,46]. Reduced signaling from TGF-β and IIS, and active DAF-3/co-SMAD and DAF-16/FOXO, converge to inhibit the formation of the DAF-8/NHR-69 pro-reproductive development complex. The cross-talk and feedback loops amplify a specific signal, dauer or reproductive fates to ensure robustness.

Studying the development of a simple organism such as C. elegans can help to better understand the etiology behind human developmental syndromes. By screening DAF-12/VDR targets for dauer phenotypes, we found that CHD-7/CHD7/8 pleiotropically regulates both TGF-β pathways (i.e., DAF-7 and Sma/Mab), affecting dauer development and cuticle deposition, which is a modified extracellular matrix (ECM). In humans, disruptive mutations in Chd7 are associated with CHARGE syndrome, a rare and severe neurodevelopmental disorder with no known treatment [120,121], but functionally relevant targets of CHD7 related to disease pathology are still poorly understood. We found a conserved function for Chd7 in vertebrates, as Chd7 depletion in Xenopus laevis embryos leads to col2a1 downregulation, affecting the ECM [47]. Interestingly, this is necessary and sufficient to develop the craniofacial defects typically observed in CHARGE patients, suggesting that the ECM is an evolutionary conserved target for Chd7 [47].

Funding Statement

The work was supported by the Conicet [PIP 1122015 0100731 CO].

Disclosure statement

No potential conflict of interest was reported by the authors.