Influence of environmental temperature on mouth-form plasticity in Pristionchus pacificus acts through daf-11-dependent cGMP signaling

Mouth-form plasticity in the nematode Pristionchus pacificus has become a powerful system to identify the genetic and molecular mechanisms associated with phenotypic (developmental) plasticity. In particular, the identification of developmental switch genes that can sense environmental stimuli and reprogram developmental processes has confirmed long-standing evolutionary theory. Together with the associated gene regulatory networks, these developmental switch genes have been important to show that plasticity is consistent with the Modern Synthesis of evolution. However, how these genes are involved in the direct sensing of the environment, or if the switch genes act downstream of another, primary environmental sensing mechanism, remains currently unknown. Here, we study the influence of environmental temperature on mouth-form plasticity. Using forward and reverse genetic technology including CRISPR/Cas9, we show that mutations in the guanylyl cyclase Ppa-daf-11, the Ppa-daf-25/AnkMy2 and the cyclic nucleotide-gated channel Ppa-tax-2 eliminate the response to elevated temperatures. Together, our study indicates that DAF-11, DAF-25 and TAX-2 have been co-opted for environmental sensing during mouth-form plasticity regulation in P. pacificus. This work suggests that developmental switch genes integrate environmental signals including perception by cGMP signaling.


Introduction
Distinct environmental conditions can shape development, resulting in different, sometimes even alternative phenotypes (Stearns, 1989;Pigliucci, 2001). Such 'developmental (phenotypic) plasticity' has been argued to facilitate evolutionary novelty and the origin of complex traits (West-Eberhard, 2003), a claim that has subsequently been confirmed in different study systems including vertebrates and invertebrates (Levis et al., 2018;Susoy et al., 2015;Parker and Brisson, 2019;van der Burg et al., 2020). As such, developmental plasticity is of importance for both developmental biology and evolutionary biology and highlights the significance of environmental factors (Sommer, 2020). However, the mechanistic basis of developmental plasticity was long limited to hormones (Suzuki and Nijhout, 2006). Only in the last decade did studies begin to identify associated developmental switches and gene regulatory networks (GRN). To this end, studies in genetically identical organisms that reproduce by asexual reproduction or self-fertilization (hermaphrodites) have provided important molecular insight, which includes among others, work in the water flea Daphnia, the pea aphid Acyrthosiphon pisum, hymenopterans with their elaborate caste systems and the self-fertilizing nematodes

Caenorhabditis elegans and Pristionchus pacificus.
To validate the significance of plasticity for development and evolution, two lines of molecular evidence must be taken into account. First, what is the genetic and molecular basis of plastic trait development and how does it differ from 'hard-wired' developmental processes that are insensitive to environmental input? Second, how is the environmental information sensed and subsequently transmitted resulting in the formation of different phenotypes? The nematode Pristionchus pacificus has emerged as one promising model organism for the investigation of phenotypic plasticity with forward and reverse genetic tools . P. pacificus shares with C. elegans many life history traits, including hermaphroditic reproduction and a short life cycle that is completed after four days when grown on standard nematode-growth-medium (NGM) agar plates with E. coli OP50 as food source at 20 °C . C. elegans and P. pacificus are members of different nematode families, the Rhabditidae and Diplogastridae, respectively, and are thought to have separated roughly 100 million years ago Werner et al., 2018a). Given its hermaphroditic mode of reproduction, P.
pacificus is amenable to genetic manipulation including forward genetic mutagenesis, CRISPR knockouts and engineering and DNA-mediated transformation Nakayama et al., 2020;Han et al., 2020). The P. pacificus genome was originally sequenced in 2008 (Dieterich et al., 2008) with more recent single molecule re-sequencing and transcriptomic efforts, which indicated that the roughly 170 Mb genome contains approximately 28,000 genes (Rödelsperger et al., , 2019Athanasouli et al., 2020).
One of the most prominent differences between P. pacificus and C. elegans is that P. pacificus exhibits plasticity of its feeding structures with two alternative mouth-forms that contain teeth-like denticles (Bento et al., 2010). These structures represent novel and complex traits unknown from most nematodes. Specifically, adult P. pacificus express either a narrow stenostomatous (St) morph with a single flint-like dorsal tooth, or a wide eurystomatous (Eu) morph, which has a claw-like dorsal tooth and an enlarged right subventral tooth ( Figure 1A). Importantly, juveniles and adult St animals are strict bacterial feeders, whereas adult Eu animals can additionally kill other nematodes and subsequently feed on them ( Figure 1A). Thus, morphological mouth-form plasticity results in alternative behaviors and feeding strategies.
Several aspects of the mouth-form dimorphism have made it a powerful system to investigate the genetics of phenotypic plasticity. First, the discrete polyphenism allows categorical phenotyping without intermediate forms. Second, the hermaphroditic mode of reproduction mitigates any concern that genetic polymorphisms contribute to experimentally observed phenotypes. Third, hermaphroditism allows for the isolation of mutants that affect mouth-form ratios (Ragsdale et al., 2013). Finally, environmental influence can easily be manipulated under laboratory conditions. For example, when wild type worms of the P. pacificus reference strain PS312 are cultured on NGM agar plates they exhibit predominantly the Eu morph (>90% Eu), whereas they are only 10% Eu when grown in liquid culture . Thus, P.
pacificus mouth-form plasticity can be genetically and environmentally manipulated allowing for a detailed analysis of associated molecular mechanisms.
What is the genetic basis of mouth-form plasticity? A series of forward and reverse genetic investigations in the last decade has gradually resulted in the discovery of a mouth-form gene regulatory network (MF-GRN) that includes i) developmental switches, ii) at least two transcription factors of the nuclear-hormone-receptor family, and iii) their downstream targets ( Figure 1B) (Bui et al., 2018;Kieninger et al., 2016;Namdeo et al., 2018;Ragsdale et al., 2013;Serobyan et al., 2016;Sieriebriennikov et al., 2018, 2020, Bui and Ragsdale, 2019Casasa et al., 2020).
The first gene to be identified came from eurystomatous-form-defective eud-1 mutants, which remained St animals under all culturing conditions tested (Ragsdale et al., 2013). Most importantly, eud-1 is dose sensitive and thus represents a developmental switch. This finding has confirmed long-standing predictions about developmental switches as key elements of developmental plasticity, which allow environmental sensing and developmental reprogramming (West-Eberhard, 2003;Ragsdale et al., 2013;Sommer, 2020). Interestingly, eud-1 resides in a conserved 'multi-gene locus' and is surrounded by tandem-duplicated genes (nag-1 and nag-2), which promote the St morph . While nag-1 and nag-2 mutants are all-Eu, eud-1 mutant animals are all-St with eud-1 being epistatic .
All three genes, eud-1, nag-1 and nag-2 are expressed in different sensory neurons. The identification of switch genes that govern intra-generational plasticity is important to confirm that plasticity is indeed consistent with the Modern Synthesis of evolution (Baugh and Day, 2020).
However, it remained unclear if these genes are involved in the direct sensing of the environment or if they act downstream of another, primary environmental sensing mechanism.
Here, we study the influence of environmental temperature on mouth-form plasticity ( Figure 1C). We show that mouth-form plasticity in many P. pacificus isolates is strongly sensitive to the temperature of the environment. We focus on the isolate RSA635 from La Réunion island that is 70% Eu at 20°C, but only 20% Eu at 27 °C. Using a combination of forward and reverse genetic approaches, we found that cGMP signaling is involved in temperature perception, which in turn regulates mouth-form plasticity. First, mutations in the guanylyl cyclase Ppa-daf-11 and the Ppa-daf-25/AnkMy2 eliminated the response to elevated temperatures as revealed in forward genetic screens. Second, reverse genetic approaches showed that the cyclic nucleotide-gated channel Ppa-tax-2, but not the Forkhead-type transcription factor Ppa-daf-16 required for temperature perception. Together, our study indicates that the DAF-11, DAF-25 and TAX-2 orthologs have been co-opted for environmental sensing during mouth-form plasticity regulation in P. pacificus. This work suggests that developmental switch genes integrate environmental signals that at least in part are perceived by cGMP signaling and potentially other primary signaling systems.

P. pacificus shows strain-specific temperature sensitivity of mouth-form plasticity
To study the molecular mechanisms of environmental sensing during mouth-form plasticity regulation in P. pacificus, we investigated the influence of temperature as an abiotic factor for several reasons. First, temperature is known to be involved in the regulation of plastic traits in a diversity of animals and plants (Pigliucci, 2001). Second, the molecular mechanisms of temperature perception in development and physiology have been intensively studied in several model organisms, including comprehensive investigations in C. elegans (for review see, Goodman and Sengupta, 2019). Finally, P. pacificus exhibits higher temperature tolerance than C. elegans with several tested wild isolates being capable of reproducing at 30 °C . Therefore, we first compared the mouth-form ratios of the wild type strain P. pacificus PS312 (RS2333) grown at 10 °C, 15 °C, 20 °C, 25 °C and 27°C on standard NGM agar plates ( Figure 1C). At most temperatures we observed more than 80% Eu animals, which means the observed temperature effect was not very strong (Figure2). Only at temperatures below 15 °C, when worms grow very slowly, we found less than 50% of animals having the Eu mouth form.
These findings suggest that temperature has a limited effect on mouth-form ratio in P. pacificus RS2333, which might result from an original adaptation of this genotype or from domestication processes. Note that RS2333 is a direct derivate of the original PS312 strain that has been isolated in Pasadena, California in 1988 and has been in laboratory culture ever since . Next, we investigated 10 natural isolates of P. pacificus that cover the large genetic diversity of P. pacificus, including several isolates from the tropical island La Réunion in the Indian ocean (Rödelsperger et al., 2014, McGaughran et al., 2016. Most noticeably, none of these strains has undergone a long period of domestication in the laboratory and most of them have been frozen in the first 20 generations after isolation. Indeed, testing the same set of temperatures in these strains revealed a strong sensitivity of mouth-form plasticity to environmental temperature (Figure2). Specifically, many strains show a bell-like curve with the highest percentage of Eu animals observed at 20 °C, such as in RSA622, RSA635 and RSA645.
In other strains, the highest Eu ratios are found at 15 °C or 25 °C culture conditions. Additionally, some strains, like RS5160, do not form more than 50% Eu animals at any temperature. We conclude that in the large majority of P. pacificus strains environmental temperature has an influence on mouth-form plasticity, largely in a strain-specific pattern, and that this effect can likely be subjected to genetic investigations.

Forward genetic screens identify temperature sensing-deficient mutants
To initiate genetic analysis of temperature perception of P. pacificus mouth-form plasticity, we selected strain RSA635 with its bell-shaped temperature response curve. At 20 °C, around 70% of the animals express the Eu mouth form, whereas only 20% are Eu at the two extreme temperatures 10 °C and 27°C ( Figure 2, Table 1A). We performed EMS mutagenesis with RSA635 J4 juveniles using the standard P. pacificus mutagenesis protocol (Aurilio and Srinivasan, 2015). Subsequently, we examined the F3 progeny of 2,400 F2 animals for high Eu mouth-form ratios after culturing the F3 generation at 27 °C. We isolated 13 mutant candidates with >70% Eu animals at 27 °C. After backcrossing, we recovered 11 mutant lines with elevated Eu mouth-form ratios. Several of the mutant strains, however, showed high variability in mouthform ratios. Three of these 11 mutant lines, tu715, tu719 and tu724, were 100% Eu at 27 °C (Table 1B). This phenotype is similar to the phenotype of mutations in the nag-1 and nag-2 genes in the P. pacificus wild type, which form the multi-gene switch locus .
To test whether these mutant lines affect the mouth-form switch, we phenotyped them at 20 °C and performed whole genome sequencing (WGS) to identify potential causal mutations.
Indeed, all three mutant lines were 100% Eu at 20 °C and thus, the effect on mouth-form ratio in these mutant lines is temperature-independent (Table 1B). Additionally, WGS identified mutations in nag-1 in all three mutant lines. Specifically, tu715 has a nonsense mutation, whereas tu719 and tu724 result in nonsynonymous changes in Ppa-nag-1 (Table 1B). Together, these experiments indicate that mutants with altered temperature perception during plasticity regulation can be isolated in P. pacificus RSA635. Furthermore, these findings suggest that the multi-gene locus controlling mouth-form switching is conserved among strains and acts in a temperature-independent manner.

tu716 is caused by a mutation in daf-25, an ortholog of mammalian Ankmy2
From the remaining eight mutant lines with elevated Eu mouth-form ratios at 27 °C, we selected two lines for further analysis. These mutants, tu716 and tu722, showed the most consistent phenotype over multiple generations. For example, tu716 is 99% Eu at 27 °C and 90% Eu at 20 °C, a phenotype that is less extreme than the one observed in the three Ppa-nag-1 alleles (Table 1C). We mapped tu716 through a modified bulk segregant analysis regime (see Materials and Methods for details). In short, we sequenced 23 mutant animals after backcrossing with RSA635 wild type. We generated Tn5 single-worm libraries for all 23 mutant animals, includinga similar number of non-mutant control animals from the same cross. After sequencing, we performed SNP enrichment in the mutant batch and identified a potential mutation in the P. pacificus ortholog of daf-25. In C. elegans, daf-25 mutants show a temperature-sensitive dauer formation-constitutive (Daf-c) phenotype (Jensen et al., 2010). daf-25 is localized in cilia of sensory neurons and encodes the ortholog of the mammalian Ankmy2, a MYND domain protein of unknown function (Jensen et al., 2010).
tu716 exhibits a missense mutation in the predicted protein pocket of Ppa-DAF-25 resulting in a V > E amino acid change (Table 1C). To verify that the temperature sensing defect of tu716 is indeed due to the mutation in Ppa-daf-25, we generated additional alleles by CRISPR knockout, a method that has been applied successfully in previous studies (Moreno et al., 2017). We isolated two CRISPR-induced mutants in Ppa-daf-25 using a guide RNA in the fourth exon, close to the site of the mutation in the original tu716allele. The new alleles, Ppa-daf-25 (tu1516) and Ppa-daf-25 (tu1517), have an 11 bp insertion and an 8 bp deletion, respectively (Table 1C). Both alleles are highly Eu at 27 °C, similar to Ppa-daf-25 (tu716) (Table 1C). However, none of the Ppa-daf-25 alleles displayed the Daf-c phenotype. We conclude that Ppa-daf-25 is involved in temperature perception of mouthform plasticity. These results are of particular interest because in C. elegans, DAF-25 is known to be required for the proper localization of the guanyl cyclase DAF-11 in cGMP signaling in cilia (Jensen et al., 2010).

A daf-11/guanylyl cyclase mutant is high temperature insensitive
Next, we used a similar mapping strategy for the second mutant with a strong temperature-sensing defect. tu722 is 93% Eu at 27 °C and 93% Eu at 20 °C (Table 1). Using the same bulk segregant analysis regime we found a mutation in Ppa-daf-11 in exon 18, resulting in a premature stop codon (Table 1D). This observation would be consistent with the known interaction of daf-25 and daf-11 in C. elegans and with the results described above. We used a CRISPR knockout approach to confirm this observation by generating additional mutants in Ppa-daf-11. Specifically, we used one sgRNA to induce mutations close to the N-terminus of the gene in exon 3 and a second sgRNA in exon 18, the same exon that harbors the tu722 mutation. The sgRNA in exon 3 resulted in five mutant F2 lines with deletions in Ppa-daf-11. However, all these lines were sterile so that no homozygous mutant line could be generated. This finding suggests that loss-of-function or strong reduction-of-function alleles of Ppa-daf-11 cannot be kept as homozygous mutant lines. In contrast, we isolated three viable alleles with mutations in exon 18. Specifically, Ppa-daf-11 (tu1438) and Ppa-daf-11 (tu1439) have two bp deletions and Ppa-daf-11 (tu1440) has a four bp insertion, all of which result in premature stop codons. All three of these mutant lines show strongly elevated Eu mouth-form frequencies at 27 °C, similar to the original Ppa-daf-11 (tu722) allele (Table 1D). Thus, the guanylyl cyclase Ppa-daf-11 is involved in temperature perception during mouth-form plasticity regulation. The fact that Ppa-daf-11 and Ppa-daf-25 have similar phenotypes is consistent with the previously described role of Cel-DAF-25 in the proper localization of Cel-DAF-11 to cilia in C. elegans (Jensen et al., 2010).
Together, these findings allow two conclusions. First, cGMP signaling involving Ppa-daf-11 and Ppa-daf-25 is required for temperature sensitivity of mouth-form plasticity in P. pacificus.
Second, the daf-11/daf-25 module has been co-opted during nematode evolution for regulating the influence of temperature on mouth-form plasticity in P. pacificus.

Ppa-daf-11 is expressed in multiple amphid neurons
In C. elegans, daf-11 mutants are Daf-c, forming dauer larvae in the absence of dauer-inducing conditions (Riddle et al., 1981;Thomas et al., 1993). Additionally, Cel-daf-11 mutants exhibit defects in several chemosensory responses, including the attraction to volatile odorants such as isoamyl alcohol (Vowels and Thomas, 1994). Consistent with most of these phenotypes, Cel-daf-11 is expressed in the five pairs of amphid neurons ASI, ASJ, ASK, AWB and AWC (Birnby et al., 2000). To study if the co-option of Ppa-daf-11 for temperature perception during mouth-form plasticity regulation involved novel expression patterns, we generated Ppa-daf-11p::rfp reporter lines. We used a 2 kb promoter fragment and obtained two transgenic lines tuEx329 and tuEx330 with 30% and nearly 100% transmission rate, respectively. Ppa-daf-11p::rfp is expressed in five pairs of amphid neurons (Figure3). According to the recent neuroanatomical study of the chemosensory system of P. pacificus (Hong et al., 2019), these cells are likely the amphid neurons AM8(ASJ), AM1(ASH), AM3(AWA), AM5(ASE) and AM4(ASK) (Figure 3). These assignments suggest similarity and divergence of daf-11 expression in amphid neurons, a finding that is related to previous observations. For example, Ppa-odr-7 and Ppa-odr-3 also show partial conservation and divergence of gene expression (Hong et al., 2019).
Specifically, Ppa-daf-11 and Ppa-odr-3 are expressed in the AWA homolog rather than the AWC-equivalent neurons. It is notable that the absence of the wing-type C. elegans amphid neurons (AWA, AWB, AWC) in P. pacificus is shared with all other nematodes that have been studied by EM reconstruction (Hong et al., 2019).

Ppa-tax-2 but not Ppa-daf-16 is required for temperature sensing during mouth-form regulation
Next, we wanted to identify additional components of the gene regulatory network involved in temperature perception. We used a candidate gene approach and targeted genes through CRISPR-mediated gene knockouts. First, we selected the FOXO transcription factor daf-16 because it is known as a terminal regulator of many developmental and physiological processes including dauer formation in C. elegans (Kenyon, 2010). The analysis of Ppa-daf-16 for a potential role in mouth-form plasticity is important because previous studies did not find any mouth-form associated phenotype (Ogawa et al., 2011). While Ppa-daf-16 mutants are dauer formation-defective (daf-d), like in C. elegans, no change in mouth-form ratios was observed in these mutants. However, these mutants in the P. pacificus RS2333 background showed nominal temperature response ( Figure 2). Therefore, we generated two novel Ppa-daf-16 mutant alleles in the RSA635 background, tu1514 and tu1515. When we analyzed the mouth-form ratio of these mutantsraised at 27 °C, we did not observe any difference from RSA635 wild type animals (Table 1E). However, when cultured at 20 °C, both Ppa-daf-16 alleles showed only 43% and 51% Eu animals (Table 1E). Given that in C. elegans dauer development, daf-16 is negatively regulated by daf-11-dependent cGMP signaling, we made a Ppa-daf-16 (tu1515), Ppa-daf-11 (tu1440) double mutant. These double mutantshad a high Eu phenotype at 27 °C similar to Ppa-daf-11 single mutants (Table 1F), suggesting that Ppa-daf-16 is not involved in the regulation of mouth-form plasticity in P. pacificus.
Finally, we wanted to know if Ppa-daf-11 and Ppa-daf-25 function through a canonical cGMP signaling pathway, or alternatively, through a different molecular network.
Work in C. elegans indicated that different cGMP signaling pathways involve different guanylyl cyclases that converge on the two cyclic nucleotide-gated channels tax-2 and tax-4 (Aoki and Mori, 2015). Therefore, we generated twoPpa-tax-2 mutants in the RSA635 background. Indeed, Ppa-tax-2 (tu1291) and Ppa-tax-2 (tu1292) mutant animals show elevated Eu mouth-form ratios when cultured at 27 °C similar to Ppa-daf-11 and Ppa-daf-25 mutants (Table 1G). Thus, Ppa-daf-11 acts through a canonical cGMP signaling pathway in temperature perception during mouth-form plasticity regulation.

Discussion
This study presents a developmental genetic analysis to elucidate the influence of environmental temperature on mouth-form plasticity in P. pacificus. Our findings reveal that temperature has a strong effect on the mouth-form decision. Together with previous studies, these results indicate that P. pacificus decides which mouth-form to execute mainly in response to pheromones (Bose et al., 2012;Werner et al., 2018b), nutrient conditions  and temperature (this study). Thus, like dauer formation in C. elegans (Golden and Riddle, 1984) and P. pacificus (Ogawa et al., 2009), mouth-form plasticity is regulated by a set of biotic and abiotic factors. This study produced in two major conclusions.
First, our study indicates that temperature perception acts mainly through cGMP signaling and the Ppa-DAF-25/Ankmy2 -Ppa-DAF-11 module. In C. elegans, the guanylyl cyclase-encoding gene daf-11 was isolated as a 'group 1 dauer' mutant with a strong Daf-c phenotype (Thomas et al., 1993). Additionally, Cel-daf-11 mutants have defects in chemosensation, i.e. they do not respond to isoamyl alcohol (Vowels and Thomas, 1994) and were later shown to have complex defects in ASJ sensory neurons (Schackwitz et al., 1996). The localization of the guanylyl cyclase to chemosensory cilia requires the MYND/Ankmy2 domain protein DAF-25 and consequently, C. elegans mutants in daf-11 and daf-25 show similar Daf-c phenotypes (Jensen et al., 2010).
The fact that we have identified Ppa-daf-11 and Ppa-daf-25 in the same screen for temperature influence on mouth-form plasticity suggests that both proteins have similar and conserved functions in P. pacificus. Phylogenetically, mouth-form plasticity represents a more recently evolved trait than dauer formation. Specifically, mouth-form plasticity is restricted to the Diplogastridae family, whereas the potential to form dauer larvae under adverse conditions is much more widespread in nematodes. Therefore, the DAF-11/DAF-25 module has likely been co-opted to control temperature perception in mouth-form plasticity.
Co-option is a well-established mechanism by which genes become integrated in the regulation of novel developmental and physiological traits (True and Carroll, 2002). Indeed, co-option was already identified as a major principle in evolutionary developmental biology when the first systematic studies of molecular comparative developmental biology were performed in insects, nematodes and vertebrates (Raff, 1996).
Second, DAF-11 is one of more than 30 guanylyl cyclases in the genomes of C. elegans and P. pacificus, many of which are 1:1 orthologous (Hong et al., 2019). In C.
elegans a subfamily of guanylyl cyclases, gcy-8, gcy-18 and gcy-23, have a major additional role in temperature sensation primarily through the amphid finger neuron AFD (Inada et al., 2006). Indeed, besides the influence of temperature on certain developmental and chemosensory processes as indicated above, C. elegans is also able to acclimate to its cultivation temperature. The resulting thermotaxis behavior becomes obvious on thermal gradients when worms prefer the temperature to which they have been previously adapted (for review see Aoki and Mori, 2015). Interestingly, the three AFD-specific guanylyl cyclases GCY-8, GCY-18 and GCY-23 are the primary thermosensors as indicated by ectopic expression of these genes conferring thermosensory properties to diverse cell types (Takeishi et al., 2016). The cyclic nucleotide-gated channels TAX-2 and TAX-4 act downstream of guanylyl cyclases in the perception of environmental temperature in various processes (Komatsu et al., 1996). Consequently, tax-2 and tax-4 mutants in C. elegans show abnormalities in multiple sensory behaviors with thermotaxis phenotypes similar to the gcy-8, gcy-18, gcy-23 triple mutants, as well as dauer formation and chemotaxis phenotypes similar to daf-11 mutants. Our candidate gene approach indicates that Ppa-TAX-2 is also involved in the transmission of cGMP signaling during P. pacificus mouth-form control.
These findings suggest that a major temperature perception network consisting of Ppa-DAF-11, Ppa-DAF-25, Ppa-TAX-2 and presumably other factors acts as one first step in the environmental regulation of mouth-form plasticity in P. pacificus. We speculate that other factors control the influence of pheromones and nutrients on mouth-form plasticity. We hypothesize that all relevant environmental cues are integrated at the level of the switch network, consisting of the sulfatase EUD-1 and the N-acetyl-glucosaminidases NAG-1 and NAG-2 . However, the exact molecular mechanism of this regulatory interaction awaits future analysis. Our finding that mutations in the FORKHEADtype transcription factor Ppa-daf-16 do not share the mouth-form phenotypes of Ppa-daf-11 and Ppa-tax-2 at culture conditions at 27 °C may indicate the involvement of additional factors, which might even act independently of transcriptional regulation.
Finally, we note that we have not observed any spontaneous dauer formation in Ppadaf-25 and Ppa-daf-11 mutants on culturing plates, neither at 20 °C nor at 27 °C. These observations are superficially similar to previous findings that the regulation of dauer larvae formation has undergone substantial changes in the lineages leading to P. pacificus and C.
elegans. For example, Ppa-daf-19, the RTX master transcriptional regulator of ciliogenesis has no dauer phenotype in the RS2333 background (Moreno, Lenuzzi et al., 2018). To clarify this point further, more detailed analyses of dauer formation have to be performed using the various available dauer-induction protocols in P. pacificus (Markov et al., 2016;Werner et al., 2018b) -a project which is clearly outside of the scope of the current manuscript.
In conclusion, our study establishes the influence of environmental temperature on mouth-form plasticity in P. pacificus and reveals the co-option of cGMP signaling as one mechanism of environmental sensing. This work adds important information to the complexity of the gene regulatory network involved in the regulation of this novel and complex trait. The identification of a temperature perception network adds an important step in the elucidation of the molecular mechanisms associated with this novel trait. A proper understanding of the proximate developmental mechanisms is crucial for a full acceptance of developmental plasticity as an important mechanism for evolution.

Acknowledgment
The study was funded by the Max Planck Society. The authors declare no conflict of interest.
We thank members of the Sommer lab for discussion and assistance throughout these studies.

Author contributions
Conceptualization
pacificus were used for studying natural variation of mouth-form. The population genetic analysis of these strains has been described by McGaughran and co-workers (2016). All strains were maintained using standard methods (Pires daSilva, 2018).

Mouth-form phenotyping
Mouth-form phenotyping was scored as previously described (Serobyan et al., 2013

Mutagenesis screen for temperature insensitive mutants
We screened for temperature insensitive mutants using Ethyl methanesulfonate (EMS) in P.
pacificus RSA635. We generated around 1,200 gametes and screened 2,400 homozygous F2 lines. F2 animals were transferred from 20°C to 27°C in the Memmert incubator on NGM plates with 200-300µl OP50. The mouth form of the F3 progeny was scored after 5-7 days.

Bulk segregant analysis and whole-genome sequencing analysis
For bulk segregant analysis, tu716 and tu722 were backcrossed with wildtype animals. From each F1 backcrossed animal, four F2 worms were transferred to 27°C and the mouth-form was phenotyped after 6 days. After one generation of recovery on 20°C, individual lines were transferred again to 27°C and mouth form was scored in their progeny after 6 days.
Lines with a highly Eu, or a highly St phenotype in at least three consecutive screens were used for sequencing. For that purpose, 1-3 worms were collected in lysis buffer. Whole- searching for sites with at least 5X coverage where the major allele frequency was >90% but differed between the mutant and the control sample. Functional classification of differential variants was performed as described previously (Sieriebriennikov et al. 2020).
To hybridize crRNA and tracrRNA, 10 µl of the 100 µM stock of each molecule were combined, followed by a denaturation step at 95°C for 5 minutes and annealing at room temperature for another 5 minutes. Five µl of the hybridization product was combined with 1 µl of 20 µM Cas9 protein (New England Biolabs) at room temperaturefor 5 minutes. The mixture was diluted with Tris-EDTA buffer to a total volume of 25 µl and injected into the gonad of young adults. Molecular lesions were detected in F1 progeny by Sanger sequencing.

Genetic transformation
To generate a Ppa-daf-11reporter construct, we amplified the upstream sequence of the insilico identified Ppa-daf-11 start codon to the 3'UTR of the next gene (794bp). We amplified TurboRFP and the 3'UTR of the ribosomal gene Ppa-rpl-23 from pUC19-based plasmid stocks. All primers were obtained from Thermo Fisher Scientific. Cloning was done using Gibson Assembly® Cloning Kit (New England Biolabs). The final product was amplified as a linear construct and confirmed by Sanger sequencing. The Primers used are listed in Supplementary Table 1. For all amplification steps we used PrimeStar GxL polymerase (Takara). Prior to injection, the linear construct and genomic carrier DNA of RSA635 were incubated for two hours with PstI (New England BioLabs). The construct (60 ng/µl) was injected with the co-injection marker egl-20promoter::TurboRFP (10 ng/µl) and genomic DNA (1800 ng/µl) to one or both gonadal arms in early adult hermaphrodites. Images were acquired using a Leica TCS SP8 confocal system and were analyzed using ImageJ.

Statistical analysis
All replicates of mouth-form phenotypes in mutant lines and wildtype RSA635 were compared using Fisher's exact test. p-values were corrected using the FDR method. All mutant lines and wildtype lines were regularly observed over the course of several months to confirm the stability of the phenotype. See Supplementary materials for complete statistical analysis (Table S1-3). In figure 2, Eu percentages for each line and temperature were plotted using geom_smooth function of ggplot2 R package.