Description

The C. elegans dauer is an alternative L3 stage larva that forms under harsh environmental conditions, including low food, high temperature, and high concentration of constitutively secreted dauer pheromone. Genetic screens identified genes conferring dauer-constitutive and dauer-defective phenotypes (Daf-c and Daf-d, respectively; Hu, 2007). Double mutant analysis using principles of epistasis and parallelism ordered genes controlling the dauer process into a network (Gottlieb and Ruvkun, 1994; Thomas et al., 1993). Molecular genetic cloning of genes provided identities with similarity to orthologs in Drosophila and mammals. Taken together, these approaches arrived at a model of four main signaling axes controlling entry into dauer: upstream and parallel TGF-beta (DAF-7, mutated to Daf-c) and receptor guanylyl cyclase (DAF-11, mutated to Daf-c) signals reflect parallel processing by sensory neurons, revealed by laser ablation experiments (Birnby et al., 2000; Ren et al., 1996; Schackwitz et al., 1996). Downstream, serial Insulin/IGF-like growth factor receptor (DAF-2, mutated to Daf-c; (Kimura et al., 1997) and nuclear hormone receptor (DAF-12/NHR, mutated to Daf-d; (Antebi et al., 2000) signals control and execute tissue-specific changes in the animal (Fig. 1C, DAF-12 not shown). Mutants for each signaling axis also control diverse developmental and metabolic outputs in addition to the dauer decision. The four-axis model of signaling control of dauer formation neglects potential positive- and negative-feedback loops and is thus likely reductive. Still, these approaches have provided a robust framework for further investigation into the control of the dauer developmental decision by sensory and endocrine signaling modalities.

Genetic screens for mutations that suppress the Daf-c phenotypes conferred by mutations in the DAF-7/TGF-beta signaling cascade identified expected proteins that confer Daf-d phenotypes when mutated (Inoue and Thomas, 2000). Mutants for transcription factors downstream in the TGF-beta signal, DAF-3/co-Smad and DAF-5/Sno-ski (da Graca et al., 2004; Patterson et al., 1997; Tewari et al., 2004), completely suppress Daf-c mutations in the TGF-beta signal. The DAF-16/FoxO transcription factor, downstream of DAF-2/InsR, confers a partial dauer phenotype in double mutant combinations with Daf-c components of TGF-beta signaling (Lin et al., 1997; Ogg et al., 1997). And mutations in DAF-12/NHR, thought to be the most downstream player in the dauer regulatory network, completely suppress mutations in the TGF-beta group that confer a Daf-c phenotype.

Yet this screen also identified mutations – suppressors of constitutive dauer (Scd) – that partially but not completely suppressed the Daf-c phenotype of mutant TGF-beta group genes. These mutations defined three novel genes: scd-1, scd-2 and scd-3 (Inoue and Thomas, 2000). One of these, scd-2, encodes a protein orthologous to Anaplastic Lymphoma Kinase (ALK), a receptor tyrosine kinase that in humans is a proto-oncogene (Reiner et al., 2008). SCD-2/ALK and its putative growth factor ligand HEN-1/Jeb also regulate diverse sensory signals (Ishihara et al., 2002; Kitazono et al., 2017; Shinkai et al., 2011; Wolfe et al., 2019). Mutations in scd-2 at 25˚C strongly but not completely suppressed the DAF-c phenotype of mutant DAF-8/R-Smad, moderately suppressed the Daf-c phenotype of mutant DAF-7/TGF-beta, weakly but consistently suppressed the Daf-c phenotype of mutant DAF-11/rGC, and failed to suppress the Daf-c phenotype of mutant DAF-2/InsR. Through screening to test whether mutations in candidate genes conferred Scd interactions similar to those of mutant SCD-2/ALK, its putative ligand HEN-1/Jeb, adaptor protein SOC-1/DOS/GAB, and ERK5/MAP Kinase SMA-5 were also identified as conferring similar Scd phenotypes. This model was further supported by showing that expression from a transgenic promoter::GFP fusion repressed by DAF-3/Co-Smad (Thatcher et al., 1999) was regulated by mutations in the putative SCD-2/ALK signaling cascade. Taken together, these results suggested that SCD-2/ALK and functionally related genes that confer a similar Scd phenotype when mutated collaborate with the main DAF-7/TGF-beta cascade to co-regulate dauer-regulating genes throughout the animal (Reiner et al., 2008).

A fourth Scd gene, scd-4, was defined by a single allele, sa321, which mapped to chromosome V and thus may have been allelic with scd-2 or soc-1. We previously mapped the phenotype of suppression of daf-7 phenotype of sa321 to the unc-62–dpy-11 interval on chromosome V, excluding the possibility that sa321 was an allele of scd-2, which is located to the right of dpy-11. sa321 also complemented the suppression phenotypes of soc-1(n1789) and scd-2(y386), suggesting that sa321 defined a novel Scd gene, scd-4 (Reiner et al., 2008). sa321 does not alter other phenotypes on plates conferred by mutations in TGF-beta like daf-7(e1372): egg-laying defective (Egl), clumping on the plate (Cpy), and dark intestine (Din; not shown; Thomas et al., 1993).

We sequenced the whole genome of the strain JT7478 daf-8(sa234); scd-4(sa321). In the genetic interval of unc-62–dpy-11, we identified non-synonymous mutations in three genes: a G245E mutation in mlk-1, an A189G mutation in ncx-2, and an I5498V mutation in ttn-1. Of these, MLK-1, a MAP3 Kinase most similar to human MLK3 (mixed lineage kinase) and Drosophila slipper (slpr) functions in signal transduction and was a parsimonious candidate for a Scd gene based on our knowledge of SCD-2 as a receptor tyrosine kinase. With Sanger sequencing we confirmed the DNA lesion in mlk-1 of G734A, resulting in the G245E amino acid change in the Ser/Thr kinase domain of MLK-1 (Fig. 1A, B).

To validate that scd-4 is actually mlk-1, we used a deletion allele in mlk-1, km19 (Mizuno et al., 2004). As with mutations in hen-1 and scd-2, sa321 and km19 mutant animals are superficially wild-type when grown on plates. Like sa321 and mutations in scd-2, at 25˚C km19 causes strong but not complete suppression of mutant daf-8, moderate suppression of mutant daf-7, very weak suppression of mutant daf-11, and no suppression of mutant daf-2 (see above; Fig. 1C). Also like scd-2(y386), mlk-1(km19) restored GFP expression from the cuIs2 reporter repressed by daf-7(e1372) (Fig. 1E, quantified in 1F). Furthermore, as observed with certain mutated components of the putative SCD-2/ALK cascade, sa321 and km19 do not alter the Egl, Cpy and Din phenotypes of daf-7(e1372), suggesting that the interaction between MLK-1 and DAF-7/TGF-beta signaling is specific to dauer formation. Thus, we conclude that scd-4(sa321) is an allele of mlk-1. Given their shared mutant phenotypes and identity as signaling molecules, we hypothesize that MLK-1/MAP3K functions downstream of SCD-2/ALK to regulate dauer formation in conjunction with TGF-beta (Fig. 1D), but not other TGF-beta-dependent phenotypes.