Social and sexual behaviors in C. elegans: The first fifty years

Abstract

For the first 25 years after the landmark 1974 paper that launched the field, most C. elegans biologists were content to think of their subjects as solitary creatures. C. elegans presented no shortage of fascinating biological problems, but some of the features that led Brenner to settle on this species—in particular, its free-living, self-fertilizing lifestyle—also seemed to reduce its potential for interesting social behavior. That perspective soon changed, with the last two decades bringing remarkable progress in identifying and understanding the complex interactions between worms. The growing appreciation that C. elegans behavior can only be meaningfully understood in the context of its ecology and evolution ensures that the coming years will see similarly exciting progress.

In January 1996, when I made my first worm pick, the idea that worms might interact with each other in interesting ways was just beginning to be appreciated. Of course, signaling between worms wasn’t unknown; on the contrary, it was well-established that soluble, worm-derived population density cues could trigger entry into the long-lived, stress-resistant dauer stage. Classic studies by Don Riddle demonstrated the involvement of a pheromonal cue in this process (Golden and Riddle, 1982, 1984, 1985) and studies by Riddle, Jim Thomas, Cori Bargmann, and others had explored the genetic control of dauer entry and its underlying neural basis (Albert et al., 1981; Bargmann and Horvitz, 1991; Riddle et al., 1981; Thomas et al., 1993; Vowels and Thomas, 1992). The possibility that worms might use pheromones for other purposes, though, hadn’t received much attention.

At the time, meaningful behavioral interactions between worms were thought to be limited to copulation. Katherine Liu and Paul Sternberg’s landmark 1995 paper on the sensory control of male mating marked the beginning of careful studies of the neural and genetic mechanisms underlying worm sexual behavior (Liu and Sternberg, 1995) (Figure 1). However, this behavior was thought to be driven largely by mechanosensory cues; at that point, there had been no reports that males could find mates via diffusible signals. Further, the sex-specificity of these behaviors was thought to be driven by sex-specific circuits, with little attention to the possibility that sex-specific modulation of shared circuits might also have a role. (Progress in understanding male copulatory behavior itself will not be discussed here; for recent reviews of this topic, see (Barr et al., 2018; Emmons, 2018).)

Figure 1.

C. elegans mating behavior. The male (above) is engaged in “scanning” behavior, in which it moves backwards while holding the ventral side of its body against the hermaphrodite (below), searching for the vulval opening.

Three’s Company: Social feeding in C. elegans

Over the following decade or so, several key discoveries overturned the idea that copulation was the only interesting behavioral interaction between worms. The first of these came in 1998, with Mario de Bono and Cori Bargmann’s finding that many wild isolates of C. elegans, including the highly divergent Hawaiian strain CB4856, exhibited “social feeding” behavior: animals aggregate at the edges of the bacterial lawn, where the food tends to be a bit thicker, and feed in groups that can range in size from several to several hundred (de Bono and Bargmann, 1998) (Figure 2). This behavior, also called “bordering” or “clumping,” had been noted before, both in wild strains and in some dauer-constitutive mutants (Hodgkin and Doniach, 1997; Thomas et al., 1993). de Bono and Bargmann’s contribution, a seminal one, was to identify a molecular basis for this behavior: they demonstrated that the non-social, solitary behavior seen in the laboratory strain N2 resulted from a gain-of-function variant in the neuropeptide receptor npr-1.

Figure 2.

C. elegans social feeding behavior. (A) Solitary N2 hermaphrodites disperse on a bacterial lawn. (B) Social npr-1 hermaphrodites form aggregates, particularly at the border of the bacterial lawn. Reproduced with permission from (de Bono and Bargmann, 1998).

The discovery of the regulation of social behavior by npr-1 launched a long series of papers from the Bargmann and de Bono groups that provided a paradigm for understanding the relationships between genes, circuits, evolution, and behavior in C. elegans. An important early clue to the neural mechanisms underlying aggregation was that it depended on sensory information from both amphid chemosensory neurons and internal gas-sensing neurons (Coates and de Bono, 2002; de Bono et al., 2002). Subsequent work showed that the N2 allele of npr-1 blunts O₂ avoidance; in wild strains carrying the ancestral npr-1 allele, strong O₂ aversion causes them to clump when cultured on plates, as animals seek relief from high ambient O₂ in the low-oxygen environment of the aggregate (Cheung et al., 2004; Gray et al., 2004).

npr-1 influences O₂ avoidance, as well as other aggregation-related behaviors, by modulating a hub-and-spoke circuit in which the interneuron RMG regulates and integrates signals from the many sensory neurons to which it is connected, including the O₂-sensing URX neurons (Busch et al., 2012; Fenk and de Bono, 2017; Jang et al., 2012; Jang et al., 2017; Laurent et al., 2015; Macosko et al., 2009). However, variation in npr-1 is not the only source of the difference in aggregation between N2 and wild strains: N2 also harbors a loss-of-function allele of the neuroglobin glb-5, which acts in O₂-sensing neurons to tune their sensory responses (Bendesky et al., 2012; McGrath et al., 2009; Oda et al., 2017; Persson et al., 2009). Interestingly, the N2 alleles of npr-1 and glb-5 were found to be laboratory-derived variants; it was thought that they became fixed as a result of artificial selection against aggregation during routine culture (McGrath et al., 2009). However, recent work has shown that it is more likely that npr-1(gf) and glb-5(lf) increase fitness in the laboratory by modulating food intake and reproductive timing (Zhao et al., 2018). Even though the difference in social behavior between N2 and wild strains does not represent true natural variation, there is no doubt that this series of studies has provided numerous foundational insights into C. elegans neurobiology and neurogenetics.

A modular chemical language

Another series of important advances during this time concerned the nature of C. elegans chemical communication. We now know that this species produces a remarkably complex set of signaling molecules, but in the early 2000s, it was apparent only that worms produced one or more soluble compounds that could trigger young larvae to commit to the dauer developmental program. One of the first indications that there was more to the story was a 2002 report from Paul Sternberg’s lab demonstrating that adult hermaphrodites produced a diffusible cue that elicited a male-specific behavioral response (Simon and Sternberg, 2002). This was somewhat controversial; others had been unable to detect such a cue in C. elegans and had proposed that ancestral pheromones produced by Caenorhabditis females had been lost in C. elegans with the transition to self-fertility (Chasnov and Chow, 2002; Chasnov et al., 2007). Though this impression persisted in the literature (Frezal and Felix, 2015), it is now clear that C. elegans hermaphrodites produce at least two distinct chemical classes of sex pheromone that can attract or retain males, and that males produce compounds that have behavioral and physiological effects on hermaphrodites.

The first insights into the molecular nature of C. elegans pheromones came in 2005, with the identification by Young-Ki Paik’s group of “daumone,” purified as a dauer-inducing activity from 300L (!) of C. elegans conditioned media (Jeong et al., 2005). Daumone, now called ascr#1, is a fatty acid derivative of the dideoxy sugar ascarylose and a member of a larger class of compounds called ascarosides, first isolated from parasitic nematodes in the early twentieth century (Ludewig and Schroeder, 2013; Park et al., 2019). Subsequent activity-guided fractionation in John Clardy’s lab identified additional ascarosides—ascr#2, ascr#3, and ascr#5—whose dauer-inducing activity is significantly more potent than that of ascr#1 (Butcher et al., 2007; Butcher et al., 2008).

Remarkably, ascarosides also affect the behavior and physiology of C. elegans adults. This was first shown by Jagan Srinivasan, working in Paul Sternberg’s lab, who purified hermaphrodite-derived signals capable of attracting males. Strikingly, in a collaboration with Frank Schroeder, it was found that the active components were ascarosides: a mixture of ascr#2, ascr#3, and ascr#4 potently retained males but had little effect on hermaphrodites (Srinivasan et al., 2008). This important 2008 paper also showed that the CEM and ASK neurons, the former male-specific but the latter sex-shared, are important for this effect.

Since then, a great deal of progress has been made by many groups in exploring the vast chemical space of ascarosides produced by C. elegans and characterizing their biological activities. Multiple recent reviews, including one in this issue, have covered this exciting work (Butcher, 2017; McGrath and Ruvinsky, 2019; Muirhead and Srinivasan, 2020; Park et al., 2019; Schroeder, 2015; von Reuss, 2018). Among the most important advances have been the demonstration that ascaroside production varies according to many aspects of worm physiology, including sex, developmental stage, and nutritional status, as well as the identification of mechanisms underlying this variation (Artyukhin et al., 2013; Faghih et al., 2020; Izrayelit et al., 2012; Joo et al., 2010; Joo et al., 2016; Kaplan et al., 2011; Panda et al., 2017; Zhang et al., 2015; Zhang et al., 2016; Zhang et al., 2018; Zhou et al., 2018; Zhou et al., 2019). Ascarosides can have multiple behavioral effects on worms, inducing aggregation, attraction/retention, aversion, and promotion of foraging, depending on their chemical structures, concentrations, and interactions; moreover, the response of the receiver depends on its sex, stage, previous experience, genetic background, and other factors (Aprison and Ruvinsky, 2019; Borne et al., 2017; Dong et al., 2016; Fagan et al., 2018; Greene et al., 2016; Hong et al., 2017; Izrayelit et al., 2012; Jang et al., 2012; Lee et al., 2019; Macosko et al., 2009; Pungaliya et al., 2009; Ryu et al., 2018; Scott et al., 2017; Sims et al., 2016; Srinivasan et al., 2008; Srinivasan et al., 2012; von Reuss et al., 2012; Zhang et al., 2017). (Plasticity in the aversive responses to ascarosides is the subject of another review in this issue (Cheon et al., 2020).) Furthermore, exposure to ascarosides modulates other worm sensory behaviors (Wu et al., 2019; Yamada et al., 2010; Yoshimizu et al., 2018) and can also cause physiological changes, including altered germline proliferation and germ cell function, increased stress resistance and lifespan, and changes in lipid metabolism (Aprison and Ruvinsky, 2015, 2016; Hussey et al., 2017; Ludewig et al., 2013; McKnight et al., 2014).

Ascarosides are not the only pheromones that mediate social behaviors. Using conditioned media, several groups have described hermaphrodite-derived factors that attract males but are likely not ascarosides. One of these papers, from Jamie White and Erik Jorgensen in 2007, showed that male-specific features of the shared chemosensory neurons AWA and AWC were important for the attraction of males to non-ascaroside pheromones produced by hermaphrodites; together with a companion paper from my lab, this was the first demonstration that genetic sex could modulate the function of sex-shared neurons (Lee and Portman, 2007; White et al., 2007). Others have shown that hermaphrodites produce volatile, non-ascaroside pheromones as a function of their germline status (Leighton et al., 2014) and that the chemoreceptor srd-1 acts in the AWA neurons to generate male-specific attraction to non-ascaroside pheromones (Wan et al., 2019).

How else might chemical signals influence worm social behaviors? Fascinatingly, some species of nematophagous fungi can lure unsuspecting C. elegans victims by producing compounds that mimic sex pheromones (Hsueh et al., 2017; Yang et al., 2020). C. elegans might also use pheromones to avoid predators, as it can also respond to alarm pheromones produced by injured nematodes (Zhou et al., 2017) and to sulfolipids produced by Pristionchus (Liu et al., 2018). Indeed, many nematodes produce ascarosides, raising the possibility that C. elegans has evolved complex responses to heterospecific competitors or predators (Choe et al., 2012). Another review in this issue explores the interesting possibility that interactions between Pristionchus, C. elegans, and their bacterial food sources might represent aggressive behavior, a social interaction that has received little attention in nematodes (Quach and Chalasani, 2020).

For years, the relevance of inter-worm behaviors observed in the lab was uncertain, as precious little was known about C. elegans ecology. Thankfully, this is changing: our understanding of C. elegans ecology and evolution is advancing rapidly, and there is a growing appreciation of the complexity of its lifestyle in the wild (Frezal and Felix, 2015; Viney and Harvey, 2017). Viewing social behaviors through this lens is essential for achieving a nuanced understanding of their proximate and ultimate causes. Over the coming decades, progress toward this goal is sure to benefit from the interdisciplinary, integrative style of investigation that has been a hallmark of the C. elegans community since its inception.