Ascaroside signaling in C. elegans^*

For more than 20 years after the initial description of the dauer pheromone as a small molecule signal by Golden and Riddle (Golden and Riddle, 1982), its chemical identity remained enigmatic. Finally, in 2005 Jeong et al. described the isolation and identification of “daumone” (also “ascr#1”, or “C7”: see Section 2.7. for ascaroside nomenclature), an ascaroside featuring a seven-carbon carboxylic acid side chain, as one component of the dauer pheromone (Jeong et al., 2005). Isolation of ascr#1 was based on activity-guided fractionation of C. elegans liquid culture extracts, using dauer formation for tracking activity. However, because of its relatively low specific activity, ascr#1 could not account for most of the dauer-inducing activity of C. elegans liquid culture extracts (“natural dauer pheromone”) (Butcher et al., 2007; Gallo and Riddle, 2009; Jeong et al., 2005). For example, Gallo and Riddle compared the physiological properties of ascr#1 with those of natural dauer pheromone and found that at the concentrations required to induce dauer ascr#1 has significant toxicity (Gallo and Riddle, 2009). These observations suggested that additional dauer pheromone components must exist, which was verified by a series of subsequent studies that showed that the dauer pheromone consists of a highly complex mixture of structurally varied ascarosides. Using activity-guided fractionation approaches, Clardy and co-workers identified two additional ascarosides with stronger dauer-inducing activity, featuring a six-carbon side chain bearing a ketone functionality (“C6”, ascr#2) and a nine-carbon α,β-unsaturated carboxylic acid side chain (“C9”, ascr#3) (Butcher et al., 2007) (Figure 2). A fourth ascaroside, bearing a three-carbon side chain (“C3” or ascr#5) was found to synergize with ascr#2 and ascr#3 in dauer formation (Butcher et al., 2008) (Figure 2). The synergistic activity of dauer pheromone components suggested that ascaroside perception may involve several different receptors with different affinities for different ascarosides. Furthermore, it was shown that dauer pheromone composition depends on cultivation conditions: dauer pheromone extract derived from liquid cultures grown at 20°C contained much less ascr#1 than extracts from cultures grown at 25°C, indicating that environmental conditions influence the molecular composition of the dauer pheromone cocktail (Butcher et al., 2007; Jeong et al., 2005; Lee et al., 2010).

The finding that the dauer pheromone consists of a synergistic mixture explained the difficulties experienced with identification of the chemical nature of the pheromone via activity-guided fractionation: separation of different components that synergize must lead to partial loss of activity, greatly complicating the task of attributing activity to individual compounds. Therefore, Pungaliya et al. explored applicability of NMR spectroscopy-based comparative metabolomics to complete identification of the dauer pheromone's constituents (Pungaliya et al., 2009). These studies revealed several additional ascarosides, including ascr#8, a potent component of the dauer pheromone, whose chemical structure differs significantly from those of the previously identified ascarosides. Ascr#8 includes a p-aminobenzoic acid moiety, forming an amide with a seven-carbon α,β-unsaturated fatty acid similar to the side chains found in ascr#1 or ascr#3 (Figure 2). p-Aminobenzoic acid is uncommon among known primary and secondary metabolites; except for it forming an integral part of folate. Additional ascarosides identified via NMR-based comparative metabolomics include ascr#7, representing an α,β-unsaturated analog of ascr#1 (daumone), and the two diastereomeric alcohols ascr#6.1 and ascr#6.2; however, none of these compounds was shown to confer significant additional dauer-inducing activity. In a parallel effort, Butcher et al. identified another intriguing structural variant, the indole ascaroside icas#9 (also “C5” or “indole ascaroside C5”) featuring an unusual indole carboxy moiety attached to the ascarylose sugar (Figure 2). The activity profile of icas#9 differs from that of other ascarosides in that its activity declines at higher concentrations, possibly suggesting a different mode of receptor activation (Butcher et al., 2009).

Two recent studies employing NMR spectroscopy and mass spectrometry based comparative metabolomics suggest that the number of dauer pheromone components may be even larger. Srinivasan et al. showed that C. elegans produce an entire series of indole ascarosides, among which icas#3 and the earlier described icas#9 are the most abundant (Srinivasan et al., 2012). These indole ascarosides were shown to function as potent aggregation pheromones (see Section 2.4), and it seems likely that they also modulate dauer formation. Subsequently, von Reuss et al. used mass spectrometry-based metabolomics to investigate ascaroside biosynthesis (see Section 4.1) which led to identification of a large number of additional ascarosides, including several structural derivatives of the strongly dauer inducing ascr#3 (von Reuss et al., 2012). Examples include the saturated derivatives ascr#10 and oscr#10, the latter of which features a side chain that bears the ascarylose at the ultimate carbon (ω-functionalization) instead of the usual penultimate carbon (ω-1-functionalization) as in ascr#10 (von Reuss et al., 2012) (Figure 3). This study also revealed a larger number of indole ascarosides as well as ascarosides featuring a p-hydroxybenzoyl- and (E)-2-methyl-2-butenoyl side chains (Figures 1 and 3). Some of these compounds have been shown to strongly affect worm behavior (von Reuss et al., 2012). Whether these additional ascarosides contribute significantly to dauer formation, inhibit dauer recovery, or synergize with other components of the dauer pheromone has not been investigated.

Using natural dauer pheromone, it has been demonstrated that the dauer pheromone signal is detected and integrated by specific chemosensory head neurons, including ADF, ASG, ASI, ASJ, and, to a lesser extent, ASK neurons (Bargmann, 2006). When exposed to pheromone, these neurons block or in a few cases initiate the production of secondary messengers including TGF-ß, cGMP, serotonin and insulin-like peptides (Figure 4). These neuroendocrine pathways transduce the dauer signal to elicit systemic responses in various tissues (Riddle and Albert, 1997). Most of these experiments have not yet been repeated with synthetic ascarosides or ascaroside mixtures in place of natural pheromone. In one study, Kim et al. showed that synthetic ascarosides repress daf-7 TGF-ß expression similar to natural dauer pheromone, with ascr#2, ascr#3, and ascr#5 being more potent than ascr#1, confirming that different ascarosides have different potencies (Kim et al., 2009). However, it remains largely unknown whether different ascarosides are associated with different pathways (and neurons), or instead all ascarosides signal via the same pathways.

The dauer pheromone signal finally converges on the cytochrome P450 DAF-9 and on the nuclear hormone receptor DAF-12 (Antebi et al., 2000; Gerisch et al., 2001; Lee and Schroeder, 2012). Under dauer inducing conditions, DAF-9 expression is strongly reduced or abolished and the steroidal ligands of DAF-12, the dafachronic acids, are not produced. As a result, the unliganded DAF-12 recruits its co-factor DIN-1 and triggers expression of dauer specific gene (Figure 4) (Bethke et al., 2009; Ludewig et al., 2004; Motola et al., 2006). Accordingly, exposure to natural dauer pheromone downregulates DAF-9 expression and thus dafachronic acid biosynthesis; however, this has not yet been demonstrated for synthetic ascaroside blends (Schaedel et al. 2012). Current knowledge of ascaroside perception and their receptors is summarized in Section 3.

C. elegans males display a broad spectrum of sex-specific behaviors that relate to mating (Barr and Garcia, 2006; Chasnov and Chow, 2002; Chasnov et al., 2007). In 2002, Simon and Sternberg showed that hermaphrodites produce a chemical cue, likely a small molecule signal, that attracts males (Simon and Sternberg, 2002). The existence of a male-attracting pheromone was further confirmed by studies of White et al., who showed that males are attracted to metabolite extracts from hermaphrodite liquid cultures (White et al., 2007). Laser-ablation studies showed that perception of the attractive signal required the male-specific cephalic companion (CEM) neurons, along with the AWA and AWC olfactory neurons. White et al. further showed that male response to hermaphrodite pheromone is mediated by a TRPV (transient receptor potential vanilloid) channel dependent on osm-9, ocr-1, and ocr-2. (White et al., 2007). Chasnov et al. compared male attraction of several different Caenorhabditis species and showed that C. elegans males are more strongly attracted to females of other Caenorhabditis species than to C. elegans hermaphrodites, suggesting that C. elegans hermaphrodites may have lost the ability to produce some male-attracting compounds, or produce these in smaller amounts than females from other Caenorhabditis species (Chasnov et al., 2007).

For chemical characterization of the male attracting signal, Srinivasan et al. pursued activity-guided fractionation of C. elegans metabolite extracts (Srinivasan et al., 2008), but in contrast to identification of the dauer pheromone, special measures were taken to avoid contamination with bacterial small molecules. This greatly simplified the chemotaxis assays as worms are attracted to many bacterial small molecules. Using synchronized batches of worms, the time window during which the male-attracting substance(s) were produced was determined. It was found that L4, young adult, and adult hermaphrodites produce the male-attracting factor, with the greatest amounts in L4 larvae and young adults (Srinivasan et al., 2008). Chemical analyses revealed that male-attracting activity is due to two major dauer pheromone constituents, ascr#2 and ascr#3. These two ascarosides strongly attract males at concentrations far below those required for dauer induction. The activity of ascr#2 and ascr#3 is further enhanced by another ascaroside, ascr#4, which chemically represents a glucosylated ascr#2 (Figure 5). Mixtures of ascr#2, ascr#3, and ascr#4 are much more potently attractive than the individual components, confirming that these compounds act in synergy. In contrast to their activity profiles in dauer formation, male-attracting activity of ascr#2 and ascr#3 decreases at higher concentrations, resulting in a bell-shaped activity profile. The bimodal activity profiles obtained for ascr#2 and ascr#3 in the male attraction assay suggest receptor desensitization at high compound concentrations, whereas the saturated dose-response curves obtained for ascr#2 and ascr#3 in the dauer formation assays suggests that that receptor desensitization may not be involved in dauer formation.

C. elegans males, but not hermaphrodites are attracted to mixtures of ascr#2 and ascr#3 at picomolar to low nanomolar concentrations. At higher, dauer inducing concentrations, males are no longer attracted whereas wild-type hermaphrodites show strong repulsion, indicating yet another function of ascarosides in social signaling (Srinivasan et al., 2008). The role of ascr#2, ascr#3, ascr#5 and other ascarosides for hermaphrodite avoidance and aggregation behavior was further investigated as described in the following section (Macosko et al., 2009; Srinivasan et al., 2012).

As activity-guided fractionation had revealed that the male-attracting signal consists of a synergistic blend of ascarosides, it seemed possible that additional components may have been missed. Testing a series of additional ascarosides identified via NMR-based comparative metabolomics revealed that the dauer-inducing ascr#8 synergizes with ascr#2 and ascr#3, strongly enhancing male attraction (Pungaliya et al., 2009). In contrast, several other ascarosides were found to be completely inactive in the male-attraction assay. For example, ascr#1 and ascr#7 are both inactive, although the chemically closely related ascr#3, alongside ascr#8, constitutes one of the two strongest male attractants (Pungaliya et al., 2009). These structure-activity relationships further suggested that ascaroside perception involves multiple, structure-specific receptors.

Male response to ascr#3 was shown to require both the sex-specific CEM neurons and the amphid single-ciliated sensory neuron type K (ASK) (Srinivasan et al., 2008). Furthermore, ascr#3 dependent male attraction was abolished in the general sensory mutants osm-6 and osm-3, which are expressed in ciliated neurons and are also required for dauer pheromone perception (Vowels and Thomas, 1992). In summary, these studies demonstrate that ascaroside signaling regulates both dauer formation and mating behavior in a concentration dependent manner, thus connecting reproductive and developmental pathways.

C. elegans exhibits natural variation in its foraging behavior. Some strains (e.g., the laboratory strain N2) disperse individually on a bacterial lawn and show a generally low tendency to aggregate, whereas most wild-type strains (for example, the Hawaii strain CB4856) aggregate where bacteria are most abundant (de Bono and Bargmann, 1998; Hodgkin and Doniach, 1997). These behavioral variants are generally referred to as ‘‘solitary’’ and ‘‘social,’’ respectively (de Bono and Bargmann, 1998; Rogers et al., 2003). These differences in aggregation behavior are associated with two different alleles of the neuropeptide Y receptor homolog NPR-1 that differ only at a single amino acid position: solitary strains such as N2 express a high-activity variant of NPR-1 (215-valine), whereas aggregating strains such as CB4856 express a low-activity variant of NPR-1 (215-phenylalanine). The strong loss-of-function mutants npr-1(ad609) and npr-1(ky13), which were generated in the N2 background, also show a high tendency to aggregate (de Bono and Bargmann, 1998; Hodgkin and Doniach, 1997). The interneuron RMG, the central site of action of NPR-1, is proposed to serve as a central hub computing aggregation and attraction signals originating from several different sensory neurons (Macosko et al., 2009), including the ASK neurons, which had been shown to be required for male attraction to ascarosides (Srinivasan et al., 2008).

Macosko et al. showed that loss of function of npr-1 strongly affects hermaphrodite response to ascarosides (Macosko et al., 2009). Whereas NPR-1 high activity strains such as N2 are strongly repulsed by ascr#2, ascr#3, and ascr#5 (Figure 6), mixtures of these compounds become attractive in npr-1 loss-of-function mutants. These npr-1 dependent effects of ascarosides are mediated by signaling from the ASK sensory neurons to the RMG interneuron where NPR-1 is expressed (Macosko et al., 2009). Thus ASK sensory neurons mediate attraction of males to ascr#2 and ascr#3 as well as attraction of npr-1 loss-of-function hermaphrodites to ascr#2, ascr#3, and ascr#5. However, it remained unclear how social wild-type hermaphrodites, carrying a low-activity variant of NPR-1 (215-phenylalanine), would respond to these ascarosides. A subsequent study showed that, in contrast to npr-1 loss-of-function mutants, social wild-type hermaphrodites are either not attracted or show weak repulsion to mixtures of ascr#2, ascr#3, and ascr#5 (Srinivasan et al., 2012).

The starkly different responses of males and hermaphrodites to ascr#3 (and perhaps other ascarosides) appear to result in part from antagonism of several different synaptic pathways, involving both the ADL and ASK sensory neurons (Jang et al., 2012). Additionally, hermaphrodite repulsion by ascarosides may result from sex-specific changes in the ASI neurons during development, which have been shown to result in repression of male-specific attraction behavior in adult hermaphrodites (White and Jorgensen, 2012), although the role of ascarosides in the observed attraction behaviors was not studied.

The indole ascaroside icas#9 (“indolecarboxy ascaroside C5”) was first discovered via activity-guided fractionation of dauer inducing metabolite fractions (Butcher et al., 2009) (see Section 2.1). Subsequent comparative metabolomic analyses of wild-type and daf-22 mutant metabolomes revealed an entire series of ascarosides carrying an indolecarboxy moiety attached to position 4 of the ascarylose (Srinivasan et al., 2012). Indole ascarosides are much less abundant than the earlier discovered non-indole ascarosides (von Reuss et al., 2012). Notably, the indole ascarosides’ quantitative distribution is distinctly different from that of the corresponding non-indole ascarosides, suggesting that incorporation of the indole unit is strongly regulated (Srinivasan et al., 2012). Stable isotope incorporation experiments demonstrated that the indole carboxy moiety is derived from the amino acid tryptophan (Srinivasan et al., 2012).

Behavioral assays showed that at high concentrations, the indole ascarosides icas#3 and icas#9 attract both hermaphrodites and males; whereas at lower, physiological concentrations only hermaphrodites are attracted (Srinivasan et al., 2012). Significant hermaphrodite attraction was observed at concentrations as low as 100 fM (femtomolar). In contrast to non-indole ascarosides, icas#3 is strongly attractive to hermaphrodites from all tested social and solitary strains, including npr-1 loss-of-function mutants. In addition, icas#3 increases hermaphrodite aggregation on food at concentrations as low as 10 fM (Srinivasan et al., 2012).

These results showed that specific structural variants of ascarosides are associated with specific functions, as addition of an indole carboxy group to repulsive ascr#3 changes the signaling properties such that the indole-modified version icas#3 functions as a strong attractant (Srinivasan et al., 2012) (Figure 7). Srinivasan et al. further showed that ascr#3 and icas#3 are competing signals. Physiological mixtures of ascr#3 and icas#3 are attractive at low total compound concentrations but repulsive at higher concentrations (Srinivasan et al., 2012).

Attraction of hermaphrodites by icas#3 requires the ASK sensory neurons, which are also involved in responses to the generally repulsive ascr#3 (Srinivasan et al., 2012). However, the finding that icas#3-mediated attraction and aggregation are not reduced by high NPR-1 activity indicated that icas#3 responses involve signaling pathways that are partially distinct from those integrating responses to non-indole ascarosides, e.g., ascr#3, and other types of aggregation-inducing stimuli, e.g., low oxygen levels. This hypothesis was supported by the observation that hermaphrodites lacking the RMG interneuron, which coordinates other aggregation responses via NPR-1, were still attracted to icas#3 (Srinivasan et al., 2012).

In summary, indole ascarosides function as highly potent aggregation pheromones. Whether hermaphrodite attraction to indole ascarosides is additive or synergistic, as is the case for non-indole ascarosides in promoting dauer or male attraction, has not been investigated. Additional aggregation-inducing ascarosides, of which some exceed the indole ascarosides’ in potency, have recently been identified via mass spectrometry-based comparative metabolomics (von Reuss et al., 2012). For example, the p-hydroxy benzoyl ascaroside hbas#3 (Figure 8) attracts hermaphrodites at concentrations as low as 10 fM (femtomolar) (von Reuss et al., 2012).

Ascaroside production has been studied mostly by using mixed-gender cultures, which effectively profiled ascaroside production in hermaphrodites which account for more than 99% of wild-type worms under typical lab conditions (Chasnov and Chow, 2002). Izrayelit et al. showed that male C. elegans also produce ascarosides and that their ascaroside profiles differ starkly from those of hermaphrodites (Izrayelit et al., 2012). Male ascaroside profiles are dominated by the saturated ascr#10, whereas hermaphrodites produce the corresponding α,β-unsaturated ascaroside, ascr#3, as major component. This seemingly minor difference of chemical structures (Figure 9) is associated with profound differences in biological activity. Whereas hermaphrodite-produced ascr#3 repels hermaphrodites and attracts males, male-produced ascr#10 strongly retains and attracts hermaphrodites. Hermaphrodites respond to extremely low quantities of ascr#10 and it appears that the amounts of ascr#10 excreted by a single male may be sufficient to affect hermaphrodite behavior. The observed differences between male and hermaphrodite ascaroside production appear to result from sex-specific expression of specific components of peroxisomal β-oxidation (the role of peroxisomal β-oxidation in ascaroside biosynthesis is discussed in Section 4).

In addition to directly affecting worm behavior, previous ascaroside exposure affects chemotaxis responses to other odorants. Yamada et al. noted that wild-type worms’ response to the odorant benzaldehyde depends on population density (Yamada et al., 2010). For worms raised at low population densities, benzaldehyde remained attractive even after pre-exposure to this odorant, whereas for worms grown at higher population density pre-exposure to benzaldehyde resulted in loss of attraction. Notably, this plasticity of olfactory responses was absent in daf-22 worms, which are defective in ascaroside biosynthesis, suggesting that ascaroside exposure enhances olfactory plasticity. This hypothesis was confirmed by addition of natural dauer pheromone or synthetic ascr#2, ascr#3, or ascr#5, which restored olfactory plasticity to daf-22 mutant worms. Dauer pheromone and synthetic ascarosides downregulate expression of a signaling peptide, SNET-1, that negatively regulates olfactory plasticity (Yamada et al., 2010). These results provide evidence for small-molecule conditioning of developing worms (or their progenitors) and connect population density signaling via ascarosides with endogenous peptide signaling that regulates chemotactic behavior. Similarly, earlier work by Peckol et al. had shown that exposure to natural dauer pheromone increases expression of three predicted chemosensory receptor genes srd-1, str-2, and str-3, in the ASI chemosensory neurons, suggesting that experience-dependent change in chemosensory receptor gene expression may modify olfactory behaviors (Peckol et al., 2001).

Following the identification of the first ascaroside-based signaling molecules, compounds were named simply by referring to the lengths of their side chain, e.g., the 6-carbon ascr#2 was named “C6” and the 9-carbon ascr#3 was named “C9” (Butcher et al., 2007). In another naming system, “daumone-1”, “daumone-2”, and “daumone-3” have been used to refer to the dauer-inducing ascr#1, ascr#2, and ascr#3 (Jeong et al., 2009). With the identification of now more than 100 additional ascarosides these naming systems became impractical (von Reuss et al., 2012). For example, there are at least six different simple ascarosides that feature a 9-carbon side chains: ascr#3, ascr#10, oscr#3, oscr#10, bhas#10, and bhos#10, making the name “C9” inconvenient and potentially misleading (Figure 10). Furthermore, names such as “C3", “C6", and “C9” are unsuitable for text mining. Lastly, a large number of modified ascarosides featuring indole-3-carboxy, p-hydroxybenzoyl, 2-methyl-2-butenoyl, glucosyl, p-aminobenzoyl moieties have been identified for which simple, unambiguous names must be found (von Reuss et al., 2012).

Therefore, a new naming system was developed in collaboration with WormBase (Srinivasan et al., 2012; von Reuss et al., 2012). The Small Molecule IDentifiers, SMIDs, were designed as searchable, gene-style identifiers consisting of four lower case non-italicized letters that refer to the general structural class of the compound, followed by a pound sign and a number, e.g., ascr#3, icas#7, or hbas#10. In collaboration with WormBase, SMIDs will be assigned for all small molecules newly identified from C. elegans and other nematodes. The SMID database (www.smid-db.org) serves as a publicly available reference for all assigned SMIDs. SMID-DB provides chemical structures, molecular formulas, a full list of references and other data for all SMIDs.

Peroxisomal β-oxidation is a 4-step process that shortens long-chain fatty acids (LCFA) and very long-chain fatty acids (VLCFA, chain length > C₂₄) iteratively by two carbon units, producing acetyl-CoA that enters the citric acid cycle (Figure 12). Peroxisomal fatty acid β-oxidation is believed to cease at a chain length of about 6-12 carbons, with β-oxidation of shorter chained fatty acids proceeding in the mitochondria (Reddy and Hashimoto, 2001). In ascaroside biosynthesis, the acyl-CoA oxidase ACOX-1 participates in the first step of peroxisomal side chain β-oxidation, introducing α,β-unsaturation of long-chained precursors (Joo et al., 2010; von Reuss et al., 2012). Mass spectrometry-based comparative metabolomics showed that in acox-1 mutants production of (ω-1)-functionalized ascarosides with less than 9 carbons was largely abolished (von Reuss et al., 2012). In addition, amounts of the 9-carbon α,β-unsaturated ascr#3, usually the most abundant ascaroside in the wild-type excretome, was greatly reduced. Instead large quantities of ascr#10, the saturated derivative of ascr#3 and a homologous series of longer-chained saturated ascarosides was observed (von Reuss et al., 2012).

Notably, the acox-1 mutation affected biosynthesis of (ω)-functionalized ascarosides differently than (ω-1)-functionalized ascarosides. In acox-1 mutants, production of ascr#5, the most abundant (ω)-functionalized ascaroside in wild type, is completely abolished, whereas production of longer chain (ω)-functionalized ascarosides with 5–13-carbon side chains, e.g., oscr#9, is strongly increased. Thus in acox-1 mutants chain-shortening of (ω)-functionalized ascarosides stalls at a chain length of 5 carbons whereas chain-shortening of (ω-1)-functionalized ascarosides ceases already at a chain length of 9 carbons. These results strongly suggest a model for ascaroside biosynthesis in which side-chain oxygenation, and perhaps ascarylose attachment, precedes peroxisomal β-oxidation (von Reuss et al., 2012). Because biosynthesis of short and medium-chained ascarosides is not completely abolished in acox-1 loss-of-function mutants, it is assumed that other, yet unidentified acyl-CoA oxidases participate in ascaroside biosynthesis (Joo et al., 2010; von Reuss et al., 2012). Furthermore, these results implicate ACOX-1 in peroxisomal β-oxidation of short-chain fatty acids, in contrast to the notion that β-oxidation of short chain fatty acids occur primarily in the mitochondria (Reddy and Hashimoto, 2001).

Whereas in vertebrates and Drosophila the next two steps of iterative peroxisomal chain shortening, hydration of the double bond and subsequent dehydrogenation to the β-ketoacyl-CoA ester, are catalyzed by one protein, e.g., MFE-2, these two enzymatic functions are associated with two separate proteins in C. elegans, MAOC-1 and DHS-28 (Figure 12) (Haataja et al., 2011; von Reuss et al., 2012; Zhang et al., 2010). Hydration of the double bond in the α,β-unsaturated side chains produced by ACOX-1 is catalyzed by the peroxisomal 2-enoyl-CoA hydratase MAOC-1. Correspondingly, maoc-1 loss-of-function mutants accumulate medium and long-chain α,β-unsaturated ascarosides, along with smaller amounts of saturated and β-hydroxylated derivatives. In contrast to acox-1 mutants, biosynthesis of short-chain ascarosides dominant in wild-type worms is completely abolished in maoc-1 worms (von Reuss et al., 2012).

Similarly, wild-type ascarosides are completely absent from the dhs-28 and daf-22 metabolomes (Butcher et al., 2009; von Reuss et al., 2012). Comparative metabolomic analyses of dhs-28 metabolome samples revealed build-up of β-hydroxylated medium- and long-chain ascarosides, which likely derive from corresponding β-hydroxyacyl-CoA derivatives that are the substrates of the β-hydroxyacyl-CoA dehydrogenase DHS-28 (Figures 12 and 13). In addition, dhs-28 worms produce smaller amounts of medium- and long-chain saturated and α,β-unsaturated ascarosides. daf-22 worms produce only very small amounts of medium- and long-chain ascarosides, most of which have saturated or β-hydroxylated side chains (von Reuss et al., 2012). Some of these long-chain ascarosides have been shown to have weak dauer inducing activity and may account for the small amount of dauer-inducing activity occasionally present in dhs-28 and daf-22 metabolite extracts (Butcher et al., 2009; von Reuss et al., 2012). daf-22 worms additionally contain larger quantities of very long chain ascarosides similar to the ascaroside lipids identified from Ascaris egg shells (Bartley et al., 1996; Jezyk and Fairbairn, 1967; Pungaliya et al., 2009). These very long-chain ascarosides may represent the ultimate precursors of all or some of the ascarosides described in this section.

Comparison on the ascaroside profiles of males and hermaphrodites indicates that peroxisomal β-oxidation in C. elegans is regulated in a sex-specific manner (Izrayelit et al., 2012). Male ascaroside profiles are dominated by ascr#10, whereas hermaphrodites produce large amounts of ascr#3, distinguished from ascr#10 solely by the incorporation of a double bond, instead. These differences in ascaroside profiles correlate with sex-specific differences in the expression levels of peroxisomal β-oxidation genes.

Short-chain indole ascarosides are absent from the maoc-1, dhs-28, and daf-22 mutant metabolomes, indicating that biosynthesis of the side chains of indole ascarosides and non-indole ascarosides proceeds via the same pathway (von Reuss et al., 2012). Correspondingly, acox-1, maoc-1, and dhs-28 worms produce medium- and long-chain indole ascarosides with saturated, α,β-unsaturated, and β-hydroxylated side chains similar to those of the non-indole ascarosides found in these mutants (von Reuss et al., 2012). Using worms grown in axenic cultures, the indole carboxy moiety in wild-type indole ascarosides was shown to be derived from tryptophan (Srinivasan et al., 2012), and incubation of ascr#3 with daf-22 worms showed that non-indole ascarosides serve as precursors for indole ascarosides (von Reuss et al., 2012). Similarly, p-hydroxybenzoyl ascarosides (e.g., hbas#3) and (E)-2-methyl-2-butenyl ascarosides (e.g., mbas#3) could be derived from the amino acids tyrosine and isoleucine, respectively. Comparison of the relative amounts of indole and non-indole ascarosides in wild-type worms indicates that indole attachment is tightly regulated. For example, ascr#9 is much less abundant than ascr#3, whereas the corresponding indole ascarosides icas#9 and icas#3 are of roughly equal abundance (von Reuss et al., 2012). Given that conversion of ascarosides into their indole carboxy derivatives is associated with a dramatic change in biological activity (see Section 2.4), it seems likely that specific mechanisms control indole ascaroside formation. However, enzymes catalyzing attachment of the indole carboxy moiety have not been identified.

Whereas the biosynthesis of dideoxysugar L-ascarylose in bacteria is well understood (Thorson et al., 1994), ascarylose biosynthesis in C. elegans has not been investigated in detail. Detection of ascarosides in axenic C. elegans liquid cultures demonstrated that worms produce ascarylose endogenously (Srinivasan et al., 2012), and the C. elegans genome includes several homologues of bacterial genes in this pathway (von Reuss et al., 2012), for example, ascE from Yersinia pseudotuberculosis (Thorson et al., 1994). Similarly, enzymes involved in hydroxylation and glycosylation of the VLCFA precursors of the ascaroside side-chains remain to be identified. Taken together, current knowledge allows proposing a preliminary model for ascaroside biosynthesis (Figure 14).

Production of ascaroside pheromones by C. elegans is stage dependent (Kaplan et al., 2011). Total amounts of ascarosides excreted increases throughout development from L1 to adult worms. Excretion of the two main components of the male-attracting signal, ascr#3 and ascr#8, peaks at the young adult stage (Kaplan et al., 2011). Under dauer-inducing conditions, excretion of the strongly dauer-inducing ascr#2 peaks right before entry into the dauer stage. Dauer worms do not appear to excrete ascarosides, but contain small amounts of ascr#1 and ascr#3 (Kaplan et al., 2011).

Temperature affects the ratio of ascr#1 and ascr#3 production (Joo et al., 2010; Lee et al., 2010) as well as the amount of ascr#5 excreted (Butcher et al., 2008). Comparison of the ascaroside profiles of well-fed and starved C. elegans wild-type cultures showed that starved worms excrete significantly larger amounts of the (ω–1)-linked ascr#3 than well-fed worms, relative to the amounts of ω-linked ascr#5 (von Reuss et al., 2012). Similarly, analysis of dhs-28 mutant liquid cultures revealed that excretion of long-chain (ω–1)-oxygenated ascarosides in starved dhs-28 cultures was about 5 times higher than those of well-fed dhs-28 cultures (von Reuss et al., 2012). Since the ω-linked ascr#5 affects worm behavior differently than the major (ω–1)-linked ascarosides, e.g., ascr#3 (Butcher et al., 2008; Macosko et al., 2009), it appears that ascaroside signaling is actively regulated in response to starvation via changes in the ratio of (ω–1) to (ω)-functionalization of the VLCFA precursors of the ascaroside side chains (von Reuss et al., 2012). Increasingly sensitive analytical techniques will enable more detailed studies of ascaroside production of different life stages under different environmental conditions.

Comparison of the ascaroside profiles of liquid culture supernatant extracts (worm excretome) and worm body extracts showed that ascaroside are differentially released (Kaplan et al., 2011; von Reuss et al., 2012). Saturated ascarosides are retained in the worm bodies to a much greater extent than their α,β-unsaturated derivatives. In addition, worm bodies contain large quantities of ascaroside O-glycoside esters, for example the ascr#10 glucoside glas#10, which are absent from media supernatant samples (von Reuss et al., 2012).