Working with starved L1 larvae of C. elegans and C. briggsae we noticed that these two species behave quite differently in starvation. First, C. elegans adults stop laying eggs after exhausting bacterial food, which eventually leads to internal hatching and bagging. C. briggsae do not show this behavior. This difference has been observed before (McCulloch and Gems, 2003). Second, at high enough density of worms, arrested C. elegans L1s aggregate on agar plates after several days of starvation (Fig. 1a). C. briggsae L1s do not form aggregates (Fig. 1b). Aggregation may serve several purposes ranging from decrease of surface to volume ratio and use of diffusible “public goods” to sharing information about quality of the environment. Third, survival of starved C. elegans L1s strongly depends on their density – the higher the worm density, the longer they survive (Fig. 1c) (Artyukhin et al., 2013). This holds true for starvation on plates as well as in suspension. Survival of C. briggsae L1s is independent of worm density (Fig. 1d). We believe that the aggregation and the density dependence are naturally connected as they seem to be correlated across Caenorhabditis species. Forth, starved C. elegans L1s release plenty of ascarosides (Fig. 1e,f). C. briggsae L1s release far less, both in variety (Fig. 1e) and in amounts (Fig. 1f). More than 100 ascarosides have been found in C. elegans (von Reuss et al., 2012), and although we do not know physiological functions for many of them, functions of the ones that we do know and the high degree of structural conservation in other nematodes suggest that ascarosides constitute a vital part of nematode chemical language. Hence, we interpret the data in Fig. 1e,f as suggesting that C. elegans L1s “talk” more with each other in starvation than C. briggsae. Finally, C. elegans is one of a couple Caenorhabditis species susceptible to environmental RNAi (Winston et al., 2007), which has been speculated to play a role in communication between organisms (perhaps, conspecifics) (Whangbo and Hunter, 2008). Based on everything above, we speculate that in unfavorable conditions (starvation) C. elegans tend to become more social than C. briggsae. In variable and unpredictable environments genotypic fitness can be maximized either by reducing individual-level variance in fitness or by reducing between-individual correlations in fitness (or some combination of the two) (Starrfelt and Kokko, 2012). Aggregating or social species may choose to minimize individual-level variance by having a mechanism that helps them to adjust to unfavorable conditions, in part through collective behavior. Non-aggregating species may use dispersal as a strategy to minimize between-individual correlations. Based on this hypothesis we would predict that lack of aggregation and density dependence in C. briggsae implies that their starved L1s hardly ever find themselves at high density in nature and the optimal strategy for them is to disperse and actively look for food.
During larval development C. elegans has two distinct time points where it can arrest development until environmental conditions become more favorable. The first one is L1 arrest, which occurs when worms hatch in the absence of food. The second decision point is later, around L1-L2 molt, and if conditions are poor at this time (little food, crowding, and high temperature) worms gain the potential to become dauers. L1 arrest and dauer formation are not mutually exclusive and, in principle, a worm can go through both of them sequentially. However, it is not known whether L1 arrest and dauer formation are coupled, i.e., if experiencing L1 arrest affects a worm’s decision to become a dauer. A naïve prediction would be that worms that experienced L1 starvation should be more susceptible to become dauers since they were already exposed to a harsh environment.
We studied dauer formation of N2 worms in liquid and found that in reality worms that experienced L1 arrest and starvation were less likely to become dauers. In a typical experiment, eggs obtained from an egg prep were incubated at 20 0C in S-complete with a small amount of bacterial food (HB101) and dauer formation was assessed 6 days later based on survival in 1% SDS (Pungaliya et al., 2009). When we let eggs hatch without food and added it 1-4 days later, keeping other conditions the same, significantly fewer worms became dauers (Figure 1). This was not due to slower development after starvation and incomplete dauer formation. We also found that this effect reflected an internal state of the worm rather than exposure to L1 exometabolome during L1 arrest since worms starved as L1s at low and high density later formed dauers with the same probability (Figure 1, during the dauer formation worms were at the same low density of 1 worm/µl to minimize the effect of released chemicals, see below).
Starved L1 worms release numerous metabolites, including several ascarosides, and we anticipated that they might promote dauer formation. Indeed, when S-complete was substituted with conditioned medium from high-density L1 larvae starved for 24 h, more worms became dauers (Fig. 2), consistent with the known dauer-inducing effect of crowding mediated by ascarosides (Ludewig and Schroeder, 2013). However, this result apparently contradicts the starvation effect described above. On one hand, when a worm itself experienced L1 arrest but received no signal from other worms (low density condition), it was less likely to become a dauer and preferred to try to grow to adult. On the other hand, a worm that had not experienced starvation itself but was exposed to chemical signals released by other starved L1 worms, was more likely to become a dauer. The next obvious experiment is to combine these two opposite effects and see which one is stronger. The result of this test (Fig. 2) is that the external starvation signal (mediated by L1 conditioned medium) overrides the internal one (L1 arrest). When making a decision whether to become a dauer, apparently C. elegans worms trust the communal voice more than their own.
While the dauer-inducing effect of L1-conditioned medium is intuitively easy to accept, the negative effect of L1 starvation on subsequent dauer formation is less clear. A recently published model of contrast effects (McNamara et al., 2013) may provide a theoretical framework for understanding this phenomenon. But to keep us from being too comfortable with our simplistic views nature presented us with another puzzle: L1 arrest does not negatively affect dauer formation in C. briggsae.
Pungaliya C, Srinivasan J, Fox BW, Malik RU, Ludewig AH, Sternberg PW, and Schroeder FC. (2009). A shortcut to identifying small molecule signals that regulate behavior and development in Caenorhabditis elegans. Proc. Natl Acad. Sci. U. S. A. 106, 7708-7713.
The C. elegans genome encodes a large family of putative insulin-like peptides comprising 40 genes (Pierce et al., 2001; Li et al. 2003). In spite of great interest in the insulin-like pathway, given its function in dauer formation, aging, and L1 arrest, the function of most insulin-like peptides remains uncharacterized. There is evidence, however, that they function as either agonists or antagonists of the insulin-like receptor DAF-2, promoting development or arrest, respectively (Pierce et al., 2001). The lack of loss-of-function phenotypes for most insulin-like peptides points towards redundancy, although to what extent remains unknown. Recent findings of Ritter et al. indicate that no single insulin is the sole agonist or antagonist for coordinating dauer formation through the DAF-2 receptor and that there may be complex patterns of redundancies between multiple insulins (Ritter et al., 2013).
We performed a microarray analysis comparing the gene expression between 24h starved L3 stage N2 and fully fed animals of the same age (Figure 1). The starved animals displayed a delay in growth and were still the size of L3 worms, whereas the fully fed animals developed normally to late L4 stage at the time of sampling. 37 of the insulin-like peptide genes (ins) were represented on the Agilent C. elegans Gene Expression Microarrays (design ID 020186).
25 of the 40 known ins genes were differentially expressed more than twofold in the starved animals (Figure 2). Remarkably, 21 insulins are upregulated under L3 starvation while only 4 are downregulated. The strongest increase in expression was observed for ins-35 (39.7 fold), ins-20 (11.4 fold), ins-16 (6.4 fold) and ins-33 (5.7 fold); the strongest decrease in ins-34 (35.4 fold), ins-37 (10.4 fold), ins-31 (4.6 fold) and ins-7 (3.9 fold). The high number of differentially regulated insulins seems to favor of a possible functional redundancy between these peptides, given that DAF-2 is the likely receptor for most of them. The strong change in transcript level of many of these peptides is indicative of their importance with regard to starvation survival.
INS-4, INS-6, INS-7 and DAF-28 have been suggested to be DAF-2 agonists (Pierce et al., 2001; Hua et al., 2003; Murphy et al. 2003), while INS-1, INS-17, INS-18, INS-33 and INS-35 represent antagonists (Pierce et al., 2001; Liu et al., 2004). Our data is consistent with a possible agonistic role for INS-7 (downregulated) and antagonistic role for INS-1, INS-17, INS-18, INS-33 and INS-35 (upregulated) with regard to L3 starvation and its resulting developmental effects. INS-4 and DAF-28 transcript levels, however, were not differentially regulated, and INS-6 was upregulated, indicating it might act as an antagonist under the condition of L3 starvation. The data also suggests possible new agonistic and antagonistic roles for many other insulin-like peptides, for example: INS-31, INS-34 and INS-37 as agonists, inhibiting developmental arrest; and INS-16, INS-20, INS-24, INS-25, INS-30 and INS-36 as antagonists, stimulating developmental arrest. A functional role for these insulins was not described before and might be specific to the condition of L3 starvation.
Different groups of insulins might therefore act differently depending on the physiological and environmental conditions, a concept that was also recently described in the ‘‘block design’’ by Ritter et al. (2013).
Hua QX, Nakagawa SH, Wilken J, Ramos RR, Jia W, Bass J, and Weiss MA. (2003). A divergent INS protein in Caenorhabditis elegans structurally resembles human insulin and activates the human insulin receptor. Genes Dev. 17, 826-831.
Murphy CT, McCarroll SA, Bargmann CI, Fraser A, Kamath RS, Ahringer J, Li H, and Kenyon C. (2003). Genes that act downstream of DAF16 to influence the lifespan of Caenorhabditis elegans. Nature 424, 277-283.
Pierce SB, Costa M, Wisotzkey R, et al. (2001). Regulation of DAF-2 receptor signaling by human insulin and ins-1, a member of the unusually large and diverse C. elegans insulin gene family. Genes Dev. 15, 672-686.
Survival of L1 arrested larvae in the absence of food is a common experiment in C elegans lifespan studies. A typical protocol involves resuspension of freshly bleached worm eggs in sterile inorganic buffer. Newly hatched worms arrest due to lack of food and their starvation survival in the following days is monitored by taking aliquots and counting live worms. It is tempting to think that such well defined and simple experimental conditions leave little room for variability. In reality, the outcome of this experiment is sensitive to a number of often neglected factors ranging from worm maintenance history dating several generations back to small temperature fluctuations during starvation. Recently, we found additional sources of artifacts in L1 survival assays.
When starved L1 larvae are incubated in M9 buffer in plastic tubes (e.g. polypropylene 15-ml tubes), worms tend to adhere to hydrophobic plastic walls, which decreases apparent worm counts in suspension. This effect becomes particularly noticeable at low worm densities (< 3 worms/μl, Fig. 1) We found that precoating tubes with bovine serum albumin (1% aqueous solution for an hour followed by 3 water rinses) renders plastic hydrophilic and greatly reduces the sticking problem. BSA solution and water for rinses should be sterile.
When the starvation experiment is performed in glass tubes or vials, worm sticking is not a problem, since clean glass is hydrophilic. The danger in this case comes from the cap, since seemingly inert material lining the cap can significantly affect survival rate if it comes in contact with liquid. We found that L1 worms starved in glass tubes with rubber-lined caps reproducibly survived starvation longer then worms in plastic tubes of the same volume (Fig. 2) or glass tubes with teflon-lined caps (Fig. 3). We detected several compounds in water from tubes with rubber-lined caps and identified one of them as 2-mercaptobenzothiazole (2-MBT), which was present at ca. 0.5 μM. 2-MBT, used as an accelerator in the vulcanization of rubber, is commonly seen in rubber leachates (Reepmeyer and Juhl, 1983). We speculate that longer survival is due to a hormetic effect of one or several chemicals leaching from rubber, which would be toxic at higher concentrations. However, low concentrations of synthetic 2-MBT alone were not sufficient to reproducibly extend L1 starvation survival (at 3 μM and above 2-MBT is toxic to worms). Other components of the rubber leachate may be necessary for the effect.
These results demonstrate artifacts that can both decrease and increase apparent survival rates in L1 starvation.