New Results

A memory of recent oxygen experience switches pheromone valence in C. elegans

Lorenz A. Fenk, Mario de Bono
doi: https://doi.org/10.1101/107524

ABSTRACT

Animals adjust their behavioral priorities according to momentary needs and prior experience. We show that C. elegans changes how it processes sensory information according to the oxygen environment it experienced recently. C.elegans acclimated to 7% O2 are aroused by CO2 and repelled by pheromones that attract animals acclimated to 21% O2. This behavioral plasticity arises from prolonged activity differences in a circuit that continuously signals O2 levels. A sustained change in the activity of O2 sensing neurons reprograms the properties of their post-synaptic partners, the RMG hub interneurons. RMG is gap-junctionally coupled to the ASK and ADL pheromone sensors that respectively drive pheromone attraction and repulsion. Prior O2 experience has opposite effects on the pheromone responsiveness of these neurons. These circuit changes provide a physiological correlate of altered pheromone valence. Our results suggest C. elegans stores a memory of recent O2 experience in the RMG circuit and illustrate how a circuit is flexibly sculpted to guide behavioral decisions in a context-dependent manner.

SIGNIFICANCE STATEMENT

Animals use memories of their recent environment to regulate their behavioral priorities. The basis for this cross-modal, experience-dependent plasticity is poorly understood. C. elegans feeds on bacteria in rotting fruit. It monitors O2 levels, and switches behavioral state when O2 approaches 21%. We show that C. elegans’ memory of its recent O2 environment reconfigures how it processes sensory information. Pheromones that attract animals acclimated to 21% O2 repel animals acclimated to 7% O2. O2 memory is encoded in the activity history of a circuit that continuously signals O2 levels. This circuit is connected to neurons driving pheromone attraction and repulsion. O2 experience changes the pheromone responsiveness of these sensors and their post-synaptic targets, correlating with the switch in pheromone valence.

INTRODUCTION

The body comprises multiple highly integrated subsystems working together to sustain life from moment-to-moment and over long time scales (1). Much of this coordination involves dynamically interacting neural circuits that optimize responses to current circumstances by taking into account sensory input, organismal state, and previous experience (2–8). Circuit crosstalk enables animals to adjust their behavioral priorities in response to a changing environment e.g. variation in temperature, humidity, day length, or oxygen (O2) levels (9–13). While some behavioral adjustments can be rapid (14, 15), others develop over time, as animals adapt to changed conditions. How animals store information about their recent environment, and use this information to modify behavioral choices is poorly understood.

The compact nervous system of Caenorhabditis elegans, which comprises only 302 uniquely identifiable neurons (wormwiring.org) (16), provides an opportunity to study the links between prior environmental experience, circuit plasticity, and behavioral change. This nematode is adapted to a life feeding on bacteria in rotting fruit (17, 18). It has sensory receptors for odors, tastants, pheromones, and respiratory gases, as well as temperature, mechanical, and noxious cues (19–22). Despite this simplicity, the mechanisms by which its nervous system marshals information about past and present sensory experience to shape behavioral priorities have largely not been dissected. While valuable (23), the anatomical connectome is insufficient to explain or predict neuronal network function (24, 25), partly because neuromodulators can dynamically reconfigure and specify functional circuits (26–28).

When ambient O2 approaches 21% C.elegans wild isolates become persistently aroused and burrow to escape the surface (29–31). This state switch is driven by tonically signalling O2 receptors called URX, AQR and PQR (32, 33) whose activity increases sharply when O2 approaches 21% (29, 31, 34, 35). The URX neurons are connected by gap junctions and reciprocal synapses to the RMG interneurons, and tonically stimulate RMG to promote escape from 21% O2 (wormwiring.org)(16, 36). URX and RMG are both peptidergic, and at 21% O2 tonically release neuropeptides (29, 36). RMG is connected by gap junctions to several other sensory neurons besides URX, including pheromone sensors (16, 37). Whether information communicated from URX to RMG about the O2 environment modulates other sensory responses is unknown.

Here, we show that acclimating C. elegans to different O2 environments gradually reconfigures its response to sensory cues. Animals acclimated to 7% O2 but not 21% O2 are aroused by CO2. Pheromones that attract animals acclimated at 21% O2 repel animals acclimated to 7% O2. These changes are driven by experience-dependent remodelling of URX O2 sensors, RMG interneurons, and the ASK and ADL pheromone sensors.

RESULTS

Acclimation to different O2 environments reprograms CO2 responses

C. elegans escape 21% O2, which signals that animals are at the surface, and accumulate at 7% O2, which indicates that animals are burrowed (29, 31, 32). We speculated that C. elegans gradually change their sensory preferences when shifted between these two environments.

To test our hypothesis, we first examined responses to CO2. CO2 is aversive to C. elegans, and its concentration rises as O2 levels fall, due to respiration. Animals escaping 21% O2 will thus often encounter high CO2, creating conflicting drives that we thought could be ecologically significant. Previous work showed that C. elegans immediately suppresses CO2 avoidance when O2 levels approach 21%, due to increased tonic signalling from URX O2 sensors (38–41). We speculated that not only current but also prior O2 experience remodels C. elegans’ CO2 responses. To test this, we kept wild isolates from California, France, and Hawaii overnight at 21% or 7% O2, and compared their responses to 3% CO2 on a thin lawn of bacteria kept at 7% O2. After halting briefly, animals acclimated to 7% O2 became persistently aroused at 3% CO2 (Fig. S1A–C), unlike animals acclimated to 21% O2.

To probe this plasticity, we studied the N2 lab strain. Unlike natural isolates, N2 is aroused by 3% CO2 regardless of prior O2 experience (Fig. 1A). In N2, output from the RMG interneurons, a major relay of the circuit signalling 21% O2, is blocked by a hyperactive neuropeptide receptor, NPR-1 215V (36, 37). Natural C. elegans isolates have a less active receptor, NPR-1 215F, which does not block RMG output (42). Does this account for altered CO2 responses? Disrupting npr-1 caused N2 animals to behave like natural isolates, and to inhibit CO2-evoked arousal when acclimated to 21% O2 (Fig. 1A). The effects of acclimating npr-1 animals to 7% O2 developed over 16 hours, and were reversed within 3 hours if animals were transferred to 21% O2 (Fig. S2A, B). Selectively expressing NPR-1 215V in RMG interneurons prevented npr-1 animals from acclimating to 21% O2 (Fig. 1B), and disrupting the GCY-35 soluble guanylate cyclase, a molecular O2 sensor in URX required for the URX – RMG circuit to signal 21% O2 (29, 31, 32, 34, 43) had the same effect (Fig. 1C).

Figure 1.
Figure 1. Recent O2 experience regulates CO2-evoked arousal

(A) N2 animals and npr-1 animals acclimated to 7% O2 exhibit a robust and persistent increase in speed when CO2 levels rise to 3%, whereas, npr-1 animals acclimated to 21% O2 do not. n = 247–302 animals. **** p < 0.0001; ns, not significant; Wilcoxon signed-rank test. In this and subsequent figures solid lines indicate the mean and shaded areas represent the standard error of the mean (S. E. M). Black bars indicate time intervals used for statistical comparisons (boxplots). Assays were performed in the presence of food and background O2 was kept at 7%. (B) Selective expression of NPR-1 215V in RMG, (C) knocking out gcy-35, or (D) RNAi-mediated knockdown of the carboxypeptidase E EGL-21 in RMG, prevent npr-1 animals acclimated to 21% O2 from suppressing CO2-evoked arousal. n = 104–235. Boxes in this and all subsequent panels show the median (black line) and extend from the 25th to 75th percentiles and whiskers represent 10th to 90th percentiles.

The NPR-1 215V receptor inhibits RMG peptidergic transmission (36). We speculated that circuitry effects of prior O2 experience might reflect prolonged differences in RMG peptidergic release. To test this we selectively knocked down the carboxypeptidase E ortholog egl-21 in RMG using RNAi. Processing of most C. elegans neuropeptides depends on EGL-21 (44). RMG-knockdown of egl-21 prevented npr-1 animals from acclimating to 21% O2 (Fig. 1D). These data suggest that neuropeptide release from RMG is required for C. elegans acclimated to 21% O2 to suppress CO2-evoked arousal.

Pheromone valence changes with prior O2 experience

We studied how prior O2 experience alters CO2 responses because of the special relationship between these gases. C. elegans has, however, many CO2-responsive neurons, complicating analysis of how persistent differences in RMG activity alter the CO2 circuits (39, 45, 46). Several studies have reported differences in the sensory responses of N2 and npr-1 mutants associated with altered RMG function (36, 37, 47, 48). We speculated that at least some of these differences could reflect a diminished capacity of N2 animals to acclimate to 21% O2 due to reduced neurosecretion from RMG.

One such behavior is pheromone preference (37). Select pheromone blends attract npr-1 hermaphrodites but repel N2 hermaphrodites (37, 49). We replicated these observations using an equimolar 10 nM mix of asc-ωC3 (ascaroside C3), asc-C6-MK (ascaroside C6), and asc-ΔC9 (ascaroside C9) pheromones, (37, 50–52) (Fig. 2A). We then asked if acclimating N2 and npr-1 hermaphrodites overnight in different O2 environments altered their pheromone response. We assayed animals at 21% O2. Whereas npr-1 animals acclimated to 21% O2 were attracted to the pheromone mix, npr-1 animals acclimated to 7% O2 robustly avoided it (Fig. 2B). Acclimating N2 animals at different O2 levels did not alter pheromone avoidance (Fig. 2B), recapitulating our observations with CO2 (Fig. 2SC) Disrupting gcy-35 switched the pheromone attraction exhibited by npr-1 animals acclimated at 21% O2 into repulsion (Fig. 2B). In summary, reducing the activity of O2 sensing circuitry for prolonged periods of time – either via environmental or genetic manipulation – transforms pheromone attraction to pheromone avoidance.

Figure 2.
Figure 2. Pheromone valence changes with prior O2 experience

(A) Quadrant assay for pheromone preference (after (37)). (B) Behavioral responses to an equimolar 10 nM mix of C3, C6 and C9 ascaroside pheromones. npr-1 animals acclimated to 21% O2 are attracted to the pheromone whereas siblings acclimated to 7% O2 robustly avoid it. N2 avoid pheromones irrespective of whether they have been acclimated to 7% or 21% O2. The soluble guanylate cyclase GCY-35 is required for normal O2 responses and pheromone attraction in npr-1 animals acclimated at 21% O2. ** p < 0.01; ns, not significant; One-way ANOVA with Tukey’s multiple comparisons test. n = 8 assays each. (C) Previous O2 experience sculpts pheromone responses in ASK sensory neurons. Acclimation to 7% O2 reduces pheromone-evoked Ca2+ responses in ASK, consistent with altered behavioral preference. (D) Quantification of data shown in (C). Heat maps in this and all subsequent figures show individual Ca2+ responses. n = 35–36 animals. **** p < 0.0001; ** p < 0.01; *p < 0.05 ns, not significant; Mann-Whitney U test.

O2 experience changes pheromone responses in ASK neurons

How does prior O2 experience switch pheromone valence? The altered behavior must reflect some lasting change in the circuitry that couples sensory detection to motor output. The principal neurons driving pheromone attraction are the ASK ciliated head neurons. ASK responds to pheromone with a decrease in Ca2+ (the ‘ON’ response) that quickly returns to above baseline when pheromone is removed (the ‘OFF’ response). The pheromone-evoked Ca2+ response in ASK is bigger in npr-1 animals compared to N2 animals, a difference thought to contribute to the opposite pheromone preference of these strains (37). Does prior O2 experience change the responsiveness of ASK to pheromones? To test this, we measured pheromone-evoked Ca2+ responses in ASK using the ratiometric Ca2+ indicator YC3.60. Overnight acclimation at 7% O2 attenuated the ASK pheromone response in npr-1 animals to levels found in N2 (Fig. 2C and 2D). Thus, prior O2 experience alters ASK pheromone responses, commensurate with a change in behavioural preference.

Peptidergic feedback heightens RMG responsiveness to 21% O2 after sustained exposure to 21% O2

RMG interneurons are connected to both the URX O2 receptors and the ASK pheromone sensors via gap junctions (wormwiring.org)(16, 37). A simple prediction made by our data is that the response properties of RMG change when npr-1 animals are acclimated to different O2 levels, and this alters the properties of ASK. To explore this we compared the RMG Ca2+ responses evoked by 21% O2 in npr-1 animals acclimated to 21% and 7% O2. Animals acclimated to 7% O2 showed significantly smaller RMG responses than those acclimated to 21% O2 (Fig. 3A and 3B). Persistent exposure to 21% O2 increases RMG responses to this stimulus.

To probe this change in RMG properties we compared the URX Ca2+ responses evoked by 21% O2 in animals acclimated to 21% and 7% O2. URX drives RMG responses (36). URX responses were smaller in animals acclimated to 7% O2 (Fig. 3C and 3D), suggesting changes in RMG properties partly reflect plasticity in URX. In addition, acclimating npr-1 animals to 21% O2 was unable to increase RMG responsiveness to 21% O2 if we selectively knocked down peptidergic transmission from RMG by RNAi of EGL-21 CPE (Fig. 3A and 3B). These data suggest there is a positive feedback loop by which tonic peptidergic signalling from RMG in npr-1 animals kept at 21% O2 increases RMG responsiveness to 21% O2. Experience-dependent plasticity in RMG and URX represent neural correlates of acclimation to different O2 environments.

Figure 3.
Figure 3. Peptidergic feedback regulates RMG properties and pheromone preference

(A) Acclimation to 7% O2, or knockdown of egl-21, similarly reduce RMG Ca2+ responses evoked by a 21% O2 stimulus. (B) Quantification of data shown in (A). n = 20–21 animals. * p < 0.05; ns, not significant; Mann-Whitney U test. (C) Acclimation to 7% O2 reduces URX Ca2+ responses evoked by a 21% O2 stimulus. (D) Quantification of data shown in (C). n = 38–39 animals. * p < 0.05; Mann-Whitney U test. (E) Knockdown of egl-21 in RMG diminishes pheromone-evoked Ca2+ responses in ASK to levels observed in N2. (F) Quantification of data shown in (E). n = 20–21 animals. * p < 0.05; ** p < 0.01; **** p <0.0001; ns, not significant; Mann-Whitney U test. (G) RNAi knockdown of egl-21 in RMG prevents npr-1 animals acclimated to 21% O2 from being attracted to pheromone. n = 12 assays each. * p < 0.05; One-way ANOVA followed by Dunnett’s multiple comparisons test.

Importantly, RNAi knockdown of EGL-21 in RMG altered pheromone responses of npr-1 animals acclimated to 21% O2, reducing pheromone-evoked Ca2+ responses in ASK to N2-like levels (Fig. 3E and 3F), and conferring robust pheromone avoidance (Fig. 3G). Thus, peptidergic signalling from RMG mediates multiple effects of acclimation to 21% O2: an increase tonic Ca2+ response to 21% O2 in RMG, a bigger ASK response to pheromone cues, and decreased C. elegans avoidance of pheromone.

Communication between neurons in the RMG circuit

The neuroanatomy suggests RMG is gap-junctionally connected to multiple sensory neurons, including ASK, the ADL and ASH nociceptors, the AWB olfactory neurons, and the IL2 chemo/mechanoreceptors (Fig. 4A) (wormwiring.org)(16). Changes in RMG may therefore influence the signalling properties of each of these neurons, and vice-versa. Previous studies suggest that the O2-sensing URX neurons cooperate with the nociceptive ADL and ASH neurons, and the ASK pheromone sensors, to promote C. elegans aggregation and escape from 21% O2 (33, 36, 37, 53). However, in the absence of physiological data it is unclear what information RMG neurons receive or transmit, apart from tonic O2 input from URX (29)(36). We asked if O2-evoked responses in RMG propagated to ADL and ASK. The wiring diagram suggests ASK and ADL are connected to RMG exclusively via gap junctions. ASK and ADL each showed O2-evoked Ca2+ responses in npr-1 animals (Fig. S3A–S3D). We also imaged RMG responses evoked by the pheromone mix we used to stimulate ASK (Fig. 4B). RMG responded with Ca2+ dynamics similar to those observed in ASK (Fig. 4B), suggesting information can flow from ASK to RMG. These results support a hub-and-spoke model in which different sensory inputs are integrated through gap junctions with the RMG hub (37).

Figure 4.
Figure 4. RMG hub neurons respond to pheromones and alter their response according to recent O2 experience

(A) Circuit showing connections between RMG interneurons and O2-sensing, nociceptive, and pheromone-sensing neurons. (B) An equimolar (100 nM) mix of C3, C6, and C9 ascarosides inhibit RMG. n = 57 animals. (C) RMG shows robust pheromone responses in both npr-1 and N2 animals. (D) Quantification of data shown in (C). n = 35–36 animals. ns, not significant; Mann-Whitney U test. (C) Acclimation to 7% O2 alters RMG properties and diminishes both ON-and OFF-responses to pheromone addition and removal. (D) Quantification of data shown in (C). n = 36 animals each. *** p < 0.001; **** p < 0.0001; Mann-Whitney U test. (E–G) Acclimation to 7% O2 enhances ADL pheromone responses and acute pheromone repulsion. npr-1 animals show decreased avoidance of the C9 ascaroside compared to N2 when grown under standard conditions, but not when acclimated to 7% O2. Plotted are the avoidance index (E) and fraction of animals reversing (E’), in response to a drop of diluted C9 (10 nM) applied to the nose. n = 260–280 animals each. * p < 0.05; *** p < 0.001; **** p < 0.0001; ns, not significant; One-way ANOVA followed by Tukey’s multiple comparisons test. (F) The Ca2+ responses evoked in ADL by 10 nM C9 pheromone are larger in npr-1 animals acclimated to 7% O2 compared to siblings acclimated at 21% O2. (G) Quantification of data plotted in (F). n = 23– 24 animals. ** p < 0.01; Mann-Whitney U test. (H) Model.

NPR-1 215V signalling has been proposed to silence the hub-and-spoke circuit (24, 27). One attractive model is that signalling from the neuropeptide receptor closes RMG gap junctions (37, 49). To investigate this, we first compared pheromone-evoked Ca2+ responses in RMG in N2 and npr-1 animals, but did not observe any significant differences (Fig. 4C and 4D). We then compared O2-evoked responses in ASK, and also did not observe differences between the two genotypes (Fig. S3A and S3B). By contrast, npr-1 but not N2 animals displayed a strong O2-evoked response in ADL neurons, (Fig. S3C and S3D); this response, unlike the ADL pheromone response (see below), did not require the TRPV1 ortholog OCR-2 (Fig. S4A–S4D). Although other interpretations are possible, a simple model to explain our data is that NPR-1 215V signalling in RMG affects different gap junctions differently, inhibiting RMG – ADL communication but having smaller or no effects on the RMG - ASK connection.

O2 experience sculpts RMG and ADL pheromone responses

The pheromone attraction mediated by ASK neurons and promoted by RMG signalling is proposed to antagonize pheromone avoidance driven by the ADL neurons in a push-pull mechanism (49). The relative strength of these arms determines the animal’s response. We found that acclimating npr-1 animals to 7% O2 greatly reduced pheromone-evoked responses in RMG compared to animals kept at 21% O2 (Fig. 4C and 4D). Thus, acclimation to 7% O2 weakens both the ASK (Fig. 2C and 2D) and RMG circuit elements that drive attraction to pheromone.

Given the neuroanatomy, and the ability of RMG to influence Ca2+ in ADL, changes in pheromone-evoked ASK – RMG responses associated with acclimation to different O2 levels might alter pheromone-evoked responses in ADL. ADL neurons are activated by the ascaroside C9, and a drop of C9 increases the probability of animals reversing (49). In this behavioral paradigm the fraction of animals reversing provides a measure of the pheromone’s repulsiveness, and is significantly higher in N2 than npr-1 animals at low pheromone concentrations (10 nM). Higher concentrations of C9 elicit strong repulsion irrespective of npr-1 genotype (49). We confirmed that N2 animals showed enhanced repulsion from 10 nM C9 compared to npr-1 animals (Fig. 4E). We then showed that npr-1 animals acclimated overnight to 7% O2 enhanced their avoidance of C9, and behaved indistinguishably from N2 (Fig. 4E). The avoidance index (A.I.) used in this assay (49, 54) is calculated as [(fraction reversing to pheromone) − (fraction reversing to buffer alone)], and any change in the A.I. could reflect an altered response to the buffer rather than to C9. Consistent with enhanced pheromone avoidance, npr-1 animals reversed more in response to C9 if they were acclimated to 7% O2 (Fig. 4E’).

Pheromone-evoked Ca2+ responses in ADL neurons were previously characterized using 100 nM C9, a concentration that elicits strong and comparable repulsion in N2 and npr-1 animals (49). By using GCaMP6 var500 we could record ADL responses to 10 nM C9, and assess the impact of previous O2 experience under conditions similar to those used in behavioral assays. npr-1 animals acclimated to 7% O2 showed significantly bigger ADL Ca2+ responses compared to siblings acclimated to 21% O2 (Fig. 4F and 4G). Together our data suggest that acclimation to 7% O2 simultaneously weakens the ASK and RMG circuit elements that drive attraction to pheromone, and strengthens the ADL pheromone response driving repulsion, thereby switching the animal’s behavioral choice.

DISCUSSION

Unfavorable environments can evoke slow, sustained changes in behavioral priorities that reflect an altered internal state. The neural mechanisms mediating such integrative, experience-dependent plasticity are poorly understood. C. elegans persistently attempts to escape 21% O2 (29), presumably because this O2 concentration signals unfavorable surface exposure (31, 32). We find that the O2 milieu experienced recently by C. elegans changes the way it processes sensory information. Pheromones that attract C. elegans acclimated to 21% O2 repel animals acclimated to 7% O2; 3% CO2 triggers sustained arousal in animals acclimated to 7% O2 but has comparatively little effect in animals acclimated to 21% O2.

A memory of previous O2 experience arises from prolonged differences in the activity of a tonically active circuit. Exposure to 21% O2 tonically stimulates the URX O2 receptor neurons and their synaptic partners, the RMG interneurons. Sustained stimulation of URX and RMG at 21% O2 increases their response to 21% O2. The reprogramming of RMG requires peptidergic signaling competence in this interneuron. Our data suggest a simple model in which over time, sustained peptide release from RMG at 21% O2 feeds back to alter RMG properties. In animals kept at 7% O2 peptide release from RMG is low, disrupting the feedback. In this neural integrator model hysteresis in the build up and decay of peptide signaling accounts for the time delays as animals acclimate to 7% or 21% O2. We previously showed that neuropeptide expression in RMG is positively coupled to neurosecretion from RMG (Laurent et al., 2015), consistent with a positive feedback loop in this interneuron. Tonic circuit activity is common in brains. We speculate that such circuits will often store information about their activity history, and potentially about the animal’s experience, by incorporating peptidergic positive feedback loops.

RMG has neuroanatomical gap junctions not only with URX, but also with the ASK and ADL pheromone sensors (16, 37). This arrangement suggests that information can be integrated across the circuit (Macosko et al., 2009), but physiological data was hitherto absent. We show that ASK and ADL show O2-evoked Ca2+ responses, and that acclimating animals to different O2 levels alters O2 and / or pheromone-evoked responses in each of the URX, RMG, ASK and ADL neurons. Inhibiting peptidergic transmission from RMG prevents RMG and ASK neurons from changing their pheromone responsive properties in animals acclimated to 21% O2; it also prevents the experience dependent switch in pheromone valence.

Changes in the pheromone-evoked responses of ASK and ADL neurons are consistent with changes in RMG changing communication across the network. For example, in animals acclimated to 21% O2 pheromone-evoked responses in ASK could inhibit ADL pheromone responses, whereas in animals acclimated at 7% O2 this communication may be less potent. While this is plausible, we cannot exclude that the intrinsic properties of several neurons in the circuit are altered by O2 experience.

We see parallels between our observations and a Drosophila study showing that repeated presentation of an aversive shadow cue leads to a persistent change in behavioral state that scales with the number and frequency of the presentations (55, 56). Our findings are also reminiscent of ‘latent modulation’ in the feeding network of Aplysia, where the history of activation in some circuit elements has a lasting effect on subsequent responses, most likely by changing neuronal excitability through peptidergic modulation (57, 58).

Why should C. elegans reconfigure its sensory responses according to prior O2 experience? It is tempting to speculate about a behavioral hierarchy (59) that gives priority to escape from 21% O2, and that dominates over sensory drives that could hinder escape from the surface. Animals at the surface may gradually suppress their aversion to CO2 to facilitate escape to low O2/high CO2 environments. Once the threat of exposure at the surface recedes, strong aversive responses to CO2 again become adaptive. In a boom-and-bust species like C. elegans (18), pheromones may be aversive because they predict an unsustainable population density. However, if escaping the surface is more important than accumulating in a crowded environment, attraction towards pheromones may be transiently adaptive because crowded environments predict reduced O2. Irrespective of the precise selective advantage(s), our data suggest C. elegans can adopt alternate persistent internal states according to the length of time they have been exposed to aversive or preferred O2 levels. In these states neural circuits process sensory information differently, changing the animal’s behavioral priorities.

AUTHOR CONTRIBUTIONS

L.F and M.d. B designed experiments, L.F. performed the experiments, L.F. and M.d.B analysed the data and wrote the manuscript.

MATERIALS AND METHODS

Strains

Strains were grown and maintained under standard conditions with E. coli OP50 as food (60). To cultivate animals in specific O2 environments we used a Coy O2 control glove box (Coy, Michigan, USA).

Behavioral Assays

Locomotion assays were performed as described previously (45, 46). 20–25 adult hermaphrodites were picked to NGM plates seeded 16–20 h earlier with 20 µL of E. coli OP50 grown in 2 xTY. To create a behavioral arena with a defined atmosphere we lowered a 1 cm × 1 cm × 200 µm deep polydimethylsiloxane (PDMS) chamber on top of the worms, with inlets connected to a PHD 2000 Infusion syringe pump (Harvard apparatus). We pumped in defined gas mixtures (BOC, UK) that were humidified using a sintered gas bubbler (SciLabware, UK) at a flow rate of 3.0 mL/min. Movies were recorded at 2 frames/s using a Point Gray Grasshopper camera mounted on a Leica M165FC dissecting microscope, and were analysed using custom-written Matlab software (36) to detect omega turns and calculate instantaneous speed.

Chemotaxis to pheromones was assayed essentially as described previously (37, 61), using four-quadrant Petri plates (Falcon X plate, Becton Dickinson Labware, USA). For each assay 200 worms were picked to a fresh seeded plate for 2–3 hours, washed three times with chemotaxis buffer, and placed at the centre of a 10 cm assay plate with pheromones in alternating quadrants. Animals were scored after ~ 15 min and a chemotaxis index calculated as (number of animals on pheromone quadrants – number of animals on buffer quadrants) / (total number of animals). We used an equimolar (10 nm) mix of the ascarosides C3, C6, and C9. Assays were repeated on at least 4 different days.

Acute C9 avoidance was examined in the presence of food, using the drop test (62) and as described by Jang and colleagues (49, 54). Responses were scored as reversals if animals initiated a backward movement within 4 s after stimulation that was equal or longer than half their body length. The fraction reversing is given by (number of animals that make a long reversal) / (number if total animals tested); the effect size or avoidance index was calculated as (fraction reversing to pheromone) – (fraction reversing to buffer alone).

Ca2+ Imaging

Ca2+ imaging was performed as described previously (37, 49, 63), using microfluidic devices (MicroKosmos, Ann Arbor, MI) to immobilize animals and either a 1:1:1 ratio of three ascarosides (C3, C6, C9) or C9 alone, at the concentration indicated (10nm – nm). For O2 and CO2 experiments, worms were glued to agarose pads (2% in M9 buffer, 1 mM CaCl2) using Dermabond tissue adhesive, with the nose and tail immersed in M9 buffer (45, 46). All imaging was performed on an inverted microscope (Axiovert; Zeiss) with a 40× C-Apochromat lens (water immersion, N.A. 1.0), and MetaMorph acquisition software (Molecular Devices). Recordings were at two frames/s with a 100 ms exposure time. Photobleaching was minimized using optical density filter 2.0 or 1.5. For ratiometric imaging experiments we used an excitation filter (Chroma) to restrict illumination to the cyan channel, and a beam splitter (Optical Insights) to separate the cyan and yellow emission light. Animals were pre-exposed to excitation light for ~ 1 min in all experiments. A custom-written Matlab script was used to analyze image stacks(36).

Molecular Biology and Generation of Transgenic Lines

Expression constructs were made using the MultiSite Gateway Three-Fragment Vector Construct Kit (Life Technologies). Promoters used include: sra-9 (3 kb; ASK), sre-1 (4 kb; ADL), flp-21 and ncs-1 (RMG). Promoter fragments were amplified from genomic DNA and cloned in the first position of the Gateway system, genes of interest in the second position, and the unc-54 3′ UTR or the SL2∷mCherry sequence in the third position. Constructs were injected at 30–55 ng/µL, with a coinjection marker (unc-122∷RFP or unc-122∷GFP) at 50–60 ng/µL.

Statistical methods

Statistical analyses used Prism 6 (GraphPad) and MATLAB (MathWorks). No statistical method was used to predetermine sample size, which are similar to those generally employed in the field. Exact tests used are indicated in figure legends; imaging and locomotion data were analyzed using a non-parametric Mann-Whitney U or Wilcoxon signed-rank test.

SUPPLEMENTAL FIGURES

Figure S1.
Figure S1. Acclimation to different O2 environments reprograms CO2 responses in natural C. elegans isolates.

(A-C)Wild strains modulate their CO2 response according to recent O2 experience. These strains encode NPR-1 215F, the natural low activity isoform of NPR-1. n = 116–171.

Figure S2.
Figure S2. Time line and reversibility of acclimation to different O2 levels.

(A and B) Mean speed of npr-1 animals at 3% CO2, plotted against time exposed to 7% or atmospheric (~ 21%) O2. (A) Acclimation to 7% happens gradually and animals continue to increase their speed over many hours. n = 151–171. (B) Acclimation is reversed rapidly, and after ≤ 3 hours animals behave like siblings grown at 21% O2. n = 138–188. Error bars in (A and B) andshaded regions in (A’ and B’) represent S. E. M. **** p < 0.0001; *** p < 0.001; ns, not significant; Kruskal–Wallis ANOVA with Dunn’s multiple comparisons test. (C) N2 are strongly aroused by a 3% CO2 stimulus, irrespective of whether they have been acclimated at 21% or 7% O2. n = 462–518 animals. **** p < 0.0001; Wilcoxon signed-rank test.

Figure S3.
Figure S3. ASK and ADL sensory neuron spokes respond to O2.

(A) O2-evoked Ca2+ responses in ASK do not differ between N2 and npr-1 animals. (B) Quantification of data plotted in (A). n = 21–24 animals; Mann-Whitney U test. Blue shading indicates a shift from 7% to 21% O2. (C) ADL sensory neurons show robust responses to a 21% O2 stimulus in npr-1 but not in N2 animals. (D) Quantification of data shown in (C). n = 30 animals each. **** p < 0.0001; Mann-Whitney U test. Note different Ca2+ sensors were used to image ASK (YC3.60) and ADL (GCaMP6).

Figure S4.
Figure S4. The OCR-2 TRPV channel is required for ADL responses to pheromone C9 but not to 21% O2.

(A) OCR-2 is not required for ADL responses to a 21% O2 stimulus, although Ca2+ appears to rise less sharply in mutants. (B) Quantification of data plotted in (A). n = 21–22 animals each; Mann-Whitney U test. Blue shading indicates a shift from 7% to 21% O2. (C) OCR-2 is required cell autonomously for ADL response to the C9 ascaroside. (D) Quantification of data shown in (C). n = 11– 12 animals each. **** p < 0.0001; Mann-Whitney U test. Grey shading indicates stimulation with 10 nM C9.

ACKOWLEDGEMENTS

We thank Rebecca Butcher for ascarosides, the Caenorhabditis Genetics Centre for strains, and members of the de Bono and Schafer labs for advice and comments. This work was supported by the European Research Council (AdG 269058) and the Medical Research Council (UK).

Footnotes

  • Classification: Major category - Biological Sciences, Minor category - Neuroscience; Genetics

REFERENCES

  1. 1.↵
    LeDoux J (2012) Rethinking the emotional brain. Neuron 73(4):653–676.
    OpenUrlCrossRefPubMedWeb of Science
  2. 2.↵
    Saeki S, Yamamoto M, Iino Y (2001) Plasticity of chemotaxis revealed by paired presentation of a chemoattractant and starvation in the nematode Caenorhabditis elegans. J Exp Biol 204(Pt 10):1757–1764.
    OpenUrlAbstract
  3. 3.
    Kindt KS et al. (2007) Dopamine mediates context-dependent modulation of sensory plasticity in C. elegans. Neuron 55(4):662–676.
    OpenUrlCrossRefPubMedWeb of Science
  4. 4.
    Su CY, Wang JW (2014) Modulation of neural circuits: how stimulus context shapes innate behavior in Drosophila. Curr Opin Neurobiol 29:9–16.
    OpenUrlCrossRefPubMed
  5. 5.
    Hoopfer ED (2016) Neural control of aggression in Drosophila. Curr Opin Neurobiol 38:109–118.
    OpenUrlCrossRef
  6. 6.
    Cohn R, Morantte I, Ruta V (2015) Coordinated and Compartmentalized Neuromodulation Shapes Sensory Processing in Drosophila. Cell 163(7):1742–1755.
    OpenUrlCrossRefPubMed
  7. 7.
    Owald D, Waddell S (2015) Olfactory learning skews mushroom body output pathways to steer behavioral choice in Drosophila. Curr Opin Neurobiol 35:178–184.
    OpenUrlCrossRefPubMed
  8. 8.↵
    Stowers L, Liberles SD (2016) State-dependent responses to sex pheromones in mouse. Curr Opin Neurobiol 38:74–79.
    OpenUrl
  9. 9.↵
    Chaffee RR, Roberts JC (1971) Temperature acclimation in birds and mammals. Annu Rev Physiol 33:155–202.
    OpenUrlCrossRefPubMedWeb of Science
  10. 10.
    Lewis L, Ayers J (2014) Temperature preference and acclimation in the Jonah Crab, Cancer borealis. Journal of Experimental Marine Biology and Ecology
  11. 11.
    Chown SL, Sørensen JG, Terblanche JS (2011) Water loss in insects: an environmental change perspective. J Insect Physiol 57(8):1070–1084.
    OpenUrlCrossRefPubMedWeb of Science
  12. 12.
    Tauber E, Kyriacou BP (2001) Insect photoperiodism and circadian clocks: models and mechanisms. J Biol Rhythms 16(4):381–390.
    OpenUrlCrossRefPubMedWeb of Science
  13. 13.↵
    Prabhakar NR et al. (2009) Long-term regulation of carotid body function: acclimatization and adaptation–invited article. Adv Exp Med Biol 648:307–317.
    OpenUrlCrossRefPubMedWeb of Science
  14. 14.↵
    Turner SL, Ray A (2009) Modification of CO2 avoidance behaviour in Drosophila by inhibitory odorants. Nature 461(7261):277–281.
    OpenUrlCrossRefPubMedWeb of Science
  15. 15.↵
    Grosjean Y et al. (2011) An olfactory receptor for food-derived odours promotes male courtship in Drosophila. Nature 478(7368):236–240.
    OpenUrlCrossRefPubMedWeb of Science
  16. 16.↵
    White JG, Southgate E, Thomson JN, Brenner S (1986) The structure of the nervous system of the nematode Caenorhabditis elegans. Philosophical Transactions of the Royal Society of London B(314):1–340.
    OpenUrlCrossRefPubMed
  17. 17.↵
    Felix MA, Braendle C (2010) The natural history of Caenorhabditis elegans. Curr Biol 20(22):R965–9.
    OpenUrlCrossRefPubMedWeb of Science
  18. 18.↵
    Frézal L, Félix MA (2015) C. elegans outside the Petri dish. Elife 4
  19. 19.↵
    Rengarajan S, Hallem EA (2016) Olfactory circuits and behaviors of nematodes. Curr Opin Neurobiol 41:136–148.
    OpenUrl
  20. 20.
    Goodman MB (2006) Mechanosensation. WormBook 1–14.
  21. 21.
    Bargmann CI (2006) Chemosensation in C. elegans. WormBook 1–29.
  22. 22.↵
    Aoki I, Mori I (2015) Molecular biology of thermosensory transduction in C. elegans. Curr Opin Neurobiol 34:117–124.
    OpenUrlCrossRefPubMed
  23. 23.↵
    Denk W, Briggman KL, Helmstaedter M (2012) Structural neurobiology: missing link to a mechanistic understanding of neural computation. Nat Rev Neurosci 13(5):351–358.
    OpenUrlCrossRefPubMed
  24. 24.↵
    Bargmann CI, Marder E (2013) From the connectome to brain function. Nat Methods 10(6):483–490.
    OpenUrlCrossRefPubMedWeb of Science
  25. 25.↵
    Koch C, Laurent G (1999) Complexity and the nervous system. Science 284(5411):96–98.
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    Marder E (2012) Neuromodulation of neuronal circuits: back to the future. Neuron 76(1):1–11.
    OpenUrlCrossRefPubMedWeb of Science
  27. 27.↵
    Bargmann CI (2012) Beyond the connectome: how neuromodulators shape neural circuits. Bioessays 34(6):458–465.
    OpenUrlCrossRefPubMedWeb of Science
  28. 28.↵
    Graebner AK, Iyer M, Carter ME (2015) Understanding how discrete populations of hypothalamic neurons orchestrate complicated behavioral states. Front Syst Neurosci 9:111.
    OpenUrl
  29. 29.↵
    Busch KE et al. (2012) Tonic signaling from O(2) sensors sets neural circuit activity and behavioral state. Nat Neurosci 15(4):581–591.
    OpenUrlCrossRefPubMed
  30. 30.
    Rogers C, Persson A, Cheung B, de Bono M (2006) Behavioral motifs and neural pathways coordinating O2 responses and aggregation in C. elegans. Curr Biol 16(7):649–659.
    OpenUrlCrossRefPubMedWeb of Science
  31. 31.↵
    Persson A et al. (2009) Natural variation in a neural globin tunes oxygen sensing in wild Caenorhabditis elegans. Nature 458(7241):1030–1033.
    OpenUrlCrossRefPubMedWeb of Science
  32. 32.↵
    Gray JM et al. (2004) Oxygen sensation and social feeding mediated by a C. elegans guanylate cyclase homologue. Nature 430(6997):317–322.
    OpenUrlCrossRefPubMedWeb of Science
  33. 33.↵
    Coates JC, De Bono M (2002) Antagonistic pathways in neurons exposed to body fluid regulate social feeding in Caenorhabditis elegans. Nature 419(6910):925–929.
    OpenUrlCrossRefPubMedWeb of Science
  34. 34.↵
    Zimmer M et al. (2009) Neurons detect increases and decreases in oxygen levels using distinct guanylate cyclases. Neuron 61(6):865–879.
    OpenUrlCrossRefPubMedWeb of Science
  35. 35.↵
    McGrath PT et al. (2009) Quantitative mapping of a digenic behavioral trait implicates globin variation in C. elegans sensory behaviors. Neuron 61(5):692–699.
    OpenUrlCrossRefPubMedWeb of Science
  36. 36.↵
    Laurent P et al. (2015) Decoding a neural circuit controlling global animal state in C. elegans. Elife 4
  37. 37.↵
    Macosko EZ et al. (2009) A hub-and-spoke circuit drives pheromone attraction and social behaviour in C. elegans. Nature 458(7242):1171–1175.
    OpenUrlCrossRefPubMedWeb of Science
  38. 38.↵
    Bretscher AJ, Busch KE, de Bono M (2008) A carbon dioxide avoidance behavior is integrated with responses to ambient oxygen and food in Caenorhabditis elegans. Proc Natl Acad Sci U S A 105(23):8044–8049.
    OpenUrlAbstract/FREE Full Text
  39. 39.↵
    Hallem EA, Sternberg PW (2008) Acute carbon dioxide avoidance in Caenorhabditis elegans. Proc Natl Acad Sci U S A 105(23):8038–8043.
    OpenUrlAbstract/FREE Full Text
  40. 40.
    Kodama-Namba E et al. (2013) Cross-modulation of homeostatic responses to temperature, oxygen and carbon dioxide in C. elegans. PLoS Genet 9(12):e1004011.
    OpenUrlCrossRefPubMed
  41. 41.↵
    Carrillo MA, Guillermin ML, Rengarajan S, Okubo RP, Hallem EA (2013) O2-Sensing Neurons Control CO2 Response in C. elegans. J Neurosci 33(23):9675–9683.
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    De Bono M, Bargmann CI (1998) Natural variation in a neuropeptide Y receptor homolog modifies social behavior and food response in C. elegans. Cell 94(5):679–689.
    OpenUrlCrossRefPubMedWeb of Science
  43. 43.↵
    Cheung BH, Arellano-Carbajal F, Rybicki I, De Bono M (2004) Soluble Guanylate Cyclases Act in Neurons Exposed to the Body Fluid to Promote C. elegans Aggregation Behavior. Curr Biol 14(12):1105–1111.
    OpenUrlCrossRefPubMedWeb of Science
  44. 44.↵
    Husson SJ et al. (2007) Impaired processing of FLP and NLP peptides in carboxypeptidase E (EGL-21)-deficient Caenorhabditis elegans as analyzed by mass spectrometry. J Neurochem 102(1):246–260.
    OpenUrlCrossRefPubMedWeb of Science
  45. 45.↵
    Fenk LA, De Bono M (2015) Environmental CO2 inhibits Caenorhabditis elegans egg-laying by modulating olfactory neurons and evokes widespread changes in neural activity. Proc Natl Acad Sci U S A 112(27):E3525–34.
    OpenUrlAbstract/FREE Full Text
  46. 46.↵
    Bretscher AJ et al. (2011) Temperature, Oxygen, and Salt-Sensing Neurons in C. elegans Are Carbon Dioxide Sensors that Control Avoidance Behavior. Neuron 69(6):1099–1113.
    OpenUrlCrossRefPubMedWeb of Science
  47. 47.↵
    Glauser DA et al. (2011) Heat avoidance is regulated by transient receptor potential (TRP) channels and a neuropeptide signaling pathway in Caenorhabditis elegans. Genetics 188(1):91–103.
    OpenUrlAbstract/FREE Full Text
  48. 48.↵
    Choi S, Chatzigeorgiou M, Taylor KP, Schafer WR, Kaplan JM (2013) Analysis of NPR-1 reveals a circuit mechanism for behavioral quiescence in C. elegans. Neuron 78(5):869–880.
    OpenUrlCrossRefPubMed
  49. 49.↵
    Jang H et al. (2012) Neuromodulatory state and sex specify alternative behaviors through antagonistic synaptic pathways in C. elegans. Neuron 75(4):585–592.
    OpenUrlCrossRefPubMedWeb of Science
  50. 50.↵
    Butcher RA, Fujita M, Schroeder FC, Clardy J (2007) Small-molecule pheromones that control dauer development in Caenorhabditis elegans. Nat Chem Biol 3(7):420–422.
    OpenUrlCrossRefPubMedWeb of Science
  51. 51.
    Butcher RA, Ragains JR, Kim E, Clardy J (2008) A potent dauer pheromone component in Caenorhabditis elegans that acts synergistically with other components. Proc Natl Acad Sci U S A 105(38):14288–14292.
    OpenUrlAbstract/FREE Full Text
  52. 52.↵
    Zhang X, Noguez JH, Zhou Y, Butcher RA (2013) Analysis of ascarosides from Caenorhabditis elegans using mass spectrometry and NMR spectroscopy. Methods Mol Biol 1068:71–92.
    OpenUrlCrossRefPubMedWeb of Science
  53. 53.↵
    De Bono M, Tobin DM, Davis MW, Avery L, Bargmann CI (2002) Social feeding in Caenorhabditis elegans is induced by neurons that detect aversive stimuli. Nature 419(6910):899–903.
    OpenUrlCrossRefPubMed
  54. 54.↵
    Jang H, Bargmann CI (2013) Acute behavioral responses to pheromones in C. elegans (adult behaviors: attraction, repulsion). Methods Mol Biol 1068:285–292.
    OpenUrl
  55. 55.↵
    Gibson WT et al. (2015) Behavioral responses to a repetitive visual threat stimulus express a persistent state of defensive arousal in Drosophila. Curr Biol 25(11):1401–1415.
    OpenUrlCrossRefPubMed
  56. 56.↵
    Anderson DJ (2016) Circuit modules linking internal states and social behaviour in flies and mice. Nat Rev Neurosci 17(11):692–704.
    OpenUrlCrossRefPubMed
  57. 57.↵
    Dacks AM, Weiss KR (2013) Latent modulation: a basis for non-disruptive promotion of two incompatible behaviors by a single network state. J Neurosci 33(9):3786–3798.
    OpenUrlAbstract/FREE Full Text
  58. 58.↵
    Marder E, O’Leary T, Shruti S (2014) Neuromodulation of circuits with variable parameters: single neurons and small circuits reveal principles of state-dependent and robust neuromodulation. Annu Rev Neurosci 37:329–346.
    OpenUrlCrossRefPubMed
  59. 59.↵
    Tyrrell T (1993) The use of hierarchies for action selection. Adaptive Behavior
  60. 60.↵
    Brenner S (1974) The genetics of Caenorhabditis elegans. Genetics 77(1):71–94.
    OpenUrlAbstract/FREE Full Text
  61. 61.↵
    Hart AC (2006) Behavior. WormBook 1–87.
  62. 62.↵
    Hilliard MA, Bargmann CI, Bazzicalupo P (2002) C. elegans responds to chemical repellents by integrating sensory inputs from the head and the tail. Curr Biol 12(9):730–734.
    OpenUrlCrossRefPubMedWeb of Science
  63. 63.↵
    Chronis N, Zimmer M, Bargmann CI (2007) Microfluidics for in vivo imaging of neuronal and behavioral activity in Caenorhabditis elegans. Nat Methods 4(9):727–731.
    OpenUrlCrossRefPubMedWeb of Science
Back to top
Posted February 10, 2017.
Download PDF
Email
A memory of recent oxygen experience switches pheromone valence in C. elegans
Lorenz A. Fenk, Mario de Bono
bioRxiv 107524; doi: https://doi.org/10.1101/107524
Reddit logo Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area