PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of elifeeLifeRecent contentAbout eLifeFor authorsSign up for alerts
 
eLife. 2017; 6: e22896.
PMCID: PMC5370180

Dopamine signaling tunes spatial pattern selectivity in C. elegans

Oliver Hobert, Reviewing editor
Oliver Hobert, Howard Hughes Medical Institute, Columbia University, United States;

Abstract

Animals with complex brains can discriminate the spatial arrangement of physical features in the environment. It is unknown whether such sensitivity to spatial patterns can be accomplished in simpler nervous systems that lack long-range sensory modalities such as vision and hearing. Here we show that the nematode Caenorhabditis elegans can discriminate spatial patterns in its surroundings, despite having a nervous system of only 302 neurons. This spatial pattern selectivity requires touch-dependent dopamine signaling, including the mechanosensory TRP-4 channel in dopaminergic neurons and the D2-like dopamine receptor DOP-3. We find that spatial pattern selectivity varies significantly among C. elegans wild isolates. Electrophysiological recordings show that natural variations in TRP-4 reduce the mechanosensitivity of dopaminergic neurons. Polymorphic substitutions in either TRP-4 or DOP-3 alter the selectivity of spatial patterns. Together, these results demonstrate an ancestral role for dopamine signaling in tuning spatial pattern preferences in a simple nervous system.

DOI: http://dx.doi.org/10.7554/eLife.22896.001

Research Organism: C. elegans

Introduction

Animals must relentlessly explore their surroundings in search of food, water, and shelter, in order to survive. Animals with complex nervous systems utilize multiple sensory modalities such as vision, touch, and hearing to perceive spatial patterns that reflect the physical structure of their environment. But can animals with much simpler nervous systems detect and discriminate spatial patterns in their environment as well? To address this question, we investigated the roundworm C. elegans, which feeds and reproduces in decomposing fruits and plant stems (Barrière and Félix, 2014). C. elegans has only 302 neurons and lack vision and hearing, mainly relying on smell and touch to sense their environment. C. elegans uses the sense of touch (or mechanosensation) to detect physical contact with external objects. For example, touch to the worm body or nose elicits reflex withdrawal responses (Chalfie and Sulston, 1981Chalfie and Wolinsky, 1990; Kaplan and Horvitz, 1993; Sulston et al., 1975; Way and Chalfie, 1989), which allows the worms to quickly avoid potential danger. Beyond reflex, mechanosensation could also provide important information about the physical structure of the environment, aiding in navigation and spatial orientation. Indeed, mechanosensory detection of food elicits a slowing in C. elegans locomotion, increasing the time spent where food is present (Sawin et al., 2000). However, it remains unknown whether C. elegans mechanosensation is limited to such binary distinctions between the presence and absence of food, or whether it enables more general and refined perception of particular spatial patterns, and if so, how this depends on context and experience. These questions are important for determining to what extent a simple animal can sense the spatial arrangement of physical features in its surroundings, providing general insights about how animals perceive spatial patterns.

Here, we report that C. elegans has the ability to discriminate spatial patterns using mechanosensory-dependent dopamine signaling. This spatial pattern selectivity is not a fixed behavior. Instead, it is modified by experience and by internal states such as starvation. Among C. elegans wild isolates, the absolute degree of spatial pattern preference is tuned by natural variations in the mechanosensory channel TRP-4 and the dopamine receptor DOP-3, leading to differential behavioral responses to identical environmental settings. These results suggest that the sense of the physical surroundings may undergo rapid evolutionary expansion to diversify the choice of preferred habitat, and perhaps to benefit the survival of the species.

Results

To study how spatial patterns in its surroundings may impact worm behavior, we examined worm locomotion in microfluidic chambers comprising four patterns of PDMS (polydimethylsiloxane) pillars, designated quad-chambers (Figure 1A and Figure 1—figure supplement 1A). Pillar density increased in the four patterns from zero pillars (Pattern I) to the highest density of pillars (Pattern IV). We found that the Bristol N2 strain exhibits a strong preference for pattern IV (Figure 1B). Approximately 63% of N2 worms were found in the section containing pattern IV 60 min after entering the chamber. By contrast, we observed only 9–15% of worms in each of the other sections I, II and III (Figure 1C) with lower pillar densities. The strong preference for pattern IV was also indicated by a high positive value of a spatial pattern preference index (Figure 1C; Methods). To investigate the robustness of this pattern selectivity, we quantified N2 preference for pattern IV at various time points (Figure 1—figure supplement 1B). Pattern IV preference was very stable between 60 and 90 min, but was reduced at 120 min. This decrease was likely due to starvation, as addition of food (OP50 bacteria) into the chamber prolonged the preference for pattern IV, which lasted at least 180 min (Figure 1—figure supplement 1C). Consistent with this notion, worms that had already been starved prior to the experiment exhibited a significant reduction in pattern IV preference (Figure 1—figure supplement 1D), indicating that feeding state (starvation) can override an inherent preference for pattern selectivity.

Figure 1.
C. elegans exhibits spatial pattern selectivity.

To test whether pattern selectivity is based on individual choice or group behavior, we performed cumulative single-worm analyses. Single-worm pattern preference index values were similar to those obtained in population experiments (Figure 1—figure supplement 2A), indicating that pattern selectivity is an individual rather than group behavior. To identify the sensory modality used for pattern selection, we analyzed mutant worms that have defects in either oxygen or touch sensation. C. elegans exhibits aggregation behavior at optimal oxygen levels (Cheung et al., 2004; Gray et al., 2004). We thus examined whether oxygen sensation might play a role in the observed clustering within pattern IV. To this end, we tested mutant worms that lack the GCY-35 soluble guanylate cyclase and are defective in oxygen-dependent aggregation, bordering, and social feeding behaviors (Cheung et al., 2004; Gray et al., 2004). However, these mutant gcy-35 worms continued to show preference for pattern IV (Figure 1D), indicating that pattern IV selectivity was independent of possible variations in oxygen levels and oxygen sensing. By contrast, mutant worms with dysfunctional MEC-10 DEG/ENaC channels showed a significant reduction in their preference for pattern IV (Figure 1E). To understand where MEC-10 functions, we performed cell-specific rescue experiments. As expected, single-copy Pmec-10::mec-10 transgenes that drive MEC-10 expression in the six touch receptor neurons and in PVD and FLP neurons fully restored the preference for pattern IV. Expression of MEC-10 under Pegl-44 in FLP and HSN neurons (Wu et al., 2001) also rescued the defects in mec-10(tm1552) mutant worms. By contrast, MEC-10 expression in the six touch receptor neurons (driven by Pmec-18) did not recover the preference for pattern IV. Together, these data indicate that MEC-10 acts in the polymodal nociceptive FLP neurons to support the selection of pattern IV.

To determine whether pattern selectivity depends on the particular absolute pillar density of pattern IV, we used a chamber consisting of patterns I and III only (Figure 1—figure supplement 2D). In this chamber, N2 worms showed a clear preference for pattern III (Figure 1—figure supplement 2E), even though this is a non-favored pattern in the quad-chamber. We therefore conclude that pattern selectivity is based on comparison between different spatial patterns i.e., environmental context, rather than on the specific structural arrangement of pattern IV.

We next investigated the strategies that worms use to select pattern IV in the quad-chamber. First, we considered local effects occurring at the boundaries between adjacent patterns. We tracked worms that left pattern IV and entered into the less desired pattern III (Figure 2A), and found that these worms re-crossed back into the preferred pattern IV at a much higher rate than worms already within pattern IV (Figure 2B). We also found an increase in both reversal and turning rates when worms cross the IV-to-III boundary (Figure 2C). Therefore, worms sense local changes in pattern density, and respond quickly to unfavorable changes by reorientation, reinforcing our finding of their ability to make decisions based on comparisons of spatial density. Second, we examined whether worm locomotion within pattern IV was affected by the presence of neighboring patterns. We hypothesized that preceding experience of alternative spatial configurations would have a lasting impact on worm behavior within preferred pattern IV. To test this hypothesis, we designed an additional chamber (designated uni-chamber; Figure 2D) containing only a pattern IV pillar configuration throughout its surface and no other patterns. We compared the speed and pause rates of worms within pattern IV in both quad- and uni- chambers. We found that worms in pattern IV of the quad-chamber moved more slowly and paused more than worms in the uni-chamber (Figure 2E–F). Such slowing and pausing could prolong the stay of worms in pattern IV. Together, these data suggest that the clustering of worms in pattern IV is due to both real-time responses to pattern changes upon boundary crossing and to longer-lasting effects caused by the experience of other less desired patterns. Thus, without vision or hearing, C. elegans can take advantage of mechanosensory information to compare and select between spatial patterns in its environment.

Figure 2.
Boundary experience impacts C. elegans spatial pattern selectivity.

We next set to explore the molecular basis for the observed spatial pattern selectivity. Several features of this behavior prompted us to investigate the contribution of dopamine signaling. In C. elegans, dopamine regulates a variety of behaviors including locomotion, mechanosensation, food availability, touch habituation, and the swim-to-crawl transition (Calhoun et al., 2015; Chase et al., 2004; Hills et al., 2004; Kindt et al., 2007; Sanyal et al., 2004; Sawin et al., 2000; Vidal-Gadea et al., 2011). In particular, dopamine signals decrease locomotion rates when worms enter a food source (a lawn of bacteria) (Sawin et al., 2000), and increase returning frequency when worms leave the bacteria lawn (Hills et al., 2004). To test whether dopamine is required also for spatial pattern selectivity, we examined cat-2(e1112) animals, which lack the CAT-2 tyrosine hydroxylase and are defective in dopamine biosynthesis (Lints and Emmons, 1999). We observed that cat-2 mutant animals have disrupted pattern selectivity, as they showed reduced preference for pattern IV in the quad-chambers (Figure 3A). Addition of exogenous dopamine restored the accumulation of cat-2 mutants within pattern IV (Figure 3A), demonstrating that the defect in spatial pattern selectivity of cat-2 mutants is due to the lack of dopamine. To determine the dopamine receptor responsible for the selectivity, we analyzed mutant worms with impaired dopamine receptors. We found that disrupting the D2-like receptor DOP-3, but not other dopamine receptors (DOP-1, DOP-2, and DOP-4), led to a reduced preference for pattern IV (Figure 3B). DOP-3 is expressed in a variety of neurons in the head, ventral cord and tail (Chase et al., 2004). The site-of-action of DOP-3 for pattern selection is unknown.

Figure 3.
Spatial pattern selectivity of C. elegans requires dopamine signaling.

To identify the molecular sensor that links mechanosensation, dopamine, and spatial pattern preference, we studied mutant worms that lack TRP-4, a stretch receptor expressed in the CEP, ADE, and PDE dopaminergic neurons and the DVA and DVC interneurons (Hu et al., 2011; Kang et al., 2010; Li et al., 2006). TRP-4 is the pore-forming subunit of mechanotransduction channels (Kang et al., 2010; Li et al., 2006). Mutant trp-4(ok1605) worms significantly reduced their preference for pattern IV (Figure 3C) and did not show the enhanced returning behavior after crossing the boundary between section IV and III (Figure 3D). To confirm that the defects in pattern selection are due to the lack of TRP-4 function, we generated a second allele trp-4(pek220) using CRISPR/CAS9 and homologous recombination (Figure 3—figure supplement 1). We found that trp-4(pek220) mutant worms exhibit similar defects as trp-4(ok1605) worms, confirming that the pattern IV preference requires TRP-4. The defects of trp-4 single mutants were similar to trp-4; cat-2 and trp-4; dop-3 double mutants, indicating that trp-4 acts in the same dopamine pathway as cat-2 and dop-3 (Figure 3E). Single-copy transgenes that express TRP-4 in dopaminergic neurons restored the preference for pattern IV (Figure 3C), further demonstrating a role of TRP-4 in dopaminergic neurons in mediating spatial pattern selectivity. Together, these data elucidate a signaling mechanism for the stretch receptor TRP-4, dopamine, and the dopamine receptor DOP-3, coupling spatial pattern detection with the behavioral preference of animals.

Recent studies have shown that natural genetic variations influence how C. elegans interacts with environmental cues such as food, oxygen, microbial pathogens, and neighboring worms (Bendesky and Bargmann, 2011; Bendesky et al., 2012, 2011; Chang et al., 2011; Persson et al., 2009). The physical structure of the environment is another important feature that varies with habitat. We therefore hypothesized that C. elegans strains from different natural habitats might differ in their selectivity of spatial patterns. We thus examined a wild isolate of C. elegans, Hawaiian CB4856 (designated HW) in the spatial pattern preference assay. Interestingly, we found that unlike N2, HW worms failed to aggregate in pattern IV of the quad-chambers (Figure 4A, B), and their returning behavior at boundaries was significantly reduced in comparison to N2 worms (Figure 4C). In addition, HW worms exhibited similar traveling speed and pause rates in uni- and quad-chambers (Figure 4—figure supplement 1), indicating that mechanosensory experience had little impact on HW behavior. Several studies have shown that behavioral differences between N2 and HW could be explained by laboratory-derived genetic changes (e.g., allelic variations in npr-1 and glb-5; reviewed in Sterken et al. (2015). To test whether this is the case for spatial pattern selectivity, we examined the LSJ1 strain, a sibling of N2 that was cultivated separately after isolation. We found that the LSJ1 strain showed similar behavior to N2 worms in the spatial pattern preference assay (Figure 4B), suggesting that the difference in spatial pattern selectivity is not due to domestication-derived alleles.

Figure 4.
Hawaiian strain does not select pattern IV.

Interestingly, the stretch receptor TRP-4 in the HW strain carries three amino acid substitutions (M36I, G41S, and E67G; Figure 5A) (Thompson et al., 2013). The trp-4 gene is located on chromosome I in C. elegans. An N2/HW hybrid strain that carries the HW Chromosome I in an N2 background behaved like the HW strain in the spatial pattern preference assay. However, hybrid strains that carry other HW chromosomes in the N2 background were indistinguishable from N2 worms (Figure 5B), with the exception of those with HW Chromosome X, which showed a minor but significant reduction in the preference for pattern IV. These data are consistent with a role of trp-4 in determining the natural variation in spatial pattern selectivity. They also suggest that other gene(s) on Chromosome X may contribute to the polymorphic behavior. Next, we isolated recombinant inbred lines (RIL) from a cross between N2 and N2/HW hybrids that carry the HW Chromosome I. As shown in Figure 5C, two RIL strains behaved differently in the spatial pattern preference assay; one that carries the N2 trp-4 gene showed preference for pattern IV, but another one with HW trp-4 did not. To confirm the role of TRP-4 in spatial preference polymorphism, we performed rescue experiments using single-copy transgenes expressing either N2 or HW trp-4 cDNA in dopaminergic neurons. N2 trp-4 cDNA fully restored the selectivity defects of the trp-4(ok1605) mutant worms. By contrast, HW trp-4 cDNA showed only low levels of rescue (Figure 5D). These data confirm that N2 TRP-4 is required for the spatial pattern preference observed in the Bristol strains. To test whether N2 TRP-4 is sufficient for endowing the HW strains with spatial pattern selectivity, we overexpressed N2 TRP-4 in HW worms. Because TRP channels are expected to form tetramers, we reasoned that overexpression of N2 TRP-4 might replace the HW TRP-4 channels. Indeed, we found that N2 TRP-4 but not HW TRP-4 overexpression caused HW worms to show spatial pattern selectivity similar to that of N2 (Figure 5E).

Figure 5.
Polymorphisms in the mechanotransduction TRP-4 channel contribute to the variation of spatial pattern selectivity.

The substitution of three residues in HW TRP-4 occurs at the N-terminal region that is adjacent to Ankyrin repeats (Figure 5A). In flies, the Ankyrin repeats of NOMPC (the TRP-4 homologue) directly bind microtubules, and they are essential for mechanogating of the NOMPC channels (Zhang et al., 2015). To understand whether the three residue substitutions alter TRP-4 properties, we recorded electrophysiological responses of the dopaminergic CEP neurons to small mechanical displacements (Figure 6A) (Kang et al., 2010). The mechanoreceptor current in CEP neurons is mediated by TRP-4, and the amplitude of the current depends on the strength of the applied mechanical force. In agreement with a previous study (Kang et al., 2010), a displacement of 0.5 μm was sufficient to elicit TRP-4 mediated current in N2 worms, and the current was saturated at 3 μm displacement (Figure 6B). Interestingly, the sensitivity of HW TRP-4 to mechanical stimuli was significantly reduced. Larger mechanical displacement was required to generate currents with similar amplitudes (Figure 6A–B). These results are consistent with the findings that the N-terminal region of the TRP-4 homologue NOMPC is required for mechanogating in flies (Zhang et al., 2015). It is worth noting that, when mechanical stimuli are sufficiently strong, the HW TRP-4 is fully active (Figure 6C). Consistently, we found that an even higher density of pillars (novel pattern V) enhanced the enrichment of HW worms, but not of trp-4(ok1605) mutants (Figure 6D). Together, these findings indicate that amino acid substitutions in HW TRP-4 reduce the sensitivity to mechanical stimuli, which could subsequently alter spatial pattern perception in HW worms. Interestingly, whereas a single-copy HW TRP-4 transgene failed to restore spatial preference (Figure 5E), it completely rescued swimming/crawling transition in the trp-4(ok1605) mutant worms (Figure 5—figure supplement 1), suggesting that various types of TRP-4 dependent behaviors are tuned to different levels of mechanogating sensitivity.

Figure 6.
Polymorphisms in TRP-4 alter mechanosensitivity of dopaminergic neurons.

Finally, we analyzed spatial pattern selectivity of ten additional wild isolates, for which complete genome sequences are available (Thompson et al., 2013). We found that five wild isolates exhibited a spatial preference like N2. Two isolates (AB1 and JU397) with substituted residues in TRP-4 exhibited defects in preference for pattern IV in the quad-chambers (Figure 7A). The TRP-4 channels in AB1 and JU397 each carry two substitutions (G41S and E67G), suggesting that the third substitution M36I in the HW strain is not required. A third isolate (CB4854) also carries two substitutions (E67G and I152M) in TRP-4; however, these worms did not show significant defects in the spatial pattern preference assay (Figure 7A). The most parsimonious explanation is that the G41S substitution is the key residue that modulates TRP-4 mechanosensitivity. However, it is also possible that E67G is important, but the combination of E67G and I152M suppresses the impact of E67G.

Figure 7.
Natural variations in either TRP-4 or DOP-3 tune spatial pattern selectivity in C. elegans.

Intriguingly, three wild isolates (CB4853, JU300, and JU322) showed altered spatial pattern selectivity despite having N2-like TRP-4 sequence (Figure 7A). Closer examination of their genome sequences revealed that all three strains carry a substituted residue in DOP-3 at the same position (A273G). Upon editing the dop-3 gene in N2 worms using CRISPR/Cas9 (Friedland et al., 2013), we found that the edited worms with the A273G substitution lost their spatial pattern selectivity (Figure 7B). These experiments indicate that the DOP-3 A273G mutation is sufficient to impair spatial pattern preference even with fully functional TRP-4. Therefore, the diversity of spatial pattern selectivity may arise from variations in either the detector that senses changes in the external world (e.g., TRP-4), or the neuromodulator signals that define the internal state of animals (e.g., dopamine and its receptors).

Discussion

These results demonstrate that C. elegans is able to utilize touch sensation to discern different spatial patterns in the environment. Previous studies have shown that nematodes can evaluate their environment by sensing food (Calhoun et al., 2015) and chemicals (Albrecht and Bargmann, 2011). Our findings add a new dimension to the ‘mind’ of the worm by demonstrating a capability to perceive physical patterns in the environment. In many animal species, touch perception plays a key role in guiding exploration of the physical world. For example, insects carry antennae to collect tactile information during locomotion, rodents use whiskers to detect spatial settings, and sea mammals use facial hairs to trace changes in water flow patterns caused by swimming fish. Using a substantially simpler nervous system, C. elegans is able to detect spatial patterns, and to use mechanosensory information about the overall environment to form spatial pattern preferences. While primitive, this sense of physical space in worms is highly dynamic – it is tightly regulated by experience (Figure 2E,F), spatial context (Figure 1—figure supplement 2D,E), and food availability (Figure 1—figure supplement 1C,D). Together, our findings elucidate a basic but surprising sense of physical patterns in the surrounding environment in a simple nervous system of 302 neurons.

The selectivity of spatial patterns in C. elegans requires dopamine, which may serve as a ‘reward’ signal that connects the sensation of spatial patterns with innate preferences (Schultz et al., 1997). In higher animals, the existence of sensory modalities such as vision and hearing and associated brain structures, substantially expand the richness and quality of available information about the physical structure of their environment. Nevertheless, together with previous findings implicating dopamine in spatial attention (Bellgrove et al., 2007) and spatial memory (Murphy et al., 1996), as well as in disorders such as agoraphobia in humans (Perna et al., 2011), our study suggests that the coupling between dopamine signaling and a sense of spatial features is an ancestral feature of animal nervous systems.

Materials and methods

Nematode strains and growth

C. elegans strains were maintained under standard conditions at 20°C on nematode growth medium (NGM) agar plates (Brenner, 1974). These plates were seeded with E. coli OP50 lawns. The following strains were used in this study: (a) Wild type strains – AB1 (Adelaide, AU), CB4853 (Altadena, US), CB4854 (Altadena, USA), CB4856 (Hawaii, US), LSJ1 (Bristol, UK), JU258 (Madeira, PT), JU300 (Indre, FR), JU322 (Merlet, FR), JU397 (Hermanville, FR), MY1 (Emsland, DE), MY14 (Mecklenbeck, DE), and N2 (Bristol, UK); (b) Mutant strains – VC1141 trp-4(ok1605), RB1052 trpa-1(ok999), CB1112 cat-2(e1112), LX645 dop-1(vs100), LX702 dop-2(vs105), LX703 dop-3(vs106), FG58 dop-4(tm1392), BJH2033 trp-4(ok1605); dop-3(vs106); BJH2034 trp-4(ok1605); cat-2(e1112); CX6448 gcy-35(ok769); RB1115 mec-10(ok1104); ZB2551 mec-10(tm1552); (c) Chromosome substitution strains – WE5236 [I, CB4856 >N2], GN67 [II, CB4856 >N2], GN68 [III, CB4856 >N2], GN69 [IV, CB4856 >N2], GN70 [V, CB4856 >N2], GN231 [X, CB4856 >N2]; (d) Single-copy MosSCI strains – BJH189 trp-4(ok1605); pekSi8 [cb-unc-119(+),Pdat-1::trp-4N2]II; BJH190 trp-4(ok1605); pekSi9[cb-unc-119(+),Pdat-1::trp-4HW]II; BJH976 mec-10(tm1552); pekSi102[NeoR(+),Pegl-44::mec-10]; BJH977 mec-10(tm1552); pekSi103[NeoR(+),Pmec-18::mec-10]; BJH978 mec-10(tm1552); pekSi104[NeoR(+),Pmec-10::mec-10]; (e) Strains with engineered missense mutation by CRISPR-Cas9 – BJH701 dop-3(pek201[A273G]); BJH720 trp-4(pek220[GFP+hygR]); (f) Strains with UV-integrated transgenes – BJH186 pekIs85 [Pdat-1::trp-4N2] in CB4856, BJH187 pekIs86 [Pdat-1::HW trp-4HW] in CB4856.

Transgenes

Single copy transgenes were generated using the MosSCI technique (Frøkjær-Jensen et al., 2012; Frøkjaer-Jensen et al., 2008). Transgenes were inserted into the ttTi5605 Mos1 site on chromosome II. Multi-copy integrated transgenic strains for trp-4 overexpression were generated by microinjection of plasmids (BJP-B289: Pdat-1::trp-4N2 and BJP-B290: Pdat-1::trp-4HW) at 50 ng µl−1 into the HW CB4856 strain. The co-injection markers were Pvha-6::mRuby (BJP-B169) and Pvha-6::GFP (BJP-B197). Blank vector pBluescript was used as an injection filler to bring final DNA concentration to 100 ng µl−1. Integration was obtained by UV irradiation of worms carrying the extrachromosomal arrays. Transgenic worms were outcrossed four times.

CRISPR/CAS9

To construct the Cas9-sgRNA expression plasmid for making the trp-4 deletion allele pek220, we replaced the gRNA sequence in the plasmid pJW1219 (Addgene; [Ward, 2015]) with the following sequence (5′-GTAAATGTCCCAGAACGCGG-3′) using the overlapping PCR method. To construct the repair construct for homologous recombination, we inserted the left trp-4 recombination arm, a disruption cassette containing GFP::tbb-2utr + Prps-0::hygR::unc-54utr, and the right trp-4 recombination arm into pPD49.26 between SpeI and SalI and sites. The left trp-4 recombination arm (~710 bp) was amplified using the forward primer 5’-CTTGCATGCCTGCAGGTCGACTTTGGAATTTCCTTTGATACAGCAACCCAC-3’ and the reverse primer 5’-GTTCTTCTCCTTTACTGGAGGTGTGGGGATTCATCTGG-3’. The right trp-4 recombination arm (~510 bp) was amplified using the forward primer 5’-GGCTCCTTGCGTTCATCTTCATGCACACGGCACATAGGACAGAG-3’ and the reverse primer 5’-GGGCCCGTACGGCCGACTAGTCTTGGGAGTTTCCTCCCTTTTCC-3’.

Fabrication of PDMS microfluidic chambers

Microfluidic chambers were fabricated using soft lithography with polydimethyl siloxane (PDMS, Sylgard 184, Dow Corning, MI). Silicon wafers were patterned with various designs using SU8-100 (Stanford microfluidic foundry, CA) and photolithography. PDMS was poured on top of the patterned silicon wafer, and was cured at 80°C for 2 hr. Cured PDMS was carefully lifted away from the silicon wafer. Sharpened stub needles (0.070 mm OD, 22G) were used to create the fluid inlet and outlet paths on the PDMS device. A plasma cleaner PDC-32G (Harrick Scientific, NY) was used to treat the PDMS device for 1 min. Glass slides (1 mm thick) were immediately bound to the devices after treating. Distilled water was added to the devices within 5 min. Before loading the worms, the devices were washed with an ample amount of M9 buffer.

Behavior assay for spatial pattern preference

Worms were synchronized for the behavioral assays. Fifty gravid adult worms were transferred onto a 10 cm NGM plate seeded with OP50 to lay eggs for an hour. Then adult worms were removed by picking. Eggs were grown at 20°C for 72 hr to reach the young adult stage for behavioral tests. The PDMS chambers were washed with M9 buffer. Synchronized worms (72 hr) were transferred using glass Pasteur pipettes into Eppendorf tubes. After settling by gravity, worms were washed once to remove residual bacteria OP50. Approximately 30–70 worms were loaded into the inlet, and were allowed to freely explore the chamber. The location of the worms was recorded for 10 s using a Moticam 2500 CCD camera mounted on a stereo microscope (Leica, Germany) at various time points as indicated in the study.

Worm preference for each pattern was quantified as the percentage of worms occupying that pattern in the quad-chambers and, in addition, using a spatial pattern preference index. The index was calculated for each pattern using the following formula: [# of worms in one pattern – (# of worms in the other three patterns/3)] ÷ total # of worms. An index value of 0 indicates in all patterns equal distribution among the different patterns. An index of −0.33 for a certain pattern indicates no worms found in that pattern. An index of 1 for a certain pattern indicates that all worms are present in that pattern. Each experiment was replicated 3–5 times. ‘n’ indicates the number of independent repeats of a particular experiment. Results were analyzed in a blind manner. To analyze the speed of locomotion and the pause rate, worm behavior in quad-chambers (between 58 to 61 min) was recorded using a Moticam 2500 digital camera that was mounted on a Leica S6D microscope. Videos were processed offline using MATLAB scripts developed by Albrecht and Bargmann (2011). The animal’s locomotory state (forward, reverse, and turn) at boundaries was determined by visual inspection of the video.

Statistical analysis

The following statistic methods were used to analyze behavioral data. For single pairwise comparisons, we used the Student's t test. For analyses that require comparisons between multiple means to a control mean, we used one-way ANOVA followed by Dunnett's multiple comparison tests. For analyses that require comparisons among all possible pairs of means, we used the Tukey-Kramer method. Statistical analysis was performed using Igor Pro 6. Mean and standard error of the mean (SEM) are reported. Results were considered statistically significant when the p-value was <0.05.

Electrophysiological recording

Young adult worms were immobilized on sylgard-coated coverslips using a cyanoacrylate glue Histoacryl Blue (Aesculap, PA). Worm dissection was performed in extracellular solution (Dong et al., 2015). To expose the cell body of the CEP neuron, a small piece of the worm cuticle was cut open as previously described (Kang et al., 2010). Dissected animals were mounted onto a fixed stage upright fluorescence BX51WI microscope (Olympus, Japan) equipped with a 60x/1.0NA water immersion objective lens, Nomarski-DIC optics, an X-cite 200DC light source (Excelitas, MA), and a ORCA-Flash2.8 CMOS camera (Hamamatsu Corp., Japan). All worms carried a Pdat-1::GFP transgene (Lints and Emmons, 1999) enabling CEP identification by green fluorescence.

Recording pipettes were pulled from borosilicate glass capillaries using a P-97 micropipette puller (Sutter Instruments, CA). These pipettes had resistances of ~15 MΩ when filled with internal solution (135 CH3O3SCs, 5 CsCl, 5 MgCl2, 5 EGTA, 0.25 CaCl2, 10 HEPES and 5 Na2ATP, adjusted to pH 7.2 using CsOH). Extracellular solution contained (in mM) 150 NaCl, 5 KCl, 1 CaCl2, 5 MgCl2, 10 glucose and 10 HEPES, titrated to pH 7.3 with NaOH, 330 mOsm with sucrose. All chemicals were purchased from Sigma.

Whole-cell patch clamp recordings (holding potential at −75 mV) were carried out by using HEKA EPC-10 double amplifier controlled by HEKA Patchmaster software at 20°C. The junction potential was measured by ±10 mv 5 ms square testing signals. After the pipette tip was inserted into the bath solution, the junction potential was corrected using the ‘Auto V0’ function, which calls a procedure for automatic zeroing of the pipette current. Just before the pipette touched the cell membrane for the seal, the ‘Auto V0’ was used again to make sure the junction potential was well corrected. The currents were amplified and acquired using an EPC-10 amplifier (HEKA, Germany). The signal was digitized at 20 kHz and filtered at 2 kHz using Patchmaster (HEKA, Germany). A glass stimulus probe was mounted to a piezo-driven actuator (Physik Instrumente, Germany) held by a micromanipulator. Displacement was triggered and controlled by the piezo amplifier E625 (PI), which is synchronized with programmed signals from HEKA Patchmaster software. A voltage step of 222 mV was used to trigger each 1 µm displacement of the probe. Series resistance was compensated to 70%. Electrophysiological data were analyzed using Igor Pro 6 (Wavemetrics, OR).

Acknowledgements

We thank Drs. Michael Ailion, Harmit Malik, Irini Topalidou, and members of the Bai lab for comments. We thank the C. elegans Genetics Stock Center, Addgene, Dr. William Schafer, Dr. Irini Topalidou, and Dr. Erik Andersen for strains and reagents. We thank Ms. Shangjun Zhao for technical assistance. This work was supported by funds from the National Institutes of Health (R01-NS085214), Hartwell Innovation Fund, and FHCRC New Development Grant to JB.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Funding Information

This paper was supported by the following grants:

  • http://dx.doi.org/10.13039/100000052NIH Office of the Director R01-NS085214 to Jihong Bai.
  • http://dx.doi.org/10.13039/100006792Hartwell Foundation 2015 Hartwell Innovation Pilot Funds to Jihong Bai.

Additional information

Competing interests

The authors declare that no competing interests exist.

Author contributions

BH, Conceptualization, Data curation, Formal analysis, Investigation, Writing—original draft, Designed and carried out experiments, collected the data and assisted with preparing the manuscript.

YD, Data curation, Validation, Investigation, Designed and carried out electrophysiology experiments, Collected and analyzed the data.

LZ, Validation, Methodology, Designed and generated transgenic C. elegans strains, Assisted with behavioral testing and validation.

YL, Assisted with behavioral testing, data collection, data analysis, and validation.

IR, Writing—original draft, Writing—review and editing, Assisted with preparing and revising the manuscript.

JB, Conceptualization, Supervision, Funding acquisition, Investigation, Methodology, Writing—original draft, Writing—review and editing, Prepare and revise the manuscript, designed all experiments, and supervised the project.

References

  • Albrecht DR, Bargmann CI. High-content behavioral analysis of Caenorhabditis elegans in precise spatiotemporal chemical environments. Nature Methods. 2011;8:599–605. doi: 10.1038/nmeth.1630. [PMC free article] [PubMed] [Cross Ref]
  • Barrière A, Félix MA. Isolation of C. elegans and related Nematodes. WormBook. 2014:1–19. doi: 10.1895/wormbook.1.115.2. [PubMed] [Cross Ref]
  • Bellgrove MA, Chambers CD, Johnson KA, Daibhis A, Daly M, Hawi Z, Lambert D, Gill M, Robertson IH. Dopaminergic genotype biases spatial attention in healthy children. Molecular Psychiatry. 2007;12:786–792. doi: 10.1038/sj.mp.4002022. [PubMed] [Cross Ref]
  • Bendesky A, Bargmann CI. Genetic contributions to behavioural diversity at the gene-environment interface. Nature Reviews Genetics. 2011;12:809–820. doi: 10.1038/nrg3065. [PubMed] [Cross Ref]
  • Bendesky A, Pitts J, Rockman MV, Chen WC, Tan MW, Kruglyak L, Bargmann CI. Long-range regulatory polymorphisms affecting a GABA receptor constitute a quantitative trait locus (QTL) for social behavior in Caenorhabditis elegans. PLoS Genetics. 2012;8:e1003157 doi: 10.1371/journal.pgen.1003157. [PMC free article] [PubMed] [Cross Ref]
  • Bendesky A, Tsunozaki M, Rockman MV, Kruglyak L, Bargmann CI. Catecholamine receptor polymorphisms affect decision-making in C. elegans. Nature. 2011;472:313–318. doi: 10.1038/nature09821. [PMC free article] [PubMed] [Cross Ref]
  • Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974;77:71–94. [PubMed]
  • Calhoun AJ, Tong A, Pokala N, Fitzpatrick JA, Sharpee TO, Chalasani SH. Neural mechanisms for evaluating environmental variability in Caenorhabditis elegans. Neuron. 2015;86:428–441. doi: 10.1016/j.neuron.2015.03.026. [PMC free article] [PubMed] [Cross Ref]
  • Chalfie M, Sulston J. Developmental genetics of the mechanosensory neurons of Caenorhabditis elegans. Developmental Biology. 1981;82:358–370. doi: 10.1016/0012-1606(81)90459-0. [PubMed] [Cross Ref]
  • Chalfie M, Wolinsky E. The identification and suppression of inherited neurodegeneration in Caenorhabditis elegans. Nature. 1990;345:410–416. doi: 10.1038/345410a0. [PubMed] [Cross Ref]
  • Chang HC, Paek J, Kim DH. Natural polymorphisms in C. elegans HECW-1 E3 ligase affect pathogen avoidance behaviour. Nature. 2011;480:525–529. doi: 10.1038/nature10643. [PMC free article] [PubMed] [Cross Ref]
  • Chase DL, Pepper JS, Koelle MR. Mechanism of extrasynaptic dopamine signaling in Caenorhabditis elegans. Nature Neuroscience. 2004;7:1096–1103. doi: 10.1038/nn1316. [PubMed] [Cross Ref]
  • Cheung BH, Arellano-Carbajal F, Rybicki I, de Bono M. Soluble guanylate cyclases act in neurons exposed to the body fluid to promote C. elegans aggregation behavior. Current Biology. 2004;14:1105–1111. doi: 10.1016/j.cub.2004.06.027. [PubMed] [Cross Ref]
  • Dong Y, Gou Y, Li Y, Liu Y, Bai J. Synaptojanin cooperates in vivo with endophilin through an unexpected mechanism. eLife. 2015;4:e05660 doi: 10.7554/eLife.05660. [PMC free article] [PubMed] [Cross Ref]
  • Frøkjær-Jensen C, Davis MW, Ailion M, Jorgensen EM. Improved Mos1-mediated transgenesis in C. elegans. Nature Methods. 2012;9:117–118. doi: 10.1038/nmeth.1865. [PMC free article] [PubMed] [Cross Ref]
  • Frøkjaer-Jensen C, Davis MW, Hopkins CE, Newman BJ, Thummel JM, Olesen SP, Grunnet M, Jorgensen EM. Single-copy insertion of transgenes in Caenorhabditis elegans. Nature Genetics. 2008;40:1375–1383. doi: 10.1038/ng.248. [PMC free article] [PubMed] [Cross Ref]
  • Friedland AE, Tzur YB, Esvelt KM, Colaiácovo MP, Church GM, Calarco JA. Heritable genome editing in C. elegans via a CRISPR-Cas9 system. Nature Methods. 2013;10:741–743. doi: 10.1038/nmeth.2532. [PMC free article] [PubMed] [Cross Ref]
  • Gray JM, Karow DS, Lu H, Chang AJ, Chang JS, Ellis RE, Marletta MA, Bargmann CI. Oxygen sensation and social feeding mediated by a C. elegans guanylate cyclase homologue. Nature. 2004;430:317–322. doi: 10.1038/nature02714. [PubMed] [Cross Ref]
  • Hills T, Brockie PJ, Maricq AV. Dopamine and glutamate control area-restricted search behavior in Caenorhabditis elegans. Journal of Neuroscience. 2004;24:1217–1225. doi: 10.1523/JNEUROSCI.1569-03.2004. [PubMed] [Cross Ref]
  • Hu Z, Pym EC, Babu K, Vashlishan Murray AB, Kaplan JM. A neuropeptide-mediated stretch response links muscle contraction to changes in neurotransmitter release. Neuron. 2011;71:92–102. doi: 10.1016/j.neuron.2011.04.021. [PMC free article] [PubMed] [Cross Ref]
  • Kang L, Gao J, Schafer WR, Xie Z, Xu XZ. C. elegans TRP family protein TRP-4 is a pore-forming subunit of a native mechanotransduction channel. Neuron. 2010;67:381–391. doi: 10.1016/j.neuron.2010.06.032. [PMC free article] [PubMed] [Cross Ref]
  • Kaplan JM, Horvitz HR. A dual mechanosensory and chemosensory neuron in Caenorhabditis elegans. PNAS. 1993;90:2227–2231. doi: 10.1073/pnas.90.6.2227. [PubMed] [Cross Ref]
  • Kindt KS, Quast KB, Giles AC, De S, Hendrey D, Nicastro I, Rankin CH, Schafer WR. Dopamine mediates context-dependent modulation of sensory plasticity in C. elegans. Neuron. 2007;55:662–676. doi: 10.1016/j.neuron.2007.07.023. [PubMed] [Cross Ref]
  • Li W, Feng Z, Sternberg PW, Xu XZ. A C. elegans stretch receptor neuron revealed by a mechanosensitive TRP channel homologue. Nature. 2006;440:684–687. doi: 10.1038/nature04538. [PMC free article] [PubMed] [Cross Ref]
  • Lints R, Emmons SW. Patterning of dopaminergic neurotransmitter identity among Caenorhabditis elegans ray sensory neurons by a TGFbeta family signaling pathway and a hox gene. Development. 1999;126:5819–5831. [PubMed]
  • Murphy BL, Arnsten AF, Jentsch JD, Roth RH. Dopamine and spatial working memory in rats and monkeys: pharmacological reversal of stress-induced impairment. Journal of Neuroscience. 1996;16:7768–7775. [PubMed]
  • Perna G, Daccò S, Menotti R, Caldirola D. Antianxiety medications for the treatment of complex agoraphobia: pharmacological interventions for a behavioral condition. Neuropsychiatric Disease and Treatment. 2011;7:621–637. doi: 10.2147/NDT.S12979. [PMC free article] [PubMed] [Cross Ref]
  • Persson A, Gross E, Laurent P, Busch KE, Bretes H, de Bono M. Natural variation in a neural globin tunes oxygen sensing in wild Caenorhabditis elegans. Nature. 2009;458:1030–1033. doi: 10.1038/nature07820. [PubMed] [Cross Ref]
  • Sanyal S, Wintle RF, Kindt KS, Nuttley WM, Arvan R, Fitzmaurice P, Bigras E, Merz DC, Hébert TE, van der Kooy D, Schafer WR, Culotti JG, Van Tol HH. Dopamine modulates the plasticity of mechanosensory responses in Caenorhabditis elegans. The EMBO Journal. 2004;23:473–482. doi: 10.1038/sj.emboj.7600057. [PubMed] [Cross Ref]
  • Sawin ER, Ranganathan R, Horvitz HR. C. elegans locomotory rate is modulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway. Neuron. 2000;26:619–631. doi: 10.1016/S0896-6273(00)81199-X. [PubMed] [Cross Ref]
  • Schultz W, Dayan P, Montague PR. A neural substrate of prediction and reward. Science. 1997;275:1593–1599. doi: 10.1126/science.275.5306.1593. [PubMed] [Cross Ref]
  • Sterken MG, Snoek LB, Kammenga JE, Andersen EC. The laboratory domestication of Caenorhabditis elegans. Trends in Genetics. 2015;31:224–231. doi: 10.1016/j.tig.2015.02.009. [PMC free article] [PubMed] [Cross Ref]
  • Sulston J, Dew M, Brenner S. Dopaminergic neurons in the nematode Caenorhabditis elegans. The Journal of Comparative Neurology. 1975;163:215–226. doi: 10.1002/cne.901630207. [PubMed] [Cross Ref]
  • Thompson O, Edgley M, Strasbourger P, Flibotte S, Ewing B, Adair R, Au V, Chaudhry I, Fernando L, Hutter H, Kieffer A, Lau J, Lee N, Miller A, Raymant G, Shen B, Shendure J, Taylor J, Turner EH, Hillier LW, Moerman DG, Waterston RH. The million mutation project: a new approach to genetics in Caenorhabditis elegans. Genome Research. 2013;23:1749–1762. doi: 10.1101/gr.157651.113. [PubMed] [Cross Ref]
  • Vidal-Gadea A, Topper S, Young L, Crisp A, Kressin L, Elbel E, Maples T, Brauner M, Erbguth K, Axelrod A, Gottschalk A, Siegel D, Pierce-Shimomura JT. Caenorhabditis elegans selects distinct crawling and swimming gaits via dopamine and serotonin. PNAS. 2011;108:17504–17509. doi: 10.1073/pnas.1108673108. [PubMed] [Cross Ref]
  • Ward JD. Rapid and precise engineering of the Caenorhabditis elegans genome with lethal mutation co-conversion and inactivation of NHEJ repair. Genetics. 2015;199:363–377. doi: 10.1534/genetics.114.172361. [PMC free article] [PubMed] [Cross Ref]
  • Way JC, Chalfie M. The mec-3 gene of Caenorhabditis elegans requires its own product for maintained expression and is expressed in three neuronal cell types. Genes & Development. 1989;3:1823–1833. doi: 10.1101/gad.3.12a.1823. [PubMed] [Cross Ref]
  • Wu J, Duggan A, Chalfie M. Inhibition of touch cell fate by egl-44 and egl-46 in C. elegans. Genes & Development. 2001;15:789–802. doi: 10.1101/gad.857401. [PubMed] [Cross Ref]
  • Zhang W, Cheng LE, Kittelmann M, Li J, Petkovic M, Cheng T, Jin P, Guo Z, Göpfert MC, Jan LY, Jan YN. Ankyrin repeats convey force to gate the NOMPC mechanotransduction channel. Cell. 2015;162:1391–1403. doi: 10.1016/j.cell.2015.08.024. [PMC free article] [PubMed] [Cross Ref]
2017; 6: e22896.

Decision letter

Oliver Hobert, Reviewing editor
Oliver Hobert, Howard Hughes Medical Institute, Columbia University, United States;

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "Dopamine Signaling Tunes Spatial Pattern Selectivity in C. elegans" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Eve Marder as the Senior Editor. The reviewers have opted to remain anonymous.

The comments of the two reviewers can be found below. You will see that the reviewers have convergent comments that I request you to address. In essence these are the two sets of key experiments that we would want to see in a revised version:

1) Validate their results with (null) alleles of mec-10 and trp-4.

2) Rescue of the mec-10 phenotype to determine its currently very unclear focus of action (determination of focus of action of dop-3 does not need to be performed, but possible site of action in VNC MNs should be discussed).

Reviewer #1:

In this paper, the authors demonstrate a role for dopaminergic neurons in C. elegans in a behavior though which worms preferentially accumulate in environments with particular mechanosensory characteristics. Specifically, they use specially fabricated microfluidic environments to show that worms prefer regions with a high density of mechanical obstacles, and that this behavior requires the mechanosensory trp channel trp-4 and the dopamine receptor dop-3. They also show that natural variation in these two genes contributes to variability in the observed behavior; in the case of trp-4, they go on to show that the genetic variants have altered sensitivity of trp-4-mediated mechanotransdution.

Overall, this is an interesting story that brings together a variety of experimental threads to provide a convincingly simple and straightforward explanation of a seemingly complex behavior. I do think there are a few small missing pieces which if added to the study would make the molecular and cellular explanation of the behavior significantly more complete.

1) In the manuscript, the authors state that "worms sense localchanges in pattern density, and respond quickly to unfavorable changes by reorientation," (Results). However, I don't think the authors present any direct evidence that reorientation (i.e. turning/reversal rates) change when crossing from higher to lower obstacle density. In principle, could the accumulation in more dense environments simply involve an orthokinesis mechanism, by which locomotion rates are higher in low-density and lower in high-density environments? This seems particularly relevant as dopamine (and dop-3) are known to control speed but not to my knowledge reorientation. It would be good for the authors to directly test turning/reversal rates (with a correction for the known differences in speed).

2) Although there is a significant amount of data in the paper about the role of dop-3, they surprisingly never do cell-specific rescue experiments to see where the DOP-3 receptor is acting. The previous work of Chase et al. suggests it could function in ventral cord motorneurons, and it would be straightforward to see if expression there is sufficient to rescue their phenotypes. This could fill in a missing gap to provide a satisfying molecular/circuit model for the behavior.

3) The authors show that mec-10 mutants are defective in the behavior and cite this as evidence that mechanosensation is involved. But MEC-10 is not known to play a direct role in mechanosensation by the dopamine neurons and indeed is not even expressed in those cells. The only plausible explanation I can think of is that since mec-10 functions in the FLP neurons, and the FLPs and CEPs are known to modulate each other's activity through a gap junction circuit of nose touch neurons (Chatzigeorgiou 2011, Rabinowitch 2013), mec-10 could influence CEP activity indirectly through this circuit. They could test this easily by seeing if the mec-10 phenotype rescues under an FLP promoter. As with point 2, this could tie what now seems like a random observation into a convincing circuit model.

Reviewer #2:

The manuscript from Han et al. evaluates the ability of C. elegans nematodes to use surface texture as a navigational aid (a process called thigmotaxis in other literature). The study introduces a behavioral assay for texture preference based on an arena fabricated using soft-lithography techniques and evaluates the effect of mutations in two ion channel genes-mec-10 and trp-4. Additionally, the work links mechanosensation to dopaminergic signaling and its reception by the DOP-3 receptor. The mec-10 gene encodes a pore formingsubunit of the sensory mechanotransduction channel in the gentle touch receptor neurons (TRNs) and the trp-4 gene encodes a pore-forming subunit of the sensory mechanotransduction channel expressed by texture-sensing and proprioceptive neurons (CEP, ADE, PDE). The TRP-4 protein is homologous to Drosophila NOMPC and is predicted to form a non-selective cation channel activated by mechanical force. The main findings are that C. elegans nematodes perform thigmotaxis, that this behavior differs among wild isolates and depends on TRP-4 ion channels as well as dopaminergic signaling.

The data are presented in a clear and engaging manner, certain findings rely on mutant alleles that do not cleanly or clearly eliminate the cellular or molecular function in question (see below). For this reason, this reviewer believes that additional experimental work is essential to support the current conclusions.

1) The role of mec-10-expressing mechanoreceptor neurons in quadrant assays (part 1)

The preference defect in mec-10(ok1104) mutants is as strong as that found for trp-4(ok1605) mutants (compare Figure 1—figure supplement 2C with Figure 5D). Thus, it is puzzling that the authors report the mec-10 result as supplementary material and even more puzzling that they have chosen to focus exclusively on trp-4. Given that these genes are expressed in non-overlapping sets of mechanoreceptor neurons, it seems essential to understand determine whether or not these genes act together to mediate the observed thigmotaxis behaviors.

2) The role of mec-10-expressing mechanoreceptor neurons in quadrant assays (part 2)

The authors study mec-10(ok1104) and find that these animals have altered responses to pillar arrays. This conclusion is predicated on the notion that the ok1104 mutation disables mechanosensation – (Results): "mutant worms with dysfunctional MEC-10 DEG/ENaC mechanosensory channels showed a significant reduction in their preference for pattern IV (Figure 1—figure supplement 2C), demonstrating that spatial pattern selectivity requires proper mechanosensation." But, the situation is much more complex. First, mec-10 is expressed in three classes of mechanoreceptor neurons, the touch receptor neurons (TRN), PVD and FLP. Which neurons are responsible for this effect? Second, the ok1104 allele has very mild effects on TRN-mediated mechanosensation. In fact, Arnadottir et al. (J Neurosci, 2011) showed that ok1104 mutants retain nearly normal behavioral responses to touch and mechanoreceptor currents. Additionally, the ok1104 allele deletes 143bp of the gene and allows for the transcription of at least a single transcript encoding an abberant protein. Thus, the ok1104 allele is unlikely to be null. Thus, the observed defects in behavioral assays could be due to the expression of a defective protein in the TRNs, PVD, or FLP. At a minimum, the authors need to communicate this complexity and uncertainty to readers. It would be better, however, to evaluate additional alleles, including tm1552, a different deletion in mec-10 and u253 a deletion in mec-4 that eliminates mechanosensory currents in the TRNs and behavioral sensitivity.

3) The contribution of trp-4 to responses to pillar assays

There are three alleles encoding deletions in the trp-4 gene available from the CGC stock center: gk341, sy695, and ok1605. It is unfortunate that the authors have chosen ok1605 for this study-this allele encodes an in-frame deletion that may not be null. Thus, the behavioral deficit observed could be the result of an aberrant TRP-4 protein rather than the absence of TRP-4. The interpretation of the authors' examination of trp-4 and its genetic polymorphisms depends on resolving this uncertainty.

4) Differences in current-displacement curves in N2 and HW

The recordings in Figure 6 are presented in an ambiguous manner and critical information about the recording conditions are missing.

Did the authors record from CEP neurons in wild-type N2 and HW worms? Or, were recordings performed in transgenic N2 animals expressing TRP-4(N2) and TRP-4(HW) from transgenes?

If the latter, were the recordings performed in a wild-type N2 background or in trp-4(ok1605)? What holding potential used to record currents? Please add this information the figure and/or its legend.

Where voltages were corrected for junction potentials?

Do the maximal currents differ between N2 and HW? Please show pooled data for absolute current amplitude for 5µm displacements that saturate the current-displacement curves in both cases.

The answers to these questions affect interpretation, especially in light of point#3 above. The authors ascribe the difference displacement sensitivity to polymorphisms in the trp-4 gene. While this inference is consistent with the available data, it is not the only possible interpretation. For example, if the recordings were in N2 and HW genetic backgrounds-variations in accessory proteins could just as easily account for the observed effects.

5) Variation in behavioral preferences among wild isolates

The difference between N2 and HW is striking and the authors' efforts to understand this source of this phenotypic difference are a terrific first start. However, their inferences about the causal role played by SNPs in the trp-4 gene extend beyond the data available. For instance, since trp-4 is not the only single gene that can mutate to alter this behavior (see point #1 above and the defect that appears to be linked to the X chromosome), the exclusive focus on trp-4 is not really warranted. In other words, the phenotypic differences between wild isolates may be a combined function of differences in trp-4 function and other factors.

6) Behavioral mechanism of preferences in quadrant assay

Throughout the description of the quadrant assay, this reviewer was struck by the concept of preference. Given the authors' finding that animals move more slowly in the denser patterns, isn't this arena essentially a speed trap? (i.e. animals accumulate in pattern IV because once they enter this zone, they move more slowly. The fact that this preference declines in longer assays – Figure 1—figure supplement 1 – supports this alternative interpretation). It would be useful to distinguish between the apparent preference and how this preference is produced.

[…] Reviewer #1:

[…] 1) In the manuscript, the authors state that "worms sense localchanges in pattern density, and respond quickly to unfavorable changes by reorientation," (Results). However, I don't think the authors present any direct evidence that reorientation (i.e. turning/reversal rates) change when crossing from higher to lower obstacle density. In principle, could the accumulation in more dense environments simply involve an orthokinesis mechanism, by which locomotion rates are higher in low-density and lower in high-density environments? This seems particularly relevant as dopamine (and dop-3) are known to control speed but not to my knowledge reorientation. It would be good for the authors to directly test turning/reversal rates (with a correction for the known differences in speed).

In the revised manuscript, we have added new results showing that worms reorient at the boundaries (increased turns and reversals; Figure 2C). Dopamine has been shown to increase the frequency of high-angled turns during area-restricted search behavior (Hill et al. 2004). Consistent with a role of DOP-3 in reorientation, a previous study suggests that dop-3 mutant animals make longer-lasting turns, which result in curvier trajectories (Stephens et al., 2010).

2) Although there is a significant amount of data in the paper about the role of dop-3, they surprisingly never do cell-specific rescue experiments to see where the DOP-3 receptor is acting. The previous work of Chase et al. suggests it could function in ventral cord motorneurons, and it would be straightforward to see if expression there is sufficient to rescue their phenotypes. This could fill in a missing gap to provide a satisfying molecular/circuit model for the behavior.

DOP-3 is strongly expressed in GABAergic motoneurons, as well as many neurons in the head and the tail. To ask if DOP-3 functions in GABAergic neurons, we have expressed single-copy Punc-47::dop-3 transgenes in dop-3(vs106) mutant worms. However, DOP-3 in GABAergic neurons fails to rescue the defects, indicating that DOP-3 expression in GABAergic motoneurons is not sufficient to support spatial pattern selection. Because these negative results do not demonstrate the site-of-action of DOP-3, we have not included them in the manuscript. However, if the reviewer feels that these data are essential, we can add these results into the manuscript.

The broad expression pattern of DOP-3 has made it difficult to precisely identify the site-of action. We agree with the reviewer that cell-specific experiments will fill in the missing knowledge gap. We continue to perform these experiments, but it will require much more effort before knowing the answer.

3) The authors show that mec-10 mutants are defective in the behavior and cite this as evidence that mechanosensation is involved. But MEC-10 is not known to play a direct role in mechanosensation by the dopamine neurons and indeed is not even expressed in those cells. The only plausible explanation I can think of is that since mec-10 functions in the FLP neurons, and the FLPs and CEPs are known to modulate each other's activity through a gap junction circuit of nose touch neurons (Chatzigeorgiou 2011, Rabinowitch 2013), mec-10 could influence CEP activity indirectly through this circuit. They could test this easily by seeing if the mec-10 phenotype rescues under an FLP promoter. As with point 2, this could tie what now seems like a random observation into a convincing circuit model.

We thank the reviewer for constructive insights. To understand where MEC-10 functions, we have carried out cell-specific rescue experiments (Figure 1F). We find that single-copy Pegl-44::mec-10 transgenes fully rescue the selection defects in mec-10(tm1552) mutant worms. By contrast, single-copy Pmec-18::mec-10 transgenes fail to rescue. These results confirm the reviewer’s prediction that MEC-10 functions in FLPs but not in TRNs.

The text has been modified as follows: “By contrast, mutant worms with dysfunctional MEC-10 DEG/ENaC channels showed a significant reduction in their preference for pattern IV (Figure 1E). […] Together, these data indicate that MEC-10 acts in the polymodal nociceptive FLP neurons to support the selection of pattern IV.”

Reviewer #2:

[…] 1) The role of mec-10-expressing mechanoreceptor neurons in quadrant assays (part 1)

The preference defect in mec-10(ok1104) mutants is as strong as that found for trp-4(ok1605) mutants (compare Figure 1—figure supplement 2C with Figure 5D). Thus, it is puzzling that the authors report the mec-10 result as supplementary material and even more puzzling that they have chosen to focus exclusively on trp-4. Given that these genes are expressed in non-overlapping sets of mechanoreceptor neurons, it seems essential to understand determine whether or not these genes act together to mediate the observed thigmotaxis behaviors.

The main reason for our focus on TRP-4 is due to the tight link between Chromosome 1 and the changes in pattern IV preference. While we have also observed that chromosome X is linked to behavior changes, the small alteration in the hybrid strain (Figure 5B) has made it very difficult to further identify the linkage region. Therefore, we have focused on the traceable linkage in the trp-4 region on Chromosome 1. We agree with the reviewer that trp-4 is not the only change between N2 and CB4856 strains. Indeed, we have stated in the previous manuscript that “other gene(s) on Chromosome X may contribute to the polymorphic behavior”.

We have carried out several experiments to further investigate the role of MEC-10. First, we have studied another mec-10 allele tm1552. Our results confirm that disruption of MEC-10 reduces the preference for pattern IV. Second, cell-specific rescue experiments show that MEC-10 functions in FLP neurons (please see response to point 3 by reviewer 1). These data are now presented in Figure 1F in the revised manuscript.

2) The role of mec-10-expressing mechanoreceptor neurons in quadrant assays (part 2)

The authors study mec-10(ok1104) and find that these animals have altered responses to pillar arrays. This conclusion is predicated on the notion that the ok1104 mutation disables mechanosensation – (Results): "mutant worms with dysfunctional MEC-10 DEG/ENaC mechanosensory channels showed a significant reduction in their preference for pattern IV (Figure 1—figure supplement 2C), demonstrating that spatial pattern selectivity requires proper mechanosensation." But, the situation is much more complex. First, mec-10 is expressed in three classes of mechanoreceptor neurons, the touch receptor neurons (TRN), PVD and FLP. Which neurons are responsible for this effect?

In the revised manuscript, we include a set of results from cell-specific rescue experiments (Figure 1F). As we discussed above, our results show that MEC-10 functions in FLPs.

Second, the ok1104 allele has very mild effects on TRN-mediated mechanosensation. In fact, Arnadottir et al. (J Neurosci, 2011) showed that ok1104 mutants retain nearly normal behavioral responses to touch and mechanoreceptor currents. Additionally, the ok1104 allele deletes 143bp of the gene and allows for the transcription of at least a single transcript encoding an abberant protein. Thus, the ok1104 allele is unlikely to be null. Thus, the observed defects in behavioral assays could be due to the expression of a defective protein in the TRNs, PVD, or FLP. At a minimum, the authors need to communicate this complexity and uncertainty to readers. It would be better, however, to evaluate additional alleles, including tm1552, a different deletion in mec-10 and u253 a deletion in mec-4 that eliminates mechanosensory currents in the TRNs and behavioral sensitivity.

We have evaluated the mec-10(tm1552) allele, as suggested by the reviewer. We show that the preference for pattern IV is disrupted in this putative null allele (Figure 1E).

3) The contribution of trp-4 to responses to pillar assays

There are three alleles encoding deletions in the trp-4 gene available from the CGC stock center: gk341, sy695, and ok1605. It is unfortunate that the authors have chosen ok1605 for this study-this allele encodes an in-frame deletion that may not be null. Thus, the behavioral deficit observed could be the result of an aberrant TRP-4 protein rather than the absence of TRP-4. The interpretation of the authors' examination of trp-4 and its genetic polymorphisms depends on resolving this uncertainty.

To address this question, we have generated a new trp-4 allele (pek220) using CRISPR/CAS9 and homologous recombination methods. As shown in Figure 3—figure supplemental 1, a disruption cassette that contains gfp::tbb-2utr and Prps-0::hygR::unc-54utr was inserted into trp-4 exon 2 (Figure 3—figure supplement 1). Insertion of this disruption cassette results in a deletion and an early stop codon that put trp-4 out-of-frame after the amino acid residue 53. The mutant trp-4(pek220) worms exhibit similar defects in the behavioral test as trp-4(ok1605) worms (Figure 3C), confirming that pattern IV preference requires TRP-4. Furthermore, we have shown that the behavioral defect in trp-4(ok1605) worms can be fully rescued by trp-4 cDNA expression in dopaminergic neurons. Therefore, the behavioral defects of trp-4(ok1605) worms are due to the loss of TRP-4 function.

4) Differences in current-displacement curves in N2 and HW

The recordings in Figure 6 are presented in an ambiguous manner and critical information about the recording conditions are missing.

We have modified the text to clarify the recording conditions, as detailed below.

Did the authors record from CEP neurons in wild-type N2 and HW worms? Or, were recordings performed in transgenic N2 animals expressing TRP-4(N2) and TRP-4(HW) from transgenes?

If the latter, were the recordings performed in a wild-type N2 background or in trp-4(ok1605)?

The recordings were performed using wild type N2 and HW strains. We have clarified this point by stating “Representative traces of mechanoreceptor currents in wild type N2 and HW dopaminergic CEP neurons …” in the legend for Figure 6A.

What holding potential used to record currents? Please add this information the figure and/or its legend.

We have added the following statement “Whole-cell patch clamp recordings (holding potential at −75 mV) were carried out at 20°C” in the figure legend.

Where voltages were corrected for junction potentials?

The following text was added to the Methods section: “The junction potential was measured by ± 10mv 5ms square testing signals. After the pipette tip was inserted into the bath solution, the junction potential was corrected using the “Auto V0” function, which calls a procedure for automatic zeroing of the pipette current. Just before the pipette touched the cell membrane for seal, the “Auto V0” was used again to make sure the junction potential was well corrected.”

Do the maximal currents differ between N2 and HW? Please show pooled data for absolute current amplitude for 5µm displacements that saturate the current-displacement curves in both cases.

The answers to these questions affect interpretation, especially in light of point #3 above. The authors ascribe the difference displacement sensitivity to polymorphisms in the trp-4 gene. While this inference is consistent with the available data, it is not the only possible interpretation. For example, if the recordings were in N2 and HW genetic backgrounds-variations in accessory proteins could just as easily account for the observed effects.

As suggested by the reviewer, we have included a comparison of absolute current amplitudes in N2 and HW worms in Figure 6B. No significant difference was detected, which suggests that the HW TRP-4 is fully active when mechanical stimuli are sufficiently strong.

Because results from electrophysiological recordings could be explained by changes in accessory proteins, we also assayed N2 and HW cDNAs for their ability to rescue behavioral defects. We show that N2 and HW trp-4 have a significant difference in their ability to rescue the behavioral defects in trp-4(ok1605) mutant worms (Figure 5D). We also show that transgenic HW worms that carry N2 trp-4 exhibit increased preference of pattern IV when compared to those HW worms expressing HW trp-4 (Figure 5E). These results are unlikely due to changes in accessory proteins, as worms with the same genetic background were used for comparison.

5) Variation in behavioral preferences among wild isolates

The difference between N2 and HW is striking and the authors' efforts to understand this source of this phenotypic difference are a terrific first start. However, their inferences about the causal role played by SNPs in the trp-4 gene extend beyond the data available. For instance, since trp-4 is not the only single gene that can mutate to alter this behavior (see point #1 above and the defect that appears to be linked to the X chromosome), the exclusive focus on trp-4 is not really warranted. In other words, the phenotypic differences between wild isolates may be a combined function of differences in trp-4 function and other factors.

As discussed above, we agree that trp-4 is unlikely to be the only single gene accounting for the phenotypic differences among wild isolates. First, we have stated that, “other gene(s) on Chromosome X may contribute to the polymorphic behavior”. Second, we have shown that an additional SNP in DOP-3 is likely to cause behavioral changes in some wild isolates (Figure 7).

The main reason of our TRP-4 focus is the tight link between chromosome 1 and the reduced preference for pattern IV. This link is traceable and we found that it ties to trp-4 in the case of the HW strain. In general, we agree with the reviewer that many factors may contribute to the phenotypic differences. What our data have shown is that SNPs in trp-4 and dop-3 produce traceable and significant changes.

6) Behavioral mechanism of preferences in quadrant assay

Throughout the description of the quadrant assay, this reviewer was struck by the concept of preference. Given the authors' finding that animals move more slowly in the denser patterns, isn't this arena essentially a speed trap? (i.e. animals accumulate in pattern IV because once they enter this zone, they move more slowly. The fact that this preference declines in longer assays – Figure 1—figure supplement 1 – supports this alternative interpretation). It would be useful to distinguish between the apparent preference and how this preference is produced.

We find that behavioral mechanisms in the quadrant assay are complex and context-dependent. Both reversal and turning rates increase once worms step out of the pattern IV (Figure 2C), suggesting that additional mechanisms besides locomotion speed are involved. Interestingly, the reduction in locomotion speed in pattern IV is context-dependent, as we find that neighboring patterns have a significant impact on worm locomotion (Figure 2E, F). Furthermore, the selection for pattern IV depends on feeding state, suggesting that this behavior is not fixed to the physical design of the patterns. Based on these observations, we propose that several behavioral strategies including changes in speed and turning are engaged to produce the plastic behavior of pattern selection.


Articles from eLife are provided here courtesy of eLife Sciences Publications, Ltd