Abstract
Tree-like neurites are crucial for receiving information into neurons. It is assumed that nurturing affects the structure and function of dendrites, yet the evidence is scarce, and the mechanisms are unknown. To study whether mechanosensory experience affects dendritic morphology, we use natural mechanical stimulation of the Caenorhabditis elegans’ polymodal PVD neurons, induced by physical contacts between individuals. We found that animal isolation affects the dendritic tree structure of the PVD. Moreover, developmentally isolated animals show a decrease in their ability to respond to harsh touch. The structural and behavioral plasticity following mechanosensory deprivation are functionally independent of each other and are mediated by an array of evolutionary conserved amiloride-sensitive epithelial sodium channels (degenerins). Our results suggest an activity-dependent homeostatic mechanism for dendritic structural plasticity, acting downstream to mechanosensory activation of degenerins.
Main Text
The general structure of the nervous system has been known for over a century. Groundbreaking studies on synaptic plasticity and its underlying mechanisms have shown that before birth and in adult animals, brain waves of neural activity are needed for synaptic remodeling 1–4. In contrast, the molecular mechanisms responsible for structural plasticity of dendritic trees, as an output of different sensory signals, especially during adulthood, are poorly understood 5–7.
Mechanistic understanding of experience-dependent dendritic structural plasticity is focused on activity sensation by calcium channels and N-methyl-D-aspartate (NMDA) receptors, that induce downstream signaling cascades including Rho family of small GTPases, calcium metabolism and microtubule stability 8–11. Since neurological disorders like autism, Down syndrome, fragile X syndrome, and schizophrenia are characterized by abnormal dendritic spine structures, uncovering the molecular basis of dendritic tree instability during development and adulthood, is of great importance 12.
The morphology of the C. elegans’ PVD bilateral neurons is composed of repetitive and spatially organised structural units that resemble candelabra (Fig. 1A) 13. These structural properties provide a useful platform to study dendritic morphogenesis. While some of the genetically programmed molecular mechanisms responsible for the development, morphogenesis and regeneration of PVD’s dendritic trees are known 14–20, the influence of nurture, e.g. sensory experience, on its structure and function during development and in adulthood remain unexplored. The PVD mediates three sensory modalities: response to harsh touch 21, response to low temperatures 21 and proprioception 22. Rose et al 23 found that deprivation of mechanosensory stimulation, generated by colliding conspecifics in the growing plate, resulted in reduced response to gentle tap stimulation and modified glutamatergic signalling in gentle touch circuits. Here, we adapted this natural mechanosensory stimulation paradigm to identify the mechanism that couples mechanosensory activity to structural and functional plasticity, focusing on degenerins/epithelial Na+ channels (DEG/ENaCs) expressed in the PVD 21, 24. These channels form homo- and hetero-trimers involved in mechanosensation and cognitive functions like learning and memory 25–28. We investigate how mechanosensory experience, mediated by activity through DEG/ENaCs, affects structural plasticity of the PVD dendritic trees in adult C. elegans. Sensory and social isolation affect the behavior and fitness of diverse animals 29, including primates 30. Our findings on the plasticity of stereotypic dendritic trees of polymodal somatosensory neurons in adult nematodes, reveal mechanisms of dendritic plasticity induced by mechanosensation that may be conserved.
Experience induces behavioral plasticity
To evaluate PVD activity we tested one of its modal functions using a behavioral assay that measures escape from the noxious mechanical stimulus by prodding with a platinum wire (harsh touch assay; Fig. 1A) (24). To study whether mechanosensory deprivation affects the nociceptive functions of the PVD 13, 31, we isolated embryos into single plates, from hatching through adulthood and compared their behavioral response to harsh touch against adults that were grown for 72 h on crowded plates (Fig. 1B). We found that ~40% of isolated wildtype (WT) animals responded to harsh touch, compared to ~80% of animals grown in crowded plates (Fig. 1C). To exclude the possibility that gentle touch neurons are involved in this behavioral response, we isolated mec-4 mutants, which are insensitive to gentle touch 32, and obtained similar results. As a negative control, we used mec-3 mutants, that are touch insensitive (31) and found a similarly low response for both groups (Fig. 1C). Thus, isolation reduces the response to noxious stimuli during adulthood in a process that is independent of gentle touch.
To investigate whether the isolation-induced decreased response is a PVD-dependent function, we used optogenetic stimulation 24. We found that isolation reduced the percentage of worms responding to optogenetic stimulation of the PVD (Fig. 1D), indicating that the plasticity in the response is acting downstream to PVD activation and mechanosensory channels.
Since mechanosensory experience can be induced not only by conspecifics present on the plate, but also by solid inert objects 33, we used the mec-4 strain to compare responses of isolated worms to responses of isolated worms grown in the presence of glass beads. Worms grown in isolation with beads had a similar response compared to animals grown in crowded plates (Fig. S1a). We then used plates pre-stimulated with pheromones (Fig. 1E) and osm-6 mutants (defective sensory cilia, including the PDE neuron 34 that reduces also harsh touch responses 35 (Fig. S1b)). For both experiments we found that isolation induced a reduction in the response to harsh touch. Thus, behavioral plasticity for harsh touch stimulation of the PVD is dependent on mechanosensory stimulation during development and is chemosensory-independent 35.
MEC-10 affects behavioral plasticity
To study the genetic mechanisms of plasticity during nociceptive response, we performed a candidate gene screen for degenerins (DEG/ENaC) and transient receptor potential (TRP) channels that are expressed in the PVD and affect its behavioral responses 9, 20, 36. The harsh touch response following isolation was reduced in WT, degt-1 and gtl-1 mutants, suggesting that these channels are not directly involved in behavioral plasticity following isolation (Fig. 1F). In contrast, for del-1, asic-1 and mec-10 mutants the difference between isolated and crowded conditions was undetectable, indicating that they are required for behavioral plasticity. Interestingly, while the harsh touch response of asic-1 mutants was high and similar to crowded WT worms, the response for mec-10 mutants was low, similar to isolated WT animals. To test whether the response to harsh touch is dependent on DEG/ENaC activity, we used DEG/ENaC blocker-amiloride for worms that were grown in crowded plates. We found that the response to harsh touch was not affected by continuous growth in the presence of amiloride (Fig. S1C). This result supports the idea that the combinatorial activities of multiple amiloride-sensitive epithelial sodium channels have positive and negative effects on the response to harsh touch (Fig. 1F).
Since MEC-10 is expressed in the PVD and is responsible for its response to harsh touch 21, we asked how it mediates the behavioral plasticity following mechanosensory experience in the crowded, but not isolated conditions. To test whether the activity of MEC-10 is required autonomously in the PVD, we performed a cell-specific rescue experiment, with a plasmid encoding for MEC-10, under a PVD-specific promoter 13, 37. We found that expression of MEC-10 in the PVD rescues the reduction in response to harsh touch in crowded mec-10 mutants, indicating that it acts cell autonomously to modulate behavioral plasticity of mechanosensory signals (Fig. 1G) 21.
To determine whether isolation affects the response to harsh touch during development or in adults, we isolated young adults for 24 h and found that isolation of young adults did not affect the response to harsh touch (Fig. 1H). Thus, the isolation-induced reduction in response to harsh touch is MEC-10-dependent and is determined during development.
Experience affects morphology via MEC-10
Sensory experience drives synaptic plasticity in the nervous system 3, 4, 38. To test whether experience regulates the morphology of the dendritic tree of the PVD and its function as a nociceptor 39, we used the mechanical deprivation paradigm (Fig. 1B) and examined morphological features of the PVD (Fig. 2A): (1) The fraction of ectopic (excess) out of non-ectopic branches (those that form the classical candelabrum), (2) the percentage of straight quaternary branches and (3) the percentage of self-avoidance defects between adjacent candelabra.
We found that isolation of both WT and him-5 (the WT background for several strains after cross) animals for 72 h, increased the fraction of ectopic branches compared with age matched crowded worms (Fig 2B), and reduced the percentage of straight quaternary branches (Fig 2C). Moreover, isolation increased self-avoidance defects (Fig. 2D-G, Fig. S2). These results show that the PVD dendritic tree is sensitive to the sensory signals generated by other worms on the plate. Using our quantitative assays to determine changes in the structure of the PVD neuron following mechanosensory isolation, we decided to investigate whether these morphological phenotypes are induced via MEC-10 mechanoreceptor; we compared the PVD arborisation patterns between crowded and isolated mec-10 mutants. We found that crowded mec-10 animals were characterized by isolated-like morphological features (more ectopic branches, fewer straight quaternary branches and more self-avoidance defects, compared to crowded him-5 animals; Fig. 2B-G). These results suggest that PVD morphology is affected by sensory experience in a mec-10-dependent pathway. To study whether MEC-10 acts cell autonomously to mediate morphological plasticity, we used the PVDp::MEC-10 line in a mec-10 background and found a reduced fraction in the ectopic branches and an increased percentage in the number of straight quaternary branches, compared to age-matched mec-10 animals (Fig. 2H-I). Interestingly, expression of MEC-10 in the PVD had no effect on the amount of self-avoidance defects (Fig. 2J-M), suggesting that it acts cell non-autonomously to mediate loss of self-avoidance. Activity and sodium influx via the DEG/ENaC UNC-8 promotes synapse elimination in C. elegans 40. To test how global inhibition of DEG/ENaCs affects the morphology of the PVD neuron, we compared crowded worms that were grown on plates with amiloride to control worms. We found that blocking DEG/ENaCs by amiloride increased the fraction of ectopic branching and decreased the percentage of straight quaternary branches, without affecting self-avoidance (Fig. S3). Thus, the structure of the PVD is sensitive to activity and cation influx via DEG/ENaCs. To determine whether the morphological effect on PVD structure is also mediated by mechanical (like friction 41 and collisions 21), rather than chemical cues, we used glass beads under isolated conditions, to test the effect on the morphology. We found that mechano-stimulation induced by glass beads did not restore the fraction of ectopic branching, but it did significantly increase the number of straight quaternary branches and decreased the percentage of self-avoidance defects (compared with isolated animals without beads, Fig. 3A-E). Thus, mechanosensory stimuli induce morphological plasticity in the PVD via DEG/ENaCs.
Similarly to Fig. 2B-D, animal isolation onto pheromone-conditioned plates 42, increased the fraction of ectopic branching, decreased the percentage of straight quaternary branches and increased the percentage of self-avoidance defects (Fig. S4). These results indicate that the effect of isolation on the structure of the PVD is not mediated by chemical signals. Additionally, we looked at mec-4 gentle-touch-insensitive mutants 23, 32 and confirmed that isolation caused a similar change to PVD structure as WT (Fig.S5, compare with Fig. 2B-D). In summary, mechanosensory experience controls morphological plasticity of PVD dendritic trees via MEC-10 activity and independently of MEC-4.
Adult isolation induces tree complexity
Isolation of eggs for 72 h comprises ~48 h of development followed by ~24 h of adulthood. To determine whether the isolation-induced morphological phenotypes of the PVD (Fig. 2A) occur during development or alternatively during adult maintenance, we compared crowded and isolated WT worms after 48 h, as young adults. We found a small but significant difference only for the percentage of straight terminal 4ry branches (Fig. S6), suggesting that isolation affects maintenance of PVD morphology during adulthood. To determine directly the period when mechanical isolation acts to cause structural phenotypes, we studied the morphology of the PVD in worms that were grown in crowded plates and then isolated as young adults for 2, 5 or 24 h. Isolation of adults for 2 h had no significant consequence on the structure of the PVD (Fig. S7A-C), while isolation of adults for 5 h reduced the percentage of quaternary straight branches compared to crowded worms (Fig. S7D-F). In contrast, worms isolated as adults for 24 h (Fig. 3F) showed increased PVD complexity, similar to worms grown under continuous isolation of eggs for 72 h (Fig. 2B-G, Fig. 3G-K). Thus, the architecture of the adult PVD is sensitive to the duration of mechanosensory signals during adulthood.
Structure and function are independent
To test the model proposing that there is a causal link between the morphology of the dendritic tree of the PVD and its function as a nociceptor 39 we followed the isolation protocol described in Fig. 1B for seven combinations of DEG/ENaC mutants and analyzed their response to harsh touch (Fig. S8) and their PVD structure (Fig. S9). To compare the morphological features of different DEG-ENaC genotypes and treatments (crowded, isolated; Fig. S9), we used discriminant analysis as a supervised classification method to combine all the morphological phenotypes analyzed (Fig. 4A). In addition to the data based on morphological similarities between groups, we superimposed the information obtained from behavioral harsh touch experiments for DEG/ENaC mutants (from Fig. S8) on the morphological clustering. We utilized a binary like distribution of the worms in terms of the response to harsh touch (<45% for isolated versus >65% for crowded, as shown in Fig. S8). We found no correlation between the morphology of the PVD and the response to harsh touch when testing the different combinations of genotypes and treatments. In particular, isolated mec-10;degt-1 mutants show crowded-like morphology with isolated-like behavioral response (Fig. 4A and Fig. S8, S9). These findings suggest that these two proteins are required for the isolation-induced morphological modification of the PVD and represent an example in which the response to harsh touch is independent of the structural alteration of the PVD. Since MEC-10 is important for laminar shear stress 41 and touch 21 we suggest that this channel and its possible partners from the DEG/ENaC family, such as DEGT-1, ASIC-1 and DEL-1 25, 43, 44, affect the structure and the function of the PVD independently, in an experience-dependent manner and contrary to the original hypothesis by Hall and Treinin 39.
Additional line of evidence supporting the independence was demonstrated for isolation of young adult worms for 24 h (Fig. 3F). This isolation affects the structure of the PVD (Fig. 3G-K) but has no effect on the response to harsh touch (Fig. 1H). Finally, to directly demonstrate that these two features are independent, we analyzed harsh touch responses of individual animals and then assayed each individual animal for its PVD morphology. We compared the dendritic morphology of individually isolated worms that either responded to harsh touch or didn’t. We found that the three morphological parameters were unchanged (Fig. 4B-F). Thus, analysis at the level of individual worms failed to demonstrate a correlation between the morphology and the response to harsh touch. In summary, the morphological and behavioral phenotypes were independently affected by experience via degenerins. We cannot exclude the possibility that other functions of the PVD, like the response to low temperatures 21 and proprioception 22 are more tightly associated to the structure of the PVD.
MEC-10 localization is experience-dependent
Since MEC-10 and DEGT-1 tend to co-localize within the PVD 21 we analyzed the genetic interaction between these two proteins, during mechanosensory experience. Differential localization of degenerins can affect both the behavioral response to harsh touch and also the structural properties of the neuron. We hypothesized that changes in the localization patterns of DEG/ENaC can account for plasticity at both the behavioral (Fig. 1) and the structural level (Figs. 2, 3). MEC-10 localization in the plasma membrane and in intracellular vesicular compartments of the axon and the quaternary branches was reduced after isolation (Fig. 5A-D, Movies S1 and S2). We also tested the localization pattern of the DEGT-1 21 in the PVD. In contrast to MEC-10 (Figs. 5A-D), DEGT-1 localization is reduced only in the cell body following isolation (Fig. 5E-H). Furthermore, degt-1 mutants reduced the amount of MEC-10, and more importantly, abrogated the isolation-induced reduction in MEC-10 localization at the quaternary branches and the axon (Fig. 5A-D). In the reciprocal experiment, DEGT-1 localization at the cell body was affected by mec-10 mutation, as isolated worms exhibit increased localization compared to WT isolated worms (Fig. 5E-H). Thus, mechanosensory experience induced plasticity in the localization pattern of MEC-10 and DEGT-1. Such experience-dependent plasticity is proposed to be part of the mechanism that locally modulates dendritic and axonal properties, to affect both the structure and the function of the PVD, respectively (Fig. 6).
Discussion
From the evolutionary point of view, dendritic trees and their structural complexity remain mysterious entities, despite many efforts to understand the contribution of arborization complexity to dendritic physiology 45.
Previous research has demonstrated both cell autonomous 13, 14, 46 and cell non-autonomous 15, 16 mechanisms that regulate PVD’s dendritic morphogenesis during development. Few studies have also focused on regeneration and aging effects on the tree structure of the PVD revealing plastic mechanisms during the adult stage 17, 18, similar to what has been shown for Drosophila sensory neurons 47. We have uncovered here that nurture, manifested as mechanosensory experience, activates mechanotransduction signaling, via DEG/ENaCs amiloride-sensitive activity, to maintain the structure of the dendritic tree in adults. Figure 6 shows our working model, where the amount of mechanosensory stimulation, in crowded or isolated conditions, affects the localization of MEC-10 in different compartments of the PVD. Localization of MEC-10, probably by forming high order complexes with other DEG/ENaC, can affect the structure of the PVD at the level of dendritic tree. Thus, experience induced structural plasticity of the dendrites is homeostatic at the dendritic branch level. In contrast, we demonstrate that the response to harsh touch is modulated by the duration and timing of mechanosensory experience and by the presence of degenerins during development. Behavioral plasticity is probably a synaptic property, mediated by DEG/ENaC 48 and related to neurotransmission modulation, independently of the structure of the PVD dendritic tree. These structural and behavioral plasticity are separated in time (adulthood vs. development) and space (dendrite vs. axon).
Somatosensory activation in vertebrates plays a prominent role in shaping the structural and functional properties of dendritic spines, mainly studied in the central nervous system 49–52. Here we suggest that degenerins mediate mechanosensation-induced dendritic growth in sensory dendrites. In contrast to mammalian cortical neurons, much less is known about their degree of plasticity and the molecular mechanisms utilized during adulthood. We suggest that the dendritic plasticity we described bears resemblance to the activity-dependent effect of glutamatergic signaling and NMDA receptors. Activity via NMDA affects dendritic spines as an upstream mechanism of cell signaling, resulting in structural modifications 53–55. We propose that degenerins mediate mechanosensory signaling sensation, by activating cationic gradients 56, leading to activation of downstream intracellular signaling pathways 8–11, which in turn enable local, actin-mediated 3, 57, 58 structural plasticity in the PVD dendritic branches. In parallel, it is possible that DEG/ENaCs are also modulating pre/postsynaptic homeostatic signaling in the harsh touch circuit, as has been shown in neuromuscular junctions 59.
Thus, we demonstrated mechanosignaling functions for the combinatorial actions of DEG/ENaCs that act as mediators of experience-induced structural plasticity of sensory dendritic trees.
Author contributions
SI and BP conceived and designed the experiments. SI performed the experiments. SI and BP analyzed the data and wrote the paper.
Author information
Competing interests
None
Materials and Methods
Strains
Nematode strains were maintained according to standard protocols 60. The list of the strains is presented in Supplementary Table 1. Strains of the DEG/ENaCs family obtained from the CGC (JPS282: asic-1(ok415) I; ZB2551: mec-10(tm1552) X; VC2633: degt-1(ok3307) V) were crossed with BP709: IS[hmnEx133](ser-2Prom3::kaede). The validation of F2 homozygotes for the DEG/ENaCs (including single, double and triple mutants) was performed by PCR amplification of the deleted region in the genome.
Primers for Multiplex PCR
asic-1(ok415) I: Forward-1: 5’ aactggtgtggccacttcaactttc 3’; Forward-2: 5’ aaggtttcagatgatcgcgtagtcaag 3’; Reverse: 5’ catttctcttcttccgtcagcgc 3’.
mec-10(tm1552) X: Forward-1: 5’ acacggctccttcttgagttccga 3’; Forward-2: 5’ attcggtttcctcctcttcttccaatgc 3’; Reverse: 5’ cgtttttttcagcgccctttcctgca 3’
degt-1(ok3307) V: Forward-1: 5’cgagtagctgattatcaaaaagtcctcga 3’; Forward-2: 5’ cggatattccagcattggcgaa 3’; Reverse: 5’ ttccccgttgatcttctatgtattaca 3’.
Spinning disk confocal microscopy
Prior to imaging, the worms were placed on an agar pad (10%), mounted on top of a microscope glass slide, with 1μl of polystyrene beads (100 nm diameter, from Polysciences, Inc.) for their mechanical immobilization, followed by sealing with a coverslip for complete physical immobilization 61 and taken to imaging: The PVD neuron was visualized by spinning disk confocal microscope, CSU-X, with Nikon eclipse Ti system and iXon3 camera (ANDOR).
Sequential z-series stacks (0.35 μm) from each worm were taken, with Plan Fluor 40X (NA 1.3). The cell body of the PVD was positioned in the middle of the field and the anterior and posterior branches were revealed. Images were captured with MetaMorph, version 7.8.1.0. Figures were prepared with Adobe Illustrator CS version 11.0.0.
Data analysis
The analysis of the PVD structure was performed for the ~200 μm surrounding the cell body. The pictures in TIFF format were analysed with ImageJ, version 1.48 (NIH). The images were converted to their negative form using the “convert lookup table (LUT)” function in ImageJ.
Maximal intensity projection was used for each PVD image.
Ectopic/excess branching defined as described previously 45. Briefly, ectopic branches defined as those branches that are not part of a wildtype “ideal” candelabrum (menorah) of a late L4 or young adult and those that sprouted at an incorrect position (non-quaternary terminal ends), as illustrated with dashed lines for ectopic branching in Fig. 2A. The total number of ectopic branching was calculated and presented as a fraction, out of non-ectopic branches.
The geometry of each quaternary branch (“candle”) was defined in the following manner: Straight geometry- all the pixels that constitute the branch are positioned on a straight line generated with ImageJ. The width of the line (1 pixel) was constant for the entire sets of experiments. The number of straight quaternary branches were divided by the total quaternary branches for each worm and presented as percentage. The analysis was done only for worms which did not move through the Z stack. Moving worms were excluded from the experiment.
Self-avoidance defects- the number of events where two adjacent candelabra overlapped (no gap formation), was divided by the total number of gaps between the candelabra within the frame (Fig. 2A). The self-avoidance values are presented as percentage.
Behavioral procedures
Harsh touch assay
After 72 h, adult worms (both isolated and those from the crowded plate as described below) were transferred using an eyelash separately to a new agar plate (each worm was transferred to a different plate) with 150 μl of fresh OP50 (about 16 h after seeding), in order to avoid accumulation of OP50 on the edge of the bacterial lawn, which might interfere with harsh touch response measurement. After ~45 minutes in the plate, the worms were prodded with a platinum wire posterior to the vulva, above the interface between worm’s body and the agar plate 31. The number of responses to harsh touch was counted every 10 sec. The non-responding worms were defined after two harsh touch events that were observed sequentially without response. More than one response was considered as responsive animal. The percentage of responding worms was calculated for each genotype and treatment. The experimenter was blind to both the genotype and the treatment-crowded or isolated.
Isolation of embryos
The worm isolation procedure was based on previous work 23 with few modifications, as indicated. Isolated animals were grown on 5 cm agar plate with 150 μl of OP50 E. coli, while crowded plate worms were grown on 600-700 μl. Adding sufficient amount of OP50, to avoid starvation of the worms in the crowded plate is important. The embryos and adult worms were isolated with platinum wire. The plates were sealed with one layer of Parafilm M and placed into a plastic box, at 20 °C, for the entire experiment. Three experimental groups were used for the 72 h (96 h of experiment was performed only for mec-4 worms, since they were L4-very young adults at 72 h) isolation experiment: (1) Single isolated embryos. (2) Crowded worms- the progeny of 30 young, non-starved, adults (approximately, 7,000-9,000 worms in different developmental stages, without approaching starvation). (3) Crowded adult worms that were isolated for certain amount of time as adults.
After 48/72/96 h (according to the experiment), age matched worms from each group were transferred to 10% agar pad slides for imaging, as described 23.
Isolation of adults
Isolation for 2 h
Crowded plate - the progeny of 30 adults per plate was used. After ~70 h in the crowded plate the worms were isolated using an eyelash into a new plate, with 150 μl of OP50, for 2 h, followed by imaging of the PVD.
Isolation for 5 h
Crowded plate - the progeny of 30 adults per plate was used. After ~67 h in the crowded plate the worms were isolated using an eyelash into a new plate, with 150 μl of OP50, for 5 h, followed by imaging of the PVD.
Isolation for 24 h
Crowded plate - the progeny of 30 adults per plate was used. After ~48 h in the crowded plate the worms (young adults) were isolated using an eyelash into new plates, with 150 μl of OP50, for 24 h, followed by imaging of the PVD or measurement of harsh touch response. The age of the worms was similar to the worms in experiments of embryo isolation for 72 h experiments (Fig. 1B). The isolated worms were compared to age matched worms from the crowded plate (their original plate).
Optogenetic stimulation
Crowded and isolated worms (ZX819: lite-1(ce314) X; xIs12[pF49H12.4::ChR2::mCherry; pF49H12.4::GFP]) were grown on OP50 with 100 μM of All Trans Retinal, in order to have functional channel Rhodopsin 24. After 72 h the worms were singly mounted with eyelash on a chunk (1cm × 1cm) of agar that was mounted on a microscope glass slide. The agar contained fresh but dry OP50, with 100 μM All Trans Retinal. About 30 minutes following the transfer the worms were tested for the response to light. Worms were stimulated with the spinning disk confocal microscope, CSU-X, with Nikon eclipse Ti system and iXon3 camera (ANDOR), at 488 nm wavelength, with laser intensity of 40% and exposure time of 100 ms, with 10X Plan Fluor (NA 0.3) for ~1 second and the forward acceleration response was measured. The experimenter determined the presence or the absence of forward acceleration response to light activation.
Isolation with chemical stimulation
The glp-4(bn2) mutants which are sterile at 25 °C were used (~40 worms for each plate), in order to prepare conditioned/chemically stimulated plates 42 prior to isolation of the ser-2Prom3::Kaede (BP709) strain. The glp-4 mutants were transferred at early larval stage (L1, L2) to a new agar plate for 96 h at 25 °C. After the removal of the glp-4 mutants, the embryo isolation procedure was used, as described before at 20 °C.
Isolation with glass beads
Single embryos were isolated to agar plates with 150 μl OP50 and 2.5 g of glass beads (1mm diameter) placed on the 150 μl OP50 lawn in the middle of the plate. The worms were isolated for 72 h and tested for response to harsh touch as described above.
Pharmacology
Amiloride hydrochloride hydrate (Sigma, #A7410) 1M stock solution in DMSO was stored at −20°C. A final concentration of 3 mM amiloride in 0.03 % DMSO was prepared in OP50 bacteria and seeded on NGM plates. 650 μl OP50 were seeded on each plate. As a control, 0.03 % DMSO was added to OP50 bacteria. For each plate (control 0.03% DMSO or 3 mM amiloride) 30 non-starved adult worms were added. After 72 h at 20 °C the progeny of the 30 adults were tested as young adults for their PVD morphology and their response to posterior harsh touch, as described in the previous sections.
Analysis of DEG/ENaCs localization in the PVD
Two DEG/ENaCs proteins, PF49H12.4::MEC-10::mCherry and PF49H12.4::DEGT-1::mCherry (the plasmids were provided by W. Schafer’s lab, 21 were analysed for their localization in the PVD, by comparing crowded to isolated worms, in a similar behavioral assay as described in Fig. 1b. The presence of the co-injection marker Punc-122::gfp was a prerequisite for analysis. For each individual worm a stack of images was examined and maximal intensity projection was performed. The images were encoded so the analysis was performed in a blind manner. The cell body of the PVD localized at the center of the region of interest with 60X Apochromat (NA 1.40) and the presence or absence of fluorescent signal was examined in three compartments: The cell body, the quaternary branches and the axon of the PVD. The results are shown as percentage of worms that localize the protein at any compartment.
Rescue strain
Worms from BP1022 (mec-10(tm1552) X; Is[hmnEx133](ser-2Prom3::kaede); him-5(e1490) V) were injected into the gonad with a rescuing plasmid for MEC-10, with a PVD specific promoter (pWRS825: ser-2Prom3::mec-10 genomic) kindly provided by W. Schafer’s lab 21. The injection mix contained myo-2::gfp (20 ng/μl) as a co-transformation marker and pWRS825 (80 ng/μl). For both behavioral and structural characterization the him-5;mec-10 strains, with and without rescuing plasmid, shared the same crowded plate and were differentiated by the presence or absence of the co-injection marker myo-2::gfp.
Statistics
For the morphological characterization of the PVD the results are expressed as means (blue circle) ± standard error of the mean (s.e.m.). In the boxplot (first, third quartiles) the upper whisker extends from the hinge to the highest value that is within 1.5 * IQR (inter-quartile range), the distance between the first and third quartiles. The lower whisker extends from the hinge to the lowest value within 1.5 * IQR of the hinge. The statistical analyses were performed with SPSS software (IBM, version 20) and “R package”. Two-tailed tests were performed for the entire data sets. Since for many experiments the distribution of the data was not normal, a-parametric tests were used: Mann Whitney test for comparison between independent groups. Kruskal-Wallis test was used for multiple comparisons for more than two groups. For proportions (percentage worms) ± standard error of proportion was calculated. Fisher’s exact test was used for analysis of differences in proportions. To estimate the variability in proportion we calculated the Standard Error of Proportion: The dot plot figures were prepared with “R package”, the bar charts with Excel software.
Discriminant analysis
Eight different strains (WT and seven DEG/ENaCs), with two treatments (crowded, isolated worms) for each strain were analyzed for Linear discriminant analysis for morphological characteristics, to evaluate similarity between different strains and treatments. Each worm in the data set was characterized by the three morphological characteristics (the fraction of ectopic branching, the percentage straight quaternary branches and the percentage self-avoidance defects). The centroid for morphological characterization was calculated for each condition and represented in square. Data from independent harsh touch experiments are shown for each group. The analysis was performed using SPSS 20.
Acknowledgments
We thank our lab members for their intellectual and technical support. Yael Iosilevskii, Veronika Kravtsov, Anna Meledin, Meital Oren-Suissa, Tom Shemesh, Shay Stern, Yehuda Salzberg and Alon Zaslaver for critically reading and commenting on the manuscript. Ehud Ahissar, Dan Cassel, Michel Labouesse and Kang Shen for fruitful discussions. William Schafer, Max Heiman and Alexander Gottschalk for plasmids and strains. Some strains were provided by the Caenorhabditis Genetics Center (CGC), which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). This work was supported by grants from the Israel Science Foundation (442/12 and 257/17), Adelis Fund (2023479), and the Ministry of Science Technology and Space (3-13022).