Plasticity of the Electrical Connectome of C. elegans

Abstract

The specific patterns and functional properties of electrical synapses of a nervous system are defined by the neuron-specific complement of electrical synapse constituents. We systematically examined the molecular composition of the electrical connectome of the nematode C. elegans through a genome- and nervous-system-wide analysis of the expression patterns of the invertebrate electrical synapse constituents, the innexins. We observe highly complex combinatorial expression patterns throughout the nervous system and found that these patterns change in a strikingly neuron-type-specific manner throughout the nervous system when animals enter an insulin-controlled diapause arrest stage under harsh environmental conditions, the dauer stage. By analyzing several individual synapses, we demonstrate that dauer-specific electrical synapse remodeling is responsible for specific aspects of the altered locomotory and chemosensory behavior of dauers. We describe an intersectional gene regulatory mechanism involving terminal selector and FoxO transcription factors mediating dynamic innexin expression plasticity in a neuron-type- and environment-specific manner.

Keywords: Caenorhabditis elegans; connectome; electrical synapse; innexins; synaptic plasticity.

Fig. 1:. Innexin gene expression in non-dauer and dauer.

(A) Schematics of subunit composition in different kinds of gap junction channels. Hemichannels formed by innexins can also be octameric (Oshima et al., 2016). (B) Expression of all neuronally expressed innexin genes in all neuronal classes. Color of each box represents whether the innexin gene is expressed in both non-dauer and dauer (yellow), only in non-dauer (red) or only in dauer stage (green). Asterisks indicate developmental changes in innexin expression among non-dauer stages. See also Table S2. (C) Abundance of each innexin gene in non-dauer and dauer nervous system. (D) Distribution of number of innexin genes in neuron classes in non-dauer and dauer. See also Fig.S1-3, Table S1,S2.

Fig. 2:. Changes in innexin gene expression in dauer.

(A) Electrical connectome at the adult stage, as inferred from the serial section EM reconstruction. Neuronal classes are represented as circles and electrical synapses are shown as gray lines. Thicknesses of lines were weighted according to the volume of each connection (measured by the total number of EM sections in which a particular synapse was observed; as per www.wormwiring.org). The number of innexin genes expressed in each neuronal class is indicated by the overlaid pie chart. Color of each pie chart section represents whether the innexin is expressed in both non-dauer and dauer (yellow), only in non-dauer (red) or only in dauer (green). (B-D) Neuronal identities were determined by expression of either cho-1 (otIs544) or unc-47 (otIs564) reporters. (B) unc-7 fosmid reporter (otEx7106) is expressed in IL2DL/R, IL2VL/R and IL2L/R in non-dauer, but selectively down-regulated in IL2L/R in dauer. (C) che-7 fosmid reporter (otEx7112) is selectively turned on in lateral IL2L/R neurons in dauer. (D) unc-7 fosmid reporter (otEx7106) is expressed in I2, but not in NSM in non-dauer. In dauer, unc-7 expression disappeared in I2, while turned on in NSM. See also Fig. S2.

Fig. 3:. Dauer-induced expression of inx-6 in AIB generates new gap junctions with che-7.

(A) Schematics of inx-6 transcriptional reporter allele, inx-6(ot804) and GFP-tagged inx-6 translational reporter allele, inx-6(ot805). (B) inx-6 reporter allele, inx-6(ot804) is additionally turned on in AIB interneurons in dauer that disappears in post-dauer stage. inx-6 allele is also expressed in AIB in L1d. (C) smFISH against endogenous inx-6 mRNA (magenta) in Fed-L1 and L1d. AIB was marked by eat-4::yfp (otIs388) expression (green). Inset shows enlargement of AIB. (D) Quantification of inx-6 smFISH data. Each circle represents the number of AIB associated smFISH puncta in a single animal, red lines indicate the mean and rectangles indicate S.E.M. Wilcoxon rank-sum tests p-values (n.s. = non significant and ****p<0.0001). (E) Expression of INX-6 puncta (green), in inx-6(ot805) dauer. npr-9p::tagrfp (otIs643) expression marks AIB processes (magenta). Arrowheads mark INX-6::GFP at the crossover points of AIBL and AIBR. Asterisks mark INX-6 puncta in pharyngeal muscles. (F) Expression of INX-6 puncta (green) on AIBL (magenta) in daf-7(e1372) dauers (inx-6(ot805); daf-7(e1372); otIs643), where AIBR was ablated. In absence of AIBR, INX-6 punctum at the AIBL-AIBR crossover point (right box) disappear, while has no effect on INX-6 puncta in other region (left box). See Fig. 3E for control (pre-ablation) image. Four INX-6 puncta that do not overlap with AIB-processes are among pharyngeal muscles. (G) Schematic of major INX-6 puncta (green circles) on AIB. Arrowheads mark INX-6 puncta at the AIBL-AIBR crossover points. (H) Schematic of the effect of AIBR-ablation on INX-6 puncta on the remaining AIBL neuron. See Fig. 3G for control (pre-ablation) schematic.

Fig. 4:. inx-6 and che-7 forms gap junction between AIB and BAG.

(A) o-localization of INX-6::GFP (green) and CHE-7::TagRFP puncta (magenta) in two putative gap junctions on AIB (yellow circle) in dauer (ot805; otEx6486). (B) Schematic of CHE-7 puncta (magenta circles) localization with INX-6 puncta (green circles) on AIB. Arrowheads mark INX-6 puncta at the AIBL-AIBR crossover points. (C) INX-6 puncta that colocalized with CHE-7 were lost in che-7(ok2373); inx-6(ot805) mutant dauers (magenta circles). INX-6 puncta at the AIBL-AIBR crossover points remained unaffected (arrowheads). (D) Quantification of INX-6 puncta on AIB in che-7(ok2373) dauers. (E) Schematics of 5’ cis-regulatory element analysis of che-7 and reporter transgenes used for cell-specific che-7 expression. Transgenic lines were created in che-7(ok2373); inx-6(ot805) background. Results show average for all transgenic lines. (F-G) INX-6 puncta (green) that co-localize with CHE-7 in dauer (dotted circles) do not colocalize with ASG (ot805;oyIs47) and AWC (ot805;otIs263) axons (magenta). (H) In che-7(ok2373) dauer, expression of CHE-7::TAGRFP (magenta) only in BAG (green) rescues CHE-7 associated INX-6 puncta (green) that also shows colocalization with CHE-7::TAGRFP (arrowheads). See Fig.4C for control image showing INX-6 puncta in che-7(ok2373) dauer. (inx-6(ot805); che-7(ok2373); otEx2487) (I) INX-6 puncta (green) in dauer co-localize with BAG axon (magenta). (ot805; otEx7230) (J) INX-6 puncta (green) in L1d co-localize with BAG axon (magenta). For BAG-axon image clarity, a projection of non-continuous z-sections was shown. (ot805; otEx7230) (K) BAG axon (green) comes in contact with AIB process (magenta) at similar positions where INX-6 puncta co-localize with BAG axon. Inset shows enlargement of the contact point. (otIs643; otEx7230) (L, O and P) INX-6 puncta present at the AIBL-AIBR crossover points are referred as ‘AIB crossover point’ and INX-6 puncta that co-localize with CHE-7 and BAG axon are referred as ‘che-7 colocalizing’. (L) Quantification of INX-6 puncta on AIB in ets-5(tm1794) mutant dauers. (M) TEM prints from wild-type adult hermaphrodite ‘N2U’ showing adjacency of BAG (pseudo colored in red) and AIB (pseudo colored in green) processes at the site where dauer specific INX-6-CHE-7 putative electrical synapses are formed. These images were collected in MRC/LMB and annotated images were obtained from www.wormimage.org, courtesy of David Hall. Prints shown here are, Left: N2U_094 and right: N2U_116. (N) AIB-specific ectopic expression of INX-6 in non-dauer stage (L3) results in INX-6 puncta (green) along the AIB (magenta). Red circles mark INX-6 at the AIBL-AIBR crossover points. - Yellow circles mark INX-6 at the site where dauer specific INX-6-CHE-7 electrical synapses are formed. (otTi19; otIs643) (O) Quantification of ectopic INX-6 puncta in non-dauer otTi19 animals. (P) Quantification of ectopic INX-6 puncta in non-dauer che-7(ok2373); otTi19 animals.

Fig. 5:. Locomotory behavior is remodeled in dauer.

(A) Principal Component Analysis of dauer (red), fed-L3 (yellow) and starved-L3s (green) based on 195 locomotory behavior feature data (Listed in table S3). Circles represent individual animals (n^dauer = 40, n^Fed-L3 = 33, n^starved-L3 = 26). Component 1 and 2 account for ~40% of the variation in the locomotory behaviors. (B) Variance explained by first ten PCs. Black line represents cumulative variance explained. (C) Schematics of foraging (nose bend), head bend, reversal through an omega turn and coiling behaviors. (D-K) Comparison of dauer, fed- and starved-L3 locomotion using Wormtracker (see Methods for details). Each circle represents the experimental mean of a single animal. Red lines indicate the mean of means and rectangles indicate S.E.M. Wilcoxon rank-sum tests and False-Discovery Rate q-values for each comparison: n.s. = non significant, *q<0.05, **q<0.01, ***q<0.001, ****q<0.0001. (Behavioral feature time ratio = total time spent performing particular behavior/total time of the assay) See also Table S3 and S4.

Fig. 6:. Loss and gain of innexin expression in dauers affects locomotory and CO₂-attraction behavior.

(A-I) Locomotion assay using Wormtracker (see Methods for details). Each circle represents the experimental mean of a single animal. Red lines indicate the mean of means and rectangles indicate S.E.M. Wilcoxon rank-sum test p-values for each comparison: n.s. = non significant, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. (Behavioral feature time ratio = total time spent executing particular behavior/total assay time). (A-F) Locomotion of dauer animals. (A-C) Dauers with inx-6 temperature sensitive allele, inx-6(rr5), at restrictive temperature and AIB-ablated dauers [peIs578 (Wang et al., 2017)] show significant increase in pausing and decrease in forward motion. (D-F) Dauers that specifically lack inx-6 expression in AIB, inx-6(AIB OFF) allele (see Fig. 8) and che-7(ok2373) dauers show significant increase in pausing and decrease in forward motion, but show no effect in backward motion. (G-I) inx-6(AIB OFF) starved-L3s show no significant difference in locomotion. AIB-specific ectopic expression of INX-6 in starved-L3 (otTi19) affects pausing and forward motion. (J) Schematic of CO₂-chemotaxis assay (See Methods for details). (K-L) Each circle represents chemotaxis index calculated from a single assay. Red lines indicate the mean and rectangles indicate S.E.M. Wilcoxon rank-sum tests p-values for each comparison: n.s. = non significant, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. (K) inx-6(AIB OFF) or che-7(ok2373) mutant young adults show no effect in CO₂-repulsion. Ectopic expression of INX-6 in AIB (otTi19) in young adults do not affect CO₂-repulsion. (L) inx-6(AIB OFF) and che-7(ok2373) dauers show reduced CO₂-attraction. Dauers with ablated BAG neurons (kyIs536; kyIs538) do not chemotax to CO₂ gradient. See also Fig. S4. (M) Expression of the unc-7 fosmid reporter (otEx7106) is lost in AVA in dauer, while continue to be expressed in RMD and AVE. cho-1 (otIs544) expression identifies the neuron. (N-O) Locomotion of two independent transgenic lines where unc-7 is ectopically expressed in AVA in dauers (Transgenic Line-2: otEx7250 and Line-3: otEx7251). See also Fig. S4.

Fig. 7:. UNC-42 and DAF-16/FOXO regulate spatiotemporal expression of inx-6 in AIB.

(A) Schematics of 5’ cis-regulatory element analysis of inx-6. Results show average for all transgenic lines. Deletion of a putative UNC-42 binding site (TAATTA) in the 5’ upstream regulatory region resulted in complete loss of inx-6 expression in AIB. (B-J) For each graph, circles represent the expression of corresponding reporter in a single animal and red lines indicate the mean. Expression of reporters are scored as: ON = similar to control, DIM = reduced expression and OFF = no expression. (B) UNC-42 affects inx-6 reporter (otIs484) expression in AIB in dauer. (C) unc-42 fosmid reporter (wgIs173) (green) is expressed in AIB in all stages. eat-4 (otIs518) (red) expression was used for neuronal identification. (D-F) Expression of eat-4, npr-9 and odr-2 in AIB were affected in unc-42(e419) mutant animals. (G) Deletion of a putative UNC-42 binding site in inx-6(ot840) animals, results in loss of inx-6 expression specifically in AIB in dauer and L1d. See Fig.3B for control images. (H) Quantification of inx-6 fosmid reporter (otIs473) expression in dauer. inx-6 expression is lost in AIB in daf-7(e1372); daf-16(mgDf50) double mutant dauers. (I) AIB specific degradation of DAF-16 in auxin-treated dauers, results in loss of DAF-16::mNeonGreen as well as inx-6 (yellow) expression in AIB. In EtOH-treated control dauers, expression of DAF-16, as well as inx-6 are maintained. Due to substantial overlap of mNG and YFP emission spectra these two expressions could not be separately imaged. [inx-6(ot804); daf-2(e1370); daf-16(ot853); otEx7309] (J) Quantification of results shown in panel I.