Fig 1.
Presynaptic neurons control the activity of the sleep-active RIS neuron.
(A) Sample trace of RIS activity and worm locomotion behavior outside of and during lethargus. RIS has no strong calcium transients outside of lethargus but shows strong activity transients during lethargus. Upon RIS activation, worms enter sleep bouts. (S1 Data, Sheet 1A). (B) Presynaptic neurons activate or inhibit RIS outside of and during lethargus. For statistical calculations, neural activities before the stimulation period (0–0.95 min) were compared to activity levels during the stimulation period (1–1.95 min). *p < 0.05, **p < 0.01, ***p < 0.001, Wilcoxon signed rank test. (S1 Data, Sheet 1B). (C) RIS activity decreases upon optogenetic PVC and RIM hyperpolarization. Statistical calculations were performed as described in panel B, but in experiments in which SDQL was stimulated, baseline activity levels were calculated over the time interval from 0.6 to 0.95 min. Baseline activity levels were calculated starting from 0.6 min as baseline activity levels were instable before that time point. *p < 0.05, **p < 0.01, Wilcoxon signed rank test. (S1 Data, Sheet 1C). (D) Circuit model of the RIS presynaptic regulatory network. Activating synaptic input is shown as green arrows, inhibitory synaptic input is shown as red arrows, and unclear synaptic input is shown as black arrow. Gap junctions are indicated as black connections. Neurons that are presynaptic to RIS present mostly activators. PVC is essential for lethargus-specific RIS activation. RIM can inhibit RIS through tyramine and FLP-18 and can activate RIS with glutamate. ΔF/F, change of fluorescence over baseline; FLP-18, FMRF-Like Peptide 18; GCaMP, genetically encoded calcium indicator; n.s., not significant; PVC, Posterior Ventral cord neuron class name; RIM, Ring Interneuron M class name; RIS, Ring Interneuron S class name; SDQL, Posterior lateral interneuron class name—left cell.
Fig 2.
PVC is an RIS activator that becomes resistant to inhibition during lethargus, but PVC activation is not sufficient for sleep induction.
(A) Simultaneous PVC and RIS GCaMP traces aligned to RIS peaks in fixed L1 lethargus worms. PVC activates before the RIS peak and stays active until the peak. *p < 0.05, **p < 0.01, Wilcoxon signed rank test. (S1 Data, Sheet 2A). (B) PVC hyperpolarization inactivates RIS and induces behavioral activity. PVC hyperpolarization was performed by expressing ArchT under the zk637.11 promoter. In contrast to the nmr-1 promoter, the zk637.11 promoter lacks expression in head command interneurons. **p < 0.01, ***p < 0.001, Wilcoxon signed rank test for GCaMP and speed, Fisher’s exact test for sleep fraction. (S1 Data, Sheet 2B). (C) During lethargus, PVC becomes resistant to inhibition. Outside of lethargus, its inhibition is stronger and continues beyond the end of optogenetic stimulation. During lethargus, PVC activity levels return back to baseline already during the stimulation period. *p < 0.05, **p < 0.01, Wilcoxon signed rank test. (S1 Data, Sheet 2C). (D) PVC activation translates into mostly a forward mobilization in L1 lethargus. *p < 0.05, ***p < 0.001, Wilcoxon signed rank test for Speed. Fisher’s exact test for fraction of direction. (S1 Data, Sheet 2D). ArchT, archaerhodopsin from Halorubrum strain TP009; ATR, all-trans-retinal; ΔF/F, change of fluorescence over baseline; GCaMP, genetically encoded calcium indicator; n.s., not significant; PVC, Posterior Ventral cord neuron class name; RIS, Ring Interneuron S class name.
Fig 3.
RIS and PVC activate each other, forming a positive feedback loop.
(A–E) RIS depolarization leads to a strong PVC depolarization outside of and during lethargus. This PVC depolarization is almost abolished in flp-11(tm2706), and it is significantly reduced in AVE-ablated worms. *p < 0.05, **p < 0.01, Wilcoxon signed rank test (S1 Data, Sheets 3A, 3B, 3C-E). (F) AVE-ablated worms show increased sleep. AVA-ablated worms do not show a significant sleep phenotype. Shown are sleep fractions during lethargus. *p < 0.05, Kolmogorov-Smirnov test (S1 Data, Sheet 3F). (G) RIS does not reach the same activation levels in aptf-1(gk794) and flp-11(tm2706) mutants compared to wild-type worms. aptf-1(gk794) and flp-11(tm2706) mutants neither immobilize nor sleep during RIS activation. ***p < 0.001, Welch test (S1 Data, Sheet 3G-I). (H) flp-11(tm2706) mutants have significantly fewer wide RIS peaks. aptf-1(gk794) mutants display the same amount of wide RIS peaks as wild-type worms. **p < 0.01, Kolmogorov-Smirnov test (S1 Data, Sheet 3G-I). (I) flp-11(tm2706) and aptf-1(gk794) mutants do not show sleep during lethargus. **p < 0.01, Kolmogorov-Smirnov test (S1 Data, Sheet 3G-I). (J) A circuit model for the positive feedback loop between RIS and PVC. Activating synaptic input is shown as green arrows, inhibitory synaptic input is shown as red arrows, and gap junctions are indicated as black connections. During wakefulness, reverse command interneurons inhibit PVC so that PVC does not activate RIS. During lethargus, PVC directly activates RIS, which then inhibits reverse command interneurons through FLP-11. This may speculatively disinhibit PVC, leading to a positive feedback. ΔF/F, change of fluorescence over baseline; FLP-11, FMRF-Like Peptide 11; GCaMP, genetically encoded calcium indicator; n.s., not significant.
Fig 4.
RIM activity peaks prior to sleep bouts, but RIM activation is not sufficient for sleep induction.
(A) RIM activates prior to sleep bouts. *p < 0.05, Wilcoxon signed rank test (S1 Data, Sheet 4A). (B–D) RIM-ablated worms have an increased sleep-bout frequency, while the sleep fraction and bout duration are not significantly changed during L1 lethargus. RIM was genetically ablated by expressing egl-1 under the tdc-1 promoter. *p < 0.05, Kolmogorov-Smirnov test (S1 Data, Sheet 4B-D). (E) RIM depolarization leads to increased mobility and reverse motion. *p < 0.05, ***p < 0.001, Wilcoxon signed rank test for speed. Fisher’s exact test for fraction of direction (S1 Data, Sheet 4E). (F–G) During lethargus, RIM becomes resistant to activation. RIM was optogenetically activated using ReaChR expressed under the tdc-1 promoter. Outside of lethargus, its activation is stronger (F). Activity levels during the stimulation period were quantified by subtracting baseline activity levels from levels during the stimulation period (G). *p < 0.05, **p < 0.01, Wilcoxon signed rank test for GCaMP and Kolmogorov-Smirnov test for quantification of stimulation levels (S1 Data, Sheet 4F-G). ATR, all-trans-retinal; ΔF/F, change of fluorescence over baseline; GCaMP, genetically encoded calcium indicator; n.s., not significant; ReaChR, red-activatable channelrhodopsin; RIM, Ring Interneuron M class name.
Fig 5.
The locomotion interneuron circuit controls RIS activation and sleep.
(A) Command interneurons are responsible for the majority of sleep. Command interneurons were genetically ablated by expressing ICE or egl-1 under the nmr-1 promoter. Command interneurons-ablated worms display a massive loss-of-sleep phenotype. ***p < 0.001, Welch test (S1 Data, Sheet 5A). (B) Hyperpolarization of command interneurons causes RIS inhibition and suppresses sleep. During lethargus, the hyperpolarization is followed by a strong post-stimulation activation of RIS. **p < 0.01, ***p < 0.001, Wilcoxon signed rank test for GCaMP and speed, Fisher’s exact test for sleep fraction (S1 Data, Sheet 5B). ΔF/F, change of fluorescence over baseline; GCaMP, genetically encoded calcium indicator; ICE, Caspase-1/Interleukin-1 converting enzyme; n.s., not significant; RIS, Ring interneuron S class name.
Fig 6.
RIS inhibition causes homeostatic rebound activation.
(A–B) A blue light stimulus leads to awakening and mobilization of C. elegans. Worms that go back to sleep after the stimulus show an activation rebound: pan-neuronal inhibition below baseline levels and RIS activation above baseline levels; “lethargus mobilizing” refers to animals that stayed awake and active during the post-stimulus time; “lethargus nonmobilizing” refers to animals that went back to sleep after the stimulation. *p < 0.05, **p < 0.01, ***p < 0.001, Wilcoxon signed rank test for GCaMP and speed, Fisher’s exact test for sleep fraction (S1 Data, Sheet 6A and 6B). (C) RIS shows rebound activation following hyperpolarization. Behavioral and brain activity measurements correlate throughout the whole experiment. *p < 0.05, **p < 0.01, ***p < 0.001, Wilcoxon signed rank test for GCaMP and speed, Fisher’s exact test for sleep fraction (S1 Data, Sheet 6C). (D–E) Dose-response curve of optogenetic RIS hyperpolarization with different stimulus lengths. RIS activation rebound transients saturate with increasing length of inhibition. Worms not showing a rebound activation transient after RIS optogenetic hyperpolarization were excluded from the analysis. Numbers of worms not responding were as follows: (1) In experiments in which RIS was optogenetically inhibited for 48 s, all worms showed an RIS rebound activation transient. (2) In experiments in which RIS was optogenetically inhibited for 5 min, 1 out of 7 worms did not show a RIS rebound activation transient. (3) In experiments in which RIS was optogenetically inhibited for 10 min, 1 out of 13 worms did not show an RIS rebound activation transient. Curve in D was fitted as an asymptotic function, and curve in E was fitted as a BoxLucas1 function (S1 Data, Sheet 6D, E). ΔF/F, fluorescence change over baseline; GCaMP, genetically encoded calcium sensor; n.s., not significant; R, fluorescence of GCaMP divided by fluorescence of mKate2; RIS, Ring Interneuron S class name.
Fig 7.
The dampening of neural and behavioral baseline activity levels during lethargus is independent of RIS function.
Reduction of command interneuron activity levels during lethargus occurs in wild-type worms and aptf-1(gk794) mutants. In the wild-type condition, activity levels are reduced to −0.16 ± 0.02. In the mutant condition, activity levels are reduced −0.08 ± 0.02. **p < 0.01, Wilcoxon signed rank test (S1 Data, Sheet 7A and 7B). ΔF/F, fluorescence change over baseline; GCaMP, genetically encoded calcium indicator; RIS, Ring Interneuron S.
Fig 8.
Arousing stimulation inhibits RIS and sleep through RIM.
(A) ASH depolarization in wild-type worms leads to RIS inhibition and RIM activation, sleep suppression, and mobilization. *p < 0.05, **p < 0.01, ***p < 0.001, Wilcoxon signed rank test for GCaMP and speed, Fisher’s exact test for sleep fraction (S1 Data, Sheet 8A, B). (B) ASH depolarization in RIM-ablated worms leads to weaker sleep suppression, mobilization, and RIS activation. *p < 0.05, **p < 0.01, ***p < 0.001, Wilcoxon signed rank test for GCaMP and speed, Fisher’s exact test for sleep fraction (S1 Data, Sheet 8A, B). (C) The response direction following ASH activation in wild-type worms is predominantly reverse, while in RIM-ablated worms it is predominantly forward. ***p < 0.001, Fisher’s exact test (S1 Data, Sheet 8C). (D) A circuit model for RIS regulation through arousal by ASH. Activating synaptic input is shown as green arrows, inhibitory synaptic input is shown as red arrows, and gap junctions are indicated as black connections. RIM could serve as a synchronizer of AVE and AVA to regulate PVC and therefore RIS inhibition. Additionally, RIM could inhibit RIS directly. ΔF/F, fluorescence change over baseline; GCaMP, genetically encoded calcium indicator; n.s., not significant.
Fig 9.
A circuit model for RIS activation through locomotion interneurons.
(A) Activating synaptic input is shown as green arrows, inhibitory synaptic input is shown as red arrows, and gap junctions are indicated as black connections. Outside of lethargus, the nervous system cycles between forward and reverse states. RIS is not activated sufficiently to cause a sleep bout, neither during the forward state during which PVC is active nor during the reversal state during which RIM is active. The locomotion circuit activates RIS briefly to cause a locomotion pause at the transition from forward to reverse movement. Speculatively, the circuit that controls RIS during sleep also controls RIS during locomotion pauses. (B) During lethargus motion bouts, the nervous system still cycles between forward and reverse states. Baseline activity and excitability in RIM are reduced, and PVC becomes resistant to inhibition and more potent to activate RIS. These changes in locomotor interneurons shift the balance to favor strong RIS activation and induction of a sleep bout, a process that may involve simultaneous activation from multiple neurons, including RIM and PVC. Such an overlap activation of RIS by otherwise mutually exclusive neurons could occur at the transition from forward to reverse locomotion states. Perhaps, RIS activation and sleep could occur similarly at the transition from reverse to forward locomotion states.