cpx-1 mutants display impaired locomotion

The C. elegans genome contains 2 complexin genes: cpx-1 and cpx-2. cpx-1 encodes the CPX-1 protein, which has a relatively high degree of homology to mammalian complexins and is considered the major complexin functioning in the nervous system in C. elegans [19,22]. The CPX-1 protein consists of four functional domains, the N-terminal domain (NTD), accessory domain (AD), central α-helix (CH), and C-terminal domain (CTD) (Fig 1A), each of which play distinct functional roles [16,19,23,26,49–52]. To examine the contribution of CPX-1, we investigated motor behavioral defects caused by cpx-1 mutations via various assays.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. cpx-1 mutants display defective locomotion and increased body curvature.

(A) Cartoon representation of the structure of CPX-1, indicating the NTD, AD, CH, and CTD. Mutation sites of ok1552 (green line), syb3665 (blue stars), and syb3584 (brown line) are labeled. (B) Motility measured by thrashing number per minute across the wild type, cpx-1(ok1552), cpx-1(KY/AA, syb3665), and cpx-1(Δ12, syb3584). n ≥ 40 animals. One-way ANOVA was performed (F_{(3, 156)} = 308.1, P < 0.0001). (C) Schematic representation of the method used to measure body curvature. (i) Each free-moving worm was imaged for 20 s at 10 Hz, and 200 images showing the continuous locomotion of the worm were obtained. (ii) The worm body in each frame was divided into 32 body segments, and the curvature was calculated via MATLAB. The curvatures of the body segments at 15%–85% (ϕ6—ϕ28) were used for subsequent analysis. (iii) The curvature obtained in (ii) is plotted as a color map. Each pixel represents the average of 200 images of the curvature of a body segment. (D) Representative body bending states (top) and curvature (bottom) in the indicated genotypes. (E) Distribution and quantification of body curvature at the indicated strains. Each point represents the average of the 15%–85% curvature of the body segments for 20-s-locomotion for each worm. n ≥ 35 animals. One-way ANOVA was performed (F_{(2, 109)} = 30.66, P < 0.0001). (F) Quantification of locomotion velocities in wild-type, cpx-1(ok1552), and cpx-1(Δ12, syb3584) worms. n ≥ 33 animals. One-way ANOVA was performed (F_{(2, 126)} = 285.5, P < 0.0001). Data are presented as box-and-whisker plots, with the median (central line), 25th–75th percentile (bounds of the box), and the 5th–95th percentiles (whiskers) are indicated in (B, E, F). Student’s t test was performed for comparisons of 2 groups, ### P < 0.001. One-way ANOVA was used for comparisons of multiple groups, followed by Tukey’s range test, *** P < 0.001; ns, not significant. The error bars represent the standard error of the mean (SEM). N = 3 independent replicates. All the raw data associated with this figure are available in S1 Data. AD, accessory domain; CH, central α-helix; CTD, C-terminal domain; NTD, N-terminal domain.

https://doi.org/10.1371/journal.pbio.3003023.g001

We first found that a null mutant allele, cpx-1 (ok1552), which contains a large fragment deletion of sequences encoding the N-terminal domain, accessory domain, central α-helix, and part of the C-terminal domain (Figs 1A and S1A), showed a reduced ability to coordinate body oscillations during thrashing assays (Fig 1B). Next, we used an aldicarb-induced paralysis assay to evaluate whether the synaptic transmission of cpx-1 mutants is altered. Aldicarb is an acetylcholinesterase inhibitor that prevents the removal of acetylcholine at the synapse [53,54]. Increased aldicarb sensitivity is commonly associated with heightened excitatory transmission or diminished inhibitory transmission in C. elegans, fundamentally stemming from a disruption in the E/I balance [54,55]. Compared with wild-type N2 animals, the cpx-1(ok1552) mutant exhibited significantly greater aldicarb sensitivity (S1B). These findings suggest that motor behavior is impaired in the absence of CPX-1.

To further explore the function of CPX-1, we generated 2 additional knock-in (KI) mutants via CRISPR/Cas9 via the proposed essential domains. The first mutation involves the deletion of the C-terminal 12 residues (Δ12, syb3584) (S1A Fig), which are suggested to be critical for its clamping function in suppressing spontaneous release [23–26]. The second mutant, KY/AA (syb3665), targets the central α-helix (S1A Fig), which is a key domain for SNARE binding [19,30,49,56]. We compared the thrashing and aldicarb phenotypes of these 2 KI mutants with those of wild-type and cpx-1(ok1552) worms. cpx-1(Δ12, syb3584) exhibited significant but mild thrashing defects and heightened aldicarb sensitivity (Figs 1B and S1B). Conversely, cpx-1(KY/AA, syb3665) worms presented thrashing defects and aldicarb sensitivity identical to those of cpx-1(ok1552) mutants (Figs 1B and S1B). These behavioral defects further demonstrate that CPX-1 is essential for regulating locomotion coordination.

Additionally, cpx-1 (ok1552) worms presented a body bending defect (S1 and S2 Movies). Quantitative analysis of the average body curvature in the free-moving worms (Fig 1C) revealed that the cpx-1(ok1552) mutant presented a hypercontraction muscle phenotype, similar to that observed in the aldicarb assay and when cholinergic motor neurons were overactivated [54,57]. To exclude errors due to unstable head and tail oscillations, the curvature was calculated for 15% to 85% of the body segments (Fig 1Cii). A significant increase in body curvature was also observed in cpx-1(Δ12) worms (Fig 1D and 1E). Consequently, increased body bending was accompanied by severe locomotion defects, with a significant reduction in the speed of movement (Fig 1F). These results demonstrate that CPX-1 regulates locomotion coordination and suggest that there is an underlying imbalance between excitatory and inhibitory synaptic transmission in cpx-1 mutants.

cpx-1 mutations lead to an E/I imbalance in synaptic transmission

Most electrophysiological analyses of synaptic transmission in C. elegans have been limited to excitatory transmission [19,22,47,48], leaving the mechanism of inhibitory synaptic release less well understood. We examined the roles of complexin/CPX-1 in both excitatory and inhibitory neuromuscular junctions (NMJs) in C. elegans. Both spontaneous release and evoked release were measured. We used cpx-1 mutants with zxIs6 (excitatory synapse) or zxIs3 (inhibitory synapse) lines to measure exocytosis via optogenetics [57]. The selective expression of channelrhodopsin in excitatory cholinergic or inhibitory GABAergic neurons enabled reliable measurement of exocytosis underlying neurotransmission.

cpx-1 null mutants reportedly exhibit enhanced spontaneous release and reduced evoked release from excitatory motor neurons [19,22]. Compared with wild-type animals, cpx-1(ok1552) increased the frequency and amplitude of miniature postsynaptic currents (mPSCs) (Fig 2A–2C). Conversely, the evoked excitatory postsynaptic current (eEPSC) recorded in cpx-1(ok1552) mutants was drastically lower than that recorded in wild-type animals (Fig 2D–2F). Thus, C. elegans CPX-1 aligns with the widely accepted perspective of its dual functions—facilitating evoked release while inhibiting spontaneous release at excitatory synapses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. cpx-1 mutations lead to an imbalance between excitatory and inhibitory synaptic transmission.

(A) Representative traces of mPSCs recorded from wild-type, cpx-1(ok1552), and cpx-1(Δ12) worms at different timescales. Recordings were made from body wall muscle cells at −60 mV. (B, C) Quantification of the mPSC frequency (B) and amplitude (C) in each genotype. n ≥ 11 animals. One-way ANOVA was performed (F_{(2, 45)} = 26.01, P < 0.0001 for B; F_{(2, 45)} = 11.94, P < 0.0001 for C). (D) Representative traces of eEPSCs recorded from the indicated genotypes in the zxIs6 [Punc-17::ChR2::YFP + lin-15(+)] transgenic background with 10 ms blue light illumination (blue arrow). The lower trace of cpx-1(ok1552) (green) shows a zoomed-in region from the upper trace. (E, F) Quantification of eEPSC amplitude (E) and charge transfer (F) at different strains. The numbers after Δ represent the ratio of the average values of WT and cpx-1 (ok1552). n ≥ 13 animals. One-way ANOVA was performed (F_{(2, 40)} = 65.01, P < 0.0001 for E; F_{(2, 40)} = 50.07, P < 0.0001 for F). (G) Representative traces of mIPSCs recorded from wild-type, cpx-1(ok1552), and cpx-1(Δ12) worms. The muscles were held at −10 mV with the 0.5 mM ionotropic acetylcholine receptor blocker D-tubocurarine (d-TBC). (H, I) Quantification of the mIPSC frequency (H) and amplitude (I) in each genotype. n ≥ 10 animals. One-way ANOVA was performed (F_{(2, 33)} = 2.313, P = 0.1148 for H; F_{(2, 33)} = 2.442, P = 0.1026 for I). (J) Representative traces of inhibitory postsynaptic currents (eIPSCs) recorded from the indicated strains in the zxIs3 [Punc-47::ChR2::YFP + lin-15(+)] transgenic background with 10 ms blue light illumination (blue arrow). (K, L) Quantification of the eIPSC amplitude (K) and charge transfer (L) at the indicated strains. The numbers after Δ represent the ratio of the average values of WT and cpx-1 (ok1552). n ≥ 9 animals. One-way ANOVA was performed (F_{(2, 32)} = 43.37, P < 0.0001 for K; F_{(2, 32)} = 8.225, P = 0.0013 for L). (M) Schematic representation illustrating the imbalance of excitatory and inhibitory synaptic transmission due to cpx-1 mutations and resulting behavioral defects. The data are presented as box-and-whisker plots, with the median (central line), 25th–75th percentile (bounds of the box), and 5th–95th percentile (whiskers) indicated. One-way ANOVA was used for comparisons of multiple groups, followed by Tukey’s range test, * P < 0.05; *** P < 0.001; ns, not significant. The error bars represent the SEM. N = 3 independent replicates. All the raw data associated with this figure are available in S1 Data. eEPSC, evoked excitatory postsynaptic current; eIPSC, evoked inhibitory postsynaptic current; mIPSC, miniature inhibitory postsynaptic current; mPSC, miniature postsynaptic current.

https://doi.org/10.1371/journal.pbio.3003023.g002

Given that C. elegans NMJs receive both cholinergic and GABAergic inputs simultaneously, mixed excitatory (predominantly) and inhibitory (minor) mPSCs were collected under our recording conditions [58]. To isolate GABAergic inhibitory transmission in cpx-1(ok1552), we employed D-tubocurarine (d-TBC, 0.5 mM), a broad-spectrum blocker of C. elegans ionotropic AChRs, to isolate miniature inhibitory postsynaptic currents (mIPSCs) [58,59]. Interestingly, compared with those in wild-type worms, neither the frequency nor the amplitude of mIPSCs was altered in cpx-1(ok1552) mutants (Fig 2G–2I). Thus, the increased mPSCs observed in the cpx-1 mutant could be attributed to its function in excitatory synapses. Nevertheless, evoked inhibitory postsynaptic currents (eIPSCs) were decreased in the cpx-1(ok1552) mutant (Fig 2J–2L). Therefore, unlike excitatory NMJs, where a lack of CPX-1 results in decreased evoked release coupled with enhanced spontaneous release, inhibitory synapses do not exhibit this coupling.

To delve deeper into the distinct impacts of cpx-1(ok1552) on excitatory and inhibitory transmission, we conducted a comparative analysis of its preferential influence on the 2 synapses. While mIPSCs were unaltered in cpx-1(ok1552) mutant animals, the frequency of mPSCs increased, indicating that spontaneous release at the NMJ without CPX-1 was driven to a heightened excitatory state (Fig 2M, spontaneous release). For evoked release, we analyzed the relative magnitude changes (Δ = I_WT/I_cpx-1) and found that the peak amplitude of eEPSC decreased by a factor of Δ = 11.56, whereas the reduction in eIPSC was only Δ = 3.04 (Fig 2E and 2K) in cpx-1(ok1552) mutants. This yielded a notable 3.8-fold disparity in the decrease in excitatory evoked transmission compared with inhibition. A similar trend was observed when we compared the relative change in transferred charges, which showed a 4.1-fold disparity between eEPSCs (Δ = 6.57) and eIPSCs (Δ = 1.59) (Fig 2F and 2L). Therefore, evoked transmission in cpx-1(ok1552) mutants preferentially weakened excitatory strength compared with inhibition (Fig 2M, evoked release).

Collectively, these findings indicate that cpx-1 mutation enhances the frequency and amplitude of miniature PSCs but not inhibitory miniature PSCs. These findings imply that complexin/CPX-1 had a more pronounced influence on excitatory synapses than on inhibitory synapses (E > I) at the neuromuscular junction. Thus, cpx-1 mutations lead to an imbalance between excitatory and inhibitory synaptic transmission, which in turn leads to motor behavioral deficits (Fig 2M). Considering that most components of the release machinery, including complexin, participate in both excitatory and inhibitory synapses, the predominant function of CPX-1 in excitatory synapses rather than inhibitory synapses suggests a coordinated or balanced mechanism between CPX-1 and other release machinery (see Discussion).

Clamping-specific mutant cpx-1(Δ12) prominently enhances spontaneous excitatory release but does not affect evoked release

The decrease in excitatory synchronous release may have been due to the increase in spontaneous release depleting the primed vesicles [60]. However, the opposite regulation by CPX-1 of spontaneous and evoked release indicates that its functional domain mechanism should be segregated [19,22,61]. To date, no clamping-specific mutant that preserves evoked release has been identified in C. elegans. Therefore, we aimed to generate such a mutant in this study. Compared with the cpx-1(ok1552) mutant, the cpx-1(Δ12) mutant presented better motility (Fig 1B and S2 and S3 Movies) and heightened aldicarb sensitivity (S1B Fig); both phenotypes are consistent with genetic lesions selectively disrupting the clamping ability of CPX-1 without altering evoked synaptic transmission.

By measuring synaptic transmission via electrophysiology, we found that cpx-1(Δ12) retained normal evoked release in both excitatory (Fig 2D–2F) and inhibitory synapses (Fig 2J–2L). As expected with high aldicarb sensitivity, cpx-1(Δ12) strongly increased the frequency and amplitude of mPSCs (Fig 2A–2C). Similar to cpx-1(ok1552) worms, cpx-1(Δ12) mutants did not alter the frequency or amplitude of mIPSCs (Fig 2G–2I). Given that evoked release was normal in cpx-1(Δ12), we demonstrated that evoked and spontaneous release can be segregated in complexin/CPX-1, where increased spontaneous activity does not necessarily deplete primed vesicles for synchronous release.

Furthermore, cpx-1(Δ12) worms specifically promoted spontaneous release in excitatory synapses without affecting inhibitory neurotransmission, leading to an E-I imbalance that dominated excitatory transmission (Fig 2M). In summary, CPX-1 exerted a stronger modulatory effect on excitatory synapses than on inhibitory synapses in C. elegans.

CAPS/UNC-31 differentially regulates evoked excitatory and inhibitory neurotransmission

We further investigated other regulatory partners in these 2 synapses. Prominent proteins in the presynaptic secretion machinery, such as Munc13, Munc18, and Synaptotagmins, have been extensively studied [27,62–64]. However, the impact of unc-31 mutants on excitatory evoked release has yielded contradictory results [42,46–48], and their phenotype in inhibitory synapses has not been thoroughly investigated. Here, we examined the unc-31(e928)-null mutant to evaluate the role of UNC-31 in E/I balance regulation.

First, we assessed the body curvature of unc-31(e928)-null mutants and found that it was significantly lower than that of the wild type (Fig 3A–3C and S1 and S4 Movies). The unc-31 mutant animals exhibited a rigid and straight posture, suggesting excessive muscle relaxation, in contrast to the cpx-1 mutants (S2 and S4 Movies). Electrophysiological analyses revealed that unc-31(e928) did not impact spontaneous release in either excitatory or inhibitory synapses, as evidenced by the unchanged frequency and amplitude of mPSCs and mIPSCs (Fig 3D–3F and 3J–3L). However, it decreased both eEPSC and eIPSC (Fig 3G–3I and 3M–3O). These findings diverged from the observations of Speese and colleagues, who noted no alterations in excitatory evoked release [42], but were consistent with Gracheva and colleagues, who reported a reduction in eEPSCs [47]. The relative decreases in eEPSC amplitude (Δ = 2.15, Fig 3H) and charge transfer (Δ = 3.95, Fig 3I) were greater than those of eIPSC (Δ = 1.36, Fig 3N; Δ = 1.54, Fig 3O) when I_WT/I_unc-31 was analyzed. Therefore, unc-31(e928) tilted the E/I balance toward a stronger inhibitory side (Fig 3P), potentially explaining the phenotype of unc-31(e928) mutants, which exhibited a rigid and straight posture indicative of stronger muscle relaxation [41]. Furthermore, the defect in excitatory evoked release in the unc-31(e928) mutant could be rescued by expressing UNC-31 cDNA in cholinergic motor neurons (S2 Fig), suggesting that the UNC-31 regulation is largely cell autonomous. Thus, similar to that of CPX-1, the loss of UNC-31 induced different phenotypes in excitatory and inhibitory synapses, with both proteins predominantly regulating excitatory synapses but having a lesser impact on inhibitory synapses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. UNC-31/CAPS differentially regulates evoked cholinergic and GABAergic neurotransmission.

(A) Representative body bending states (top) and curvature (bottom) in wild-type and unc-31(e928) animals. (B, C) Quantification (B) and distribution (C) of body curvature in WT and unc-31 mutant animals. n ≥ 37 animals. (D) Representative traces of mPSCs recorded from WT, unc-31(e928), egl-3(tm1377), and unc-31(e928); egl-3(tm1377) worms at different timescales. (E, F) Quantification of the mPSC frequency (E) and amplitude (F) across different genotypes. n ≥ 10 animals. One-way ANOVA was performed (F_{(3, 37)} = 3.028, P = 0.0415 for E; F_{(3, 37)} = 2.065, P = 0.1216 for F). (G–I) Representative traces and quantification of evoked EPSCs recorded from the indicated strains. The numbers after Δ represent the ratio of the average values of WT and unc-31 (e928). n ≥ 9 animals. One-way ANOVA was performed (F_{(3, 43)} = 19.95, P < 0.0001 for H; F_{(3, 43)} = 39.33, P < 0.0001 for I). (J) Representative traces of mIPSCs recorded from the indicated strains are shown at different timescales. (K, L) Quantification of mIPSC frequency (K) and amplitude (L) among different genotypes. n ≥ 10 animals. One-way ANOVA was performed (F_{(3, 37)} = 1.896, P = 0.1472 for K; F_{(3, 37)} = 1.466, P = 0.2395 for L). (M–O) Representative traces and quantification of evoked IPSCs recorded from the indicated strains. The numbers after Δ represent the ratio of the average values of WT and unc-31 (e928). n ≥ 9 animals. One-way ANOVA was performed (F_{(3, 40)} = 3.709, P = 0.0191 for N; F_{(3, 40)} = 5.71, P = 0.0024 for O). (P) Schematic representation illustrating the imbalance of excitatory and inhibitory synaptic transmission in the unc-31 mutant. The data are presented as box-and-whisker plots, with the median (central line), 25th–75th percentile (bounds of the box), and 5th–95th percentile (whiskers) indicated. Student’s t test was performed for comparisons of 2 groups, and one-way ANOVA was used for comparisons of multiple groups. Tukey’s range test was subsequently performed, * P < 0.05; *** P < 0.001; ns, not significant. The error bars represent the SEM. N = 3 independent replicates. All the raw data associated with this figure are available in S1 Data. mIPSC, miniature inhibitory postsynaptic current; mPSC, miniature postsynaptic current.

https://doi.org/10.1371/journal.pbio.3003023.g003

Regulation of synaptic vesicle exocytosis by UNC-31 is independent of peptidergic signaling

The alteration in synaptic transmission observed by the deletion of CAPS/UNC-31 has been considered secondary to the dysregulation of DCV release [32,42]. However, the differential reduction in eEPSCs and eIPSCs without affecting spontaneous release of unc-31(e928) in our work suggests the direct involvement of UNC-31 in synaptic vesicle exocytosis. To test this hypothesis, we examined whether the phenotypes of unc-31 null mutants would be occluded by introducing a mutation in EGL-3. EGL-3 is a homolog of mammalian type 2 kex2/subtilisin-like proprotein convertase (KPC-2) and is required for the processing and subsequent maturation of most C. elegans neuropeptides [65]. The egl-3(tm1377)-null mutant did not show changes in spontaneous release in either excitatory or inhibitory NMJs (Fig 3D–3F and 3J–3L). Importantly, in contrast to the unc-31 mutation, the egl-3(tm1377) mutation did not decrease evoked release from either excitatory or inhibitory synapses (Fig 3G–3I and 3M–3O). These findings suggest that the synaptic transmission defects observed in unc-31 mutants are at least partially independent of EGL-3-processed peptidergic signaling. Indeed, the unc-31(e928); egl-3(tm1377) double mutant showed no additional changes in eEPSCs or eIPSCs compared with the unc-31(e928) single mutant (Fig 3G–3I and 3M–3O).

We further tested the effects of other key enzymes that process peptides, such as KPC-1, AEX-5, and BLI-4, on synaptic vesicle release. KPC-1 is homologous to furin proprotein convertase and is known to process insulin-like peptides [66] and regulate dendritic branching in C. elegans [67,68]. aex-5 encodes a proprotein convertase involved in the secretion of multiple neuropeptides [69–71], whereas bli-4 encodes kex2/subtilisin-like endoproteases in the proprotein convertase family [72]. However, none of these proprotein convertase family mutants caused spontaneous or evoked release (S3 Fig). Therefore, disruption of neuropeptide synthesis does not negatively affect synaptic transmission, suggesting that the modulation of evoked release by UNC-31 is independent of and extends beyond its involvement in DCV regulation.

unc-31 suppressed the enhanced spontaneous release of cpx-1 mutants at excitatory synapses

Both UNC-31 and CPX-1 are known to differentially regulate synaptic transmission at excitatory synapses. Understanding their interplay could reveal cooperative or antagonistic mechanisms for balancing synaptic vesicle exocytosis. We generated the cpx-1; unc-31 double mutant and found that it exhibited a striking decrease in body curvature and motility (S5 Movie), which was worse than those of the cpx-1(ok1552) single mutants (S2 Movie). Specifically, the increased body curvature caused by the cpx-1 mutation was reduced to less than that of the wild-type worms after the unc-31 mutation was introduced (Fig 4A–4C). The double mutants exhibited a posture more like that of unc-31(e928) mutants. Additionally, the thrashing frequency in M9 liquid and the free-moving speed on NGM plates further decreased and were even more severe than those in each single mutant (S4 Fig). This strong suppression of the cpx-1 behavioral phenotype by the unc-31 mutation implies a potential functional interplay between these proteins in synaptic transmission.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. UNC-31 regulates enhanced spontaneous release in the excitatory synapses of cpx-1 mutants.

(A) Representative body bending states (top) and color map of curvature (bottom) in wild-type, cpx-1(ok1552), cpx-1(ok1552); unc-31(e928), cpx-1(Δ12), and cpx-1(Δ12); unc-31(e928) worms. (B, C) Quantification (B) and distribution (C) of body curvature under the indicated strains. n ≥ 39 animals. One-way ANOVA was performed (F_{(4, 195)} = 47.65, P < 0.0001 for B). (D) Representative traces of eEPSCs recorded from the indicated genotypes. The lower traces of cpx-1(ok1552) (green) and cpx-1(ok1552); unc-31(e928) (blue) show a zoomed-in region from the upper traces. (E) Representative traces of mPSCs recorded from the mentioned genotypes, displayed at different timescales. (F, G) Quantification of eEPSC amplitude (F) and charge transfer (G) at each strain. n ≥ 12 animals. One-way ANOVA was performed (F_{(4, 64)} = 108.8, P < 0.0001 for E; F_{(4, 64)} = 80.13, P < 0.0001 for G). (H, I) Quantification of the mPSC frequency (H) and amplitude (I) in each genotype. n ≥ 10 animals. One-way ANOVA was performed (F_{(4, 62)} = 8.581, P < 0.0001 for H; F_{(4, 62)} = 18.25, P < 0.0001 for I). (J) Summary graph showing the differential regulation of excitatory and inhibitory synaptic vesicle release in the indicated genotypes. The data are presented as box-and-whisker plots, with the median (central line), 25th–75th percentile (bounds of the box), and 5th–95th percentile (whiskers) indicated. Student’s t test was performed for comparisons of 2 groups, ## P < 0.01; ### P < 0.001. One-way ANOVA was used for comparisons of multiple groups, followed by Tukey’s range test, * P < 0.05; ** P < 0.01; *** P < 0.001; ns, not significant. The error bars represent the SEM. N = 3 independent replicates. All the raw data associated with this figure are available in S1 Data.

https://doi.org/10.1371/journal.pbio.3003023.g004

We then analyzed excitatory and inhibitory synaptic transmission in the double mutants. Consistent with their motility activity, both eEPSC (Fig 4D, 4F, and 4G) and eIPSC (S5A–S5C) activities were further decreased in cpx-1(ok1552); unc-31(e928) double mutants compared with each single mutant, demonstrating an additive or synergistic effect of unc-31 and cpx-1 mutations on evoked release. Similar to the single mutants, the cpx-1(ok1552); unc-31(e928) double mutant decreased evoked release in the excitatory synapses more strongly than in the inhibitory synapses (Fig 4J). Therefore, these 2 proteins appear to regulate evoked release through distinct pathways in both types of synapses.

Unexpectedly, the enhanced spontaneous release in the excitatory synapses observed in cpx-1(ok1552) was abolished in the cpx-1(ok1552); unc-31(e928) double mutant (Fig 4E, 4H, and 4I). In contrast, mIPSCs were unaffected in the double mutant (S5D–S5F), which is consistent with the lack of effects of unc-31(e928) and cpx-1(ok1552) on mIPSCs. Hence, the unc-31 null mutation restored the E/I imbalance of the cpx-1 mutant during spontaneous release. However, regarding evoked release, the more severe impairment of excitatory synaptic transmission in the double mutant shifted the balance toward inhibition, leading to excessive muscle relaxation and reduced body curvature (Fig 4J).

Next, we examined whether the unc-31 null mutation also suppressed the enhanced spontaneous release of the CPX-1 clamping-specific mutant cpx-1(Δ12). Consistent with our observations in the cpx-1(ok1552) null mutant, the increase in mPSC in the cpx-1(Δ12) mutant was abolished in the cpx-1(Δ12); unc-31(e928) double mutant (Fig 4E, 4H and 4I). Additionally, the cpx-1(Δ12) clamping mutation further decreased both eEPSC (Fig 4D, 4F and 4G) and eIPSC (S5A–S5C Fig) in the unc-31(e928)-null background. Similarly, cpx-1(Δ12); unc-31 double mutant exhibited a decrease in body curvature and motility (Fig 4A–4C and S6 Movie) compare to those of the cpx-1(Δ12) single mutants (S3 Movie). This result confirms that UNC-31 was indispensable for the enhanced spontaneous release induced by the removal of the CPX-1 clamping function.

The suppression of cpx-1 by unc-31 relies on the presence of external Ca²⁺

The spontaneous release of synaptic vesicles (SVs) requires external Ca²⁺. We sought to examine the necessity of external Ca²⁺ on the enhanced mPSC in the cpx-1 mutant and the role of UNC-31. At 5 mM external Ca²⁺, the mPSC frequency in the cpx-1 mutant was approximately twice that in the wild-type animals (Fig 5A and 5B). The removal of external Ca²⁺ (0 mM) reduced the mPSC frequency in both cpx-1 mutant and wild-type animals. However, the mPSC frequency and amplitude of the cpx-1 mutant were still significantly greater than those of the wild-type animals (Fig 5A–5C). Compared with 5 mM external Ca²⁺, pre-incubation with 1 mM BAPTA-AM in Ca²⁺-free medium lowered the mPSC frequency approximately threefold in both wild-type and cpx-1 animals (Fig 5A and 5B). However, the increased mPSC amplitude in the cpx-1 mutant remained unchanged, indicating independence from external Ca²⁺ (Fig 5C). Thus, consistent with prior studies [31], the enhanced mPSCs observed in the cpx-1 mutant are Ca²⁺ dependent.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Ca²⁺-dependent suppression of the enhanced mPSCs in cpx-1 mutants by unc-31.

(A) Representative traces of mPSCs recorded from wild-type and cpx-1(ok1552) worms in a bath solution containing 5 mM [Ca²⁺]_o, 0 mM [Ca²⁺]_o, or 0 mM [Ca²⁺]_o with 1 mM BAPTA-AM. (B, C) Quantification of the mPSC frequency (B) and amplitude (C) in the WT and cpx-1(ok1552) strains in various bath solutions. An evident Ca²⁺ dependence of enhanced mPSCs in cpx-1 mutants was observed. n ≥ 8 animals. One-way ANOVA was performed (F_{(2, 24)} = 10.26, P = 0.0006 for WT, and F_{(2, 22)} = 13.43, P = 0.0002 for cpx-1 in B; F_{(2, 24)} = 0.9251, P = 0.4102 for WT, and F_{(2, 22)} = 1.568, P = 0.2309 for cpx-1 in C). (D) Representative traces of mPSCs recorded from WT, unc-31(e928), cpx-1(ok1552), and cpx-1(ok1552); unc-31(e928) worms in different Ca²⁺ bath solutions. (E, F) Quantification of the mPSC frequency (E) and amplitude (F) in response to the indicated strains. The suppression of the enhanced spontaneous release in cpx-1 mutants by unc-31 was apparent only in the presence of [Ca²⁺]_o. n ≥ 9 animals. One-way ANOVA was performed (F_{(3, 33)} = 13.55, P < 0.0001 for 0 mM; F_{(3, 33)} = 19.09, P < 0.0001 for 0.5 mM; F_{(3, 33)} = 13.95, P < 0.0001 for 1.5 mM; F_{(3, 33)} = 8.934, P = 0.0002 for 5 mM in E. F_{(3, 33)} = 9.683, P < 0.0001 for 0 mM; F_{(3, 33)} = 11.14, P < 0.0001 for 0.5 mM; F_{(3, 33)} = 12.94, P < 0.0001 for 1.5 mM; F_{(3, 33)} = 15.23, P < 0.0001 for 5 mM in F). (G) Pull-down experiments indicate no detectable direct binding of full-length CAPS (CAPS-FL) to GST-Complexin, regardless of the presence or absence of Ca²⁺. The dashed box highlights the absence of CAPS-FL in the pull-down. (H) Top: Binding analysis of complexin and CAPS-FL to GST-SNARE complexes, whether assembled or disassembled by boiling, in both the presence and absence of Ca²⁺. Bottom: A magnified view of the region outlined by the dashed box in the top panel, showing the specific bands corresponding to CAPS, marked by red asterisks. The data are presented as box-and-whisker plots, with the median (central line), 25th–75th percentile (bounds of the box), and 5th–95th percentile (whiskers) indicated. One-way ANOVA was performed, * P < 0.05; ** P < 0.01; *** P < 0.001; ns, not significant. The data are presented as the means ± SEMs. N = 3 independent replicates. All the raw data associated with this figure are available in S1 Data. Western blot can be found in S1 Raw Images. mPSC, miniature postsynaptic current.

https://doi.org/10.1371/journal.pbio.3003023.g005

Given that UNC-31 plays a crucial role in maintaining the enhanced spontaneous release observed in cpx-1 mutants, we hypothesized that UNC-31 functions as a secondary Ca²⁺ sensor in addition to synaptotagmin-1, as previously suggested from complexin knockdown experiments [31]. To test this hypothesis, we investigated the Ca²⁺ dependency of the regulatory effect of UNC-31 on cpx-1 mutants. At external Ca²⁺ concentrations ranging from 0.5 mM to 1.5 mM, the wild-type and unc-31 single mutants presented comparable mPSC patterns. In contrast, the increased mPSC frequency and amplitude observed in the cpx-1 mutant were abolished in the cpx-1; unc-31 double mutant at the same external Ca²⁺ concentration, which was consistent with observations at 5 mM external Ca²⁺. Unexpectedly, at zero external Ca²⁺, the cpx-1; unc-31 double mutant displayed a similar level of enhanced mPSC frequency and amplitude as the cpx-1 mutant did (Fig 5D–5F). These results imply that the suppression of the cpx-1 mutant by the unc-31 mutation depends on the presence of external Ca²⁺. Alternatively, CAPS/UNC-31 could function as a Ca²⁺-sensor responsible for spontaneous release, a role that is hindered by the presence of complexin/CPX-1.

CAPS/UNC-31 and complexin/CPX-1 do not bind directly

Next, to investigate whether the co-regulation of spontaneous release by CAPS/UNC-31 and complexin/CPX-1 is due to a direct interaction, we conducted a series of experiments. Previous studies have shown that CAPS can bind to the SNARE proteins like syntaxin-1, SNAP-25, and VAMP-2 through its Munc13 homology domain-1 (MHD1), which is essential for CAPS function in dense-core vesicle exocytosis [73,74]. Additionally, recent findings indicate that CAPS can directly interact with Munc13-1 via the DAMH domain [44]. However, no interaction between CAPS and complexin has been reported.

To test for a potential direct interaction, we generated a full-length CAPS and performed glutathione S-transferase (GST) pull-down experiments using GST-complexin (see Materials and methods). No CAPS was pulled down in the presence or absence of Ca²⁺ (Fig 5G), suggesting that no direct binding between CAPS and complexin.

We further speculated that any interaction might be dependent on other components of the exocytosis machinery, such as the SNARE complex. Additional GST pull-down experiments using GST-SNARE (either assembled or disassembled) showed that while both complexin and trace amounts of full-length CAPS could bind to the SNARE complex, there was no strong evidence of direct binding between CAPS and complexin (Fig 5H). The presence of CAPS did not enhance SNARE binding to complexin, nor did complexin enhance CAPS binding to SNARE, although weak, calcium-dependent binding between CAPS and the SNARE complex was observed (Fig 5H).

These results suggest that CAPS and complexin do not directly bind to each other. Their regulation of synaptic vesicle exocytosis is likely mediated through indirect interactions involving other components of the exocytosis machinery.

Activity-dependent synaptic plasticity was suppressed by cpx-1 and rescued by unc-31

Complexin plays a crucial role in regulating vesicle pool size, including the readily releasable pool, which is essential for short-term synaptic response changes [21]. To further investigate the conserved role of UNC-31’s dependence on CPX-1 in regulating synaptic neurotransmission, we utilized optogenetic stimulation protocols to induce activity-dependent synaptic plasticity specifically at excitatory synapses [57,75]. We employed 2 protocols: one involving 10 trains of short (0.1 s) light pulses delivered every 0.5 s (2 Hz) to mimic modest nerve stimulation, and another using 10 trains of longer (1 s) light pulses delivered every 5 s (0.2 Hz) to simulate tonic stimulation [76,77] (Fig 6A). These protocols targeted excitatory neuromuscular junctions through the expression of channelrhodopsin-2 (ChR2) in cholinergic motor neurons. We found that in wild-type animals, the modest stimulation induced a transient increase in miniature postsynaptic current (mPSC) frequency, which peaked at 10 s post-stimulation and rapidly returned to baseline (Fig 6A and 6B). By contrast, tonic stimulation resulted in a sustained increase in spontaneous release, as shown by a significantly elevated mPSC frequency lasting over 40 s (Fig 6A and 6B). These results indicate that spontaneous mPSCs demonstrates stimulus-dependent synaptic plasticity.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. The UNC-31 requirement for stimulus-dependent spontaneous release increased.

(A) Sample traces of persistent mPSCs before and after stimulation with repeated blue light illumination (blue bars) from wild-type worms. Modest stimulation: 10 × 0.1 s pulses at 2 Hz. Tonic stimulation: 10 × 1 s pulses at 0.2 Hz. Zoomed-in views of individual minis traces are presented in the gray rounded rectangle, extracted from the indicated regions of the upper panel (lines 1–5). (B) Ratio between mPSC frequency post- and pre-blue light illumination in (A). The blue bar depicts the duration of 10 consecutive illuminations. n ≥ 8 animals. (C) Sample traces of persistent mPSCs before and after tonic stimulation from wild-type (i), cpx-1(ok1552) (ii), unc-31(e928) (iii), and cpx-1(ok1552); unc-31(e928) (iv) worms. (D) Ratio between mPSC frequency post- and pre-blue light illumination in (C). n ≥ 8 animals. Two-way ANOVA was performed, * P < 0.05; ** P < 0.01; *** P < 0.001; ns, not significant. The error bars represent the SEMs. N = 3 independent replicates. All the raw data associated with this figure are available in S1 Data. mPSC, miniature postsynaptic current.

https://doi.org/10.1371/journal.pbio.3003023.g006

We then tested whether synaptic plasticity was affected in cpx-1 and unc-31 mutants. Given the sustained changes observed with tonic stimulation, we focused on analyzing mPSC frequency at different time intervals following this protocol. Unexpectedly, in the cpx-1(ok1552) mutant, tonic stimulation not only failed to increase mPSC frequency but instead caused a significant reduction, which persisted for over 40 s post-stimulation (Fig 6C and 6D). The mechanism behind this inhibition is unclear, but it could be due to a substantial depletion of synaptic vesicles, which cannot be replenished promptly in the presynaptic readily releasable pool following repeated tonic stimuli [22]. More interesting, the mPSC inhibition seen in cpx-1(ok1552) mutants was suppressed in the presence of unc-31 mutation. In the cpx-1(ok1552); unc-31(e928) double mutant, the response was neither enhanced nor inhibited, resembling the unc-31 mutant alone (Fig 6C and 6D) [57]. These results show that the activity-dependent synaptic inhibition observed in cpx-1 mutants also relies on UNC-31, further supporting the conserved role of CAPS/UNC-31 in modulating complexin/CPX-1 regulation of synaptic neurotransmission.

In summary, both complexin/CPX-1 and CAPS/UNC-31 showed a preferential regulation of excitatory synapses. The dependence of complexin/CPX-1 regulation on CAPS/UNC-31 extends beyond spontaneous release to also include stimulus-dependent synaptic plasticity.

Strains

The complete lists of constructs, primers, transgenic lines, and strains generated or acquired for this study are provided in S1–S3 Tables. PHX3665 cpx-1(syb3665) I and PHX3584 cpx-1(syb3584) I were generated by SunyBiotech via CRISPR/Cas9 sequence knock-in. The generation of transgenic strains that carry extrachromosomal arrays (gaaEx) involved the co-injection of plasmid DNA with a marker. All C. elegans strains were cultured on standard nematode growth medium (NGM) plates seeded with OP50 and maintained at 22°C. Unless otherwise noted, the wild-type animal refers to the Bristol N2 strain. Only hermaphrodite worms were used for the experiments.

Molecular biology

All expression plasmids were constructed with a three-fragment multisite gateway system (Invitrogen, Thermo Fisher Scientific, Waltham, Massachusetts, United States of America). Three entry clones, A, B, and C, were recombined into the pDEST R4-R3 Vector II destination vectors via standard attL-attR (LR) recombination reactions to generate expression clones.

The entry clone A slot1 was generated via the In-Fusion method via the ClonExpress One Step Cloning Kit (Vazyme, Nanjing). The acetylcholine neuron promoter unc-17 (3262 bp) fragment was amplified from N2 C. elegans genomic DNA to replace the rab-3 fragment in the standard BP reaction-generated entry clone A.

All entry clones B slot2 and C slot3 were generated via BP recombination reactions. The full-length cDNA (4126 bp) encoding UNC-31 was combined into the pDONR221 donor vector as slot2. The fluorescent protein sl2d-Wcherry was combined with the pDONR-P2R-P3 donor vector to generate slot3.

Aldicarb assays

Aldicarb sensitivity was assessed in synchronously growing adult worms placed on non-seeded 30 mm NGM plates containing 1 mM aldicarb. Over a 4- or 24-h period, worms were monitored for paralysis at 15- or 30-min intervals. Worms were considered paralyzed when there was no movement or pharyngeal pumping in response to 3 taps to the head and tail with a platinum wire. Once paralyzed, the worms were removed from the plate. Six sets of 15 to 20 worms were examined for each strain.

Thrashing assays

The motility of each strain was determined by counting the thrashing rate of C. elegans in liquid M9 medium. Briefly, worms were bleached to release their eggs and were then synchronously grown to young adulthood. Young adult worms were placed in a 60 μl drop of M9 buffer on a 30 mm Petri dish cover. After a 2-min recovery period, the worms were video recorded for 2 min via an OMAX A3580U camera on a dissecting microscope with OMAX ToupView software. The number of thrashes per minute was manually counted and averaged within each strain. A thrash was defined as a complete bend in the opposite direction at the midpoint of the body. In each experiment, at least 40 worms were measured for each strain.

Velocity and body curvature assays

Velocity and body curvature assays were performed with L4 stage animals. Plates for behavioral assays were prepared from 60 mm NGM plates seeded with a thin layer of OP50. Before the experiment, the OP50 lawn was spread evenly across the plate with a sterile bent glass rod. A single worm was then gently transferred to an assay plate with eyebrows. One minute after acclimatization, a 20-s movie of spontaneous locomotion was recorded on a modified stereomicroscope (Axio Zoom V16, Zeiss) with a digital camera (acA2500-60um, Basler). All images were captured with a 10× objective at 10 Hz.

Behavioral parameter analysis after imaging was performed via an in-house written MATLAB (MathWorks, Inc., Natick, Massachusetts, USA) script. The central line was used for tracking. The midpoint was used to calculate the velocity between each frame. Images for curvature analysis from each animal were divided into 32 body segments. The curvatures at each point along the worm centerline k(s) can be calculated with the coordinate of each point (x(s), y(s)) via the formula , where s is the normalized location along the centerline (head = 0, tail = 1), and the unit of k is pixel^(-1). Then, k is normalized with the length of the worm body L, resulting in the dimensionless k~(s) via the formula . To exclude errors due to uncoordinated head and tail oscillations, the curvature was calculated for 15% to 85% of the body segments. At least 40 worms of each strain were imaged and analyzed in each set of experiments.

In situ electrophysiology

Dissection and recording were carried out via protocols and solutions described in previous studies [58,109]. Briefly, 1- or 2-day-old hermaphrodite adults were glued (Histoacryl Blue, Braun) to a Sylgard (Dow Corning, USA)-coated cover glass covered with bath solution. Under a DIC microscope, semi-fixed worms were dissected dorsally with a glass pipette, the cuticle flap was flipped and gently glued (WORMGLU, GluStitch Inc.) to the opposite side, and the intact ventral body muscles were exposed after the viscera were cleaned. Anterior body wall muscle cells were patched with 4–6 MΩ resistant borosilicate pipettes (World Precision Instruments, USA), which were pulled with a micropipette puller P-1000 (Sutter). Membrane currents were recorded in the whole-cell configuration via PULSE software with an EPC-9 patch clamp amplifier (HEKA, Germany). The data were digitized at 10 kHz and filtered at 2.6 kHz. The pipette mixture contained the following (in mM): K-gluconate 115; KCl 25; CaCl₂ 0.1; MgCl₂ 5; BAPTA 1; HEPES 10; Na₂ATP 5; Na₂GTP 0.5; cAMP 0.5; cGMP 0.5, pH 7.2 with KOH, ~320 mOsm. cAMP and cGMP were included to maintain the activity and longevity of the preparation. The bath solution consisted of the following (in mM): NaCl 150; KCl 5; CaCl₂ 5; MgCl₂ 1; glucose 10; sucrose 5; and HEPES 15, pH 7.3 with NaOH, ~330 mOsm. Different extracellular Ca²⁺ bath solutions were prepared according to the bath solution with different CaCl₂ concentrations (0, 0.5, 1.5, or 5 mM). Zero Ca²⁺ with BAPTA solution consists of an additional 1 mM BAPTA in the 0 Ca²⁺ bath solution to further chelate and reduce the amount of extracellular free Ca²⁺ [31]. The muscle cells were held at −60 mV when the mPSCs or eEPSCs were recorded. To isolate mIPSCs or eIPSCs, recordings were performed with a holding potential of −10 mV, with 0.5 mM D-tubocurarine (d-TBC) included in the bath solution to block all acetylcholine receptors [58,59]. Chemicals were obtained from Sigma unless stated otherwise. The experiments were performed at room temperature (20 to 22°C).

Optogenetics

NGM plates were seeded with OP50 bacteria containing 500 μm all-trans retinal (Sigma). The seeded retinal plates were kept in darkness at 4°C. L4 worms were transferred from regular plates to retinal plates and then grown for 24 h in the dark before electrophysiological recording. Light stimulation of zxIs6 or zxIs3 was performed with an LED light source at a wavelength of 460 ± 5 nm (13.75 mW/cm²) to evoke postsynaptic currents [57,110]. The light pulse was triggered by PULSE software with a duration of 10 ms unless otherwise stated. In Fig 6, 10 light stimuli of 0.1 s are applied at a frequency of 2 Hz (modest), or 10 light stimuli of 1 s are applied at a frequency of 0.2 Hz (tonic) to evoke a series of eEPSCs.

Plasmid, protein purification, and GST pull-down experiments

Full-length CAPS-1 was expressed and purified as previously described [44]. The plasmid pFastBac-CAPS-1 for eukaryotic expression was generated by sub-cloning CAPS-1 from a pcDNA3.1-CAPS-1 plasmid into the pFastBacHtB vector (Invitrogen), which contains a polyhedron promoter (for high-level expression of recombinant protein in insect cells) before the start codon and encodes an N-terminal TEV cleavable 6× His-tag. Full-length CAPS-1 was expressed in Sf9 insect cells as described previously [111]. Sf9 cells were cultured in ESF 921 medium (Expression Systems) at 27°C at 110 rpm. Sf9 insect cells were infected with CAPS-1 baculovirus, harvested approximately 72 h post infection, and resuspended in lysis buffer (50 mM Tris (pH 8.0), 300 mM NaCl, and 10 mM imidazole). The cells were lysed and then centrifuged at 16,000 rpm for 30 min, and the supernatant was incubated with Ni²⁺-NTA agarose (QIAGEN) at 4°C for 2 to 3 h. The beads were washed with lysis buffer supplemented with an additional 10 to 30 mM imidazole. The protein was eluted with 300 mM imidazole and further purified by gel filtration in buffer containing 25 mM HEPES (pH 7.4), 150 mM KCl, and 10% glycerol (v/v).

Full-length rat complexin (residues 1–134), full-length human SNAP-25 (SN25) (residues 1–206), and the cytoplasmic domain of Syb2 (residues 29–96) were inserted into the pGEX-6p-1 vector (GE Healthcare), which contains a PreScission Protease cleavable N-terminal GST tag. All the recombinant GST fusion proteins described above were expressed in the E. coli BL21 (DE3) strain cultured in LB media at 37°C to an OD600 of 0.8 and were induced with 0.4 mM IPTG at 25°C for 16 h. The cells were lysed in Tris buffer A [20 mM Tris (pH 8.0) and 1 M NaCl] supplemented with 0.5% Triton X-100, 5 mM EDTA, and 1 mM PMSF and then disrupted via an AH-1500 Nano Homogenize Machine (ATS Engineering, Inc.) at 800 bar 3 times at 4°C. The cell lysates were subsequently centrifuged at 16,000 rpm for 30 min in a JA-25.50 rotor (Beckman Coulter) at 4°C. The supernatants were incubated with 2 ml of glutathione-Sepharose beads (Amersham-Pharmacia Biotech) at 4°C for 3 h. The bound proteins were washed with Tris buffer A. The GST tag was cleaved from the proteins by GST-fused PreScission protease (10 U/mg protein) at 4°C overnight on the beads. For purification of GST-Complexin-1 and GST-Syb2, proteins were eluted with Tris buffer B [20 mM Tris (pH 8.0) and 150 mM NaCl] containing 20 mM L-glutathione (GSH; Amresco). The proteins were further purified by size-exclusion chromatography on a Superdex 75 pg 16/600 column (GE Healthcare) and stored in Tris buffer B.

The cytoplasmic domain of rat Syx1 (residues 1–261) was inserted into the pET28a vector (Novagen). The cells were lysed in Tris buffer A containing 0.5% Triton X-100, 5 mM EDTA, and 1 mM PMSF and then centrifuged at 16,000 rpm for 30 min, after which the supernatant was incubated with Ni²⁺-NTA agarose (QIAGEN) at 4°C for 2 h. The beads were washed with the lysis buffer described above supplemented with an additional 30 mM imidazole. The protein was eluted with the lysis buffer described above supplemented with 300 mM imidazole. The protein was further purified by gel filtration in Tris buffer B. The original, uncropped, and minimally adjusted images is uploaded as S1 Raw Images.

Statistical analysis

Two-tailed Student’s t tests were used to compare 2 groups of data. One-way or two-way ANOVA was used to compare significant differences between more than 2 groups. P < 0.05 was considered statistically significant; *, **, and *** denote P < 0.05, P < 0.01, and P < 0.001, respectively. Graphing and subsequent analysis were performed via Igor Pro (WaveMetrics), Clampfit (Molecular Devices), GraphPad Prism 8 (GraphPad Software Inc., USA), MATLAB (MathWorks), ImageJ (National Institutes of Health), and Excel (Microsoft, USA). For behavior analysis and electrophysiology, each recording trace was obtained from a different animal. Unless otherwise specified, the data are presented as the mean ± SEM.