All-optical presynaptic plasticity induction by photoactivated adenylyl cyclase targeted to axon terminals

Summary

Intracellular signaling plays essential roles in various cell types. In the central nervous system, signaling cascades are strictly regulated in a spatiotemporally specific manner to govern brain function; for example, presynaptic cyclic adenosine monophosphate (cAMP) can enhance the probability of neurotransmitter release. In the last decade, channelrhodopsin-2 has been engineered for subcellular targeting using localization tags, but optogenetic tools for intracellular signaling are not well developed. Therefore, we engineered a selective presynaptic fusion tag for photoactivated adenylyl cyclase (bPAC-Syn1a) and found its high localization at presynaptic terminals. Furthermore, an all-optical electrophysiological method revealed rapid and robust short-term potentiation by bPAC-Syn1a at brain stem-amygdala synapses in acute brain slices. Additionally, bPAC-Syn1a modulated mouse immobility behavior. These results indicate that bPAC-Syn1a can manipulate presynaptic cAMP signaling in vitro and in vivo. The all-optical manipulation technique developed in this study can help further elucidate the dynamic regulation of various cellular functions.

Graphical abstract

Highlights

• •bPAC-Syn1a is highly localized at presynaptic terminals
• •All-optical method combined with bPAC-Syn1a allows robust short-term potentiation
• •Short-term potentiation by bPAC-Syn1a is mediated through presynaptic mechanisms
• •In vivo manipulation of bPAC-Syn1a in presynaptic terminals modulates mouse behavior

Motivation

Manipulating intracellular signaling in a spatially restricted manner is useful for regulating neuronal functions, including synaptic plasticity. However, while light-sensitive channels such as channelrhodopsin-2 have improved subcellular targeting using localization tags, the light-sensitive signaling molecules regulating second messengers are not well developed. Here, we developed a presynaptic photoactivated adenylyl cyclase by fusing it with a presynaptic molecule, synapsin1a, and established an all-optical electrophysiological method by combining C1ChrimsonSA, a red-shifted mutant derivative of Chrimson. This enabled us to all-optically manipulate intracellular signaling and synaptic activities in a pathway-specific manner and induce presynaptic short-term plasticity.

Manipulating local cAMP signaling is useful for regulating neuronal functions, including synaptic plasticity. Nagase et al. engineer a presynaptic bPAC, bPAC-Syn1a, that is highly localized at presynapses and develop an all-optical synaptic potentiation method with bPAC-Syn1a. This all-optical manipulation technique can help further elucidate the dynamic regulation of neuronal functions.

Introduction

In the central nervous system, neuronal functions are actively and precisely regulated by intracellular signaling through second messengers, such as cyclic adenosine monophosphate (cAMP), inositol triphosphate, diacylglycerol, and Ca²⁺. For example, synaptic plasticity is induced by synaptic activity and the subsequent activation of intracellular signaling pathways, leading to adaptive behaviors, such as learning and memory.¹ This signaling activation is spatially and temporally restricted within the activated synapses so that only synapses that have experienced activity are subject to synaptic modification, leading to input specificity, which is an essential signature of Hebbian plasticity.^2,3 Interestingly, intracellular signaling can bidirectionally regulate synaptic plasticity, and the occurrence of long-term potentiation (LTP) or long-term depression (LTD) is determined by the rate and degree of Ca²⁺ rise, even though the Ca²⁺ signaling is the same.^3,4 Therefore, approaches that modulate signaling cascades with spatiotemporal precision are essential.

Optogenetic tools, such as channelrhodopsin-2 (ChR2), allow us to manipulate neuronal activities specific to projecting pathways or genetically defined populations with high temporal resolution; they are widely used to reveal the physiological significance of neuronal circuits and synaptic plasticity in various brain regions.^5,6,7 Recently, several tags have been developed to localize ChR2 to specific subcellular domains within neurons:^8,9 ChR2-Kv2.1 for a cell body,¹⁰ ChR2-MBD for dendrites,¹¹ SA-Ch for spines,¹² ChR2-MVIBD¹³ and axChR2¹⁴ for axons, and ChR2-mGluR2-PA for presynaptic terminals¹⁵ have been produced. These localization tags improve the spatial and/or temporal accuracy of optogenetic manipulation.

Recently, various light-sensitive signaling molecules that regulate second messengers have been reported; particularly, photoactivated adenylyl cyclase (PAC),^16,17 which can increase intracellular cAMP by light illumination, has been utilized to regulate the following specific neuronal functions: peripheral axonal regeneration in zebrafish,¹⁸ axonal branching and elongation of cultured neurons,¹⁹ pituitary glucocorticoid release in zebrafish,²⁰ release of synaptic vesicles and dense core vesicles in C. elegans,²¹ and LTP induction by postsynaptic or astrocytic cAMP signaling in the mouse hippocampus.^22,23 Presynaptic cAMP signaling also causes LTP in the hippocampus and amygdala^24,25,26 and modulates short-term plasticity, which regulates synaptic efficacy through a change in neurotransmitter release probability.²⁷ Intriguingly, a recent study demonstrated that PAC from Beggiatoa (bPAC) fused to the synaptic vesicle protein synaptophysin significantly enhanced synaptic localization in cultured hippocampal neurons.²⁸ Furthermore, optic stimulation of synaptophysin-bPAC induces LTP at mossy fiber-CA3 synapses but not at Schaffer collateral-CA1 synapses in acute hippocampal slices. Although the previous study developed the presynaptic bPAC, its presynaptic localization was evaluated using the synapse-to-dendrite ratio in cultured neurons, which was not a direct and fine assessment for presynaptic terminal localization, and quantitative analysis of in vivo expression was not performed. To date, studies have identified various molecules localized at presynaptic terminals, including downstream targets of cAMP signaling, suggesting that other presynaptic molecules may also serve as localization tags for bPAC, which can further enhance presynaptic terminal localization. Therefore, it is worth comparing the characteristics and efficacy of the types of localized bPACs in their localization and regulating synaptic functions.

This study aimed to develop a presynaptic bPAC by fusing it with synapsin1a (bPAC-Syn1a). We applied bPAC-Syn1a to a long projection pathway from the brain stem to the amygdala. Furthermore, we developed an all-optical electrophysiological method by combining C1ChrimsonSA, a red-shifted mutant derivative of Chrimson,²⁹ and bPAC-Syn1a for the manipulation of synaptic transmission and intracellular signaling, respectively, because conventional electrical stimulation cannot perform pathway-specific synaptic stimulation. We demonstrated the effects of bPAC-Syn1a on synaptic transmission in acute brain slices from mice. Additionally, we applied bPAC-Syn1a in vivo and examined the behavioral effects.

Results

Subcellular localization of bPAC by fusing it to trafficking sequence or synaptic vesicle-related proteins

Using bPAC-S27A (hereafter referred to as bPAC for simplicity), which shows decreased dark activity,³⁰ we developed several bPAC tools with different localization preferences. First, to localize bPAC just below the plasma membrane, similar to the mouse endogenous transmembrane adenylyl cyclase (AC), we generated the Pal-bPAC-myc-2A-tDimer by inserting the palmitoylation sequence of GAP43³¹ at the N terminus of bPAC-myc. Non-tagged control bPAC-myc or Pal-bPAC-myc was expressed in hippocampal primary cultured neurons using an adeno-associated virus (AAV), and the membrane-to-cytoplasm ratio across the dendrites was calculated (Figures 1A and S1A). We found that Pal-bPAC-myc showed a higher membrane-to-cytoplasm ratio than the control bPAC-myc, but not in tDimer. To localize bPAC to the axon terminals, we fused synapsin1a, a synaptic vesicle-related protein that is a protein kinase A (PKA) target, to the C terminus of bPAC-EYFP (bPAC-Syn1a). Based on a previous report,²⁸ we constructed synaptophysin-EYFP-bPAC (Syp-bPAC) (Figures 1B and S1B). The control bPAC was diffused throughout the nerve cells, whereas both bPAC-Syn1a and Syp-bPAC were distributed in the puncta. In addition, both puncta were found to co-localize with the active-zone protein Bassoon. However, in Syp-bPAC, there were punctum-like aggregates that did not overlap with Bassoon. To evaluate the differences in localization, we analyzed Pearson’s R value between the EYFP signal of each bPAC and the subcellular markers MAP2 (dendrites) or Bassoon (presynaptic terminals). Syp-bPAC and bPAC-Syn1a showed lower Pearson’s R values with MAP2 than that of the control bPAC. In particular, the Pearson’s R value of bPAC-Syn1a with MAP2 is also significantly lower than that of Syp-bPAC. Furthermore, bPAC-Syn1a showed a higher Pearson’s R value with Bassoon than that of control bPAC and Syp-bPAC. Altogether, we determined that bPAC-Syn1a was mostly localized to axon terminals and not to other subcellular regions.

Figure 1.

Subcellular localization of tagged bPACs

(A) Top: schematic of the control bPAC-myc and Pal-bPAC-myc expression vectors. Center: representative images of control bPAC-myc- and Pal-bPAC-myc-expressing hippocampal primary neurons as observed on immunocytochemistry. The green signal indicates anti-myc fluorescence. The magenta signal indicates the fluorescence of tDimer. White lines indicate the line at which the site was scanned. Scale bar, 10 μm. Bottom: membrane-to-cytosol ratio of myc and tDimer signals. Pal-bPAC-myc was localized to the membrane, whereas control bPAC-myc was diffused throughout the neuron. Sample sizes of the study groups are as follows: control bPAC-myc, n = 11; Pal-bPAC-myc, n = 9 neurons. ∗∗∗∗p < 0.0001 (unpaired t-test). tDimer, p = 0.7676 (unpaired t-test).

(B) Top: schematic of the control bPAC, Syp-bPAC, and bPAC-Syn1a expression vectors. Center: representative images of hippocampal primary neurons expressing control bPAC, Syp-bPAC, and bPAC-Syn1a as observed on immunocytochemistry (scale bar, 10 μm). A green signal indicates EYFP fluorescence. Magenta and gray indicate anti-Bassoon and anti-MAP2 immunostained signals, respectively. The regions within white dashed lines are expanded in the insets (scale bar, 2 μm). Bottom: bPAC-Syn1a significantly increased the co-localization of EYFP with Bassoon, as indicated by the R value compared with control bPAC (one-way ANOVA with post-hoc Bonferroni’s multiple comparison test: control bPAC vs. Syp-bPAC, p = 0.0627; control bPAC vs. bPAC-Syn1a, p < 0.0001; Syp-bPAC vs. bPAC-Syn1a, p = 0.0441), while bPAC-Syn1a and Syp-bPAC significantly decreased the co-localization with MAP2 compared with control bPAC (one-way ANOVA with post-hoc Bonferroni’s multiple comparison test: control bPAC vs. Syp-bPAC, p < 0.0001; control bPAC vs. bPAC-Syn1a, p < 0.0001; Syp-bPAC vs. bPAC-Syn1a, p = 0.0454). Sample sizes of the study groups are as follows: control bPAC, n = 23; Syp-bPAC, n = 31; bPAC-Syn1a, n = 24 neurons. Results are presented as mean ± SEM. ∗∗∗∗p < 0.0001, ∗p < 0.05 (one-way ANOVA with post-hoc Bonferroni’s multiple-comparisons test).

Comparison of expression at the injection and projection sites

Next, we applied these bPAC tools in vivo and compared their expression levels at the injection and projection sites in the lateral parabrachial nucleus (PB)-central amygdala (CeA) pathway, which is a long projection suitable for analyzing presynaptically localized optogenetic tools.¹⁵ This pathway also shows presynaptic plasticity.²⁵ The bPAC tool-expressing AAVs were injected into the PB (Figure 2A). To unify the conditions of the tag-fused bPACs, the palmitoylation sequence was fused to the initiation codon of bPAC-EYFP. In addition to the bPAC tools, we coexpressed C1ChrimsonSA-tdTomato, a red-light-activated cation channel,²⁹ which was used for optical synaptic stimulation in subsequent electrophysiological experiments. C1ChrimsonSA-tdTomato expression was comparable in all groups at both the injection (PB) and projection sites (CeA) (Figures 2B and S2). In contrast, the expression levels of Syp-bPAC and bPAC-Syn1a at the injection site were markedly lower than those of the control bPAC and Pal-bPAC (Figure 2C), suggesting a remarkably limited expression of Syp-bPAC and bPAC-Syn1a in the somata and dendrites. In particular, the expression of bPAC-Syn1a at the injection site tended to be lower than that of Syp-bPAC. At the projection site, Syp-bPAC and bPAC-Syn1a also showed significantly lower expressions compared with the others, although they were comparable with each other (Figures 2B and 2C). However, whereas the fluorescence intensity of Syp-bPAC at the projection site relative to the injection site was significantly higher than that of the control bPAC and Pal-bPAC, bPAC-Syn1a showed an even higher relative intensity at the projection site than Syp-bPAC (Figure 2D). These results suggest that, although Syp-bPAC is preferentially distributed in the distal axons and axon terminals compared with the control bPAC and Pal-bPAC, bPAC-Syn1a further enhances its selectivity.

Figure 2.

Comparison of expression of bPAC tools at the injection and projection sites

(A) Schematic of AAV injection and histological analysis. AAVs were applied to the long projecting pathway from the brain stem to the subcortical nuclei (PB, parabrachial nucleus; CeA, central amygdala).

(B) Representative images of C1ChrimsonSA-tdTomato native fluorescence (top, magenta) and anti-EYFP-immunostained signals of control bPAC, Pal-bPAC, Syp-bPAC, or bPAC-Syn1a (bottom, green) at the injection (left) and projection sites (right). Images of the injection and projection sites were captured under identical conditions. Brightness and contrast were adjusted equally across the groups. Representative images from each group were obtained for the same mouse. Scale bars, 200 μm.

(C) Summary of EYFP fluorescence intensities at the injection (left) and projection (right) sites.

(D) Summary of relative EYFP fluorescence intensities (projection site per injection site). Data are presented as mean ± SEM. Sample sizes of the study groups are as follows: control bPAC, n = 8; Pal-bPAC, n = 11; Syp-bPAC, n = 6; bPAC-Syn1a, n = 10 injection sites. ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001 (one-way ANOVA with post hoc Tukey’s multiple-comparisons test).

Preferential presynaptic expression of bPAC-Syn1a

The expression patterns of bPAC tools at the projection site were examined in more detail using immunostaining for the presynaptic marker vesicular glutamate transporter 2 (VGLUT2).^32,33 The bPAC tools were uniquely expressed in the projecting axons and axon terminals in the amygdala (Figure 3A). Control bPAC and Pal-bPAC were expressed in axon-like fibers and some bouton-like structures co-localized with VGLUT2, while Syp-bPAC expression was observed in many puncta, some of which co-localized with VGLUT2 and some did not. However, bPAC-Syn1a was mainly expressed in the bouton-like puncta and was colocalized with VGLUT2 (Figures 3A and S3A). To examine the differences in expression patterns, EYFP-labeled branch lengths were quantified. A shorter branch length indicated a higher proportion of punctum expression because the skeletonized punctum length is short (Figure S3B; STAR Methods). The branch lengths of Syp-bPAC and bPAC-Syn1a were significantly shorter than those of the control bPAC and Pal-bPAC (Figures 3B and S3C), indicating that the Syp-bPAC and bPAC-Syn1a expression is confined to punctum structures. Moreover, we calculated the proportion of VGLUT2-positive EYFP signals to verify the presynaptic localization of the bPAC tools and found that the proportion of VGLUT2-positive EYFP signals in bPAC-Syn1a was significantly higher than that in the control bPAC and Pal-bPAC, whereas that in Syp-bPAC tended to increase, but not significantly (Figure 3C). These results suggested that bPAC-Syn1a is preferentially localized in the presynaptic terminals.

Figure 3.

Expression patterns of bPAC tools at the projection site

(A) Representative images of anti-EYFP immunostaining signals in each group (top, green). The center images were merged with anti-VGLUT2 immunostained signals (magenta; scale bars, 5 μm). High-magnification images are shown at the bottom (scale bars, 2 μm). Arrowheads indicate typical double-immunostained bouton-like signals.

(B) Summary of branch lengths.

(C) Ratio of EYFP and VGLUT2 doubleimmunostained signals to the entire EYFP signal. Data are presented as mean ± SEM. Sample sizes of the study groups are as follows: control bPAC, n = 3; Pal-bPAC, n = 4; Syp-bPAC, n = 4; bPAC-Syn1a, n = 4 injection sites. ∗∗∗∗p < 0.0001, ∗p < 0.05 (one-way ANOVA with post-hoc Tukey’s multiple-comparisons test).

bPAC-Syn1a rapidly and markedly potentiated synaptic transmission

Histological analysis of bPAC-Syn1a showed higher presynaptic localization than other bPAC tools. To evaluate the functional significance of bPAC-Syn1a in regulating synaptic activity, the effects of presynaptic photoactivation of bPAC tools on synaptic transmission were electrophysiologically and all-optically analyzed in the PB-CeA pathways in acute brain slices. Excitatory postsynaptic currents (EPSCs) were evoked by red light using C1ChrimsonSA, a mutant with red-shifted absorption and reduced blue-light sensitivity²⁹ (Figure 4A). Activating bPAC tools, especially bPAC-Syn1a, rapidly and drastically potentiated EPSCs after blue light application (Figures 4B and S4A). While the EPSC amplitude during the first 2 min of blue light application increased in all groups examined, the bPAC-Syn1a group demonstrated a synaptic potentiation of greater than 300% on average—the highest among all the groups (Figure 4C). The potentiation degrees in the Syp-bPAC and bPAC-Syn1a groups were significantly higher than that in the EYFP group, where EPSCs were also increased by blue light through unknown mechanisms. Furthermore, synaptic potentiation by bPAC tools lasted for at least 10 min during blue light application and recovered to baseline levels after cessation of blue light (Figures 4B and 4C). EPSC amplitudes during the 38^th–40^th min (corresponding to the 28^th- to 30^th-min period after blue light application) were 107.3% ± 5.6% in EYFP, 94.5% ± 11.6% in control bPAC, 104.4% ± 8.5% in Pal-bPAC, 97.6% ± 4.8% in Syp-bPAC, and 115.9% ± 10.9% in bPAC-Syn1a (EYFP, p = 0.2786; control bPAC, p = 0.1049; Pal-bPAC, p = 0.9372; Syp-bPAC, p = 0.6905; bPAC-Syn1a, p = 0.2086; Mann-Whitney U test compared with baseline). Furthermore, to examine whether this short-term synaptic potentiation was mediated by PKA, the effect of the PKA inhibitor H-89²⁵ was analyzed. We found that H-89 (10 μM) significantly attenuated the extent of increase in the EPSC amplitude induced by bPAC-Syn1a activation by blue light compared with the dimethyl sulfoxide (DMSO) solvent control (p = 0.0290, Mann-Whitney U test; Figures S4B–S4D). These results suggest that bPAC-Syn1a can cause rapid and manifest short-term synaptic potentiation in a PKA-dependent manner.

Figure 4.

Synaptic potentiation induced by photoactivation of bPAC tools

(A) Schematic of AAV injection and all-optical manipulation and electrophysiological analysis. Intracellular signaling and synaptic activity were activated by blue and red light stimuli, respectively.

(B) Representative traces (average of six consecutive traces) of EPSC evoked by red LED light (every 20 s; duration, 5 ms; pink square) in each group (gray, control bPAC; green, Pal-bPAC; blue, Syp-bPAC; red, bPAC-Syn1a) at each time point (from left to right: before blue light application; during the first 2 min of blue light application, at the 8^th–10^th min of blue-light application, and at the 38^th–40^th min after the start of the 10-min blue light application). The blue light was applied every 10 s (duration, 500 ms). Scale bars, 200 pA and 10 ms.

(C) Time course of the EPSC amplitude evoked by red light in each group. EPSC amplitudes were averaged per minute and plotted. The horizontal bar indicates the application of blue-light stimuli (500 ms, 0.1 Hz).

(D) Summary of EPSC amplitude during the first 2 min (left) and at the 8^th–10^th min (right) of blue-light application. Data are presented as mean ± SEM. Sample sizes of the study groups were as follows: EYFP, n = 8; control bPAC, n = 8; Pal-bPAC, n = 6; Syp-bPAC, n = 5; bPAC-Syn1a, n = 7 neurons. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05 (using Mann-Whitney U test when compared with before blue light application). #p < 0.05 (Kruskal-Wallis test with post hoc Dunn’s multiple comparison test when compared with EYFP).

Short-term synaptic potentiation by bPAC-Syn1a was mediated through presynaptic mechanisms

To examine whether the EPSC potentiation caused by the bPAC tools (Figure 4) was due to changes in the presynaptic release probability, we recorded two successive EPSCs at the 100-ms inter-stimulus interval and compared the paired-pulse ratio (PPR) before and during blue light stimulation (Figure 5A). During stimulation, PPR was significantly decreased in all groups (Figure 5B). The extent of the decrease in PPR was not significantly different among the groups; however, a reduction in PPR was constantly observed in the Syp-bPAC and bPAC-Syn1a group. These results suggest that short-term synaptic potentiation by Syp-bPAC and bPAC-Syn1a photoactivation is caused by an increase in the presynaptic release probability. Because the degree of change in EPSC amplitude (Figure 4C) and PPR (Figure 5B) displayed cell-to-cell variations, especially in the Syp-bPAC and bPAC-Syn1a groups, we compared the effects on EPSC amplitude and PPR among different cell types. According to their firing patterns, the cells recorded in the CeA were classified into three types as follows: late spike (LS) and regular spike (RS) (Figure S5A; STAR methods). The proportions were 63.6% and 36.4% for the LS and RS cells, respectively (Figure S5B). There were no significant differences in the degree of EPSC potentiation or PPR reduction by Syp-bPAC and bPAC-Syn1a photoactivation between the cell types (Figures S5C and S5D).

Figure 5.

The mechanisms of synaptic potentiation using bPAC tools

(A) Representative scaled traces (average of six consecutive traces) of EPSC evoked by red light (two successive stimuli with a 100-ms inter-stimulus interval) before (left) and during (right) blue-light application in each group (gray, control bPAC; green, Pal-bPAC; blue, Syp-bPAC; red, bPAC-Syn1a). The traces are scaled to the first EPSC.

(B) Summary of changes in PPR in each group. Data are presented as mean ± SEM. Sample sizes of the study groups are as follows: EYFP, n = 8; control bPAC, n = 8; Pal-bPAC, n = 6; Syp-bPAC, n = 5; bPAC-Syn1a, n = 7 neurons. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05 (using Mann-Whitney U test when compared with before blue light application). Although there was a significant difference in the change in PPR between the groups (p = 0.0403, Kruskal-Wallis test), the post hoc Dunn’s multiple comparison test showed that none of the groups exhibited significant differences when compared with EYFP.

In vivo activation of bPAC-Syn1a in presynaptic terminals modulated mouse behavior

Finally, we assessed whether the photoactivation of bPAC-Syn1a at presynaptic terminals was functional in vivo. Optogenetic stimulation of PB terminals in the CeA with ChR2 generates several unconditioned aversive responses, including immobile behavior.³⁴ After AAV infection of the bilateral PB and dual light-emitting diode (LED) cannula implantation in the CeA (Figure 6A) and a subsequent habituation session, we applied photostimulation to freely behaving mice in a test chamber throughout the test session. EYFP and control bPAC were used as controls. No significant differences were observed in immobile behavior between the groups during the habituation session, in which no photostimulation was applied (Figure 6B). Mice in all groups explored the chamber at the beginning of the test session and then gradually exhibited some immobile behavior (Figure 6C). The immobility index increased over time in the bPAC-Syn1a group upon photostimulation, which was significantly differed from that in EYFP and control bPAC groups (Figure 6D). The immobility index during the latter half of the session was significantly higher in the bPAC-Syn1a group than that in the EYFP and control bPAC groups (Figure 6E). These results indicate that bPAC-Syn1a has practical applications for in vivo manipulation of mouse behavior.

Figure 6.

Behavioral effects on the optogenetic stimulation of bPAC-Syn1a in the CeA-projecting PB terminals

(A) Schematic of AAV injection and placement of an LED cannula unit (left) and a behavioral test (bottom right). The blue-light stimuli (every 5 s; duration, 1 s) were applied for 240 s. The top right shows a representative EYFP fluorescence image of bPAC-Syn1a in the CeA (projection site). BLA, basolateral amygdala.

(B) Summary of immobile behavior in a habituation session without light stimulation. n.s., not significantly different (p = 0.1410, ANOVA).

(C) Heatmaps showing the instantaneous immobile behavior of each mouse in each group (top, EYFP; center, control bPAC; bottom, bPAC-Syn1a). Sample sizes of the study groups are as follows: EYFP, n = 8; control bPAC, n = 5; bPAC-Syn1a, n = 8 mice.

(D) Time course of immobile behavior every 60 s. ∗∗p < 0.01 (two-way ANOVA with post hoc Tukey’s multiple-comparisons test: main effect of group, p = 0.0012; main effect of time, p = 0.0006; interaction, p = 0.1798; two-way ANOVA).

(E) Summary of immobile behavior for the 120^th–240^th s during the session. ∗p < 0.05 (ANOVA followed by Tukey’s multiple-comparisons test). Data are presented as mean ± SEM.

Discussion

In this study, we developed a presynaptic tag fused to bPAC, called bPAC-Syn1a. bPAC-Syn1a showed high presynaptic terminal localization compared with the previously reported presynaptic bPAC,²⁸ and it was sufficient to induce rapid and marked short-term synaptic potentiation through presynaptic mechanisms in acute brain slices and to manipulate in vivo mouse behavior by terminal photoactivation. Furthermore, by combining bPAC-Syn1a and C1ChrimsonSA, we developed all-optical manipulation of intracellular signaling and synaptic activity in a pathway-specific manner. Our findings suggest that bPAC-Syn1a would not only be useful for selectively activating the cAMP signaling cascade at presynaptic terminals in vitro and in vivo but also as an optical tool in combination with optogenetic synaptic stimulation.

Presynaptic cAMP signaling regulates synaptic plasticity at certain synapses, such as the mossy fiber-CA3 and PB-CeA synapses.^25,35 To locally manipulate presynaptic cAMP signaling by light, we fused bPAC, which has a small molecular size and long duration of activity among PACs,¹⁷ with synapsin1a, a presynaptic protein. Compared with the previously reported bPAC fused with synaptophysin, bPAC-Syn1a had higher localization to synaptic terminals in cultured and mouse brain neurons (Figures 1, 2, and 3). Both types of presynaptic bPACs successfully potentiated synaptic transmission in acute brain slices (Figure 4). While it remains unclear whether previous Syp-bPAC can be used in vivo, we indeed succeeded in modulating mouse behavior with bPAC-Syn1a (Figure 6). Recently, bPAC-F198Y, with lower dark activity than bPAC-S27A, was reported.³⁶ bPAC-F198Y-Syn1a may be useful to limit the risk of unexpected activation of the cAMP cascade.

A downstream target of cAMP signaling, PKA, forms a closed functional compartment with G protein-coupled receptors, AC, and phosphodiesterases via A-kinase anchoring proteins (AKAPs) in the cell.³⁷ As an AKAP that localizes PKA at presynaptic terminals, AKAP7 has been investigated for its role in localizing PKA at the presynaptic terminal of the mossy fiber of the hippocampal dentate gyrus,³⁸ but little is known about other synapses. The activation of presynaptic PKA phosphorylates many presynaptic proteins, including synapsin1a,³⁹ and affects synaptic terminals, which promote vesicle release, such as an increase in docked vesicles.⁴⁰ For example, phosphorylation of SNAP25 by PKA increases synaptic vesicle priming pools,⁴¹ whereas syntaphilin and synapsin1a increase synaptic vesicle release probability by interacting with syntaxin or dissociating from synaptic vesicles, respectively.^42,43 Phosphorylation of presynaptic molecules by PKA affects the dynamics of synaptic vesicles, and the large increase in EPSC amplitude and decrease in PPR observed in the present study may be because of the simultaneous phosphorylation of these molecules. We showed that the short-term synaptic potentiation by bPAC-Syn1a was significantly attenuated by a PKA inhibitor (Figure S4). cAMP signaling also acts on other downstream cascades that affect presynaptic plasticity.^44,45 Identifying the molecular mechanisms underlying bPAC-induced potentiation is an interesting topic for future studies.

The PB-CeA pathway displays presynaptic LTP, which requires cAMP-dependent PKA activation.²⁵ However, we observed pronounced but transient synaptic potentiation by bPAC-Syn1a photoactivation, and the EPSC amplitude returned to baseline levels within 30 min after photostimulation (Figure 4). This discrepancy may result from differences in the experimental configurations. The previous study used pharmacological approaches to activate the AC and PKA, which affect both presynaptic and postsynaptic neurons.²⁵ However, we could selectively express bPAC-Syn1a at the presynaptic terminals and activate cAMP signaling via photostimulation. Furthermore, they used conventional electrical stimulation for synaptic transmission, which cannot avoid stimulating other passing fibers or local circuit neurons. We used C1ChrimsonSA for optogenetic stimulation, which allowed for pathway-specific stimulation. Therefore, presynaptic LTP induction at the PB-CeA synapses might require presynaptic intracellular signaling and other synaptic activation. Interestingly, EPSC amplitudes were also increased in the EYFP group (Figure 4). Even though the red-shifted variant C1ChrimsonSA was used, non-negligible blue-light absorption was observed. Thus, blue light may directly affect C1ChrimsonSA. Although the elucidation of the mechanisms is an interesting future study, increasing the calcium permeability or channel conductance of C1ChrimsonSA, or shifting C1ChrimsonSA to a more excitable state using blue light, might contribute to the increment in EPSC amplitude. Chrimson can be used for the selective optical stimulation of two distinct pathways.^46,47 To the best of our knowledge, this is the first report of the all-optical manipulation of intracellular signaling activation and synaptic activity. Optogenetic tools that regulate second messengers, except cAMP, are also being developed.^48,49 All-optical manipulation using these tools and mutant derivatives of Chrimson, as in this study, would be advantageous for revealing the dynamic regulation of cellular functions by neuronal activity and signaling cascades in a pathway-specific manner. Notably, despite the emergence of bistable G protein-coupled receptors (GPCRs) that modulate various intracellular signaling pathways, the extremely high light sensitivity of GPCRs makes it nearly impossible to avoid cross-activation of other optogenetic tools with commonly used LED and band-pass-based stimulation devices. Our all-optical manipulation has advantages in this regard because of the suppression of unnecessary overexpression of bPAC-Syn1a. Development of new optogenetic tools for regulating other intracellular signaling and/or strategies for regulating their suitable localization are warranted. Because blue light directly affected C1ChrimsonSA and increased synaptic transmission, as discussed above, the development of more selective and red-shifted novel opsins for stimulating synaptic activities is also necessary for precise all-optical manipulation.

To confirm whether bPAC-Syn1a is practically applicable in vivo, we conducted optogenetic manipulation in freely behaving mice and found that optogenetic activation of bPAC-Syn1a in vivo enhanced immobility behavior (Figure 6). Optogenetic stimulation of the PB terminals in the caudal CeA with ChR2 increased freezing behavior during photoactivation.³⁴ The PB is considered an alert center,^50,51 and calcitonin gene-related, peptide-expressing PB neurons, which project to the CeA, are activated by aversive stimuli, such as pain, heat, and itch, and visceral stimuli.⁵² In the alert state due to exposure to a new environment, PB-CeA synaptic transmission might be potentiated via the activation of cAMP signaling at presynaptic terminals by bPAC-Syn1, resulting in increased immobility behavior. The PB-CeA pathway transmits aversive information to the forebrain, and optogenetic stimulation of the PB-CeA terminals forms artificial fear memories without any foot shocks.⁵³ It would be interesting to use our bPAC tools to study the role of presynaptic cAMP signaling in the PB-CeA pathway in the formation, discrimination, and extinction of fear memory.

Compared with presynaptic ChR2 stimulation, presynaptic cAMP elevation by bPAC-Syn1a has some possible advantages. Presynaptic depolarization with ChR2 induces action potentials and would activate various downstream signaling pathways. Thus, direct activation of specific signaling is useful for manipulating a specific signaling cascade and examining its molecular mechanisms. Furthermore, presynaptic ChR2 stimulation can result in antidromic activation and ectopic stimulation of other projecting areas. In this regard, presynaptic cAMP elevation with presynaptic bPAC, such as bPAC-Syn1a, may stimulate only specific axon terminals. Because cAMP signaling is also downstream of some neuromodulators and GPCRs, presynaptic bPAC, but not ChR2, can be used to reveal their roles.

Localized cAMP signaling has been suggested to be important for the regulation of various cellular functions,⁵⁴ but the physiological significance remains unclear. Subcellularly localized bPAC is a useful tool for addressing this issue. Notably, by using bPACs targeted to the plasma membrane, endosomes, or cytoplasm, it has been demonstrated that proper transcriptional responses depend on the subcellular location of cAMP production.⁵⁵ The selective activation of cAMP signaling at the axon terminals by bPAC-Syn1a can be applied to induce various plasticities. Functional plasticity, such as LTP and LTD, underlies not only learning and memory but also pathological states, such as pain and stress disorders,^56,57 and depends on presynaptic cAMP signaling. cAMP signaling also regulates structural plasticity, such as axonal elongation and terminal arborization.⁵⁸ Axonal growth and repair can be facilitated using bPAC,^18,19 suggesting that bPAC-Syn1a could be a useful regenerative tool for spinal cord injury. Axon terminal structural plasticity is also involved in diverse behaviors, including female sexual behavior and social authentication.^59,60 bPAC-Syn1a may also be applicable as a tool to regulate axon terminal structural dynamics and related behaviors. Intriguingly, while presynaptic cAMP elevation by bPAC promoted the release of synaptic vesicles and dense core vesicles, presynaptic depolarization by ChR2 induced only synaptic vesicle release,²¹ suggesting that bPAC-Syn1a may be useful for analyzing the regulation of dense core vesicles. It might be interesting to confirm that bPAC-Syn1 also targeted presynaptic structures in the axons from neuromodulators. In particular, noradrenergic fibers tend to have different synaptic organization with no postsynaptic compartment and varicosity-like structures.

To conclude, our study suggests that bPAC-Syn1a is a tool that is suitable for selective manipulation of presynaptic cAMP signaling in vitro and in vivo. The all-optical manipulation technique developed in our study could help further elucidate the dynamic regulation of cellular functions by neuronal activity and signaling cascades in a pathway-specific manner.

Limitations of the study

cAMP signaling activates not only PKA but also other downstream cascades for regulating various neuronal functions.^44,45 Therefore, although bPAC-Syn1a was suitable for PKA-dependent short-term plasticity in this study, there is a possibility that the subcellularly targeted bPACs with other localization tags might more efficiently manipulate neuronal functions depending on the downstream cascades. Furthermore, it should be noted that a weak but significant increase in EPSC amplitude was also observed in the EYFP group. Thus, C1ChrimsonSA alone can slightly potentiate synaptic transmission by blue light. This feature may complicate the use of our all-optical method for fine manipulation of synaptic plasticity in some regions. Further enhancement of the wavelength selectivity of C1CrimsonSA or the development of a new opsin with high specificity for red light is required.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit monoclonal anti-myc	Cell Signaling	Cat# 2278; RRID: AB_490778
Rabbit monoclonal anti-MAP2	GeneTex	Cat# GTX133109; RRID: AB_2886830
Mouse monoclonal anti-Bassoon	ENZO Life Sciences	Cat# ADI-VAM-PS003; RRID: AB_10618753
Goat anti-Mouse IgG Alexa Fluor 568	Thermo Fisher Scientific	Cat# A-11004; RRID: AB_2534072
Goat anti-Rabbit IgG Alexa Fluor 633	Thermo Fisher Scientific	Cat# A-21070; RRID: AB_2535731
Rabbit polyclonal anti-GFP	MBL International	Cat# 598; RRID: AB_591819
Guinea pig polyclonal anti-VGLUT2	Frontier Institute	Cat# VGlut2-GP; RRID: AB_2571621
Goat polyclonal anti-rabbit IgG Alexa Fluor 488	Thermo Fisher Scientific	Cat# A11034; RRID: AB_2576217
Goat anti-guinea pig IgG Biotinylated	Vector Laboratories	Cat# BA-7000; RRID: AB_2336132
Streptavidin, Alexa Fluor 594 conjugate	Thermo Fisher Scientific	Cat# S11227
Bacterial and virus strains
AAV DJ-Syn-bPAC-2A-tDimer	Addgene	Addgene plasmid #85398
AAV DJ-Syn-bPAC-EYFP	This paper	N/A
AAV DJ-Syn-Pal-bPAC-2A-tDimer	This paper	N/A
AAV DJ-Syn-Pal-bPAC-EYFP	This paper	N/A
AAV DJ-Syn-synaptophysin-EYFP-bPAC	This paper	N/A
AAV DJ-Syn-bPAC-EYFP-synapsin1a	This paper	N/A
AAV DJ-Syn-EYFP	This paper	N/A
AAV DJ-Syn-C1ChrimsonSA-tdTomato	This paper	N/A
Chemicals, peptides, and recombinant proteins
Picrotoxin	Sigma-Aldrich	Cat# P1675
H-89	Sigma-Aldrich	Cat# B1427
Experimental models: Cell lines
AAV-293	Funakoshi	Cat#: AAV-100; RRID: CVCL_6871
Experimental models: Organisms/strains
Rat: Wistar	CLEA JAPAN	N/A
Mouse: C57BL/6JmsSlc	Japan SLC	N/A
Recombinant DNA
pAAV hSyn1 bPAC S27A cMyc 2A tDimer	Addgene	Addgene plasmid #85398
pAAV-Syn-ChR2-EYFP	Hamada et al.15	N/A
pAAV-Syn-EYFP	Hamada et al.15	N/A
pAAV-Syn-bPAC-EYFP	This paper	N/A
pAAV-Syn-Pal-bPAC-2A-tDimer	This paper	N/A
pAAV-Syn-Pal-bPAC-EYFP	This paper	N/A
pAAV-Syn-synaptophysin-EYFP-bPAC	This paper	N/A
pAAV-Syn-bPAC-EYFP-synapsin1a	This paper	N/A
pAAV-Syn-CrimsonR-tdTomato	Addgene	Addgene plasmid #59171
pAAV-Syn-C1ChrimsonSA-tdTomato	This paper	N/A
Software and algorithms
Fiji	https://imagej.net/software/fiji/	RRID: SCR_002285
GraphPad Prism 6 and 9	GraphPad Software	RRID: SCR_002798
IGOR Pro 7	WaveMetrics	RRID: SCR_000325
LabChart	ADInstruments	RRID: SCR_023643
SutterPatch	Sutter Instrument	N/A
Time FZ software	O’Hara & Co., Ltd	N/A

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Ayako M. Watabe (awatabe@jikei.ac.jp).

Materials availability

The AAVs and plasmids generated in this study are available from the corresponding authors upon request.

Data and code availability

• •All data reported in this paper will be shared by the lead contact upon request.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Experimental model and study participant details

Ethics

All recombinant DNA and animal experiments in this study were performed following regulations and guidelines for the care and use of experimental animals and approved by the Institutional Committee for the Care and Use of Experimental Animals at the University of Yamanashi (protocol number: A30-21), and by the Institutional Animal Care and Use Committee of the Jikei University (approval number: 2017-048 and 2021-045). All experiments were compiled with the Guidelines for the Proper Conduct of Animal Experiments of the Science Council of Japan (2006).

Animals

Male C57BL/6J mice (Japan SLC, Inc., Shizuoka, Japan) were group housed (3–5 mice per cage) on a 12 h light/12 h dark cycle and provided with food (CE2, CLEA Japan) and water ad libitum. A previous study using male and female mice did not report sex differences of the effects of Syp-bPAC,²⁸ suggesting that those with bPAC-Syn1a might be limited.

Primary cultures

The preparation of rat hippocampal primary neurons was previously described.¹⁵ The hippocampi were dissected from Wistar rat embryonic at day 18 brain and dissociated by papain at 37°C. The neurons were then plated on poly-D-lysine (PDL) coated 24-well plates at 3,000 cells/well. Cultures were maintained in a Neurobasal medium containing B-27 supplement and GlutaMAX (Gibco) and kept in an incubator at 37°C with 5% CO₂.

Method details

AAV vector plasmid constructs and preparation

The bPAC expression AAV vector, named bPAC-2A-tDimer in the manuscript, was derived from pAAV hSyn1 bPAC S27A cMyc 2A tDimer, a gift from Thomas Oertner (Addgene plasmid #85398; http://n2t.net/addgene:85398; RRID: Addgene_85398). Pal-bPAC-myc-2A-tDimer was generated by fusing the mouse GAP43 palmitoylation sequence (1–20 amino acid)³¹ at the initiation codon of the bPAC-myc-2A-tDimer. YFP expression AAV vector was constructed by deleting the ChR2 sequence from pAAV-Syn-ChR2-EYFP.¹⁵ The bPAC-EYFP was constructed by inserting a synthesized bPAC S27A with modifying codon sequence reducing restriction enzyme recognition sites into the EYFP upstream site of pAAV-Syn-EYFP. Pal-bPAC-EYFP and synaptophysin-EYFP-bPAC, and bPAC-EYFP-synapsin1a were constructed by inserting the above-mentioned palmitoylation sequence, mouse synaptophysin sequence (1–924 nt derived from GenBank: X95818), or mouse synapsin1a sequence (130–2250 nt derived from NCBI Reference Sequence: NM_013680) into the start codon of bPAC-EYFP or EYFP-bPAC, or the stop codon of bPAC-EYFP plasmid, respectively. C1ChrimsonSA-tdTomato was created by modifying pAAV-Syn-ChrimsonR-tdTomato (Addgene plasmid #59171; http://n2t.net/addgene:59171; RRID: Addgene_59171) based on the previous report.²⁹ In detail, 1–79 amino acid, 169th Ser, and 176th Arg of ChrimsonR were replaced into 1–76 amino acid of CrChR1, Ala, and Lys, respectively.

The preparation of AAV vectors was previously described.¹⁵ First, the AAV vector solution was prepared by transfecting pAAVs as above, pAAV-DJ, and pHelper into AAV293 cells using PEI-Max (Polysciences, Inc.). After 4 days, cells were centrifuged at 1,600 × g for 5 min and resuspended with 150 μL of phosphate-buffered saline (PBS). Then, cells were lysed by freezing and thawing four times using liquid N₂, followed by incubation in Benzonase at 45°C for 15 min. Next, the crude lysates were centrifuged at 21,500 × g for 10 min at 4°C and the vector-containing supernatants were collected. This step was repeated three times, and a total of 450 μL of AAV solution was collected. Finally, the AAV titers were confirmed by RT-PCR and stored at −80°C.

Evaluation of localization in hippocampal primary neuron culture

Cells were infected with AAVs (2 × 10⁷ vg/well) at 3 days in vitro (DIV) and left for at least 10 days before preparing for immunostaining. For the immunostaining, the cells were washed with PBS and fixed with 4% paraformaldehyde in PBS for 15 min. After washing with PBS, cells were permeabilized with 0.25% Triton X-100 in PBS for 15 min, blocked with 1% BSA, and 0.25% Triton X-100 in PBS for 30 min. Cells were incubated with the following primary antibodies, anti-myc (Cell Signaling), or anti-MAP2 (GeneTex) and anti-Bassoon (ENZO Life Sciences) diluted in blocking buffer for 2 h at room temperature, followed by Alexa-Fluor conjugated secondary antibodies. Images of stained cells were captured with confocal laser microscopy (FV1200, Olympus) equipped with a 60× oil immersion objective lens (UPLSAPO60X, NA 1.35) with a confocal aperture of 300 and 3× digital zoom. The line scan across the dendrites was performed, with the membrane at both ends and the cytoplasm at the center, to determine the membrane per cytoplasm ratio of myc and tDimer. Pearson’s R value between EYFP and subcellular markers was calculated using the coloc2 plug-in in Fiji. Images were background-subtracted using a 15-micron rolling ball algorithm, and threshold regression selected Bisection.

Surgery for AAV microinjection

Surgery for AAV injection was performed as previously described.^15,61 Male C57BL/6J mice (4–5 weeks old; Japan SLC) were anesthetized with a mixture of medetomidine hydrochloride (0.75 mg/kg; Zenoaq), midazolam (4.0 mg/kg; Astellas), and butorphanol tartrate (5.0 mg/kg; Meiji Seika Pharma) and fixed in a stereotaxic instrument. AAV vectors were bilaterally injected into the lateral PB (1.25 mm to lateral, 6.4 mm to posterior from bregma, and 3.2 mm deep with a 20° anterior to posterior angle) using a Hamilton microsyringe (1701RN Neuros Syringe, 33 G, 10 μL, Hamilton) at an injection speed of 50 nL/min controlled with a microinjection syringe pump (UMP3, World Precision Instruments). AAV DJ-Syn-bPAC-EYFP (histological and electrophysiological analysis, 2.62 × 10⁷ to 4.25 × 10⁸ vg/μL; behavioral analysis, 4.25 × 10⁸ vg/μL), AAV DJ-Syn-Pal-bPAC-2A-tDimer (electrophysiological analysis, 3.90 × 10⁸ to 2.07 × 10⁹ vg/μL), AAV DJ-Syn-Pal-bPAC-EYFP (histological analysis, 4.08 × 10⁷ to 3.01 × 10⁸ vg/μL; electrophysiological analysis, 3.01 × 10⁸ vg/μL), AAV DJ-Syn-synaptophysin-EYFP-bPAC (histological and electrophysiological analysis, 3.47 × 10⁸ vg/μL), AAV DJ-Syn-bPAC-EYFP-synapsin1a (histological analysis, 5.79 × 10⁷ vg/μL; electrophysiological and behavioral analysis, 5.79 × 10⁷ to 8.98 × 10⁸ vg/μL) and AAV DJ-Syn-EYFP (electrophysiological analysis, 3.62 × 10⁷ to 2.64 × 10⁸ vg/μL; behavioral analysis, 1.77 × 10⁸ vg/μL) was used to express bPAC tools or EYFP (250 nL). The mixture of these AAV and AAV DJ-Syn-C1ChrimsonSA-tdTomato (250 nL; 6.44 × 10⁷ to 1.54 × 10⁹ vg/μL) was injected at a volume of 500 nL for co-expression. In some electrophysiological experiments, AAV DJ-Syn-Pal-bPAC-2A-tDimer was used instead of AAV DJ-Syn-Pal-bPAC-EYFP. We did not find any detectable differences between the effects of these AAVs; thus, these data were pooled. The injection syringes were left in place for 10 min before the withdrawal.

Histology of in vivo expression and localization

First, AAV-injected mice were deeply anesthetized with a mixture of medetomidine hydrochloride (0.75 mg/kg), midazolam (4.0 mg/kg), and butorphanol tartrate (5.0 mg/kg) at least 5 weeks after microinjection and transcardially perfused with phosphate-buffered saline (PBS) and 4% paraformaldehyde (PFA). The brains were then removed, post-fixed in 4% PFA at least one overnight at 4°C and cut into coronal section (50-μm thickness) with a vibrating blade slicer (VT1200S, Leica). The brain sections were incubated overnight at 37°C with a rabbit antibody against GFP (1:500; No. 598, MBL) in 0.05 M Tris-buffered saline (TBS) containing 10% normal goat serum, 2% bovine serum albumin and 0.5% Triton X-100 (NGS). A guinea pig (GP) antibody against VGLUT2 (1:200; MSFR106290, Nittobo) was also applied in the double immunostaining. After washing in PBS, the sections were incubated with the goat anti-GP IgG antibody (H + L), Biotinylated (1:200; BA-7000, Vector Laboratories) in NGS at room temperature for 2 h. Subsequently washing in PBS, sections were incubated with a goat anti-rabbit Alexa Fluor 488 (1:200; A11034, Thermo Fisher Scientific) and Streptavidin, Alexa Fluor 594 (1:500; S11227, Thermo Fisher Scientific) in NGS at room temperature for 2 h. Finally, the sections were washed in PBS and mounted on slides with ProLong Glass Antifade Mountant (P36980, Thermo Fisher).

Images of EYFP immunofluorescence and native tdTomato fluorescence at the injection and projection sites were acquired with a confocal microscopy (FV3000, Olympus) with a 10× objective lens (UPLXAPO10X, NA 0.4; Olympus) at excitation wavelengths of 488 and 561 nm with the same setting for each image. Three z stack images were obtained at a 3.90-μm interval and analyzed as a maximum intensity projection image. The average intensity at the three points of region of interest (ROI; 100 × 100 μm for the injection site; 50 × 50 μm for the projection site) was quantified. In the double immunostaining for EYFP and VGLUT2, nine z stack images were obtained with a 60× oil immersion objective lens (UPLSAPO60X, NA 1.42; Olympus) at a 0.41-μm interval. Maximum intensity projection images were produced every three images. Immunofluorescent images of EYFP and VGLUT2 were binarized by auto threshold Li and Triangle methods, respectively, in Fiji after the application of Gaussian blur (r = 0.5 pixels) and particles >0.20 μm² were quantified. Brach length was calculated using the Skeletonize and the Analyze Skeleton plugins.

Electrophysiology in acute brain slices

First, for at least 5 weeks after AAV injection, the mice were deeply anesthetized with 5% isoflurane, and the brains containing the amygdala were quickly removed. Next, coronal brain slices (300-μm thick) were cut in an ice-cold cutting solution consisting of (in mM) 92 N-Methyl-D-glucamine, 2.5 KCl, 0.5 CaCl₂, 10 MgSO₄, 1.25 NaH₂PO₄, 3 sodium pyruvate, 12 N-acetyl-Lcysteine, 25 D-glucose, 5 L-ascorbic acid, 20 HEPES, and 30 NaHCO₃, bubbled with 95% O₂ + 5% CO₂ (pH of approximately 7.4; osmolality of approximately 290 mOsm/kg) with a vibrating microtome (VT1200S, Leica) and incubated in the cutting solution at approximately 34°C for 10–15 min. The slices were then kept at room temperature (20°C–25°C) in standard artificial cerebrospinal fluid (ACSF) consisting of (in mM) 125 NaCl, 3 KCl, 2 CaCl₂, 1.3 MgCl₂, 1.25 NaH₂PO₄, 10 D-glucose, 0.4 L-ascorbic acid, and 25 NaHCO₃, bubbled with 95% O₂ + 5% CO₂ (pH of approximately 7.4; osmolality of approximately 300 mOsm/kg) until the recordings.

Whole-cell patch-clamp recordings were performed on the CeA neurons visually identified under an upright microscope with oblique illumination (BX-51WI, Olympus). Each slice was continuously perfused at a rate of 1.5–2.5 mL/min with standard ACSF at approximately 30°C during the recording chamber. Patch-clamp electrodes (4–8MΩ) were made of borosilicate glass pipettes (1B150F-4, World Precision Instruments). The composition of the internal solution was (in mM): 122.5 potassium gluconate, 10 HEPES, 17.5 KCl, 0.2 EGTA, 8 NaCl, 2 MgATP, and 0.3 NaGTP (pH 7.2; osmolarity, 290–300 mOsm). Neurons were held at −60 mV in voltage-clamp recordings. Picrotoxin (100 μM) was added to the ACSF to isolate EPSCs. The PKA inhibitor (H-89, Sigma-Aldrich) was dissolved in DMSO at 20 mM and subsequently dissolved in ACSF to obtain the final concentration (10 μM). After pre-incubation in ACSF containing H-89 or only 0.05% DMSO solvent as a control for 30–40 min, EPSCs were recorded in the presence of H-89 or DMSO, respectively. The membrane current and potential were recorded with an amplifier (MultiClamp 700B, Molecular Devices; DOUBLE IPA, Sutter Instrument), filtered at 2 kHz, and digitized at 10 kHz with a 16-bit resolution using a PowerLab interface (AD Instruments) together with timing pulses for light stimulation. Photostimulation was applied by a multi-LED source using blue (455 or 470 nm; M455L3, M470L3, M470L4, Thorlabs) and red (625 nm; M625L3, M625L4, Thorlabs) LED delivered to the entire field through a ×40 water-immersion objective lens (LUMPLFLN40XW, NA 0.8; Olympus). Light intensities were measured using a digital optical power meter (9742-10/3664; Hioki, Nagano, Japan) and as follows: blue light, 3.2–4.5 mW/mm²; red light, 15.2–25.7 mW/mm². The timing and duration of photostimulation were controlled by Master-8 (A.M.P.I.). The bPAC was activated by 500-ms blue light at 0.1 Hz. The C1ChrimsonSA was activated by 5-ms red light at 0.05 Hz. The red and blue light were applied with a slightly offset: while the timing of blue light stimulation every 10 s and red light stimulation every 20 s overlapped, blue light was applied 100 ms after red light. Paired-pulse ratio (PPR) was defined as the ratio of the second EPSC amplitude to the first EPSC amplitude in response to two stimuli with a 100-ms inter-stimulus interval. Only records with amplitudes of greater than −20 pA were analyzed. The series resistance was continuously monitored by hyperpolarizing steps (−10 mV), and recordings were discarded if it changed by >20% within an experiment. To record intrinsic firing patterns, depolarizing current pulses (1 s; 25 pA step) were applied in current-clamp recordings. Late spiking (LS) neurons and Regular spiking (RS) neurons were discriminated from the first spike onset in response to the step current injection. The spike onset was >300 ms, classified as the LS neurons, and it was shorter than 300 ms, which we did as the RS neurons. The recorded membrane currents and potentials were analyzed offline with Igor Pro 7 (WaveMetrics).

Behavioral test

First, AAV-injected mice were recovered for 7 days from the cannulation surgery in which a bilateral LED cannula unit consisting of dual optical fibers (diameter, 0.25 mm; length, 4.5 mm; spacing, 6.0 mm) attached to a LED body (blue, 470 nm) was stereotactically inserted to the CeA (1.3 mm posterior to bregma). At least 2 days before the behavioral session, mice were conducted handling and attaching the optogenetic receivers for 1–3 min per day, for consecutive 3 days. The day before the behavioral test, the mice were habituated for 12 min with the same apparatus as in the test session, but without photostimulation. On the behavioral test, an infrared light-driven wireless LED unit Teleopto receiver (2 g; TeleR-2-P, Bio Research Center) was attached to the LED cannula unit of the mouse in the home cage. After 15 min resting time, mice were placed into the context (square shapes, 170 mm width × 100 mm depth × 100 mm height, 200 Lux, 60 dB background white noise) on the inside wall of the sound-attenuating box (CL-M3, O’Hara & Co., Ltd., Japan). After being placed in the context, mice received optogenetic LED illumination (light duration, 1 s; frequency, 0.2 Hz; LED power, 4.5 mW) controlled by an infrared light-driven remote controller (Teleopto remote controller, Bio Research Center) for 4 min. Light pulses were precisely controlled by a programmable stimulator (Master-8, A.M.P.I.) and Time FZ software (O’Hara & Co., Ltd). Behavioral analyses were conducted in a manner blinded to the mouse groups. Mouse behavior was captured using a digital camera at 2 frame/s, and immobility behavior was analyzed using Time FZ software (O’Hara & Co., Ltd), a package based on NIH Image. The movement of the mouse was detected by pixel-to-pixel subtraction between two subsequent frames, and the behavior at each frame was defined as “immobile” when the total number of pixels with a detectable frame-to-frame difference was <30 pixels. The identification of immobility was pre-optimized by two independent human observers using C57BL/6J mice. The immobility index was calculated as follows: immobility index = the number of immobile frames/the number of total frames.

Quantification and statistical analysis

Data in the text and figures are expressed as mean ± standard error of the mean (SEM) and sample size (n). We used appropriate statistical tests with post-hoc analyses when applicable, such as the unpaired t-test, Mann–Whitney U test, one- or two-way ANOVAs with Bonferroni’s or Tukey’s multiple comparison tests, and Kruskal–Wallis test with Dunn’s multiple comparison test. Statistical analyses were performed using GraphPad Prism 6 and 9 (GraphPad Software, La Jolla, CA, USA). p < 0.05 was considered statistically significant.

Published: March 22, 2024