CAPS1 stabilizes the state of readily releasable synaptic vesicles to fusion competence at CA3–CA1 synapses in adult hippocampus

Calcium-dependent activator protein for secretion 1 (CAPS1) regulates exocytosis of dense-core vesicles in neuroendocrine cells and of synaptic vesicles in neurons. However, the synaptic function of CAPS1 in the mature brain is unclear because Caps1 knockout (KO) results in neonatal death. Here, using forebrain-specific Caps1 conditional KO (cKO) mice, we demonstrate, for the first time, a critical role of CAPS1 in adult synapses. The amplitude of synaptic transmission at CA3–CA1 synapses was strongly reduced, and paired-pulse facilitation was significantly increased, in acute hippocampal slices from cKO mice compared with control mice, suggesting a perturbation in presynaptic function. Morphological analysis revealed an accumulation of synaptic vesicles in the presynapse without any overall morphological change. Interestingly, however, the percentage of docked vesicles was markedly decreased in the Caps1 cKO. Taken together, our findings suggest that CAPS1 stabilizes the state of readily releasable synaptic vesicles, thereby enhancing neurotransmitter release at hippocampal synapses.

In this study, we examined the role of CAPS1 in the exocytosis of SVs using forebrain-specific Caps1 conditional KO (cKO) mice that are able to mature to adulthood 28 (Supplemental Fig. S1). Our results show that CAPS1 deficiency decreases activity-dependent SV release events in vivo at CA3-CA1 synapses in adult hippocampal slices. In addition, it causes the accumulation of SVs near the active zone but reduces the number of SVs at the plasma membrane of presynaptic terminals. Collectively, our results for the first time indicate that CAPS1 stabilizes the state of readily-releasable SVs at mature synapses in the adult hippocampus.

Discussion
Neurotransmitter release plays a key role in cognitive processes and behaviour, and is tightly regulated across the sequential stages of SV exocytosis. Although the SNARE fusion machinery and numerous associated proteins regulate exocytosis 29 , the underlying molecular mechanisms are not fully understood. In the present study, using forebrain-specific Caps1 cKO mice and acute Caps1 deletion in primary neurons, we demonstrate that CAPS1 is required for the proper exocytosis of SVs at CA3-CA1 synapses in the adult hippocampus.
A previous study showed a reduction in EPSC amplitude in primary hippocampal microisland cultures prepared from Caps1/2 double KO (DKO) mice 18 . In the present study, stimulus-induced fEPSPs were significantly reduced in mature CA3-CA1 synapses in acute hippocampal slices prepared from adult Caps1 cKO mice, suggesting that CAPS1 is involved in basal synaptic transmission in the mature hippocampus. Although a previous report using hippocampal microisland cultures revealed paired-pulse depression in autaptic synapses 18 , PPR was significantly increased in Caps1/2 DKO neurons compared with control neurons 18 . This latter observation is in agreement with our present finding of a robust enhancement of PPF at mature CA3-CA1 synapses in Caps1 cKO mice, to roughly twice that in control animals for ISIs of 50-300 ms. Thus, our findings suggest that CAPS1 regulates the release probability of SVs in mature hippocampal synapses in adult mice. Furthermore, time-lapse fluorescence imaging of recycling pool vesicles in neurons with acute Caps1 deletion showed a significant reduction in the exocytosis of SVs (Supplemental Fig. S2). Collectively, these results show that the impaired synaptic release observed in the cKO is caused by the loss of CAPS1 function, rather than the result of an indirect developmental effect. Therefore, our findings suggest that CAPS1 regulates not only the secretion of DCVs containing neuropeptides, peptide hormones and monoamines 7,8,10,11,[19][20][21]26 , but also the release of glutamate via SV exocytosis at CA3-CA1 synapses in the adult hippocampus, as in vitro experiment reported previously 18 .
Several presynaptic changes may underlie the reduction in fEPSP, including shrinkage of presynaptic boutons 30 , reduction in active zone size 31 , decrease in SV number 31 , reduction of Ca 2+ sensitivity 32 , and reduction in SV release probability 33 (resulting from a decrease in readily releasable vesicle pool size 34 and/or the number of docked vesicles 35 ). Ultrastructural analysis of CA3-CA1 synapses using 2D-TEM and 3D-SEM revealed no overall changes in presynaptic area, volume, active zone length and area in adult Caps1 cKO animals. The total number of SVs was significantly increased; however, docked SVs were markedly decreased in Caps1 cKO mice. Interestingly, proximal SVs, the SVs within 50 nm of the active zone, were significantly increased in the Caps1 cKO. This suggests that the distribution of SVs is regulated, at least in part, by CAPS1.
Our ultra-structural studies of KO synapses showed a decrease in number of "docked vesicles", defined as vesicles attached to the presynaptic membrane, but an increase in number of "undocked vesicles", defined distal and proximal to the presynaptic membrane. Within "docked vesicles" we defined we could not distinguish between the "docking" and "priming" steps of synaptic vesicles during exocytosis event. However, our electrophysiological studies of KO synapses revealed a decrease in release probability compared to WT, suggesting a decrease in fusion-competent vesicles. Taken together, we suggest that CAPS1 is crucial for at least fusion competent and/or readily releasable vesicles that are distributed on the presynaptic membrane. Therefore, a CAPS1 deficiency may cause the accumulation of SVs in the presynaptic terminals of CAPS1 cKO mice. Similar morphological findings, especially for docked and proximal SVs, were reported by Imig and colleagues using organotypic slice cultures prepared from E18 Caps1/2 DKO mice 27 , although the remarkable accumulation of SVs at presynaptic terminals in Caps1 cKO cells observed in the present study was not detected in the culture system in their study 27 . This disparity may be caused by differences between acutely prepared slices and organotypic slice cultures 36 . Indeed, although docked vesicles may be severely reduced in Caps1/2 DKO organotypic slice cultures 27 , morphological and electrophysiological properties can be influenced by the culture environment 36 . In the present study, release probability and magnitude were reduced in Caps1 cKO mature hippocampal synapses, suggesting that SVs might accumulate in presynaptic terminals because of the loss of CAPS1.
CAPS1 has a domain that is homologous with the Munc13 family of proteins, suggesting that it may function as a priming factor for SV membrane fusion 37 . Indeed, mice lacking Munc13 3 and RIM1 38 (another presynaptic protein that regulates Ca 2+ -dependent SV release) also exhibit a reduction in release probability, although without SV accumulation in presynaptic terminals 27 . Thus, CAPS1 may prime SVs for exocytosis, independently or in concert with Munc13 and RIM1. CAPS1 may regulate not only SV exocytosis, but also SV endocytosis following exocytosis. CAPS1 interacts with plasma membrane phosphatidylinositol 4,5-bisphosphate (PIP 2 ) 39 , which has been reported to be involved in vesicle endocytosis in addition to exocytosis 40 . Thus, CAPS1 may also have the ability to regulate vesicular endocytosis, which may in part account for the significant reduction of recycling SVs in Caps1 acute KO synapses (Supplemental Fig. S2F). Further study is required to clarify the functions of CAPS1 in the exocytotic and endocytotic pathways.
In conclusion, our findings suggest that CAPS1 plays a critical role in synaptic transmission by regulating the recruitment and/or access of SVs to the active zone. Thus, CAPS1 is involved in the fusion competency of SVs. CAPS1 may also be involved in the trafficking of SVs in both the exocytotic and endocytotic pathways. Our study using the forebrain-specific Caps1 cKO mouse suggests that CAPS1 is not essential for the release of glutamate, but enhances synaptic transmission at CA3-CA1 synapses in the hippocampus. Conventional CAPS1 KO mice die soon after birth, suggesting that CAPS1 may have numerous critical functions that are dependent on cell type and the specific cargo released.

Methods
Animals. All experimental protocols were evaluated and approved by the Regulation for Animal Research at Tokyo University of Science. All experiments were conducted in accordance with the Regulations for Animal Research at the Tokyo University Science. The generation of forebrain-specific Caps1 cKO C57BL/6 mice has been described previously 28 . Briefly, Caps1 flox/− females were crossed with Caps1 and Emx1-Cre heterozygote (Caps1 +/− /Emx1 Cre/wt ) males to produce control (Caps1 flox/− /Emx1 wt/wt ) and Caps1 cKO (Caps1 flox/− /Emx1 Cre/wt ) offspring.
Hippocampal acute slice preparation. Hippocampal slices were prepared as described previously 15 .
Electrophysiology. Electrophysiological recordings were performed as described previously 15 . All electrophysiological recordings were performed in ACSF at 26 °C. ACSF was exchanged at a rate of 1 ml/min. A bipolar tungsten-stimulating electrode (WPI) was placed in the CA1 stratum radiatum region. fEPSPs were recorded from the CA1 stratum radiatum following 0.05-Hz test pulses. The recording electrode was set in a glass pipette (Harvard Apparatus) filled with ACSF. Electrical signals were amplified using a MultiClamp 700A (Molecular Devices) and digitized at 10 kHz and filtered at 2 KHz using a Digidata 1440 system with pCLAMP10 software (Molecular Devices).
Electron microscopy. Electron microscopic analysis was performed as described previously 41 with partial modification. Eight-week-old Caps1 cKO and control mice were anaesthetized with CO 2 gas and perfused with PBS (0.9% NaCl in 0.1 M phosphate buffer) and then with modified Karnovsky fixative (0.8% paraformaldehyde (PFA) [TAAB], 1.5% glutaraldehyde [Nacalai] in 0.15 M phosphate buffer) (the procedure was performed by investigators blinded to genotype). The brains were removed and post-fixed in 4% PFA in 0.1 M phosphate buffer at 4 °C overnight. The fixed brains were cut into coronal sections (100 µm thick) using a vibratome. The slices are placed in cacodylate buffer containing 2% OsO 4 and 1.5% potassium ferrocyanide for 1 h at room temperature, followed by subsequent treatments with 1% thiocarbohydrazide solution for 20 min and the second osimium staining (2% aqueous OsO 4 ) for 30 min both at room temperature. The slices were then placed in 2% aqueous uranyl acetate at 4 °C overnight. The slices were subsequently treated with lead aspartate solution (0.066 g of lead nitrate in 10 ml of 0.003 M aspartic acid, pH 5.5) at 60 °C for 30 min. The slices were dehydrated with a graded ethanol series (70, 90, 100 and 100%, 10 min each at 0 °C) and mounted with Durcupan/Araldite. Then, 40-nm serial sections were prepared using an UltracutT microtome (Leica). Images of the CA1 stratum radiatum were collected, using one animal each for control and Caps1 cKO, on a JEOL 1400CX electron microscope (JEOL) with t-test. (D) Area of the active zone in control and Caps1 cKO synapses; n = 19 and 22 active zones for control and Caps1 cKO, respectively. P = 0.26, Student's t-test. (E) Representative 3D reconstructed images of SVs around the active zone in control and Caps1 cKO synapses. The active zone is colored orange. Docked, proximal and other SVs are colored magenta, green and pale blue, respectively. (F) Number of docked SVs; n = 19 and 22 active zones for control and Caps1 cKO, respectively. *P < 0.05, Student's t-test. (G) Number of proximal SVs; n = 19 and 22 active zones for control and Caps1 cKO, respectively. *P < 0.05, Student's t-test. (H) Number of distal SVs; n = 17 and 19 synapses for control and Caps1 cKO, respectively. *P < 0.05, Student's t-test.

Statistical analysis.
If not stated otherwise, data are expressed as mean ± SEM. Differences between data sets were assessed using two-tailed Student's t-test for unpaired data, analysis of covariance (ANCOVA) for continuous variables, Kolmogorov-Smirnov test for discretely distributed data, and one-way ANOVA with post hoc Tukey-Kramer test for multiple data sets. All the data were collected and analysed using a double-blind approach.