Lack of the Transient Receptor Potential Vanilloid 1 Shifts Cannabinoid-Dependent Excitatory Synaptic Plasticity in the Dentate Gyrus of the Mouse Brain Hippocampus

The transient receptor potential vanilloid 1 (TRPV1) participates in synaptic functions in the brain. In the dentate gyrus, post-synaptic TRPV1 in the granule cell (GC) dendritic spines mediates a type of long-term depression (LTD) of the excitatory medial perforant path (MPP) synapses independent of pre-synaptic cannabinoid CB1 receptors. As CB1 receptors also mediate LTD at these synapses, both CB1 and TRPV1 might be influencing the activity of each other acting from opposite synaptic sites. We tested this hypothesis in the MPP–GC synapses of mice lacking TRPV1 (TRPV1-/-). Unlike wild-type (WT) mice, low-frequency stimulation (10 min at 10 Hz) of TRPV1-/- MPP fibers elicited a form of long-term potentiation (LTP) that was dependent on (1) CB1 receptors, (2) the endocannabinoid 2-arachidonoylglycerol (2-AG), (3) rearrangement of actin filaments, and (4) nitric oxide signaling. These functional changes were associated with an increase in the maximum binding efficacy of guanosine-5′-O-(3-[35S]thiotriphosphate) ([35S]GTPγS) stimulated by the CB1 receptor agonist CP 55,940, and a significant decrease in receptor basal activation in the TRPV1-/- hippocampus. Finally, TRPV1-/- hippocampal synaptosomes showed an augmented level of the guanine nucleotide-binding (G) Gαi1, Gαi2, and Gαi3 protein alpha subunits. Altogether, the lack of TRPV1 modifies CB1 receptor signaling in the dentate gyrus and causes the shift from CB1 receptor-mediated LTD to LTP at the MPP–GC synapses.

The transient receptor potential vanilloid 1 (TRPV1) participates in synaptic functions in the brain. In the dentate gyrus, post-synaptic TRPV1 in the granule cell (GC) dendritic spines mediates a type of long-term depression (LTD) of the excitatory medial perforant path (MPP) synapses independent of pre-synaptic cannabinoid CB 1 receptors. As CB 1 receptors also mediate LTD at these synapses, both CB 1 and TRPV1 might be influencing the activity of each other acting from opposite synaptic sites. We tested this hypothesis in the MPP-GC synapses of mice lacking TRPV1 (TRPV1-/-). Unlike wild-type (WT) mice, low-frequency stimulation (10 min at 10 Hz) of TRPV1-/-MPP fibers elicited a form of long-term potentiation (LTP) that was dependent on (1) CB 1 receptors, (2) the endocannabinoid 2-arachidonoylglycerol (2-AG), (3) rearrangement of actin filaments, and (4) nitric oxide signaling. These functional changes were associated with an increase in the maximum binding efficacy of guanosine-5 ′ -O-(3-[ 35 S]thiotriphosphate) ([ 35 S]GTPγS) stimulated by the CB 1 receptor agonist CP 55,940, and a significant decrease in receptor basal activation in the TRPV1-/-hippocampus. Finally, TRPV1-/-hippocampal synaptosomes showed an augmented level of the guanine nucleotide-binding (G) Gα i1 , Gα i2 , and Gα i3 protein alpha subunits. Altogether, the lack of TRPV1 modifies CB 1 receptor signaling in the dentate gyrus and causes the shift from CB 1 receptor-mediated LTD to LTP at the MPP-GC synapses. Keywords: endovanilloid system, CB 1 receptor, excitatory synapses, long-term potentiation, G proteins INTRODUCTION Cannabinoid functions in the brain are typically associated with the activation of cannabinoid CB 1 receptors (Kano et al., 2009;Katona and Freund, 2012;Pertwee, 2015;Lu and Mackie, 2016). Additionally, endogenous, plant-derived, and synthetic cannabinoids could have other molecular targets. In particular, several cannabinoid effects depend on the activation of members of the transient receptor potential (TRP) channel family (De Petrocellis et al., 2011;Morales and Reggio, 2017;Muller et al., 2019). For example, the synthetic cannabinoid WIN 55,212-2 exerts analgesic effects by acting on the transient receptor potential vanilloid 1 (TRPV1) Patwardhan et al., 2006;Ruparel et al., 2011).

Animals
All protocols were approved by the Committee of Ethics for Animal Welfare of the University of the Basque Country (CEEA/M20/2015/105; CEIAB/M30/2015/106) and were in accordance with the European Communities Council Directive of September 22, 2010 (2010/63/EU) and Spanish regulations (Real Decreto 53/2013, BOE 08-02-2013. All efforts were made to minimize pain and suffering and to reduce the number of animals used. Seven-/eight-week-old male TRPV1-/-mice (n = 43) and their WT littermates (TRPV1+/+) (n = 21) were used. The TRPV1-/-mice were derived from heterozygous breeding pairs as described previously in the study by Egaña-Huguet et al. (2021). Mice were housed in pairs or groups of maximum three littermates in standard Plexiglas cages (17 × 14.3 × 36.3 cm), and before the experiments were conducted, they were allowed to acclimate to the environment for at least 1 week. They were maintained at standard conditions with food and tap water ad libitum throughout all experiments and in a room with a constant temperature (22 • C), and kept in a 12:12 h light/dark cycle with lights off at 9:00 p.m.

Slice Preparation for Electrophysiology
TRPV1-/-and WT mice were anesthetized by the inhalation of isoflurane. After decapitation, their brains were rapidly removed and placed on ice-cold sucrose-based solution that contained the following components (in mM): 87 NaCl, 75 sucrose, 25 glucose, 7 MgCl 2 , 2.5 KCl, 0.5 CaCl 2 , and 1.25 NaH 2 PO 4 . Coronal sections (300 µm thick) were cut with a vibratome (Leica Microsystems S.L.U.), then were recovered at 32-35 • C, and were superfused (2 mL/min) in the recording chamber with artificial cerebrospinal fluid (ACSF) containing the following components (in mM): 130 NaCl, 11 glucose, 1.2 MgCl 2 , 2.5 KCl, 2.4 CaCl 2 , 1.2 NaH 2 PO 4 , and 23 NaHCO 3 , equilibrated with 95% O 2 /5% CO 2 . All experiments were carried out at 32-35 • C. The superfusion medium contained picrotoxin (PTX) (100 µM). All drugs were added to their final concentration in the superfusion medium. For the extracellular field recordings, a glass recording pipette was filled with ACSF. The stimulation electrode was placed in the MPP (middle one-third of the ML) or LPP (outer one-third of the ML), and the recording pipette was always placed in the inner one-third of the dentate ML (mossy cell fiber layer).
A low-frequency stimulation (LFS, 10 min at 10 Hz) was applied to induce endocannabinoid-dependent excitatory LTD (eCB-eLTD) of glutamatergic inputs following the recording of a steady baseline in the presence of drugs (Puente et al., 2011;Peñasco et al., 2019). The fEPSP slope, area, and amplitude were measured (graphs depict the area). MPP stimulation was confirmed by the group II mGluRs agonist LY354740. Consistent with previous reports (Macek et al., 1996;Chiu and Castillo, 2008;Chávez et al., 2010), 1 µM of LY354740 strongly reduced MPP-fEPSPs by 59.60 ± 1.451% 10 min after the drug application (n = 4, * * p < 0.002) (data not shown). The magnitude of the eCB-eLTD after the LFS stimulation was calculated as the percentage change between baseline (averaged excitatory responses for 10 min before LFS) and last 10 min of stable responses, normally at 30 min after the end of the LFS. The slices used for each experimental condition (n) were obtained from at least three mice.

Extracellular Field Recordings
To evoke field excitatory post-synaptic potential responses (fEPSPs), repetitive control stimuli were delivered at 0.1 Hz (Stimulus Isolator ISU 165, Cibertec, Spain; controlled by a Master-8, A.M.P.I.). An Axopatch-200B (Axon Instruments/Molecular Devices, Union City, CA, United States) was used to record the data filtered at 1-2 kHz, digitized at 5 kHz on a DigiData 1440A interface (Axon Instruments/Molecular Devices, Union City, CA, United States). Data were collected on a PC using Clampex 10.0 (Axon Instruments/Molecular Devices, Union City, CA, United States) and analyzed using Clampfit 10.0 (Axon Instruments/Molecular Devices, Union City, CA, United States). At the start of each experiment, an input-output curve was constructed. A stimulation intensity was selected for baseline measurements that yielded between 40 and 60% of the maximal amplitude response.

Hippocampal Membrane Preparation
Synaptosomes were prepared as previously described by the study by Garro et al. (2001). TRPV1-/-and WT mice (n = 7 each) were anesthetized with isoflurane and decapitated; the brains were removed and placed on ice-cold 0.32 M sucrose, pH 7.4, containing 80 and 20 mM NaH 2 PO 4 (sucrose phosphate buffer) with protease inhibitors (iodoacetamide 50 µM, PMSF 1 mM). The hippocampal tissue was minced and homogenized in 10 volumes of sucrose/phosphate buffer using a motor-driven Potter Teflon glass homogenizer (motor speed 800 rpm; 10 up and down strokes; mortar cooled in an ice-water mixture throughout). The homogenate was centrifuged at 1,000 × g for 10 min, and the obtained pellet (P1) was re-suspended and pelleted. The supernatants (S1 + S1 ′ ) were pelleted at 15,000 × g (P2) and re-suspended in the homogenization buffer to a final volume of 16 mL. The suspension was layered directly onto the tubes containing 8 ml of 1.2 M sucrose/phosphate buffer and centrifuged at 180,000 × g for 20 min. The material retained at the gradient interface was carefully collected with a Pasteur pipette and diluted with the ice-cold 0.32 M sucrose/phosphate buffer to a final volume of 16 ml. The diluted suspension was then layered onto 8 ml of 0.8 M sucrose/phosphate buffer and centrifuged as described above. The obtained pellet was resuspended in the ice-cold phosphate buffer, pH 7.5, and aliquoted in microcentrifuge tubes. Aliquots were then centrifuged at 40,000 × g for 30 min, the supernatants were aspirated, and the pellets corresponding to the nerve terminal membranes were stored at −80 • C. The protein content was determined using the Bio-Rad dye reagent with bovine γ-globulin as a standard.

[ 35 S]GTPγS Binding Assays
Hippocampal extracts (25 µg protein) were thawed and incubated at 30 • C for 2 h in [ 35 S]GTPγS-incubation buffer (0.5 nM [ 35 S]GTPγS, 1 mM EGTA, 3 mM MgCl 2 , 100 mM NaCl, 0.2 mM DTT, 50 µM GDP, and 50 mM Tris-HCl, pH 7.4). The CB 1 receptor agonist CP 55,940 (10 −11 -10 −5 M, eight concentrations) was added to determine the receptor-stimulated [ 35 S]GTPγS binding. Non-specific binding was defined in the presence of 10 µM unlabeled GTPγS. Basal binding was assumed to be the specific [ 35 S]GTPγS binding in the absence of an agonist. The reactions were terminated by the rapid vacuum and filtration through Whatman GF/C glass fiber filters, and the remaining bound radioactivity was measured by the liquid scintillation spectrophotometry. For the analysis of data, individual CP 55,940 concentration-response curves were fitted by the non-linear regression method to the four-parameter Hill where E denotes the effect, log [A] the logarithm of the concentration of agonist, nH the midpoint slope, LogEC 50 the logarithm of the midpoint location parameter, and E max and Basal the upper and lower asymptotes, respectively. When required, simultaneous model fitting with parameter sharing across the datasets was performed using GraphPad Prism (GraphPad Prism, GraphPad Software Inc, San Diego, United States). The EC 50 values were transformed into pEC 50 (-LogEC 50 ) as EC 50 , and affinity constants obtained experimentally are log-normally distributed. Therefore, statistical analysis was performed accordingly (Christopoulos, 1998).

Drugs and Chemicals
All drugs for performing electrophysiological studies were dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich) and were added at the final concentration to the superfusion medium. CP 55,940, AMG9810, LY354740, AM251, URB597, JZL 184, tetrahydrolipstatin (THL), latrunculin A (LAT-A), and PTX were purchased from Tocris BioScience (Bristol, United Kingdom). S-Nitroso-N-acetylpenicillamine (SNAP) was purchased from Abcam (Cambridge, MA). All drugs were perfused at least 20 min before the LFS protocol apart from JZL 184. JZL 184 was preincubated at least 1 h before LFS protocol.

Experimental Design and Statistical Analysis
All values are given as mean ± SEM with p-values and sample size (n). Shapiro-Wilk and Kolmogorov-Smirnov tests were used to confirm the normality of the data. Statistical significance between groups was tested using parametric or non-parametric two-tailed Student's t-test as required. The significance level was set at p < 0.05 for all comparisons. All statistical tests were performed with GraphPad Prism (GraphPad Prism, GraphPad Software Inc, San Diego, United States).
Agonist's maximal efficacy, which is defined as the maximal difference between specific [35S]GTPyS binding in the presence and absence of agonist, was calculated as the percentage of basal activation. However, if the data were not normalized to their corresponding basal values, Emax differences were not statistically significant (see captions and the inset of Figure 4C).

DISCUSSION
The aim of this investigation was to study the impact of the genetic deletion of TRPV1 on the CB 1 receptor functionality and synaptic plasticity in the hippocampal dentate gyrus. Particularly, the TRPV1-/-mice have (1) an increase in the CB 1 receptor coupling efficacy and (2) a shift from CB 1 receptor-dependent LTD to LTP at the MPP-granule cell (GC) synapses.
We observed an increase in the CB 1 receptor-related Gα i1 , Gα i2 , and Gαi3 subunits in synaptosomal fractions of TRPV1-/-. Furthermore, CB 1 receptor functionality was altered in TRPV1-/-as the receptor showed a significant higher maximum coupling efficacy and a lower basal activation. These changes should be interpreted in the context of the decrease in CB 1 receptors in the synaptosomal fractions (Egaña-Huguet et al., 2021). In fact, taking into account that [ 35 S]GTPγS binding assays reflect a primary response of the system, the decrease in CB 1 receptors could be expected to produce a decrease in maximal responses. However, the achievement of similar maximal responses in TRPV1-/-synaptosomal fractions together with a decrease in basal levels reveals an increase in the coupling efficacy induced by the agonist. Similarly, the decrease in the constitutive activation of G-proteins could be linked to the decrease in CB 1 receptor expression. Actually, the CB 1 receptor-induced suppression of the fEPSP revealed at the MPP-GC synapses (Peñasco et al., 2019) was not observed in TRPV1-/-. However, although CB 1 receptors located on the glutamatergic synapses are tightly coupled to G protein signaling (Steindel et al., 2013), basal activation in the hippocampus could not only correspond to CB 1 receptors as Gi/o proteins are also coupled to other metabotropic receptors besides the CB 1 receptor (Conn and Pin, 1997). In addition, a drastic increase in the cannabinoid receptor-interacting protein 1a (CRIP1a) was detected previously in TRPV1-/-hippocampus (Egaña-Huguet et al., 2021). CRIP1a reduces the agonist-stimulated CB 1 receptor internalization and attenuates CB 1 receptor signaling, thus increasing neurotransmitter release (Booth et al., 2019;Oliver et al., 2020). Furthermore, CRIP1a overexpression in N18TG2 cells produces a robust stimulation of [ 35 S]GTPγS binding to Gα i1 and Gα i2 subunits (Blume et al., 2015). Therefore, taking into account that TRPV1-/-courses with an increased Gα i expression in the hippocampus, the potentiation observed in the [ 35 S]GTPγS binding after CP 55,940 stimulation might be related to the CRIP1a rise (Blume et al., 2015;Booth et al., 2019;Oliver et al., 2020).
The present results also show a CB 1 receptor-dependent shift to MPP-LTP in TRPV1-/- (Figure 5) after applying the LFS that elicits MPP-LTD under normal conditions (Peñasco et al., 2019;Fontaine et al., 2020). Interestingly, this type of potentiation is independent of NMDAR signaling, although eCB-eLTD requires NMDA receptor activation at other synapses (Sjöström et al., 2003;Bender et al., 2006;Lutzu and Castillo, 2021). These results highlight the importance of the crosstalk between CB 1 and TRPV1 signaling in this form of synaptic plasticity.
The MPP-LTP in TRPV1-/-was abolished in the presence of DAGL or MAGL inhibitors. This suggests that 2-AG regulation may be a limiting factor for this kind of synaptic plasticity. Consequently, as observed for the eCB-eLTD at the MPP-GC synapses (Peñasco et al., 2019), CB 1 receptor desensitization caused by high 2-AG concentration (Chanda et al., 2010;Schlosburg et al., 2010) could be responsible for the MPP-LTP blockade. Altogether, our findings suggest that 2-AG acting on CB 1 receptors mediates MPP-LTD and MPP-LTP in WT and TRPV1-/-, respectively, in which levels and timing of 2-AG availability might be playing a role in this switch from LTD to LTP (Cui et al., 2018).
The MPP-LTP in TRPV1-/-might stand on the ability of LFS (1 Hz) of the MPP-GC synapses to induce CB 1 receptorindependent LTD that relies on AEA, post-synaptic TRPV1, and AMPA receptor internalization (Chávez et al., 2010). However, MPP-LTP in TRPV1-/-was unaffected by the FAAH inhibitor URB597 (2 µM), indicating that AEA would not be involved. Actually, the conspicuous TRPV1 localization in the GC dendritic spines post-synaptic to the perforant path synaptic terminals in the outer two-thirds of the dentate ML (Puente FIGURE 5 | Summary of the main findings. The absence of TRPV1 shifts the cannabinoid CB 1 receptor-dependent long-term depression to long-term potentiation at the excitatory medial perforant path-granule cell synapses in the mouse dentate molecular layer. et al., 2015) endorses TRPV1-mediated synaptic plasticity at the MPP-GC synapses. The differences observed between eCB-eLTD at the pre-synaptic (Peñasco et al., 2019;Fontaine et al., 2020) or post-synaptic sites of the MPP-GC synapses (Chávez et al., 2010) might be regarded as a synergistic effect of CB 1 receptors reducing the glutamate release (Peñasco et al., 2019) and TRPV1 promoting AMPA receptor internalization (Chávez et al., 2010). Then, the change in CB 1 receptors in the absence of TRPV1 (Egaña-Huguet et al., 2021) could be at the base of the shift from MPP-LTD to MPP-LTP. However, HFS (100 Hz) of the LPP-GC synapses actually triggers a CB 1 receptormediated LTP that requires post-synaptic NMDA receptors and the production of mGluR5-dependent 2-AG (Wang et al., 2016). Furthermore, CB 1 receptor activation at LPP synaptic terminals causes the assembly of LAT-sensitive actin filaments resulting in an increased release of glutamate (Wang et al., 2016(Wang et al., , 2018. In our study, MPP-LTP in TRPV1-/-significantly decreased in the presence of LAT-A, suggesting a similar mechanism. Furthermore, our LFS triggered a similar LTP at the LPP-GC synapses in both WT (Wang et al., 2016(Wang et al., , 2018 and TRPV1-/-, suggesting that the CB 1 receptor-mediated LPP-LTP observed in our model is independent of TRPV1. Moreover, the MPP-LTP triggered in WT by pharmacological TRPV1 antagonism was significantly reduced by the CB 1 receptor antagonist AM251 but it was unaffected by LAT-A, indicating that the MPP-LTP upon chemical TRPV1 blockade shares the CB 1 receptor participation but not the intracellular signaling cascades turned on in the absence of TRPV1 at the MPP-GC synapses. In this sense, the NO donor SNAP blocked the MPP-LTP in TRPV1-/-. As TRPV1 is highly permeable to Ca 2+ ions (Caterina et al., 1997;Caterina and Julius, 2001), it is plausible that reduced intracellular calcium caused by the absence of TRPV1 at the MPP-GC dendritic spine synapses would damp post-synaptic NO synthase and, therefore, decrease NO production needed to support pre-synaptic LTD in the hippocampus (Reyes-Harde et al., 1999). Under normal conditions, NO activates the pre-synaptic cGMP-dependent protein kinase (PKG) known to phosphorylate and inactivate the small GTPase RhoA (Sawada et al., 2001), as well as to regulate actin cytoskeleton (for review, Francis et al., 2010). Thus, the lack of NO in the absence of TRPV1 would eventually lead to the signaling of pre-synaptic molecular pathways that ultimately would enhance the release of glutamate (Wang et al., 2018) endorsing MPP-LTP. The possibility of indirect effects on MPP-LTP by CB 1 receptors in other cells like astrocytes should also be considered, as astroglial CB 1 receptors promote the excitatory LTD by favoring the local glutamate availability (Han et al., 2012) and endorse LTP at the distant excitatory synapses (Araque et al., 2017).
Previous studies have shown that TRPV1-/-mice exhibit learning and conditioned fear deficits as well as anxiety-like behaviors, which were related to a decrease in excitatory LTP at CA1 synapses (Marsch et al., 2007). Also, the disappearance of CB 1 receptor-dependent LTD at the MPP synapses in the adult brain after exposure to intermittent ethanol intake during the adolescence and the associated recognition memory deficits were rescued by increasing 2-AG (Peñasco et al., 2020). So the shift from CB 1 receptor-dependent MPP-LTD to MPP-LTP in TRPV1-/-mice might be affecting memory as the hippocampus in general (Eichenbaum et al., 2012), and the MPP synapses, in particular, are involved in spatial memory processing (Fyhn et al., 2004;Hargreaves et al., 2005).
Altogether, the biochemical and anatomical changes taking place in the endocannabinoid system of TRPV1-/-mice (Egaña-Huguet et al., 2021), together with the increase in Gα i1 , Gα i2 , and Gα i3 proteins, the low basal CB 1 receptor activation, the high CB 1 receptor coupling efficacy, and the shift from MPP-LTD to MPP-LTP demonstrated in this study support a functional crosstalk between TRPV1 and CB 1 receptors in the dentate gyrus ( Figure 5).

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
The animal study was reviewed and approved by Committee of Ethics for Animal Welfare of the University of the Basque Country (CEEA/M20/2015/105; CEIAB/M30/2015/106).

AUTHOR CONTRIBUTIONS
JE-H: conceptualization, methodology, formal analysis, data curation, visualization, validation, investigation, and writing original draft. MS-E: methodology, formal analysis, investigation, visualization, validation, and writing original draft. SA and IB-D: methodology, visualization, and investigation. ES-G: writing original draft. GG: methodology, investigation, data curation, and writing-review and editing. SB: methodology, investigation, and data curation. JS: funding acquisition and writing-review and editing. IG: resources and investigation. NP and IE: conceptualization, methodology, formal analysis, and data curation. PG: conceptualization, supervision, project administration, funding acquisition, and writing-review and editing. All authors contributed to the article and approved the submitted version.

ACKNOWLEDGMENTS
We thank all members of PG laboratory for their helpful comments, suggestions, and discussions during the performance of this study.