The atypical adhesion GPCR ADGRA1 controls hippocampal inhibitory circuit function

SUMMARY

Neural circuits contain a diverse array of inhibitory interneurons that control information processing. The cell surface receptors and signaling pathways that modulate cell-type-specific inhibitory function are unclear. Here, we identify the orphan G protein-coupled receptor (GPCR) ADGRA1 as essential for hippocampal PV inhibitory synaptic function. ADGRA1 is selectively enriched in hippocampal PV interneurons and localizes to a subset of synapses. ADGRA1 deletion in PV interneurons impairs intrinsic excitability and reduces inhibitory synaptic strength onto dentate gyrus granule cells. ADGRA1 engages several downstream G proteins, notably Gα13, a pathway important for the establishment of hippocampal PV interneuron synaptic networks. These results identify an orphan receptor pathway selective for fast-spiking PV neuron function and expand our understanding of the signaling mechanisms that establish hippocampal inhibitory circuits.

In brief

Diverse arrays of interneurons shape circuit function and information processing in the brain. The cell surface receptors and signaling pathways that mediate the establishment and function of these interneuronal circuits are incompletely understood. Tosun et al. identify a synaptic orphan GPCR critical for PV interneuron function and learning and memory.

Graphical Abstract

INTRODUCTION

Mammalian neural circuitry contains an array of physiologically and morphologically diverse inhibitory interneurons that are critical modulators of information processing.¹ While representing only ~10%–15% of the total hippocampal neuron population, interneurons shape circuit function via networks of synapse-specific connectivity.² Their cell-type- and synapse-specific functions are essential for learning and memory. Moreover, experience and neuronal activity shape the plasticity of interneuronal synapses in a cell-type-specific manner.³ Interneuron dysfunctions are highly associated with neurological disorders, including autism spectrum disorder (ASD), Rett syndrome, and epileptic encephalopathies.⁴ Despite the importance of diverse inhibitory synaptic connections, the cell surface receptors and signaling pathways that control the establishment and functional properties of these cell types remain unclear.

Increasing evidence supports important roles of adhesion G protein-coupled receptors (aGPCRs) in neurodevelopment and physiology.^3–7 aGPCRs are a large GPCR class that display both extracellular adhesion and intracellular signaling functions.^8,9 A signature feature of the aGPCR class is a membrane-proximal GAIN (GPCR autoproteolysis-inducing) domain.¹⁰ The GAIN domain has been a focus of aGPCR signal activation mechanisms.^11,12 The GAIN domain of many aGPCRs contains an autoproteolysis site that can generate a membrane-proximal self-tethered agonist (TA), also known as a Stachel peptide.¹³ Signaling by the C-terminal fragment (CTF) GPCR can occur via removal of the N-terminal fragment (NTF) and exposure of the TA or through a more tunable mechanism involving the NTF.¹⁴

All aGPCRs exhibit a GAIN domain, with one exception: ADGRA1 (GPR123).⁸ ADGRA1 lacks both N-terminal extracellular adhesion domains and the GAIN domain, suggesting unique functions and signaling mechanisms compared to other aGPCRs. ADGRA1 exhibits a 7-transmembrane (7-TM) GPCR followed by a relatively large cytoplasmic tail (268 amino acids in mice) with a C-terminal PDZ-binding motif but lacks extensive extracellular regions characteristic of other aGPCRs. Despite these features, ADGRA1 is evolutionarily conserved and exhibits 37% sequence identity with ADGRA2 (GPR124) and 44% with ADGRA3 (GPR125) in the 7-TM region and C-terminal tail. Unlike ADGRA1, ADGRA2 and ADGRA3 contain extensive extracellular adhesion modules composed of leucine-rich repeats and immunoglobulin domains.⁹ Thus, ADGRA1 likely lost these N-terminal features over evolution and may exhibit unique functional and signaling properties.

Despite these interesting properties, the mechanistic functions of ADGRA1 remain unclear. Elevated Adgra1 expression has been identified in human bladder and breast cancer studies^15,16 and has been shown to regulate maintenance and acquisition of pluripotency in human induced pluripotent stem cells (hiPSCs).¹⁷ Adgra1 is highly expressed in the mammalian CNS,¹⁸ and proteomic analysis found it enriched in postsynaptic fractions, supporting synaptic localization.¹⁹ Functional studies using constitutive Adgra1-knockout (KO) mice found reduced body weight and deficiencies in metabolism and thermogenesis, most likely due to dysfunctions in the hypothalamus.²⁰ Further analysis of constitutive KO mice determined altered anxiety-like behaviors in mutant male mice.²¹ However, the cell-type-specific and putative synaptic role of ADGRA1 in mammalian circuits remains unknown.

Our studies identify ADGRA1 as highly enriched in hippocampal PV interneurons. Selective ADGRA1 deletion from these cell types impairs fast-spiking PV excitability and reduces inhibitory synaptic function. ADGRA1 activates several G proteins, notably Gα13, an important signaling pathway for inhibitory synaptic function, and partially co-localizes and co-expresses with Gα13 in neurons. Collectively, these studies identify a synaptic orphan receptor that controls inhibitory synaptic function in a cell-type-specific manner.

RESULTS

ADGRA1 is enriched in hippocampal PV-positive interneurons and localizes to synapses

To examine aGPCR expression in hippocampal inhibitory neurons, we crossed Cre-inducible RiboTag (HA-Rpl22) mice to the vGAT-Cre driver line (Figures 1A–1C). This approach enables immunoprecipitation of ribosome-bound mRNAs in a cell-type-specific manner.²² We performed HA immunoprecipitations from postnatal day 30 (P30) hippocampal tissue and analyzed the enrichment of aGPCRs and control transcripts relative to input mRNA samples using RT-qPCR (Figures 1A–1C). Interestingly, several aGPCRs were enriched in hippocampal interneurons, including Adgra1/Gpr123, Adgrc3/Celsr3, Adgrf2/Gpr111, and Adgrg6/Gpr126 (Figure 1B). Control transcripts confirmed the enrichment of inhibitory markers in immunoprecipitated samples (PV, SST, and Slc32a1/vGAT) and de-enrichment of glial (Gfap) and excitatory neuronal (Slc17a7/vGlut1) markers (Figure 1C). Of these aGPCRs, Adgra1 has not been studied in a cell-type-specific manner in the brain. Next, we conducted RNA in situ hybridizations to examine the spatial expression pattern of Adgra1 relative to other cell-type molecular markers (Figures 1D–1I and S1–S3). Adgra1 was highly expressed in sparse sub-populations of cells throughout the P30 hippocampus (Figures 1D and 1E). Interestingly, of the 33 murine aGPCRs, Adgra1 is the only member that lacks N-terminal adhesion domains and the GAIN domain (Figure 1F). We analyzed the developmental time course of Adgra1 expression throughout the hippocampus and found that it was expressed at negligible levels shortly after birth and increased during postnatal development (Figures S1A and S1B). We next conducted double RNA in situ/immunohistochemistry experiments to determine if the sparse population of Adgra1-expressing cells is neurons or glia (Figures S1C–S1F). Adgra1 expression overlapped with cells positive for the neuronal marker NeuN, supporting high expression in a sub-population of hippocampal neurons (Figures S1C–S1F). These results support that Adgra1 is an orphan aGPCR enriched in hippocampal interneurons.

Figure 1. The atypical orphan aGPCR Adgra1 is enriched in hippocampal interneurons.

(A) Representative sections of HA tag immunohistochemistry from vGAT-Cre/RiboTag/Ai14 mice.

(B) RiboTag enrichment measurements from the postnatal day 30 (P30) hippocampus of vGAT-Cre/RiboTag mice. The relative enrichment of each target in the HA RiboTag immunoprecipitation was compared to the input sample from the same tissue.

(C) RiboTag enrichment of interneuronal (Pvalb, SST, and Slc32a1/vGAT), glial (Gfap), or excitatory neuronal (Slc17a7/vGlut1) transcripts.

(D and E) Adgra1 is expressed in sparse sub-populations of cells in the postnatal hippocampus.

(D) Representative RNA in situ of the P30 hippocampus labeled for Adgra1 together with DAPI.

(E) Representative high-magnification images of RNA in situ for Adgra1 in the dentate gyrus (DG), CA3, and CA1.

(F) Domain organization of Adgra1 compared to other aGPCRs.

(G–I) Adgra1 is enriched in hippocampal PV+ interneurons.

(G) RNA in situ hybridizations for Adgra1 together with PV.

(H) High-magnification images of Adgra1 and PV RNA in situ in indicated hippocampal sub-regions.

(I) Quantification of Adgra1 RNA intensity within indicated hippocampal cell types.

Numerical data are the means ± SEM from 3–6 independent biological replicates (mice). See Figures S1–S3 for additional analysis of Adgra1 expression in the brain. Statistical significance was determined via two-way ANOVA with Tukey’s multiple comparisons (*p < 0.05 and ***p < 0.001).

To determine the subtype of interneuron, we performed RNA in situ experiments in the P30 hippocampus for Adgra1 together with markers including PV, SST, CCK, and Calb2 (Figures 1G–1I, S2, and S3). Adgra1 expression was highly co-localized with PV+ cells (Figures 1G–1I). Adgra1 was moderately expressed in SST+ cells and excluded from CCK- and Calb2-expressing cells, suggesting that Adgra1 is predominantly selective for PV subtypes (Figures 1G–1I, S2, and S3). Adgra1 was also expressed in distinct patterns in other brain regions, including cortical layer 5–6 and the thalamus, and therefore may also have important functions in these areas (Figure S3). These results show that Adgra1 is an orphan aGPCR selectively enriched in hippocampal PV interneurons.

We next assessed the subcellular localization of ADGRA1 in hippocampal neurons. Given the absence of reliable antibodies for ADGRA1, we expressed HA-tagged ADGRA1 in primary hippocampal cultures (Figures 2 and S4). We conducted either sparse transfection to examine cell-autonomous localization or lentiviral transduction for population-level expression (Figures 2A–2H). Neurons sparsely receiving HA-ADGRA1 overexpression displayed surface HA signals along both MAP2-labeled dendrites and AnkG-labeled axon initial segments, suggesting subcellular localization to both pre- and postsynaptic sites (Figure 2A). We next used lentiviral transduction to compare the cell surface localization of wild-type (WT) ADGRA1 with that of a PDZ-binding motif truncation (ADGRA1-ΔPDZ) to determine if this sequence is important for localization (Figures 2B–2H). Surface ADGRA1 formed puncta that partially co-localized with both pre- and postsynaptic markers (Figures 2C–2H). ADGRA1-ΔPDZ localized to synapses comparable to the WT, suggesting that other sequence features are responsible for synaptic localization. Next, to assess localization in interneurons, we co-labeled hippocampal cultures for surface HA and GAD67 (Figures S4A and S4B). The C-terminal PDZ-binding motif was dispensable for HA-ADGRA1 localization in both GAD67+ and GAD67− neurons (Figures S4A and S4B). We subsequently examined HA-ADGRA1 localization ex vivo by delivering Cre-inducible HA-ADGRA1 AAVs into the dentate gyrus (DG) of PV-Cre or SST-Cre mice (Figures 2I–2K and S4C). HA-ADGRA1 localized with inhibitory vGAT slightly higher in PV+ neurons than in SST+ neurons (Figures 2I–2K and S4C). Collectively, these results show that ADGRA1 is a synaptic GPCR and its expression is selectively enriched in hippocampal PV interneurons.

Figure 2. ADGRA1 is a synaptic orphan GPCR.

(A) Cell-autonomous localization of HA-ADGRA1 using sparse transfection. Hippocampal cultures were co-labeled for HA, the somatodendritic marker MAP2, and the axon initial segment (AIS) marker AnkG.

(B) Representative neuron expressing HA-ADGRA1 via lentiviral transduction.

(C–H) Immunocytochemistry for lentivirally transduced HA-tagged ADGRA1 in primary hippocampal neurons.

(C) Example dendrites immunolabeled for surface HA-ADGRA1, followed by Bassoon and MAP2.

(D) Co-localization of lentiviral HA-ADGRA1 or a PDZ truncation (HA-ADGRA1-ΔPDZ) and presynaptic Bassoon.

(E and F) Similar to (C) and (D) except for excitatory postsynaptic SHANK2.

(G and H) Similar to (C) and (D) except for inhibitory postsynaptic gephyrin.

(I–K) Expression of HA-ADGRA1 in DG PV+ and SST+ neurons.

(I) HA immunolabeling following AAV hSyn DIO HA-ADGRA1 injections into the DG of PV-Cre/Ai14 mice.

(J) Similar to (I) except for SST-Cre/Ai14 mice.

(K) Quantification of HA/vGAT co-localization from experiments in (I) and (J).

Numerical data are the means ± SEM. Statistical significance was assessed via a two-tailed t test and one-way ANOVA with post hoc Tukey tests (*p < 0.05). See Figure S4 for additional localization studies.

ADGRA1 is dispensable for synaptic function in excitatory DG GCs

We subsequently sought to elucidate the cell-type-specific function of ADGRA1. We generated Adgra1 floxed mice (Adgra1 conditional KO [cKO]) and validated this allele by generating primary hippocampal cultures and transducing them with lentiviral Cre or inactive ΔCre as a control (Figure S5). This approach enables highly efficient Cre delivery to obtain a robust KO across the cell population. Adgra1 cKO cultures infected with Cre lacked floxed exon 6 relative to ΔCre conditions, supporting effective deletion with this allele (Figure S5). Transcripts for several other aGPCRs were unaltered (Figure S5; Table S1). Next, we crossed Adgra1 cKO models to several Cre drivers to obtain cell-type-specific deletion. First, since we observed relatively low signals of Adgra1 expression throughout the hippocampal granule cell (GC) and pyramidal cell layers (Figure 1), we crossed Adgra1 cKO mice to the EMX1-Cre driver to delete ADGRA1 from forebrain excitatory neurons and glia (Figures 3, S6, and S7).²³ In parallel, we crossed Adgra1 cKO models to the PV-Cre or SST-Cre driver lines to delete ADGRA1 specifically in PV or SST interneurons (Figures 4, 5, 6, 7, and S8–S12). We focused on the GC circuitry of the DG, given the importance of PV and SST inhibitory synaptic function in this area, as well as its well-characterized patterns of synaptic connectivity.^24,25

Figure 3. Neuronal morphology, synapse density, and synaptic transmission are preserved in DG GCs lacking ADGRA1.

(A) Experimental diagram.

(B–D) Morphological analyses of DG GCs with biocytin/streptavidin loading.

(B) Representative spines from Ctl or EMX-cKO GCs.

(C) Quantification of dendritic spine densities in GCs.

(D) Quantification of average total dendrite length from DG GCs.

(E–G) Analysis of mEPSCs from Ctl or EMX-cKO GCs.

(E) Representative mEPSC traces.

(F) Cumulative probability plot of inter-event intervals and summary graph (inset) of the mean mEPSC frequency.

(G) Cumulative probability plot and summary graph (inset) of mEPSC amplitude measurements.

(H–J) Analysis of mIPSCs from Ctl or EMX-cKO GCs.

(H) Representative mIPSC traces.

(I) Cumulative probability plot of inter-event intervals and summary graph (inset) of the mean mIPSC frequency.

(J) Cumulative probability plot and summary graph (inset) of mIPSC amplitude measurements.

(K–P) Immunohistochemical analysis of synapses in the DG.

(K) Immunolabeling for presynaptic Bassoon together with postsynaptic excitatory Homer1 in the hippocampal DG molecular layer (ML), granule cell layer (GCL), or CA3 stratum lucidum.

(L) Quantification of Bassoon puncta density in indicated sub-regions.

(M) Quantification of Homer1 puncta density in indicated sub-regions.

(N–P) Similar to (K)–(M) except for vGAT/SPO.

Numerical data are the means ± SEM or cumulative histograms. See Figure S5 for characterization of the Adgra1 cKO mouse allele, Figure S6 for additional morphological parameters, and Figure S7 for quantification of immunohistochemistry. Statistical significance was determined via a two-tailed t test, one-way ANOVA with post hoc Tukey tests, or Kolmogorov-Smirnov test (cumulative histograms).

Figure 4. ADGRA1 is essential for maintaining the intrinsic excitability of PV+ neurons.

(A) Experimental diagram. PV+ cells in the DG were identified using the Ai14 tdTomato reporter.

(B) Representative traces from current-clamp recordings at 600 pA current steps.

(C) Quantification of action potential (AP) frequency as a function of current injection steps in PV+ DG neurons.

(D) Average input resistance from current-clamp recordings.

(E) Average capacitance measurements from current-clamp recordings.

(F) Average rheobase measurements from current-clamp recordings.

(G–I) Spontaneous postsynaptic currents (sPSCs) monitored from PV+ interneurons.

(G) Representative sPSC traces.

(H) Cumulative probability plot of inter-event intervals and summary graph (inset) of the mean sPSC frequency.

(I) Cumulative probability plot and summary graph (inset) of sPSC amplitude measurements.

Numerical data are the means ± SEM or cumulative histograms. Statistical significance was determined via a two-tailed t test or two-way ANOVA (*p < 0.05, **p < 0.01, and ***p < 0.001).

Figure 5. ADGRA1 deletion in PV+ neurons impairs inhibitory synaptic output onto DG GCs.

(A) Diagram of experimental approach.

(B–D) Analysis of spontaneous mIPSCs from indicated conditions.

(B) Representative mIPSC traces from DG GCs.

(C) Cumulative probability plot of inter-event intervals and summary graph (inset) of the mean mIPSC frequency.

(D) Cumulative probability plot and summary graph (inset) of mIPSC amplitude measurements.

(E–H) Evaluation of eIPSCs in Ctl or PV-cKO GCs.

(E) Representative eIPSC traces.

(F) eIPSC amplitude in response to increasing extracellular stimulation.

(G) Representative eIPSC paired-pulse ratio (PPR) traces.

(H) Quantification of PPR from DG GCs in Ctl or PV-cKO mice.

(I) Representative images of Syt2-labeled presynaptic PV terminals in the indicated hippocampal sub-regions.

(J) Quantification of Syt2 puncta density in the indicated hippocampal sub-regions.

(K) Analysis of Syt2-positive puncta density on PV-Ai14-labeled soma.

(L and M) Assessment of open field behavior in Ctl or PV-cKO mice.

(L) Average distance traveled over a 60 min open field trial.

(M) Distance traveled over time during the 60 min open field trial.

(N and O) Cued learning from Pavlovian fear conditioning studies.

(N) Quantification of the percentage of time spent freezing before or after presentation of the cue stimulus following fear conditioning.

(O) Average percentage of time spent freezing during the habituation period or presentation of the cue.

Numerical data are the means ± SEM or cumulative histograms. See Figure S8 for characterization of Cre driver lines, Figure S9 for additional electrophysiological parameters, and Figure S10 for additional behavioral characterization of PV-cKO mice. Statistical significance was determined via a two-tailed t test, one-way ANOVA with post hoc Tukey tests, two-way ANOVA, or Kolmogorov-Smirnov test (*p < 0.05, **p < 0.01, and ***p < 0.001).

Figure 6. Adgra1 deletion in SST interneurons modestly disrupts inhibitory input onto DG GCs.

(A) Diagram of experimental approach.

(B–D) Analysis of spontaneous mIPSCs from Ctl or SST-cKO GCs.

(B) Representative mIPSC traces from GCs.

(C) Cumulative probability plot of inter-event intervals and summary graph (inset) of the mean mIPSC frequency.

(D) Cumulative probability plot and summary graph (inset) of mIPSC amplitude measurements.

(E–H) eIPSC measurements in Ctl or SST-cKO GCs.

(E) Representative eIPSC traces.

(F) eIPSC amplitude in response to increasing extracellular stimulation.

(G) Representative eIPSC paired-pulse ratio (PPR) traces.

(H) Quantification of PPR from GCs in Ctl or SST-cKO mice.

(I and J) Assessment of open field behavior in Ctl or SST-cKO mice.

(I) Average distance traveled over a 60 min open field trial.

(J) Distance traveled over time during the 60 min open field trial. (K and L) Cued learning from Pavlovian fear conditioning.

(K) Quantification of the percentage of time spent freezing before or after presentation of the cue stimulus following fear conditioning.

(L) Average percentage of time spent freezing during the habituation period or presentation of the cue.

Numerical data are the means ± SEM or cumulative histograms. See Figure S11 for additional electrophysiological parameters and Figure S12 for additional behavior studies. Statistical significance was determined via a two-tailed t test or Kolmogorov-Smirnov test (*p < 0.05 and ***p < 0.001).

Figure 7. ADGRA1 engages several G proteins, including Gα13.

(A) TRUPATH BRET2 analysis of full-length ADGRA1. A LPHN3 construct with either the Gαi1 or Gα13 TRUPATH sensors was used as a negative or positive control, respectively.

(B) Similar to (A) except using an ADGRA1 mutant lacking the short N-terminal extracellular sequence prior to the 7-TM GPCR.

(C–E) ADGRA1 plasmid dose-response curves using the indicated TRUPATH BRET2 biosensors.

(C) ADGRA1 plasmid dose-response experiments for Gα11.

(D) Similar to (C) except for Gα15.

(E) Similar to (C) except for Gα13.

(F) Representative immunocytochemistry for surface HA-ADGRA1 and intracellular Gα13 and MAP2 in the indicated conditions.

(G) Representative MAP2-labeled dendrite co-labeled for surface HA and intracellular Gα13.

(H) Quantification of HA/Gα13 co-localization from experiments in (F) and (G).

(I–L) Analysis of Adgra1/Gα13 expression in PV+ and SST+ neurons of the entorhinal cortical-hippocampal circuit.

(I) Representative RNA in situ for PV/Adgra1/Gα13 in P30 sections.

(J) Representative high-magnification images in the indicated areas.

(K) Quantification of the percentage of total PV cells that also express Adgra1 and the percentage of PV cells that express both Adgra1 and Gα13.

(L) Similar to (K) except for SST+ neurons.

Numerical data are the means ± SEM. See Figure S13 for additional characterization of ADGRA1 constructs and quantification of ADGRA1/Gα13 co-localization and expression. Statistical significance was determined via one-way ANOVA with post hoc Tukey test (**p < 0.01 and ***p < 0.001).

We performed whole-cell patch-clamp electrophysiology and obtained recordings from GCs in EMX-cKO or littermate control (Ctl) acute brain slices. We first measured miniature excitatory and inhibitory postsynaptic currents (mEPSCs and mIPSCs). We filled cells during recordings with biocytin to enable subsequent 3D reconstruction and morphological analysis of their dendritic arborizations and dendritic spines (Figures 3A–3J, S6A, and S6B). GCs lacking ADGRA1 exhibited no detectable alterations in dendritic spine density and dendrite length and complexity (Figures 3B–3D, S6A, and S6B). Moreover, EMX-cKO GCs displayed no significant amplitude and frequency changes in spontaneous synaptic transmission (Figures 3E–3J). We subsequently analyzed synaptic density using immunohistochemistry for pre- and postsynaptic markers (Figures 3K–3P and S7). We detected no significant differences in the density or intensity of excitatory or inhibitory synaptic markers or synaptoporin (SPO), a marker for large mossy fiber terminals (LMTs) formed by GCs (Figures 3K–P and S7). Thus, ADGRA1 has a minor synaptic role within excitatory GCs of the DG.

ADGRA1 is essential for hippocampal PV cell intrinsic excitability and inhibitory synaptic function

We next analyzed PV-Cre/Adgra1 cKO (PV-cKO) models using imaging, electrophysiology, and behavior (Figures 4, 5, 6, and S8–S10). While these Cre drivers are well characterized and routinely used, we first confirmed their specificity relative to endogenous PV (Figure S8). We then conducted current-clamp recordings to measure the intrinsic excitability of PV+ neurons in the DG (Figure 4). We observed firing patterns characteristic of fast-spiking interneurons, including non-accommodating firing, action potential trains, and maximal firing frequencies up to 120 Hz, in both control and experimental groups. However, Adgra1-deficient PV+ neurons displayed significant alterations in passive and active membrane properties. Adgra1 deletion substantially reduced the firing frequency of DG fast-spiking interneurons (Figures 4B and 4C). The reduced firing frequency in PV-cKO PV+ neurons was accompanied by a significant decrease in input resistance and an increase in membrane capacitance. We also observed a slight increase in rheobase and action potential threshold. Together, these findings suggest reduced intrinsic excitability in PV interneurons and indicate a role for ADGRA1 in PV cell excitability. We also observed slight depolarization of the resting membrane potential in PV-cKO PV+ cells. Despite this more depolarized baseline, ADGRA1-deficient PV cells exhibited reduced excitability, suggesting impaired spike initiation (Table 1). We measured spontaneous postsynaptic currents (sPSCs) onto PV+ cells while holding at − 70 mV and found no significant changes in sPSC frequency or amplitude (Figures 4G–4I). Collectively, these alterations shift fast-spiking interneurons into a hypoexcitable state, resulting in reduced firing output. Our observations support a cell-autonomous role for ADGRA1 in maintaining the intrinsic properties required for fast-spiking interneuron function in the DG.

Table 1.

Electrophysiological properties of PV interneurons of the dentate gyrus

	Pvcre(+)/Ai14homo	Pvcre(+)/Adgra1 homo/Ai14 homo	p value
Access resistance (MΩ)	20.63 ± 2.029	19.15 ± 2.029	0.4748
AP amplitude (mV)	63.95 ± 2.662	63.53 ± 2.662	0.8766
AP decay time (ms)	0.9364 ± 0.07523	0.8833 ± 0.07523	0.4886
AP half-width (ms)	0.7364 ± 0.04325	0.7 ± 0.04325	0.41
AP rise time (ms)	0.3909 ± 0.02855	0.3917 ± 0.02855	0.9791
AP threshold (mV)	− 41.12 ± 1.533	− 37.69 ± 1.533	0.036
Fast AHP amplitude (mV)	16.47 ± 1.523	19.05 ± 1.523	0.1044
Input resistance (MΩ)	121.4 ± 11.34	87.32 ± 11.34	0.0065
Medium AHP amplitude (mV)	14.58 ± 1.484	14.92 ± 1.484	0.8175
Membrane capacitance (pF)	86.76 ± 9.064	110.7 ± 9.064	0.015
Membrane resistance (MΩ)	115.1 ± 11.61	67.9 ± 11.61	0.0005
Resting membrane potential (mV)	− 69.64 ± 1.770	− 65.65 ± 1.770	0.0344
Rheobase (pA)	222.7 ± 33.77	287.5 ± 33.77	0.0688
Voltage sag (mV)	1.027 ± 0.3585	0.5317 ± 0.3585	0.1812

Results are shown as the mean ± SEM. Parameters are from current-clamp recordings in Figure 4. AHP, afterhyperpolarization.

We then measured PV inhibitory synaptic output onto DG GCs (Figures 5A and S9). Interestingly, GCs from PV-cKO acute slices displayed a significantly decreased mIPSC frequency but increased amplitude (Figures 5B–5D and S9A). We used rescue approaches to determine the specificity of these effects on pre-synaptic PV+ neurons by injecting Cre-dependent ADGRA1-encoding AAVs into the DG of PV-cKO mice. This approach rescued alterations in mIPSCs onto DG GCs, supporting a role for ADGRA1 in presynaptic PV+ cells that target DG GCs (Figures 5B–5D and S9A). To discern the effect on inhibitory synaptic strength compared to presynaptic release probability, we next analyzed evoked IPSCs (eIPSCs) and eIPSC paired-pulse ratios (PPRs) (Figures 5E–5H and S9). PV-cKO GCs exhibited significantly lower eIPSC amplitudes, particularly at increasing stimulation strengths (Figures 5E and 5F). However, PV-cKO GCs displayed no changes in the PPR or coefficient of variation in eIPSCs, supporting that presynaptic release probability is preserved (Figures 5G, 5H, S9B, and S9C). These results show that ADGRA1 plays a critical role in controlling the presynaptic inhibitory strength of PV interneurons onto GCs.

Importantly, despite this decrease in inhibitory strength, the overall density of PV terminals labeled with synaptotagmin-2 (Syt2) was unaltered throughout the hippocampus, suggesting a functional but not developmental role of ADGRA1 in presynaptic PV interneurons (Figures 5I–5K). We then assessed the consequences of ADGRA1 deletion in PV interneurons on learning and memory (Figures 5L–5O and S10). Open field behavior was unaltered in PV-cKO models (Figures 5L, 5M, S10A, and S10B). However, ADGRA1 deletion in PV interneurons produced an impairment of Pavlovian fear conditioning (Figures 5N, 5O, S10C, and S10D). This deficit was selective for cued learning, while contextual learning was unaltered (Figures S10C and S10D). Given that alterations in hippocampal excitatory/inhibitory balance can generate seizures, we analyzed seizure susceptibility using PTZ (pentylenetetrazole)-mediated seizure-induction paradigms but found no significant changes (Figures S10E–S10I), supporting a role of ADGRA1 function in circuits involved with learning and memory.

Given that ADGRA1 is also expressed in SST interneurons (Figures 1I and S2), we assessed ADGRA1 synaptic function at SST-GC synapses by analyzing SST-Cre/Adgra1 cKO mice (Figures 6, S11, and S12). Spontaneous mIPSC frequency was reduced and eIPSC amplitudes were moderately reduced in GCs, while release probability measured via the PPR was unaltered (Figures 6B–6H and S11A–S11C). Open field behavior, fear conditioning, and seizure susceptibility were intact in SST-Cre/Adgra1 cKO mice (Figures 6I–6L and S12). Collectively, these results support that ADGRA1 controls cell-type-specific inhibitory synaptic strength predominantly in PV+ neurons onto DG GCs.

ADGRA1 activates several G proteins, including Gα13

Of the aGPCRs, ADGRA1 is the only member that lacks an extensive extracellular region harboring adhesion domains and the GAIN domain (Figure 1F). To obtain initial insights into ADGRA1 signaling, we assessed which G proteins ADGRA1 is capable of activating using the complete panel of TRUPATH BRET2 biosensors (Figure 7A).²⁶ Given ligands for ADGRA1 are unknown, we examined the ability of overexpressed ADGRA1 to activate the complete panel of 14 TRUPATH Gαβγ sensors relative to empty-vector-transfected controls. Full-length ADGRA1 activated several G proteins, most notably Gα13 (Figure 7A). While the short extracellular sequence of ADGRA1 has no homology to the aGPCR TA, we determined if this extracellular sequence is required for G protein activation by replacing it with a short glycine linker (ΔN-ADGRA1) (Figure 7B). We first determined the relative expression level of overexpressed full-length ADGRA1 compared to ΔN-ADGRA1 using immunocytochemistry and immunoblotting (Figures S13A–S13D). The ΔN-ADGRA1 mutant was expressed at comparable levels to full-length ADGRA1 (Figures S13A–S13D). The G protein coupling profile of ΔN-ADGRA1 was similar to full-length ADGRA1, suggesting that this extracellular sequence is not involved in basal G protein activation (Figure 7B). We subsequently validated ADGRA1 G protein coupling by performing full-length ADGRA1 plasmid dose-response experiments to measure the relationship between ADGRA1 plasmid copy number and BRET2 response (Figures 7C–7E). Gα11, Gα15, and Gα13 all exhibited a plasmid copy-number-dependent change in BRET2, supporting the specificity of these measurements (Figures 7C–7E). Our previous studies found that Gα13 partially co-localizes with synaptic markers and has important roles in PV synaptic function.²⁷ We next examined the co-localization of ADGRA1 with endogenous Gα13 (Figures 7F–7H). HA-ADGRA1 partially co-localized with Gα13 in primary hippocampal neurons (Figures 7F–7H). Similar to co-localization studies with synaptic markers (Figures 2 and S4), deletion of the PDZ-binding motif preserved this co-localization. We then assessed expression of Adgra1 and Gα13 in PV and SST neurons of the entorhinal cortical-hippocampal circuit (Figures 7I–7L and S13E). Over 95% of PV and SST neurons also expressed Adgra1 and Gα13 (Figures 7I–7L and S13E). Collectively, these experiments show that ADGRA1 activates several G proteins, including Gα13, and co-localizes and co-expresses with Gα13 in hippocampal neurons.

DISCUSSION

The cell surface receptors and signaling pathways that establish inhibitory circuitry remain poorly understood. Our studies identify ADGRA1 as a synaptic receptor highly enriched in hippocampal PV interneurons. ADGRA1 controls the intrinsic excitability of fast-spiking PV neurons and inhibitory synaptic strength onto DG GCs. Furthermore, ADGRA1 activates several G proteins, including Gα13. Recent studies found that the Gα13 signaling pathway is involved in establishing hippocampal PV inhibitory circuitry.²⁷ We postulate that ADGRA1 modulates a Gα13-dependent signaling pathway in PV neurons that tunes excitability, thereby influencing PV synaptic output onto post-synaptic targets. Our results suggest that ADGRA1 controls PV synapse function rather than their overall density in the DG circuit. Thus, ADGRA1 likely functions in mature hippocampal inhibitory circuits rather than during development.

The endogenous ligands for ADGRA1 are unknown. Contrary to all other aGPCRs, ADGRA1 lacks a GAIN domain and therefore cannot utilize the primary modes of aGPCR activation mechanisms previously studied involving either the TA or conformational coupling of the NTF and CTF.²⁸ We found that the short extracellular region of ADGRA1 is not involved in basal G protein activation. ADGRA1 may be activated by small-molecule ligands similar to other GPCRs. For example, studies identified that the aGPCR GPR97 is activated by glucocorticoids.²⁹ ADGRA1 may use similar steroid ligands independent of the GAIN and TA. Alternatively, ADGRA1 may form heterodimers with other GPCRs that are necessary for signaling function. Yet another potential mechanism involves ADGRA1 acting as a constitutively active receptor that is instead regulated through spatial restriction of its localization. Future studies identifying how ADGRA1 is activated will advance our understanding of both inhibitory circuit function and aGPCR activation mechanisms.

While ADGRA1 lacks an extensive extracellular region, it contains a large intracellular C-terminal tail with a PDZ-binding motif. This elaborate tail region is likely involved in ADGRA1 synaptic function. Other aGPCRs, including latrophilins (Lphns), exhibit extensive C-terminal tails essential for their functions. For example, the intracellular tail of Lphn serves as a scaffold for several protein-protein interactions, including postsynaptic excitatory SHANK proteins.^30,31 Lphn3 mutants lacking the C-terminal tail are incapable of rescuing Lphn3-dependent synaptic deficits.⁷ The large intracellular tail of ADGRA1 likely also functions as a scaffold for several protein-protein interactions and may be important for ADGRA1 subcellular localization. We found that deleting the C-terminal PDZ-binding motif had no effect on ADGRA1 synaptic localization. Thus, other sequence features in the large C-terminal tail are likely responsible for the subcellular targeting of ADGRA1. Interestingly, the C-terminal tail of closely related ADGRA3 has been shown to interact with DLG1/SAP97, a scaffold in the MAGUK family that contains three PDZ domains.³² Thus, the ADGRA intracellular tail may function as a scaffold for multiple protein-protein interactions that cluster at the synapse.

We were unable to definitively determine the subcellular localization of endogenous ADGRA1 due to the lack of a reliable antibody. Future studies will be required to determine its endogenous localization and protein-protein interaction partners. Moreover, while we observed localization of ADGRA1 at synapses, consistent with previous studies,²¹ we were unable to definitively measure pre- or postsynaptic localization due to the resolution limits of confocal microscopy. Our studies suggest that ADGRA1 may be present at both pre- and postsynaptic sites, but studies using super-resolution imaging will be necessary to resolve this.

While we determined that ADGRA1 functions in PV interneurons, its physiological function in other brain regions is incompletely understood. We found that ADGRA1 deletion using PV-Cre selectively impaired cued but not contextual fear conditioning. Cued fear conditioning is thought to be hippocampal independent,³³ suggesting that this impairment is due to ADGRA1 function in PV neurons in other areas or multiple circuits simultaneously. Moreover, our mRNA expression analysis shows that ADGRA1 is expressed in various other brain regions, including cortical layer 5/6 and the thalamus. ADGRA1 is likely expressed in cell types other than PV+ neurons within these brain regions. ADGRA1 may engage distinct signal transduction pathways within these different cell types or may interact with different sets of protein-protein interactions within these cells to direct specific functions.

Collectively, these studies identify an orphan receptor selective to PV hippocampal interneurons that is essential for their intrinsic excitability and synaptic output. Further knowledge of the cell surface receptors and downstream signaling pathways responsible for synapse-specific circuit establishment will increase our understanding of how diverse cell types and synaptic connections process information and generate behaviors.

Limitations of the study

Our studies examine the localization of overexpressed HA-tagged ADGRA1 due to the lack of reliable antibodies available for endogenous ADGRA1. Future studies overcoming these (technical barriers will be necessary to determine the precise subcellular localization of endogenous ADGRA1. Moreover, our studies used confocal microscopy, and additional super-resolution studies will be essential toward further defining the synaptic localization of endogenous ADGRA1.

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Richard C. Sando, (richard.sando@vanderbilt.edu).

Materials availability

All materials generated in this study will be openly shared upon request.

STAR★METHODS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Mice

Mice were weaned at 18–21 days of age and housed in groups of 2–5 on a 12 h light/dark cycle with food and water ad libitum. Vanderbilt Animal Housing Facility: All procedures conformed to National Institutes of Health Guidelines for the Care and Use of Laboratory Mice and were approved by the Vanderbilt University Administrative Panel on Laboratory Animal Care. Primary hippocampal cultures were generated from P0 pups from Adgra1 cKO and C57BL/6J mice (Jax, Cat#000664). For generation of Adgra1 cKO mice, sperm was obtained from KOMP MMRRC (Stock number 046501-UCD), and the line was generated at the Vanderbilt Genome Editing Facility. C57BL/6J female mice (Jax, Cat#664) were superovulated for in vitro fertilization using the CARD method (Takeo et al., 2019, Methods Mol Biol³⁵). There was a 52% fertilization rate, and 100 embryos were transferred generating 25 pups. The following CRE driver lines and reporter lines were used: EMX-Cre, B6.129S2-Emx1^tm1(cre)Krj/J (Jax, Cat#5628); PV-Cre, B6.129P2- Pvalb^tm1(cre)Arbr/J (Jax #017320); SST-Cre, Sst^{tm2.1(cre)Zjh}/J (Jax, Cat#013044); Ai14, B6.Cg-Gt(ROSA)26Sor^{tm14(CAG-tdTomato)Hze}/J (Jax #007914).

Cell lines

GKO HEK293^36,37 cells used for all TRUPATH BRET2 assays were originally a kind gift from Asuka Inoue (Tokyo University, Japan), and were provided to our studies as a kind gift from Drs. Vsevolod Gurevich and Chen Zheng (Vanderbilt University). GKO HEK293 cells were maintained in DMEM (Gibco, Cat#11995065) containing 10% FBS (Gibco, Cat#16000044), 1X Penicillin-Streptomycin (Corning, Cat# MT30002Cl), and 1× MEM Non-essential Amino Acid (NeA) Solution (Sigma, Cat# M7145) at 37°C and 5% CO₂. HEK293T cells (ATCC # CRL-11268 were maintained in DMEM (Gibco, Cat#11995065) containing 10% FBS (Gibco, Cat#16000044), 1X Penicillin-Streptomycin (Corning, Cat#MT30002Cl) at 37°C and 5% CO₂. Cell lines were maintained for a maximum of 25 passages.

Primary hippocampal cultures

For immunocytochemistry, cover glasses (#0, 12 mm, Carolina Biological Supply Company, Cat#633009) were placed into 24-well plates (Genesee, Cat#25-107MP) and coated for 2 h with 100 μL of 50 μg/mL poly-D-lysine (Gibco, Cat#A38904-01) in a 37°C tissue culture incubator. Excess poly-D-lysine was removed, coverslips were washed 3× with sterile dH₂O and then dried for 30 min. For primary cells, bilateral hippocampi were dissected from P0 mouse pups and collected tissue was dissociated by papain (Worthington Biochemical Corporation, Cat#LS003126) digestion for 20 min at 37°C. After the digestion, the excess papain solution was removed and tissue was washed twice with plating media containing: 5% fetal bovine serum (Life Technologies, Cat#16000044), B27 (Gibco, Cat#17504044), 1:50), 0.4% glucose, and 2 mM glutamine in 1× MEM. Immediately after the second wash, tissue was triturated using a 1000 μL pipette. Next, cells were filtered through a 70 μm cell strainer (Corning, Cat#431751), and plated at a density of 40,000 cells per dish/well in 1 mL plating media. Culture media was exchanged 24 h later (at DIV1) to growth medium, which contained 5% fetal bovine serum, B27, 2 mM glutamine in Neurobasal A (Gibco, Cat#10888022). Cytosine β-D-arabinofuranoside (Sigma, Cat#C6645) was added to a final concentration of 4 μM on DIV3 along with a 50% growth media exchange. Primary hippocampal cultures were infected with respective lentiviral conditions at DIV1, and staining experiments were conducted between DIV10-14.

METHOD DETAILS

Plasmids

Mouse ADGRA1 overexpression cDNAs used in HEK293T cell experiments were encoded in the pEB Multi-Neo vector (Wako Chemicals, Japan), and empty vector controls were empty pEB Multi-Neo. Mouse ADGRA1 corresponded to Uniprot #Q8C4G9 and contained an N-terminal HA tag for expression studies. ΔN-ADGRA1 contained the first 22 residues of ADGRA1 (MTQWDLKTVLSLPQYPGEFLHP) replaced with an 8× glycine flexible sequence. Lentiviral NLS-GFP-Cre or NLS-GFP-ΔCre were encoded in a 3^rd generation lentiviral backbone and driven by the EF1α promoter. All molecular cloning was conducted with the In-Fusion Assembly system (Takara, Cat#638948). Cloning oligonucleotides were synthesized via Integrated DNA Technologies (IDT).

Antibodies

The following antibodies and reagents were used at the indicated concentrations (IHC-immunohistochemistry, ICC-immunocyto-chemistry, IB-immunoblot): anti-HA rabbit (Cell Signaling Technologies, Cat# 3724, 1:1,000, ICC, 1:2,000 IB), anti-Homer1 rabbit (Synaptic Systems, Cat#160003, 1:500 IHC), anti-Bassoon mouse (AbCam, Cat#ab82958, 1:500 IHC,1:1,000 ICC), anti-SHANK2 guinea pig (Synaptic Systems, Cat#162204, 1:1,000 ICC, 1:500 IHC), anti-Syn1/2 rabbit (Synaptic Systems, Cat#106002, 1:500 IHC), anti-VGAT guinea pig (Synaptic Systems, Cat#131004, 1:500 IHC), anti-Synaptoporin rabbit (Synaptic Systems, Cat#102002, 1:500 IHC), anti-Gephyrin mouse (1:2,000 ICC, Synaptic Systems, Cat#147111) anti-Parvalbumin rabbit (SWANT, Cat#PV27a, 1:2,000 IHC), anti-Synaptotagmin-2 (Synaptic Systems, Cat#105225,1:500 IHC), anti-GAD67 mouse (Millipore Sigma, Cat#AB5062P, 1:1,000 ICC), anti-βactin mouse (Sigma, Cat#A1978, 1:10,000 IB) and corresponding fluorescently-conjugated goat secondary antibodies from Life Technologies (1:1,000).

Sparse transfection

Transfections were conducted when primary hippocampal neurons were DIV 5–6. One μg of each plasmid was mixed in a total of 15 μL volume of 0.267 M CaCl₂. The DNA/CaCl₂ mix was then gradually added dropwise to 15 μL of 2× HBS and gently vortexed. The mixture was allowed to sit at room temperature for 30 min. Prior to the end of the incubation period, growth medium was removed from primary neurons and was saved in unused wells. The neurons were washed with prewarmed 1 mL blank DMEM and then 0.5 mL blank DMEM was added to each well. Twenty μL of the transfection mix was added to each well, gently rocked for proper distribution in the medium, and then incubated for 30 min at 37°C. After the incubation was completed, the neurons were washed with 1 mL blank DMEM and then the conditioned growth medium was returned to each well.

TRUPATH BRET2

HEK G protein K.O cells were plated into 12-well plates at a density of 3-4 x 10⁵ cells in 1 mL per well. HEK G K.O. media contained 1× DMEM (Gibco, Cat# 11995065) plus 10% FBS (Gibco, Cat#16000044), 1X Penicillin-Streptomycin (Corning, Cat#MT30002Cl) with 1× MEM Non-essential Amino Acid (NeA) Solution (Sigma, Cat# M7145). Cells were co-transfected 16–24 h after plating with receptor-of-interest and TRUPATH plasmids at 1:1:1:1 DNA ratio (receptor:Gα-RLuc8:Gβ:Gγ-GFP2) via TransIT-2020 (Mirus, Cat#MIR5400). Each condition required 97 μL of room temperature 1× Opti-MEM (Gibco, Cat#31985070), 1 μL each DNA plasmid at 1 μg/μL concentration), and 3 μL of room temperature and gently vortexed TransIT-2020 reagent. The TransIT-2020:DNA complexes mixtures were gently mixed via pipetting 10 times and incubated at room temperature for 20 min before adding dropwise in the well. The plate was rocked gently side to side and incubated at 37°C for 24 h before harvesting. In each well, media was aspirated, and cells were washed with 1 mL warm PBS. Cells were detached with 300 μL warm Versene (Gibco, Cat#15040066) and incubated at 37°C for 5 min then resuspended via gentle pipetting 10 times. Cells were plated in complete DMEM containing 1× NeA at 200 μL with a density of 30,000–50,000 cells per well in Matrigel-coated 96-well assay plate. Each experimental condition was plated into three separate wells within the 96-well assay plate. BRET2 assays were performed 48 h after transfection. In each well, media was aspirated, and cells were incubated in 80 μL of 1× Hanks’ balanced Salt Solution (Gibco, Cat#14175095) with 20 mM HEPES (Sigma, Cat#H3375, pH 7.4) and 10 μL 100 μM Coelenterazine-400a (NanoLight Technologies, Cat#340) diluted in PBS. After 10 min of incubation, BRET2 intensities were measured using a BERTHOLD TriStar² LB 942 Multimode Reader with Deep Blue C filter (410nm) and GFP2 filter (515 nm). The BRET2 ratio was obtained by calculating the ratio of GFP2 signal to Deep Blue C signal per well. The BRET2 ratio of the three wells per condition were then averaged. Net BRET2 was subsequently calculated by subtracting the BRET2 ratio of cells expressing donor only (Gα-RLuc8) from the BRET2 ratio of each respective experimental condition. Net BRET2 differences were then compared as described in the Figures, by subtracting Net BRET2 ratios of conditions overexpressing indicated receptors compared to empty vector (EV). For plasmid copy-dependent BRET2 experiments, conditions were transfected in a 12-well plate format with the same total amount of plasmid DNA and varying copies of experimental plasmid, adjusted to the same total amount with empty vector (pEB-multi).

RiboTag immunoprecipitations

Hippocampal tissue was rapidly dissected, gently transferred to a microcentrifuge tube and flash frozen in liquid nitrogen. Tissue was then stored at −80°C until homogenization. The benchtop surface and all supplies were sprayed with RNAse zap (Invitrogen, Cat#AM9780) prior to homogenization. The Homogenization Buffer (HB), Supplemented Homogenization Buffer (HB-S), and Lysis Buffer (LB) solutions were prepared as follows. HB: 1% IGEpal (Spectrum Chemical MFG Corp, Cat#I1112), 100 mM KCl (Sigma, Cat#P5405-500g), 50 mM Tris, pH 7.4 (Sigma, Cat#T6066-500G), and 12 mM MgCl₂ (Sigma-Aldrich, Cat#63069-100mL) to final volume with nuclease-free dH₂O. HB-S: Previously prepared HB, 100 μg/mL Cycloheximide (5 mg/mL, Sigma, Cat#C7698-1G), 1:100 dilution Mammalian Protease Inhibitor Cocktail (Sigma-Aldrich, Cat#P8340-1mL), 1 mg/mL Heparin (50 mg/mL, sodium salt from porcine intestine, (Sigma, Cat#H3393-100KU), 40 U/mL RNAsin (Promega, Cat#N251A), 1 mM DTT (Sigma, Cat#D0632-10G). LB:1% β-Mercaptoethanol in RLT Plus Buffer (Qiagen, Cat#1030963; Sigma, Cat#M3148-100mL). Once solutions were prepared, tissue was homogenized in 10% weight/volume of HB-S in a Dounce homogenizer (Kimble, Cat#885301-0007). The homogenate was transferred to fresh microcentrifuge tubes and spun for 10 min at 4°C at 10,000 × g. The supernatant was collected and 50 μL of supernatant was transferred into a new tube and 350 μL of the previously mentioned LB added to generate the Input sample, which was flash frozen in liquid nitrogen. Pierce Anti-HA magnetic beads (Thermo Scientific, Cat#88836) 30 μL per sample was placed into a microcentrifuge tube in a magnetic rack and the storage buffer was removed. Beads were subsequently washed in 1 mL of HB-S on a gentle rotator for 10 min at 4°C. Washed beads were resuspended with HB-S. The remainder of the tissue homogenate was added to magnetic beads and incubated on a gentle rotator at 4°C overnight. After the incubation, the beads were then washed 3 × 10 min in High Salt Buffer (HSB) containing 300 mM KCl, 1% IGEpal, 50 mM Tris (pH 7.4), 12 mM MgCl₂, 100 μg/mL Cycloheximide, 1 mM DTT and nuclease-free dH₂O. 350 μL of LB was added to the beads and the samples were resuspended on a vortex mixer on low speed for 30 s. Samples were placed back on the magnetic rack and the IP supernatant was collected. RNA was subsequently purified from Input and IP samples using the RNeasy Micro Kit (Qiagen, #74134), eluting in 20 μL of RNase-free water. The purified eluted RNA was converted to cDNA using SuperScript IV (Invitrogen, Cat#18090050) with random hexamers.

RT-qPCR

RNA was extracted from primary hippocampal neurons between DIV10-12. Each well was briefly washed with 1× PBS, followed by the addition of 250 μL Trizol/well. The plate was rocked for 5 min at RT, and all wells/samples were pooled into a single tube, bringing the final volume to 1 mL with additional Trizol. The mixture was incubated at room temperature for 5 min, then 200 μL chloroform was added and mixed gently by inverting the tube 20 times. After 3-min incubation at RT, samples were centrifuged at 12,000 × g for 15 min at 4°C. The upper aqueous phase (~500 μL) was carefully collected without disturbing the interphase and transferred to a new tube. An equal volume (500 μL) of 2-propanol was added and mixed by inversion 20 times, followed by a 10-min incubation at RT. The RNA was precipitated by centrifugation at 12,000 × g for 10 min at 4°C. The supernatant was discarded, and the pellet was washed two times with 500 μL of 70% ethanol by spinning at 7,000 × g for 5 min at 4°C. After removing all remaining supernatant, the pellet was immediately resuspended in 100 μL of RNase-free dH₂O pre-warmed to 55°C, then incubated at 55°C for 10 min. During this incubation, Buffer RLT with β-Mercaptoethanol (10 μL per 1 mL RLT) was prepared. Finally, the tube was gently tapped to fully dissolve the RNA pellet, and 350 μL of the RLT/β-Mercaptoethanol solution was added to the sample. Total RNA was isolated using the RNeasy Micro Kit (Qiagen, Cat#74134), and then used for cDNA synthesis with Super Script IV (Invitrogen, Cat#18090050) with random hexamers. qPCR with PowerUp SYBR Green Master Mix (Applied Biosystems, Cat#A25742) was used for quantification of relative gene expression. GAPDH expression was used to normalize among samples. qPCR primer pairs (PrimeTime, IDT) were used mentioned in Table S1.

Surface labeling and immunocytochemistry

Cells were washed briefly once with PBS, fixed with 4% PFA (Electron Microscopy Sciences, Cat#15714)/4% sucrose/PBS for 20 min at 4°C, and washed 3 × 5 min in PBS. After the final wash, blocking buffer containing 4% BSA (Sigma, Cat#10735086001)/3% normal goat serum (Jackson Immunoresearch, Cat#005000121)/PBS was applied for 30 min. For surface immunolabeling of HA-ADGRA1, samples were subsequently incubated overnight at 4°C with anti-HA rabbit (Cell Signaling Technologies, Cat#3724) primary antibody solution diluted in blocking buffer. The next day, samples were washed 3 times with PBS, followed by permeabilization with 0.2% Triton X-100/PBS for 5 min at room temperature. After permeabilization, samples were transferred into blocking buffer and incubated at RT for 30 min. For total labeling, the following primary antibodies were used: anti-MAP2 chicken (1:5,000; EnCor, Cat#CPCA-MAP2), anti-SHANK2 guinea pig (1:1,000; Synaptic Systems, Cat#162204), anti-Bassoon mouse (1:1000; AbCam, Cat#AB82958), anti-GNA13 mouse (1:1000; Proteintech, Cat#67188-1), anti-GAD67 mouse (1:1,000; Millipore Sigma, Cat#AB5062P) and anti-Gephyrin mouse (1:2,000; Synaptic Systems, Cat#147111). Samples were incubated with primary antibodies for 2 h at RT. Samples were then washed 5 × 5 min in PBS, incubated with fluorescently conjugated secondary antibodies diluted in blocking buffer for 1 h at room temperature. Corresponding secondary antibodies were goat anti-chicken 647 (1:1,000; ThermoFisher, Cat#A21449), goat anti-rabbit 488 (1:1,000 ThermoFisher, Cat# A11034), goat anti-guinea pig 555 (1:1,000; ThermoFisher, Cat#A21435), and goat anti-mouse 546 (1:1,000; ThermoFisher, Cat#A11003). Samples were labeled with DAPI (Sigma, Cat#10236276001) diluted into PBS for 5 min, then washed 4 × 5 min with PBS. Samples were mounted on UltraClear microscope slides (Denville Scientific, Cat# M1021) using 10 μL ProLong Gold antifade reagent (Invitrogen, Cat#P36930) per coverslip.

Immunoblotting

Primary hippocampal cultures were briefly washed 1X with PBS, and samples were collected in sample buffer containing 312.5 mM Tris-HCl pH 6.8, 10% SDS, 50% glycerol, 12.5% 2-mercaptoethanol, bromophenol blue, and protease inhibitors (Roche, Cat# 11873580001) and run on 12% SDS-PAGE gels (BioRad mini-protean TGX, Cat#4561024) at 30 mA/gel constant current. The HiMark Prestained Protein Standard (Invitrogen, Cat# LC5699) was used as a protein molecular weight ladder. Protein was transferred onto nitrocellulose transfer membrane in transfer buffer (25.1 mM Tris, 192 mM glycine, 20% methanol) at 200 mA constant current for 2 h at 4°C. Membranes were blocked in 4% bovine serum albumin (BSA, Sigma, Cat#10735086001)/TBST (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 0.05% Tween 20) for 1 h at room temperature, incubated in primary antibodies diluted into 4% BSA/TBST overnight at 4°C (anti-HA rabbit 1:2,000; anti-βactin mouse 1:10,000), washed 3 × 5 min in TBST, incubated in corresponding secondary antibodies (Licor IRDye 800CW donkey anti-mouse, Cat#92632212; anti-rabbit #92632213) diluted 1:10,000 into TBST, washed 5 × 5 min in TBST, and imaged on a Licor Odyssey system.

Immunohistochemistry

Mice were briefly anesthetized with isoflurane and transcardially perfused with 10 mL room temperature heparinized (10 U/mL, Sigma, Cat#H3393) PBS, followed by 25 mL room temperature 4% PFA/PBS. The brains were post-fixed for 2 h at 4°C, washed with PBS, cryoprotected in gradients of 10% sucrose/PBS, 20% sucrose/PBS, and 30% sucrose/PBS, rapidly embedded in OCT (Fisher, Cat#23730571), and sliced on a cryostat at 25 μm. The 25 μm thick, free-floating sections were washed 1 × 5 min with PBS/0.1% Triton X-100, blocked for 1 h at room temperature in blocking buffer containing 4% BSA (Sigma Cat#10735086001)/3% normal goat serum (Jackson Immunoresearch Cat#005000121)/0.1% Triton X-100/PBS and incubated overnight at 4°C with primary antibodies diluted in blocking solution. The following primary antibodies were used: anti-vGAT guinea pig (1:1,000; Synaptic Systems Cat#131004), anti-Bassoon mouse (1:500; AbCam, Cat#AB82958), anti-Synaptoporin rabbit (1:500; Synaptic Systems, Cat#102002), anti Syn1/2 rabbit (1:500; Synaptic Systems Cat#106002), Anti-Synaptotagmin-2 guinea pig (1:500; Synaptic Systems, Cat#105225), anti-Homer1 rabbit (1:500; Synaptic Systems, Cat#160003), anti-Parvalbumin rabbit (1:2,000; Swant, Cat#PV27a), and anti-SHANK2 guinea pig (1:500; Synaptic Systems, Cat#162204). Primary antibody incubation was followed by 3 × 5 min washes in PBS/0.1% Triton X-100 and 2 h room temperature incubation with corresponding fluorescently labeled secondary antibodies in blocking buffer. Samples were labeled with DAPI (Sigma, Cat#10236276001) diluted into PBS/0.1% Triton X-100 for 15 min, washed 5 × 5 min with PBS/0.1% Triton X-100, and mounted on glass slides coated in 0.1% Triton X-100/PBS, dried briefly, and covered with ProLong Gold antifade reagent (Invitrogen, Cat#P36930) and cover glass (Corning, Cat#2980-246).

Lentivirus production for culture experiments

Lentiviruses were packaged in HEK293T cells from ATCC (CRL-11268). For lentiviral production, co-transfection of the expression shuttle vector and the three helper plasmids (pRSV-REV, pMDLg/pRRE and vesicular stomatitis virus G protein (VSVG)) was done with FuGENE6 (Promega E2691) using 2.5 μg of each plasmid per 9.6 cm². Lentiviral-containing medium was collected 48 h after transfection, briefly spun down 5,000 xg for 5 min for removal of cellular debris and then stored at 4°C. The LV genomic titer was estimated using PowerUp SYBR Green Master Mix for qPCR (Applied Biosystems, A25742) with the following primers: F - ccactgctgtgccttggaatgc, and R - aatttctctgtcccactccatccag. Shuttle plasmids at 10× serial dilutions (1 × 10⁵ − 1 × 10⁹ copies/mL) were used for generating a standard curve. After quantification, the LVs were directly applied to primary neuron culture medium.

RNA in situ hybridizations

Wild-type C57BL/6J (Jackson, Cat#000664) mice were taken from their home cages at postnatal day (P) 5, P10, P21, and P30, and whole brain tissue was collected in the following manner. Brains were rapidly dissected following brief anesthesia with either ice (P5) or isoflurane (P10, P21, P30) and placed in a rectangular cryomold (Epredia Peel-a-way, Cat#18–30) which was flash-frozen in liquid N₂ for 15 s to allow for indirect exposure of the tissue with liquid N₂. The brain was subsequently embedded in O.C.T. Compound (Fisher Healthcare, Cat#4585) within a second cryomold using a bath of 2-methylbutane (Sigma-Aldrich, Cat#M32631-46) chilled with dry ice. Once frozen, the blocks were stored at −80°C until cryosectioning. The frozen blocks were removed from the −80°C freezer and allowed to equilibrate in the cryostat (Leica CM 1950, Cat#047742456) at −20°C for 1 h. The blade for slicing (Sakura, Cat#4689), forceps, razor blade for block trimming, and paintbrushes for manipulating sections, and the anti-roll plate (Leica, Cat#14047742497) were also all placed in the cryostat and allowed to equilibrate. Tissue was sectioned at 15 μm and mounted directly onto room temperature Diamond White Glass microscope slides (Globe Scientific Inc., white frosted 25 × 75 × 1mm, charged +/+, Cat#1358W). Once mounted, the slides were kept in the cryostat until all sectioning was complete. Sections were sub-sequently dried at −20°C for 1 h, then stored at −80°C.

Tissue Pre-Treatment: The RNAscope Multiplex Fluorescent manual assay (Advanced Cell Diagnostics, Cat#323100) was carried out using the Fresh Frozen sample preparation according to the manufacturer’s protocol as described below. The RNAscope Hydrogen Peroxide (Advanced Cell Diagnostics, Cat#322335) and RNAscope Protease IV (Advanced Cell Diagnostics, Cat#322336) reagents were set out on the benchtop to equilibrate to room temperature. Sections were removed from the −80°C freezer and placed immediately into ice-cold 4% PFA (Electron Microscopy Sciences, Cat#15714)/PBS (MP Biomedicals, Cat#2810306) within a glass slide holder (Epredia RA Lamb Glass Coplin Jar, Fisher, Cat#E94). Slides were incubated at 4°C for 15 min to fix the tissue and subsequently washed twice with 1× PBS. Slides were subsequently dehydrated in the following ethanol (Decon Laboratories, Inc., 200 proof, Cat#2705HC) gradient: 50% EtOH/ddH₂O for 5 min, 70% EtOH/ddH₂O for 5 min, followed by two treatments with 100% EtOH (50 mL of each treatment). After the final 100% step, the slides were placed section side up on a paper towel and allowed to dry for 5 min. Then a hydrophobic pen (IHC World, super pap pen, Cat#SPR0905) was used to draw a barrier around each section, which dried at room temperature for 5 min. While the barriers were drying, the HybEZ Humidity Control Tray with lid (Advanced Cell Diagnostics, Cat#310012) was prepared. A sheet of HybEZ Humidifying Paper (Advanced Cell Diagnostics, Cat#310025) was placed on the bottom of the tray and sprayed with ddH₂O until damp. The EZ-Batch Slide Holder (Advanced Cell Diagnostics, Cat#310017) was placed inside the humidity control tray and the slides were placed in the holder. Three drops of hydrogen peroxide were added to each section. The cover was placed over the humidity control tray and the slides were left to incubate for 10 min at room temperature. Once the incubation was complete, the slides were washed twice with ddH₂O, removing excess liquid after each wash with a vacuum aspirator. Slides were reinserted into the slide holder and 4 drops of Protease IV were added to each section, followed by incubation at room temperature for 30 min. While the slides were incubating the RNAscope Probes were prepared in the following manner. Probes (Adgra1-C3 #492281, PV-C2 #421931, SST-C2 #404631, CCK-C2 #402271, Calb2-C2 #313641) were placed in a heat block at 40°C for 10 min, and subsequently removed from the heat and incubated at room temperature for 10 min. The probes were combined to form a probe mix consisting of 100 μL of probe diluent per section, and 2 μL each of the C2 and C3 probes per section, respectively. The HybEZ II Oven (Advanced Cell Diagnostics, Cat#321720) was then pre-warmed to 40°C. Once the Protease IV incubation was complete, the slides were washed twice with 1× PBS, placed back in the slide holder and humidity control tray and 100 μL of probe mix were added to each section. The tray was placed in the HybEZ Oven for 2 h at 40°C to hybridize the probes. While the incubation was occurring, the 1× RNAscope wash buffer solution (Advanced Cell Diagnostics, #310091) and 5× Saline Sodium Citrate (SSC) buffer were prepared. The 20× SSC stock contained 175.3 g of NaCl (Fisher Chemical, certified ACS, crystalline, #S271-1) and 88.2 g of sodium citrate (Fisher Chemical, dihydrate, granular, Cat#S279-500) in ddH₂O, pH 7.0. Following the 2-h period, slides were washed twice with 1× RNAscope wash buffer and placed in 5× SSC buffer overnight at 4°C.

RNAscope Multiplex Fluorescent Assay: The following reagents were equilibrated at room temperature for 1 h: RNAscope Multi-plex FL v2 AMP1 (Advanced Cell Diagnostics, 323101), RNAscope Multiplex FL v2 AMP2 (Advanced Cell Diagnostics, 323102), RNA-scope Multiplex FL v2 AMP3 (Advanced Cell Diagnostics, 323103), RNAscope Multiplex FL v2 HRP C1 (Advanced Cell Diagnostics, 323104), RNAscope Multiplex FL v2 HRP C2 (Advanced Cell Diagnostics, 323105), RNAscope Multiplex FL v2 HRP C3 (Advanced Cell Diagnostics, 323106) and RNAscope Multiplex FL v2 HRP Blocker (Advanced Cell Diagnostics, 323107). While this equilibration was occurring, the HybEZ Oven was equilibrated to 40°C. Slides were removed from 5× SSC and washed twice with RNAscope wash buffer. Three drops of the AMP1 were applied to each section and the humidity control tray was placed back in the oven where it incubated for 30 min at 40°C. Slides were subsequently washed twice with wash buffer and placed back in the slide holder and 3 drops of AMP2 were applied to each section and left to incubate in the oven at 40°C for 30 min. Slides were washed twice and treated with 3 drops of AMP3 at 40°C for 15 min. While the AMP3 incubation was occurring, the dye solutions were prepared as follows. The dyes were prepared at a concentration of 1:1,000 by combining 1000 μL of TSA Buffer (Advanced Cell Diagnostics, Cat#322809) with 1 μL of Opal 520 Reagent (in DMSO, Akoya Biosciences, Cat#OP-001001) and Opal 690 Reagent (in DMSO, Akoya Biosciences, Cat#OP-001006) respectively.

Once the 15-min AMP3 incubation was complete the slides were washed and inserted back into the slide holder, then 3 drops of HRPC1 were applied to each slide, and that was allowed to incubate for 15 min at 40°C. Once that was complete the slides were washed and returned to the slide holder and 3 drops of the HRP Blocker were added to each section. This was allowed to incubate for 15 min at 40°C. Slides were washed and 150 μL of the Opal 520 dye mix was added to each section. This was allowed to incubate at 40°C for 30 min. This process was then repeated for the C3 channel, which was treated with Opal 690 dye.

Counterstaining Mounting and Imaging: Four drops of RNAscope DAPI (Advanced Cell Diagnostics, Cat#323108) were added to each section for the purpose of counterstaining and left to sit at room temperature for 30 s. The DAPI was gently tapped off the slide and 50 μL of Prolong Gold antifade reagent (Invitrogen, Cat#P36930) was added inside the barrier but not directly touching the section, avoiding bubbles. A glass coverslip (Corning, 24 × 60 mm, Cat#2975-246) was lowered onto the slide, slides were allowed to dry overnight in a dark slide box (Fisher Brand, Cat#03-448-4) at 4°C before imaging. Slide boxes were stored in the cold room a 4°C for long term storage. Three separate mice were analyzed for each postnatal age, and quantitative data depicts the average values from three mice.

Double immunohistochemistry/RNA in situ hybridizations

NeuN or GFAP immunohistochemistry/Adgra1 RNA in situ experiments were conducted in the following manner, essentially as described in the manufacturer’s protocol (Advanced Cell Diagnostics Cat#323180 and Cat#323100). Tissue collection, sectioning, and pretreatment were conducted as described above for standard RNA in situ experiments up until the initial 10-min room temperature hydrogen peroxide treatment. Following hydrogen peroxide treatment, slides were washed twice with ddH₂O followed by once with 1x-PBS-T (PBS with 0.1% Tween 20). The slides were returned to the slide holder and 150 μL of the primary antibody (anti-NeuN Mouse, EMD Millipore Corp., Cat#MAB377; anti-GFAP, Invitrogen, Cat#PA1-10004) diluted in RNAscope Co-Detection Antibody Diluent (Advanced Cell Diagnostics, Cat#323160) in a 1:500 concentration was added to each section. Slides were incubated at 4°C overnight in the humidity control tray.

Post-primary Fixation and Protease Treatment: After incubation with the primary antibody, slides were washed three times with 1x-PBS-T at room temperature. Then slides were submerged in 10% Neutral Buffered Formalin (Sigma-Aldrich, Cat#65346-85) for 30 min at room temperature. Following that incubation, slides were washed four times with PBS-T. Slides were subsequently placed back into the humidity control tray, and 4 drops of Protease 4 were added and incubated for exactly 30 min at room temperature. After incubation the slides were washed three times with ddH₂O. The RNAscope Multiplex fluorescent assay was then performed, as described above. Following the last HRP blocker step in the RNAscope Multiplex assay, immunofluorescence for NeuN or GFAP was performed. For NeuN, HRP-conjugated goat anti-mouse secondary antibody (Biotium #20400-1mL) was diluted in Co-Detection Antibody Diluent (Advanced Cell Diagnostics, #323160) at a 1:500 concentration was added to completely cover the sections and allowed to incubate at room temperature for 30 min. For GFAP, goat anti-chicken Alexa Fluor 647 (Invitrogen # A21449) was diluted in Co-Detection Antibody Diluent at a 1:500 concentration and allowed to incubate at room temperature for 30 min. Slides were subsequently washed twice with 1× PBS-T. For NeuN, 150 μL of the previously prepared Opal dye (Akoya Biosciences) was added to the slides and incubated for 10 min at room temperature. Then the slides were washed twice with 1× PBS-T and ready for counterstaining and mounting as described above for standard in situs.

Confocal imaging of fixed samples

Images were acquired using a Nikon A1r resonant scanning Eclipse Ti2 HD25 confocal microscope with a 10× (Nikon #MRD00105, CFI60 Plan Apochromat Lambda, N.A. 0.45), 20× (Nikon #MRD00205, CFI60 Plan Apochromat Lambda, N.A. 0.75), and 60× (Nikon #MRD01605, CFI60 Plan Apochromat Lambda, N.A. 1.4) objectives, operated by NIS-24 Elements AR v4.5 acquisition software. Laser intensities and acquisition settings were established for individual channels and applied to entire experiments, and images were collected at the following resolution: 10× − 1.73 μm/pixel, 20× − 0.62 μm/pixel, 60× − 0.29 μm/pixel, 60× with deconvolution − 0.07 μm/pixel. Brightness was adjusted uniformly across all pixels for a given experiment for Figure visualization purposes. Images were pseudocolored for Figure visualization purposes. Quantification of fluorescence intensities was conducted by imaging 3–5 image frames per biological replicate, which were averaged to generate a single biological replicate value.

Image analysis (immunocytochemistry)

Images were analyzed using NIS Elements AR 5.42 (Nikon) with a custom-made pipeline in NIS Elements GA3 (Nikon). For synaptic puncta detection, binary masks were applied to each channel after background subtraction with rolling ball and binary mask settings were maintained across the images and experiments. Objects detected smaller than 0.010 μm filtered out. For synaptic puncta co-localization, the centroid coordinate of each object was calculated. The distance between individual puncta was calculated on MATLAB based on centroid approximation. Colocalization was defined if the distance between two centroid coordinates was less than 500 nm. Percent colocalization was calculated as the proportion of colocalized puncta count relative to total puncta count per channel. All ICC data were collected and analyzed blindly.

Image analysis (immunohistochemistry)

Images were analyzed using NIS Elements AR 5.42 (Nikon) with a custom-made pipeline in NIS Elements GA3 (Nikon). For synaptic puncta detection, binary masks were applied to each channel after background subtraction with rolling ball and binary mask settings were maintained across the images and experiments. Objects detected smaller than 0.010 μm were filtered out. Object count, volume, mean intensity were measured with binary processing. For analyzing Synaptotagmin-2 (Syt-2) puncta located on PV soma, binary masks are applied to maximum intensity projection images. Next, the Syt-2 channel was processed using a rolling ball background subtraction. Subsequently, a region of interest was determined via the Ai14 reporter. The PV cell soma object area was then calculated. Object count, object intensity and object area were calculated from the region-of-intertest. Objects detected smaller than 0.010 μm filtered out. All IHC data were collected and analyzed blindly.

Morphological analysis of granule cells

Biocytin-filled cells were imaged on a Nikon A1r resonant scanning Eclipse Ti2 HD25 confocal microscope as described above. Images for Imaris 3D reconstructions were taken at 60× magnification. X-Y stitches together with Z-stacks were collected to encompass the entire cell at a Z-step size of 0.175 μm. For dendritic spine quantification, each filled cell was imaged at 60× magnification to examine dendritic spines in the molecular layer of the dentate gyrus. For each cell, multiple dendritic branches were examined and analyzed to obtain average spine density. For 3D Imaris reconstructions, images were converted to Imaris compatible format (.imd). Images were reconstructed into 3D surface using the “surfaces” tool and processed further with a default background subtraction level. The region of interest was defined to include the full cell in all 3 dimensions. Dendrites were reconstructed as “filaments” by choosing the “autopath no spine function”. The program was trained through multiple iterations until accurate detection and reconstruction of dendrites in all three dimensions. Quantitative analysis of the cell was exported from the software. Filament Segment Length in microns was the sum of all segment lengths within the entire filament graph. Filament No. of Sholl Intersections was defined as the number of segment intersections on concentric spheres (1.0 μm), defining segment spatial distribution as a function of distance from the beginning point (soma). Filament No. of Segment Branches was calculated by following a segment from the segment beginning point (soma) along all segments for the defined distance, and then counting how many different branches have been reached. Imaris calculates this as: Distances 0 to n * DistanceIncrement, where DistanceIncrement has a default of value of 1 μm and n is dependent on the maximum segment terminal point distance to the beginning point of the filament.

Acute slice electrophysiology

Mice were deeply anesthetized with isoflurane, decapitated and their brains were quickly removed and placed into ice-cold solution containing (in mM): 228 Sucrose, 2.5 KCl, 1 NaH₂PO4, 26 NaHCO₃, 0.5 CaCl₂, 7 MgSO₄, 11 D-Glucose saturated with 95% O₂/5% CO₂. Transverse hippocampal slices (300 μm thick) were cut by a vibratome (Leica VT 1200S) and transferred to a holding chamber containing artificial cerebrospinal fluid (ACSF, in mM): 119 NaCl, 2.5 KCl, 1 NaH₂PO₄, 26 NaHCO₃, 2.5 CaCl₂, 1.3 MgSO₄, 11 D-Glucose, ~290 mOsm. Slices were recovered at 32°C for 30 min, followed by recovery at room temperature for 1 h in the same holding chamber. Acute slices were transferred to a recording chamber continuously perfused with oxygenated ACSF (1.5 mL/min) maintained at 32°C. Whole cell recordings were performed from the granule cell layer of the dentate gyrus. For whole-cell patch-clamp experiments, the patch pipettes were pulled from borosilicate glass capillary tubes (World Precision Instruments, Cat#TW150-4) using a PC-100 pipette puller (Narishige PC-100). The resistance of pipettes filled with whole cell pipette solution varied between 3 and 5 MΩ. Synaptic currents were monitored with a Multiclamp 700B amplifier (Molecular Devices) synchronized with Clampex 11.2 data acquisition software (Molecular Devices). Electrophysiological data were digitized with Digidata 1550B (Molecular Devices). The recording rig contained a Nikon Eclipse FN1 microscope controlled via NIS Elements software with 4X (CFI60 Plan Fluor 4X objective lens, N.A. 0.13) and 40X (CFI60 Apochromat 40X water dipping lens, N.A. 0.8), pco Edge 4.2 LT sCMOS camera (Cat#77067009), and Sutter micromanipulators (MPC-200). Fluorescent reporters were visualized with different wavelengths from an Aura III light engine (Lumencor). For voltage-clamp recordings of excitatory transmission, a whole-cell pipette solution was used containing (in mM) 135 Cs-Methanesulfonate, 8 CsCl, 10 HEPES, 0.25 EGTA, 0.3 Na₂GTP, 2 MgATP, 7 phosphocreatine, 0.1 Spermine (pH 7.3, adjusted with CsOH and 302 mOsm). For voltage-clamp recordings of inhibitory transmission, a whole cell pipette solution was used containing (in mM) 146 CsCl, 10 HEPES, 0.25 EGTA, 2 MgATP, 0.3 Na₂GTP, 7 phosphocreatine, 0.1 Spermine (pH 7.3, adjusted with CsOH and 296 mOsm). The external bath solution contained (in mM) 140 NaCl, 5 KCl, 2 CaCl₂, 0.8 MgCl₂, 10 HEPES, and 10 glucose (pH 7.35, adjusted with NaOH). AMPAR- and NMDAR-excitatory postsynaptic currents (EPSCs) were pharmacologically isolated by adding the γ-aminobutyric acid receptor blocker picrotoxin (50 μM; Tocris, Cat#1128) to the extracellular bath solution. Inhibitory postsynaptic currents (IPSCs) were isolated by adding AP5 (50 μM; Tocris Cat#0106) and CNQX (10 μM; Tocris, Cat#1045) to the extracellular bath solution to block NMDA and AMPA receptors, respectively. Spontaneous miniature postsynaptic currents (mEPSCs and mIPSCs) were monitored in the presence of tetrodotoxin (1 μM; Tocris, Cat#1069) to block action potential triggered neurotransmitter release. Whole cell patch clamp experiments were conducted while holding the cells at −70 mV. Synaptic currents were sampled at 10 kHz and analyzed offline using Clampfit 11.2 software (Molecular Devices). Miniature events were monitored for 5 min 10 s and the last 4 min 45 s of each trace was analyzed. Miniature events were analyzed using the template matching search. Events smaller than 5 pA were excluded and each event was visually inspected for inclusion or rejection. For voltage-clamp recordings of evoked postsynaptic currents (eIPSCs) QX-314 (1 mM; Tocris, Cat#1014), a blocker of voltage gated Na+ channels, was added to the previously described inhibitory whole cell patch pipette solutions. Local stimulation was provided for eIPSCs recordings using a concentric bipolar electrode (FHC, Cat#CBAEB75) immersed into external bath solution and controlled by Sutter micromanipulator (MPC-200). The concentric bipolar electrode was placed in the granule cell layer and was kept at a consistent distance (150 μm) with minimal variability from each cell to deliver stimulation to the field. Stimulation intensities were set to 50-100-200-400 μA in increasing order. For measuring paired-pulse ratio (PPR), paired pulses with 400 μA intensity were delivered with the bipolar electrode with the following order of intervals: 25-50-100-250-500 ms. Five sweeps were measured for each inter-stimulus interval. Between interval changes, 30 s period was allowed for the recovery of the neurotransmitter pool. For current-clamp recordings of PV neurons K gluconate-based whole cell pipette solution was used containing (in mM) 115 K-gluconate, 20 KCl, 10 HEPES, 0.5 EGTA, 0.3 Na₂GTP, 2 MgATP, 0.1 Spermine (pH 7.3, adjusted with KOH and 291 mOsm). Coronal hippocampal slices (260 μm thick) were cut by a vibratome (Leica VT 1200S) and recovered like previously mentioned. PV cells were visualized with the fluorescent reporter Ai14. PV cells were recorded initially in voltage clamp mode to monitor spontaneous activity. Intrinsic excitability was assessed by evoking action potentials (AP) with current injection steps. 1 s square current pulses were injected into the patched cell, beginning at −100 pA and increasing in 50 pA steps to a max current injection of 850 pA. To quantify spike frequency (Hz), the number of spikes was divided by the length of the current pulse (1 s). Input-output curves were drawn based on spike frequency. Rheobase was calculated as the minimum current injection that evoked an action potential. The resting membrane potential was measured as the membrane potential at steady state. Action potential threshold, amplitude, and after hyperpolarization were calculated with Easy Electrophysiology automated software. Input resistance was calculated by dividing the difference between the average voltage response to a −100 pA injection (measured at steady state) and the baseline voltage (resting membrane potential) to injected current. The voltage sag was determined by the difference between the minimum voltage at the peak deflection to a −100 pA current injection and the voltage of the steady-state response. For biocytin fill, internal solution was made as above except with the addition of 2 mg/mL Biocytin (Sigma, Cat#B4261). After the recordings, the patch pipette was gently removed, and the slices were transferred to PBS in a 24-well plate. The PBS was immediately exchanged with 4% PFA/PBS and slices were incubated overnight at 4°C, followed by 5 × 5 min washes with PBS. Samples were permeabilized for 30 min in 0.3% Triton X-100/PBS at room temperature, blocked for 1 h at room temperature in 5% normal goat serum/0.1% Triton X-100/PBS, and incubated with Streptavidin Alexa Fluor 647 conjugate (ThermoFisher, Cat#S21374) diluted 1:1,000 into blocking buffer for 90 min at room temperature. Samples were labeled with DAPI (Sigma, Cat#10236276001) diluted into PBS for 5 min. Samples were subsequently washed 4 × 5 min with PBS and mounted as described for immunohistochemistry.

Stereotactic injections

Stereotactic injections were performed on P21-25 mice anesthetized with 1–5% vaporized isoflurane. The mice received a preoperative subcutaneous injection of 0.05 mg/g meloxicam (PCAA 55–4476) and twice again at 24 and 48 h post procedure. Mouse heads were shaved and cleaned with Betadine followed by 70% ethanol. Heads were secured to a stereotactic rig (Stoelting Digital Lab Standard with mouse and neonate adaptor), and lubricant was applied to eyes (Puralube Vet Ointment). A small incision was made through the scalp with sterilized tools. Viral solution was injected with a glass pipette at a flow rate of 0.9 μL/h and a total volume of 0.2 μL per injection. Viral solution was delivered using a syringe pump (World Precision Instruments, Cat#SP100I), Hamilton 1701RNR 10 μL Syringe (Cat#80065) with 18-gauge syringe needle (Hamilton, RN NDL 18- gauge S Cat#7804-06) completely continuous with mineral oil to a sharp beveled glass pipette (Warner Instruments G120F-4). Coordinates used for bilateral dentate gyrus injections were A/P −1.80 mm, M/L ± 1.30 mm, and D/V −2.05 mm. To prevent capillary action, the injection pipette was left at the injection site for 10 min post-injection, then slowly raised +0.05 mm dorsally and left 2 min before being slowly removed. Incisions were sutured and mice were removed from rig. Mice were monitored in a fresh cage where they were kept warm until full recovery. For neonatal injections in rescue experiments, pups were anesthetized for 5 min on ice. Rescue AAVs were co-injected with AAV encoding mClover3 as an injection site marker. Heads were secured to a stereotactic rig with neonatal adaptor (Stoelting Digital Lab Standard with mouse and neonate adaptor). Viral solution was injected using a previously mentioned syringe pump system. Injections made with a glass pipette at a flow rate of 60 μL/h and a total volume of 0.3 μL per injection. Coordinates were determined based on lambda for dentate gyrus bilaterally. Three serial injections made with the following coordinates: A/P −1.00 mm, M/L ± 1.00 mm and respective D/V coordinates −1.80 mm, −1.50 mm and −1.20 mm to cover the region including dentate gyrus. Efficiency and localization of viral injection was confirmed by a Nikon Eclipse FN1 microscope.

Mouse behavior

Previously mentioned ADGRA1 PV and SST cKO mouse lines were used for the behavior assays. All mice were aged to 8 weeks, and approximately equal numbers of male and female mice were used. Control and cKO mice were randomly distributed across cages and experimenters were blinded to the genotype of the mice throughout the experiments. All animals were housed in a temperature-and humidity-controlled housing facility and were kept on a 12:12 h light cycle. All procedures were approved by the Vanderbilt Institutional Animal Care and Use Committee. All behavioral assays were conducted at the Vanderbilt Murine Neurobehavioral Laboratory Core. Mice were acclimated to the facility at least two weeks before the experiments. Assays were run at the same time of the day between 7 a.m. and noon. Assays were conducted in the following order: Open field assay, fear conditioning, seizure induction.

Open field assay

Exploratory locomotor activity was measured in chambers measuring 27 × 27 cm (ENV-510; MED Associates, Georgia, VT, USA). Chambers were housed in sound-attenuating cases to restrict surrounding stimuli during testing. Infrared beams and detectors were used to record the movement in the open field arena. Locomotor activity was measured over 60 min. Overall activity in the box, rearing count, and distance traveled in the center area of the box were measured. These parameters were measured automatically by beam breaks. Time spent exploring the center area (19.05 × 19.05 cm) versus periphery (50% of surface area) of the chamber was also calculated.

Fear conditioning assay

Mice were placed in a sound-attenuating chamber with a wire grid floor capable of transmitting an electric shock. Mice and movements were monitored by cameras fixed to the doors. On the first day, mice were trained to associate an auditory tone with a small electric shock. During training trials (total duration 8 min) a 30 s tone was played at the end of which a small shock was administered (2 s, 0.5 mA). The tone shock pairing was repeated during the training trial 3 times. On the subsequent test day, the mouse was exposed to the same chamber for a 4-min trial (no tone, no shock). In this context-retrieval trial time spent immobile was reported as time freezing. One hr later, a different context was presented to conduct the cue retrieval (memory of the tone) trial. Room lighting was changed to ambient red light. The chamber was altered using a white curved plexiglass wall and floor insertion, and a 10% vanilla smell was placed in an inaccessible container. Following a 2 min exploration of the novel context the tone was played for 2 min, and the freezing response was measured. Freezing behavior was monitored automatically by software (VideoFreeze, Med Associates, USA) and time spent immobile was reported as time freezing.

PTZ-induced seizure severity assay

Pentylenetetrazol (PTZ) (Sigma-Aldrich, Cat#P6500-25G) was freshly reconstituted on the day of the assay in normal saline to 5 mg/mL. Mice were weighed immediately before the assay, and the appropriate dose was calculated. Two doses were selected for PTZ and administered via intraperitoneal (i.p.) injection: 35 mg/kg and 50 mg/kg. Behavioral observations and video recordings were started immediately after PTZ administration and continued for 30 min. After the test period, mice were either returned to their home cages or euthanized. Cages were not returned to the housing facility until at least 3 h of monitoring after drug administration and confirmation that no further ictal phases were observed. If status epilepticus was observed (continuous seizure activity longer than 90 s) the assay was terminated and mice were euthanized. Seizure severity was scored according to a modified Racine scale (Table S2). Behavioral recordings were divided into 5-min intervals over the 30-min observation period. Each interval was scored based on the highest seizure activity that was observed. Behavioral scoring was performed by an experimenter who was blinded to the experimental conditions of the mice. For seizure latency measurements, the time of onset of the first observable seizure behavior was recorded.

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistics

All data are expressed as means ± SEM and represent the results of at least three independent biological replicates, as indicated within each Figure Legend and as open circles within bar graphs. Statistical significance was determined using the two-tailed Student’s t test, one-way ANOVA with following post hoc Tukey tests for multiple comparisons, or two-way ANOVA with following post hoc tests for multiple comparisons, as indicated in the Figure Legends. Data analysis and statistics were performed with Microsoft Excel, MATLAB, GraphPad Prism 8.0 and GraphPad Prism 9.0.

Supplementary Material

Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2026.117255.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-HA rabbit	Cell Signaling Technologies	Cat#3724; RRID:AB_1549585
Anti-HA mouse	BioLegend	Cat#901513; RRID:AB_2565335
Anti-Homer1 rabbit	Synaptic Systems	Cat#160003; RRID:AB_887730
Anti-AnkG rabbit	Synaptic Systems	Cat#386003; RRID:AB_2661876
Anti-MAP2 chicken	Encor	Cat#CPCA-MAP2; RRID:AB_2138173
Anti-GAD67 mouse	Sigma-Aldrich	Cat#MAB5406; RRID:AB_2278725
Anti-SHANK2 guinea pig	Synaptic Systems	Cat#162204; RRID:AB_2619861
Anti-Syn1/2 rabbit	Synaptic Systems	Cat#106002; RRID:AB_887804
Anti-Gephyrin mouse	Synaptic Systems	Cat#147111; RRID:AB_2619837
Anti-Synaptotagmin-2 guinea pig	Synaptic Systems	Cat#105225; RRID:AB_2744654
Anti-Vgat guinea pig	Synaptic Systems	Cat#131004; RRID:AB_887873
Anti-Synaptoporin rabbit	Synaptic Systems	Cat#102002; RRID:AB_887841
Anti-Bassoon mouse	AbCam	Cat#AB82958; RRID:AB_1860018
Anti-GNA13 mouse	Proteintech	Cat#67188-1; RRID:AB_2882483
Anti-βactin mouse	Sigma	Cat#A1978; RRID:AB_476692
Anti-NeuN mouse	EMD Millipore	Cat#MAB377; RRID:AB_2298772
Anti-GFAP chicken	Invitrogen	Cat#PA-10004; RRID:AB_1074620
Goat anti-mouse Alexa Fluor 488	ThermoFisher	Cat#A11001; RRID:AB_2534069
Goat anti-mouse Alexa Fluor 546	ThermoFisher	Cat#A11003; RRID:AB_2534071
Goat anti-mouse Alexa Fluor 647	ThermoFisher	Cat#A21236; RRID:AB_2535805
Goat anti-rabbit Alexa Fluor 488	ThermoFisher	Cat#A11034; RRID:AB_2576217
Goat anti-rabbit Alexa Fluor 546	ThermoFisher	Cat#A11010; RRID:AB_2534077
Goat anti-rabbit Alexa Fluor 647	ThermoFisher	Cat#A21245; RRID:AB_2535813
Goat anti-chicken Alexa Fluor 647	ThermoFisher	Cat#A21449; RRID:AB_2535866
Goat anti-guinea pig Alexa Fluor 647	ThermoFisher	Cat#A21450; RRID:AB_2535867
Goat anti-guinea pig Alexa Fluor 555	ThermoFisher	Cat#A21435; RRID:AB_2535856
Phalloidin Alexa 647	ThermoFisher	Cat#A22287
Licor IRDye 800CW donkey anti-mouse	LiCor	Cat#92632212
Licor IRDye 800CW donkey anti-rabbit	LiCor	Cat#92632213
Mouse Adgra1 RNAscope probe	Advanced Cell Diagnostics	Cat#492281
Mouse Parvalbumin RNAscope probe	Advanced Cell Diagnostics	Cat#421931
Mouse Somatostatin RNAscope probe	Advanced Cell Diagnostics	Cat#404631
Mouse Cck RNAscope probe	Advanced Cell Diagnostics	Cat#402271
Mouse Calb2 RNAscope probe	Advanced Cell Diagnostics	Cat#313641
Opal 520	Akoya Biosciences	Cat#OP-001001
Opal 690	Akoya Biosciences	Cat#OP-001006
HRP-conjugated goat anti-mouse secondary antibody	Biotium	Cat#20400-1mL
RNAscope Mutliplex Fluorescent manual assay	Advanced Cell Diagnostics	Cat#323100
RNA-protein Co-detection ancillary kit	Advanced Cell Diagnostics	Cat#323180
Bacterial and virus strains
DH10β	ThermoFisher	Cat#18297010
Stellar cells	Takara	Cat#636767
Chemicals, peptides, and recombinant proteins
Bovine Serum Albumin Fraction V	Roche	Cat#10735086001
Normal Goat Serum	Jackson Immunoresearch	Cat#005000121
Bovine Serum Albumin	Sigma	Cat#A3803
DAPI	Roche	Cat#10236276001
DMEM	Gibco	Cat#11995065
Fetal Bovine Serum	Gibco	Cat#16000044
Hanks’ Balanced Salt Solution	Gibco	Cat#14175095
HEPES	Sigma	Cat#H3375
Matrigel Membrane Matrix	ThermoFisher	Cat#CB-40234
Poly-D-lysine	Gibco	Cat#A38904-01
MEM Non-essential Amino Acid Solution	Sigma	Cat#M7145
Papain	Worthington Biochemical Corporation	Cat#LS003126
Opti-MEM	Gibco	Cat#31985070
Paraformaldehyde	Electron Microscopy Science	Cat#15714
Penicillin/Streptomycin	Corning	Cat#MT30002Cl
B-27 supplement	Gibco	Cat#17504044
Neurobasal A	Gibco	Cat#10888022
GlutaMAX Supplement	Gibco	Cat#35050061
TransIT-2020	Mirus	Cat#MIR5400
Cytosine β-D-arabinofuranoside	Sigma	Cat#C6645
Versene	Gibco	Cat#15040066
FuGENE6	Promega	Cat#E2691
Heparin sodium salt	Sigma	Cat#H3393
OCT embedding compound	Fisher	Cat#23730571
D-AP5	Tocris	Cat#0106
CNQX disodium salt	Tocris	Cat#1045
Tetrodotoxin citrate	Tocris	Cat#1069
Picrotoxin	Tocris	Cat#1128
Adenosine 5′-Triphosphate magnesium	Sigma	Cat#A9187
Guanosine 5′-Triphosphate sodium	Sigma	Cat#G8877
Phosphocreatine disodium hydrate	Sigma	Cat#P7936
Cesium methanesulfonate	Sigma	Cat#C1426
Cesium chloride	Sigma	Cat#289329
Biocytin	Sigma	Cat#B4261
Streptavidin Alexa 647 conjugate	ThermoFisher	Cat#S21374
QX-314 bromide	Tocris	Cat#1014
Spermine	Sigma	Cat#S4264
Superscript IV Reverse Transcriptase	Invitrogen	Cat#18091050
In-Fusion Assembly system	Takara	Cat#638948
PowerUp SYBR Green Master Mix	Applied Biosystems	Cat#A25742
Coelenterazine-400a	NanoLight Technologies	Cat#340
Prolong Gold Antifade Reagent	Invitrogen	Cat#P36930
Experimental models: Cell lines
HEK293T	ATCC	CRL-11268
GKO HEK293	Kind gift from Asuka Inoue, Tokyo University	N/A
Experimental Models: Organisms/Strains
C57/BL6J	Jackson Laboratories	000664
Ssttm2.1(cre)Zjh/J	Jackson Laboratories	013044
B6.129P2-Pvalbtm1(cre)Arbr/J	Jackson Laboratories	017320
B6.129S2-Emx1tm1(cre)Krj/J	Jackson Laboratories	5628
B6.Cg-Gt(ROSA)26Sortm14(CAG-tdTomato)Hze/J	Jackson Laboratories	7914
Adgra1 cKO	Generated at Vanderbilt Genome Editing Facility	046501-UCD
Oligonucleotides
	See Table S1	N/A
Recombinant DNA
TRUPATH kit	Olsen et al.28	Addgene Cat#1000000163
pEB mouse ΔN-ADGRA1-HA	This study	N/A
pEB mouse ADGRA1-HA	This study	N/A
pEB PAR1-LPHN3	Bui et al.34	N/A
Software and algorithms
SnapGene	GSL Biotech	Downloads - SnapGene
NIS-Elements AR 5.42.04	Nikon	Software Resources | NIS-Elements | Software | Microscope Products | Nikon Instruments Inc
ImageJ	National Institutes of Health	Fiji Downloads
Imaris	Oxford Instruments	Microscopy Image Analysis Software - Imaris - Oxford Instruments
Adobe Photoshop	Adobe	Download Creative Cloud desktop app
Adobe Illustrator	Adobe	Download Creative Cloud desktop app
Easy Electrophysiology	Easy Electrophysiology	Download | Easy Electrophys
Graphpad Prism 8.0, 9.0	Graphpad	Prism - GraphPad
MATLAB R2025a	Mathworks	Download and Install MATLAB - MATLAB & Simulink

Highlights.

• Adgra1 is a synaptic orphan GPCR enriched in hippocampal PV+ neurons

• Adgra1 controls the intrinsic excitability of hippocampal PV+ neurons

• Adgra1 deletion impairs PV inhibition onto postsynaptic targets and learning

• Adgra1 activates several G protein pathways, including Gα13

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.