Introduction

Proper control of neuronal activity and synaptic transmission is critical for normal nervous system function, and accordingly, multiple neurotransmitters, neuromodulators, and signaling pathways have been found to participate in their regulation. However, neurons do not respond to these factors in isolation. To produce appropriate responses, neurons need to integrate multiple inputs.

Here we identify an interaction between two pathways known to play important roles in regulating neuronal and synaptic activity: (1) group 1 metabotropic glutamate receptors (mGluRs) and (2) the nuclear hormone receptor family member, retinoid X receptor gamma (RXRγ). Specifically, we found that animals lacking RXRγ exhibited impairments in group 1 mGluR-mediated synaptic transmission and synaptic plasticity, together with impairments in group 1 mGluR-mediated behaviors. These data suggest an interaction between these pathways in normal learning and memory and in the control of disease-relevant behaviors including Parkinson’s disease-related motor impairments, and Schizophrenia-related executive function.

RXRγ is one of three retinoid X receptors (α, β, γ) found in mice and humans with partially overlap** expression patterns and functions. RXRs are members of the nuclear receptor superfamily and form heterodimers with other nuclear receptor family members that bind to discrete DNA sequences to regulate transcription in response to ligand binding (reviewed inCalcium imaging

Calcium imaging was performed on proximal apical dendrites of CA1 pyramidal cells in acute hippocampal slice preparations as described previously96. A customized two-photon laser-scanning Olympus BX61WI microscope with a 60×/1.1 nA objective was used to detect Ca2+ signals. A Mai/Tai laser (Solid-State Laser, Mountain View, CA) tuned to 820 nm was used for excitation, and image acquisition and storage was controlled by Olympus Fluoview FV300 software (Olympus, Melville, NY). In the transfluorescence pathway, a 565 nm dichroic mirror was used to separate green and red fluorescence, passed through HQ525/50 and HQ605/50 emission filters, to eliminate transmitted or reflected excitation light (Chroma Technology, Rockingham, VT), and detected simultaneously by two photomultiplier tubes. Neurons were loaded with dyes through the patch pipette for 20 min before commencing image acquisition. Alexa Fluor-594 was used to outline neuronal dendritic structure, and Calcium Green-1 (100 µM in the patch pipette) to detect [Ca2+] changes. To measure [Ca2+] dynamics, fluorescence was collected by scanning at 4–5 Hz in a surface-scanning mode (XYT), or 1 kHz in XT mode, and averaged from specified structures to obtain F(t). Baseline fluorescence (F0) was calculated as the average of four images obtained immediately prior to bath application of 30 μM DHPG. Images of the same cells were obtained after 5, 10, 15 and 20 min of DHPG application. Normalized values for each cell at each time point were calculated by dividing the fluorescent signals at a given time point by the average fluorescent signal for that cell obtained prior to drug treatment (F0). No EGTA was added to the internal solution for [Ca2+] imaging.

Western blots

Hippocampi were homogenized by sonication in 50 mM Tris pH 7.4/2% SDS followed by incubation at 95 °C for 5 min. Extracts were run in duplicate and resolved by denaturing PAGE and blotted to nylon membrane (Immobilon-FL, Millipore). Blocking and antibody incubations were for 1 h at room temperature in Tris-buffered saline, 0.05% tween 20, 3% non-fat dry milk. Washes were carried out TBS-Tween without milk. Proteins were detected using commercially available antibodies against mGluR1 (Millipore AB1551), mGluR5 (Millipore AB5675), and α-tubulin (Sigma T9026) in conjunction with dye conjugated anti-mouse or rabbit secondary antibodies (Li-COR, Lincoln, NE). Blots were scanned and integrated band intensities determined using an Odyssey infrared imaging system. mGluR1 or mGluR5 band intensities were normalized to the band intensity for tubulin in that lane as a loading control and data were presented as percent of the average normalized control value.

Q-PCR

Total hippocampal RNA was prepared using Trizol Reagent and first strand cDNAs were prepared using the Superscript III first strand cDNA synthesis supermix for Q-PCR (Invitrogen) as per manufacturers instructions. Quantitect primer pairs were purchased from Qiagen. All primer pairs showed a linear response of Log(template dilution) vs. cycle threshold for template dilutions over 5 orders of magnitude. Optimal annealing temperatures, template dilutions and corresponding efficiencies were determined experimentally for each primer pair, and melting curves determined for each reaction showed a single peak under these conditions. Relative expression levels were calculated for each sample using the experimentally determined efficiency and the number of cycles to reach a florescence threshold set manually within the linear portion of a plot of the Log(florescence intensity) vs. cycle number. Values were normalized to those obtained using a primer pair specific for GAPDH run in parallel for each sample.

Immunohistochemistry

Brains were fixed by transcardial perfusion in 4%PFA dissolved in 100 mM phosphate buffer pH 7.4. 30 μM sections were cut using a Vibratome slicer. Floating sections were permeablized for 30 min at 4 °C in Tris-buffered saline + 0.2% triton and blocked for 1 h at room temp in Tris-buffered saline + 10% fetal bovine serum. Primary antibody incubations were carried out overnight at 4 °C in blocking buffer, secondary antibody incubations were carried out for 1 h at room temperature also in blocking buffer, and washes were TBS + 0.2% triton. Primary antibodies were from Millipore as above, and an Alexa-568 anti-rabbit from Invitrogen was used as a secondary. Stained sections were mounted in Fluorsave aqueous mounting medium (Calbiochem) and images were obtained by confocal microscopy (Olympus Fluoview FV1000 and associated software).

Open field

Behavior in a novel open field was assessed using Plexiglas activity chambers (model ENV- 520; Med Associates, St. Albans, Vermont) (43.2 cm long × 43.2 cm wide × 30.5 cm high). Mice were placed in the open field and activity was recorded for 60 min. Behavioral measures were calculated using the Activity Monitor program (Med Associates) and “center” was defined as an area beginning 10 cm from the walls.

Elevated plus maze

Animals were placed on an elevated plus maze apparatus for 6 min and their position was tracked using an overhead digital video camera and Anymaze software (Stoelting, Wood Dale, IL). The apparatus consisted of a plus shaped track with arms 18 cm long and 5 cm wide. Two arms were enclosed on 3 sides and the entire apparatus was elevated 50 cm above the bench top and surrounded by a white curtain under ambient room lighting.

Accelerating rotarod

Mice were tested on an accelerating rotarod apparatus (Letica LE8200) over a period of 5 days. The first day consisted of 3 pre-training trials in which animals were habituated to the stationary apparatus 2 times for 3 min each, and then for an additional 3 min at 4 rpm. Animals that fell from the apparatus during these trials were returned to the rod until 3 min had elapsed. On subsequent days, animals were given 3 × 5 min trials per day with the rotation speed ramped from 4 to 20 rpm over the course of the trial. Animals that fell from the rod during these trials were returned to the rod once to discourage them from learning to fall as an escape strategy, however only the latency to the first fall was used for analysis. For all trails on all days, the intertrial interval was approximately 30 min during which animals were returned to their home cages.

Fear conditioning

All fear conditioning trials were carried out in conditioning chambers placed in sound attenuating chambers (Med Associates, St Albans, VT) and freezing during each trial was monitored continuously using a video tracking and analysis system (Freezeframe, Actimetrix Software, Wilmette, IL). On day one animals were given a 3 min training trial consisting of 2 min of pre-exposure followed by a 2 s 0.6 mA continuous foot shock and 58 s of post-shock exposure. On subsequent days animals were reintroduced to the conditioning chamber in the absence of shock for 10 min. Percent of time spent freezing during the first 2 min of these trails was used to monitor extinction of the freezing response.

Morris water maze

The experiment was carried out in a circular pool 120 cm in diameter filled with water made opaque with white paint. A video tracking system (HVS Image VP-118) was used to record and analyze each animal’s behavior. A 10 cm square platform submerged 1 cm below the water surface was used as an escape platform and in the visible version of the task, a colored syringe barrel was affixed to the center to mark its location. For all phases of the experiment, 4 trials of no more than 120 s were given each day at 20 min inter-trial intervals. Animals that exceeded the 120 s limit were guided to the platform before being returned to their home cage. A single “sha**” trial in which animals were placed on the platform for 15 s, was performed before training on the first day of the visible, hidden and reversal phases of the task. Probe trials of 60 s, during which the platform was removed from the water maze, were performed at the end of training on days H5 and R5. In the visible platform phase of the task, the pool was surrounded by a curtain and the marked platform was placed in a different location on all four trials of a day. In the hidden and reversal phases, the curtain was removed to reveal the room cues and the hidden platform was placed in a fixed location for each phase.

Spatial and non-spatial object recognition

For all object recognition experiments, the apparatus consisted of an opaque plexiglass arena 71 cm square and 36 cm high. The arena was surrounded by a white curtain, and two prominent visual cues were affixed to two of the arena walls. Objects were glass, metal or ceramic items fixed in place with double stick tape (eg. padlock, glass chess piece, metal bottle cap, ceramic drawer knob). All objects, the arena and floor were cleaned with non-alcoholic antiseptic wipes between each trial. To habituate the animals to the testing environment, animals were handled in the testing room on the three days prior to testing. Animal movements were tracked and analyzed using Noldus Ethovision XT software and an overhead-mounted video camera. For the novel object version of the task animals were placed in the empty arena for 5 min first. Then three 5 min training trails were conducted in which three different objects were placed equidistantly in a diagonal line across the arena. Finally a 5 min. testing trial was conducted in which one of two peripheral objects was replaced with a new object. The delay between the empty arena trial and all of the training trials were 10 min, and the delay between the final training trial and the testing trial was 3 h. The spatial version of this task was carried out similarly except that instead of replacing one of the peripheral objects in the testing trail, it was moved to one of the two empty corners so that the three objects were now arranged in a “V”.

Prepulse inhibition

Testing was performed essentially as described in97 and98. Animals were placed in startle chambers (SR-Lab startle response system, San Diego Instruments, San Diego, CA) containing a high-frequency speaker for generation of acoustic stimuli and a piezoelectric accelerometer to measure animal movements in response to stimulus presentations. A continuous level of 65 dB background white noise was presented through the experiment. The testing protocol consisted of 6 presentations of a 40 ms broadband 120 dB startle tone, followed by 10 pseudo-randomly interleaved presentations of background alone, prepulse trials consisting of stimuli at 3, 6 or 12 dB above background delivered 100 ms before the 40 ms broadband 120 dB startle tone, and non-prepulse trials in which the 120 dB startle tone was presented alone. This protocol was followed by another 6 presentations of the 120 dB startle tone alone. The interval between trials varied pseudorandomly between 15 and 30 s. PPI for each animal at a particular prepulse interval was calculated as the difference between the average startle magnitude on interspersed startle tone-alone trials and prepulse trials divided by the average startle magnitude on the same startle tone-alone trials.

Y-maze spontaneous alternation

Mice were tested individually for spontaneous alternation during a single 6 min trial in a custom fabricated Y-maze apparatus. The maze consisted of three arms 40 cm in length, 17 cm high and 4 cm wide, separated by 120 degrees located on a table in the center of the testing room under dim lighting. Animals were placed at the end of one of the arms in a psedorandomly assigned manner. Movements were recorded by an overhead mounted video camera and monitored by an observer blinded to genotype. Arm entries were scored when all four of the animal’s paws entered an arm. The percent alternation was calculated as the number of triads containing entries into all three arms divided by the total number of triads (the total number of arms entered minus 2) × 100. Average number entries for all groups was 24.2 with a standard deviation of 5.5. Three RXRγ knockout and 2 wild-type mice with 13 or fewer arm entries were excluded from the analysis.

Behavioral sensitization to cocaine

Testing was conducted in the same apparatus used for open field behavior testing. Immediately before each 30 min trial, animals were given a single intraperitoneal injection of either 15 mg/kg cocaine or saline only. The testing protocol consisted of 3 days of habituation (1–3) in which all animals were administered saline only. Habituation was followed by 5 sensitization trials, conducted on days 4–8 and day 20, in which either cocaine or saline was administered as indicated. On day 22 all animals were tested again after receiving saline injections.

Statistical analysis

Tests for statistical significance between groups were performed using Prism 7 (Graphpad Software, San Diego, CA, USA). Student’s unpaired, two-tailed t-tests were used for experiments involving 2 groups, and paired t-test for experiments in a single group comparing before to after drug application. 2-way ANOVA comparisons—with or without repeated measures as dictated by the experimental design—were used for analysis of all experiments involving multiple groups. Post-hoc comparisons for multiple group experiments were performed using Dunnett’s tests when comparisons were made to a single control group, and Tukey’s tests when comparisons were made among multiple groups. The results of these tests together with their associated P values are listed in the legends for their corresponding figure.