Abstract
Methamphetamine (METH) is a widely abused psychoactive drug that readily establishes reward memories contributing to METH relapse. The medial prefrontal cortex (mPFC) is central to cognition, motivation, reward, and emotion, and the hippocampus is critically involved in reward memory. The mPFC possesses an enormous variety of projection neurons. However, the direct projection from the mPFC to the hippocampus involved in METH addiction has not been studied well. To explore the role of a mPFC-hippocampus pathway of regulating METH reward memory, conditioned place preference (CPP) was used to detect reward memory and recombinant adeno-associated virus 2/9s (rAAV2/9s) were used to label neurons, identify projections, and optogenetically explore involvement of the male mice mPFC-hippocampus pathway in regulating METH-CPP. We found that a novel prelimbic prefrontal cortex (PrL) projection directly to the dorsal hippocampus CA1 (dCA1) regulated CPP induced by METH. Moreover, optogenetic activation or inhibition and silencing the PrL to dCA1 glutamatergic pathway with tetanus neurotoxin (TeNT) modulated METH-CPP. Our results reveal a PrL to dCA1 glutamatergic pathway that regulates METH-CPP and could serve as a potential target for treating METH use disorder.
Significance Statement
This study elucidated the intricate molecular, circuit, and functional architecture of the prelimbic prefrontal cortex to the dorsal hippocampus CA 1 and identify CaMKIIα-expressing glutamatergic neurons in the medial prefrontal cortex as a key inversely driver of reward and methamphetamine addiction. These findings open new avenues for exploring how the prefrontal cortex to the hippocampus regulates reward and addiction.
Introduction
Methamphetamine (METH) is an illicit widely abused psychoactive drug that has a major impact on cortical and striatolimbic brain regions regulating reward behaviors, including the frontal cortex, the striatum, and the hippocampus (Volkow and Morales, 2015). The medial prefrontal cortex (mPFC) is important for motivated behaviors (Euston et al., 2012) including processing the value of a reward signal (Otis et al., 2017; Lak et al., 2020) and reward seeking (Moorman and Aston-Jones, 2015; Kim et al., 2017). The mPFC dysfunction is implicated in the loss of control over drug use (Goldstein and Volkow, 2011; Parsegian et al., 2011) and is consistently implicated in substance use disorder (SUD; Tzschentke and Schmidt, 2003; Kalivas et al., 2005). More broadly, the mPFC acts as a driver of goal-directed behavior (Szczepanski and Knight, 2014) and is crucial not only for reward processing but also attention and memory (Riga et al., 2014). Glutamatergic projections from pyramidal neurons in the mPFC to striatolimbic brain nuclei are involved in drug reward, including projections to the hippocampus (Anastasiades and Carter, 2021). The hippocampus is a limbic structure that is deeply involved in episodic memory and plays a pivotal role in spatial navigation (Takahashi et al., 2024) based on features of the environment, including sensory cues and the location of rewards (Robinson et al., 2020). A small population of hippocampus CA1 and subiculum neurons encode for reward location (Gauthier and Tank, 2018), and the dorsal hippocampus CA1 (dCA1) is critical for reward and memory (Hyman et al., 2006; Wirtshafter and Wilson, 2020). Accordingly, the hippocampus is implicated in the reinstatement of psychostimulant self-administration and the rewarding properties of METH (Sun and Giocomo, 2022). Moreover, METH increases extracellular glutamate in the hippocampus (Rocher and Gardier, 2001) that can activate ionotropic receptors, such as N-methyl-D-aspartate receptor (NMDA) and α-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid receptor (AMPA) and both receptors are involved in METH addiction (Simões et al., 2008). Given the importance of both mPFC and dCA1 in reward seeking, the role of this direct projection in regulating METH reward was examined in the present study using METH-induced conditioned place preference (CPP).
Materials and Methods
Animals
Male C57BL/6 mice (8 weeks old and weighing 18.0–25.0 g) were purchased from Chongqing Medical University. All mice were housed in cages, four per cage or one individually housed during photostimulation behavioral assays under a 12 h light/dark cycle with food and water ad libitum in a 21–23°C, humidity-controlled room. All experimental procedures were performed between 7:00 A.M. and 7:00 P.M. All experimental procedures were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 80-23, revised 1978). All experiments were approved by the Institutional Animal Care and Use Committee of Kunming University of Science and Technology and the Committee on Animal Use and Protection of Yunnan province (No. LA2008305). All efforts were made to minimize the number of animals used and their suffering. Male mice were used in the current study if not indicated otherwise.
Viral vectors
The rAAV2/9-hSyn-mCherry-WPREs-hGH pA, rAAV2/9-hSyn-hChR2 (H134R)-p2A-mCherry-WPREs-hGH pA used for optogenetic experiments, and rAAV2/9-hSyn-eNpHR3.0-p2A-mCherry-WPREs-hGH pA used for neuronal tracing and cell labelling experiments were from Hanbio. The rAAV2/9-CaMKⅡα-DIO-mCherry-WPREs-hGH pA (6.36 × 1012 vg/ml), rAAV2/9-CaMKⅡα-DIO-hChR2 (H134R)-p2A-mCherry-WPREs-hGH pA, rAAV2/9-CaMKⅡα-DIO-eNpHR3.0-p2A-mCherry-WPREs-hGH pA, rAAV2/9-CaMKⅡα-DIO-TeNT-p2A-mCherry-WPREs pA, and rAAV2/Retro-hSyn-Cre-EYFP-WPRE-hGH pA were purchased from BrainVTA. The titer of vectors range was 2.00–5.00 × 1012 vg/ml.
Drugs and reagents
METH dissolved in saline was intraperitoneally injected at a dose of 2.5 mg/kg, which was same as our previous study (Huang et al., 2018). METH was provided by the Public Security Bureau of Yunnan Province, Kunming, China and purity validated using a gas chromatography-mass spectrometer. In this study, the information on primary antibody used was as follows: VGLUT1 antibody (Rabbit, 1:500, Abcam, ab227805), GluN2B antibody (Rabbit, 1:1,000, Abcam, ab65783), GluR1 antibody (Rabbit, 1:2,000, Abcam, ab109450), Beta actin (β-actin) antibody (Mouse, 1:10,000, Proteintech, HRP-66009) with appropriate secondary antibodies Goat Anti-Rabbit IgG H&L (HRP; Rabbit, 1:10,000, Abcam, ab205718), and Goat Anti-Mouse IgG H&L (HRP; Mouse, 1:10,000, Abcam, ab205719) and Immobilon western chemilum HRP substrate (1:1, Millipore, WBKLS0500) for Western blotting analysis (WB). c-Fos antibody (Rabbit, 1:200, Abcam, ab222699), VGLUT1 antibody (Rabbit, 1:200, Abcam, ab227805), Anti-mCherry antibody (Rabbit, 1:500, Abcam, ab167453), DAT antibody (Rabbit, 1:1,000, Proteintech, 22524-1-AP) with appropriate secondary antibodies Goat carrying Alexa Fluor 488 (goat anti-Rabbit, 1:500, Abcam, ab150077), Alexa Fluor 647 (goat anti-Rabbit, 1:1000, Abcam, ab150079) for immunofluorescence staining.
Stereotaxic surgery
Mice were anesthetized with intraperitoneal injection of pentobarbital (80 mg/kg) and then fixed into a stereotaxic frame. Standard surgery was performed to expose the brain surface above the PrL and dCA1. The head hair was shaved, the skin was sterilized using iodine solution, and the eyes were smeared with erythromycin eye cream.
For viral injection, rAAVs were stereotaxically delivered to the following brain regions (coordinates relative to bregma site in parentheses). Coordinates used for site injection were PrL: anterior-posterior (AP): +2.68 mm, medial-lateral (ML): ±0.40 mm, dorsal-ventral (DV): −1.50 mm and dCA1: AP: −2.30 mm; ML: ±1.80 mm; DV: −1.25 mm according to The Mouse Brain in Stereotaxic Coordinates (5th Edition) by George Paxinos and Keith B.J. Franklin. The rAAVs were bilaterally injected with a glass pipette (10 μl syringe and a 36 gauge blunt needle) connected to Nanoliter Injector (RWD Life Science) at a slow flow rate of 50 nl/min to avoid potential damage of local brain tissue, a total unilateral volume of 200 nl. The pipette was remained in the brain for at least 10 min after microinjection to avoid backflow and was pulled out slowly. The incision was stitched by using a surgical suture.
One week following microinjection of rAAVs, bilateral optical fibers (200 µm diameter, 0.37 NA) were stereotaxically implanted 0.1 mm above the dCA1, designed position. Microscrews were secured adjacent to the fibers, and dental cement was applied to fix both fibers and screws. Protective sleeves were subsequently installed to protect optical fibers until the behavioral experiments began. Mice were put on a warm pad until fully recovered from anesthesia and given iodophor disinfection to the surgical wound. Benzylpenicillin sodium (50,000 U/kg, HAPHARM GROUP) was given using intramuscular injection per day for 3 d. Following surgery, the mice were allowed to recover and housed at least 1 week before behavioral experiments and histological analyses. Behavioral testing and histological analyses were conducted at least 3 weeks after AAV injection.
Conditioned place preference
After viral vectors injection or fiber implantation, the mice were individually housed for at least 2 weeks before the behavioral tests. The conditioned place preference behavioral (CPP test) apparatus consisted of two compartments (15 cm × 15 cm × 30 cm) with differences in visual (black or white wall) and tactile (smooth or rough floor) cues, divided by a sliding door. Mice were put into the CPP apparatus and the pre-CPP score was detected during the behavioral tests of pretest. METH-CPP protocol was applied as our previous study, including habituation, pretest, training, and posttest sessions (Huang et al., 2018). During the habituation phase on Days 1 and 2, mice were released from the middle of the conditioning apparatus and allowed to freely explore the full extent of the CPP apparatus for 20 min per day. During pretest (20 min) on Day 3, the initial place preference was determined by recording the time that the mice spent in each compartment. During training on Days 4, 6, 8, and 10, mice were treated with METH (2.5 mg/kg, i.p.) or saline for 20 min. On Days 5, 7, 9, and 11, mice were given saline and confined to the opposite conditioning compartment for 20 min. Posttest on Day 12, mice were put into the conditioning apparatus to freely move for 20 min without drug treatment. Time that the mice spent in each compartment was quantified to determine preference. The CPP score refers to the difference between the length of time spent by each mouse in the drug-paired compartment during the pretest and posttest phases. During posttest phase, mice locomotion was also measured by mice moving distance, and results showed that there is no difference in locomotion between saline and METH groups. The first pretest assessed baseline place preference, followed by AAV injection on 4 weeks; mice that exhibited a preference exceeding preference which spent ≥780 or ≤420 s of the total time (1,200 s) in one side or had too few crossings (≤20 times) were excluded to rule out congenital behavioral abnormalities before the operation. The second pretest (pretest on Day 3) is to evaluate the effect of surgery on place preference and served as the control measurement prior to METH training. Thus, this second test corresponds to the subsequent conditioning days. CPP design is biased based on the time when mice stay in two chambers during pretest; the less side is paired with stimulation (drug or laser). The connecting lines in the behavioral data plots (Fig. 2I) represent the pre- and posttest CPP changes for each one mouse. Before behavioral measurements, mice were randomly assigned to either the experimental or the vehicle group.
In vivo photostimulation
For photostimulation during behavioral assays, a 473 nm (blue light) or 595 nm laser (yellow light) intelligent wired photogenetic system (inper) was connected to a patch cord with a pair of FC/PC connectors on each end. This patch cord was connected through a fiber-optic rotary joint, which allows free rotation of the fiber for mice to move freely (Newdoon Technology). The ceramic fiber-optic cannulas implanted in the mouse (200 μm in diameter, 0.37 NA; Zhou et al., 2019) were connected to the optic patch cord using ceramic mating sleeves (Newdoon Technology). Blue light (473 nm, 3 mW) was delivered with 15 ms light pulses at 20 Hz for 20 min, and yellow light was delivered at constant 5 mW for 20 min during the posttest in METH matching side. The laser power was measured at the tip of the optic fiber using an Optical Power Monitor (Thorlabs).
Immunofluorescence staining
Mice were anesthetized and transcardially perfused with 0.9% saline for 7 min followed by 4% paraformaldehyde (PFA) dissolved in 0.1 M phosphate-buffered saline (PBS) buffer, pH 7.5, for 5 min. Brains were postfixed in 4% PFA at 4°C for 24 h and then transferred to 15% for 1 d and then 30% sucrose solution for 2 d. Then brains tissues were coated with embedding reagent, Tissue-Tek O.C.T. Compound (SAKURA), a formulation of clear, water-soluble glycols and resins and then stored at −20°C. Slices were frozen sectioned into 20 μm thick.
For immunofluorescence staining, each slice was washed three times with 0.1 M PBS and blocked using 5% goat serum in PBST (containing 0.1% Tween 20) for 1 h followed by incubation with primary antibody at 4°C overnight. After rinsing with PBS, the slices were incubated with fluorescence-conjugated secondary antibody at room temperature for 2 h. Finally, the slices were coverslipped on antifluorescence attenuation tablets including DAPI (4′,6-diamidino-2-phenylindole; Solarbio). All analyses were performed blinded to the experimental conditions. Images were captured using a Nikon AX fluorescence confocal microscope.
For c-Fos immunohistochemistry
On Day 12, after the post-CPP test was performed, mice were placed in the home cage for 90 min. The mice were anesthetized with 1.2% pentobarbital sodium reagent and treated with 0.9% saline and 4% PFA reagent for cardiac perfusion. For the expression of c-Fos, the whole brain was taken and fixed with 4% PFA for 24 h. After fixing, tissues were dehydrated in ethanol solution of different concentrations for 2 h, mixed with paraffin wax, and stored overnight. The coronal surface of the brain tissue was sliced (7 μm), and the slices were spread, patch and baked (60°C, 2 h) until the wax melted, and then were dewaxed with xylene. The slices were dehydrated in different concentrations of ethanol for 5 min, were soaked in distilled water for 5 min, and rinsed with TPBS. During antigen repair, sections were placed in 0.01 M citric acid buffer, pH 6.0 (95°C, 15 min) and rinsed with TPBS. Then the slices were sealed with 5% bovine serum (37°C, 30 min). The slices were incubated with a standard c-Fos primary antibody (1:1,000, 4°C, overnight) and followed by an HRP-conjugated goat anti-Rabbit IgG secondary antibody (1:2,000, 37°C, 30 min) labeled with horseradish peroxidase biotin. Last color developing agent (DAB) was added for color development (lasting 10 s). Microscopic observation was in bright field.
Western blotting analysis
The mice were sacrificed with cervical dislocation immediately after the behavioral test was completed. The brain tissue was rapidly extracted by combining maps of the mouse brain [The Mouse Brain in Stereotaxic Coordinates and the Allen Brain Reference Atlases (Adult Mouse data)]. The mPFC and dCA1 were dissected bilaterally on an ice-cold plate and stored at −80°C immediately. Protein lysates were prepared using a solubilizing solution [20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 1 mM ethylene diamine tetra acetic acid (EDTA), 1 mM phenylmethanesulfonyl fluoride, 1 mM ethylene glycol bis (2-aminoethylether)-N,N′-tetraacetic acid (EGTA), 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM Na3VO4, 1 mM β-glycerol phosphate, and 1 mg/ml leupeptin]. Protein concentration was determined using a BCA protein assay kit reagent (Beyotime). Samples were separated on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) for β-actin, VGLUT1, GluN2B, and GluR1 and transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore). The PVDF membrane was rinsed with PBS, and nonspecific sites were blocked by incubating the membrane with blocking buffer (TPBS containing 10% nonfat milk) for 1 d at 4°C and then incubated with primary antibodies overnight at 4°C followed by appropriate secondary antibodies goat anti-Rabbit IgG H&L (HRP; Abcam) for 1 h at room temperature. Rinsing was performed in each session using PBS for three times for 5 min. The epitope was visualized by an ECL Western blot detection kit (Millipore).
Electrophysiological recordings
Four weeks after virus infusion, the experimental mice were anesthetized and perfused with oxygen-saturated (95% O2 and 5% CO2) ice-cold NMDG-based artificial cerebrospinal fluid (ACSF), which contained (in mM): 96 NMDG, 2.5 KCl, 1.25 NaH2PO4, 25 NaHCO3, 20 HEPES, 25 glucose, 2 thiourea, 5 Na-ascorbate, 3 Na-pyruvate, 3 myo-inositol, 0.01 taurine, 12 N-acetyl-ʟ-cysteine, 0.5 CaCl2·4H2O, and 10 MgSO4·7H2O (pH 7.5). The brain was quickly dissected and placed in oxygen-saturated cold ACSF and then sliced coronally (250 μm) with using a vibrating blade microtome (Leica VT1200s). Brain slices containing mPFC and dCA1 were incubated in a holding chamber filled with high sucrose concentration ACSF at 32°C for 30 min. The brain slices were then transferred to standard ACSF (in mM; 130 NaCl, 2.5 KCl, 1.25 NaH2PO2, 2 MgCl2, 2 CaCl2, 26 NaHCO2, and 12.5 glucose) and incubated for at least 1 h at room temperature.
For the cell recording, the brain slices were transferred to the recording chamber, which was continuously perfused with fully oxygenated standard ACSF at a rate of 2 ml/min. Whole-cell patch-clamp recordings were using a MultiClamp 700B amplifier (Molecular Devices) and a 1440A digital-to-analog converter (Molecular Devices). The patch electrodes (4–5 MΩ resistance) used for recording were pulled by a P-97 puller (Sutter Instrument) and filled with an internal solution containing (in mM): 10 HEPES, 0.2 EGTA, 4 ATP-Mg, 0.3 Na-GTP, 4 KCl, 5 Na2-phosphocreatine, 13.43 biocytin, pH 7.3; osmolality 290–300 mOsm/kg).
To examine the efficiency of ChR2 expressed in the mPFC CaMKIIα neurons, the neurons were recorded in current-clamp mode and were given blue light pulses at frequencies of 5, 10, and 20 Hz, respectively (3–5 mW, pulse width 2 ms) of each sweep. To record postsynaptic currents of the mPFC excitatory neurons projecting to the dCA1, the same blue light pulse (3–5 mW, pulse width 2 ms) was delivered to evoke presynaptic glutamate release in the dCA1. The dCA1 neurons were held at −70 mV in voltage clamping mode, and neurons were recorded for 10–30 sweeps. Tetrodotoxin (TTX, sodium channel blocker, 1 μM, MedChemExpress) and 4-aminopyridine (4-AP, voltage-gated potassium channel blocker, 100 μM, MedChemExpress) were used to validate that the mPFC-dCA1 pathway is a monosynaptic connection, and 6,7-quinoxalinecarbonitrile dinitroquinoxaline-2,3-dione (CNQX, AMPA receptor antagonist, 10 µM, MedChemExpress) and D-norvaline, 5-phosphono (D-AP5, NMDA receptor antagonist, 50 µM, MedChemExpress) were used to prove that the mPFC-dCA1 pathway was an excitatory connection.
Imaging
Microscopic observation by 20× objectives in bright-field observation was used for c-Fos immunohistochemistry. Confocal fluorescence images were acquired using a Nikon AX confocal laser scanning microscope with 10×, 20×, and 60× oil objectives for imaging immunofluorescence-stained neurons and NIS-Elements Viewer image analysis software (Nikon). Neuronal fluorescence was imaged using a Nikon TE2000-S fluorescence microscope equipped with 10× and 20× objectives. Coronal scans of the whole brain were performed using an OLYMPUS VS200 scanning microscope with 10× and 20× objectives for imaging fluorescent neurons and analyzed by OlyVIA Viewer image analysis software (OLYMPUS).
Statistical analysis
Band density analysis was performed using the ImageJ software. Statistical analysis was performed using GraphPad Prism 9.0 software. One-way or two-way ANOVAs were followed by a Bonferroni post hoc multiple-comparison test to compare saline and METH or vehicle and treated groups. The significance threshold was held at p = 0.05, p < 0.05 was considered to indicate a statistically significant difference, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. ns, not significant. Data were presented as mean ± SEM.
Results
Direct mPFC projection to the dCA1
To investigate projections from the prelimbic prefrontal cortex (PrL) to other brain regions, recombinant adeno-associated virus 2/9 (rAAV2/9)-human Synapsin I (hSyn)-mCherry, an anterograde non-transsynaptic tracer (Sun et al., 2023) were stereotaxically injected (Fig. S1A) into the mPFC in mice (Fig. S1B). Mice were returned to the home cage for 2 weeks to allow for maximum virus expression (Fig. S1C).
We found that the fluorescent protein (mCherry, red) was expressed in PrL neurons (Fig. S1D) and in axon terminals in the dCA1 (Fig. S1E; also verified by anti- mCherry in Fig. S1F–H). mCherry was also seen in more established PrL axon terminal fields, such as olfactory bulb (onl), striatum (STRd), and others (Fig. S2). These anterograde tracing data clearly identify a projection from the PrL to the dCA1 in mice.
The PrL to the dCA1 top-down regulates CPP
Activity of dCA1 is necessary for drug-induced CPP (Hyman et al., 2006), a common and reliable behavioral assay to assess the rewarding effects of drugs. Since we found that the dCA1 received input from the PrL, we examined place preference after selectively activating or inhibiting the PrL to dCA1 pathway with optogenetic stimulation of channelrhodopsin-2 (ChR2), a blue light-sensitive (473 nm) ion channel fluorescent protein (ChR2-mCherry), or the silencing opsin halorhodopsin (eNpHR3.0-mCherry), a yellow light-sensitive (595 nm) ion channel. rAAV2/9-hSyn-hChR2 (H134R)-mCherry or rAAV2/9-hSyn-eNpHR3.0-mCherry (Fig. 1A) were microinjected into the PrL, and optical fibers were implanted into the dCA1 (Fig. 1B,C). Mice were trained for CPP by applying optogenetic stimulation as shown in Figure 1D. Expression of mCherry in the PrL and in the dCA1 was shown in Figure 1E,F and the location of optical fiber implantation in the dCA1 was shown in Figure 1G. After surgery pretest, habituation, and CPP pretest sessions, mice were given 20 min daily CPP training session 2 weeks after surgery and rAAV2/9s administration. Real-time place preference (RTPP) tracking was used to detect mice place preference (Fig. 1H). On the first 3 d, mice freely moved to either side. Optical stimulation was applied when mice were placed into one side of the CPP apparatus (light-paired), whereas the light was turned off when mice were placed in the other side every other day for 8 d. During the posttest session mice had ad libitum access to both sides of the chamber, training with optogenetic activation of the PrL to dCA1 pathway prevented the development of CPP (preference) for the chamber paired with light stimulation and optogenetic inhibition increased the time spent in the chamber paired with light (ChR2 = −25.79 ± 43.66 s; eNpHR3.0 = 88.30 ± 36.62 s; Vehicle = −121.79 ± 30.02 s, respectively; Fig. 1I). On the test day 90 min after optical stimulation (473 nm, 3.0 mW), mice were sacrificed, and tissues were used to detect the expression of c-Fos, a marker of neuronal activation (Joo et al., 2016). c-Fos expression was quantified in the PrL and the dCA1 of mice (Fig. 1J–O). Change in c-Fos protein expression occurred in both the PrL (83.33% positive overlap) and the dCA1 (38.18% positive overlap) following optical stimulation (Fig. 1P). These results demonstrated that the PrL to the dCA1 pathway was activated after ChR2 optical stimulation and that the pathway negatively regulated place preference, inversely enhanced place preference after eNpHR3.0 optical stimulation.
The PrL to the dCA1 top-down regulates CPP. A, Schematic diagram of the major components of anterograde tracing rAAVs. B, C, Diagram of the anterograde virus strategy with injection site into the PrL and optical fiber implantation site in the dCA1 using 473 nm, blue light (B), or 595 nm, yellow light, respectively (C). D, Schematic of the experimental timeline of the optogenetic experiment to stimulate the PrL to the dCA1 pathway detection of c-Fos expression following virus expression and CPP-like behavioral experiments trained in combined with optogenetic stimulation on the 12th day. Laser stimulation (blue) or inhibition (yellow) was on the same side of chamber based on its preference in pretest. On Days 5, 7, 9 and 11, the arrow indicating “no” opto stimulation is that the mice were fitted with optical fiber sleeves but were not given light stimulation. E, Representative images showing the expression of the ChR2-mCherry fluorescent expressions in the coronal PrL. The red solid wire frame represents a magnified view of a specific area (left), the white dashed box represents the peripheral brain region structures of PrL (middle), and the white arrow indicates the neuronal cell body merging ChR2-mCherry with DAPI in the PrL (right). Scale bars: 100 μm, 50 μm, 20 μm. F, Representative image of the ChR2-mCherry fluorescent from PrL neurons to target synapse in the dCA1. The red solid wireframe represents a magnified view of a specific area (left), the white dashed box represents the peripheral brain region structures of dCA1 (middle), and the white arrow indicates the terminals from the mPFC neurons (right). Scale bars: 100 μm, 50 μm, 20 μm. G, Diagram of optic fiber embedded into dCA1. Scale bar, 200 μm. H, Behavioral trajectory graphs during administration of light stimuli. I, Quantification of CPP behavioral trajectories during given light stimuli [Vehicle, n = 5 mice, ChR2, n = 5 mice, eNpHR3.0, n = 4 mice; two-way ANOVA, interaction, F(2,24) = 6.549, p = 0.0054; CPP (pre-, post-no, post-with opto), F(2,24) = 3.838, p = 0.0358, groups (Vehicle, ChR2, eNpHR3.0), F(1,24) = 9.200, p = 0.0057, Bonferroni's multiple-comparison test]. J, c-Fos expression in the PrL 90 min after a light stimulus of 473 nm on the 12th day. Scale bar, 100 μm. K, The c-Fos quantification in the PrL (Vehicle, n = 6 mice, ChR2, n = 6 mice). p < 0.01; paired t test. L, The c-Fos expression in the dCA1 90 min after an optic stimulus of 473 nm on the 12th day. Scale bar, 100 μm. M, The c-Fos quantification in the dCA1 (Vehicle, n = 6 mice, ChR2, n = 6 mice). p < 0.01; paired t test. N, O, The fluorescent schematic of costaining of c-Fos (green), ChR2-mCherry (red), and DAPI (blue) in the PrL (N) and dCA1 (O). Scale bar, 20 μm. P, Quantification of fluorescent costaining of c-Fos with ChR2-mCherry in the PrL after optical stimulation in the PrL (above) leading to c-Fos expression in the dCA1 (below). Data in I, K, and M are presented as mean ± SEM. Two-sided statistical tests were used. *p < 0.05, **p < 0.01; ns, not significant.
The PrL to the dCA1 optogenetic regulation on METH-CPP
The above data showed that activity in the PrL to the dCA1 pathway might regulate place preference in mice. To test the activity of the PrL to the dCA1 pathway in regulating METH-CPP, we employed the same rAAV2/9s expression (Fig. 2A), rAAV2/9s injection in the PrL and projection to the dCA1 (see Fig. 2B for ChR2 and Fig. 2C for eNpHR3.0). After rAAV2/9s injection 3 weeks, mice were habituated for 2 d before the pre-CPP measurement and then injected with METH (2.5 mg/kg) or saline on the alternate day for 8 d. METH induced CPP in mice of vehicle and either ChR2 or eNpHR3.0 expressing mice. However, when optical stimulation was applied 24 h later followed by detection of METH-CPP (Fig. 2D), stimulation of ChR2 in the PrL-dCA1 reduced CPP, while stimulation of eNpHR3.0 in the PrL-dCA1 potentiated METH-CPP (Fig. 2E). These results showed that activating top-down of PrL-dCA1 suppressed METH-CPP, whereas optogenetic inhibition on this pathway enhanced METH-CPP (ChR2 = −119.58 ± 52.08 s; eNpHR3.0 = 204.37 ± 50.67 s; Vehicle = 46.80 ± 36.55 s, respectively). Note the overlap of mCherry with neuron soma-specific DAPI in the PrL injection site and the overlap between the mCherry terminals and DAPI in the dCA1 (Fig. 2F,G). In either experiment, METH-CPP was unaffected by stimulation in vehicle mice transfected with rAAV2/9. The experiment shown in Figure 2D could contain an ordering confound since the optogenetic posttest session was always administered following the control posttest session for METH-CPP. Using the experimental schematic for the rAAV2/9s and METH-CPP in Figure 2H where vehicle posttest was compared with optogenetic posttest only, akin to Figure 2E, METH-CPP was induced in vehicle mice (117.98 ± 42.50 s), blocked by optical stimulation in the ChR2-expressing mice (−84.39 ± 29.41 s) and potentiated in the eNpHR3.0-expressing mice (256.14 ± 32.64 s; Fig. 2I). Figure 2, J and K, shows anatomical site of rAAV2/9s injection in the PrL and terminal mCherry in the dCA1. These results are consistent with the above results (Fig. 2E), confirming the same conclusions.
The PrL-dCA1 optogenetic regulation on METH-CPP. A, Schematic diagram of the major components of anterograde tracing rAAVs. B, C, Diagram of anterograde virus strategy to selective injection site in the PrL and optical fiber implantation site in the dCA1 by light stimuli of 473 nm, blue light (B), and 595 nm, yellow light (C), respectively. D, Schematic of the experimental timeline of METH-CPP test on the 12th day (no opto) and optical stimulation on the 13th day. E, Quantification of METH-CPP expression in Vehicle (n = 8 mice) versus ChR2 (n = 9 mice) versus eNpHR3.0 expressing mice (n = 7 mice) behavioral experiments trained in combined with light stimulation on the 13th day [with opto; two-way ANOVA with repeated measures over session, interaction, F(4,63) = 7.695, p < 0.0001; CPP (pre-, post-no, post-with opto), F(2,63) = 10.01, p = 0.0002, groups (Vehicle, ChR2, eNpHR3.0), F(2,63) = 21.47, p < 0.0001, Bonferroni's multiple-comparisons test]. F, Representative images showing the expression of the ChR2-mCherry fluorescent expressions in the coronal PrL (n = 3 mice). The red solid wireframe represents a magnified view of a specific area (left), the white dashed box represents the peripheral brain region structures of the PrL (middle), and the white arrow indicates the neuronal cell body merging ChR2-mCherry with DAPI in the PrL (right). Scale bars: 100 μm, 50 μm, 20 μm. G, Representative image of the ChR2-mCherry fluorescent from PrL neurons to target synapse in the dCA1 (n = 3 mice). The red solid wireframe represents a magnified view of a specific area (left), the white dashed box represents the peripheral brain region structures of dCA1 (middle), and the white arrow indicates the terminals from the mPFC neurons (right). Scale bars: 100 μm, 50 μm, 20 μm. H, Schematic of the experimental timeline of METH-CPP test and light stimulation on the 12th day (with opto) combining with the optogenetic experiment to stimulate the PrL to dCA1 pathway. I, Quantification of METH-CPP expression in Vehicle (n = 7 mice) versus ChR2 (n = 8 mice) versus eNpHR3.0 expressing mice (n = 6 mice) behavioral experiments trained in combined with light stimulation on the 12th day [with opto; two-way ANOVA, interaction, F(4,63) = 7.695, p < 0.0001; CPP (pre-, post-no, post-with opto), F(2,63) = 10.01, p = 0.0002, groups (Vehicle, ChR2, eNpHR3.0), F(2,63) = 21.47, p < 0.0001, Bonferroni's multiple-comparisons test]. J, Representative images showing the expression of the eNpHR3.0-mCherry fluorescent expressions in the coronal PrL (n = 3 mice). The red solid wireframe represents a magnified view of a specific area (left), the white dashed box represents structures of PrL (middle), and the white arrow indicates the neuronal cell body merging eNpHR3.0-mCherry with DAPI in the PrL (right). Scale bars: 100 μm, 50 μm, 20 μm. K, Representative image of the eNpHR3.0-mCherry fluorescent from PrL neurons to target synapse in the dCA1 (n = 3 mice). The red solid wireframe represents a magnified view of a specific area (left), the white dashed box represents structures of dCA1 (middle), and the white arrow indicates the terminals from the mPFC neurons (right). Scale bars: 100 μm, 50 μm, 20 μm. Data in E and I are presented as mean ± SEM. Two-sided statistical tests were used. *p < 0.05, **p < 0.01, ****p < 0.0001; ns, not significant.
Optogenetic activation or inhibition of glutamatergic PrL to dCA1 inhibits or enhances METH-CPP
To characterize the type of projection between the PrL and the dCA1 and its role in regulating CPP, we employed a retrograde rAAV2/Retro-hSyn-Cre-EYFP into the dCA1 in mice combined with injecting rAAV2/9-Ca2+/calmodulin calcium-calmodulin (CaM)-dependent protein kinase II alpha (CaMKⅡα)-double-floxed inverse orientation (DIO)-mCherry into the PrL (Fig. 3A–C). Fluorescent protein expression was observed in the PrL and the dCA1, and mCherry and EYFP were coexpressed in the PrL (Fig. 3D) and the dCA1 (Fig. 3E). The PrL glutamatergic axon terminal regions were also observed and other brain regions, from collateralizing glutamatergic PrL neurons were also identified (Fig. S3).
The glutamatergic PrL to the dCA1 regulation on METH-CPP. A, Schematic diagram of the major components of anterograde tracing rAAVs and rAAV2/Retro. B, C, Diagram of anterograde virus strategy to selective injection site in the PrL and retrograde virus in the dCA1 by light stimuli of 473 nm, blue light (B), and 595 nm, yellow light (C), respectively. D, E, Fluorescence expression of CaMKII-mCherry, Retro-Cre-EYFP, and DAPI in coronal section scanning located in PrL (n = 3, D) and the dCA1 (n = 3, E) and the peripheral brain region structures, at different distances from bregma showing different expression profiles (left), a coronal fluorescence boxed in red (middle), and a localized fluorescence zoom in image boxed (right). FrA, frontal association cortex; MO, medial orbital cortex; VO, ventral orbital cortex; LO, lateral orbital cortex; M2, secondary motor cortex. Scale bar: 500 μm, 50 μm. F, Schematic of the experimental timeline of METH-CPP test and laser stimulation on the 12th day to stimulate the PrL to the dCA1 pathway. G, Quantification of METH-CPP in Vehicle (n = 6 mice) versus ChR2 (n = 7 mice) versus eNpHR3.0 expressing mice (n = 8 mice) combined with light stimulation on the 12th day [two-way ANOVA, interaction, F(4,63) = 7.695, p < 0.0001; CPP (pre-, post-no, post-with opto), F(2,63) = 10.01, p = 0.0002, groups (Vehicle, ChR2, eNpHR3.0), F(2,63) = 21.47, p < 0.0001, Bonferroni's multiple-comparisons test], *p < 0.05, **p < 0.01, ****p < 0.0001; ns, not significant. H, Schematic of the experimental timeline of whole-cell patch recordings and optical stimulation. I, Representative images showing the expression of the ChR2-mCherry fluorescent expressions in the coronal PrL. Scale bars: 100 μm, 20 μm. J, Schematic of current-clamp recording from PrL mCherry neurons. The blue lines indicate light delivery. K, Action potential recording in blue light pulses at 5, 10, and 20 Hz reliably activated mCherry-expressing neurons in the PrL. L, Representative images showing the expression of the ChR2-mCherry fiber fluorescent expressions in the coronal dCA1. Scale bars: 100 μm, 20 μm. M, Schematic of voltage-clamp recording on dCA1 neurons combining blue light with PrL fibers. N, Representative EPSC traces recorded from dCA1 neurons by optical stimulation of PrL fibers in the dCA1 in sequence of bath application of ACSF, TTX, TTX + 4-AP, and TTX + 4-AP + CNQX + D-AP5.
Figure 3F showed the experimental schematic for the AAV microinjection, optical stimulation, and CPP. We detected the CPP after rAAV2/9s expression and METH administration in mice receiving rAAV2/9-CaMKⅡα-DIO-mCherry without optical stimulation. METH-CPP was suppressed in mice receiving optogenetic activation of CaMKIIα-ChR2 and enhanced in mice receiving optogenetic inhibition of CaMKIIα-eNpHR3.0 (Fig. 3G; ChR2 = −124.81 ± 34.56 s; eNpHR3.0 = 151.93 ± 24.37 s; Vehicle = 70.73 ± 18.20 s, respectively). The expression of ChR2-mCherry and eNpHR3.0-mCherry in the PrL and the projection into the dCA1 are shown in Figure S4. These results further validate the presence of a glutamatergic projection from the PrL to the dCA1 that regulates METH-CPP in mice.
The PrL glutamatergic neurons are functionally connected with the dCA1
Given the above-identified anatomical projections from the PrL to the dCA1, we next investigated the type of synaptic connectivity between these regions. We infused an adeno-associated virus (AAV) containing mCherry under the control of CaMKIIα promotor (AAV-CaMKIIα-mCherry) into the PrL to induce mCherry expression in excitatory neurons and optogenetically stimulated the glutamatergic PrL to dCA1 pathway (Fig. 3H). The ChR2 fused mCherry (AAV-CaMKIIα-ChR2-mCherry) was expressed in PrL neurons (Fig. 3I) and AAV-retro-hSyn-Cre-EYFP was expressed in dCA1 neurons (Fig. 3L). Using whole-cell current-clamp recordings (Fig. 3J), we observed that blue light pulses reliably induced pulse-locked action potentials in the ChR2-expressing PrL neurons (Fig. 3K). This observation verified the efficacy of ChR2 expression in PrL neurons. We then proceeded to determine whether PrL neuronal axons make functional synaptic contacts with dCA1 neurons. Photostimulation of ChR2-expressing PrL axons in the dCA1 reliably evoked excitatory postsynaptic currents (EPSCs) in recorded dCA1 neurons (Fig. 3M,N). Moreover, the EPSCs were completely blocked by the voltage-gated sodium channel blocker tetrodotoxin (TTX) and were restored by additional bath application of the potassium channel blocker 4-aminopyridine (4-AP). Moreover, the restored EPSCs were eliminated by the AMPA receptor antagonist, CNQX, and the NMDA receptor antagonist, D-AP5 (Fig. 3N). The above sequence of experiments indicates that the dCA1 neurons are monosynaptically innervated by the glutamatergic presynaptic axons originating from the PrL neurons.
Silencing the PrL to the dCA1 glutamatergic pathway enhances METH-CPP
To further study the necessity of the PrL to the dCA1 glutamatergic pathway in METH-CPP, the effect of silencing the PrL to the dCA1 glutamatergic pathway was further investigated using tetanus toxin (TeNT) to impede release of glutamatergic synaptic vesicles. A previous study showed that the expression of TeNT efficiently silenced the basolateral amygdala (BLA) to nucleus accumbens (NAc) pathway by blocking synaptic transmission through electrophysiological experiment (Zhou et al., 2022). The rAAV2/9s containing rAAV2/9-CaMKⅡα-mCherry (Fig. 4A) were injected into the PrL by stereotaxic surgery, rAAV2/Retro-hSyn-Cre-EYFP into the dCA1 (Fig. 4B). Experimental schematic for rAAV2/9s and METH-CPP procedure were described in Figure 4C. On Test day 12, METH-CPP was expressed in control mice of vehicle rAAV2/9 and was significantly enhanced in mice of TeNT (Fig. 4D; METH = 51.66 ± 27.58 s; METH-TeNT = 158.31 ± 17.71 s, respectively). These results showed that silencing the PrL to dCA1 glutamatergic transmission significantly enhanced METH-CPP. Confocal images showed expression of mCherry in the PrL (Fig. 4E) and the projection to the dCA1 (Fig. 4F). This result is consistent with Figure 3G.
Silencing glutamatergic PrL-dCA1 further promotes METH-CPP. A, Schematic diagram of the major components of anterograde tracing rAAVs and rAAV2/Retro. B, Diagram of anterograde virus strategy to selective injection site in the PrL and retrograde virus in the dCA1 to silence the PrL to the dCA1 glutamatergic pathway with TeNT. C, Schematic of the experimental timeline of METH-CPP test to stimulate the PrL to dCA1 pathway. D, METH-CPP expression of Vehicle (n = 12 mice) versus TeNT (n = 8 mice) expressing mice in saline (n = 19 mice) versus METH groups [n = 18 mice; two-way ANOVA, interaction, F(3,66) = 9.006, p < 0.0001; CPP (Saline, METH), F(3,66) = 14.67, p < 0.0001; treated (Vehicle, TeNT), F(1,66) = 31.75, p < 0.0001, Bonferroni's multiple-comparisons test; *p < 0.05, **p < 0.01, ****p < 0.0001; ns, not significant]. E, Representative images showing expression of the TeNT-mCherry fluorescence in coronal diagrams of PrL (n = 3 mice). The red solid wireframe represents a magnified view of a specific area (left), the white dashed box represents structures of PrL (middle), and the white arrow indicates the neuronal cell body merging TeNT-mCherry with DAPI in the PrL (right). Scale bars: 100 μm, 50 μm, 20 μm. F, Representative image of the TeNT-mCherry fluorescent from PrL neurons to target synapse in the dCA1 (n = 3 mice). The red solid wireframe represents a magnified view of a specific area (left), the white dashed box represents structures of the dCA1 (middle), and the white arrow indicates the terminals merging TeNT-mCherry from the mPFC neurons with EYFP and DAPI in the PrL (right). Scale bars: 100 μm, 50 μm, 20 μm. Data in D was presented as mean ± SEM.
Effect of inhibiting glutamate release on METH-CPP alterations on glutamate transmission-related proteins
Given the potent interaction of TeNT with METH-CPP, we sought to determine if TeNT altered METH-induced changes in glutamate receptor and transport-related proteins in the PrL and the dCA1. After inducing METH-CPP, we first used immunofluorescent labeling for the vesicular glutamate transporter1 (VGLUT1) in the PrL and the dCA1 (Fig. 5A,C). Figure 5, B and D, shows that TeNT did not alter METH-CPP-associated increases in VGLUT1 in either the PrL or the dCA1 but did reduce the levels of VGLUT1 in both brain regions in saline-trained mice. We next conducted Western blotting of PrL and dCA1 tissue to quantify levels of protein subunits of NMDAR (GluN2b) and AMPAR (GluR1), along with VGLUT1, after METH-CPP, with or without TeNT expression (Fig. 5E–K). The expression of all three proteins was increased by METH in the PrL and the dCA1 in vehicle mice regardless of TeNT expression, but in saline mice TeNT reduced expression of all three proteins in both brain regions. These results showed that while reduced basal levels of glutamate transmission-related receptor and transporter proteins, TeNT did not alter the capacity of METH-CPP to elevate the proteins.
The expression of NMDAR2B, AMPAR and VGLUT1 in the PrL and the dCA1. A, Confirmation by VGLUT1-immunofluorescence in 568 nm (red) of the PrL. L, lateral; D, dorsal; M, medial; V, ventral. Scale bars: 100 μm, 20 μm. B, VGLUT1 relative fluorescent expression in the PrL by two-way ANOVA [interaction, F(1,8) = 3.010, p = 0.1210; Vehicle (Saline, METH), F(1,8) = 60.36, p < 0.0001; treated (Vehicle, TeNT), F(1,8) = 341.5, p < 0.0001, Bonferroni's multiple-comparisons test]. C, Confirmation by VGLUT1-immunofluorescence in 640 nm (white) of the dCA1. Scale bars: 100 μm, 20 μm. D, VGLUT1 relative fluorescent expression in the dCA1 by two-way ANOVA [interaction, F(1,8) = 11.76, p = 0.0090; Vehicle (Saline, METH), F(1,8) = 62.24, p < 0.0001; treated (Vehicle, TeNT), F(1,8) = 752.9, p < 0.0001, Bonferroni's multiple-comparisons test]. E, The expression of GluN2B, GluR1, and VGLUT1 in the PrL and in the dCA1 by Western blotting. F–H, The relative expression on GluN2B, GluR1, and VGLUT1 in the PrL using a two-way ANOVA, interaction, F(1,18) = 1.415, p = 0.2497; Vehicle (Saline, METH), F(1,18) = 2.869, p = 0.1075; treated (Vehicle, TeNT), F(1,18) = 84.61, p < 0.0001, Bonferroni's multiple-comparisons test; G, ANOVA, interaction, F(1,19) = 1.864, p = 0.1881; Vehicle (Saline, METH), F(1,19) = 7.467, p = 0.0132; treated (Vehicle, TeNT), F(1,19) = 49.68, p < 0.0001; H, ANOVA, interaction, F(1,18) = 5.709, p = 0.0280; Vehicle (Saline, METH), F(1,18) = 18.71, p = 0.0004; treated (Vehicle, TeNT), F(1,18) = 71.06, p < 0.0001. I–K, The relative expression on GluN2B, GluR1, VGLUT1 in the dCA1. I, ANOVA, interaction, F(1,20) = 0.9132, p = 0.3507; Vehicle (Saline, METH), F(1,20) = 11.33, p = 0.0031; treated (Vehicle, TeNT), F(1,18) = 119.9, p < 0.0001; J, ANOVA, interaction, F(1,20) = 0.0967, p = 0.7590; Vehicle (Saline, METH), F(1,20) = 10.45, p = 0.0042; treated (n.c, TeNT), F(1,20) = 113.7, p < 0.0001; K, ANOVA, interaction, F(1,20) = 10.57, p = 0.0040; Vehicle (Saline, METH), F(1,20) = 29.16, p < 0.0001; treated (n.c, TeNT), F(1,20) = 99.83, p < 0.0001. Data in B, D, F–K are presented as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001; n.c, negative vehicle group.
Discussion
We identified a monosynaptic glutamatergic pathway from the mPFC to the dorsal hippocampus that regulated place preference and METH-induced place preference.
The prefrontal cortex involved in METH addiction
The prefrontal cortex (PFC) is a key “hub” in associative learning events. Dysregulation of the PFC and specific subregions of PFC also contributes to distortions in reward learning associated with SUDs (Zhang et al., 2019). METH use disorder (MUD) results in gross perturbations in executive function and other aspects of cognition governed by the PFC, and these dysfunctions relate with the PFC metabolic hypoactivity (Goldstein and Volkow, 2011) that can arise from perturbations in the PFC excitatory glutamate transmission (McCann et al., 2008). Involvement of the PFC glutamate transmission is evidenced by enduring reductions in glutamate/glutamine content in METH-abstinent individuals (Ernst and Chang, 2008). The PFC hypoactivity is also linked to compulsive seeking in SUDs (Chen et al., 2013). In rodents, METH injection is sufficient to impact extracellular glutamate within both the cell body and terminal regions of excitatory corticostriatal projections (Han et al., 2012) that are highly implicated in SUD neurocircuitry (London et al., 2015). Thus, increasing the PFC activity or reversing the hypoactivity of PFC could suppress METH addiction. The CPP results here are explained as in Figure 1H. During the Posttest day 12, mice had ad libitum access to both sides of the chamber with optogenetic fibers, which are regulated by laser. The result in Figure 1H showed that there was no time difference in two chambers in mice of vehicle. The time when mice were in the chamber paired with blue light (ChR2) stimulation decreased compared with time when mice in the chamber had no light stimulation; however, the time when mice were paired with yellow light (eNpHR3.0) stimulation increased compared with time when mice in the chamber with no light stimulation (Fig. 1I). Although c-Fos expression was not quantified when stimulation of eNpHR3.0 in the PrL-dCA1 was performed, a previous study showed that the expression of c-Fos was decreased when laser-on treatment with eNpHR3.0 in the dCA3 was given to alter cocaine memory extinction (Qi et al., 2022). We believe our results are more consistent suppressed behavior rather than aversion. Activating ChR2 in the PrL-dCA1 inhibited METH-CPP expression compared with vehicle. There is no significant difference between the pre-CPP and the post-CPP with opto in the ChR2 group (Fig. 2E). These suggests optogenetic activation of the mPFC-dCA1 inhibited METH-CPP, whereas optogenetic inhibition of this pathway enhanced METH-CPP (Fig. 2).
Top-down regulation from the PFC to the hippocampus involved in METH-CPP
The PFC is postulated to communicate with many cortical and subcortical brain regions, providing top-down control of behavior (Guise and Shapiro, 2017). Among these projections, the hippocamp-prefrontal pathway is unique as it connects rostral and caudal ends of the cerebral cortex. The hippocampus forms spatial codes that are critical for navigation (Robinson et al., 2020). These codes are based on features of the environment, including sensory cues and the location of rewards (Muller and Kubie, 1987; Leutgeb et al., 2005; Gauthier and Tank, 2018). Interactions between the PFC and the hippocampus play a key role in cognitive and emotional behaviors and contextual learning (Eichenbaum, 2017). Although anatomical connections between the PFC and the hippocampus are known (Naber and Witter, 1998; Thierry et al., 2000), knowledge about direct anatomical projections mediating this top-down control has been lacking. The currently available study of direct excitatory-to-excitatory connectivity from the PFC to the hippocampus is a sparse projection from the anterior cingulate cortex (ACC), which is related with control of memory retrieval (Rajasethupathy et al., 2015). Optical inhibition of the afferent glutamatergic projection to the accumbens core (NAcore) from the PrL has been reported to block reinstated cocaine-seeking behavior (Stefanik et al., 2016). Whereas a glutamatergic projection from the infralimbic cortex (IL) to nucleus accumbens shell (NAcSh) is necessary for extinction training to suppress cocaine seeking (Peters et al., 2008). Thus, the pathway of the PFC to the hippocampus may be involved in METH seeking. We will examine the effects of the PrL to dCA1 optogenetic regulation on METH-CPP extinction and METH-CPP reinstatement in future studies. However, a top-down prefrontal pathway consisting of GABAergic projections targets disinhibitory microcircuits, thereby enhancing signals and network dynamics underlying exploratory behavior (Malik et al., 2022). Our data showed that the PrL-dCA1 projection is glutamatergic and monosynaptic, which regulates METH-CPP (Fig. 3). Thus, optogenetically activating this circuit reduced, while inhibition potentiated the expression of METH-CPP.
The glutamate-related molecules involved in METH-CPP
Increasing or decreasing NMDAR and AMPAR subunit trafficking to the membrane of PFC neurons during repeated exposure to psychostimulants leads to maladaptive plasticity, cognitive decline and addiction (Bernheim et al., 2016; Bisagno et al., 2016). Studies demonstrate that METH exposure elevates GluN2B surface expression in the mPFC and enhances neuronal activity in the NAc (Mishra et al., 2017). Another study demonstrates that METH exposure reduces glutamate receptor and vesicular protein levels in the dorsomedial striatum (DMS) but increases them in the dorsolateral striatum (DLS; Furlong et al., 2018). AMPA adaptations also contribute cocaine and METH incubation (Scheyer et al., 2016). The vesicular glutamate transporter (VGLUT) is responsible for glutamate uptake by vesicles (Boulland et al., 2004; Travaglia et al., 2018), and pharmacological regulation of VGLUT expression alters glutamate release and glutamate-dependent synaptic plasticity involved in SUDs (Goldstein and Volkow, 2011). Repeated METH exposure produces an enduring deficit in NMDAR function within the PFC that contributes to the neurocognitive pathologies characteristic of MUD (Goldstein and Volkow, 2011). Our results showed that expression of GluN2B, GluR1 and, VGLUT1 was increased by METH in both the PrL and the hippocampus, and the basal levels of GluN2B, GluR1, and VGLUT1 in TeNT silencing mice was lower than that in vehicle mice, which suggests enduring deficit in glutamate function contributed to augmenting METH-CPP (Figs. 4, 5). Tonic activation of the hippocampus via the excitatory pathway from the mPFC is critical in the development of METH-CPP (Han et al., 2014). Since the mesocorticolimbic dopamine (DA) system (Wise and Robble, 2020) is a prominent focus in research on reward processing and drug addiction, and METH increases GluN2B synaptic currents in the midbrain dopamine neurons (Li et al., 2017), we also quantified the dopamine transporter (DAT) in the dCA1, but the DAT expression in the dCA1 was not altered (Fig. S5). This result further supports top-down glutamatergic PrL to dCA1 role in regulating METH-CPP.
Activating the PrL to the dCA1 glutamatergic pathway inhibits METH-CPP, posing the possibility that manipulating activity in PrL to dCA1 projection may be therapeutically consequential in MUD. Activity in the mPFC has been shown to influence the acquisition, recall, and extinction of opiate-related reward memories, showing unique patterns of tonic and phasic activity patterns during these separate components of the opiate-related reward learning and memory recall (Sun et al., 2011). Also, cocaine exposure induces hypofrontality, which is related to the compulsive craving for cocaine use (Sasase et al., 2019). Moreover, GluN2B-containing NMDAR-mediated synaptic plasticity in the dorsal hippocampus is related to cocaine seeking (Werner et al., 2020). Thus, this projection can be expected to play the same role in regulating cocaine and morphine rewards and seeking.
There are a few studies on the PrL-dCA1 impact on the ventral hippocampus. Rajasethupathy et al. identified a monosynaptic projection from the PFC (predominantly the ACC) to the hippocampal CA3–CA1 region. They demonstrated that stimulating these prefrontal-hippocampal projections during learning induced a sparse class of neurons in the CA2/CA3. These neurons exhibit strong correlation with the local network and drive synchronous population activity events (Rajasethupathy et al., 2015). In Avigan's study, rewarded episodes were found to engage context-associated rules in the mPFC, which in turn activated selective prospective memory codes in both the dorsal hippocampus and the ventral hippocampus. The dorsal-ventral hippocampus circuits and their mPFC interactions are essential for flexible spatial learning (Avigan et al., 2020). Therefore, the PrL-dCA1 projection may have local networks and interactions with the ventral hippocampus in specific behaviors.
Summary
Our results show that there is a monosynaptic glutamatergic pathway from the PrL to the dCA1. Activation of the PrL to the dCA1 glutamatergic pathway via ChR2 inhibits the expression of METH-CPP, whereas inhibition this pathway enhances METH-CPP expression via eNpHR3.0 and TeNT blockade of glutamate release. Thus, this is a novel brain circuit that regulates contextual learning in a necessary and sufficient manner.
Footnotes
This work was supported by National Natural Science Foundation of China (No. U2002220, 82371271). The innovation team of stress and disorder in nervous system in Yunnan Province (No. 202305AS350011). We thank Professor Lin Xu, Rongrong Mao, and Bin Gou for their helpful suggestions and discussion.
The authors declare no competing financial interests.
This paper contains supplemental material available at: https://doi.org/10.1523/JNEUROSCI.0374-25.2025
- Correspondence should be addressed to Peter W. Kalivas at kalivasp{at}musc.edu or Jie Bai at jiebai662001{at}126.com.
