Using Optogenetics to Reverse Neuroplasticity and Inhibit Cocaine Seeking in Rats

Summary

The methods described here outline a procedure used to optogenetically reverse cocaine-induced plasticity in a behaviorally-relevant circuit in rats. Sustained low-frequency optical stimulation of thalamo-amygdala synapses induces long-term depression (LTD). In vivo optogenetically-induced LTD in cocaine-experienced rats resulted in the subsequent attenuation of cue-motivated drug seeking.

Abstract

This protocol demonstrates the steps needed to use optogenetic tools to reverse cocaine-induced plasticity at thalamo-amygdala circuits to reduce subsequent cocaine seeking behaviors in the rat. In our research, we had found that when rats self-administer intravenous cocaine paired with an audiovisual cue, synapses formed at inputs from the medial geniculate nucleus of the thalamus (MGN) onto principal neurons of the lateral amygdala (LA) become stronger as the cue-cocaine association is learned. We hypothesized that reversal of the cocaine-induced plasticity at these synapses would reduce cue-motivated cocaine seeking behavior. In order to accomplish this type of neuromodulation in vivo, we wanted to induce synaptic long-term depression (LTD), which decreases the strength of MGN-LA synapses. To this end, we used optogenetics, which allows neuromodulation of brain circuits using light. The excitatory opsin oChiEF was expressed on presynaptic MGN terminals in the LA by infusing an AAV containing oChiEF into the MGN. Optical fibers were then implanted in the LA and 473 nm laser light was pulsed at a frequency of 1 Hz for 15 minutes to induce LTD and reverse cocaine induced plasticity. This manipulation produces a long-lasting reduction in the ability of cues associated with cocaine to induce drug seeking actions.

Introduction

Substance abuse is a very serious public health issue in the U.S. and worldwide. Despite decades of intense research, there are very few effective therapeutic options^1,2. A major setback to treatment is the fact that chronic drug use generates long-term associative memories between environmental cues and the drug itself. Re-exposure to drug-related cues drives physiological and behavioral responses that motivate continued drug use and relapse³. A novel therapeutic strategy is to enact memory-based treatments that aim to manipulate the circuits involved in regulating drug-cue associations. Recently, it was observed that synapses in the lateral amygdala (LA), specifically those arising from the medial geniculate nucleus (MGN) of the thalamus, are strengthened by repeated cue-associated cocaine self-administration, and that this potentiation can support cocaine seeking behavior^4,5. Therefore, it was proposed that cue-induced reinstatement could be attenuated by reversing plasticity at MGN-LA synapses.

The ability to precisely target the synaptic plasticity of a specific brain circuit has been a major challenge to the field. Traditional pharmacological tools have had some success in decreasing relapse behaviors, but are limited by the inability to manipulate individual synapses. However, the recent development of in vivo optogenetics has provided the tools needed to overcome these limitations and control neural pathways with temporal and spatial precision^6,7,8. By expressing light-sensitive opsins in a specific brain circuit, laser light can then be used to activate or inhibit the circuit. Frequency-dependent optical stimulation can be utilized to specifically manipulate the synaptic plasticity of the circuit in a behaving animal.

This manuscript outlines the procedure taken to manipulate the behaviorally-relevant MGN-LA circuit using in vivo optogenetics. First, the excitatory opsin oChIEF was expressed in the MGN and optical fibers were bilaterally implanted in the LA. Animals were then trained to self-administer cocaine in a cue-dependent fashion, which potentiates the MGN-LA pathway. Next, sustained, low frequency stimulation with 473 nm laser light was used to produce circuit-specific LTD. Reversing the plasticity induced by cocaine use generated a long-lasting reduction in the capacity of cues to trigger actions that are associated with drug seeking behavior.

Protocol

The experiments described in this protocol were consistent with the guidelines set forth by the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the University of Pittsburgh's Institutional Animal Care and Use Committee. All procedures were performed using adult, naïve Sprague-Dawley rats that weighed 275-325 g upon arrival.

1. Construction of Optic Fiber Implants and Patch Cables

1. Prepare optic fiber implants following previously published protocols⁹. Experiments described in this protocol used 200 µm core fiber (0.5 NA) and Ø1.25 mm Multimode LC/PC Ceramic ferrule, Ø230 µm hole size.
   1. Use a Dremel tool to score the lower third of a ferrule (closest to the flat end of the ferrule). Scoring the ferrules helps them stay attached to dental cement, increasing the likelihood that they will remain secure throughout the entire extent of experimentation.
   2. Use wire cutters to cut ~35 mm of fiber. Use fiber stripping tool to strip ~25 mm of fiber, leaving 10 mm unexposed.
   3. Prepare heat-curable epoxy according to manufacturer's instructions. Dissolve 1 g of resin powder in 100 mg of hardener compound. Attach a blunted 25-gauge needle to a 1 mL syringe. Fill the syringe with epoxy and attach a blunted 25-gauge needle tip.
   4. Use a vice or clamp to securely hold the ferrule with the flat side facing up and the convex side facing down. With the epoxy-filled syringe, add one drop of epoxy to the flat side of the ferrule, using caution to wipe excess epoxy from the sides of the ferrule.
   5. Insert the stripped portion of fiber through the ferrule allowing an extra 5 mm of stripped fiber exposed. In the case of LA implants, the fiber will be implanted 7.9 mm ventral to bregma, so the exposed length of unstripped fiber should be ~13 mm.
   6. Cure the epoxy with a heat gun for about 30-40 s until it turns black/amber in color.
   7. Score the fiber directly at the interface of the convex end of the ferrule with diamond knife and use a finger to gently tap the fiber off.
   8. Polish the convex end of the ferrule by holding with a hemostat, being sure to apply even pressure, and making 20 circular rotations each on a series of polishing paper (from high grade to low grade; 5, 3, 1, 0.3 µm).
   9. Secure unstripped fiber to table using tape and score stripped fiber, leaving an extra 2 mm beyond the ventral coordinate (For LA implants, final length of fiber is ~10 mm). Use a hemostat to pull on the ferrule and evenly break the fiber where it has been scored. Be careful not to cut fiber completely when scoring, or else the core of the fiber will be damaged.
2. Build patch cables that are compatible with optic fiber implants. Custom designed patch cables were purchased (See Table of Materials). Alternatively, patch cables can be constructed following previously published protocols⁹.
   NOTE: The diameter and NA of the ferrule fiber and patch cable fiber must match at the coupling junction to prevent excess loss of light, which can result in a failure to sufficiently stimulate neural activity.
3. Measure the light output through the patch cable and optic fiber implants by attaching patch cable/optic fiber to an appropriate laser light source (473 nm, 1 mW output) and measuring output with a light sensor. A successfully constructed fiber will emit a concentric circle of light and have no more than 30% light loss.

2. Rodent Intravenous Catheterization, Virus Delivery, and Optic Fiber Implantation

1. Prepare animal for surgery.
   1. Fully anesthetize rats with anesthetic of choice based on institutional guidelines. One option is ketamine hydrochloride (87.5-100 mg/kg, i.m.) and xylazine hydrochloride (5 mg/kg, i.m.). Ensure that the rat is fully anesthetized by checking for lack of a toe pinch reflex.
      CAUTION: Ketamine is a controlled substance that must be handled according to institutional guidelines. 
      NOTE: Intramuscular injection of anesthetics is used in this study as it produced a more rapid and reliable anesthesia induction than intraperitoneal injection. Continuously monitor the rat's respiration and responsiveness and provide thermal support throughout the surgery. 
   2. Shave a large area of the rat's back (upper back from just above the shoulder blades to the middle of the back) as well as the area of the neck underneath the right forelimb, and the scalp.
   3. Place the rat in the surgical area and apply puralube (artificial tears) to the eyes. Inject a body weight volume of carprofen (analgesic) subcutaneously (s.c.) through the skin of the upper back, then inject 5 mL of Lactated Ringer's solution s.c. through the skin of the lower back.
   4. Sanitize all surgical sites by wetting a piece of sterile gauze with betadine and wiping it down the shaved surgical area using a circular motion. Then repeat the process with 70% ethanol. Repeat this alternating cycle three times. 
2. Perform intravenous catheter implantation according to previously published protocols^4,10.
   NOTE: A surgical drape is not used during this surgery to avoid irritation during catheter implantation. Sterilize all instruments and equipment before use. Use sterile gloves and change the gloves if any non-sterile surface is contacted.
3. Immediately following catheter implantations, secure the rat in a stereotaxic frame to perform AAV injections.
   1. Deliver a subcutaneous (s.c) injection of 2% lidocaine (0.2-0.3 mL) to the scalp as a local anesthetic.
      NOTE: Local anesthetic is not used during the intravenous catheter implantation to avoid alterations to the surgical outcomes. 
   2. Connect a 26-gauge stainless steel injection cannula to a Hamilton syringe filled with 1 µL of concentrated AAV solution: either AAV5-hSyn-tdTomato or AAV5-hSyn-oChIEF-tdTomato
      NOTE: oChIEF is a variant of the blue-light sensitive opsin channelrhodopsin (ChR2), that can respond to a wide range of frequencies^8,11, and therefore has utility for the low-frequency LTD experiments discussed in this protocol, but also for high-frequency LTP experiments (not discussed here). The oChIEF construct was donated by Dr. Roger Tsien and processed for packaging and purification by the Duke Viral Vector Core. At least 3-4 weeks is needed between the injection day and the day of LTD induction to allow for optimal virus expression in MGN axon terminals.
      CAUTION: Generally, AAV is regarded as a Biosafety Level 1 (BSL-1) organism, with low risk of self-infection unless a helper virus is used in its production. Its use requires IACUC approval and proper PPE must be used at all times in accordance with institutional guidelines to limit unnecessary exposure.
   3. Using a scalpel, make a 0.5 mm incision from the front to the rear of the skull, and remove overlying tissue to expose the surface of the skull.
   4. Level the rat's head in the anterior/posterior axis and zero stereotaxic coordinates to bregma.
   5. Drill three small holes through the skull using a Dremel tool equipped with a small drill bit. Use screwdriver to firmly mount stainless steel screws (M2x4 965-A2) in place.
      NOTE: Screws are necessary for proper binding of dental cement and creation of sturdy, long-lasting headcaps. Position of screws should be spread across the anterior-posterior axis of the skull, and away from AAV injection site.
   6. Drill bilateral holes for injection of AAV based on the coordinates from the Rat Brain Atlas (Watson and Paxinos)¹² for the medial portion of the medial geniculate nucleus (MGN); in mm from bregma, AP: -5.4; ML: ±3.0; DV: -6.6. Slowly lower injection cannulae (4 mm/min) until positioned in the MGN. Inject concentrated AAV solution at a rate of 0.1 µL/min.
   7. Leave injection cannula in place for 5 min after infusions are complete to allow for diffusion away from the cannula and then slowly withdraw cannula from the brain.
4. Immediately following virus injections, continue to implant optic fibers^4,9 targeting MGN-LA terminals.
   1. Use a Dremel tool to drill bilateral holes for optic fiber implants targeting the lateral amygdala (in mm from bregma, AP: -3.0; ML ±5.1).
   2. Use forceps to grab the ferrule of the optic fiber implant and fix it to the stereotaxic adapters so that they are securely held in place.
   3. Slowly lower fibers at a rate of 2 mm/min, until the tip of the fiber sits in the dorsal portion of the LA (DV: -7.9 mm).
   4. Secure ferrules to the skull first using a thin layer of Loctite instant adhesive followed by dental cement and cover ferrules with 1.25 mm diameter ferrule sleeves and dust covers.
      NOTE: The choice of adhesive for securing the ferrules to the skull should be approved by the local or institutional animal care and use committee. Loctite is used for this study to reliably secure the ferrules to the skull after trial and error with multiple adhesives; however, available alternatives can be considered. 
5. Following surgical procedures, house rats individually, and provide free access to food and water. Provide postoperative care consistent with institutional guidelines. Flush catheters daily with saline containing gentamicin (5 mg/mL) and heparin (30 USP/mL) to maintain patency. At least 5 days post-surgery and 24 h prior to the start of behavioral experiments, food restrict rats to ~90% of their free-feeding weight.

3. Rodent Cocaine Self-Administration and Instrumental Lever Extinction

NOTE: All behavioral procedures are conducted in standard operant conditioning chambers, equipped with two retractable levers on one wall, a stimulus light above each lever, a tone generator, a house light, and an infusion pump.

1. Subject rats to daily 1-h cocaine (2 mg/mL) self-administration training sessions under an FR1 schedule of reinforcement.
   1. Place rats in operant chamber each day and allow rats to lever press. A press on the designated 'active lever' (counterbalanced across left and right levers) results in a cocaine infusion (1.0 mg/kg/infusion) and a 10 s presentation of a compound light and tone cue. A press on the designated 'inactive lever' has no programmed effects.
   2. Continue self-administration experiments for at least 10 d and until rats successfully earn at least 8 infusions/day across 3 consecutive days. Failure to reach acquisition criteria by day 20 results in exclusion from the study.
2. After acquisition criteria are successfully met, subject rats to 1 h instrumental extinction sessions for 6-10 d.
   1. Place rats in operant chambers and allow rats to freely lever press. However, responses on both the active and inactive levers have no programmed consequences.
   2. Have rats continue instrumental extinction daily until an average of < 25 lever presses over two consecutive days occurs.

4. In vivo Optogenetic Induction of LTD

NOTE: Optogenetic inhibition experiments take place 24 h after the final day of instrumental extinction.

1. Connect patch cords to a 473-nm blue laser diode via a rotary joint suspended above a clean, standard rodent housing cage with the cover removed. This setup allows rodents to freely move around the cage during the optogenetic stimulation.
2. Turn on laser diode according to operating instructions and connect to a pulse generator. Adjust settings, so that when turned on the rat will receive 900 2-ms pulses of light at 1 Hz.
   CAUTION: Proper eye protection must be used at all times while operating laser.
3. Measure the light intensity through the patch cord using a light sensor. Adjust the intensity of the laser so that light output through the patch cable is ~5-7 mW.
4. Place rats in a clean housing cage. Remove dust covers and ferrule sleeves, exposing the ferrules. Connect patch cords bilaterally to the optic fiber implants. Allow rats to explore the environment for 3 min prior to LTD induction.
5. Turn on the pulse generator to initiate optogenetic stimulation.
   NOTE: Although unlikely, if rat experiences any adverse reactions to stimulation, the experiment is immediately terminated, and rats are properly euthanized based on institutional guidelines.
6. Following LTD induction, keep rats in the cage for 3 min, and then place back in their home cages.
7. For control experiments, use the same stimulation procedure on rats that express the AAV5-tdTomato control virus. For sham experiments, attach a patch cord to the optic fiber of rats that express the AAV5-oChIEF virus, but no stimulation is delivered during a 15 min session.

5. Test the Effect of Optogenetic Stimulation on Cue-Induced Cocaine Seeking

1. 24 h after in vivo optogenetic stimulations, place rats back in operant conditioning chambers. Rats are subjected to a 1-h standard cue-induced reinstatement session to assess cocaine seeking behavior.
   NOTE: During cue-induced reinstatement, a response on the active lever yields a 10-s presentation of the cocaine-associated cue, but no cocaine infusions.
2. Give a second reinstatement test at least 1 week after the first test to determine if optogenetic LTD induction results in a long-term suppression of cocaine seeking

6. Staining, Fluorescence, and Imaging for Histological Verification of Viral Expression and Optic Fiber Placement

1. Make 1x phosphate buffered saline (PBS) and 4% paraformaldehyde (PFA). Store both solutions on ice. Total volume will depend on the number of rats in the study (~100 mL of PBS and 200 mL of PFA will be used per rat).
   CAUTION: PFA is a toxic chemical and known carcinogen. Take proper care to avoid inhalation as well as contact with the skin. Its use and disposal should be in accordance with institutional guidelines, including the use of proper PPE and a chemical flow hood.
2. Setup peristaltic pump at a flow rate of 20 mL/min. Fill tubing of pump with 1x PBS. Attach a blunted 20-gauge needle to the end of the tubing.
3. Deeply anesthetize rats with sodium pentobarbital (100 mg/kg, i.p.). Confirm the depth of anesthesia by lack of response to toe pinch before proceeding further.
   NOTE: Sodium pentobarbital is used since the perfusion is a terminal procedure.
4. Use surgical scissors to cut open the abdominal cavity of the rat below the diaphragm. Cut through the rib cage rostrally along the lateral edges to expose the rat's heart. Use hemostat to clamp the rostral portion of the rib cage away from the heart. Cut away any overlying fat tissue that surrounds the heart.
5. Insert the blunted needle through the left ventricle and up into the aorta. Cut a small hole in the right atrium to drain solution as it returns to the heart.
6. Perfuse each rat with 1x PBS for 5 min followed by 4% PFA, pH 7.4 for 10 min.
7. Decapitate the rat, extract the brain, and postfix it in 4% PFA for 24 h. Then transfer the brain to 30% sucrose solution for 2-3 d.
8. Section brains at 50 µm using a cryostat.
9. Mount all slices containing the LA or MGN onto glass slides and coverslip.
10. Image slices using an epifluorescent microscope to verify AAV-oChIEF-tdTomato expression in the MGN and its projections to the LA, as well as placement of the optic fiber above the LA.

7. Perfusion and Acute Brain Slice Preparation for Electrophysiology Experiments

NOTE: Electrophysiological experiments are performed on a subset of animals to validate the success of in vivo LTD.

1. Prepare electrophysiology solutions using the reagents listed in Tables 1-3^4,13. Adjust the pH of all solutions to 7.4 with HCl and adjust osmolarity to 300-310 mOsm/kg H₂O. Make solutions fresh prior to experiments and store at 4 °C for up to 1 week. Saturate all solutions with carbogen (95% O₂/5% CO₂) at all times during use.
2. Using isoflurane according to the local or institutional animal care and use guidelines, deeply anesthetize the rat in an enclosed euthanasia chamber.  Confirm that the animal is fully anesthetized via the toe-pinch reflex.
3. Fill a small beaker with 50 mL of ice-cold cutting solution. Fill tubing of peristaltic pump with solution and adjust flow rate to 20 mL/min. Attach a blunted 20-gauge needle to the end of the tubing.
4. Open the abdominal cavity (See steps 6.4 and 6.5) and briefly perfuse rat with cutting solution (maximum 1-2 min).
5. Following perfusion, immediately decapitate rat. Remove the brain and place it in a small beaker filled with 4 °C cutting solution for 30 s-1 min.
6. Transfer brains with a spatula and quickly fix them to the chamber of a vibratome. Remove the pia using fine forceps. Fill the chamber with 4 °C cutting solution and prepare acute coronal slices (250 µm thick) of the amygdala at a velocity of 0.37 mm/sec and a frequency of 70 Hz.
7. As slices are obtained, place each one in a holding chamber filled with cutting solution and incubate at 37 °C for 10-12 min. About 5-7 slices containing the LA can be obtained per animal.
8. Transfer slices to a beaker of room temperature (RT) holding solution and allow to recover for >30 min prior to experimentation.
   ​NOTE: Slices generally remain healthy while kept in holding solution for 4-6 h. Due to the fluorescent nature of the AAV, slices are kept in low-light conditions.

8. Ex vivo Electrophysiological Recordings

1. Prepare intracellular solutions using the reagents listed in Tables 4-5.
   NOTE: Intracellular solutions should be made in advance of experiments and can be stored long-term (3-12 months) at -80 °C or short-term (1-2 months) at -20 °C. Solutions are pH adjusted to 7.3 (with CsOH for Cs-based intracellular solution and with KOH for K-based intracellular solution). Adjust to a final osmolarity of 290-300 mOsm/kg H₂O.
2. Prepare 500 mM picrotoxin stock solution dissolved in dimethylsulfoxide (DMSO).
   NOTE: Picrotoxin stocks are aliquoted and stored at -20 °C. The day of use, aliquots are thawed and added to recording solution to a final concentration of 100 µM.
   CAUTION: Picrotoxin is a non-competitive antagonist of GABA_A receptors, so infusion of picrotoxin has a stimulative effect. It is severely toxic by oral ingestion or skin absorption. Proper PPE must be used at all times when working with picrotoxin.
3. Transfer slices to an upright microscope designed for electrophysiology experiments.
4. During experiments, continuously bath perfuse slices with recording solution that is heated to 31-33 °C.
5. Magnify the LA using a 4x objective. Identify principal neurons by morphology with a 40x water immersion lens.
6. Use a glass pipette (3-5 MΩ) filled with either a Cs-based intracellular solution (for voltage clamp experiments) or a K-based intracellular solution (for current clamp experiments) to obtain whole-cell patch clamp recordings.
7. Identify AAV-infected MGN axonal projections under fluorescence (using an RFP filter). Stimulate projections using a blue light (473 nm) DPSS laser connected to a pulse generator.
   CAUTION: To limit laser exposure, collimated laser light is coupled to a fluorescent port on the microscope and focused onto the slice through the objective.
   NOTE: In voltage clamp mode, excitatory postsynaptic currents (EPSCs) are optically-evoked at 0.1 Hz. Neurons receiving inputs from AAV-infected MGN neurons will exhibit reliable EPSCs.
8. To induce ex vivo LTD, in current clamp mode, record a stable baseline of excitatory postsynaptic potentials (EPSPs) for at least 10 min. Next, deliver 900 2-ms pulses of 473-nm light at a frequency of 1 Hz (total time = 15 min). Then continuously record EPSPs at 0.1 Hz for ≥60 min.

Representative Results

A timeline outlining the order of experiments is shown in Figure 1. Throughout behavioral experiments, the number of cocaine infusions as well as the number of responses made on the active lever serves as a measure of the intensity of cocaine-seeking behavior. During the initial days of cocaine self-administration, the number of active responses should gradually increase across each acquisition day, before stabilizing during the second week. Conversely, inactive lever responses should remain low throughout the entirety of the experiment (Figure 2A). On the first day of instrumental extinction, there is typically an increase in the number of active lever responses, as the unexpected absence of cocaine results in the escalation of cocaine-seeking behavior. However, this response will gradually decrease with subsequent sessions as rats learn the new contingency, resulting in a low and stable number of active lever responses within 6-10 d (Figure 2B). Rats that fail to reach the specified acquisition criteria in either the self-administration or instrumental extinction phases of the experiment are removed from the study, and data is not included in final analysis.

Following instrumental extinction, re-exposure to cocaine-associated cues reinstates cocaine-seeking behavior, resulting in an increase in the number of active lever responses. This increase is observed in both groups of control experiments: rats that were injected with virus lacking oChIEF (AAV control) and rats that did not receive laser stimulation (SHAM control; Figure 3A) However, in vivo optogenetic LTD of MGN-LA terminals caused a reduction in subsequent cue-induced cocaine-seeking. 24 h following optogenetic LTD induction, the number of active lever presses was significantly reduced relative to both AAV controls and SHAM controls (Figure 3A). This low level of responding was maintained during a subsequent reinstatement test 7 days later (conducted in a subset of rats) (Figure 3B), indicating a persistent reduction in cue-motivated cocaine seeking across multiple reinstatement tests.

Ex vivo electrophysiological recordings from animals exposed to optical stimulation confirmed that the attenuation in reinstatement was indeed due at least in part to a modulation of MGN-LA synaptic plasticity. This was evidenced by a decrease in optically-evoked EPSC amplitude in LA neurons following exposure to optical LTD (Figure 4A). This attenuation in EPSC amplitude was specific to neurons that received optic stimulation, as EPSC amplitude remained unchanged in SHAM-controls. Additionally, LTD was unable to be generated in slices from rats that had already received in vivo optical stimulation, but was reliably evoked in neurons from rats that underwent SHAM stimulation, as evidenced by a sustained reduction in EPSP rise slope (Figure 4B). Thus, in vivo optical stimulation appears to occlude further LTD induction in slice. While recording, it is important to measure series resistance through the duration of the recording to ensure the maintained health of the patch. Cells with a change in series resistance beyond 20% are not accepted for data analysis. This is especially important for LTD experiments which last >60 min, as changes in series resistance can influence receptor and channel dynamics. To ensure that the afferents being stimulated during electrophysiological recordings originate in the MGN, it is important to collect slices through the extent of the thalamus. This serves as validation that the cell bodies of the MGN indeed fluorescently express AAV. In addition to visual confirmation, functional validation is also necessary. Under current-clamp conditions, AAV-oChIEF-infected MGN neurons fire action potentials in response to both high and low frequencies of 473-nm-light stimulation (Figure 4C).

All behavioral results were considered provisional until viral expression and optic fiber implants were histologically verified and proper placement was confirmed (Figure 5). Lack of AAV expression in either the MGN or LA and/or those in which the optic fibers were not correctly positioned within the dorsal LA were excluded from experimental analysis, but in some instances may be included as a negative anatomical control.

Figure 1: Timeline of experimental procedure. An outline of the critical steps of the protocol, including the sequential time course and duration of each experimental phase. Please click here to view a larger version of this figure.

Figure 2: Acquisition and extinction of cocaine self-administration. (A) Animals exhibit an increasing number of cocaine infusions and active lever responses across acquisition, and a low level of inactive lever responses. (B) Following an initial boost in lever pressing on day 1 of extinction, animals decrease responding on both active and inactive levers to a low, stable level. Error bars, mean ±SEM. This figure has been modified from Rich et al. 2019⁴. Please click here to view a larger version of this figure.

Figure 3: In vivo optogenetic LTD attenuates cue-induced reinstatement. (A) Optical LTD causes a significant reduction in active lever presses during reinstatement relative to animals that received control virus or SHAM control stimulation. Two-way ANOVA, main effect of group (F_(2,27) = 7.04, P = .004) and a day x group interaction (F_(2,27) = 8.08, P = .002); Bonferroni's post hoc analysis: ***p < .001. (B) 7 days later, rats underwent a second reinstatement test, revealing a significant reduction in active lever pressing in animals that previously underwent MGN-LA LTD relative to SHAM controls. Two-way ANOVA, main effect of group (F_(1,32) = 5.04, P = .032), significant interaction (F_(1,32) = 7.69, P = .009); Bonferroni's post hoc analysis, **p < .01. Error bars, mean ±SEM, n in bars, number of rats. This figure has been modified from Rich et al. 2019⁴. Please click here to view a larger version of this figure.

Figure 4: Functional validation of in vivo low-frequency optogenetic stimulation (A) In vivo dual hemisphere LTD of MGN-LA synapses attenuates EPSC amplitude relative to SHAM-controls (Unpaired t-test, t₍₁₀₎ = 2.73, *P = .021). Inset: Sample average EPSC traces evoked at E_rev-70 mV. Scale bars: 50 ms, 200 pA, n in bars, number of rats (neurons). (B) In vivo optical LTD induction occludes ex vivo LTD. 24 h after in vivo LTD induction, amygdala slices were prepared and the same stimulation protocol was applied. EPSP rise slope at MGN-LA terminals was reduced by ex vivo optical stimulation in neurons from animals that had received in vivo SHAM stimulation, but not in neurons from animals that had received in vivo optical LTD. n in italics, number of neurons. (C) Sample current clamp recordings from AAV-oChIEF-infected MGN neurons. Action potentials were elicited by blue light stimulation (5-100 Hz). Scale bars: 100 ms, 40 mV. Error bars, mean ±SEM. This figure has been modified from Rich et al. 2019⁴. Please click here to view a larger version of this figure.

Figure 5: Histological verification of viral expression and optic fiber placements. (A) Representative microscopic images showing DAPI and AAV-oChIEF-tdTomato expression in LA (Left) and MGN (Right). Scale bar: 2 mm. (B) Schematic showing injection of AAV-oChIEF-tdTomato and throughout the anterior-posterior extent of the LA (Left) and MGN (Right), and optic fiber placements in the LA. Dark red shading shows representation of smallest acceptable virus spread, and light pink shading shows representation of largest acceptable spread. Blue circles correspond to successful optic fiber placement in both hemispheres. Black circles correspond to successful optic fiber placement in only one hemisphere. Black "X" corresponds to unsuccessful fiber placement. To be included in final analysis, rats required viral dual hemisphere expression in the LA as well as successful placement of fibers. Coordinates are in mm, posterior from bregma. This figure has been modified from Rich et al. 2019⁴. Please click here to view a larger version of this figure.

Chemical	mM	MW	g/1000 mL
N-methyl-D-glucamine	92	195.215	17.96
Potassium Chloride	2.5	74.551	0.19
Sodium Phosphate Monobasic	1.25	119.98	0.15
Sodium Bicarbonate	30	84.01	2.52
HEPES	20	238.301	4.77
D-glucose	25	180.16	4.5
Sodium Ascorbate	5	198.11	0.99
Thiourea	2	76.12	0.15
Sodium Pyruvate	3	110	0.33
Magnesium Sulfate	10	Use 5.0 mL of 2.0 M Stock
Calcium Chloride	0.5	Use 250 μL of 2.0 M Stock
1. For 1 L of solution, add salts in the order listed to 850 mL ddH2O
2. pH with concentrated HCl to 7.3-7.4 (NMDG makes a basic solution)
3. Oxygenate for 5-10 min then add MgSO4 and CaCl2
4. Bring final volutme to 1 L with ddH2O and double check final pH
5. Check osmolarity with osmometer and adjust to 300-310 mOsm/kg H2O

Table 1: List of Ingredients for Extracellular Cutting Solution. Ingredients and instructions used for the preparation of the NMDG-based extracellular cutting solution.

Chemical	mM	MW	g/1000 mL
N-methyl-D-glucamine	86	195.215	5.03
Potassium Chloride	2.5	74.551	0.19
Sodium Phosphate Monobasic	1.25	119.98	0.15
Sodium Bicarbonate	35	84.01	2.94
HEPES	20	238.301	4.77
D-glucose	25	180.16	4.5
Sodium Ascorbate	5	198.11	0.99
Thiourea	2	76.12	0.15
Sodium Pyruvate	3	110	0.33
Magnesium Sulfate	1	Use 500 μL of 2.0 M Stock
Calcium Chloride	2	Use 1000 μL of 2.0 M Stock
1. For 1 L of solution, add salts in the order listed to 850 mL ddH2O
2. pH with 1 N HCl or KOH to 7.3-7.4
3. Oxygenate for 5-10 min then add MgSO4 and CaCl2
4. Bring final volutme to 1 L with ddH2O and double check final pH
5. Check osmolarity with osmometer and adjust to 300-310 mOsm/kg H2O

Table 2: List of Ingredients for Extracellular Holding Solution. Ingredients and instructions used for the preparation of the extracellular holding solution.

Chemical	mM	MW	g/1000 mL
N-methyl-D-glucamine	119	195.215	6.95
Potassium Chloride	2.5	74.551	0.19
Sodium Phosphate Monobasic	1.25	119.98	0.15
Sodium Bicarbonate	26	84.01	2.18
HEPES	5	238.301	1.19
D-glucose	12.5	180.16	2.25
Magnesium Sulfate	1	Use 500 μL of 2.0 M Stock
Calcium Chloride	2	Use 1000 μL of 2.0 M Stock
1. For 1 L of solution, add salts in the order listed to 850 mL ddH2O
2. pH with 1 N HCl or KOH to 7.3-7.4
3. Oxygenate for 5-10 min then add MgSO4 and CaCl2
4. Bring final volutme to 1 L with ddH2O and double check final pH
5. Check osmolarity with osmometer and adjust to 300-310 mOsm/kg H2O

Table 3: List of Ingredients for Extracellular Recording Solution. Ingredients and instructions used for the preparation of the extracellular recording solution.

Chemical	mM	MW	mg/50 mL
Cesium Methanesulfonate	108	227.997	1231.3
Cesium Chloride	15	168.36	126.3
Cesium-EGTA	0.4	Add 80 μL of 250 mM Cs-EGTA
TEA-Chloride	5	165.705	41.4
HEPES	20	238.301	238.3
QX-314-Br	1	343.31	17.2
L-glutathione	1	307.323	15.4
Sodium Phosphocreatine	7.5	255.1	95.7
Mg-ATP	2.5	507.18	63.4
Na-GTP	0.25	523.18	6.5
1. Start with 40-45 mL HPLC-grade H2O
2. Keep phosphocreatine, ATP, and GTP on ice at all times.
3. Add ingredients in the order listed in the table
4. pH to 7.3 with CsOH (About 200 μL of 2 M CsOH)
5. Use osmometer to check osmolarity.
6. Add HPLC-grade H2O to a final osmolarity of about 285-290 mOsm/kg H2O
7. Prepare 500-1000 μL aliquots and store at -80 °C or -20 °C

Table 4: List of Ingredients for Cesium Methanesulfonate Intracellular Electrophysiology Solution. Ingredients and instructions used for the preparation of cesium methanesulfonate intracellular solution.

Chemical	mM	MW	mg/50 mL
Potassium Gluconate	145	234.246	1698.2
Potassium Chloride	2.5	74.56	9.3
Sodium Chloride	2.5	58.44	7.3
K-BAPTA	0.1	Add 80 μL of K-BAPTA
HEPES	10	238.301	119.2
L-glutathione	1	307.323	15.4
Sodium Phosphocreatine	7.5	255.1	95.7
Mg-ATP	2	507.18	63.4
Tris-GTP	0.25	886.59	11.1
1. Start with 40-45 mL HPLC-grade H2O
2. Keep phosphocreatine, ATP, and GTP on ice at all times.
3. Add ingredients in the order listed in the table
4. pH to 7.3 with KOH (About 200 μL of 2 M KOH)
5. Use osmometer to check osmolarity.
6. Add HPLC-grade H2O to a final osmolarity of about 285-290 mOsm/kg H2O
7. Prepare 500-1000 μL aliquots and store at -80 °C or -20 °C

Table 5: List of Ingredients for Potassium Gluconate Intracellular Electrophysiology Solution. Ingredients and instructions used for the preparation of potassium gluconate intracellular solution.

Discussion

As described above, there are several critical steps that are important for achieving the proper experimental results. The protocol will likely only be effective in animals that properly acquire cocaine self-administration, and to date, it has only been tested using the parameters outlined above. It is possible that cocaine dose, schedule of reinforcement, and cue parameters can be modified with likely little effect on behavioral outcomes, with the exception that a second-order schedule of reinforcement may lead to amygdala-independent cocaine seeking that could reduce efficacy of the procedure, though this has not been directly tested¹⁴. There are several points throughout the protocol where validating proper construction and functioning of optic fibers, will help ensure successful optical stimulation. It is necessary to properly score ferrules to prevent loss from the headcap, and polish optic fibers, and to test that loss of light output through the implants does not exceed 30%⁹. Additionally, the laser stimulation parameters are important considerations. The laser should be operated at relatively low power (5-7 mW). Sustained, low frequency stimulation is used to induce LTD, and this can be functionally validated by measuring MGN-LA synaptic strength with electrophysiological recordings. Finally, results indicated a significant reduction in cue-induced reinstatement with a final n of 10 animals per group, however, experimenters should anticipate starting with a larger n, as it is likely that some animals will need to be excluded from final analysis. It is crucial to verify the proper anatomical placement and expression of virus and optical fibers, and to only use data from animals in which histology has been verified.

Despite the robust behavioral effect on cocaine-seeking observed with this protocol, there are several limitations that must be considered. For one, the method has only been tested in rats that were trained with a single audiovisual cue paired with cocaine. It's not clear what would happen in a scenario where multiple different cues were conditioned, which would be a more accurate representation of human addiction, whereby multiple environmental stimuli become highly associated with drug use^15,16,17. Evidence from our lab indicates that the ability of optogenetic LTD to reduce drug-seeking is due to a decrease in synaptic strength that weakens drug-cue-associated memories. However, it is not clear to what extent neutral, or memories that are not associated with cocaine self-administration might be affected by this protocol. Furthermore, while the method only affects synaptic strength at one circuit, other circuits may also be important for encoding memory and/or driving cocaine seeking behavior^18,19,20. Finally, it should be noted that LTD can only be induced at synapses where virus is sufficiently expressed, likely leaving some synaptic connections unaffected by the stimulation protocol, which potentially limits behavioral impact. Moreover, evidence suggests that only small ensembles of neurons and synapses are involved in the encoding of a particular memory, giving credence to the idea that LTD induction within an entire brain region is not the best strategy for effecting behavioral change, whereas other approaches exist to specifically target cue- or contextually-active populations of neurons^21,22. Despite this, the protocol is effective at attenuating cue-motivated drug seeking, likely because the low-frequency optical stimulation limits LTD induction to synapses that have previously been potentiated by repeated cocaine-cue pairings⁴.

This protocol provides a significant advance to more commonly used optogenetic behavioral studies where the activity of neurons is activated or inhibited while the animal is performing the behavior in real-time^19,23. Instead, optogenetics is used here as a neuromodulatory tool to reverse cocaine-induced plasticity. An advantage of this method is that the optogenetic manipulation is independent of the behavioral test, such that potential confounding effects of optogenetics (e.g., local circuit effects, refractory period after light stimulation, antidromic stimulation effects, etc.²⁴ should not affect the results, thereby increasing confidence that the hypothesized neural mechanism is mediating changes in behavior. This method can therefore be utilized in a number of applications investigating how synaptic plasticity, particularly an increase in synaptic strength as occurs with LTP, relates to changes in behavior. Similar approaches might also be relevant to clinical application of neural stimulation technologies where abnormal connections in the brain driving dysfunctional behavior can be downregulated. Likewise, because the oChIEF viral construct is responsive to both low- and high-frequency stimulations, there are potential applications to these methods beyond the scope of the described experiments. For instance, optogenetically-induced LTP may be beneficial for reversing deficits in plasticity found in a wide range of neurodegenerative and neurodevelopmental disorders²⁵. Furthermore, bidirectional plasticity at MGN-LA synapses has also been directly linked to the regulation of behaviors relevant to fear-associated disorders⁸.

Modulating the specific neural circuits that support drug-motivated behaviors is essential for establishing long-term abstinence from drug use. This protocol utilizes novel advances in in vivo optogenetics to reverse plasticity of a precise neural circuit that is strengthened by repeated cocaine self-administration in the presence of environmental cues. The result of this specific neuromodulation is a decreased likelihood for subsequent cue re-exposures to trigger a cocaine-seeking response, which may have important implications for the development of future therapies for substance use disorders.

Materials

Name	Company	Catalog Number	Comments
0.9% Saline	Fisher Scientific	NC0291799
A.M.P.I. Stimulus Isolator	Iso-Flex
AAV5.hSyn.oChIEF.tdTomato	Duke Viral Vector Core (via Roger Tsien)	#268	See Lin et al., 2009; Nabavi et al., 2014
AAV5.hSyn.tdTomato (Control)	Duke Viral Vector Core Control		See Lin et al., 2009; Nabavi et al., 2014
Artificial Tears (Opthalmic Ointment)	Covetrus	70349
ATP Magnesium Salt	Fisher Scientific	A9187
Betadine	Butler Schein	38250
Calcium chloride	Fisher Scientific	C1016
Cesium chloride	Fisher Scientific	289329
Cesium hydroxide	Fisher Scientific	516988
Cesium methanesulfonate	Fisher Scientific	C1426
Cocaine HCl	NIDA Drug Supply Center	9041-001
Cryostat	Leica	CM1950
D-Glucose	Sigma-Aldrich	G8270
DMSO	Fisher Scientific	BP231-1
Dual-Channel Temperature Controller	Warner Instruments	TC-344C
EGTA	Fisher Scientific	E3889
Ethanol	University of Pittsburgh Chemistry Stockroom	200C5000
Ferrule Dust Caps	Thor Labs	CAPL	White plastic dust caps for 1.25 mm Ferrules
Ferrule Mating Sleeves	Doric Lenses	F210-3011	Sleeve_BR_1.25, Bronze, 1.25 mm ID
Ferrules	Precision Fiber Products	MM-FER2007C-2300	Ø1.25 mm Multimode LC/PC Ceramic ferrule, Ø230 μm hole size
Fiber Optic	Thor Labs	FP200URT	200 μm core multimode fiber (0.5 NA)
Fiber Optic Rotary Joint	Prizmatix	(Ordered from Amazon)	18 mm diameter, FC-FC connector for fiber
Fiber Stripping Tool	Thor Labs	T12S21
Fluoroshield with DAPI	Sigma-Aldrich	F6057
Gentamicin	Henry Schein	6913
GTP Sodium Salt	Fisher Scientific	G8877
Hamilton syringe	Hamilton	80085	10 μL volume, 26 gauge, 2 inch, point style 3
Heat Gun	Allied Electronics	972-6966	250 V, 750-800 °F
Heat-Curable Epoxy	Precision Fiber Products	PFP-353ND-8OZ
Heparin	Henry Schein	55737
HEPES	Sigma-Aldrich	H3375
Hydrochloric Acid	Fisher Scientific	219405490
Isoflurane	Henry Schein	29405
Ketamine HCl	Henry Schein	55853	Ketamine is a controlled substance and should be handled according to institutional guidelines
Lactated Ringer’s	Henry Schein	9846
Laser, driver, and laser-to-fiber coupler	OEM Laser Systems	BL-473-00100-CWM-SD-xx-LED-0	100 mW, 473-nm, diode-pumped solid-state laser (One option)
L-glutathione	Fisher Scientific	G4251
Lidocaine	Butler Schein	14583
Light Sensor	Thor Labs	PM100D	Compact energy meter console with digital display
Loctite instant adhesive	Grainger	5E207
Magnesium sulfate	Sigma-Aldrich	203726
Microelectrode Amplifier/Data Acquisition	Molecular Devices	MULTICLAMP700B / Digidata 1440A
Microinjector pump	Harvard Apparatus	70-4501	Dual syringe
Micromanipulator	Sutter Instruments	MPC-200/ROE-200
Microscope	Olympus	BX51WI	Upright microscope for electrophysiology
Microscope	Olympus	BX61VS	Epifluorescent slide-scanning microscope
N-methyl-D-glucamine	Sigma-Aldrich	M2004
Orthojet dental cement, liquid	Lang Dental	1504BLK	black
Orthojet dental cement, powder	Lang Dental	1530BLK	Contemporary powder, black
Paraformaldehyde	Sigma-Aldrich	P6148
Patch Cables	Thor Labs	FP200ERT	Multimode, FT030 Tubing
Picrotoxin	Fisher Scientific	AC131210010
Polishing Disc	Thor Labs	D50FC
Polishing Pad	Thor Labs	NRS913	9" x 13"
Polishing Paper	Thor Labs	LFG5P	5 μm grit
Polishing Paper	Thor Labs	LFG3P	3 μm grit
Polishing Paper	Thor Labs	LFG1P	1 μm grit
Polishing Paper	Thor Labs	LFG03P	0.3 μm grit
Potassium chloride	Sigma-Aldrich	P9333
Potassium hydroxide	Fisher Scientific	P5958
Potassium methanesulfonate	Fisher Scientific	83000
QX-314-Cl	Alomone Labs	Q-150
Rimadyl (Carprofen)	Henry Schein	24751
Self-Administration Chambers/Software	Med Associates	MED-NP5L-D1
Sodium bicarbonate	Sigma-Aldrich	S5761
Sodium chloride	Sigma-Aldrich	S7653
Sodium Hydroxide	Sigma-Aldrich	1064980500
Sodium L-Ascorbate	Sigma-Aldrich	A7631
Sodium Pentobarbital	Henry Schein	24352
Sodium phosphate	Sigma-Aldrich	S9638
Sodium phosphocreatine	Fisher Scientific	P7936
Sodium pyruvate	Sigma-Aldrich	P2256
Stainless steel machine screws	WW Grainger	6GB25	M2-0.40mm Machine Screw, Pan, Phillips, A2 Stainless Steel, Plain, 3 mm Length
Stereotaxic adapter for ferrules	Thor Labs	XCL
Stereotaxic Frame	Stoelting	51603
Sucrose	Sigma-Aldrich	S8501
Suture Thread	Fine Science Tools	18020-50	Silk thread; Size: 5/0, Diameter: 0.12 mm
TEA-Chloride	Fisher Scientific	T2265
Thiourea	Sigma-Aldrich	T8656
Vetbond Tissue Adhesive	Covetrus	001505
Vibratome	Leica	VT1200S
Xylazine	Butler Schein	33198