Animals: All behavior experiments were performed with wild type C57/Bl 6J mice (Jackson Laboratory, Bar Harbor, Maine, JAX Stock No: 000664). Fluorescent reporter mouse lines Drd1a-tdTomato mice (JAX Stock No: 016204) were used for labeling D1-MSNs in patch clamp electrophysiology experiments. All laboratory procedures were reviewed and approved by the Mayo Clinic Institutional Animal Care and Use Committee (IACUC) and conform to guidelines published by the NIH Guide for the Care and Use of Laboratory Animals (Department of Health and Human Services, NIH publication No. 86-23, revised 1985). Mice were housed in plastic cages under standard 12-hour light/dark cycle and conditions (21oC, humidity 45%) with ad libitum access to food and water. Animals were acclimated for at least one weeks before use. All efforts were made to minimize both the number of mice used and any discomfort that may be experienced.
Stereotactic surgery: A stereotaxic frame (Model 1900, David Kopf Instruments, Tujunga, CA) was used for implantation of custom- built stimulating electrodes and Tucker-Davis Technologies (TDT) Zif-Clip microwire arrays (TDT, Alachua, FL). Custom-built stimulating electrodes consisted of bipolar 50.8 µm diameter Teflon-insulated Pt/Ir wire (A-M systems 77600, Sequim, WA) separated 200 µm apart with exposed tips cut at a 45º angle. All surgical instruments, electrodes, and skull screws were sterilized in an autoclave or MetriCide28 (Metrex, Orange, CA, USA) cold sterilant. Analgesia was provided with buprenorphine HCL (0.05 mg/kg) preoperatively and three days postoperatively. Additionally, Ibuprofen was delivered in the animals’ drinking water two days prior to the surgery and no less than five days after for analgesia. Anaesthesia was induced in mice with isoflurane 4% followed by 1-2% isoflurane for maintenance. The depth of anesthesia was monitored using toe pinch and eye blink reflexes. A heating pad was used to maintain the subject's body temperature to 37.0±0.5 ºC throughout the duration of anesthesia. The surgical sites on mice heads were shaved and cleaned with betadine. Then, mice were placed in the stereotaxic apparatus and secured using ear bars. An incision of approximately 1.5 cm was made in the skin over the skull with a scalpel. Hydrogen peroxide was used to clean the skull surface. Sterile screws were used to provide a strong fixation of the head cap and serve as an electrical reference and ground. Screws were secured into 0.7mm holes drilled in the skull with a trephine drill bit. An additional craniotomy, approximately 2 mm x 2 mm in size, was created over the DMS to implant the Zif-Clip microwire array in a subset of animals. The dura was removed to expose brain tissue, which was irrigated with cool saline throughout the surgery. For all surgical procedures, additional burr holes were drilled at the desired coordinates described next to create a cranial window for unilateral implantation of custom-built stimulating electrodes per the Mouse Brain in Stereotaxic Coordinates 12. Stimulating electrodes were placed in the GPi (AP: -1.4 mm, ML: 1.6 mm, DV: −4.5 mm from Bregma) and ACC (AP: 0.50 mm, ML: 0.7 mm, DV: −1.85 mm from Bregma). Upon insertion and fixation of the stimulating electrodes to the skull using dental cement (Henry Schein, Melville, NY), a Zif-Clip microwire array was chronically implanted into the DMS (array center: AP: 0.70 mm, ML: 1.5 mm, DV: −2.75 mm from Bregma). Kwik-Sil silicone elastomer (World Precision Instruments, Sarasota, FL, USA) was put around the Zif-Clip microwire array and stimulating electrodes to maintain a seal between the brain and dental cement described in the next step. Finally, all components were secured with Metabond dental cement (Parkell, Edgewood, NY, USA). After the electrode implantation surgery, the animals were housed individually for one week to allow sufficient time for recovery. Animals were monitored twice daily for five days following surgery for signs of distress and infection at the surgical site. If signs of infection or distress were observed, topical polysporin antibacterial ointment was applied to the surgical site or a veterinarian was consulted for the appropriate methods of treatment. Animals were excluded from behavioral and electrophysiological analysis if they demonstrated weight loss greater than or equal to 20% body weight if they demonstrated inability to ambulate or were unable access food and water as a result of surgical procedures. Additionally, animals were excluded upon mechanical failure of dental cement headcap or chronically implanted electrodes.
Electrical stimulation parameters: Electrical stimulation was delivered through custom-built bipolar parallel chronically indwelling stimulating electrodes via a IZ2M-64 microstimulator programmed in the Synapse software (TDT, Alachua, FL). The ultra-low frequency spike-timing dependent plasticity stimulation (ULF-STDP) protocol consists of a charge-balanced, biphasic, cathodic-leading, 250 µA current stimulation with a pulse duration of 90µs delivered to the GPi and ACC with 18ms between GPi and ACC stimulation. GPi-ACC stimulation pairing was delivered at 1Hz for 10 minutes. ULF-STDP(+) is defined as postsynaptic GPi stimulation 18 ms prior to presynaptic ACC stimulation and ULF-STDP(-) is defined as presynaptic ACC stimulation 18 ms prior to postsynaptic GP stimulation. Current was delivered at a density of 941 μC/cm^2 such that Shannon safety criteria are satisfied to avoid tissue damage13.
Two-bottle choice alcohol self-administration AUD model: A standard two-bottle choice paradigm was used to assess alcohol consumption and preference62. Briefly, animals were implanted with stimulating electrodes as described in the Stereotactic Surgery section and allowed one week for recovery. Two bottles were presented daily in the animal’s home cages: one containing water and the other alcohol 63 at the onset of the dark cycle. Bottles consisted of a Hydropac valve (Avidity Science, Waterford, WI, USA) affixed the cap to a 50mL falcon conical tube (Corning, Corning, NY, USA). The position of the alcohol and water bottles was swapped daily to reduce confounds produced by location preference. For the first week, alcohol concentrations were increased every other day from 3% to 6% to 10% v/v. For the remaining duration of the experiment, mice were presented with one bottle of water and another bottle with 10% alcohol. For three alcohol exposure days prior to ULF-STDP stimulation, the animals were tethered to the stimulating system to allow habituation to the procedure. At day 25, animals were stimulated with the ULF-STDP protocol described in the methods section under Electrical stimulation parameters if mean and standard error of alcohol consumption (calculated in g/kg of body mass at be beginning of the night cycle) was less than 20% over a 3-day period after at least two weeks of alcohol exposure. The mass of each mouse, the mass of alcohol consumed, and the mass of water consumed was assessed daily using a scale. Additionally, an IR beam break sensor was placed in front of the sippers to measure the total time spent interacting with the sipper. Data were analyzed and plotted using Prism v9 software (GraphPad, San Diego, CA)
Limited-access AUD model: A limited access model of AUD capable of achieving blood alcohol content >1mg/ml in mice was used to assess acute alcohol consumption 64. Briefly, animals were implanted with stimulating and recording electrodes as described in the Stereotactic Surgery section and allowed one week for recovery. Mice were then provided with daily 2-hour access of 20% v/v ethanol in their home cages beginning three hours after the onset of the dark cycle for ten days. Immediately prior to providing alcohol on day ten, animals were stimulated with the ULF-STDP(+) protocol described in the methods section under Electrical stimulation parameters. For three alcohol exposure days prior to ULF-STDP(+) stimulation, the animals were tethered to the stimulating and recording system to allow habituation to the handling procedure. The mass of each mouse, the mass of alcohol consumed, and the mass of water consumed was assessed daily using a scale. Data were analyzed and plotted using Prism v9 software (GraphPad, San Diego, CA)
In vivo electrophysiology: Previously described standard methods were followed for in vivo electrophysiological recordings of MSNs within the DMS 59. Briefly, mice were surgically implanted with chronic indwelling Pt/Ir stimulating electrodes in the GPi and ACC along with a TDT 32-channel ZIF-Clip microwire array (TDT, Alachua, FL) placed in the DMS as described in the Stereotactic Surgery section. Mice were allowed one week to recover from surgery prior to recordings during unrestrained locomotion. Electrophysiological recordings were performed in three stages: alcohol naïve, following repeated alcohol exposure in the Limited Access AUD model, and following ULF-STDP stimulation. Multichannel extracellular recordings were performed at 50 kHz and band-pass filtered from 300 Hz to 5 kHz prior to spike thresholding and sorted using principal component analysis (PCA) via the TDT Synapse software (TDT, Alachua, FL). Mean event rates over 10-minute periods were compared immediately before and after ULF-STDP application using MATLAB (MathWorks, Inc., Natick, MA) Student’s unpaired t-test (p<0.05), Data were plotted using Prism v9 software (GraphPad, San Diego, CA). Additionally, extracellular LFPs band-pass filtered from 1-300 Hz were recorded from chronically implanted electrodes. Power spectral density of the GPi, DMS, and ACC was be analyzed following previously described methods65 across delta (1-4 Hz), theta (4-8 Hz), alpha (8-12 Hz), beta (13-30 Hz), and gamma (31-70 Hz) bands.
Ex vivo whole-cell patch clamp electrophysiology: We followed standard procedures for slice preparation and ex vivo whole cell electrophysiology to measure D1-MSN activity 10,66,67. Five week-old alcohol-naiive Drd1a-tdTomato mice (JAX Stock No: 016204) were assigned randomly to receive either ULF-STDP(+) in vivo or sham surgery while under 1-2% isoflurane anesthesisia 30 minutes prior to euthanasia for whole cell patch clamp electrophysiology experiments. Briefly, cells were clamped at -75 mV in the presence of lidocaine (0.7 mM) to record spontaneous mEPSCs. The mEPSC amplitude and frequency were compared between animals which received ULF-STDP(+) and stimulation naiive animals. Statistical significance (p<0.05) was determined via unpaired Student’s t-tests using MATLAB (MathWorks, Inc., Natick, MA). Data were plotted using Prism v9 software (GraphPad, San Diego, CA). D1-MSNs were identified by fluorescence microscopy in Drd1a-tdTomato mice.
Euthanasia and Histology: At the end of the experiments, animals were euthanized by a lethal overdose of Pentobarbital (i.p.), in accordance with the Panel on Euthanasia of the American Veterinary Medical Association (A.V.M.A.). Mice were transcardially perfused with 4% paraformaldehyde in phosphate buffered saline (PBS). Mice brains were removed and dehydrated in 30% sucrose prior to flash-freezing with dry ice and slicing into 40 μm-thick coronal sections using a sliding microtome (Leica Biosystems, Buffalo Grove, IL). Slices were mounted and counterstained with DAPI (Vector Laboratories, Burlingame, CA, USA) and visualized using bright field and fluorescence microscopy (Carl Zeiss LSM780 or Nikon Eclipse FN1) to verify electrode location. Animals were included in behavioral and electrophysiological analysis only if at least one of each bipolar electrode tips were located in both the ACC and GPi as defined by atlas boundaries.12 Animals with electrodes absent in the ACC and GPi were excluded from all behavioral and electrophysiological analysis.
Rigor and reproducibility: The observed effect size of ULF-STDP in reducing alcohol consumption was dCohen = 0.603, in which 75% of the mice (15 out of 20 mice) developed high alcohol preference in the two-bottle choice model of AUD. Animal numbers for each cohort in each behavioral experiment were determined based on an 80% statistical power, 95% confidence interval, and effect size of 0.603. Experimenters were blinded to stimulation parameters during allocation, data collection, and data analysis. Data collected were tested for normality with the Kolmogorov-Smirnov test. Comparisons of non-normal distributions were performed using nonparametric tests, substituting the Mann-Whitney U-test in place of the unpaired Student’s t-test.