Protocol for closed-loop optogenetic manipulation of cortical activity following hippocampal sharp-wave ripples in freely behaving rats

Summary

Temporally and spatially selective closed-loop perturbation of brain activity can be used to reveal the contribution of spontaneous neural events to behavior. Here, we present a protocol for optogenetic inhibition of the medial prefrontal cortex following hippocampal sharp-wave ripple events in freely behaving rats. We describe procedures for injecting a viral vector bilaterally, constructing and implanting an assembly of optical fibers and tetrodes, and performing closed-loop perturbations in rats.

For complete details on the use and execution of this protocol, please refer to den Bakker et al.¹

Graphical abstract

Highlights

• •Instructions for optogenetic inhibition of the prefrontal cortex bilaterally
• •Instructions for closed-loop manipulations following sharp-wave ripples in the hippocampus
• •Procedures for viral vector injection
• •Guidance on constructing and implanting an assembly of optical fibers and tetrodes

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Temporally and spatially selective closed-loop perturbation of brain activity can be used to reveal the contribution of spontaneous neural events to behavior. Here, we present a protocol for optogenetic inhibition of the medial prefrontal cortex following hippocampal sharp-wave ripple events in freely behaving rats. We describe procedures for injecting a viral vector bilaterally, constructing and implanting an assembly of optical fibers and tetrodes, and performing closed-loop perturbations in rats.

Before you begin

This protocol was designed to detect sharp-wave ripples in the hippocampus and use optogenetic silencing in the medial prefrontal cortex (mPFC) following each detection. Other applications are also possible, following slight modifications to the protocol. Examples of such modifications include targeting a different brain region other than the mPFC for optogenetic feedback, using optogenetic stimulation instead of inhibition, or using a different trigger than sharp-wave ripples for the initiation of optogenetic feedback.

Institutional permissions

The experimental procedures were carried out conform the protocols approved by KU Leuven animal ethics committee (P175/2020) in accordance with the European Council Directive, 2010/63/EU, regulation (EU)2019/1010. Anyone wishing to use this protocol in studies involving rodents must obtain approval from the relevant institutions.

Preparation of the implant

Timing: 5 h

• 1.3D-print all parts for construction of the implant (Figure 1) and a mold for creating micro-drive assembly (Figure 2, item I). Use a stereolithography (SLA) 3D printer with a resin that produces prints with high strength and resolution (e.g. Gray Resin, Formlabs, Cambridge, MA).

Note: For all parts, 3D print files in STL format are available in the accompanying Open Science Framework repository that is located at https://osf.io/3n9p8. Both a commercial 3D-printing service (Materialise, Leuven, Belgium) and an in-house Form 3 printer (Formlabs) have been used successfully to print the implant parts.

Note: When using a Form 3 printer, use gray resin and the highest accuracy setting (25 μm). After printing, wash the parts in IPA for 10 min (Form Wash; Formlabs), and post-cure the parts for 30 min at 60 degrees Celsius (Form Cure; Formlabs). Carefully remove the supporting scaffold and ensure that the main body and the base fit snugly.

• 2.Gather an 8-tetrode electrode interface board (EIB; Figure 2, item IV; EIB-36, Neuralynx, Bozeman, MT) that connects tetrodes to pre-amplifiers and LEDs to LED drivers, two tapered optical fibers (Figure 2, item XVI; Optogenix, Arnesano LE, Italy), two custom-made angled LEDs (595 nm; Figure 2, item XIV; Doric lenses, Canada) and an 1-mm-thick laser-cut metal shuttle (Figure 2, item XV; design file available at https://osf.io/3n9p8).

Note: Expect a lead-time of up to 6 weeks for the custom-made angled LEDs.

• 3.Gather screws, tubing and other small components (Figure 2), and collect the tools for fabrication of the implant (Figure 3).

Figure 1.

Exploded view of the main implant parts

The implant consists of a main body and a base that are protected by a protective cone and removable cap. A separate one-time-use skull connector enables reuse of the implant. The main body contains holes for holding eight independently movable micro-drives for tetrodes (right side) and one micro-drive for the optical fiber/LED assembly (left side). The base of the implant features two collector cannulas for guiding the tetrodes and optical fibers. Various holes intended for screws and tubing are labeled in the parts, together with the size drill bit for clearing the hole or the size tap for creating a thread.

Figure 2.

Implant components

I. Micro-drive mold. II. Custom micro-drive screw. III. Tetrode collector cannula (17Ga hypodermic tubing; 304H17TW, MicroGroup, Medway, MA), cut to 8-10 mm length. IV. Electrode interface board (EIB). V. 6 mil stainless steel wires (304B0006XRND, MicroGroup). VI. 5 mil stainless steel wires (304B0005XRND, MicroGroup). VII. 30Ga stainless steel tubing (304H30RW, length: 5.5 cm; MicroGroup, Medway, MA). VIII. Tetrode wire (R0800; Sandvik, Stockholm, Sweden). IX. Pin receptacle (1212-01427-87-3010-ND; DigiKey, Enschede, Netherlands). X. M1.2 screw (3 mm long; DIN84M1.2x3I; Micro-Modele, Strasbourg, France). XI. M1.2 screw (6 mm long; DIN84M1.2x6I; Micro-Modele). XII. 23Ga micro-drive cannula (304H23RW, length: 0.5518"; MicroGroup). XIII. 20Ga sleeve (304H20TW, length: 0.1575"; MicroGroup). XIV. Angled LED-optic fiber assembly with wires. XV. Metal shuttle for the optical/LED assembly. XVI. Tapered optical fibers (NA 0.66, core Ø 0.2 mm, active length 2.5 mm, implant length 20 mm). XVII. Tetrode guide cannula (polyimide tube; 141-0001; Nordson Medical, Abbeytown, Ireland). XVIII. Gold pins (small; Neuralynx).

Figure 3.

Implant fabrication tools

I. Blue light source (Radii-cal, SDI Limited, Bayswater, Australia). II. File (Key files, MXID, Germany). III. Custom micro-drive screwdriver. IV. Light curable cement (wave A2 syringes, 7500012, Dental Elite, Wambrechies, France). V. Tweezers (2ACFR.SA, IDEAL-TEK, Farnell). VI. Forceps (11252-00, Fine Science Tools, Heidelberg, Germany). VII. Scalper handle (10003-12) and scalpel blade (#11, 10011-00, Fine Science Tools). VIII. Needle-nose pliers (378M, Xcelite, Weller, Germany). IX. Pin vise hand drill (19057, SHAVIV, Israel). X. Caliper (D03195, Farnell, Germany). XI. M1.2 m tap (DP2061105, CWS TOOLS, Belgium). XII. Ø 0.8 mm, Ø 0.95 mm, Ø 1.4 mm drill bits (Twist Drill Bits, Mesee, US). XIII. Diamond cutting wheel (2615S545JB, Dremel, Farnell, Belgium). XIV. Clamp (JR. 201, Panavise, China). XV. High-viscosity liquid glue (1501202, Sekunden Kleber, R&G GmbH, Germany). XVI. Vise (FMS75, Proxxon, Luxemburg). XVII. Microscope (25 GXMXTL3TV7 Long Reach Stereo Microscope, Trinocular; GT-Vision, UK). XVIII. Cordless drill (Bosch Hammer Drill, Germany). XIX. Rotary cutting tool (Model 4000, Dremel, Germany).

Preparation of surgical instruments

Timing: 20 min

• 4.Gather all the reagents (isoflurane, lidocaine cream, Vaseline, 70% ethanol, mineral oil, saline, Baytril, eye ointment, Metacam, Vetbond (tissue adhesive), bone wax, silicone grease and iso-Betadine; see Figure 4) and equipment (shaver, 3 small beakers, ear bars, scalpel holder and scalpel, retractor, small dull forceps, surgical spoon, pointy forceps, several syringes and needles, alcohol swabs and cotton tips, #007 and #009 drill bits, bone screw clamp and bone screws, dental cement (Wave A2; SDI, Victoria, Australia) and UV light, surgical suture, clamp and scissors, surgical drapes and a holder for the implant that fits the stereotaxic frame (Model 51903, Parkland Scientific, US; see Figure 5).
• 5.
  Prepare the bone screws by cutting off the sharp tips and filing down the sharp edges.

  • a.
    For electrical ground, solder a 5 cm long insulated stainless steel wire (Cat. No. 791400, A-M Systems, Sequim, WA) to one bone screw.
  • b.
    Solder a connector pin (ED1121-ND; DigiKey, Enschede, Netherlands) to the other end of the wire that fits the pin receptacle on the implant.

Note: strip the insulation at the ends of the ground wire before soldering.

• 6.Sterilize all autoclavable equipment (surgical instruments, bone wax, cotton tips and drill bits).
• 7.Soak the bottom of the implant (the part that will be in contact with the animal; the tetrode collector cannula and the bottom of the skull connector) and at least 10 bone screws in 70% ethanol at least 20 min prior to the surgery.
• 8.Clean all table surfaces that will be used during the surgery with 70% ethanol.
• 9.
  Prepare syringes:

  • a.
    Prepare one syringe with antibacterial solution (Baytril solution).
  • b.
    Prepare two syringes with the anti-inflammatory and pain-relieving drug Metacam (pre-operative and post-operative solutions; see materials and equipment).
• 10.
  Immediately before the surgery:

  • a.
    Fill the tank of the isoflurane vaporizer.
  • b.
    Prepare the table with the stereotaxic frame and heating pad.
  • c.
    Cover surfaces with surgical covers.
  • d.
    Turn on the hot bead sterilizer (to be used during the surgery after each use of the surgical instruments).

Note: When working with isoflurane, ensure proper ventilation in the room and use the appropriate personal protective equipment.

Figure 4.

Reagents needed for performing the surgery

I. Isoflurane (Iso-Vet, 1000 mg/g, Piramal Critical Care B.V., Voorschoten, the Netherlands); II. Lidocaine crème (Xylocaine 5%, 0137398, Aspen, Durban, South Africa); III. Vaseline pure petroleum jelly (Unilever, London, UK); IV. Ethanol 70% (Thermo Fisher, Waltham, MA); V. Mineral oil (330760-1L, Sigma-Aldrich, St. Louis, MO); VI. Saline (Physio Sterop 900 mg/100 ml, Sterop, Brussels Belgium); VII. Baytril 100 mg/ml (enrofloxacine, BE-V140314, Bayer, Leverkusen, Germany); VIII. Duratears eye ointment (BE 160377, Alcon, Geneva, Switzerland); IX. Metacam 2 mg/ml (meloxicam, REG NL 105004, Boehringer Ingelheim, Ingelheim am Rhein, Germany); X. Vetbond (1469SB, 3M, Saint Paul, MN); XI. White Bone Wax 2.5g. (SS903, Surgical Specialties, Braintree, MA); XII. Silicone Lubricant (29051-45, Fine Science Tools, Heidelberg, Germany); XIII. Iso-Betadine (Mylan, Canonsburg, PA).

Figure 5.

Equipment needed for performing the surgery

I. Shaver; II. Three small beakers (20 ml, Pyrex, Rosemont, Illinois); III. Ear bars (505387, World Precision Instruments, Sarasota, FL); IV. Scalper handle (10003-12) and scalpel blade (#11, 10011-00, Fine Science Tools, Heidelberg, Germany); V. Retractor (17009-08, Fine Science Tools); VI. Dull forceps (Iris Forceps, 11064-07, Fine Science Tools); VII. Surgical spatula (10094-13, Fine Science Tools); VIII. Pointy forceps (Dumont #5SF Forceps, 11252-00, Fine Science Tools); IX. Syringes (1 ml) and needles (27 gauge); X. Alcohol prep pads (Kendall Webcol 5110, Covidien, Dublin, Ireland) and cotton tips (Sterile Absorption Spears, 18105-01, Fine Science Tools); XI. Meisinger drill bits #007 (19008-07) and #009 (19008-09; Fine Science Tools, Heidelberg, Germany); XII. Bone screw clamp (12003-15) and self-tapping bone screws (1.19 x 4.8 mm, 19010-00, Fine Science Tools, Heidelberg, Germany); XIII. Dental cement (wave A2 syringes, 7500012, Dental Elite, Wambrechies, France) and UV light (Radii-cal, SDI Limited, Bayswater, Australia); XIV. Surgical suture (Sofsilk Silk Suture, Size 4-0, CVSS683G, Covidien, Dublin, Ireland), clamp (Halsey Needle Holder, 12001-13, Fine Science Tools) and scissors (14060-11, Fine Science Tools); XV. Surgical drapes (H2 504375, MediWare); XVI. Implant holder constructed from a standard stereotaxic arm and dummy Neuralynx headstage.

Viral vector injection setup

Timing: 20 min

• 11.Aliquot the viral vector into Eppendorf tubes in samples of 6 μL. Label the samples and store in a freezer at −80 degrees Celsius until use.

Note: Be aware of biosafety when aliquoting viral vectors. Use the appropriate personal protective equipment and ensure proper ventilation by working in a laminar flow cabinet.

• 12.In preparation for surgery, gather the items and equipment needed for the viral vector injections (Parafilm, 20 μL pipette tips, 20 μL pipette, glass capillaries, stereotaxic injector and Hamilton syringe; see Figure 6A).
• 13.
  Prepare glass capillaries:

  • a.
    Pull the glass capillaries with a micropipette puller (P-2000, Sutter instrument, Novato, CA) with the heat set to 500 (representing the output power of the laser), the velocity set to 35 (one unit represents a change of one or more millivolts of transducer output, the output velocity depends on the viscosity and temperature of the glass) and the force of the pull set to 100 (each unit representing a change of 4 milliamps of current through the solenoid).
  • b.
    Cut the tip of the pulled glass capillary using a small pointy forceps under a microscope to create an opening of 30 μm (see Figures 6B and 6C).
• 14.Attach the glass pipette to a Hamilton syringe and backfill the syringe and pipette by removing the plunger and loading mineral oil from the back using a second syringe.
• 15.Use the stereotaxic pump to eject at least 5 μL of mineral oil out of the Hamilton syringe, flushing out all air bubbles and ensuring that 5 μL of the viral vector can be front-filled during the surgery.

Figure 6.

Equipment needed for the viral vector injections

(A) I. Parafilm (Amcor, Victoria, Australia); II. 20 μl pipette tips (ep Dualfilter T.I.P.S., 2-20 μl, Eppendorf, Hamburg, Germany); III. Single Channel Pipette (2-20 μl, , Eppendorf); IV. Glass capillaries (TW100-4, World Precision Instruments, Sarasota, FL); V. Stereotaxic injector (QSI Quintessential Stereotaxic Injector 53311, Stoelting, Wood Dale, Illinois); VI. Hamilton syringe (10 μL, model 701, 7635-01, Hamilton company, Reno, NV).

(B) Micropipette puller (P-2000, Sutter Instrument Company, Novato, CA).

(C) An example of a pulled glass capillary with the tip cut to an opening of ∼30 μm in diameter.

Data acquisition and online signal processing setup

Timing: 1 day

• 16.
  Set up the hardware for data acquisition and closed-loop optogenetic experiments (see Figures 7 and 8).

  • a.
    Connect the data output ports at the back of the neural data acquisition system (Digital Lynx SX, Neuralynx, Bozeman, MT) to network cards in two separate computers using glass fiber-optical network cables for fast throughput of data.

    Note: One computer runs Cheetah software (Neuralynx, Bozeman, MT) and is used for neural data acquisition and storage (‘acquisition computer’). The second computer runs Falcon software for online sharp-wave ripple detection (‘Falcon computer’).

  • b.
    Enable duplication of the data stream on the second output port of the Digilynx acquisition system by adding the following command to the Cheetah configuration file (.CFG) on the acquisition computer:

    -SendLynxSXCommand AcqSystem1

    -SetDataLink2Mode 1

    Note: This command needs to be added every time you save a new Cheetah configuration file.

  • c.
    Configure the network card in the Falcon computer with the exact same IP address and MAC address as the network card in the acquisition computer:

    • i.
      Open a command prompt on the acquisition computer and run command

      > ipconfig /all
      to find the values for the MAC address (“Physical Address”), the IPv4 Address, and the Subnet Mask for the network adapter connected to the Digilynx system.
    • ii.
      Find the interface name of your network card in the Falcon computer by running the following command in a terminal window.

      > sudo lshw -C network -short
    • iii.
      Run the following commands in a terminal window on the Falcon computer to configure the network card.

      > sudo macchanger --mac=00:15:77:4F:9B:E1 enp1s0f1

      > sudo ifconfig enp1s0f1 192.168.3.100 netmask 255.255.255.0

      Note: Make sure to replace the example MAC address (00:15:77:4F:9B:E1), IP address (192.168.3.100) and Subnet Mask (255.255.255.0) with the values found in step i. Replace enp1s0f1 in the example above with the identifier of the network interface card in your system (found in step ii).

  • d.
    Wire an Arduino Uno for the generation of TTL pulses that signal the detection of sharp-wave ripples and trigger LED activation (Figure 7).

    • i.
      Connect the Arduino Uno USB jack to a USB port on the Falcon computer (Figure 8A).
    • ii.
      Connect a wire from pin number 2 on the Arduino Uno to pin 0 of TTL I/O Port 0 on the Digilynx acquisition system (Figure 8B).

      Note: TTL pulses on this wire signal the detection of a sharp-wave ripple and their timestamps will be recorded by the Digilynx system.

    • iii.
      Connect a wire from pin number 6 on the Arduino Uno to pin 1 of TTL I/O Port 0 on the Digilynx acquisition system (Figure 8B).

      Note: TTL pulses on this wire signal LED activation and their timestamps will be recorded by the Digilynx system.

    • iv.
      Connect the ground on the Arduino Uno to the ground pins next to pins 0 and 1 of TTL I/O Port 0 on the Digilynx acquisition system (Figure 8B).
    • v.
      Connect two additional wires from pin number 6 on the Arduino Uno to the positive input ports TRIGIN1+ and TRIGIN3+ on the LED driver.

      Note: Selecting two non-adjacent ports on the LED driver reduces the risk of accidentally creating a short between the two wires (Figure 8C).

    • vi.
      Connect two additional wires from the ground on the Arduino Uno to the negative input ports TRIGIN1- and TRIGIN3- on the LED driver (Figure 8C).
  • e.
    Wire the LED driver for optogenetic LED control

    • i.
      Connect the LED driver to a USB port on the acquisition computer for configuring the light pulse durations.
    • ii.
      Connect the LED1^+/− and LED3^+/− ports on the LED driver to the Global Reference Bus connector on the Digilynx acquisition system, using a 10-Pin Stimulus Cable from Neuralynx (see Figure 8D).
  • f.
    When you are ready to start data acquisition, connect a tether and headstage from the Headstage input on the Digilynx acquisition system to the electrode interface board on the implant of the animal.

    Note: This tether is used both for the acquisition of neural data and the optogenetic LED control (see Figure 7).

    Note: Due to the weight of the Neuralynx tether, a counter-weight system in the room may be needed to perform experiments with freely moving animals. To create a counter-weight system in the room, span fishing wire tightly between the four corners at the top of the room (fixed structure). Use another fishing wire with equal weights at both sides hung from two pulleys on this fixed structure, to create a wire that can move in the X and Y dimensions through the room. Attach the Neuralynx tether to this moving wire to ensure that its weight is lifted up, while having full mobility to reach any area of the room.

• 17.Download and install the control software for the LED driver on the acquisition computer (https://www.mightexsystems.com/led-controller-download/).
• 18.Upload the following sketch to your Arduino Uno.

/∗ This sketch responds to serial commands by generating TTL pulses on selected output pins. One output pin signals a detection event, another output signals a stimulation.

The following single-character commands are supported:

o: detection and stimulation

d: stimulation only

r: detection only

∗/

int detectionPin = 2;

int stimulationPin = 6;

int allPins[] = {detectionPin, stimulationPin};

int numPins = sizeof(allPins) / sizeof(allPins[0]);

// serial port

int BAUDRATE = 9600; // bits per second

unsigned long ttlDuration = 100;

char buffer = 'n';

void setup() {

 // declare both pins as output

 pinMode(detectionPin, OUTPUT);

 pinMode(stimulationPin, OUTPUT);

  // connect to the serial port at BAUDRATE bits per second

  Serial.begin(BAUDRATE);

  Serial.setTimeout(1); // 1 ms time-out for serial readout

}

void loop() {

 Serial.readBytes(&buffer, 1);

 if (buffer=='o') { // detection + stimulation

  ttl(allPins, numPins, ttlDuration);

 } else if (buffer=='d'){ // stimulation

  ttl(stimulationPin, ttlDuration);

 } else if ( buffer=='r') { // detection

  ttl(detectionPin, ttlDuration);

 }

 buffer = 'n';

}

void ttl(int pins[], int npins, unsigned long duration) {

 for (int pin=0; pin<npins; ++pin) {

 digitalWrite(pins[pin], HIGH);

 }

 delayMicroseconds(duration);

 for (int pin=0; pin<npins; ++pin) {

 digitalWrite(pins[pin], LOW);

 }

}

void ttl(int pin, unsigned long duration) {

 digitalWrite(pin, HIGH);

 delayMicroseconds(duration);

 digitalWrite(pin, LOW);

 }

• 19.
  Set up the Falcon computer.

  • a.
    Install Falcon² and dependent packages on the Falcon computer (see https://falcon-core.readthedocs.io/en/latest/index.html).
  • b.
    Complete installation by executing the following commands in a terminal window.

    Note: During the installation, you will create a Conda environment for Falcon in order to run the necessary packages (for more information on Conda environments, see https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html).

    > conda config --append channels KloostermanLab

    > conda env create -n falcon cmake>=3.11

    > conda activate falcon

    > conda install pyyaml pyqt=5 pkgconfig appdirs fklab-cmake-gui pyzmq zeromq

    > fklab-build --gui false --path https://bitbucket.org/kloostermannerflab/falcon-core.git ∖

       --version 1.3.0 ∖

       --build_options DCMAKE_INSTALL_PREFIX="$HOME/opt/falcon-core"
  • c.
    Install visualization and control clients for Falcon by cloning or downloading the Git repository of the Falcon client project from https://bitbucket.org/kloostermannerflab/falcon-client/src/master/. Install falcon-client by executing the following commands in a terminal window in the falcon_client folder.

    > python setup.py develop

    > pip install -e.

    Note: Falcon software requires a workstation with a Linux operating system. It is recommended that the workstation has 4–8 available CPU cores for faster parallel processing of the signals in Falcon.

  • d.
    Define initial ripple detection parameters in a yaml file (called a user-side Falcon graph; for more information see https://falcon-core.readthedocs.io/en/latest/manual/graphs.html). Use the example shown in the snippet below, but change the following parameters.

    • i.
      Set the address option of the source to the IP address and the port of the network interface card on the Falcon computer (see preparation 16.c.iii).
    • ii.
      List the Digilynx recording channel(s) in the source channelmap for the detection of sharp-wave ripples in the hippocampus (hc) and for control detections in the cortex (cx). Channels are numbered with an index starting at 0.
    • iii.
      Verify the serial port on the Falcon computer to which the Arduino Uno is connected.
    • iv.
      Verify the lock-out periods that are implemented to avoid duplicate detections and detection of stimulation artefacts (Figure 9).

      Falcon:

       version: 1.0 # minimum required Falcon version for this graph

      # the full specification of the graph with all parameters

      graph: graphs://neuralynx/ripple_detection.yaml

      # user-specified options that overwrite default values

      options:

       source:

        options:

         # IP address and port that receives Digilynx data stream

         address: 192.168.3.100

         port: 26090

         # the recording channel(s) used for sharp-wave ripple detection

         # in the hippocampus (hc) and cortex (cx). Cortical channels are

         # used to control for spurious detections. Channel index starts at 0.

         channelmap:

          hp: [0,1,2,3]

          cx: [4,5,6,7]

       HIPPOCAMPUS_detector:

        options:

         # multiplier for computing detection threshold

         # the threshold is computed as mean + threshold dev x mean absolute deviation

         threshold dev: 16

         # integration time in seconds for estimating signal statistics

         smooth time: 3

         # lock-out period in milliseconds following each detection

         # to avoid duplicate detections

         analysis lockout time: 50

       CORTEX_detector:

        options:

         threshold dev: 12

         smooth time: 3

         analysis lockout time: 50

       ttl_output:

        options:

         # serial port to which the Arduino Uno is connected

         port address: /dev/ttyACM1

         stimulation_trigger:

         options:

         # enable detection-only mode (set to false to enable stimulation trigger)

         detection only mode: false

         # enable delayed stimulation trigger

         delayed mode: true

         # range of delays of stimulation trigger following detection (in milliseconds)

         # only used if delayed mode is true

         delayed range: [400, 450]

         event trigger lockout time :

          # lock-out period in milliseconds following each detection

          # that leads to a stimulation

          period: 210

          detection: true # apply lock-out at the detection time

          stimulation: false # and not at the stimulation time

          analysis lockout time:

          # lock-out period in milliseconds at the start and end of each

          # stimulation, to avoid detection of stimulation induced artefacts

          period: 50

          # start time of lock-out periods in milliseconds relative

          # to the stimulation start time.

          starting time: [0, 190]

      Note: A bug reported in the original research article¹ regarding the implementation of the event-triggered lock-out time has been fixed and should not be a problem for future users (see https://bitbucket.org/kloostermannerflab/falcon-fklab-extensions/commits/f4a0ec3901cacb4d8a98c84092d67b886174556f).

Figure 7.

Schematic of hardware set-up for closed-loop optogenetic experiments

(A) Tetrodes implanted in the brain are connected to the EIB on the implant, which is connected to a headstage with a tether at the start of each experimental session.

(B) Analog input arrives at the Digilynx acquisition system, where the signals are digitized.

(C) The digitized signals are sent to two processing computers using two fiber optical cables for fast throughput of data.

(D) The ‘acquisition computer’ is running Cheetah software for data acquisition and storage.

(E) The ‘Falcon computer’ is running Falcon software for sharp-wave ripple detection.

(F) Each time a sharp-wave ripple is detected and each time an optogenetic manipulation is desired, a command is sent to an Arduino Uno attached to the Falcon computer with a USB cable.

(G) Each detection and optogenetic manipulation is registered on the acquisition computer through a TTL pulse in the digital input port on the Digilynx acquisition system.

(H) Each time an optogenetic manipulation is desired, a TTL pulse is sent from the Arduino Uno to the LED driver.

(I) The LED driver produces a current for pre-set duration, which passes through the Digilynx acquisition system.

(J) The current passes through the tether and headstage, and activates the LEDs that are coupled to tapered optical fibers implanted in the brain, causing the timed optogenetic manipulation of brain activity.

Figure 8.

Wiring diagrams for hardware set-up

(A) Connect an Arduino Uno to a USB port on the Falcon computer.

(B) Connect the detection pin and stimulation pin on the Arduino Uno to the TTL I/O port on the Digilynx acquisition system, for keeping track of the time stamps of each detection and optogenetic manipulation.

(C) Connect the detection pin to the trigger ports on the LED driver.

(D) Connect the LED ports on the LED driver to the Global Reference Bus connector on the Digilynx acquisition system for controlling the LEDs.

Figure 9.

Detection lock-out periods

Illustration of the enforced lock-out periods as applied to the on-time (top) and delayed (bottom) inhibition conditions. Three different lock-out periods are enforced to reduce the number of spurious detections. The first is an analysis lock-out time (set to 50 ms) that is enforced after each detection to prevent another detection (of the same sharp-wave ripple) from occurring immediately. The second is an event-triggered lock-out time (set to 210 ms) that is enforced after each detection that leads to an optogenetic inhibition. The final lock-out period is designed to avoid spurious detection of artefacts that occur in the signal when the LEDs are switched on and off. This 50-ms long lock-out period is enforced at the start and end of each optogenetic manipulation.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Bacterial and virus strains
pAAV-CamKII-ArchT-GFP (titer: ≥ 1.9 × 1013 genome copies per mL)	Addgene	Cat# 99039-AAV5
Deposited data
3D design files, analysis code, and example data	Open Science Framework	https://osf.io/3n9p8
Experimental models: Organisms/strains
Long Evans rats (adult male rats weighing between 300 and 350 g, corresponding to 60–75 days)	Janvier, France	N/A
Software and algorithms
Falcon real-time software	Ciliberti & Kloosterman2	https://bitbucket.org/kloostermannerflab/falcon-core
Cheetah data acquisition software (Cheetah 6.4.1)	Neuralynx	https://neuralynx.fh-co.com/research-software/cheetah/
Mightex LED controller software	Mightex	https://www.mightexsystems.com/led-controller-download/
Other
Lidocaine crème (Xylocaine 5%)	Aspen	0137398
Mineral oil	Sigma-Aldrich	330760-1L
Isoflurane (Iso-Vet, 1000 mg/g)	Piramal Critical Care	QN01AB06
Vaseline pure petroleum jelly	Unilever	N/A
Ethanol (70%)	Thermo Fisher Scientific	T10290
Saline (Physio Sterop 900 mg/100 mL)	Sterop	0862334
White bone wax (2.5 g)	Surgical Specialties	SS903
Iso-Betadine Dermicum 10%	Mylan	3305752
Baytril 100 mg/mL (enrofloxacin)	Bayer	BE-V140314
Duratears eye ointment	Alcon	BE 160377
Metacam 2 mg/ml (meloxicam)	Boehringer Ingelheim	REG NL 105004
Vetbond	3M	1469SB
Silicone lubricant	Fine Science Tools	29051-45
Forceps	Fine Science Tools	11252-00
Scalper handle	Fine Science Tools	10003-12
Scalpel blade (#11)	Fine Science Tools	10011-00
Retractor	Fine Science Tools	17009-08
Iris forceps	Fine Science Tools	11064-07
Dumont #5SF forceps	Fine Science Tools	11252-00
Surgical spatula	Fine Science Tools	10094-13
Sterile Absorption Spears	Fine Science Tools	18105-01
Halsey needle holder	Fine Science Tools	12001-13
Surgical scissors	Fine Science Tools	14060-11
#007 Meisinger drill bits	Fine Science Tools	19008-07
#009 Meisinger drill bits	Fine Science Tools	19008-09
Self-tapping bone screws (1.19 × 4.8 mm)	Fine Science Tools	19010-00
Bone screw clamp	Fine Science Tools	12003-15
Hot bead sterilizer	Fine Science Tools	18000-45
Stereotaxic frame	Parkland Scientific	51903
Ear bars	World Precision Instruments	505387
Alcohol prep pads	Covidien	Kendall Webcol 5110
Sofsilk silk suture (size 4-0)	Covidien	CVSS683G
Surgical drapes	MediWare	H2 504375
Beakers (20 mL)	Pyrex	70000-20
Spongostan Dental	Ethicon	1706030
QSI Quintessential stereotaxic injector	Stoelting	53311
Hamilton syringe (10 μL, model 701)	Hamilton company	7635-01
Micropipette puller	Sutter Instrument Company	P-2000
Glass capillaries	World Precision Instruments	TW100-4
Single-channel pipette (2–20 μL)	Eppendorf	41121500
ep Dualfilter T.I.P.S., 2–20 μL	Eppendorf	0030078586
Parafilm	Amcor	HS234526B
17Ga hypodermic tubing	MicroGroup	304H17TW
6 mil stainless steel wires	MicroGroup	304B0006XRND
5 mil stainless steel wires	MicroGroup	304B0005XRND
30Ga stainless steel tubing (length: 5.5 cm)	MicroGroup	304H30RW
23Ga micro-drive cannula (length: 0.5518″)	MicroGroup	304H23RW
20Ga sleeve (length: 0.1575″)	MicroGroup	304H20TW
Digital Lynx SX data acquisition system	Neuralynx	31-0605-0140
8-tetrode electrode interface board (EIB-36)	Neuralynx	31-0603-0117
Gold pins (small)	Neuralynx	31-0603-0102
NanoZ adaptor (Omnetics NZ-EIB-36)	Neuralynx	31-0602-0126
Gold-plating solution	Neuralynx	31-0611-0100
Custom-made angled optical fiber coupled LEDs, 595 nm	Doric Lenses	N/A
Tapered lambda optical fiber, 0.66 NA, 2.5 mm active length, 20 mm total length	Optogenix	N/A
Dental cement (wave A2 syringes)	Dental Elite	7500012
Radii-cal blue light	SDI	5600102
Tetrode wire	Sandvik	R0800
Tetrode guide cannula (polyimide tubing)	Nordson Medical	141-0001
Pin receptacle	DigiKey	1212-01427-87-3010-ND
Connector pin	DigiKey	ED1121-ND
Stainless steel wire	A-M Systems	791400
M1.2 screw (length: 3 mm)	Micro-Modele	DIN84M1.2x3I
M1.2 screw (length: 6 mm)	Micro-Modele	DIN84M1.2x6I
Key files	MXID	300107
Tweezers	Ideal-tek	2ACFR.SA
Needle-nose pliers	Xcelite	378M
Pin vise hand drill	SHAVIV	19057
Caliper	Farnell	D03195
Diamond cutting wheel	Farnell	2615S545JB
M1.2 tap	CWS Tools	DP2061105
Twist drill bits (Ø 0.8 mm, Ø 0.95 mm, Ø 1.4 mm)	Mesee	B08Y67YRYG
High-viscosity liquid glue	R&G GmbH	1501202
Clamp	PanaVise	JR. 201
Vise	Proxxon	FMS75
Long Reach stereo microscope	GT Vision	25 GXMXTL3TV7
Rotary cutting tool	Dremel	Model 4000
Cordless hammer drill	Bosch	06039D4002
Formlabs Form 3+ SLA 3D printer	Formlabs	N/A
Gray resin	Formlabs	RS-F2-GPGR-04
Fiber-optical network cables	StarTech	FIBLCSC1
Arduino Uno Rev3	Arduino	A000066
Mightex LED controller	Mightex	N/A
NanoZ impedance testing and electroplating device	White Matter	N/A

Materials and equipment

• •Baytril solution: add 0.35 ml Baytril (100 mg/ml) in 0.65 ml saline
• •Metacam solution (post-operative): add 0.35 ml Metacam (2 mg/ml) in 0.65 ml saline
• •Metacam (pre-operative): draw up 0.35 ml Metacam (2 mg/ml)

Step-by-step method details

Assembly of implant

Timing: 2–5 days

This step covers the fabrication of an implant that incorporates eight micro-drives for tetrodes and one micro-drive for a dual optical fiber/LED assembly.

• 1.Familiarize yourself with the basic steps and procedures for the fabrication of a 3D-printed micro-drive array implant that is loaded with tetrodes.

Note: This has been demonstrated in Kloosterman et al.³ and Nguyen et al.⁴

Note: The implant design used in this protocol was adapted from Kloosterman et al.,³ and many components and fabrication steps are similar.

• 2.
  Prepare implant parts.

  • a.
    Pre-drill screw holes in the main body and base parts (holes a, d, e, f and h to be drilled in any order; labeled green in Figure 1) with a Ø 0.95 mm drill bit. Next, create a screw thread using a M1.2 tap.
  • b.
    Clear through-holes in the base, skull connector and protective cone parts (holes g, j and k; labeled blue in Figure 1) using a M1.2 drill bit.
  • c.
    Clear the eight holes on the inner ring of the main body (holes c in Figure 1) with a Ø 0.8 mm drill bit.
  • d.
    Use pliers to press fit a 20Ga sleeve into each hole.

    Note: The sleeve ensures smooth up-and-down movement of the micro-drive cannula.

    Caution: Be careful when press-fitting the sleeves as the thin walls are easily dented.
  • e.
    Drill the indicated hole of the base (Figure 1I) with a Ø 1.4 mm drill bit.

    • i.
      Insert a metal tubing (Figure 2, item III; Ø 1.5 mm) as a tetrode collector cannula, verify that it is co-linear with the vertical axis of the base.
    • ii.
      Secure it with dental cement applied to its junction with the base. Leave 5 to 7 mm of the cannula exposed.
• 3.
  Assemble the main body and base of the implant.

  • a.
    Insert 6 mil wire into eight 30Ga tetrode support cannulas.

    • i.
      Insert all cannulas into the tetrode collector cannula of the implant base such that they extend approximately 5 mm from the bottom.
    • ii.
      Fill space in between tetrode support cannulas with additional 30Ga tetrode support cannulas or 6 mil wire.
    • iii.
      Apply a small drop of high-viscosity liquid glue to tetrode support cannulas where they meet the collector cannula.

      CRITICAL: Do not let glue get inside the tetrode support cannulas.

  • b.
    Guide each of the 30Ga tetrode support cannulas through one of the small holes at the bottom of the main body, all the way through the corresponding 20Ga sleeves at the top of the main body.
  • c.
    Concurrently with step b., slowly push the main body and base together and secure them in place with three 6-mm long M1.2 screws (holes f and g in Figure 1).

    Note: To facilitate the insertion of the tetrode support cannulas into the main body, it is recommended to first extend and guide the 6 mil wires into the appropriate holes in the main body. The tetrode support cannulas will then follow the 6 mil wire as the two parts are slowly pushed together. Note further that the 6 mil wires inside the tetrode support cannulas will help to create a smooth bend without kinks.

    CRITICAL: Insert the tetrode support cannulas into the nearest hole in the main body to avoid unnecessary crossings and twisting of the cannulas that could create sharp bends.

  • d.
    Further strengthen the 30Ga tetrode support cannulas in the bottom base with dental cement, hardening by a blue-light source (wavelength: 440–480 nm).
  • e.
    For extra security, apply dental cement to the junction between the main body and the base. Add dental cement on the inside of the base to secure the 30Ga tetrode support cannulas in place.
  • f.
    Make sure that the 6 mil wires extend from both ends of the tetrode support cannulas.
  • g.
    Clamp the Dremel rotary tool (set to a speed of ∼1500 rpm) with a diamond cutting wheel into a vise and cut the tetrode support cannulas flush with the bottom of the tetrode collector cannula.

    Note: The tetrode collector cannula should extend at least 4 mm from the base after cutting.

    Caution: Be careful when using the rotary tool and wear safety goggles.
  • h.
    Pull-push the 6 mil wires inside the tetrode support cannulas to clear the opening on the bottom (cut) end.

    Note: The tip of a 30Ga needle can help to clear the opening. Make sure the 6 mil wires slide smoothly up and down inside the tetrode support cannulas before removing them. You could try 5 mil wires first if you cannot go through with 6 mil wires.

  • i.
    Make a note of the mapping between the location of each tetrode support cannula in the collector cannula and the position of the corresponding micro-drive in the implant main body.
• 4.
  Construct tetrode micro-drives.

  • a.
    Use two 6-mm long M1.2 screws to secure the electrode interface board to the top of the main body.
  • b.
    Follow Kloosterman et al.³ to fabricate eight micro-drives (Figure 10A) using the micro-drive mold (Figure 2, item I).

    Note: As a modification to the original protocol, we have found it helpful to press-fit a 20Ga sleeve into the mold to smoothly guide the 23Ga micro-drive cannula.

  • c.
    Insert the micro-drives into the implant.
  • d.
    Follow Kloosterman et al.³ to insert tetrode guide cannulas into the tetrode support cannulas and to secure the tetrode guide cannulas to the micro-drive using high-viscosity glue.
  • e.
    Follow Nguyen et al.⁴ to fabricate eight tetrodes, load the tetrodes into the micro-drive arrays and connect the tetrodes to the electrode interface board with small gold pins.
  • f.
    Measure the impedance of each electrode in the tetrodes and ensure they are properly connected (impedance below 2 MΩ).

    Note: If any electrode has an impedance that is over the limit, first try pushing down the gold pin to make a better connection or replace the tetrode.

• 5.
  Mount the protective cone.

  • a.
    Put aluminum foil around the protective cone to act as a faraday cage.
  • b.
    Insert a pin receptacle (Figure 2, item IX) into the hole of the protective cone (Figure 1, hole l) through the aluminum foil and connect it to the aluminum foil with conductive wire glue (Wire Glue).

    CRITICAL: Make sure there is no wire glue inside the socket.

  • c.
    Cover the entire aluminum foil with epoxy.
  • d.
    Reopen the 3 holes (Figure 1, hole k) of the protective cone.
  • e.
    Attach the protective cone to the implant main body using three 6-mm long M1.2 screws.
• 6.
  Construct optical fiber/LED assembly.

  • a.
    Connect two tapered optical fibers to two customized angled LEDs via thinned sleeves (customized by Doric lenses).
  • b.
    Place the two optical fiber/LED assemblies as close as possible to obtain a fiber distance of 0.9 mm and ensure through visual inspection that the optical fibers are parallel to each other.
  • c.
    Secure the fibers with light-curing dental cement.

    Note: The distance between the optical fibers was optimized for targeting the prefrontal cortex bilaterally but can be adjusted for other cortical targets.

  • d.
    Protect the tips of the optical fibers with polyimide tubing (141-005; Nordson Medical).
• 7.
  Construct optical fiber/LED micro-drive.

  • a.
    Insert a metal shuttle (Figure 2, item XV) into the micro-drive for the optical fiber/LED assembly of the main body (Figure 1) and then insert a M1.2 custom screw.
  • b.
    When the custom screw is in place, secure it by adding dental cement to the lip of the custom screw (see Figure 2 of Kloosterman et al;³) from the side hole of the micro-drive (Figure 1, hole b).

    Note: The custom screw fixed in the micro-drive for the optical fiber/LED assembly is not movable, but rotatable. Thus, the optical fiber/LED assembly is pushed down when turning the screw anticlockwise.

  • c.
    Make sure the shuttle is lowered 2 mm from the top.
  • d.
    Being careful not to touch the optical fibers, insert the optical fiber/LED assembly into the hole in the metal shuttle.
  • e.
    Check at the optical fiber exit tip (Figure 1) of the main body and ensure that the two optical fibers extend 2 mm out of the exit tip of the base. Use dental cement to secure the assembly to the shuttle.
  • f.
    Connect the two wires of the two LEDs to the stimulation S1 and S2 holes in the electrode interface board (Figure 10C) with gold pins.
  • g.
    Measure the output power of the LEDs with an optical power meter (PM100D, Thorlabs, Newton, New Jersey) as a function of the input current. For an example of an output power curve, see Table 1.
• 8.
  Clean and apply a gold coating to the tetrode tips using a NanoZ automatic impedance testing and electroplating device (White Matter, Seattle, WA).

  Note: It is recommended to complete this step immediately prior to surgery.

  • a.
    Lower all tetrodes until they extend ∼3 mm outside the cannula.
  • b.
    Connect the implant to the NanoZ device through an adaptor (Omnetics NZ-EIB-36; Neuralynx) and submerge the tetrode tips in saline solution.
  • c.
    Make sure to complete the circuit by connecting the saline bath to reference pin of the NanoZ.
  • d.
    To clean the electrode surface, pass +0.15 μA @ 1000 Hz for 1 second and target an impedance of 1 MΩ. Repeat at most 10 times.
  • e.
    Replace saline solution with gold-plating solution (Neuralynx) and plate the electrode tips with progressively smaller currents and lower target impedances according to Table 2. Repeat step 3 if necessary.

Figure 10.

An assembled implant

(A) Custom micro-drive. The left is a custom screw, the right is a 23Ga micro-drive cannula, and the white block near the screw head is cured with dental cement.

(B) An optical fiber/LED assembly, including two tapered optical fibers and two customized angled LEDs, attaching to a metal shuttle.

(C) Wiring of two LEDs to S1 and S2 of the EIB.

(D) Top view and side view of a finished implant. Note that optical fibers and tetrodes are sticking out of the two cannulas of the implant.

Table 1.

An example curve of the optical output power (in μW) as a function of the input current in increments of 50 mA

Input current (mA)	Output power (μW)
50	90
100	180
150	250
200	300

Data is obtained from the median values of 25 tested LEDs.

Table 2.

The recommended gold plating procedure to be followed consecutively from step 1 until step 3

Step	Current (μA)	Duration (second)	Target impedance (kΩ)	Runs
1	−0.150	1	500	10
2	−0.05	1	400	10
3	−0.025	1	300	10

Viral vector injection and implantation

Timing: 3–5 h

This step includes both the viral vector injections and the fixation of the implant on a rat skull, for chronic neural recordings combined with closed-loop optogenetic feedback.

• 9.
  Perform the surgical procedure.

  • a.
    Prepare the animal. We used adult male Long Evans rats weighing between 300 and 350 grams (corresponding to 60–75 days; Janvier Labs, Le Genest-Saint-Isle, France).

    • i.
      Anesthetize the animal using 4% isoflurane and 1 L/min oxygen flow in a ventilated induction chamber.
    • ii.
      Weigh the animal and shave its head, and then place it back in the induction chamber to ensure the animal remains anesthetized.
    • iii.
      Note the starting time of the surgery.
    • iv.
      Redirect the isoflurane to the respiratory mask on the stereotaxic frame.
    • v.
      Transfer the animal to the stereotaxic frame and place its nose in the respiratory mask.
    • vi.
      Inject 0.35 ml Metacam (2 mg/ml) subcutaneously to control pain and inflammation.
    • vii.
      Immobilize the head of the animal in the stereotaxic apparatus using nose clamp and ear bars with tips covered in ointment.

      Note: Make sure the animal’s head is secured in place and does not slide but can still rotate around the ear bar axis.

    • viii.
      Carefully pull the tongue out with small dull forceps to keep the airways clear, and protect the eyes with ointment and aluminum foil against dehydration and prolonged light exposure.
    • ix.
      Insert rectal temperature probe covered with Vaseline and clip the infrared sensor on one of the hind paws.
    • x.
      Connect the probe, the sensor and the heating pad to the monitoring system and maintain body temperature at 37°C.
    • xi.
      Lower the isoflurane level to ∼2%.
    • xii.
      Fill three beakers with 70% ethanol, iso-Betadine and sterile saline respectively.
    • xiii.
      Clean the scalp of the animal by alternatingly applying iso-Betadine with cotton tips and wiping it clean with alcohol swaps for a total of three times.
    • xiv.
      Apply lidocaine cream as a local anesthetic on the skin and wait 3 min for it to take effect.
    • xv.
      Gently pinch the tip of tail or the paw of the animal to check for reflexes, and increase the isoflurane level if needed.
    • xvi.
      Make a midline incision with a scalpel spanning from right behind the eyes to in between the ears.

      Note: Disinfect the surgical tools after every use by placing them in a hot bead sterilizer for 3 seconds.

    • xvii.
      Push the skin and periosteum to expose the skull, and use a retractor to keep the skull well exposed.
    • xviii.
      Use a surgical spoon and saline to clean the surface of the skull until it is white in color.
    • xix.
      Apply 0.5 ml of the antibiotic Baytril solution to the exposed skull and surrounding skin with a syringe.
    • xx.
      Score the skull with a scalpel to improve the adhesion of the dental cement at a later step.

      Note: Monitor the animal’s vital signs throughout the surgery. The heart rate should be between 300 and 400 bpm and remain stable. The respiratory rate should be between 30 and 60 bpm and remain stable. The body temperature should be between 36.0 and 37.5 degrees Celsius.

  • b.
    Drill the craniotomies and insert bone screws for anchoring the implant.

    • i.
      Determine the position of lambda and bregma with a needle attached to a stereotaxic arm and ensure these points are aligned in the horizontal plane.

      Note: Adjust the ear bars or nose clamp if necessary. Then position the needle point on bregma and define the corresponding stereotaxic coordinates as the origin.

    • ii.
      Mark the anterior-posterior and medial-lateral extremes of the craniotomies above the hippocampus and mPFC (for coordinates relative to bregma, see Figure 11).
    • iii.
      First mark points with the needle, and then drill a shallow hole using the #007 drill bit.
    • iv.
      Drill the holes for the bone screws using the #009 drill bit. Drill 8 to 10 holes in a semi-circle around the two craniotomies.

      Note: The holes for bone screws should be as close as possible to the bone ridges and not too close to the craniotomies to avoid interfering with the implant placement at a later step.

    • v.
      Use the bone screw clamp to insert each screw into the skull.

      Note: The screws should fit tightly, but should not extend beyond the skull to avoid damaging the brain (approximately 2 turns for the anterior screws and 3 turns for the posterior screws). The screw used as electrical ground is inserted in one of the posterior holes above the cerebellum (see Figure 11).

    • vi.
      One by one, thoroughly clean and dry the base of the bone screws and the surrounding skull with triangular cotton tips.
    • vii.
      Apply a small drop of Vetbond at the base of each screw. Wait several seconds for the Vetbond to set.
    • viii.
      Construct a wall of dental cement that connects the bone screws and cure it immediately so the cement does not flow and touch the surrounding skin.
    • ix.
      Apply the remaining 0.5 ml of the Baytril solution on top of the skull and surrounding skin with a syringe.
    • x.
      Using the #007 drill bit, drill a circular or oval outline of each craniotomy by connecting the four marker holes.
      Caution: Slowly and evenly thin the skull along the outline until cracks appear in the thinned bone and you reach the space between skull and dura mater.
    • xi.
      The remaining piece of bone in the center of the craniotomy should move freely when gently pressed.

      Note: Remove this piece of bone using small dull forceps and try lifting from one end to slowly tear any remaining thinned bone or connective tissue. If the piece of bone does not come out easily, you may need to thin the bone further with the drill.

    • xii.
      Create a small tear in the dura inside the craniotomies using the tip of a needle (e.g. 30G) at a shallow angle to puncture and lift the dura.
    • xiii.
      Use pointy forceps to slowly lift and tear the dura further until the center of the craniotomy is cleared.

      Note: If necessary, apply some saline into the craniotomy for better visibility. After the dura is successfully removed, apply mineral oil to prevent the brain from drying out.

      CRITICAL: Removing the bone in the center of the craniotomy and removing the dura are difficult procedures that should be done slowly with a steady hand, ensuring a clear visual view through the microscope at all times. In the case of mild or severe bleeding, a small piece of Spongostan Dental (Ethicon, Raritan, New Jersey) can be cut off and placed on the bleeding area for several minutes until the bleeding stops.

  • c.
    Inject the viral vector.

    • i.
      Take the viral vector out of the freezer and thaw on ice for 1-2 min. Pipette 6 μL onto a sterilized piece of parafilm.
    • ii.
      Position the tip of the glass pipette in the center of the viral vector droplet and withdraw 5 μL of the liquid at a speed of 1 μL/min.

      Note: Make sure that the tip of the glass pipette remains in the liquid at all times and no air bubbles enter the pipette.

    • iii.
      Position the tip of the pipette above the right hemisphere in the frontal craniotomy (AP: +2.7 mm, ML: +0.4 mm) touching the surface of the brain and set the Z coordinate of the stereotactic manipulator to zero.
    • iv.
      Gradually lower the glass pipette into the brain to a final depth of 4 millimeters.
    • v.
      Infuse 0.5 μL of the viral vector at a rate of 0.6 μL/min (corresponding to 0.1 μL/10 s, see⁵).

      Pause Point: Wait for 5 min to ensure diffusion of the viral vector into the surrounding brain tissue.

    • vi.
      Repeat step v. three more times at 3 mm, 2 mm and 1 mm depths (see Figure 11 right; raise the pipette 1 mm for each injection, infuse 0.5 μL of the viral vector at a rate of 0.6 μL/min and wait 5 min).
    • vii.
      Remove the pipette from the brain and ensure it was not clogged by dispensing a small amount of the viral vector and visually confirming that liquid flows out of pipette tip.
    • viii.
      Repeat steps iii – vii above for the left hemisphere (AP: +2.7, ML: −0.4).
  • d.
    Secure the implant.

    • i.
      Use the stereotaxic manipulator to position the implant above the craniotomies, so that the bottom of the tetrode collector cannula is lowered into the hole but does not touch the brain.
    • ii.
      Lower the optical fibers so that the tips just touch the surface of the brain.
    • iii.
      Lower the optical fibers an additional 4 millimeters (16 full turns of the custom screw).

      Note: If necessary, you can temporarily remove the implant from the stereotaxic frame for easier access to the screws. Reattach the implant to the dummy Neuralynx connector on the stereotaxic arm once the optical fibers have been lowered.

    • iv.
      Spread a layer of silicone grease in each craniotomy, so that the brain is protected from dental cement flowing in.
    • v.
      Slowly lower the implant (<1 mm/min, checking carefully that the optical fibers are not causing any bleedings) with the stereotaxic arm, ensuring that the optical fibers are inserted carefully to a depth of 4 mm.

      Note: The tetrode cannula should be aligned with the brain.

    • vi.
      Apply dental cement around the tetrode collector cannula, connecting the top of the anchoring screws to the implant.
    • vii.
      Cure cement immediately, and when the implant is secured then detach the implant from the stereotaxic manipulator.
    • viii.
      Insert the pin of the ground wire into the corresponding socket on the implant.
    • ix.
      Apply more dental cement to fill the gap between the skull and the implant completely.

      Note: Make a smooth outer wall covering all the screws to avoid irritation for the animal.

      Note: If necessary, stitch the wound at the front and back using the surgical suture.

    • x.
      Lower all tetrodes into the brain to a 1 mm depth (4 full turns of the custom screws), to avoid the tips getting stuck by dried blood at the top of the brain.
    • xi.
      Generously apply lidocaine cream around the wound using a cotton tip to numb any pain and apply iso-Betadine to prevent infections.
• 10.
  Post-operative care.

  • a.
    Turn the isoflurane down to 0% but leave oxygen flowing.
  • b.
    Inject Metacam with saline subcutaneously (use two injection sites of 0.5 ml each to minimize the volume per injection).
  • c.
    Wait for the animal to recover its reflexes and note the end time of the surgery.
  • d.
    Remove the ear bars, nose clamp, rectal probe, and paw clip, and put the animal in a cage lined with towels.
  • e.
    Once the animal can walk, weigh it, and put it back in its home cage. Give nutritional gel in case the animal is not able to chew food yet.
  • f.
    Check on the animal regularly and give a subcutaneous injection of Metacam (0.35 ml) daily for three days following the surgery.

Pause Point: Wait for two weeks after the surgery before starting experiments, to ensure the animal recovers properly, the viral vector has enough time to express, and the risk of shedding viral particles is reduced to a minimum.

Figure 11.

Stereotaxic coordinates for craniotomies and injection sites

Left panel: top view of a rat skull facing right. The coordinates of the 4 holes to drill for each craniotomy are indicated with little red crosses, and the locations of the bone screws are indicated with larger red crosses. The ground screw should be inserted in one of the posterior holes. Right panel: a coronal slice of the mPFC (AP: 2.7), where the injections are made at ML 0.4 in both hemispheres at 4,3,2 and 1 mm from brain surface.

Positioning of tetrodes in the hippocampus

Timing: 1 week

This step describes how to position the tetrodes in the hippocampal CA1 cell layer to measure sharp-wave ripples.

• 11.
  Lower the tetrodes to the CA1 region of the hippocampus.

  • a.
    Launch the acquisition software (Cheetah 6.4.1, https://neuralynx.fh-co.com/research-software/cheetah/) and inspect the initial data acquisition.
  • b.
    Remove any noisy or broken channels from the viewer and pick a quiet channel (low noise and low electrophysiological activity) as an initial reference channel.

    Note: If all channels look noisy, this is probably due to the reference channel being broken. Change the reference channel to fix this issue.

  • c.
    Move the tetrodes from the initial implantation depth (1 mm) to approximately 2.5 mm in increments of maximum 500 μm per day.

    Note: Turn each screw slowly, paying close attention to the electrophysiological signatures as you go down. Both visual and auditory feedback are very useful tools to help guide the electrode positioning. As the tetrodes are moved down, you can expect to encounter spiking activity in the cortex, bursts of activity in the deep cortex (that could be mistaken for sharp-wave ripples), and low amplitude signals in the white matter above the hippocampus.

  • d.
    Stop once you see sharp-wave ripple activity with a positive sharp-wave and large ripple amplitude.

    Note: If the sharp wave is a negative deflection, move the tetrode up slightly (in increments of 50 μm).

    CRITICAL: Keep close track of how much you have turned each tetrode on each day, and which electrophysiological signals you encountered, as knowing the history of movement of each tetrode greatly improves the success in finding sharp-wave ripples.

  • e.
    Pick one tetrode and place it in the white matter above the CA1 to serve as a final reference (find sharp-wave ripples and raise the tetrode until it is in the white matter, as evidenced by very low electrophysiological activity, approximately 250 μm above the point at which sharp-wave ripples were observed). Do this for the first tetrode that you find sharp-wave ripples on, so that this tetrode can be used as a reliable reference when adjusting the other tetrodes.
  • f.
    Pick one tetrode and place it in the cortex above the hippocampus (approximately 1–1.5 mm from the brain surface), for spurious sharp-wave ripple detections.
  • g.
    At the end of each day, raise the tetrodes up to half of the distance that they were lowered, in order to avoid tetrodes passing through the CA1 cellular layer due to overnight drift.

    Pause Point: After moving all the tetrodes (partially) down, let the tissue settle and wait until the next day before moving each tetrode again.

Closed-loop sharp-wave ripple-triggered inhibition of the mPFC

Timing: 10 min

This step describes the steps to start an experimental session to perform closed-loop perturbation of activity in the mPFC following hippocampal sharp-wave ripple detection.

• 12.
  Start online sharp-wave ripple detection using Falcon software.

  • a.
    Start the acquisition of neural data in the Cheetah software on the acquisition computer. This will also start the duplicate data stream from the Digilynx to the Falcon computer.
  • b.
    Start the Falcon server from a terminal window on the Falcon computer (either execute command falcon if the installation was done system wide, or navigate to the build folder and execute ./falcon/falcon if you installed it locally).
  • c.
    Open a second terminal window, activate the Conda environment (i.e., execute conda activate falcon) and start the control client (i.e., execute simple_client).
  • d.
    Upload the yaml file with the sharp-wave ripple detection parameters that was created previously, and start processing.
  • e.
    Open a third terminal window, activate the Conda environment (i.e., execute conda activate falcon) and start the application to visualize the ripple envelope and threshold (i.e., execute live_plot_ripple_stats).

    Note: To change the y-scale of the live visualization, manually adjust the ylim parameter in the configuration file: falcon_clients/plot_live/live_ripple_stats_config.py.

  • f.
    Based on the visual feedback, adjust the threshold in the control client to separate the noise from the sharp-wave ripples as clearly as possible (a threshold between 12 and 16 has worked reliably in our hands).
• 13.
  Set up the LED driver.

  • a.
    Turn on the LED driver and open the Mightex LED controller software on the acquisition computer.
  • b.
    Set the parameters for triggered LED activation.

    • i.
      In the Parameters Setting Selection box, check the “Trigger Setting” option.
    • ii.
      In the Trigger Mode Settings, set the desired LED activation current (up to 200 mA, corresponding to an output power of around 300 μW) and the pulse duration (200000 μs) in the first row of the table.
    • iii.
      In the Current Mode Selection box, check the TRIGGER option.
    • iv.
      Tick the box “Set All Channels (of this device)” and click on “Set Current Mode” to set all the LED channels to trigger mode.

Expected outcomes

Following this protocol, you should be able to obtain clear neural signals in the hippocampus, with a good signal to noise ratio (for an example, see Figure 12A). By slowly lowering the tetrodes into the CA1 cellular layer, you should expect to see clear sharp-wave ripple activity on minimum one, and maximum six of the tetrodes (the other two tetrodes are placed in the white matter and in the cortex, as a reference and a control for spurious sharp-wave ripple detections respectively). It is possible that tetrodes break or otherwise become unusable during the process of fabrication, surgery, recovery or lowering the tetrodes. Although it is recommended to use three electrodes for the detection of sharp-wave ripples in order to maximize the reliability of detections, only one good electrode is in theory also sufficient for accurate detections, as sharp-wave ripples occur very synchronously in time across the dorsal CA1.^6,7,8

Figure 12.

Expected results and encountered problems

(A) Three examples of data in the no inhibition condition (left panel) the on-time inhibition (middle panel) and the delayed inhibition condition (right panel). The sharp-wave ripple detections are shown with dashed gray lines, and the times of the optogenetic inhibition are indicated in orange. The SWR detections by the online algorithm occurred at a median of 40 ms following SWR onset. One tetrode is left in the cortex (in orange) as a control for spurious detections.

(B) A Nissl stained slice of the dorsal CA1 showing an induced tetrode lesion in the CA1 cellular layer.

(C) A fluorescent Nissl stained slice of the mPFC visualized with a fluorescent microscope, showing adequate expression of the viral vector (as evidenced by the expression of the fluorescent protein GFP) in both hemispheres and no clear damage of the surrounding tissue.

(D) Four examples of encountered problems, including large lesions in the mPFC (top panels), a trace of the optical fiber detected in the tissue (bottom left), and insufficient expression of the viral vector (bottom right). Note: these four animals were not included in the experiments.

After finishing the experiments and obtaining histological images from the harvested brains, it should be clearly visible that the virus is expressed bilaterally along the full depth of the mPFC (Figure 12C). Moreover, the trace of the optical fibers should not be visible and there should not be any damage due to obstructed or torn blood vessels (Figure 12D).

Quantification and statistical analysis

In order to characterize sharp-wave ripple detections by the online algorithm, and compare these detections to offline detected events, a Jupyter notebook with example data is available in the accompanying Open Science Framework repository that is located at https://osf.io/3n9p8. The example data was recorded while an animal explores a maze for 15 min. We included data from 8 different channels in the hippocampus (from three tetrodes; see Figure 13A, purple, blue and red traces) and 4 channels in the cortex above the hippocampus (see Figure 13A, orange traces). The comparison includes the match and non-match rate (Figure 13B), corresponding to the fraction of offline SWRs that were also detected and not detected online, ripples envelope (Figure 13C), time to online and offline ripple detection (Figure 13D, left), and duration of offline ripples (Figure 13D, right), detailed method see Den Bakker et al.¹

Figure 13.

Example SWR characterization analysis

(A) One second from an example segment of data (one 15-min learning block on the maze) with two SWRs detected by the offline algorithm (blue shaded area), and by the online algorithm (dashed black line).

(B) The match rate (fraction of offline SWRs that were also detected online) and non-match rate (the fraction of online detections that did not correspond to an offline detection) of this example segment of data.

(C) The average ripple envelope of this example segment of data.

(D) Left panel: the time to detection by the online algorithm (in ms; measured from the onset of the offline detection). Right panel: the durations of the detected SWRs (in ms; measured by the offline detection algorithm).

Limitations

One limitation of this protocol is that no recording electrodes are inserted into the mPFC for recording of neural activity. This would allow for confirmation of successful optogenetic inhibition in every animal. However, we found that adding recording electrodes in close vicinity to the tapered optical fibers made the assembly of the implant more challenging and time-consuming, and reduced the overall success rate of the experiments. Instead, we opted to establish the optogenetic inhibition approach in a first set of experiments and optimize the injection of the viral vector and subsequent expression of the opsin in another set of experiments with anesthetized animals. In our hands, the protocol described above resulted in 89% of experiments (16 out of 18 rats) with histologically confirmed expression of the opsin bilaterally in the mPFC. Experiments in which opsin expression was weak or incomplete, were excluded post-hoc.

Troubleshooting

Problem 1

The brain tissue at the implantation site of the optical fibers is damaged.

Potential solution

• •If there are large round lesions around the implantation site (for example Figure 12D, top panels), it is possible that the insertion speed of optical fibers is too high or there was some tremulous movement during insertion (step 9.c.iv). Ensure that the animal’s head is properly fixed while implanting the optical fibers (i.e., there is no movement due to breathing), and the stereotaxic frame allows for very slow and steady lowering of the implant. If necessary, using an electronic micromanipulator can help regulate the speed and precision of implantation.
• •If there seems to be microstructural damage, double check the amount of viral vector you are injecting (step 9.c). It has been suggested that adeno-associated viruses can cause a decrease of dendritic complexity.⁹
• •If the trace of the optical fibers is clearly visible (Figure 12D, bottom left), it is possible that the LEDs were shining too brightly, or for too long durations. Check the power output of the LEDs (step 7.g) and temperature of the optical fibers to ensure the protocol you are using is not causing burns in the tissue.

Problem 2

There is no clear viral expression in the histological images.

Potential solution

• •If there is no viral expression seen in the histological images (Figure 12D, bottom right), ensure that the correct fluorescence setting is used on the microscope, corresponding to the fluorescent protein in the viral vector. Additionally, ensure that the infusion of the viral vector is working properly and the correct amount is infused at the intended coordinates (step 9.c).
• •If the viral expression is only seen in one hemisphere, ensure that the pipette tip is not clogged in between viral vector injections (step 9.c.vii). Additionally, ensure that the pipette tip is truly aligned to the midline of the brain (step 9.c.iii), and both sets of injections are done in a different hemisphere. The superior sagittal sinus should be clearly visible and the sets of injections should be targeted to the right and left side of this blood vessel.

Problem 3

The electrophysiological signal is noisy.

Potential solution

• •Ensure that the reference signal is not the source of the noise (change reference channel; step 11.b).
• •Ensure that the EIB is properly grounded to the protective cone (step 5.b) and to the bone screw in the skull (step 9.d.viii). Applying some conductive wire glue can help to restore the connection.
• •If the noise is only apparent when the LED driver is turned on, ensure that the LED driver is properly grounded. If one of the LEDs has a problem, but the other one is fine, both LEDs can be connected to the same pins on the EIB, which should result in similar power output according to tests that we have done.
• •If the sharp-wave ripples become less clear over time, adjust the tetrodes as necessary. Tetrodes that are picking up a lot of spiking activity can lead to spurious detections. Raising the tetrode slightly (max 30 μm) can help to eliminate this problem. Tetrodes with low amplitude sharp-wave ripples may need to be inserted slightly deeper (in increments of 30 μm).

Problem 4

The Falcon software is not processing any data.

Potential solution

• •Ensure that a duplicate data stream from the Digilynx acquisition hardware is enabled (Preparation, step 16.b).
• •Ensure that the network is set up properly to receive the neural data stream (Preparation, step 16.c).
• •If Falcon is processing and detecting events, but the events are not registered in the Cheetah software and the LEDs are not turning on, ensure that the correct Arduino port is listed in the Falcon graph (Preparation, step 19.c).

Problem 5

The Falcon software is processing events, but the live plot is not showing anything.

Potential solution

• •Ensure that the y-axis is set to the right scale (step 12.e).

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Fabian Kloosterman (kloosterman.fabian@gmail.com), and Jyh-Jang Sun (g874259@gmail.com).

Technical contact

Further information and requests for technical support should be directed to Hanna den Bakker (hannadenbakker@gmail.com), Jyh-Jang Sun (g874259@gmail.com), and Marine Guyot (marine@codingresearcher.com).

Materials availability

No new materials were generated in this study.

Data and code availability

CAD files, code, and data underlying these results are available in the following Open Science Framework repository: https://osf.io/3n9p8. Falcon software is available from https://falcon-core.readthedocs.io/en/latest/index.html.

Acknowledgments

F.K. and J.-J.S. are funded by Research Foundation Flanders (FWO), Belgium, under grant number G0A5422N. F.K. is funded by FWO, Belgium, under grant number G077321N and by KU Leuven, Belgium, C1 grant C14/17/042.

Author contributions

H.d.B., J.-J.S., and M.G. established the protocol. H.d.B., J.-J.S., and F.K. wrote the original draft. F.K. supervised the study. H.d.B., J.-J.S., M.G., and F.K. reviewed and edited the manuscript.

Declaration of interests

J.-J.S. is a co-founder of Sun and Jiang Consultancy and a consultant of Atlas Neuroengineering.