Protocol for the quantitative assessment of social and nonsocial reward-seeking in mice using an automated two choice operant assay

Summary

The complexity of social behaviors makes it difficult to study the neural mechanisms that underlie them. Here, we describe an automated, low-cost two choice operant assay to directly compare social and nonsocial reward-seeking in mice. We provide instructions for the assembly of the operant chamber and conducting behavioral experiments, which include a graphical user interface (GUI)-based acquisition system that can be readily combined with various experimental manipulations. This assay allows for the characterization of social and nonsocial behaviors and their associated neural mechanisms.

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

Graphical abstract

Highlights

• •Instructions for assembling a two choice (social-sucrose) operant assay
• •Software provided for behavioral protocols and data analysis
• •Steps described for conducting behavioral experiments and neural manipulations

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

The complexity of social behaviors makes it difficult to study the neural mechanisms that underlie them. Here, we describe an automated, low-cost two choice operant assay to directly compare social and nonsocial reward-seeking in mice. We provide instructions for the assembly of the operant chamber and conducting behavioral experiments, which include a graphical user interface (GUI)-based acquisition system that can be readily combined with various experimental manipulations. This assay allows for the characterization of social and nonsocial behaviors and their associated neural mechanisms.

Before you begin

Appropriate social behaviors are imperative for the survival of many species. Affiliative social behaviors are maintained through the rewarding nature of social interactions. Additionally, animals navigate a dynamic environment in which their motivation to seek different rewarding stimuli, such as conspecifics or food, is flexibly modulated by various internal and external factors.^2,3 However, it is unclear how the brain represents competing rewarding stimuli and guides decisions to seek one reward over another.⁴ In particular, it is unknown how neural representations of social rewards compare to those of nonsocial rewards.⁵ Much progress has been made towards characterizing the neural circuits that mediate nonsocial reward-related behaviors. In contrast, less is known about how the brain represents social rewards and modulates social reward-seeking behavior.⁶

A number of behavioral assays have been developed to investigate social behavior in laboratory animals.⁷ Several of these have been adapted from nonsocial behavioral paradigms.⁸ In particular, the conditioned place preference (CPP) assay which compares the time spent by an experimental mouse in a reinforced versus nonreinforced location is now used to characterize social reward.⁹ On a social CPP assay, social reward is measured as the time spent in the location of a previous social interaction compared to a location without a previous social interaction.^10,11,12 Another assay that examines social reward is the three chamber assay which quantifies time spent in the area of a chamber with a social target versus one without a social target.¹³ In both of these assays, time spent in the previous location or proximity of a social target is used as a metric for social reward-seeking.¹⁴ However, time spent is a passive metric that does not fully capture the motivational component of social reward-seeking. To capture the full range of social behaviors, naturalistic assays, such as the resident-intruder assay, have also been used.¹⁵ Despite increasing the range of social behaviors that can be assessed, these assays make it difficult to separate the motivational and consummatory components of social behaviors.¹⁶ They also introduce other confounds, such as subjective hand-scoring of particular behaviors.¹⁷ Operant assays restrict social behaviors to the motivational component, which eliminates the need for hand scoring. Additionally, operant assays enable researchers to differentiate active reward-seeking as opposed to passive social contact measured in the social CPP or the three chamber assay. In fact, several social operant assays have recently been developed, all of which demonstrate that mice will engage in operant behavior to access a conspecific.^18,19,20,21

Here, we describe a protocol for the assembly and implementation of a two choice (social-sucrose) operant assay that allows for the direct comparison and quantification of social and nonsocial reward-seeking behavior.¹ Details regarding operant assay assembly and the experimental timeline are provided in Figures 1, 2, and S1. We demonstrate how to use a programmable state machine with an easily constructed operant arena to create a low-cost alternative to available operant assays. Using this two choice operant assay, we show that male and female C57 mice display similar levels of positive reinforcement of social and nonsocial reward-seeking behavior.

Figure 1.

Components for assembly of the two choice operant assay

(A) Schematic of the two choice operant assay in which a mouse is trained to nose-poke choice ports to access social or sucrose rewards.

Top-down view (B) and magnified images (C) of the behavioral setup with numbered labels for the components included in its assembly (described in Stage 1 of the protocol) - (1) Sucrose choice port, (2) Social choice port, (3) Sucrose reward port, (4) Sucrose reward reservoir, (5) Automated social gate, (6) Acrylic lid for social target box, (7) Social target box, (8) Platform for social target box, (9) Bpod state machine hardware, (10) Conveyor belt for social gate, (11) Arduino Uno, (12) Stepper motor, (13) IR light source, (14) Top-view mounted camera. Colored insets in (C) highlight the wiring between the stepper motor, Arduino, and the conveyor belt (in the left panel), wiring for the choice port (middle panel, right), and a front view of the social target box with its platform and acrylic cover (middle panel, bottom). The right panel in (C) shows the connections to and from the Bpod state machine.

Figure 2.

Experimental timeline to establish reward-seeking behavior in C57 mice

(A) Schematic of behavioral training timeline. Mice are water deprived and run through a series of increasingly difficult training phases within the same operant chamber. The first training phase is operant conditioning (described in Stage 5, steps 55-71), the second training phase is single port (described in Stage 6, steps 72-88), the third training phase is opposing port (described in Stage 7, steps 89-105) and the fourth training phase is partial water access (described in Stage 8, steps 106-127).

(B) Schematic of experimental timeline after completion of training. Mice are removed from water deprivation and run through the two choice operant assay on full water access (described in Stage 9, steps 128-149) and can be run through additional experimental manipulations (described in Stage 10, steps 150-171). Modified from ref.¹

Institutional permissions

All experimental procedures were approved by the Emory Institutional Animal Care and Use Committee. The users of the protocol must acquire similar permissions from their relevant institutions.

Comparison to existing social operant assays

There has been a growing interest in dissociating the motivational aspects of social reward seeking behavior from the consummatory aspects of social behavior.¹⁶ One such social operant assay identified a neural circuit from the medial amygdala to the hypothalamus that was necessary for social operant behavior.¹⁸ Another study using a social operant assay demonstrated that dopaminergic neurons in the ventral tegmental area increase their activity during the social interaction phase of the assay and that inhibiting these neurons disrupts social reward-seeking behavior.¹⁹ The aforementioned operant assays do not directly compare social and nonsocial reward-seeking within the same paradigm. A different operant assay has examined operant behavior for both conspecifics and palatable food rewards in female mice. This assay showed that CD1 female mice lever press more than C57 female mice to gain access to a familiar conspecific compared to a palatable food reward.^20,21 However, this assay relies on commercially available operant chambers which can be costly. Our two choice operant assay is built entirely from inexpensive, open source components, and we provide the requisite code for a variety of experimental manipulations. Furthermore, we show that C57 mice of both sexes will engage in operant behavior to access conspecifics and sucrose rewards on our two choice operant assay.¹

Operant assay assembly

Timing: 1–2 days

The physical components of the two choice operant assay include an acrylic chamber with social and nonsocial (sucrose) reward access zones on opposite sides and two choice ports on the wall adjacent to both reward access sites (Figures 1A and S1A: left panel). Below, we describe the steps necessary to make the port access holes in the acrylic chamber. While the assembly of the operant chamber does not require a particular skillset, we recommend getting assistance from an experienced technician when necessary. The assembly of the operant arena requires at least 2 personnel.

Note: For safety, we recommend using safety goggles or a face shield as a protective cover from acrylic scraps when drilling.

• 1.Choose a preferred side of the acrylic box for the choice ports and mark two 1″ circles that are centered 3″ from the corners, and 1″ from the base of the box (Figures 1B and 1C: numbered insets 1 and 2).
• 2.Using a graduated step drill bit (3/4″), drill through the marked circle starting from the center using moderate force and progressively increasing the speed. Stop the drill when you are about to reach the marked edges (circumference) of the circle.
• 3.File the edges of the circular hole using a round file to smooth the fluted edges created by the drill bit until you reach the marked edges. Test the fit of the choice ports in each circular hole and file further if necessary (Figure S1A: numbered inset 1).
• 4.On one of the sides adjacent to the side with the choice ports, mark a 1″ circle that is centered 6″ from the side and 1″ from the bottom for the sucrose delivery (Figure 1B: numbered inset 3). Repeat steps 2-3 to create the reward port hole.

CRITICAL: The acrylic box is prone to cracking with excessive force from the drill. In order to avoid cracks, ensure that a sharp and high-quality drill bit is used, and support the box through the process. Alternatively, you can use a 1/8″ drill bit prior to drilling the full hole in order to make a smaller hole to serve as an anchor to stabilize the 3/4″ drill bit or a 1″ hole saw drill bit to drill the holes for the ports. The acrylic box can also be substituted for a thicker one (T&T Plastic land – Catalog# - D#29T2PZZ ).

• 5.On the side opposite to the sucrose reward port, mark a 2″ × 2″ square centered 6″ from the side and 1/2″ from the bottom for the social target access.
• 6.Using a 1/8″ regular step drill bit (non-fluted), create pilot holes in the inner edge of the 2″ × 2″ square.
• 7.Insert a hot knife through the holes and drag along the corner of square marking until the acrylic piece peels off (Figure S1A: numbered inset 2).

Note: For safety, we recommend using leather gloves while using the hot knife.

CRITICAL: Take sufficient caution when using both the drill and the hot knife to cut through the box. Ensure that the hot knife is heated to 315°C before starting to cut. Additional holes using the 1/8″ drill bit can be made if the knife has difficulty moving through the acrylic.

Note: The 2″ × 2″ cutout made in steps 5-7 is covered on the outer side of the chamber by a flat sheet of aluminum fixed to a conveyor belt with a stepper motor whose movement is controlled by an Arduino Uno. This serves as the automated social gate and the steps for its assembly are given below (Figure S1A: middle panel).

• 8.Set the camera slider on a flat base (Figure S1A: numbered inset 3).

Note: Tighten the 4 red screws on the base of the slider to ensure stability. A linear actuator controlled by a stepper motor can be used instead of a camera slider to move the gate. (See problem 1).

• 9.
  Bring the platform of the camera slider to the starting point of the conveyor belt (the end of the belt has a motor attached to it).

  • a.
    If there is resistance to the movement of the camera slider, loosen the red screw attached to the platform (camera platform lock).
• 10.
  Using the L-bracket, align the 8″ × 8″ × 0.063″ aluminum sheet at right angles to the platform as shown in (Figure 1C: numbered inset 5).

  • a.
    The long shaft of the bracket sits on the platform, while the short shaft is attached to the aluminum sheet.
• 11.Use either cobalt or titanium drill bits to increase the width of the holes on the L-bracket to fit the 1/4th″ and 3/8th″ screw holes on the platform to account for the slight misalignment of screw holes between the bracket and platform.
• 12.Mark the edges of the bracket’s short shaft onto the aluminum sheet. Using sandpaper, roughen the outer surface of the L-bracket and the marked portion of the aluminum sheet to ensure proper adhesion.
• 13.Cut 2″ of epoxy putty resin and knead until a uniform color is reached. Cover the marking on the aluminum sheet with the epoxy and firmly press the bracket onto it until it is flush with the surface.

Note: Additional epoxy can be added to the sides as needed to strengthen the grip. The epoxy requires a minimum of 2 h to dry (Figure S1A: numbered inset 4).

• 14.Once dried, fit the gate to the platform using 1/4″ and 3/8″ bolts with the corresponding nuts and tighten to the desired extent (Figure S1A: numbered inset 5). (See problem 2).

Note: For safety, please use nitrile gloves when handling epoxy resin.

Note: The social target mouse is housed in a smaller acrylic chamber (3″ × 3″×3″) with one side fitted with 6 aluminum rods (3/32″ in diameter and 0.2″ apart) (Figures 1B and 1C: numbered inset 7). Spacing of bars in this manner creates a grate that allows for the transfer of chemosensory cues and social contact between the experimental mouse and social target without permitting either mouse to enter the other area. The chamber is supported by a plexiglass encasing and placed on an aluminum platform to ensure its stability. (Figures 1B and 1C: numbered inset 8). The steps below describe the assembly of the social target box along with its platform.

• 15.Cut the aluminum rods into 3″ pieces using a coping saw.
• 16.Mark points that are 0.2″ apart from each other, and 0.04″(0.1 cm) away from the edge on two open sides of the acrylic box and use a ruler to ensure that they are as aligned to each side as possible.
• 17.Using a round Dremel tip (3/32″ preferably, however, any tip size that is < 3/32″ would also work), make holes in the marked points on the two edges of the open side of the acrylic.
• 18.Using a triangular file, flatten the sides of the holes, and fit the aluminum rods in each hole (Figure S1A: numbered inset 6).
• 19.To create a placeholder for the social target box on the aluminum platform, place it 3″ from the sides of the platform, ensuring the open side of the chamber (now fit with bars) is 1″ from the edge. Mark the boundaries of the chamber on the platform.
• 20.Cut plexiglass to the measurements of the markings taken above along with an additional margin to make room for the target chamber to fit in. This is approximately 3.2″-3.4″. (See problem 3).
• 21.Cut about 4″ of epoxy putty and knead until uniform color is reached. Stick the plexiglass to the platform using putty on the outer edge of the glass. Let it dry for about 2 h (Figure S1A: numbered inset 7).
• 22.Test the fit of the social target chamber in the plexiglass enclosing and adjust accordingly.

CRITICAL: When making the smaller diameter holes on the open side of the box, excessive pressure can break the box. Going slowly and using a high-quality drill can help avoid breaks. Alternatively, a cage can be 3D-printed with the specified dimensions.

• 23.Gather all materials: Computer, acrylic operant box, automated social gate, social target box with its platform, Bpod state machine, Arduino UNO, stepper motor, two mouse behavior ports, two port interface boards, one mouse port assembly as sucrose reward port, 10 mL syringe, silicone tubes, three CAT5e ethernet cables, BNC(F) to lead wire, jumper wires (male to male, male to female, and female to male), USB micro connector cable, two 12 V AC/DC adaptor with barrel jack, USB 2.0 cable type A/B, and a female DC Cord to terminals block transformer plug.

Note: To facilitate seamless communication between the hardware components of the operant assembly, we recommend using a 64-bit Windows OS computer with a minimum of 16 GB RAM and 500 GB of SSD storage.

• 24.
  Establish the wiring between the operant box and Bpod through the following steps:

  • a.
    Connect the mouse behavior ports to the port interface boards as shown in (Figure 1C: numbered inset 1, zoomed image; Figure S1B: numbered inset 2) to make the 2 choice ports.
  • b.
    Insert the ports into the choice port holes in the operant box and secure them using double-sided Velcro tape or Scotch tape.
  • c.
    Connect the social choice port (choice port on the opposite side of the social reward zone, adjacent to sucrose reward port) to the Bpod using port 1 of the Behavioral ports with an ethernet cable.
  • d.
    Similarly, connect the sucrose choice port (choice port on the opposite side of the social zone, and the side adjacent to the sucrose reward port) to Behavioral port 3 and the sucrose reward port to Behavioral port 2 (Figure S1B: numbered inset 1 and 3).

    Note: The configuration of the two choice ports can be changed based on experimental conditions.

  • e.
    Remove the plunger of a 10 mL syringe and connect its needle end to a silicone tube (Figures 1B and 1C: numbered inset 4).
  • f.
    Connect the other end of the silicone tube to the open port of the solenoid valve on the reward port assembly.
  • g.
    Tape the syringe to the top of the operant box on the side of the sucrose reward to create an elevated reservoir that facilitates the flow of sucrose to the reward port.
  • h.
    Fill the syringe with distilled water for testing.

• 25.
  Establish the wiring between the social gate assembly, Bpod and Arduino through the following steps:

  • a.
    Connect the Bpod I/O port OUT1 to the Arduino board pin 4 and GND using a BNC(F) to lead wires (Red -> 4 and Black -> GND) (Figure S1B: numbered insets 1 and 7).
  • b.
    Connect the Arduino to the microstep driver using a male-to-male jumper wire as shown in Table 1 and the schematic in Figure S1B: numbered insets 7 and 6.
  • c.
    Lastly, connect the microstep driver to the stepper motor on the slider as shown in Table 2 and the schematic in Figure S1B: numbered inset 5. (See problem 4).
  • d.
    Place the social target box and its platform such that the side of the chamber with aluminum bars faces the 2 × 2 square side of the operant chamber (Figure 1B: numbered insets 6-8).
  • e.
    Place the aluminum social gate between the social target chamber and the open side of the operant assay (Figures 1B and 1C: numbered inset 5). Ensure that the social gate’s platform is at the starting point of the conveyor belt.
• 26.
  Establish the power supply to the Bpod, Arduino, and the microstep driver through the following steps:

  • a.
    Connect the Bpod to the computer with a USB micro connector. This serves as the primary power supply to the Bpod. Additionally, connect the Bpod to a wall-mounted power supply using a 12 V DC barrel jack (provided with purchase of the Bpod) to power the solenoid valve in the sucrose reward port.
  • b.
    To power the Arduino, connect it to the computer using a USB 2.0 cable type A/B.
  • c.
    To power the stepper motor, connect the VCC and Ground pins on the microstep driver to a female DC to terminal connector, then attach it to a 12 V AC/DC power adaptor.
  • d.
    To further control the movement resolution (precision/speed) and the current supply to the stepper motor, modify the DIP switches in the microstep driver.

Note: We recommend starting with a 1/4th microstep mode (with 800 rev/sec), and 0.5 A of current supply to the motor. This corresponds to an ON-OFF-OFF-ON-ON-ON setup for the DIP switches on the microstep driver (Figure S1B: numbered inset 6, bottom inset in red).

Table 1.

Arduino to Microstep driver pin connections

Arduino pin #	Microstep driver port
8	ENA+
2	DIR+
5	PULL+
GND	ENA-, DIR-, PULL-

Table 2.

Microstep driver to Stepper motor pin connections

Microstep driver port	Stepper motor port
A+ and A-	Coil 1
B+ and B-	Coil 2

Software setup

Timing: 1–2 h

The protocol is designed for a Windows operating system (OS), and we recommend using a 64-bit Windows 10 OS.

• 27.If MATLAB is not installed, download MATLAB 2021a here, and follow the instructions in the installer page.
• 28.Download the MATLAB API for Bpod along with Bpod protocols for the two choice operant assay, the associated Arduino code and analysis pipeline from Supplementary materials.
• 29.Extract files and folders to a preferred location on the computer.
• 30.
  Open MATLAB and add the path to the Bpod_Gen2 root folder using the “Set Path” option in the “Environment” tab. Do NOT select “Add with subfolders”.

  • a.
    To test whether Bpod can be accessed through MATLAB, run Bpod in the command window, which should open the Bpod console GUI as seen in Figure 3A.

Note: If the Bpod is successfully powered, the state machine indicator turns blue. It will turn green upon opening the GUI console on MATLAB. (See problem 5).

• 31.(Optional, if Arduino IDE is not installed) Download Arduino IDE version 1.18.19 here, and follow the instructions for installation here.
• 32.Navigate to the Arduino folder from the downloaded code package.
• 33.Open the sketch file socialGateAssembly.ino in the Arduino IDE and upload the sketch onto the Arduino board by following the instructions here.

Figure 3.

Graphical user interface (GUI) for the two choice operant assay

(A) GUI elements of the Bpod console with numbered insets corresponding to Steps 45-50 in Stage 3: Testing the operant assay. If a successful connection is established between the Bpod hardware and the computer, running the Bpod command in MATLAB will open the console GUI as shown in the left panel of (A). The numbers refer to (1) the wrench icon to access the settings menu GUI shown on the right panel of (A); (2) port icon to enable behavioral ports 1-3; (3) the LED buttons of the behavioral ports that can be toggled from the OFF (red colored button) to the ON (green colored button) state; (4) the valve button for behavioral port 2 that opens the solenoid valve to allow the flow of liquid to the port from the reward reservoir in the ON state; (5) liquid calibration icon for precise delivery of the reward; (6) BNC1 button to open and close the social gate in the ON state and OFF state respectively; (7) start/pause button that opens the protocol launch manager.

(B) The launch manager features two panels with drop-down lists to choose from one of the four protocols (blue dialog box, defaults to the Operant conditioning protocol), and select the mouse’s ID (pink dialog box, defaults to FakeSubject (a dummy subject)). The plus icon (highlighted in orange) next to the drop-down menu is used to add a new mouse id to the protocols (orange dialog box). The next button in the bottom right of the launch manager navigates to the protocol parameters window that is unique to the chosen protocol.

(C) The template for protocol parameters. Users can verify and modify the parameters associated with the protocol such as session duration, duration of the inter-trial interval, amount of sucrose reward dispensed, etc. The back button in the bottom left can be used to revert to the launch manager and the next button can be used to advance to the session summary window which is unique to each protocol.

(D) A template of the session summary window. This window features the chosen protocol settings, in-progress results featuring trial numbers, and a stairstep graph that serves as a visual representation of the trial progression in the session. At the end of the session, the session summary window is automatically saved within the data folder corresponding to the mouse ID.

Testing the operant assay

Timing: 1–2 h

CRITICAL: Our protocol is optimized for mice on a reverse light-dark cycle, and we perform our behavioral recordings in low light conditions. We recommend using an IR light to illuminate the operant arena. Ensure that both the camera and lens have high spectral sensitivity to the IR wavelength.

• 34.Place the IR light source on the side of the operant chamber facing the choice ports such that it illuminates the box and its beam faces the floor of the chamber. (Figure 1B: numbered inset 13).

Note: In our experiments, the IR light is approximately 6–8" from the operant box, however, we recommend testing the optimal distance by accounting for both the level of illumination and direction of its beam.

• 35.Check the video stream of the camera and ensure that all four edges of the chamber are clearly visible, along with the ports and the social gate. Position the chamber in the center of the camera’s view.
• 36.Set the acquisition frame rate of the camera to 40 Hz using the Pylon software.

CRITICAL: It is critical to optimize the quality of the video recording to improve the accuracy of behavioral tracking by Bonsai software. The parameters to optimize are: the vertical distance of the camera from the box, the zoom and focus of the camera which should provide a clear image of the operant arena and the mice, and the brightness/contrast controlled by the aperture.

• 37.On the MATLAB command line, open the Bpod console GUI using the command Bpod or Bpod(‘COM#’). The state machine indicator on the Bpod hardware will turn green upon successful connection to the computer. Details regarding the elements of the console GUI and their functionality can be found here.
• 38.Navigate to the Config section of the GUI and choose the settings button (Figure 3A: numbered inset 1). This opens a dialog box with menu options to configure the Bpod settings (Figure 3A: right panel). Click on the 4th panel from the left to enable the use of behavioral ports 1, 2, and 3. (Figure 3A: numbered inset 2).
• 39.Upon enabling the ports, test the LED of the choice and reward ports by toggling the LED dial buttons 1, 3, and 2 respectively from the OFF to ON state on the Manual Override panel of the GUI. (Figure 3A: numbered inset 3).
• 40.Test the flow of liquid through the reward port by toggling the VLV dial button for port 2 on the Manual Override panel of the GUI (Figure 3A: numbered inset 4). When toggled, the solenoid valve clicks, and liquid should flow through the steel tube of the port. (See problem 6).
• 41.To ensure a precise delivery of liquid from the reward port, perform liquid calibration (Figure 3A: numbered inset 5) for the solenoid valve on the reward port by following the steps here for Valve 2 using distilled water.

Note: We recommend testing 5 measurements starting from 2 ms pulses to 10 ms pulses to determine the liquid calibration curve. Once completed, replace the distilled water with 10% sucrose solution.

CRITICAL: We use 10% sucrose solution (1 gm solid sucrose / 10 mL distilled water), dispensed at 10 μL intervals, as the nonsocial reward in our operant assay. To maintain the purity and quality of the solution, we recommend using distilled water for the dilutions and preparing fresh sucrose as necessary. (See problem 7).

• 42.Test the movement of the social gate by toggling the BNCOut dial button 1 (Figure 3A: numbered inset 6). The gate should open swiftly when toggled to ON state and have a 4 s jittered movement when closing once toggled to the OFF state. (See problem 7).

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
Sucrose	IBI Scientific	IB37160
Deodorizer Epoleon N-100	Wildlife Control Supplies	WCSN100-S
Experimental models: Organisms/strains
C57BL/6J male and female mice (8–12 weeks)	Jackson Laboratory	JAX: 000664
Software and algorithms
MATLAB 2021a	MathWorks	https://www.mathworks.com
Bpod – MATLAB control software	Sanworks	https://github.com/muruganlab/Two_choice_operant_assay/tree/main/Bpod/Bpod_Gen2
Codes for behavioral protocols for the two choice assay	This study	https://doi.org/10.5281/zenodo.14975400
Arduino IDE v.1.18.19	Arduino	https://www.arduino.cc/en/software
Arduino code for the social gate assembly	This study	https://github.com/muruganlab/Two_choice_operant_assay/tree/main/Arduino/socialGateAssembly
Bonsai	Lopes et al., 2015	https://bonsai-rx.org/docs/articles/installation.html
Bonsai code for behavioral tracking in the two choice assay	This study	https://github.com/muruganlab/Two_choice_operant_assay/tree/main/Analysis/Bonsai
Pylon v.6.3	Basler	https://www.baslerweb.com/en-us/downloads/software/
Other
Numerical ear tags	Kent Scientific	INS1005-1Z
Operant box: 12″ × 12″ × 12″ acrylic chamber	Amazon	ASIN - B00KRDZWMG
Bpod r >= 1.9	Sanworks	Product ID 1035
Choice port: mouse behavior port + port interface board	Sanworks	Product ID 1002, Product ID 1008
Sucrose reward port: behavioral mouse port assembly	Sanworks	Product ID 1009
Arduino UNO R3	Amazon	ASIN - A000066
Flat sheet aluminum, 8″ × 8″ × 0.063″	McMaster-Carr	89015K186
L-bracket	Amazon	ASIN - B0768SH78H
Stepper motor driver, TB6600 4A, 9–42 V	Amazon	ASIN - B07RRB6BGQ
Motorized slider, GVM camera slider	Amazon	ASIN - B07M7G7C6T
Acrylic box 3″ × 3″ × 3″ with 4 sides	Amazon	ASIN - B07D3DSBCS
Tight tolerance, high strength, aluminum rods, 7075, 3/32″ diameter, ½′ length	McMaster-Carr	9063K29
Flat sheet aluminum 6061, 6″ × 6″ × 1/4″	McMaster-Carr	9246K423
10 mL syringe (without needle)	Amazon	ASIN - B08JZ4KVQR
Silicone tubing, 1.5 mm internal diameter	Amazon	ASIN - B01MXFB25J
Soft white diffuser sock for standard reflector	Amazon	ASIN - B01KA5JNYS
Far red (IR) light source for illumination CMVision	Amazon	ASIN - B004F9LF7E
100 h timer	Traceable	5001CC
Basler ace 2: a2A1920, 160umBAS	Basler AG	Catalog - 107820
C-mount 12–36 mm varifocal lens	B&H Photo Video	COM3Z1228CM
Tripod camera mount	Basler AG	2200000191
Camera pole clamp	Amazon	ASIN - B0018LQVIA
Acrylic sheet for the base of the operant box	Canal Plastics Center	White Opaque P95 Matte Acrylic Sheet - 18″ × 24″ / 1/8″ (0.118″ / 3.0 mm)

Step-by-step method details

Stage 1: Conducting behavioral experiments

Timing: 2–3 days

This section describes the steps for preparing to start behavioral experiments.

• 1.Group house and maintain mice on a reverse light-dark cycle.

CRITICAL: We use male and female C57 mice aged 8–12 weeks (obtained from Jackson Laboratories) for all behavioral experiments. We group house the mice with up to four additional cagemates and maintain them on a reverse light-dark cycle for the duration of the experiments. For the behavioral training phase (approximately two weeks), place each mouse on water deprivation schedule while ensuring they receive up to 1.5 mL of water outside the assay. Otherwise, provide ad libitum access to food and water to the mice during the remainder of the behavioral testing phases.

• 2.Secure numerical ear tags onto each mouse to track the identity of individual mice throughout the behavioral experiments.
• 3.At least two days prior to behavioral training, place mice on a water deprivation schedule. Track their individual weights and provide them with up to 1.5 mL of water outside of the operant assay each day.
• 4.
  Prior to initiation of behavioral experiments, ensure the functionality of the two choice operant assay.

  • a.
    Check the calibration of the sucrose valve and that it is dispensing liquid when the valve is opened.
  • b.
    Check sucrose and choice ports for infrared beam break detection.
  • c.
    Check that the gate opens and closes when BNC 2 is pushed.

Note: For issues with the sucrose reward delivery, see problem 6 and 7. For difficulties with the social gate function, see problem 8.

CRITICAL: Wipe down the operant chamber and social target box between each behavioral session with deodorizer. Clean the entire apparatus thoroughly with soapy water at the end of each day of behavioral testing. Avoid ethanol-based cleaning products which may cause the acrylic to crack.

Note: We have described the steps for assembling a single operant chamber. Each behavioral session is 1 h long, allowing for a cohort of 6–8 mice to be run per day on individual behavioral setups. This protocol can be readily extended to accommodate multiple operant chambers and the recording of simultaneous behavioral sessions. When using multiple operant chambers ensure that they are separated by a minimum of 20" and regularly wipe down the surfaces with deodorizer or soapy water for odor control. The software (installed on one computer) can support up to two operant setups namely Bpod1 and Bpod2 (selected from the dropdown menu under ‘Protocol parameters’, shown in Figures 3C, S2B, S4B, S6B, S8B, and S11B). For Bpod1 the sucrose reward port is associated with Behavioral port 2 and for Bpod2 the sucrose reward port is associated with Behavioral port 4. Depending on the experiment, multiple behavioral setups can be used at once. The data collection phase requires at least of 2 personnel.

Stage 2: Operant conditioning—first training phase

Timing: 1–2 days

This section describes how mice are trained to associate nose-poking a reward port to access a sucrose reward.

• 5.Transport mice from animal facility to behavioral testing room.
• 6.Initiate behavioral session. Open MATLAB and type Bpod into the command window.
• 7.In the protocol launch manager, select ‘Operant’ in the dropdown menu to the right of ‘Choose protocol to run’, enter the animal ID below and click the next button (Figure S2A).
• 8.On the following screen, set the protocol parameters as desired and click the next button (Figure S2B).
• 9.On the following screen, confirm that the recording settings are correct (Figure S2C).
• 10.Open Pylon software, confirm settings and ensure that operant assay is centered in the camera viewer.
• 11.Place mouse in operant assay.
• 12.Start video recording.
• 13.Start the behavioral session by pressing the play button on the Bpod GUI (Figure S2D, green box). All port LEDs will flash once to signal the start of the behavioral session.

Note: The flowchart for the code structure is shown in Figure S3.

• 14.Behavioral results from the ongoing session will appear in the pop-up window, along with a timer showing approximate time remaining (Figure S2E).
• 15.Allow the mouse to perform for 1 h in a dark, quiet room.
• 16.After an hour, stop video recording, remove mouse from behavioral assay and return to a cage.
• 17.Record number of successful sucrose trials.
• 18.Wipe down the operant chamber with deodorizer.
• 19.At the end of each day, clean the operant chamber with soap and water.
• 20.Return mice to animal facility.
• 21.Repeat daily until the mouse has reached a minimum of 50 successful sucrose trials in a single hour-long behavioral session. Advance mouse to the next training phase.

Stage 3: Single port—second training phase

Timing: 1–3 days

This section describes how mice are trained to associate an illuminated reward port with access to a sucrose reward.

• 22.Transport mice from animal facility to behavioral testing room.
• 23.Initiate behavioral session. Open MATLAB and type Bpod into the command window.
• 24.In the protocol launch manager, select ‘SinglePort’ in the dropdown menu to the right of ‘Choose protocol to run’, enter the animal ID below and click the next button. (Figure S4A).
• 25.On the following screen, set the protocol parameters as desired and click the next button. (Figure S4B).
• 26.On the following screen, confirm that the recording settings are correct. (Figure S4C).
• 27.Open Pylon software, confirm settings and ensure that operant assay is centered in the camera viewer.
• 28.Place mouse in operant assay.
• 29.Start video recording.
• 30.Start the behavioral session by pressing the play button on the Bpod GUI (Figure S4D, green box). All port LEDs will flash once to signal the start of the behavioral session.

Note: The flowchart for the code structure is shown in Figure S5.

• 31.Behavioral results from the ongoing session will appear in the pop-up window, along with a timer showing approximate time remaining. (Figure S4E).
• 32.Allow the mouse to perform for 1 h in a dark, quiet room.
• 33.After an hour, stop video recording, remove mouse from behavioral assay and return to a cage.
• 34.Record number of successful sucrose trials.
• 35.Wipe down the operant chamber with deodorizer.
• 36.At the end of each day, clean the operant chamber with soap and water.
• 37.Return mice to animal facility.
• 38.Repeat daily until the mouse has reached a minimum of 40 successful sucrose trials in a single hour-long behavioral session. Advance mouse to the next training phase.

Stage 4: Opposing port—third training phase

Timing: 2–7 days

This section describes how mice are trained to associate nose-poking an illuminated choice port with subsequent sucrose reward availability at the reward port.

• 39.Transport mice from animal facility to behavioral testing room.
• 40.Initiate behavioral session. Open MATLAB and type Bpod into the command window.
• 41.In the protocol launch manager, select ‘OpposingPort’ in the dropdown menu to the right of ‘Choose protocol to run’, enter the animal ID below and click the next button. (Figure S6A).
• 42.On the following screen, set the protocol parameters as desired and click the next button. (Figure S6B).
• 43.On the following screen, confirm that the recording settings are correct. (Figure S6C).
• 44.Open Pylon software, confirm settings and ensure that operant assay is centered in the camera viewer.
• 45.Place mouse in operant assay.
• 46.Start video recording.
• 47.Start the behavioral session by pressing the play button on the Bpod GUI (Figure S6D, green box). All port LEDs will flash once to signal the start of the behavioral session.

Note: The flowchart for the code structure is shown in Figure S7.

• 48.Behavioral results from the ongoing session will appear in the pop-up window, along with a timer showing approximate time remaining. (Figure S6E).
• 49.Allow the mouse to perform for 1 h in a dark, quiet room.
• 50.After an hour, stop video recording, remove mouse from behavioral assay and return to a cage.
• 51.Record number of successful sucrose trials.
• 52.Wipe down the operant chamber with deodorizer.
• 53.At the end of each day, clean the operant chamber with soap and water.
• 54.Return mice to animal facility.
• 55.Repeat daily until the mouse has reached a minimum of 40 successful sucrose trials in a single hour-long behavioral session. Advance mouse to the next training phase. (See problem 9 and 10).

Stage 5: Partial water access—fourth training phase

Timing: 5 days

This section describes how mice are trained to associate nose-poking individual choice ports to gain access to either a sucrose reward or a social reward (a social target in the social reward zone).

• 56.Transport mice from animal facility to behavioral testing room. Approximately 10 min prior to the initiation of the behavioral session give each mouse 500 μL of water.
• 57.Initiate behavioral session. Open MATLAB and type Bpod into the command window.
• 58.In the protocol launch manager, select ‘TwoChoice’ in the dropdown menu to the right of ‘Choose protocol to run’, enter the animal ID below and click the next button. (Figure S8A).
• 59.On the following screen, set the protocol parameters as desired and click the next button. (Figure S8B).
• 60.On the following screen, confirm that the recording settings are correct. (Figure S8C).
• 61.Open Pylon software, confirm settings and ensure that operant assay is centered in the camera viewer.
• 62.Place the social target in the social target box and position the platform so that the metal grating of the social target box is aligned to the opening in the operant chamber.
• 63.Test that the gate opens and closes properly without getting stuck on the social target chamber by clicking the BNC1 button on the Bpod GUI.
• 64.Place experimental mouse in operant assay.
• 65.Start video recording.
• 66.Start the behavioral session by pressing the play button on the Bpod GUI (Figure S8D, green box). All port LEDs will flash once to signal the start of the behavioral session.

Note: The flowchart for the code structure is shown in Figure S9.

• 67.Behavioral results from the ongoing session will appear in the pop-up window, along with a timer showing approximate time remaining. (Figure S8E).
• 68.Allow the mouse to perform for 1 h in a dark, quiet room. Check occasionally to ensure that the social target remains in the social target box and that the gate closes fully at the end of social trials.
• 69.After an hour, stop video recording, remove the experimental mouse from the behavioral assay and return to a cage.
• 70.Remove the social target from the social target box and return to a separate cage.
• 71.Record number of successful sucrose trials and social trials.
• 72.Wipe down the operant chamber and social target box with deodorizer.
• 73.At the end of each day, clean the operant chamber and social target box with soap and water.
• 74.Return mice to animal facility.
• 75.Repeat daily for 5 hour-long behavioral sessions. Analyze each behavioral session to confirm the social reward latency is decreasing with training. (Figure S10).
• 76.Once a mouse has completed 5 behavioral sessions, remove it from the water deprivation schedule and provide ad libitum water access outside of the assay.
• 77.Advance mouse to the experimental phase. (See problem 11).

Stage 6: Full water access—first experimental phase

Timing: 5 days

This section describes the experimental phase of the two choice behavioral assay where mice are removed from the water deprivation schedule and tested to determine the amount of social and nonsocial reward-seeking on the two choice operant assay under control conditions.

• 78.Transport mice from animal facility to behavioral testing room.
• 79.Initiate behavioral session. Open MATLAB and type Bpod into the command window.
• 80.In the protocol launch manager, select ‘TwoChoice’ in the dropdown menu to the right of ‘Choose protocol to run’, enter the animal ID below and click the next button. (Figure S8A).
• 81.On the following screen, set the protocol parameters as desired and click the next button. (Figure S8B).
• 82.On the following screen, confirm that the recording settings are correct. (Figure S8C).
• 83.Open Pylon software, confirm settings and ensure that operant assay is centered in the camera viewer.
• 84.Place the social target in the social target box and position the platform so that the metal grating of the social target box is aligned to the opening in the operant chamber.

CRITICAL: We use non-cagemate, age- and sex-matched conspecifics as social targets for our behavioral experiments, which resulted in approximately five additional mice per experimental animal. Other types of mice may be used as social targets (opposite-sex conspecifics, cagemates, etc.) depending on the needs of the researcher, but this has not yet been tested using our two choice operant assay. We ensure that the experimental mouse has access to a novel social target on each day of the phase and limit the use of each social target mouse to once a day.

• 85.Test that the gate opens and closes properly without getting stuck on the social target chamber by clicking the BNC1 button on the Bpod GUI.
• 86.Place experimental mouse in the operant assay.
• 87.Start video recording.
• 88.Start the behavioral session by pressing the play button on the Bpod GUI (Figure S8D, green box). All port LEDs will flash once to signal the start of the behavioral session.

Note: The flowchart for the code structure is shown in Figure S9.

• 89.Behavioral results from the ongoing session will appear in the pop-up window, along with a timer showing approximate time remaining. (Figure S8E).
• 90.Allow the mouse to perform for 1 h in a dark, quiet room. Check occasionally to ensure that the social target remains in the social target box and that the gate closes fully at the end of social trials.
• 91.After an hour, stop video recording, remove mouse from behavioral assay and return to a cage.
• 92.Remove the social target from the social target box and return to a separate cage.
• 93.Record number of successful sucrose trials and social trials.
• 94.Wipe down the operant chamber and social target box with deodorizer.
• 95.At the end of each day, clean the operant chamber and social target box with soap and water.
• 96.Return mice to animal facility.
• 97.Repeat daily for 5 hour-long behavioral sessions while mice have ad libitum access to water outside the assay.
• 98.By the third behavioral session, social and sucrose reward-seeking behavior should be stabilized and relatively equivalent. Behavioral outcomes from sessions 3-5 of full water access establish the baseline for reward-seeking behavior of mice under control conditions.
• 99.Once a baseline of reward-seeking behavior has been established, mice can be run through additional experimental manipulations on the two choice operant assay (See problem 12).

Stage 7: Additional experimental manipulations—second experimental phase

Timing: 5–7 days

This section describes how researchers can optogenetically manipulate neural activity during the two choice operant assay.

• 100.Transport mice from animal facility to behavioral testing room.
• 101.Initiate behavioral session. Open MATLAB and type Bpod into the command window.
• 102.In the protocol launch manager, select ‘TwoChoice’ in the dropdown menu to the right of ‘Choose protocol to run’, enter the animal ID below and click the next button. (Figure S11A).
• 103.On the following screen, set the protocol parameters as desired. To perform optogenetic stimulation experiments, select ‘Opto’ from the dropdown menu next to ‘Choose recording type’ and click the next button. (Figure S11B).
• 104.On the following screen, set the laser stimulation parameters as desired, including the percentage of stimulated trials. (Figure S11C).
• 105.On the following screen, confirm that the recording settings are correct. (Figure S11D).
• 106.Open Pylon software, confirm settings and ensure that operant assay is centered in the camera viewer.
• 107.Place the social target in the social target box and position the platform so that the metal grating of the social target box is aligned to the opening in the operant chamber.

CRITICAL: As mentioned in Stage 6, we used non-cagemate, age- and sex-matched conspecifics as social targets for our behavioral experiments. Additionally, we ensure that the experimental mouse has access to a novel social target on each day of the phase and limit the use of each social target mouse to once a day.

• 108.Test that the gate opens and closes properly without getting stuck on the social target chamber by clicking the BNC1 button on the Bpod GUI.
• 109.For optogenetic manipulation experiments, plug one end of a TTL cable into the BNC2 channel on the Bpod state machine and the other end into the light source.

CRITICAL: Take appropriate measures to ensure user safety when operating light sources, especially lasers. Users should be trained in safe operating practices before handling lasers.

• 110.Turn on the light source. No light should be emitted until signaled by the Bpod.
• 111.Place experimental mouse in the operant assay.
• 112.Start video recording.
• 113.Start the behavioral session by pressing the play button on the Bpod GUI (Figure S11E, green box). All port LEDs will flash once to signal the start of the behavioral session.
• 114.Behavioral results from the ongoing session will appear in the pop-up window, along with a timer showing approximate time remaining. Stimulated trials will be denoted with asterisks. (Figure S11F).
• 115.Allow the mouse to perform for 1 h in a dark, quiet room. Check occasionally to ensure that the social target remains in the social target box and that the gate closes fully at the end of social trials.
• 116.After an hour, stop video recording, remove mouse from behavioral assay and return to a cage.
• 117.Remove the social target from the social target box and return to a separate cage.
• 118.Record number of successful sucrose trials and social trials.
• 119.Wipe down the operant chamber and social target box with deodorizer.
• 120.At the end of each day, clean the operant chamber and social target box with soap and water. Turn off the light source.
• 121.Return mice to animal facility.

Expected outcomes

There are several behavioral metrics that can be used to quantify the behavioral data collected using our two choice operant assay. In particular, we provide the instructions and tools to quantify the number of successful trials, choice latency, reward latency and number of reward fails of the two trial types for each hour-long behavioral session administered daily (see quantification and statistical analysis section below). Additional metrics, such as distance traveled within the operant arena and time spent in the social zone, can also be quantified. All data collected in each behavioral session is saved in a MATLAB file that can be easily accessed and exported into a preferred file type. Additionally, the summary behavioral data files displayed during each session are automatically saved at the end of the session as a MATLAB fig file and as a jpeg file. Behavioral videos are analyzed using Bonsai software²² (Figure S10). Statistical analyses, t-tests and analyses of variance with post hoc t-tests are performed on the behavioral data using MATLAB software.

This operant assay is versatile and can be tailored to the experimental needs of the researcher. There are multiple behavioral parameters that can be manipulated. For example, we show that reward-seeking on this assay is sensitive to internal state manipulations, such that thirsty mice preferentially increase sucrose reward-seeking behavior.¹ It is also possible to change the availability and type of reward on this assay. By changing the operant assay parameters in the Bpod GUI, rewards can be also withheld on specific trials or the time and effort required to access a reward can be increased. The researcher can determine the type of social or nonsocial reward to use on this assay and substitute different social targets or liquid solutions as necessary. Specifically, different types of social rewards can be implemented by varying the familiarity or sex of conspecifics. Our operant assay can also be paired with a variety of experimental manipulations. In particular, our protocol enables neural recordings and optogenetic experiments which allows for the alignment of neural data or laser stimulation to specific behavioral events. There is also a wide range of genetic tools that are available in the C57 strain of mice. Therefore, experimenters can utilize the genetic tractability of C57 mice to study the neural circuits that underlie social and nonsocial reward-seeking on our assay.⁸ There are also numerous mouse models of neurodevelopmental and neuropsychiatric disorders, such as Autism Spectrum Disorders (ASD), which involve dysfunctional social and nonsocial reward processing that can be studied using our assay^23,24,25. The investigation of such disruptions in reward processing may provide insights into the development of therapeutic interventions for such disorders.

Using the protocol that we describe here, we have shown that male and female C57 mice can be trained to nose-poke for access to a social target and sucrose reward on the same operant assay, more so than an empty cage or novel object (Figure 4). We demonstrate how to construct the operant chamber using inexpensive components, how to automate behavioral data collection using a programmable state machine and how to perform behavioral testing in C57 mice. We show that we can alter reward-seeking behavior in both male and female mice on this two choice operant assay through optogenetic inhibition of the medial prefrontal cortex (Figure 5). We have also performed our own calcium imaging and optogenetic activation experiments in both male and female C57 mice (described in detail in Isaac et al.¹) to confirm that our operant chamber and the data acquisition software are optimized for such experiments. We provide the requisite code to perform these experiments. As a result, this automated two choice operant assay can be used to assess reward-seeking behavior and the neural circuits that underlie it.

Figure 4.

Male and female mice find both social and sucrose stimuli positively reinforcing on a two choice operant assay

(A and E) Fully trained male mice (A) complete an equivalent number of successful sucrose and social trials, while female mice (E) complete more successful social than sucrose trials. Paired t test (male: number of trials: p = 0.73; female: p = 4.91∗10⁻⁴).

(B and F) Male and female mice (B and F) show a similar choice latency for sucrose and social choices. Paired t test (male: p = 0.12; female: p = 0.93).

(F and J) Male mice (C) were slightly faster to consume sucrose reward than social reward, while female mice (G) had equivalent latencies to consume social and sucrose reward. Paired t test (male: p = 2.04∗10-8; female: p = 0.093).

(D and H) Male (D) and female (H) mice made fewer social reward fails than sucrose reward fails. Paired t test (male: p = 1.00∗10-7; female: p = 4.10∗10-6). N = 21 male mice, 12 female mice, 3 behavioral sessions per mouse.

(I) Under full water access conditions, mice run on the two choice operant assay with a social target (SG) completed fewer successful sucrose trials when compared to mice from the object group (OG) and a similar number of successful sucrose trials when compared to mice from the empty cage group (ECG). One-way ANOVA (p = 1.34∗10⁻⁸) with post-hoc t tests (SG vs. OG: p = 8.44∗10⁻⁹; SG vs. ECG: p = 0.052; OG vs. ECG: p = 0.10).

(J–L) There was no difference in sucrose choice latency (J), sucrose reward latency (K) or number of sucrose reward fails (L) between groups. One-way ANOVA (j, p = 0.080; k, p = 0.19; l, p = 0.078).

(M) Mice run on the two choice assay with a social target (SG) completed more social trials than mice run with an empty cage (ECG) or an object (OG). One-way ANOVA (p = 3.30∗10⁻¹⁰) with post-hoc t tests (SG vs. OG: p = 2.63∗10⁻⁹; SG vs. ECG: p = 1.17∗10⁻⁴; OG vs. ECG: p = 0.87).

(N and P) Across all conditions, mice showed similar social choice latency (N) and number of social reward fails (P). One-way ANOVA (n, p = 0.92; p, p = 0.44).

(R) Mice run with a social target (SG) showed similar social reward latency to mice run with an empty cage (ECG) and decreased social reward latency compared to mice run with an object (OG). One-way ANOVA (p = 2.00∗10⁻⁴) with post-hoc t tests (SG vs. OG: p = 1.04∗10⁻⁴; SG vs. ECG: p = 0.46; OG vs. ECG: p = 0.19). SG: n = 21 mice, ECG: n = 4 mice, OG: n = 10 mice, 3 sessions per mouse. Boxplots: center line denotes median, box edges indicate the 25th and 75th percentiles and whiskers extend to ± 2.7σ. ∗p < 0.05. Adapted from ref.¹

Figure 5.

Inhibition of mPFC neurons during the reward period of the two choice assay disrupts reward-seeking behavior

(A) Schematic of the viral strategy (left panel) used to label mPFC neurons with either GtACR inhibitory opsin (GtACR) or mCherry (control). Example histology showing GtACR expression (GtACR in red, DAPI labeling of cell nuclei in blue) and ferrule placement in the mPFC at 4× magnification (right panel). Scale bar: 500 μm.

(B) Reconstruction of optic ferrule placement in male (left, n = 13 mice) and female (right, n = 13 mice) mice using WholeBrain software.²⁶ Each colored dot shows the position of the optic ferrule in the Allen Mouse Brain Common Coordinate Framework. Red dots indicate GtACR mice (male: n = 7 mice; female: n = 7 mice), black dots indicate mCherry mice (male: n = 6 mice; female: n = 6 mice). Coronal slice is 1.945 mm anterior to bregma.

(C and G) Optogenetic inhibition of mPFC neurons during the reward period increases reward latency on both social and sucrose trials compared to mCherry controls in male (C) and female (G) mice. Two-factor ANOVA with virus (GtACR/mCherry) and trial type (sucrose/social) as factors (C, interaction: p = 0.67, virus: p = 0.0004, trial type: p < 0.00001; G, interaction: p = 0.83, virus: p = 2.46∗10⁻⁸, trial type: p = 2.29∗10⁻¹²) with post-hoc unpaired t tests comparing virus within trial type (C, sucrose: p = 0.0014, social: p = 0.021; G, sucrose: p = 5.46∗10⁻¹⁰, social: p = 0.0014).

(D and H) Optogenetic inhibition also increases the number of social reward fails compared to mCherry controls in male mice (D) and both sucrose and social reward fails in female mice (H). Two-factor ANOVA with virus (GtACR/mCherry) and trial type (sucrose/social) as factors (D, interaction: p = 0.50, virus: p = 0.0039, trial type: p = 0.37; H, interaction: p = 0.055, virus: p = 0.0002, trial type: p = 0.0001) with post-hoc unpaired t tests comparing virus within trial type (D, sucrose: p = 0.13, social: p = 0.0084; H, sucrose: p = 0.002, social: p = 0.026).

(E and I) Optogenetic inhibition of mPFC neurons had no effect on average distance traveled during social or sucrose trials in both sexes. Two factor ANOVA (E, interaction: p = 0.18, virus: p = 0.16, trial type: p = 3.01∗10⁻¹⁰⁶; I, interaction: p = 0.76, virus: p = 0.42, trial type: p = 1.97∗10⁻⁸⁹).

(F and J) Optogenetic inhibition of mPFC neurons caused a decrease in the time spent in the social zone during the reward period when compared to mCherry controls in male (F) but not female (J) mice. Unpaired t test (male: p = 0.013; female: p = 0.31). ∗p < 0.05. Boxplots: center line denotes median, box edges indicate the 25th and 75th percentiles and whiskers extend to ± 2.7σ. Adapted from ref.¹

Quantification and statistical analysis

Following each behavioral session, output files can be accessed and analyzed to quantify a number of behavioral metrics. The summary behavioral data from each session are saved as .fig and .jpeg files along with the corresponding session data in each Data folder (<localdirectorypath>\Two-choice-operant-assay>\Bpod>\BpodLocal>\Data>\<mouseid>>\SessionData>\). These files contain information regarding the total number of social and nonsocial trials completed during each behavioral session, as well as the number of sucrose reward fails, and stimulated versus non-stimulated trials if optogenetic experiments are being performed. The MATLAB file contains time stamps of each trial event, such as port entry, which can be used to calculate choice latency of all trial types and reward latency of sucrose trials. Videos are analyzed using Bonsai software²² (Figure S10) to track the centroid of the experimental mouse, trial start (change in light levels of choice port LEDs) and entry into social and sucrose reward zones. This analysis is combined with the time stamps from the Bpod data to quantify the social and nonsocial reward-seeking behavior of mice during each hour-long behavioral session. We define the reported behavioral metrics as follows. Trials are considered successful when the mouse nose-pokes either choice port and enters the corresponding reward zone while the reward is available. Choice latency is the time from when the choice ports illuminate to signal trial start to when a mouse nose-pokes either choice port. Reward latency is the time from when the mouse nose-pokes the choice port to when it enters the corresponding reward zone. Reward fails occur when the mouse nose-pokes the choice port and fails to enter the corresponding reward zone while the reward is available.

Limitations

This operant assay examines social and nonsocial reward-seeking. As a result, it requires additional mice as social targets along with the experimental mice that are being tested on the operant assay. In our experiments, we used non-cagemate, age- and sex-matched conspecifics for each behavioral session, so we used approximately five mice per experimental mouse. However, it is possible to use various types of social targets (opposite-sex conspecifics, cage mates, pups, aggressors, etc.) depending on the experimental design, which can further increase or decrease the number of additional mice that are needed as social targets. As with the three chamber and social CPP assays, the social behaviors on the two choice operant assay are constrained by the social target being confined within a social target box. Although the two animals can interact through the metal grating, it is difficult to determine the nature of the interactions. A full range of social behaviors can be characterized using naturalistic assays. Additionally, in the current experimental design, the training stages require water deprivation, and the behavioral outcomes are sensitive to internal state. Therefore, once mice are removed from water deprivation during the full water access condition, they should be given at least two days to establish a baseline of social and sucrose reward-seeking behavior. This can be addressed by training animals without water deprivation over a longer timescale.

Troubleshooting

Problem 1

The slider has an uneven base. Related to operant assay assembly step 8.

Potential solution

The four screws in the base of the camera slider are used to increase its height. We find that the most stable setting for the slider is with all four screws tightened in its lowest position. This can be changed if deemed necessary.

Problem 2

The L-bracket is not adhering to the gate. Related to operant assay assembly step 14.

Potential solution

We encourage sanding the L-bracket as it promotes adhesion to the aluminum. Alternatively, an unpainted L-bracket (Amazon - ASIN - B07MZLLVXC) can be used.

Problem 3

The holes created in the side of the social target box exceed the diameter of the aluminum rods used as restraints. Related to operant assay assembly step 20.

Potential solution

Use super glue to adhere the bar to the holes.

Problem 4

The configuration of pins corresponding to each coil in the stepper motor is unknown. Related to operant assay assembly step 25c.

Potential solution

If you are unaware of which pair of pins correspond to each coil on the stepper motor, then connect the ends of any two of the pins with each other using a jumper wire and check the resistance on the motor. Wires from the same coil provide more resistance to movement of the motor as compared to wires from different coils.

Problem 5

MATLAB has the error message when running the command “Bpod” “Unidentified function or variable 'Bpod'”” or “Bpod device not found”. Related to software setup step 30a

Potential solution

• •Verify that the Bpod folder has been added to the path.
• •Verify that the Bpod is plugged into the computer. If the error persists after testing the connection, identify the serial COM port that the BPOD is plugged into (Windows search bar -> Device Manager -> Ports (COM & LPT) - > Identify the COM number here) and specify the COM number while calling the Bpod function in MATLAB as Bpod(‘COM#’). Visit the Sanworks website for additional instructions on troubleshooting Bpod installation.

Problem 6

The sucrose reward port does not dispense liquid. Related to testing the operant assay step 40 and Stage 1 step 4b.

Potential solution

• •If you do not hear an audible click from the solenoid valve (sound of the valve opening), make sure that the wall mounted power source to the Bpod is connected and is adequately delivering power. If the solenoid valve is open and still does not dispense liquid, encourage the flow of liquid from the reservoir to the valve by creating negative pressure in the end connected to the input in the solenoid valve.
• •Continuous dispensing of sucrose can clog the internal circuitry of the solenoid valve. To avoid clogs, implement a flush protocol with distilled water to clear out any sucrose in the system at the end of each day and dry out the valve after completion of a behavioral study. Alternatively, using an isolated solenoid valve (The Lee company - VHS series 2-way dispense solenoid valve) designed for dispensing liquids can help prevent valve failure.

Problem 7

The silicone tubing of the sucrose reward port has overgrowth. Related to testing the operant assay step 41 and Stage 1 step 4b.

Potential solution

• •The use of sucrose solution can lead to overgrowth in the silicone tubing and the reward port valve. We recommend regularly flushing the tubes with distilled water and thoroughly cleaning them to prevent overgrowth or valve dysfunction.

Problem 8

The social reward gate does not open/does not open sufficiently. Related to testing the operant assay step 42 and Stage 1 step 4c.

Potential solution

• •If the stepper motor of the social gate is not engaged, reupload the .ino code to the Arduino board. Ensure that the microstep driver’s power source is connected.
• •If the gate does not open correctly, you can either increase the current supply to the stepper motor by altering the orientation of the last 3 DIP switches present on the microstep driver (refer to the current setting reference table on the motor), or the movement resolution (refer to the subdivision setting reference table on the motor) or both as necessary.

Problem 9

Failure to progress through training phases, especially opposing port training phase. Related to Stage 4 step 55

Potential solution

In this case, mice may be insufficiently water deprived or overtrained on the sucrose reward port. To address this, ensure that mice are appropriately water deprived. If mice do not engage with the choice port for several successive sessions, the choice port can be primed with a small drop of liquid to encourage engagement. However, make sure that this is noted and reported in the documentation of behavioral training.

Problem 10

Low engagement with sucrose reward or excessive number of sucrose reward fails. Related to Stage 4 step 55.

Potential solution

In this case, mice may have developed an aversion to the reward port due to contamination or overgrowth. To address this, clean sucrose reward port, tubing and reservoir thoroughly then remake the sucrose solution. Additionally, the sucrose port may not be detecting nose-pokes or dispensing liquid correctly. In this case, test the valve to ensure that it is dispensing sucrose, reposition the IR light and test the sensor for beam break with a gloved finger.

Problem 11

Low engagement with social reward. Related to Stage 5 step 77.

Potential solution

In this case, mice may be excessively water deprived. Track weight daily to ensure that there is no excessive weight loss with water deprivation. Engagement with social reward should increase once mice are removed from water deprivation during full water access phase.

Problem 12

Low engagement with two choice operant assay. Related to Stage 6 step 99.

Potential solution

We recommend that mice complete at least 20 total successful trials during each session. If mice continue to complete fewer than 20 successful trials in each behavioral session, consider the conditions within the behavioral room that may be affecting the stress levels of the mice. Run behavioral experiments in a dark room with as little ambient noise and experimenter intervention as possible. Identify and exclude mice that consistently fail to engage with the assay. However, this should be done sparingly and with careful consideration of all the factors that might contribute to low engagement.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Malavika Murugan (mmurug5@emory.edu).

Technical contact

Technical questions on executing this protocol should be directed to and will be answered by the technical contact, Hymavathy Balasubramanian (hymavathy.balasubramanian@emory.edu).

Materials availability

This study did not generate new unique reagents.

Data and code availability

The necessary codes for data acquisition, and analysis for the two choice operant assay can be downloaded from Zenodo: https://doi.org/10.5281/zenodo.14975400.

Acknowledgments

This work was supported by the National Institutes of Health grants R01MH130755 (M.M.) and F31MH133373 (J.I.), the Emory Conte Center Pilot Project Grant (M.M.), and the Emory University Research Committee Grant (M.M.). We would like to thank Jan Hawes and Dr. Yunmiao Wang for their technical assistance with this project.

Author contributions

M.M., J.I., and H.B. conceived the project. J.I., H.B., and N.S. designed and improved the operant behavioral setup. J.I. and S.C.K. ran the behavioral experiments. H.B., with assistance from N.S., developed the code. J.I., S.C.K., and H.B. analyzed the data. J.I., H.B., and M.M. wrote the manuscript.

Declaration of interests

The authors declare no competing interests.