Development of a Marmoset Apparatus for Automated Pulling to study cooperative behaviors

  1. Olivia C Meisner
  2. Weikang Shi
  3. Nicholas A Fagan
  4. Joel Greenwood
  5. Monika P Jadi
  6. Anirvan S Nandy  Is a corresponding author
  7. Steve WC Chang  Is a corresponding author
  1. Interdepartmental Neuroscience Program, Yale University, United States
  2. Department of Psychology, Yale University, United States
  3. Department of Neuroscience, Yale University, United States
  4. Wu Tsai Institute, Yale University, United States
  5. Kavli Institute for Neuroscience, Yale University School of Medicine, United States
  6. Department of Psychiatry, Yale University, United States
3 figures, 1 table and 1 additional file

Figures

Design and Structure of Marmoset Apparatus for Automated Pulling (MarmoAAP).

(A) CAD model of apparatus highlighting general apparatus layout, transparent testing boxes, frame structure, video cameras, and microphones. (B) CAD model of lever-motor assembly. Left: Side view of lever-motor assembly. Potentiometer-motor adaptor has been omitted in this view due to its obstruction of view of other components. Right: Front view of lever-motor assembly. (C) Photo of actual apparatus. (D) Photo of lever-motor assembly.

Quantitative behavioral measurements with Marmoset Apparatus for Automated Pulling (MarmoAAP).

(A) Quantification of average number of lever pulls performed per animal in Self-Reward and 1 s Mutual Cooperation sessions (mean ± s.e.m.; n = 10 Self-Reward sessions, 67 Cooperation sessions). Red dashed line indicates the average number of trials collected per session from a sample of non-automated cooperative pulling paradigm experiments (Cronin et al., 2005; Martin et al., 2021; Mendres, 2000; Plotnik et al., 2011; Range et al., 2019; Drea and Carter, 2009; Seed et al., 2008). (B) Event time series example from a Mutual Cooperation and Self-Reward session. Each bar represents a task-related event (tan: monkey 1 lever pull, brown: monkey 2 lever pull, blue: reward delivery). (C) Metrics quantifying performance on MarmoAAP from six example sessions (three Self-Reward, three Mutual Cooperation). Red diamond indicates the mean. Left: Success rate calculated as number of successful lever pulls divided by number of total lever pulls in a session. Middle: Rewards earned per working minute calculated as the average number of rewards earned in 60 s for cumulative time in a session in which a monkey had pulled a lever within 30 s of that time. Right: Inter-pull time calculated as the average amount of time between M1 and M2 lever pulls in a session. (D) Example of DLC2 labeled video frame as marmosets perform cooperative pulling task. (E) Quantification of gaze targets averaged across three example cooperation sessions (mean ± s.e.m., n = 2 marmosets with 4 sessions each). (F) Example vocalizations captured during behavioral sessions. (G) Peristimulus time histogram (mean ± s.e.m.; n = 1103 Cooperation lever pulls, 566 Self-Reward lever pulls) of chirp vocalizations from one marmoset across Self-Reward (gray) and Mutual Cooperation (blue) sessions. Red bar indicate time bins with significantly different call counts for Cooperation compared to Self-Reward task conditions (Wilcoxon rank sum test, p < 0.05).

Wireless neural data recordings with Marmoset Apparatus for Automated Pulling (MarmoAAP).

(A) Photo of marmosets performing Mutual Cooperation task with one marmoset performing the task with a head-mounted wireless recording system. (B) Example single-unit waveforms from one recording session of the orbitofrontal cortex (OFC). (C) Top: Peristimulus time histogram of an example single OFC neuron averaged across all self lever pulls in one session aligned to the time of lever pull registration (dashed line) (n = 49 lever pulls, bin size = 150 ms with 50 ms sliding window). Bottom: Raster plot of the same example OFC neuron relative to all self lever pulls (red line = lever pull registration) in one example session. (D) Top: Peristimulus time histogram of an example dorsolateral prefrontal cortex (dlPFC) multi-unit averaged across all self lever pulls in one session aligned to the time of lever pull registration (dashed line) (n = 49 lever pulls, bin size = 150 ms with 50 ms sliding window). Bottom: Raster plot of example dlPFC multi-unit relative to all self lever pulls (red line = lever pull registration) in one example session.

Tables

Table 1
Apparatus parts and information.
Apparatus itemPartManufacturer/part numberFunction
Pull Lever HeadBall HeadCustomLever grip suitable for marmosets
Pull Lever Shaft6-inch rodThorLabs / ER6-P4Holds lever head at appropriate height
Lever BumpersLoad-Rated Threaded BumperMcMaster-Carr/93115K121Bumper for lever to rest on when in starting position
Lever Assembly MotorServo MotorClearPath Integrated Servo System by Teknic/CPM-MCVC-3411S-RLNExerts rotational force on the lever
Motor Mounting PlateCustom-designed aluminum mounting plateCustom designed. Aluminum cut by water jetHolds servo motor to base frame
PotentiometerRotary Potentiometer—10k Ohm, LinearSparkFun Electronics/COM-09939Delivers real-time positional output of pull lever
Strain GaugeLoad cellSparkFun Electronics/SEN-14729Delivers real-time force reading of pull lever
Potentiometer-Motor AdaptorCustom designedCustom designed. Aluminum cut by water jetYokes potentiometer shaft and motor shaft
Potentiometer Mounting PlateCustom-designed aluminum mounting plateCustom designed. Aluminum cut by water jetHolds the potentiometer in position adjacent to the servo motor
Potentiometer Plate CrossbarsØ1/2" Optical Post, SS, 8–32 Setscrew, 1/4"-20 Tap, L=2"ThorLabs / TR2Fixes potentiometer mounting plate to motor mounting plate
CounterweightCustom-designed aluminum barCustom-designed aluminum bar cut by water jetOffsets pull lever weight to balance system
Control boardTeensy 3.2 USB Development BoardSparkFun Electronics/DEV-13736Microcontroller used to integrate motors, potentiometer, and strain gauge
BreadboardSolder-able BreadboardSparkFun Electronics/PRT-12070Connects Teensy microcontroller to sensors, facilitating control and computer integration
Syringe PumpSyringe PumpNew Era/DUAL-NE-1000XSyringe pump controlled by task code to deliver juice reward
FrameT-slotted railsMcMaster-Carr/47065T553Provide a scaffolding to support lever-motor assembly
CamerasGo-Pro CamerasGo-Pro/HERO10Record videos of marmosets performing the task
Audio RecorderVoice Recorder, 16 GBQZTELECTRONIC (via Amazon)Record vocalizations from marmosets performing the task

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Olivia C Meisner
  2. Weikang Shi
  3. Nicholas A Fagan
  4. Joel Greenwood
  5. Monika P Jadi
  6. Anirvan S Nandy
  7. Steve WC Chang
(2024)
Development of a Marmoset Apparatus for Automated Pulling to study cooperative behaviors
eLife 13:RP97088.
https://doi.org/10.7554/eLife.97088.3