A System for Tracking the Dynamics of Social Preference Behavior in Small Rodents

Summary

Described here is a novel automated experimental system that offers an alternative to the three-chamber test and also solves several caveats. This system supplies multiple behavioral parameters that enable rigorous analysis of small rodent behavioral dynamics during the social preference and social novelty preference tests.

Abstract

Exploring the neurobiological mechanisms of social behavior requires behavioral tests that can be applied to animal models in an unbiased and observer-independent manner. Since the beginning of the millennium, the three-chamber test has been widely used as a standard paradigm to evaluate sociability (social preference) and social novelty preference in small rodents. However, this test suffers from multiple limitations, including its dependence on spatial navigation and negligence of behavioral dynamics. Presented and validated here is a novel experimental system that offers an alternative to the three-chamber test, while also solving some of its caveats. The system requires a simple and affordable experimental apparatus and publicly available open-source analysis system, which automatically measures and analyzes multiple behavioral parameters at individual and population levels. It allows detailed analysis of the behavioral dynamics of small rodents during any social discrimination test. We demonstrate the efficiency of the system in analyzing the dynamics of social behavior during the social preference and social novelty preference tests as performed by adult male mice and rats. Moreover, we validate the ability of the system to reveal modified dynamics of social behavior in rodents following manipulations such as whisker trimming. Thus, the system allows for rigorous investigation of social behavior and dynamics in small rodent models and supports more accurate comparisons between strains, conditions, and treatments.

Introduction

Revealing the biological mechanisms underlying neurodevelopmental disorders (NDDs) is one of the main challenges in the field of neuroscience¹. Addressing this challenge requires behavioral paradigms and experimental systems that typify the behavior of rodents in a standard and unbiased manner. An influential study published more than a decade ago by Moy and colleagues² presented the three-chamber test. Since then, this test has been widely used to investigate social behavior in rodent models of NDDs. This test evaluates two innate tendencies of rodents: 1) to stay in the proximity of a social stimulus over an object (sociability, also termed social preference [SP]), and 2) to prefer the proximity of a novel social stimulus over a familiar one (social novelty preference [SNP])^3,4. Several subsequent studies suggested methods of automated analysis of the three-chamber test using computerized methods^5,6.

This test still suffers from several caveats. First, it principally examines social place preference rather than the motivation of the subject to directly interact with a social stimulus, although some groups also measure olfactory investigation (sniffing) time, either manually⁷ or using commercial computerized systems^8,9,10. Second, the three-chamber test is mostly used to measure the total time spent by the subject in each chamber, and it neglects behavioral dynamics. Finally, it relies on only one aspect of the social behavior, which is the time spent by the subject in each chamber (or sniffing time, if measured).

Here we present a novel and affordable experimental system that is an alternative to the three-chamber apparatus. It also allows performance of the same behavioral tests while solving the abovementioned caveats. The presented behavioral system automatically and directly measures the investigative behavior of a rodent towards two stimuli. Additionally, it analyzes the behavioral dynamics in an observer-independent manner. Moreover, this system measures multiple behavioral parameters and analyzes these at both individual and population levels; thus, it supports a rigorous analysis of social behavior and its dynamics during each test. Furthermore, random repositioning of the chambers in opposite corners of the arena during the various test stages neutralizes any effects of spatial memory or preference. This system may also be used for other discrimination tests, such as sex discrimination. The custom apparatus is easy to produce, and the analysis system is publicly accessible as an open-source code, thereby allowing its use in any laboratory. We demonstrate the ability of this system to measure multiple parameters of social behavior in rodent strains with distinct fur colors during the social preference and social novelty preference tests. We also validate the ability of the system to reveal modified dynamics of social behavior in rodents following manipulations, such as whisker trimming.

TrackRodent software: three algorithms were written in MATLAB (2014a-2019a) to track the experimental subject and its interactions with the stimuli. All algorithms were deposited in GitHub, found at <https://github.com/shainetser/TrackRodent>. The main aim of all four algorithms is to track the contours of the subject's body to detect any direct contact with the stimuli areas.

Body-based algorithm: this algorithm has three versions that track the contours of an unwired dark mouse on a white background (BlackMouseBodyBased), a white mouse on a dark background (WhiteMouseBodyBased), or a white rat on a dark background (WhiteRatBodyBased). The graphical user interface (GUI) of the software requires that the experimenter chooses an experiment using either mice or rats and then selects the correct code. For each version of the algorithm, there are two optional codes: one that presents the tracking process on the screen while it performs the analysis, and one that does not (hence, it runs faster and is termed "fast"). For example, the names of the relevant codes for the BlackMouseBodyBased algorithm are: "BlackMouseBodyBased23_7_14" and "BlackMouseBodyBased23_7_14_Fast". All algorithms ending with "fast" do not show the tracking online, and users must directly save the data to the results file (.mat file). All body-based algorithms require setting a single threshold ("low threshold" in the software's GUI) to detect the body of the subject.

Head-directionality based algorithm: the second algorithm, which is available only for black mice, is based on the body-based algorithm, in addition to determining head directionality. This algorithm detects the interactions of subject's head with the "stimuli" areas, thus avoiding false positives that can arise from random contacts of the subject' with these areas. For this algorithm, two detection thresholds of mouse body contours are defined: high threshold, which includes the brighter tail of black mice, and low threshold, which includes the body without tail. Thereafter, the algorithm fits an ellipsoid to the detected boundaries using the lower threshold and defines the location of the mouse head and tail (with no distinction between the two). The final discrimination between the tail and head is based on the boundaries defined by the higher threshold.

Wired animal algorithm: the third algorithm aims to minimize artifacts resulting from cables (i.e., electrical wire or optical fiber) connected to the animal, allowing analysis of the animal's behavior while connected to a cable. This algorithm has codes only for black mice and white rats. The code for rats requires the experimenter to define both low and high thresholds, while the mouse code requires only a low threshold.

Protocol

All methods described have been approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Haifa.

1. Experimental Set-up

1. Arena
   1. Construct the experimental arena for mice (Figure 1A,D) by placing a white or black (depending on the animal's color) plexiglass box (37 cm x 22 cm x 35 cm) in the middle of an acoustic chamber (60 cm x 65 cm x 80 cm, made of 2 cm thick wood coated on the inside with 2 cm thick foam). For the light, remove a strip (2 cm wide, 10 cm below the ceiling of the chamber) of the foam around the acoustic chamber (besides the door), and attach an LED strip with either red or white light bulbs. Make sure the light is as uniform as possible around the arena to prevent reflections.
   2. Construct the arena for rats similarly to the one described above for mice, with different dimensions that appropriately fit the size of Sprague-Dawley (SD) rats (Figure 1G). Place a black plexiglass box (50 cm x 50 cm x 40 cm) in the middle of an acoustic chamber (90 cm x 60 cm x 85 cm, made of 2 cm thick wood coated on the inside with 2 cm thick foam).
2. Chambers
   1. For mice, create two black or white (depending on the fur color) triangular chambers (12 cm isosceles, 35 cm height, with the floors closed) from 6 mm thick plexiglass. Locate them in two randomly selected opposite corners of the arena (Figure 1B,E). Stick a metal mesh (18 mm x 6 cm; 1 cm x 1 cm holes) in the lower part of each chamber using epoxy glue to allow direct interactions with the stimulus through the mesh (Figure 1C,F). Mark each chamber in a manner that allows discrimination from the others in a video without giving clues to the subjects (see Figure 1B,E for examples).
      NOTE: Each chamber will eventually contain a social stimulus (mouse) or object (plastic toy, 5 cm x 5 cm x 5 cm, with a distinct shape and color; Figure 1C,I insets). Let the epoxy glue smell evaporate for at least a week before usage.
   2. For rats, create two black triangular chambers (20.5 cm isosceles, 40 cm height, made of 6 mm thick plexiglass, with the floors closed) and place them in two randomly selected opposite corners of the arena (Figure 1H), each with a metal mesh (25 cm x 7 cm; 2.5 cm x 1 cm holes) covering its lower portion (Figure 1I).
3. Place a high quality monochromatic camera, equipped with a wide-angle lens, at the top of the acoustic chamber and connect it to a computer to enable clear viewing and recording of the subject's behavior using commercial software (see Table of Materials for suggestions).

2. Behavioral Paradigm

NOTE: Steps 2.1-2.7 describe the behavioral paradigm for mice. See section 2.8 for specific instructions involving rats.

1. Make sure that the cages of all animals (subjects: 2-4 month-old male mice; stimuli: 21-30 day-old juvenile mice) remain in the experimental room for at least 1 h before beginning the behavioral experiment.
2. Following the acclimation period, insert two empty chambers into the arena randomly at two opposite corners. Place the subject in the middle of the arena for 15 min of habituation. During that time, place the two social stimuli, each in a different chamber located out of the arena for habituation. Place an object (a plastic toy, 5 cm x 5 cm x 5 cm, with a distinct shape and color) in another chamber.
3. To perform the social preference (SP) test, start video recording and keep recording until the end of the test.
4. Remove the two empty chambers and immediately insert the object and one of the social stimuli, each in a distinct chamber. Locate these chambers randomly in the opposite corners of the arena that were empty during habituation. Allow the subject to interact with the stimuli for the 5 min of the SP test. At the end of the test, stop recording.
5. Following the SP test, remove the stimuli-containing chambers from the arena and leave the subject in the empty arena for 15 min. Clean the chambers from outside with 10% ethanol wipes.
6. To perform the social novelty preference (SNP) test, start video recording and insert two chambers into the arena: one containing the same social stimulus used for the SP test (familiar stimulus), and the other containing the novel social stimulus. Place these chambers randomly in two opposite corners of the arena, making sure these locations were not used for the SP test. Allow the subject to interact with the stimuli for the 5 min of the SNP test.
7. At the end of the SNP test, stop video recording, remove the subject and chambers from the arena, and place the subject back in its home cage. Leave the stimuli in the chambers for the next experiment (with another subject) or return them to their home cages. Clean the arena and chambers with running water followed by 10% ethanol and let dry.
8. Behavioral paradigm for rats
   1. For rats, repeat the behavioral paradigm described in steps 2.1-2.7, with two modifications: 1) handle the rat subjects and habituate the social stimuli to the chambers for 2 days (10 min every day) prior to the experiment; and 2) extend the SP test for 15 min to give the rats a longer period of exposure to social stimuli. Later, restrict the analysis of the SP test to the initial 5 min.
      NOTE: At least one arena and five chambers are needed in order to run a single session.

3. Using the TrackRodent GUI for Behavioral Analysis

NOTE: See the upper panel of Figure 2A for the TrackRodent GUI.

1. Open MATLAB (tested with 2014a-2019a) and choose the TrackRodent folder.
2. Add all subfolders to the working path by right-clicking on each folder and selecting Add to Path | Selected Folders and Subfolders.
3. Type TrackRodent in the command window and press Enter.
4. Upload a single file or multiple video files (AVI or MP4 format) by selecting Load session file (AVI).
5. A movie inspector, allowing the inspection of the video clip frame-by-frame, will immediately be opened for the first file in the list (Figure 2A). Use it to examine the video clip and define the first and last frames of the segment to be analyzed. Record the numbers of these frames, which will be required later. Close the window when done.
6. For inspecting additional video files, open the video inspector at any time by pressing Inspect movie and selecting a specific video file.
7. Select the species tested (mouse or rat; mouse is the default).
8. Exclude all the areas that may interrupt the tracking, in accordance with the colors of the subject and arena (black or white).
   1. To exclude a given area, press Exclude area, and after the cursor changes to a cross shape, mark all corners of the area for exclusion. When done, right-click on the mouse, then double left-click the center of the marked area. The excluded area will become a shade of red on the screen. Repeat this procedure to exclude as many areas as needed.
9. To remove an area from exclusion, press Remove excluded area, then (using the crossed cursor) click on the area to be removed from exclusion.
10. To define each chamber as a 'stimulus' area, for automatic detection of its investigation by the subject, choose the shape of the 'stimulus' area to be either polygon or elliptical by checking the appropriate box, then pressing Stimulus X (where "X" represents 1, 2, or 3). Mark the 'stimulus' areas similarly to the excluded areas, which will then become yellow in color. For changing the location of a specific 'stimulus' area, press Stimulus X again and mark the new area location (this will automatically update the location).
    NOTE: Choose the different stimuli number in a consistent manner for all files (i.e. object as stimulus 1 for all the SP test files).
11. To track the presence of the subject in a specific virtual compartment inside the arena, choose the shape of the 'compartment' area (polygon or elliptical), then press Compartment X (where "X" represents 1, 2, 3, 4, or 5). Mark the 'compartment' areas similarly to the excluded or stimuli areas, which will then become blue in color. For changing the location of a specific 'compartment' area, press Compartment X again and mark the new area location (this will automatically update the location).
12. Choose the desired algorithm (BlackMouseBodyBased was used for the video) from the list (see available algorithms in Figure 2B).
13. Write the numbers of starting and ending frames for the analysis in the corresponding edit boxes of the GUI.
14. Choose a threshold for detecting the subject body.
    NOTE: Most algorithms use the "Low" threshold only, while the head directionality-based algorithms use the "High" threshold, as well. For the "Low" threshold, choose a level that includes the mouse/rat body without the tail (as much as possible), while the "High" threshold should also include the tail. In the case of using head directionality-based algorithms, the software will determine the head location as being opposite to the tail location.
    NOTE: The software will later on ignores small objects detected using the chosen threshold.
15. To evaluate automatic detection of the subject borders for a given threshold, insert a value to the relevant threshold field and press Enter on the keyboard.
16. When choosing multiple files, move to the next file (using the Next button at the top) and select the appropriate definitions for each file. When finished, verify the parameters and area locations for all files by moving between each, using the Previous and Next buttons at the top of the GUI.
    NOTE: The definitions of all areas and parameters are specific to a given file.
17. For starting behavioral analysis of all selected files, move to the first file and press Start.
18. At the end of the analysis, a results file (.mat file) is saved for each movie in the same directory of the movie files.
    NOTE: If the slow (non-fast) version of the algorithm is used, it will also save a version of the movie with a white cross of the center of body mass, which changes its color every frame that is detected as investigatory, unless the Save analyzed movie toggle button of the GUI is unchecked. This version of the movie (saved in the same directory, with the same name as the original movie, with the suffix 'analyzed movie') can be used offline to evaluate the quality of the automatic detection performed by the system.

4. Using the TrackRodent GUI for Results Presentation

NOTE: See the lower panel of Figure 2A for results presentation.

1. To inspect the results of each movie file, press Load results file and choose the .mat files generated by the behavioral analysis.
2. Move between the toggle buttons to examine onscreen any of the following analyses: Mouse location trace (Figure 2A); Compartments occupation along session (if 'compartments' were defined, not shown); Stimuli exploration along session (Figure 2C); Total time in compartments (if 'compartments' were defined, not shown); and Total stimuli exploration time (Figure 2D).
   NOTE: 'Stimulus' areas are areas in which the software evaluates subject interaction, while 'Compartment' areas are areas in which the software tracks the presence of the subject. Stopping the analysis using the Stop analysis button will automatically save the results generated up to the last analyzed frame. For most computers, it should be possible to upload and analyze as many as 20 movies at once (depending on computer performance).

5. Using the TrackRodentPopulationSummary GUI for Population Analysis (Figure 2E)

1. Open MATLAB (tested with 2014a-2019a) and choose the TrackRodent folder.
2. Type TrackRodentPopulationSummary in the command window and press Enter.
3. Upload multiple TrackRodent results files (.mat format) by pressing Choose results files.
4. Fill in the numbers of Last frame for analysis, Test name, Stimulus 1 name, and Stimulus 2 name.
5. Choose the desired analyses from the list of optional analyses by checking all appropriate boxes.
6. Choose the Export results to a speadsheet by checking the appropriate box to extract all the results of the checked analyses as a single spreadsheet file.
7. Press Start and wait until analysis complete.
   NOTE: This concludes the analysis. The software can be used to analyze the results of as many movie files as desired, given that they were all behaviorally analyzed using the TrackRodent software. The analysis performed by the software assumes video recording at a frame rate of 30 Hz. In the case that a different frame rate was used, multiply the time by 30 and divide by the frame rate used for recording to convert it to the correct value (s).

Results

Using the system for the social preference test in C57BL/6J mice
Figure 1 shows three versions of the experimental set-up. The first version (Figure 1A-C) is designed for mice with dark fur colors, such as C57BL/6J mice. The second (Figure 1D-F) is planned for mice with bright fur colors, such as BALB/c or ICR (CD-1) mice. The third is larger (

Discussion

The experimental system described here, which was designed as an alternative to the three-chamber apparatus^2,5, allows performance of the same tests while solving some of its limitations. The use of triangular chambers, which are located in two opposite corners of the rectangular arena limits the subject-stimulus interaction area to a well-defined plane, thus enabling precise automated analysis of investigation behavior. One advantage is the use of the analysis s...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Flea3 1.3 MP Mono USB3 Vision	FLIR (formerly PointGrey)	FL3-U3-13Y3M-C	Monochromatic Camera
FlyCap 2.0	FLIR (formerly PointGrey)	FlyCapture 2.13.3.61X64	Video recording software
Home 5 minute Epoxy glue	Devocon	20845	For gluing the metal mesh to the Plexiglas stimuli chambers
Matlab 2014-2019	MathWorks	R2014a - R2019a	Programming environment
Plexiglas boards (6 mm thickBlack or white)	Melina (1990) LTD, Israel	NaN	For arena and stimuli chambers construction
Red led strips (60 leds per meter) connected to a 12 V power supply	2012topdeal eBay supplier	NaN	For illumination of the acoustic chamber