Protocol for evaluating mutualistic cooperative behavior in mice using a water-reward task assay

Summary

Social cooperation is fundamentally important for group animals but rarely studied in mice because of their natural aggressiveness. Here, we present a new water-reward assay to investigate mutualistic cooperative behavior in mice. We describe the construction of the apparatus and provide details of the procedures and analysis for investigators to characterize and quantify the mutualistic cooperative behavior. This protocol has been validated in mice and can be used for investigating mechanisms of cooperation.

For complete details on the use and execution of this protocol, please refer to Zhang et al. and Wang et al.^1,2

Graphical abstract

Highlights

• •Steps for constructing behavioral apparatus for measuring cooperative behavior
• •Instructions for detecting cooperative behavior in rodents
• •Guidance on the accurate quantification of cooperative behavior in mice
• •Can be applied to other rodent models, including neuropsychiatric-related animal models

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

Social cooperation is fundamentally important for group animals but rarely studied in mice because of their natural aggressiveness. Here, we present a new water-reward assay to investigate mutualistic cooperative behavior in mice. We describe the construction of the apparatus and provide details of the procedures and analysis for investigators to characterize and quantify the mutualistic cooperative behavior. This protocol has been validated in mice and can be used for investigating mechanisms of cooperation.

Before you begin

Cooperation is defined as two or more individuals working together to achieve a common goal. Cooperative behavior increases participants’ survival fitness and thereby promotes the reproduction of species.³ Cooperation can be briefly categorized into two groups: mutualistic cooperation, in which participants gain immediate benefits,^4,5 and reciprocal altruism, where one or some participants benefit less or even have net costs at the time of behavior but get compensated later.⁶ Reciprocal altruism is pervasive in human society but is only sparsely reported in nonhuman animals, while mutualistic cooperation is widely evidenced across various animal species.⁷ For reciprocal altruism, recent fMRI studies on human subjects identified important brain regions, such as the caudate nucleus and the orbitofrontal cortex in reward processing^8,9,10; on the other hand, although seeming simpler and more experimentally tractable, the neuronal and molecular mechanisms beneath mutualistic cooperation remain unclear. This calls for the development of animal models and suitable behavioral assays that could reliably assess animal’s cooperative ability.

Cooperation is a complex behavior with an intrinsic social dimension and at the same time requires certain levels of cognitive abilities. Many experimental studies are conducted with nonhuman primates and dolphins, in which animals are expected to pull a rope or handles or push buttons.^11,12,13 Rats are gregarious animals and are found to have generalized altruistic behavior, making it an excellent model for studying cooperative behavior.^5,14 The mouse models are extensively employed in various social and cognitive behavior studies, but their use in cooperation research are rarely reported. The mice are superior in the convenient genetic editing and the established behavior-tracking techniques. Therefore, developing mouse cooperative behavior assays is particularly attractive for mechanistic studies.

Recently, we established a new water-reward assay to investigate mutualistic cooperative behavior in mice.^1,2 This paradigm required mice to be first trained in a chamber individually, where they learn to switch on a water dispenser for drinking by staying in a particular zone. Then, two trained mice were paired and put into a chamber containing two water dispensers and two zones. They could get water only if they stayed within the respective zones simultaneously and this behavior was defined as mutualistic cooperation. Mice were water-deprived 8 h before each training or test to enhance their motivation for the task. During the testing period, we observed a consistent reduction in co-drinking latency, and a consistent increase in co-drinking number and cumulated co-drinking time each day, together reflecting a gradually enhanced cooperative performance, as stochastic co-occurrence should predict flattened curves through time.

Institutional permissions

All experiments carried out should be approved by the Institutional Animal Care and Use Committee of Nanjing Medical University. Every effort was made to minimize the number of animals used and to ensure ethical treatment. Please note that all experiments conducted using animals require permission from the relevant institutions. An air-conditioned room (22 ± 2°C) is required for the cooperative test. Also, dim light (< 5 lux) is also needed. Experimenters should stay one meter away from the apparatus at least.

Mice models

In this experiment, we used different model mice to evaluate our new cooperative behavior test (CBT). Firstly, 2-month-old C57BL/6 male mice were used and exhibited the mutualistic cooperative behaviors in our CBT experiments. Then we investigated whether gender or age has an effect on cooperation by using 2-month-old C57BL/6 male or female mice and C57BL/6 male mice of 2 to 9-month-old with CBT (see expected outcomes). We also used three models with social interactive defection, chronic social defeat stress (CSDS), APP/PS1 mice and 5×FAD mice, to evaluate the reliability of our CBT. Besides, we explored whether the total test time could be shortened by increasing the number of training sessions per day.

Experimental design

In this protocol, we established a new method to evaluate the mutualistic cooperative behaviors of rodents. We use water as a reward to promote the cooperation of the animals. Four parts, which comprise habituation, training, test and data analyses, are mainly involved in our cooperative behavior test (CBT) procedure. Mice are firstly habituated for the environment and the apparatus. During habituation, they have access to water ad libitum. Then, during the training stage, all mice are deprived of water for 8 h before training. They are introduced to apparatus one by one to learn how to open the switch for water reward (Here, we call it “task” in training period). Finally, during the testing stage, two mice are moved to the apparatus at the same time. They should learn to open the two switches simultaneously to get water for themselves.

Construction of the apparatus

Timing: ∼ 1 week

This section outlines the procedure for the construction of the apparatus needed for the cooperative behavior test.

• 1.Prepare the plastic box needed for the apparatus.

Note: The CBT apparatus is a rectangular box with a size of 30 cm (length) × 20 cm (width) × 17 cm (height). The apparatus is made of polyethylene material that is 0.4 cm thick and the walls are assembled by a super-bonding compound. The box is divided into two chambers by the mesh. The size of the upper chamber is 30 cm (length) × 20 cm (width) × 15 cm (height), and the lower one is 30 cm (length) × 20 cm (width) × 2 cm (height). The gate of the apparatus is on the side wall. Rows of holes are punched in the walls for the airflow (Figure 1).

Note: The upper chamber is used for exploration of mice and the lower one is set as a drawer so that it is convenient to clean the feces of mice.

• 2.Install the optoelectronic switches.

Note: The optoelectronic switches are set above the top wall and the mirrors are set on the mesh of the corresponding location. These optoelectronic switches (retroreflective photoelectric sensors) are the use of the test object to block or reflect beam by the synchronization loop gating circuit to detect the presence or absence of the object.

Note: The two optoelectronic switches are set above the top wall 3 cm from the edge (the side wall opposite to the gate of the apparatus).

• 3.Install the mirrors for the optoelectronic switches.

Note: The mirror is made up of a plastic reflector with 5 cm (length) × 5 cm (width) × 0.3 cm (height). The surface of the mirror is smooth and comfortable to touch.

Note: The mirrors are set on the mesh of the corresponding location. The distance between mirrors is long enough so that one mouse cannot open the two switches simultaneously.

• 4.Install the solenoid valve connecting to the straws, water bottles, and the optoelectronic switches.

Note: The solenoid valve/normally closed water inlet flow switches are used as water control valves and connected to the switches by wires and water bottles by pipes. Straws are extended into the box for drinking from the water control valves on the top wall (Figures 1 and 2).

Note: The two straws are extended into the box just next to the mirrors from the water control valves on the top wall.

Note: To evaluate cooperative behaviors, we set up some tasks for the subjects. There are two switches and two straws. The two mice must occupy the mirror to activate the corresponding optoelectronic switch. They should open the switches at the same time and therefore get two water rewards (Here, we call it “task”).

• 5.Connect the solenoid valve and the optoelectronic switches to the main switches.

Note: All equipment can be controlled by a main switch.

• 6.Test whether the apparatus operates properly.

Note: Provide enough water in the bottles. Open the main switch, occupy the two mirrors at the same time and check if the water flows from the straws. Do not occupy the mirrors or occupy only one mirror and then check if the water flow stops.

Figure 1.

Photographs of the apparatus for cooperative behavior test

Scale bar = 3 cm.

Figure 2.

Illustration of the connections between solenoid valve, pipes, straws and optoelectronic switches from the top view

Data collection system

Timing: 1 day

This section outlines the ways to collect the mice behavioral data.

• 7.Prepare the computer system and a video camera located vertically in the apparatus.

CRITICAL: The experimenters should have been trained to identify the cooperative behavior so that the data collected are valid and reproducible. If a single experimenter cannot finish the whole experiment, it is important to ensure that the same criteria must be met to identify the cooperative behavior. Here, we define the cooperative behavior as the water valve is opened and at least one mouse drinks water.

Note: A video-tracking system is an alternative. Otherwise, the data can be collected manually by experimenters who should be blinded to the groups. It is recommended to use a video system (even though cannot track the mice) because a video recording allows for the analyses after the testing stage.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
75% Ethanol (cleaning the equipment)	Various	N/A
Experimental models: Organisms/strains
Mice: male or female C57BL/6J mice, aged 2/5/9 months	The Jackson Laboratory	RRID: IMSR_JAX:000664 https://www.jax.org/strain/000664
Mice: female APP/PS1 mice, aged 5 months	The Jackson Laboratory	RRID: MMRRC_034832-JAX https://www.jax.org/strain/005864
Mice: 5×FAD C57BL6/J mice, aged 3 months	The Jackson Laboratory	RRID: MMRRC_034840-JAX https://www.jax.org/strain/006554
Mice: male CD1 mice, aged 5 months	The Jackson Laboratory	RRID: IMSR_JAX:009122 https://www.jax.org/strain/009122
Software and algorithms
GraphPad Prism	Dotmatics	www.graphpad.com
Microsoft Excel	Microsoft	www.microsoft.com
Adobe Illustrator	Adobe	www.adobe.com
TopScan	CleverSys	www.cleversysinc.com
Other
Water bottles	Various	N/A
Pipes and standard stainless-steel straw	Various	N/A
Video camera	Various	N/A
Plastic box needed for the apparatus	In-house; see “construction of the apparatus”	N/A

Step-by-step method details

Habituation

Timing: 3 days

The habituation stage makes the mice feel easy and reduce their stress during training and testing stages.

• 1.During habituation (Steps 1–2), move mice to the test room for environmental adaption for about 2 h on the first day before apparatus habituation.
• 2.In the next two days, transfer them to the chambers for apparatus habituation 5 min a day following 2 h environmental adaption.

CRITICAL: The apparatus should be cleaned up with 75% (vol/vol) ethanol between two consecutive adaptations and remember to make the main switch of the apparatus off.

Note: Handle mice by the experimenters for 10 min every day during habituation (Steps 1–2).

Note: Habituation is essential to reduce stress being introduced to the CBT apparatus. Also, this habituation can minimize the confounding effects of novelty on optoelectronic switches and straws.

Note: At these stages (Steps 1–2), the animals have access to water ad libitum.

Training

Timing: 7 days

The training stage trains the mice to finish the task and get rewards subsequently.

• 3.Begin water restriction and maintain it throughout the whole test.

Note: Our CBT protocol is based on the water reward and therefore requires animals to undergo a water deprivation schedule to ensure that they are willing to do the task for obtaining water reward.

Note: Be sure to adhere to guidelines of governmental and institutional animal welfare.

Note: During the training stage (Steps 3–10), all mice are deprived of water for 8 h before training every day.

• 4.Weigh each animal every day.

Note: Avoid weight reduction of more than 5% per day or 15% during the whole behavioral testing period and unqualified mice are excluded from the following experiments.

• 5.Set up the apparatus with the main switch on and turn on one of the optoelectronic switches.

CRITICAL: One of the two optoelectronic switches should be turned on during training days. You can open one switch by attaching the tape to it (Figure 3). When the mice occupy the mirror and activate the other optoelectronic switch, the water valve is opened and the water is given from the straw at a speed of 2–3 drops per second. When one mirror is not occupied, the water flow will stop.

Note: Make one of the two optoelectronic switches on at liberty during each trial manually to prevent the animals from habituating to only one switch.

Note: Provide enough water in the bottles so that the mice can get rewards if they finish tasks.

Note: We recommend to train the animals twice a day at least and interval between two trials last about 8 h.

• 6.Move one mouse into the apparatus and allow 5 min to explore the apparatus and learn how to finish the task and gain rewards.

CRITICAL: If the mice cannot fulfill a task during the training time, it is necessary to place them on the mirror to make the switch on and help them to get water at least 3 times during the 5-min exploration.

CRITICAL: The number of training is important for the mice to learn to use the equipment. The mice will perform better and learn faster as the number of training increased. We recommend training the animals twice a day at least.

Note: We recommend to mark the tail of each animal with a permanent marker. So it is convenient for you to distinguish the training status of an individual animal.

Note: It is recommended to use video-tracking systems to detect the animals during this period if possible.

• 7.After training termination, return animals to their respective home cages.
• 8.Clean up the switches, straws, and chambers with 75% (vol/vol) ethanol and ensure them dry before the next training.
• 9.Repeat step 7–9 until all mice have been trained.
• 10.Repeat step 4–10 twice a day for 7 days.

CRITICAL: Make sure that all the mice are trained well and can learn to use the equipment and get water rewards until the end of training days. Generally, five days are needed to reach this point. You can extend the training periods appropriately.

Note: We consider the mice to be well trained if their drinking latency reduces along with time and they drink water for at least 10 seconds in the last 2 training days.

Figure 3.

Open one optoelectronic switch by attaching the tapes to it

Left, side view; right, bottom view.

Cooperation behavior test

Timing: 5 days

This stage detects the cooperation behavior of the pair of mice.

• 11.Repeat step 4 for all test days.
• 12.Turn on the main switch and ensure that optoelectronic switches can work well.

Note: Check if there is enough water in bottles and the water rewards can be given properly.

Note: During the testing stage (Steps 11–18), all mice are deprived of water for 8 h before training every day.

Note: It is recommended to use video-tracking systems to detect the animals during this period if possible.

• 13.Clean up the switches, straws, and chamber with 75% (vol/vol) ethanol and ensure them dry before the test.
• 14.Move every pair of mice into the “mutualistic cooperation” apparatus and allow the 10-min tests for all pairs every day.

CRITICAL: Make sure that all the mice have learned to use these models for rewards. If not, repeat the training sessions again.

CRITICAL: Experimenters should stay in another room or at least 1 meter from the apparatus since the test is going on. So, we recommend a video-tracking system to record the performance of the mice. It is useful for you to analyze the results at the same time of or after the test.

Note: Three approaches can be used to choose the pair of mice. (i) Choose the two animals at random. (ii) Choose the same pair of mice for every test. (iii) Use a well-trained mouse and match all other mice to it successively. All these approaches can be adapted for different purposes. Both the first way and the second way could be used to evaluate the cooperative behavior in the group level. Compared to the second way, every two mice could be matched together in the first way, which could avoid experimental errors caused by fixed pairs. However, fixed pairs might have a good performance in the testing period. Besides, the third way could be used to evaluate cooperative behavior at the individual level and do not use the same demonstrator in a consecutive test if you adopt this approach. Here we recommend the second one for the normal test.

Note: When the two mice occupy both mirrors concurrently, the optoelectronic switches are activated, the water valve is opened and the water is given from the straw at a speed of 2–3 drops per second. When one mirror is not occupied, the water flow will stop.

• 15.Return animals to their home cages after the tests are finished.

Note: All animals are given water at libitum in their home cages at the end of the test. This process lasts during the testing stage.

• 16.Clean up the switches, straws, and chambers with 75% (vol/vol) ethanol and ensure them dry before the next test.
• 17.Make the main switch off and clean the bottles and straws until all the mice have been tested.
• 18.Repeat the test (step 10–16) once a day for 5 days.

Data analyses

Timing: 1 day

This stage provides the detail parameters for the analyses of the cooperation behavior.

• 19.Analyze the performance of mice during the test as the following parameters.

CRITICAL: The optoelectronic switches will be activated and the water valves are open when the two mice occupy both mirrors concurrently. The water control valves stay open when two mirrors are both occupied. We will consider this process as mutualistic cooperative behavior as long as the water valve is opened and at least one mouse drinks water.

Note: Three parameters are recommended for analyses, including latency to get water rewards cooperatively, the number of the tasks accomplished and the time of their cooperation.

Expected outcomes

Three parameters are used here to evaluate the cooperative ability in our CBT: the latency for the first successful co-drinking, the number of co-drinking in each trial, and the total time of co-drinking in each trial. The latency is more reflective of their cooperation ability, because their desire for water should be highest at the beginning; the co-drinking number and the total time are more reflective of the amount of water they required during each trial.

To demonstrate the efficiency of our CBT in measuring mice’s cooperative ability, we introduced 2-month-old male mice into our apparatus (Figure 4A). After the 7 days of individual training, all mice learned to accomplish the task with water rewards, as reflected by the declining latency curve (Figures 4B–4D). Next, we moved those mice to the test setup. In the “mutualistic cooperation (i)” model, we observed constantly declining latency during the testing phase. The co-drinking number ranged from 2-8 times and the total time ranged from 20-80 s during the testing phase. However, the co-drinking number and time did not show a significant increase during the 5-day test (Figures 4E–4G). We also tried the 1% sucrose solution instead of the regular water as a reward for a short training and testing phase. However, the latencies did not decrease significantly (Figures 4H–4M). Therefore, the following experiments were still carried out with regular drinking water.

Figure 4.

Establishment of cooperative behaviors test

(A) Diagram of our cooperative behaviors test apparatus. Scale bar = 2.5 cm.

(B–D) The latency to cooperation and the number and time of cooperation of mice during the training stage (latency, F_{(6, 132)} = 27.589, p < 0.001; number, F_{(6, 132)} = 17.639, p < 0.001; time, F_{(6, 132)} = 26.246, p < 0.001; n = 23).

(E–G) The latency to cooperation and the number and time of cooperation of the pair of mice during the testing stage (latency, F_{(4, 32)} = 3.304, p = 0.022; number of cooperation, F_{(4, 32)} = 0.473, p = 0.755; time of cooperation, F_{(4, 32)} = 0.586, p > 0.675; n = 9).

(H–J) The latency to cooperation and the number and time of cooperation of mice rewarded by regular or sugar implanted water during the training stage (latency, F_{(1, 53)} = 0.149, p = 0.701; number, F_{(1, 53)} = 0.741, p = 0.393; time, F_{(1, 53)} = 0.260, p = 0.613; n = 25–30).

(K–M) The latency to cooperation and the number and time of cooperation of the pair of mice rewarded by regular or sugar implanted water during the testing stage (latency, F_{(1, 25)} = 0.656, p = 0.426; number of cooperation, F_{(1, 25)} = 0.082, p = 0.777; time of cooperation, F_{(1, 25)} = 0.330, p = 0.571; n = 12–14). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. The results are shown as the mean ± SEM and tested by repeated-measures ANOVA with Bonferroni’s post hoc test.

Next, the weight of mice and general behavior tests were detected to evaluate the influence of our CBT on the mice.¹ The weight of mice with or without the CBT showed no difference (Figure 5A). The anxiety or depression related behaviors were also unchanged detected by open field test or elevated plus maze between two groups (Figures 3B–3E). There was also no significant difference in memory associated tests such as Y maze and novel object recognition test (Figures 5F–5H) (For the detailed protocols of these tests, please refer to Zhang et al. and Wang et al.^1,2). These results demonstrated that our CBT did not influence the performance of mice in these behavioral tests.

Figure 5.

Our cooperative behavior test did not influence the weight and other behavior performance of the mice

(A) The weight of mice with or without our cooperative behavior test.

(B and C) The percentage of time spent in the center area and the number of entering into the center area of the open field in mice with or without the cooperative behavior test.

(D and E) The percentage of time spent in the open arm and the number of entering into the open arm of elevated plus maze in mice with or without the cooperative behavior test.

(F and G) The percentage of time spent in the novel arm and the number of entering into the novel arm of the Y-maze in mice with or without the cooperative behavior test.

(H) The ratio of time spent on sniffing novel object in total sniff during novel object recognition test in mice with or without the cooperative behavior test. n = 10–19. The results are shown as the mean ± SEM and tested by Student’s t-test.

To reveal whether gender or age has an influence on cooperative ability or not, 2-month-old male or female mice or 2 to 9-month-old mice were tested in our CBT. The results showed that female mice have good cooperative performance in our CBT as well (Figures 6A–6F). In terms of age, 9-month-old mice had significant longer latency and less number and time of cooperation compared to 2-month-old male mice during the testing period (Figures 6G–6L).

Figure 6.

Detection of cooperative behaviors in mice with different genders and ages by our models

(A–C) The latency to cooperation and the number and time of cooperation of male and female mice during the training stage (latency, F_{(1, 44)} = 0.135, p = 0.715; number, F_{(1, 44)} = 0.190, p = 0.665; time, F_{(1, 44)} = 0.020, p = 0.888; n = 22–23).

(D–F) The latency to cooperation and the number and time of cooperation of male and female mice during the testing stage (latency, F_{(1, 17)} = 2.263, p = 0.152; number, F_{(1, 17)} = 4.325, p = 0.054; time, F_{(1, 17)} = 4.378, p = 0.053; n = 8–10).

(G–I) The latency to cooperation and the number and time of cooperation of 2, 5, and 9-month-old mice in “mutualistic cooperation” models during the training stage (latency, F_{(2, 57)} = 0.344, p = 0.710; number, F_{(2, 57)} = 0.528, p = 0.593; time, F_{(2, 57)} = 0.058, p = 0.943; n = 17–23).

(J–L) The latency to cooperation and the number and time of cooperation of 2, 5, and 9-month-old mice in “mutualistic cooperation” models during the testing stage (latency, F_{(2, 44)} = 3.694, p = 0.044; number, F_{(2, 44)} = 6.039, p = 0.009; time, F_{(2, 44)} = 10.824, p = 0.01 n = 7–9). ∗p < 0.05, ∗∗p < 0.01. The results are shown as the mean ± SEM and tested by repeated-measures ANOVA with Bonferroni’s post hoc test.

Then, three other mice models, CSDS mice, APP/PS1 mice and 5×FAD mice were used to qualify the feasibility of our CBT. Previous studies have reported that the social interactive ability was impaired in CSDS mice, 5-month-old APP/PS1 mice or 3-month-old 5×FAD mice and social interactive ability was fundamental for cooperation behaviors.^15,16 In our studies, we found that the drinking latencies during the training stage were declined in all of three mice models, including CSDS, APP/PS1, and 5×FAD mice (Figures 7A–7C, 7G–7I, 8A–8C and 8G–8I), suggesting their capability of finishing the tasks during the training stage. Most importantly, similar to the C57BL/6J WT mice, all three mice models were able to get water rewards during the 10-min testing stage, revealing by co-drinking latency and time (Figures 7D–7F, 7J–7L, 8D–8F, and 8J–8L), which indicated they could show cooperative behavior in this apparatus. However, the co-drinking latency was increased and the co-drinking number/time decreased in these three mice models compared to the respective control groups (Figures 7D–7F, 7J–7L, 8D–8F, and 8J–8L). These results revealed that our CBT is applied for testing cooperative behaviors in a wide range of mice models. Taken together, the set of procedures and results above provide a reliable and useful method to evaluate cooperative behaviors.

Figure 7.

Detection of cooperative behaviors in social ability impaired APP/PS1 and CSDS mice by our models

(A–C) The latency to cooperation and the number and time of cooperation of WT and APP/PS1 mice during the training stage (latency, F_{(1, 37)} = 0.144, p = 0.707; number, F_{(1, 37)} = 0.153, p = 0.698; time, F_{(1, 37)} = 0.634, p = 0.431; n = 18–20).

(D–F) The latency to cooperation and the number and time of cooperation of WT and APP/PS1 mice during the testing stage (latency, F_{(1, 15)} = 28.028, p < 0.001; number, F_{(1, 15)} = 64.989, p < 0.001; time, F_{(1, 15)} = 69.421, p < 0.001; n = 7–9).

(G–I) The latency to cooperation and the number and time of cooperation of WT and CSDS mice during the training stage (latency, F_{(1, 39)} = 0.855, p = 0.361; number, F_{(1, 39)} = 1.861, p = 0.181; time, F_{(1, 39)} = 0.599, p = 0.444; n = 20).

(J–L) The latency to cooperation and the number and time of cooperation of WT and CSDS mice in “mutualistic cooperation” models during the testing stage (latency, F_{(1, 39)} = 20.532, p = 0.001; number, F_{(1, 39)} = 29.301, p < 0.001; time, F_{(1, 39)} = 14.534, p = 0.002; n = 7–10). ∗∗p < 0.01, ∗∗∗p < 0.001. The results are shown as the mean ± SEM and tested by repeated-measures ANOVA with Bonferroni’s post hoc test.

Figure 8.

The cooperative performance of 5×FAD mice at 3 months was similar between 3 days of training (6 times a day) and 7 days of training (2 times a day)

(A–F) Statistical chart shows the time of drinking latency, drinking number and drinking time during the training period in WT mice and 5×FAD mice for 7 consecutive days (2 times a day) (A–C) (latency, F_{(1, 79)} = 0.154, p = 0.696; number, F_{(1, 79)} = 0.131, p = 0.718; time, F_{(1, 79)} = 0.584, p = 0.447; n = 40) or 3 consecutive days (6 times a day) (D–F) (latency, F_{(1, 34)} = 0.005, p = 0.947; number, F_{(1, 34)} = 0.915, p = 0.346; time, F_{(1, 34)} = 0.136, p = 0.715; n = 18).

(G–L) During the testing phase, 5×FAD mice trained for 7 days (G–I) or 3 days (J–L) showed impairments in cooperation performance, compared with WT mice (7 days: latency, F_{(1, 39)} = 21.954, p < 0.001; number, F_{(1, 39)} = 27.065, p < 0.001; time, F_{(1, 39)} = 33.804, p < 0.001; n = 20; 3 days: latency, F_{(1, 17)} = 5.490, p = 0.032; number, F_{(1, 17)} = 4.706, p = 0.045; time, F_{(1, 17)} = 17.099, p = 0.001; n = 9). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. The results are shown as the mean ± SEM and tested by repeated-measures ANOVA with Bonferroni’s post hoc test.

To exclude the possibility that the pair of mice completed the tasks due to their desire for water separately, but not cooperatively, the original “mutualistic cooperation (i)” and the other “mutualistic cooperation (ii)” models were used. In the “mutualistic cooperation (i)” model, the mirrors were located near the straws so that the mice could occupy the mirror and drink the water at the same time. However, in the “mutualistic cooperation (ii)” model, the mirrors were far from the straws and the mice were unable to occupy the mirrors and get the water concurrently. They were able to only drink the water drops left in the straws after activating the optoelectronic switches together. The mice could finish the cooperative task as well in the “mutualistic cooperation (ii)” model (Figures 9A–9D), which proved that the design of our CBT is useful to test cooperative behaviors, independent of the location of the switches. As expected, the tasks in “mutualistic cooperation (ii)” were more difficult than that in “mutualistic cooperation (i)”, so that the co-drinking latency was extended in “mutualistic cooperation (ii)” model (Figure 9C). However, the co-drinking time were unchanged in “mutualistic cooperation (ii)” model (Figure 9D). Then, we compared the difference between the previous cooperative test model and ours. In the previous model,^16,17 the participants were separated by a grille partition. Although they can see, hear, and smell each other, most physical contact is deprived. We used a similar grille partition in our CBT to separate the mice and detect cooperative behavior subsequently. The mice separated by the partition spent more time learning to finish the task during the training period. In the test period, the number of cooperation also decreased significantly (Figures 9E–9J). Cooperative behavior needs more than sight and smell, so our CBT provided a better place for the mice to touch to finish the cooperative tasks. However, the mechanism of this cooperative behavior requires further investigation.

Figure 9.

The modified vision of our cooperative behavior test

(A and B) The latency to cooperation and the number and time of cooperation were comparable between “mutualistic cooperation” model (i) and the “mutualistic cooperation” model (ii) during the training stage (latency, F_{(1, 39)} = 0.012, p = 0.912; number, F_{(1, 39)} = 4.013, p = 0.052; n = 19–22).

(C and D) However, the latency to cooperation was increased in modified “mutualistic cooperation (ii)” models during the testing stage (latency, F_{(1, 40)} = 12.712, p = 0.012; number, F_{(1, 40)} = 5.4, p = 0.103; n = 6–8).

(E–G) The latency to cooperation and the number and time of cooperation of the mice in models where they could touch each other or not during the training stage (latency, F_{(1, 50)} = 1.216, p = 0.276; number, F_{(1, 50)} = 0.034, p = 0.855; time, F_{(1, 50)} = 0.078, p = 0.781; n = 20–30).

(H–J) The latency to cooperation and the number and time of cooperation of the mice in models where they could touch each other or not during the testing stage (latency, F_{(1, 24)} = 4.198, p = 0.052; number, F_{(1, 24)} = 10.370, p = 0.004; time, F_{(1, 24)} = 3.351, p = 0.08; n = 12–13). ∗∗p < 0.01. The results are shown as the mean ± SEM and tested by repeated-measures ANOVA with Bonferroni’s post hoc test.

Quantification and statistical analysis

In our CBT, we mainly use three parameters to evaluate cooperative ability, including latency, the number of the tasks accomplished and the time of their cooperation. All the parameters are based on the finish of tasks. The tasks including opening the switch in training, opening the switch to get water for the other or opening switches simultaneously. Latency is defined as the time of finishing the task, firstly during training or test period. The number of the tasks accomplished is defined as the sum of numbers that they finish tasks during the training or testing period. The time of their cooperation is defined as the sum of time that they are drinking water. For comparison between two groups at a single time point, a student’s t-test may be used for datasets fitting a normal distribution and meeting requirements for parametric testing. To compare within and between treatment groups at various time points over a time course, the repeated measures ANOVA should be used. Significant interactions should be followed up by appropriate post-hoc comparisons.

Limitations

The primary limitation of our approach is that it is time-consuming. The whole procedure will take about 15 days from the training to the test period. Our longer time frame results from: 1) it took about one week for the mice to learn to use the apparatus at the frequency of 2 sessions/day. 2) test period often costs 5 days due to evaluating the improvement of cooperative ability. To address this issue, we examined whether increasing the number of training sessions per day could shorten the total number of training days. We found that the mice were able to use this apparatus when the training time was reduced from 2 sessions/day for 7 days to 6 sessions/day for 3 days (Figures 8A–8F). These mice undergoing 3-day training also showed a similar cooperative performance (Figures 8G–8L). The next limitation of our approach is water restriction. Because we used water as the reward, water restriction was needed during the training and testing period. Such water restriction could lead to the stress or weakness of the mice. However, the weight of the mice was detected during our approach and no significant changes in weight were observed between mice with or without our CBT. So, the experimenters’ experience was important to make the detriment of water restrictions under control.

Troubleshooting

Problem 1

Animals undergo significant weight loss (steps 3–4).

Potential solutions

The weight loss could be attributed to water restriction during experiments. Giving water appropriately may avoid the problem.

Problem 2

Few mice do not reach the criterion. Or most of the mice do not reach the criterion (steps 10/14).

Potential solutions

The mice could not reach the criterion if they were nervous during the experiment. If a few mice do not reach the criterion, experimenters should exclude the mouse from the analysis directly. If most of the mice do not reach the criterion, there may be several reasons. One is ineffective water restriction and you can restrict water for a long time. The other is stress. You can check the experimental conditions, including temperature, humidity, luminosity, noise, or odors related to the experimenter. Make an effort to relax the mice and avoid noise, light, and other environmental changes during the test.

Problem 3

Leaking of tubes (steps 5/12).

Potential solutions

Replace with new tubes and keep air tight in all the pipes.

Problem 4

The mice sniff the apparatus for a long time and cannot learn to open the switches (steps 6/14).

Potential solutions

Remove the feces and clean up the apparatus with 75% ethanol thoroughly to avoid interference from previous mice.

Problem 5

The mice sniff each other and cannot learn to open the switches (steps 6/14).

Potential solutions

The mice might sniff each other for a long time because of different genders. If you find a pair of mice sniffing each other all the time, stop the experiments, keep the pair of mice of the same gender and try again.

Problem 6

There is no water from the tube when the mice open the switches cooperatively (steps 11–18).

Potential solutions

You should check if there is enough water in the water bottles. In addition, you should check if the pipes and the tubes work well. If not, replace them with a new one.

Resource availability

Lead contact

Further information and requests for resources should be directed to and will be fulfilled by the lead contact, Ming Xiao (mingx@njmu.edu.cn).

Technical contact

Questions about the technical specifics of performing the protocol should be directed to and will be answered by the technical contacts, Weixi Feng (weixif@hotmail.com) and Yanli Zhang (yanlizhang0229@njmu.edu.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

This study did not generate code. Behavioral data are available from the technical contact upon request.