Set up a rodent behavior testing room

In one study, three labs ran the same mouse strains through six behavior tests at the same time, with the apparatus, the protocols and many room conditions matched.⁷ Behavior still differed from lab to lab. For some tests, how far apart the strains scored depended on which lab ran the test.⁷

The person running the test counts too. In a study that Hånell and Marklund describe, the experimenter’s identity affected hot plate test results more than genotype did.⁶

No lab can copy another lab’s room exactly. You can hold your own room, routine and people steady for every group, day and cohort, and record what you kept the same.

Picking a mouse up by its tail is aversive and raises anxiety. With a tunnel or cupped hands, mice come to the handler on their own and accept being held.^4,5

Gouveia and Hurst tested 48 female BALB/c mice handled one of three ways. Mice moved in a tunnel spent about 40% of a 60-second test interacting with the handler, against about 6% for mice picked up by the tail.⁵ Cupped mice fell in between.⁵

In a scent test that followed, where mice had to tell two urine scents apart, tunnel-handled mice explored readily and responded clearly to the scents.⁵ Tail-handled mice explored little and performed poorly, and time to get used to the arena first helped them only slightly.⁵ Cupped mice responded more variably than tunnel-handled mice.⁵

Move a mouse in a tunnel

1. Leave a tunnel in the cage

   Keep a handling tunnel in each home cage, so one is there whenever you move a mouse.⁵

2. Guide the mouse in

   Encourage the mouse into the tunnel, then cover both ends with your hands while you carry it.⁵

3. Tip it out backwards

   At the apparatus, tip the tunnel gently so the mouse slides out backwards, the same way for every mouse.⁵

To cup a mouse, first pick it up between your two hands closed loosely around it, then hold it on your open hands once it no longer jumps away.⁵ Mice new to cupping jump off the hand at first and come to the handler gradually over several sessions. A tunnel is safer for mice that might jump, such as young mice or those little used to handling.⁵

Long handling sessions are unnecessary once the method is not aversive. In Gouveia and Hurst’s first experiment, mice were handled only at cage cleaning, from three weeks of age, and the handling methods differed clearly after five cleanings.⁵ Mice in their second experiment arrived as adults and got the same brief pick-up, about 2 seconds, in ten daily sessions instead, to override their earlier handling.⁵ Tunnels have lowered anxiety-related behavior across strains and both sexes, in the light and the dark phase, even when the animals or handlers were new to handling.⁵

These handling studies were in mice. For rats and mice alike, handling animals before testing may reduce the experimenter’s effect on the results and the variability between animals.⁶

Rodents react to who is in the room. Sorge and colleagues found that exposure to male experimenters caused a stress response in mice and rats that dulled their response to pain, while exposure to female experimenters did not.⁸ T-shirts worn by men were enough to cause it, so the cue is smell. Sorge and colleagues conclude that experimenter sex can affect baseline responses in behavior tests.⁸

Odor traces from pets that hunt rodents, such as cats, stress rodents, so keep them off the clothes you test in.⁶ A person’s own stress, mood and odor can also vary from day to day.⁶

• Have one person run all the testing in a study, which is how labs usually handle experimenter effects.⁶
• Keep the observer’s presence and position the same for every animal. An observer present for some animals and absent for others may change their behavior.⁹
• If the observer stays in the room, they should move as little as possible and make no noise.⁹
• Hang a curtain or set up a room divider so the animal cannot see the experimenter.^9,10
• Where you can, leave the room while data are collected. One open field protocol asks everyone to leave, to cut noise and distraction.¹

Hånell and Marklund list the potential advantages of automated scoring as objective scores, less contact between experimenter and animal, and longer tests.⁶

Rodents avoid large, brightly lit open spaces.² Small changes in lighting, temperature, humidity, odor, noise and human activity can all change activity levels.¹

• Light each apparatus evenly, with no direct light, dark corners or shadows over it.¹
• Space the apparatus evenly around the room.¹
• Control the room’s temperature and humidity.¹
• Measure light at the floor of the apparatus with a light meter, and record it in lux with the temperature and humidity.

Rats and mice rest during the day, and testing at different points in the light-dark cycle can change the outcome.⁶ Hånell and Marklund consider a reversed light-dark cycle, which lets you test while the animals are active, likely the best choice for most behavior studies.⁶ Whatever cycle you use, test at the same time each day.¹

Noise is one of the room conditions that can change test outcomes.¹ Stressed animals tend to do worse in the water maze, so Bromley-Brits and colleagues keep temperature, light and noise constant across the whole task.¹⁰

Rodents hear ultrasound and use it to communicate, so a room that seems silent to you can still carry sound they hear.⁶ Several common mouse strains, though, have restricted hearing.⁶ Before a test that uses a sound or another cue, check that your animals can sense it. Hearing can be checked with prepulse inhibition of the acoustic startle response.⁶

• Keep talk, phones, doors and carts out of the room during sessions.
• Run one practice animal first. It checks the recording system, and it means the first study animal, like the rest, follows an earlier trial’s noises and odor cues.³

Bring the animals to the testing room in their home cages and let them settle before testing.¹ Seibenhener and Wooten’s open field protocol and Komada and colleagues’ plus maze protocol allow 30 minutes or more. Tatem and colleagues’ open field activity protocol, with beam chambers, allows about 10 minutes.^2,3,1

Pick one settling time, give it to every cage, and record it as the acclimation period. The ARRIVE guidelines ask you to report any acclimation periods.¹¹

Rodents rely on smell far more than sight, so traces one animal leaves may change what the next one does.⁶ Besides urine and feces, rodents leave fluid from their foot pads that may affect later tests.⁶ Mice see rats as predators, so clean shared equipment thoroughly when both species use it.⁶

1. Remove waste

   Pick out every fecal pellet and wipe up every spot of urine.²

2. Wipe every surface

   Wipe the floor and walls before the first animal and after each one, to remove the scent the last animal left.²

3. Let it dry

   Wait until the cleaner has fully evaporated before the next animal goes in. With 95% ethanol this took 5 to 10 minutes in one protocol.²

4. Choose a cleaner with little smell

   One plus maze lab uses super hypochlorous water, which removes odor well and has a weaker smell of its own than other cleaners.³

5. Clean after each session

   Clean every chamber thoroughly at the end of each session.¹

Equipment is easier to clean when it has no narrow passages or corners and is made of a material that cleans readily, ideally one that can be sterilized.⁶

Housing can change behavior results in several ways.⁶ Hånell and Marklund consider group housing likely the better choice for most studies, because results from animals housed alone may be less valid.⁶ Group housing has one drawback. Cage mates form a hierarchy, and lower-ranking animals have higher stress hormone levels and more anxiety-like behavior. That adds variability to the data.⁶

• Change cages on days without testing, because moving to a clean cage briefly raises stress hormones.⁶
• Use the tunnel or cupped hands at cage changes too, which gives mice the brief handling they need.⁵
• Record housing and husbandry, including enrichment, as the ARRIVE guidelines ask.¹¹

Where an animal falls in the testing order can change its score. In Komada and colleagues’ plus maze protocol, the third and fourth mouse tested in a session spent more time on the open arms than the first, so they balance trial order across genotypes.³ Hånell and Marklund add that the order in which animals from one cage are tested can add variation, so randomize the testing order.⁶

Fully random assignment can leave a cage with only control animals, which risks systematic errors. Use each cage or litter as a block instead, and assign the animals within it to groups at random.⁶ The ARRIVE guidelines also ask you to name the experimental unit, such as one animal, a litter or a cage.¹¹

Score behavior without knowing each animal’s group, and run the test blind as well where you can.^6,1 To hide the groups, have a colleague label the animals after the intervention, or replace the cage cards.⁶

Many studies run several tests on the same animals, and taking part in one test can change the results of the next.⁶ The order of the tests therefore matters. Running them on separate days may reduce the carryover, though Hånell and Marklund note this has not been verified.⁶

• Give every animal the same tests in the same order.
• Run each test on its own day where the design allows.⁶
• Report the order and the gaps, since the ARRIVE guidelines ask how you limited confounders such as the order of measurements.¹¹

When the ARRIVE guidelines were updated in 2020, their authors noted that most publications sampled in recent studies still left out details the guidelines ask for. Randomization was reported in only 30% to 40% of them, blinding in about 20% and sample size justification in less than 10%.¹¹

• The animals, with species, strain and substrain, sex, age and, if relevant, weight.¹¹
• Housing and husbandry, including enrichment.¹¹
• When, how often and where each procedure was done, with any acclimation period.¹¹
• Whether and how animals were randomized to groups, and the experimental unit.¹¹
• Who knew the group assignments during allocation, testing, scoring and analysis.¹¹
• How you limited confounders such as test order and cage position.¹¹

For a behavior study, add the room details this guide covers, such as the light level, the time of day, who ran the tests, the handling method and the cleaner.

• Results differ between cohorts. Keep the time of day, experimenter, handling method and settling time fixed, and test every group in every cohort.
• Mice avoid the handler and explore little. Switch from the tail to a tunnel, and use it at every cage change.⁵
• Scores drift through the day. Randomize the testing order and spread each group across the day.^3,6
• The first animal of the day behaves differently. Run a practice animal before the first study animal.³
• Animals react to the observer. Hide the observer behind a curtain or leave the room, and keep the setup the same for every animal.^9,1
• Mice behave oddly after rats used the apparatus. Clean it thoroughly before the mice go in.⁶
• Baselines shift when a new person joins the study. Keep one experimenter for the whole study, and record who ran each session.^6,8