An appetitive spatial working memory task for mice, using a semi-automated 8-arm radial maze

doi:10.21203/rs.3.pex-2431/v1

Method Article

An appetitive spatial working memory task for mice, using a semi-automated 8-arm radial maze

https://doi.org/10.21203/rs.3.pex-2431/v1

This work is licensed under a CC BY 4.0 License

This protocol has been posted on Protocol Exchange, an open repository of community-contributed protocols sponsored by Nature Portfolio. These protocols are posted directly on the Protocol Exchange by authors and are made freely available to the scientific community for use and comment.

Version 1

posted

You are reading this latest protocol version

This methods paper details a protocol to test spatial working memory in mice using a semi-automated radial 8-arm maze (RAM). The RAM is a partially enclosed apparatus with 8 horizontal equally spaced arms radiating from a central hub, from which access to each arm can be controlled individually by servo-controlled motorized doors. Animals start in the central hub and are allowed to explore the maze for a food reward at the end of each arm or selected arms. The RAM task was originally designed for rats but we have adapted the protocol for mice – for example, by including more habituation steps. In our protocol, all arms are initially baited with sweetened condensed milk and mice are admitted sequentially into four pseudo-randomly selected arms to collect the rewards (“forced run”) before opening all doors together to allow the mice to run freely and find the remaining four rewards (“free run”). A 5-sec time delay is introduced between the forced and free runs to challenge working memory; an error is recorded if the mouse enters any previously-visited arm during the free run. The task is complete when all rewards are recovered. After 6 days of habituation and 9 days of maze training, male C57BL/6 mice regularly achieve ≥80% success rate, defined as 4/(4+E) where E is the number of errors. This semi-automated task could in principle be combined with in vivo monitoring methods such as electrophysiology or calcium imaging.

Radial arm maze

Working memory

Mice

Radial arm maze
The land radial arm maze (RAM) is a popular and commonly used apparatus to assess spatial- and nonspatial- learning and memory in rodents^1-3. The RAM takes advantage of the natural tendency of rodents to explore the maze, find novelties and learn spatial information on a trial-and-error basis in addition to foraging. This tendency can be augmented by restricting their diet and rewarding them with a preferred food in the maze. The original RAM was designed by Olton and Samuelson¹ to assess spatial learning and memory in rats using food baits for positive reinforcement, unlike water mazes such as the Morris water maze and 8-arm radial water maze that use negative reinforcement, i.e. water immersion^1,4-6. Olton and Samuelson¹found that rats have excellent memories for remembering places they passed. Subsequently, they showed that the hippocampus is involved in spatial information processing; they made lesions at the hippocampal formation (the fimbria-fornix) of rats and tested them for their ability to recall pre-learned spatial discrimination in 8- or 17-arm versions of the RAM^7-10. Olton et al. found that hippocampal function is essential for working memory by making lesions of the entorhinal area, body of the fimbria-fornix anterior to hippocampus, septum, or postcommissural fornix to disrupt each of the major extrinsic fibre connections of the hippocampal formation^7-10 and also found that making fornix-fimbria lesion did not affect reference memory¹⁰, introducing into the field of rodent behaviour the concept of a dual memory system, which had been suggested previously by Honig¹¹ based on experiments with pigeons. Nowadays, several variants of the RAM apparatus are used e.g. with 6, 8, 12 or 16 arms and a partially enclosed rather than an open/elevated apparatus^12-14. Most present-day protocols are based on that of Olton et al.^1,10 but use simpler (6- or 8 arm) mazes^13,15-18. In modern protocols, animals are habituated to the environment by allowing them to explore the maze freely for 10-15 min per day for 3 consecutive days, followed by maze training for one to three weeks with a preferred food at the end of all arms or selected arms, testing working memory, or working memory and reference memory in the same experiment, respectively^17-19. Often these protocols are not described fully, so we thought it useful to describe our own protocol and experience in detail here.

Our semi-automated 8-arm radial maze is an enclosed maze with opaque white acrylic (plexiglass) walls but no lids (see Fig. 2a for dimensions). While in the maze, mice cannot see the experimenter, which can be a distraction with transparent-wall or open-elevated mazes, but they can see distal extramaze visual cues above the maze (see Fig. 2). In our protocol, mice have three types of habituation for 2 days each, 6 days in total. During habituation, mice become familiar with the sound and movement of the doors, the presence of food rewards at the end of each arm, and simultaneously start to form a spatial map of the maze layout (see below, section on “ Habituation and working memory task”, for details).

Behavioural differences between rats and mice
Most cognitive tasks were originally designed for rats and spatial learning/memory has been intensively studied using rats. Because of the advantage of today’s transgenic and genome modification technologies for mice, researchers often apply protocols devised for rats to mice. However, there are consistent behavioural differences between rats and mice^20-24. For example, they have preference differences (social preference by rats and food preference by mice)²⁰ and use different strategies to negotiate different maze tasks (water or land mazes)²¹. Also, it seems mice tend to take a longer time to achieve good spatial learning scores in 8-arm RAM^17,21,25, which is one reason that mice are generally regarded as inferior to rats in spatial learning tasks.

Rats and mice also differ in general behaviour/character including their responses to stress or fear management^20-24,26. For example, previous studies found poor fear extinction learning in adolescent mice^27-30 but not rats^31-34. In addition, mice are prone to stress and do not like being handled — unlike rats, which when familiar with handling apparently enjoy being tickled by the handler and try to interact with handlers for more tickling³⁵. Therefore, fear management is an important factor to consider when studying mouse behaviour. Furthermore, Borrow et al.³⁶ have reported that chronic variable stress decreases oxytocin mRNA, and increases soma volume of oxytocin neurons in the paraventricular nucleus. Additionally, recent studies have shown that tunnel handling can reduce stress and improve performances in behavioural experiments^37-39. Therefore, it is important to take care to avoid stressing mice during or before behavioural experiments.

Mice do not like to be exposed in the open – for example in the open field test or elevated-plus maze — preferring more enclosed and darker places^40,41. In the context of the RAM, the central hub is the most open space and it is from there that mice decide which arm to visit next. If they are anxious in the central hub, they might be distracted from the task and perform below par. For that reason, we decided to catch mice in one of the radial arms (see procedure part), not in the central hub, so that they do not associate the hub with a fearful experience.

When exploring the 8-arm RAM for food rewards, rats emerging from a given arm tend to turn at right angles for their next arm visit, skipping the intervening arm, and rarely visit arms sequentially (what we term “daisy-chaining”); that is to say, rats tend to visit in the order 1,3,5,7 not 1,2,3,4 ^1,42,43. In contrast, mice often adopt a daisy-chaining strategy, possibly because of their body size and flexibility. The daisy-chaining strategy is a fairly cost-effective way for mice to recover the food rewards, but it is a considerable drawback for the experimenter trying to test spatial working memory because it circumvents the purpose of the experiment⁴⁴. It is known that mice rely on distal visual cues more than proximal visual cues when navigating the maze^45-47; we found that incorporating additional proximal cues did not reduce daisy-chaining behaviour, which is consistent with the claim by Olton et al. that rats did not utilize the intramaze cues in solving the task¹. The daisy-chaining strategy favoured by mice is another reason that mice are generally considered to perform less well than rats in the RAM. Thus, we tried to change their strategy or improve their performance by the following protocol.

Training and testing working memory in the RAM

Working memory is a cognitive process that can hold a limited amount of information for a short time while it is manipulated and used in the planning and execution of a mental task. Examples include memorising the entry code for a door at a remote location or memorising the path to a specific destination. Dual N-back task and spatial span task are commonly used visuospatial working memory tasks in that human subjects are asked to memorise the location, the identity of the stimuli and their temporal order. Takeuchi et al. found that the amount of training in successful working memory tasks in humans which include dual N-back task correlates with the increased fractional anisotropy in the anterior part of the body of the corpus callosum in their study using MRI⁴⁸. In our earlier study⁴⁹ using the protocol described in this paper, we found that myelinating oligodendrocyte increase in the anterior corpus callosum and other brain areas that are thought to be involved in working memory processing including the anterior cingulate cortex which is believed to be involved in attention and the oligodendrocyte increase closely correlate with their working memory task performance, which is consistent with the study by Takeuchi et al^48-50. However, it is still necessary to study whether there are psychological process similarities between humans and mice by for example electrophysiology. Working memory is closely related to attention. It correlates with measures of “fluid intelligence” in humans and is important for learning, hence an important subject for study in its own right. Olton and Samuelson described an 8-arm radial maze (RAM) task for training and assessing spatial memory in rats¹. In the free choice experiment, they provided all maze arms with 0.1 g rat chow and the rat chow was not replenished during the test. So, 8 rewards are obtained in one test. All arms were open and rats had immediate access to all the arms during the experiment. Since food rewards were not replenished, the rats had to use working memory to avoid re-visiting previously visited arms, in order to recover as many rewards as possible before the end of the task. Each rat was placed in the central hub at the start and allowed to enter any arm at will to recover the food reward until 16 arm choices were made or 10 min elapsed. Olton et al. counted the number of distinct arms rats visited within the first 8 choices and found all rats chose an average of more than 7 distinct arms within 10 days¹. We tried a similar protocol with mice and found that mice went for a daisy chaining in a few days unlike rats and did not change their strategy afterwards, which circumvents the purpose of assessing their working memory and brought us to confine the mouse in the central hub between arm visits hoping confinement in the central hub prevents them from making a daisy chaining.

Olton et al. confined rats for 1 min or 15 sec in the central hub to interfere with any sequence of choices and it prevented them from adopting their usual tendency to progress around the maze continuously in a clockwise or anti-clockwise direction^1,43. If the choice of any arm is random, equally probable and completely independent of previous choices, the experiment is an example of the classic “occupancy problem”¹, so the probabilities p(n,8) of selecting n distinct arms within the first 8 successive arm visits are: p(1,8)=4.8x10^-7, p(2,8)=4.2x10^-4, p(3,8)=0.019, p(4,8)=0.17, p(5,8)=0.42, p(6,8)=0.32, p(7,8)=0.067, p(8,8)=0.0024. From these probabilities, one can calculate that the mean chance likelihood to choose distinct arms in 8 selections is 5.3 new arms out of 8 arm entries (Expectation value = = 5.25; 5.3/8= 66%). Olton et al. reported that, over a 10-day training period, their rats entered on average 5.7/8 distinct arms (71%; range: 4.2-6.6) over days 1-5, increasing to 7.6/ 8 distinct arms (95%; range: 7.2-8.0) during days 6-10¹, which is significantly higher than chance and indicates that rats used working memory to avoid food-depleted arms.

We tried a similar protocol with mice, confining them to the central hub for 5 sec between arm visits. We found that mice entered an average of 5.7/ 8 different arms per 8-arm trial over days 1-5 (71%; range: 5.2-5.8; eleven C57BL/6 mice) and 5.9/ 8 arms during days 6-10 (74%; range: 5.7-6.5; eleven C57BL/6 mice). Therefore, mice were unable to learn, or learned very slowly, to improve their reward-finding performance significantly above chance levels during the 10 days of the experiment. We therefore decided to abandon those protocols above as unsuitable for mice, opting instead for a “forced run/free run” paradigm, similar to that described for rats by Sasaki et al.¹⁵

In the protocol of Sasaki et al.¹⁵, all arms were baited at the beginning of the experiment, rewards were not replenished, and each rat was initially confined to the central hub of an open elevated 8-arm RAM and from there was released into each of four pre-selected arms, one arm at a time without confining in the arms and the central hub between arm visits (the “forced run”). Following the 4^th arm visit and the rat goes back to the central hub, all doors were immediately opened to allow the rat to visit any and all arms at will (the “free run”). In order to collect all the remaining food rewards in the most efficient way during the free run, animals need to remember and avoid any arms they have already visited in either the forced or free runs.

We used a very similar paradigm to that of Sasaki et al.¹⁵ (Fig. 4a), but using mice rather than rats. Sasaki et al.¹⁵ used 10 min habituation periods at the beginning of every day’s RAM training; we replaced this with a single 6-day habituation period prior to training. Another modification was that we allowed the mice to make 4 additional arm visits at the end of each trial, after they had recovered all the rewards. This was intended to reinforce the fact that rewards are not replenished in previously-visited arms. Also, since mice learn to “avoid regret”^51,52, it might discourage mice that had been “daisy-chaining” during their most recent trial from persisting with that strategy in the following trial, because daisy-chaining goes unrewarded during the 4 additional arm visits immediately prior to their starting the next trial and they will never find 4 adjacent empty food wells before finishing all 8 rewards when they do a daisy chaining. Another reason for the additional 4 arm visits was to prevent the mice from associating successful recovery of all rewards with the possibly fearful experience of being caught by the experimenter at the end of the task. Mice might have developed a fear memory of being caught during the 4 additional arm visits, which could be one of the reasons why mice spent more time towards the end of the experiment (Fig. 4h). Indeed, mice broke the daisy chaining during the additional 4 arm visits if they did it in the trial. Furthermore, mice were confined in each arm for 15 sec during the forced run and then in the central hub for 5 sec before opening all the doors to start the free run. Without confining the mouse in the central hub, mice tended to go a daisy chaining as described above.

For the forced run, we selected from a set of five 4-arm pseudorandom sequences that contain 3 or fewer adjacent arms. These were (1,3,5,7), (2,4,5,8), (1,4,5,7), (2,3,4,7) and (1,2,5,6) and their variations of different arm orders and rotations, numbering the arms from 1-8 clockwise from the top (Fig. 2c). Each mouse in a cohort undertook 6 trials (one trial = forced run +5 sec confinement in the central hub + free run) per day during the 9-day post-habituation stage of the RAM test. During the free run, if a mouse made only 4 arm visits to recover all 4 rewards remaining after the forced run, this was termed a “perfect trial”. Their performance during the 9 days of training was quantified by 1) “success rate”, calculated as 4/N, where N is the total number of arm visits needed to recover all rewards during the free run or 2) the proportion of trials that were “perfect trials” (i.e. N=4). For example, 1) if a mouse needed 6 arm visits to retrieve the 4 remaining rewards during the free run then the “success rate” would be 4/6 = 0.67 or 67% and a perfect trial would score 100% on this scale; 2) if a mouse made 5 perfect trials on the day the score would be 5/6 = 83.3 %. Using either scoring system we found that their performance was almost flat for the first 3 days of the 9-day RAM testing period but improved markedly after that (Fig. 4b, c). The average distance travelled in the free runs became shorter over time but the average speed and acceleration were fairly constant from day to day (Fig. 4d-g).

We can manually calculate the success rates that can be achieved by daisy chaining, for each of the 4-arm forced run patterns that we employed (previous paragraph), assuming that mice start their free run at random in any one of the 8 arms of the RAM. The average score over many trials that can be achieved by daisy-chaining is between 53% and 56% for the individual forced run arm patterns, or ~55% overall arm patterns. The probability of obtaining a perfect score (100%) by daisy-chaining is zero, and that by random arm visits is <1% (4/8 x 3/7 x 2/7 x 1/7 = 0.87%), assuming mice never return to the arm they have just exited. Therefore, the average daily success rate scores (>80%) and percent of perfect trial scores (>60%) reached by our mice in the final 2 days of the 9-day testing period (Fig. 4b,c) could only be achieved consistently using a working memory-based strategy.

Mice that were familiar with the task tended to loiter at the last food well, or in the central hub, once they had consumed all the rewards, and they also tended to spend more time on the extra 4 arm visits at the end of the task (Fig. 4h and movie 9), suggesting that they had learned the “rule” that each arm contained one and only one reward, and also that they could keep track of the fact that they had visited all available arms. For mice of a different genetic background than that used here (C57BL/6), results might vary because genetic background matters in behavioural experiments^53-55. However, in our previous study⁴⁹ , using mice with a mixed C57BL/6, CBA, 129P2 background, we found that their working memory task performance was greatly improved using the protocol we describe here, with fewer mice resorting to daisy-chaining during the free run. Koike et al.⁵⁶ reported that all 129 mouse inbred substrains carry a Disc-1 deletion polymorphism consisting of a 25 bp deletion in exon 6 of Disc-1 that causes a frameshift and truncation of the DISC-1 protein, affecting its functionality^57,58. Koike et al.reported that this mutation causes working memory impairment using T-maze delayed alternation task⁵⁶, another appetitively motivated win-shift maze task, possibly explaining why the control mice used in our previous study⁴⁹ had slightly lower working memory scores than the inbred C57BL/6 mice used here.

Sex differences
We only used male mice in the experiments reported here because of the various impacts of sex in cognition, learning and memory observed in different species (rodents and humans) and because the stress-coping responses differ between males and females^36,59-63. However, when interpreting results obtained with rodents for possible benefits for humans, it is important to study both sexes. Guidelines from NIH and other funding councils also say male and female animals need to be used.

Reducing anxiety in mice during handling
Handling stress is well known to influence animals physiologically and behaviourally^64-68. Also, Sandi et al demonstrated that spatial memory impairment caused by acute stress is associated with decreased expression of neural cell adhesion molecule (NCAM) in the hippocampus and prefrontal cortex by testing rats in a hippocampus-dependent water maze task using a female cat as a stressor between the last learning trial and the probe test⁶⁹, suggesting that the hippocampus and prefrontal cortex are vulnerable to high levels of stress. Therefore, it is important to reduce stress/anxiety in experimental animals to avoid unexplained variations within and between animal studies. Especially, working memory tasks require animals to focus during the task, so unnecessary distractions must be avoided. Published protocols for testing working memory in rats, or working memory and reference memory together, often do not include important details such as how, when and from where experimenters remove the animals from the maze at the end of the task. Since mice are less social than rats with the handler and are more prone to stress, such details are doubly important for experiments with mice. Rodents can develop episodic memory, which is a form of explicit long-term memory that includes memory of times, locations, and associated emotions^70-72. Therefore, we chose not to remove mice from the maze immediately after they had collected all food rewards in the RAM but instead allowed them to make 4 subsequent arm visits without gaining food rewards, our rationale being to prevent the mice from associating successful recovery of all rewards with the possibly fearful experience of being caught by the experimenter, which is important especially for the early stage of the learning by them in the RAM. In addition, the additional 4 unrewarded arm visits can reinforce the fact that food rewards are not replenished in arms that the mice have previously visited.

If the handling is not too stressful for mice, they interact with the experimenter more often³⁷. To avoid stress we lift the mouse from its home cage at the start of the RAM experiment using a paper tube in the home cage supplied when they were separated from the grouped cage for experiment³⁷, or else we allow the mouse to climb onto the experimenter’s hand voluntarily (Movie 1), when handling them in their home cage before the experiment and when bringing them from their home cage to the maze. When retrieving the mice from the maze at the end of an experiment, we do not chase them in the maze but instead block the maze arm entrance with one hand and let mice climb on it voluntarily, while covering the maze arm with the other hand to prevent the mouse from jumping out of the maze. This reduces the stress of being pursued in the maze or caught by the tail. Male experimenters are often more stress-inducing than female experimenters where mice are concerned⁷³. However, by paying attention to the stress-reducing steps we have described, we found that male experimenters could conduct the RAM tests without major problems.

Materials and Methods

Mice

We used 3-month-old male C57BL/6J mice. Mice were kept in individually ventilated cages (IVCs) on a 12hr light/12hr dark cycle (lights on from 07:00-19:00 h). The temperature in the room was 20°C – 24°C and the relative humidity was 45% – 65%. Mice were group-housed, 2 – 5 per cage with ad libitum access to standard chow and water. For dietary restriction and habituation to the reward food (condensed milk), they were separated and caged singly for one week prior to the RAM habituation steps. During this week, mice were handled to familiarise them with the experimenter, and they were weighed daily to monitor body weights and maintained between 85% and 90% of their starting weights. Mice had access to water ad libitum throughout the experiment. All mouse experiments were pre-approved by the UCL Ethical Committee and authorized under the Animals (Scientific Procedures) Act 1986 and subsequent legislation of the UK Government. !CAUTION Mouse genetic background matters for the behaviour experiments⁵³. If they have a mixed background for breeding purposes, backcrossing to C57BL/6 for 5-10 generations is recommended. Also, it is recommended by the Jackson Laboratory to refresh the genetic background of the transgenic mice every 5-10 generations by backcrossing to the inbred control strain to mitigate the impact of genetic drift between mutant/ transgenic and control strains.

Mice were provided in their home cage with sawdust as bedding material, a wooden bite-block, a paper tunnel, a paper igloo and paper nesting strips. The paper tunnel is important for mouse handling³⁷.

Mouse handling.
Familiarity between mice and experimenter is critically important and needs to be established early in the run-up to the experiment. Put your hand into the cage holding 2-5 mice when they are 2-3 months old and slowly accustom them to your touch, but without picking them up. After some minutes of this, pick a mouse up with the paper tube as demonstrated by Gouveia et al³⁷ or allow the mouse to climb on your hand voluntarily (Movie 1), lift it to chest height and then put it back in the cage again. Do this for each mouse in the cage. Do not hold for too long in the early stages or they can become agitated and jump off your hand. pCRITICAL Never pick them up by the tail, which is stressful for mice. Mice need to be well-habituated and relaxed or they will not learn well in the maze later.

Behaviour room

The room setup is shown in Fig. 2d. There are extramaze visual cues such as Zener card (also known as ESP card) symbols printed on A4 size papers put on the ceiling above the maze, a TV monitor hanging from the ceiling, a bookshelf on a wall and a metal ball above the maze which is located near the arm end, not above the central hub. A room of at least 3 meters square is large enough for this experimental setup. The maze floor is illuminated evenly with indirect lights (~80 lux at the centre of the maze). During the behaviour experiments, white noise was played at 70-75 dB, measured at the central hub, because it is known that exposure to moderate-intensity white noise leads rats to perform better in a geometrically complex maze ^74,75. White noise seems to do more than simply mask unavoidable noise in the environment; 70 dB white noise increases choline uptake in several brain regions of rats including the prefrontal cortex, hypothalamus and hippocampus (but not in the striatum)⁷⁶ and RAM working memory performance is known to depend on the integrity of cholinergic pathways among the hippocampus, frontal cortex and forebrain^6,77. pCRITICAL It is known that animals use both distal and proximal visual cues to navigate accurately in their environments. However, mice depend on distal visual cues more than proximal cues to navigate during spatial learning⁴⁵. We introduced proximal cues into our 8-arm RAM but they did not reduce daisy-chaining behaviour, whereas we found that distal cues (on the ceiling and walls of the room) did reduce daisy-chaining and were essential for spatial navigation in the RAM.

8-arm radial arm maze and door control

The 8-arm RAM with servo-driven doors, a monochromatic CCD camera and the door controller box were purchased from Tracksys Ltd (Nottingham, UK; Fig. 2a). It was placed on a bench high enough that the seated experimenter was not visible from within the maze. The doors are housed beneath the floor of the apparatus from where they are driven up and down to close and open the arms (Fig. 2b). This is beneficial if you do in vivo live imaging or electrophysiology. The doors in the maze were controlled by EthoVision XT13 software (Noldus, Wageningen, the Netherlands) and the control script was prepared at UCL by T.S. (Fig. 3) (available on request).

Food rewards

The reward provided in the arms of the RAM was sweetened condensed milk (Nestle Carnation) diluted 1:1 (v/v) with distilled water. During habituation days 3-6 in the RAM, they were rewarded only in the maze, not in the home cage, with 70 µl of the dilute condensed milk in each arm. During the working memory task, we used 40 µl per food well in each arm to prevent the mice from becoming sated and de-motivated to collect all the rewards. The food wells are ~1 cm deep and wide enough to accommodate a white plastic bijoux bottle cap containing the condensed milk. Since the food wells are deep enough, the mouse cannot see the milk droplet from >10cm away (Fig. 2b). Furthermore, we placed the milk droplet on the side of the food receptacle nearest the central hub, making it invisible from the arm entrance even if the mouse rears on its hindlimbs. !CAUTION It is important to provide equal volumes/concentrations of food reward in each arm to avoid arm bias; it is advisable therefore to pre-prepare a single large batch of condensed milk in advance, enough to last throughout the day’s experiments.

Body weight monitoring.

After the initial handling in the group cage, lift a mouse in the paper tube as demonstrated by Gouveia et al³⁷, transfer it gently to your hand and then to the weighing scale. After weighing, separate it from the group cage into a new home cage holding a paper igloo, paper tube, wooden bite block and paper nesting strips. pCRITICAL If mice are not in the normal body weight range for the mouse line you are using (20-30 grams at 8-12 weeks old: https://www.criver.com/products-services/find-model/c57bl6-mouse?region=3671 ) at the first body weight check, the mouse should be removed from the cohort because it will be very difficult to maintain its body weight in the required range later, and possibly they can be used for other behavioural experiments that do not require dietary restriction or other histological experiments.

Dietary restriction

Dietary restriction and habituation to the condensed milk reward start after mice are split from the group cage and housed singly. By single housing, we can prevent monopolizing chows by a stronger mouse or mice when they are housed in a group. C57BL/6 mice eat 3-4 g of standard chow every day⁷⁸. Put 2-3 g of chow into the cage on the first day of food restriction and observe how much their body weight drops overnight as a guide to how much food to give the following day. From then on, give a sufficient daily amount of standard chow to maintain 85-90% of their starting body weight throughout the experiment. Each day, place a 15 ml centrifuge tube cap (in any colour but not white) containing 1 ml of diluted condensed milk, as used for rewards in the RAM, on the floor of the home cage to let the mouse become habituated to its taste. Mice normally consume the milk within 5 min after placing it in the cage. !CAUTION It is better to watch from a distance while they consume the milk during the first few days, to avoid food neophobia during the experiment. pCRITICAL If body weight drops below 80% of the starting weight, the mouse normally should be culled on humanitarian grounds. This depends on the regulations in different countries, and you need to follow the guidance in your country.

8-arm radial arm maze and door control

The 8-arm RAM with servo-driven doors, a monochromatic CCD camera and the door controller box were purchased from Tracksys Ltd (Nottingham, UK; Fig. 2a). It was placed on a bench high enough that the seated experimenter was not visible from within the maze. The doors are housed beneath the floor of the apparatus from where they are driven up and down to close and open the arms (Fig. 2b). This is beneficial if you do in vivolive imaging or electrophysiology. The doors in the maze were controlled by EthoVision XT13 software (Noldus, Wageningen, the Netherlands) and the control script was prepared at UCL by T.S. (Fig. 3) (available on request).

Habituation and working memory task

After dietary restriction and habituation to the food reward, the mice undergo 6 days of habituation to the maze apparatus, in three stages. Leave them undisturbed for 5 to 10 min after taking them into the testing room. Animals should be in an optimal condition of arousal for behavioural experiments. Shake their cages a little by finger tapping 5 to 10 min before their trial starts, in order to raise their arousal states. !CAUTION If you leave mice untouched for 30 min, they can fall asleep in the cage. In this protocol, habituation steps take 6 days, which can adjust their body clock for the experiment even if you cannot reverse the light/dark cycle in your animal facility.

Habituation stage 1 (days 1 and 2; Movie 2 and Fig. 3a). Mice explore the maze freely, without food rewards, for 30 min each day. During this time the doors are operated by our inhouse program by T.S. on EthovisionXT13, to familiarize the mice with door movements and sounds. A mouse is placed in the central hub with all doors closed; after 5 sec all doors open, and mouse tracking starts. When the mouse enters an arm and its tracking point is 4 cm from the central hub, all doors are closed, then 3 sec (the time for opening and closing the doors) later the door to the occupied arm is opened again. When the mouse re-enters the central hub and its tracking point is 3 cm from the door, that door is closed and the mouse is confined in the central hub for 5 sec before opening all doors to allow the mouse a subsequent arm visit. This sequence is repeated for 30 min and the session ends when the mouse goes back to the central hub. Collect the mouse in an arm. (NOTE) After 30 min, mice might not have visited all 8 arms of the maze. This is normal and is not a problem. (NOTE) Mice tend to visit arms in an unsystematic, non-sequential pattern and move cautiously, spending uneven times in each arm. pCRITICAL Mice can climb on the maze walls and/or jump out of the maze at this stage. If this seems possible (e.g. the mouse tries to climb on the walls), it can be prevented by the experimenter making the mouse aware of his/her presence by approaching the maze and making some noise such as clothes rustling, not needing to make big sounds. Therefore, careful observation of the mouse’s behaviour in the RAM and quick action might be required. After a few repetitions of this tactic, mice will stay in the maze. If, despite everything, the mouse will not stay in the RAM during habituation, it should be removed from the analysis.

Habituation stage 2 (days 3 and 4; Movies 3, 4 and Fig. 3b). This stage is to encourage mice to explore for rewards in the maze. Once the mouse is placed in the central hub, all doors will open in 5 sec, and all doors remain open till the end. All arms are baited with 70 µl of dilute condensed milk and they are not replenished during the habituation trial. Mice are allowed to enter all the arms at least twice within 30 min, and the session ends when the mouse enters all arms twice and then goes back to the central hub. If 30 min elapse, the session ends even if the mouse did not visit all arms. Collect the mouse in an arm. Each mouse experiences this habituation step twice per day, each mouse in sequence. (NOTE) At this stage, mice may do a daisy chaining, but it is not necessary to worry because the additional 4 arm entries in the working memory task can discourage them from making a daisy chaining. !CAUTION In this step, mice can stop in the arm without consuming the reward. On day 4 they should be more fluent than on day 3, but they still can stop in the middle of the arm (Movie 3 and 4).

Habituation stage 3 (days 5 and 6; Movie 5 and Fig. 3c). Mice experience a combination of stages 1 and 2 and are allowed to enter each arm at least once. The arms are baited with food rewards as in stage 2 and the doors operated as in stage 1, except that the mice are confined to the central hub for 3 sec rather than 5 sec between arm visits. Once the mouse has visited all arms at least once and returns to the central hub or 30 min elapse (normally ends within 20 min), the trial ends. Collect the mouse in an arm. All mice in the cohort complete this habituation stage twice per day, each mouse in sequence. Until the end of day 5, they can spend variable times in the arms and can return to the central hub without touching the rewards (Fig. 1A). However, on the sixth day, animals enter each arm at least once within 7-15 min, spend similar period in each arm, which indicates animals are ready for the task (Fig. 1 A day6 lower row and 1B). pCRITICAL If any mice still try to jump out of the maze at this stage, it is best to remove them from the analysis.

Working memory task (days 7-15; Movies 6-9 and Fig. 3d). In the “forced run” phase of the task the mouse starts in the central hub with all doors closed and all arms baited with 40 µl (not 70 µl). After 5 sec the door to one of 4 pseudo-randomly selected arms is opened to allow the mouse to enter and consume the reward. The door is closed when the mouse tracking point is 3 cm inside the arm, and remains closed for 15 sec before opening to allow the mouse to re-enter the central hub, then the mouse enters the central hub and is 3 cm away from the door just passed, the door for the 2^nd pre-selected arm opens, and so on until all 4 selected arms have been visited and the rewards consumed. After a 5-sec confinement in the central hub, all doors are opened together to allow the mouse to enter arms at will to find and consume the remaining 4 rewards (the “free run”). The task is terminated when the mouse collects all 8 food rewards. During the free run, the experimenter needs to watch the mouse to record its performance and to allow it 4 extra arm visits after the mouse finishes collecting all the rewards. At the end of the trial, the experimenter collects the mouse in the final arm (not the central hub) by allowing it to climb onto his/her hand voluntarily while using his/her other hand above the RAM to prevent the mouse from jumping. The time taken for one mouse to go through the whole procedure once is usually 5-10 min. All mice in the cohort go through this procedure one after another and the whole sequence is repeated 6 times per mouse per day for the 9 days of the task. (NOTE) The average duration of the four extra arm visits should become longer in later trials, as mice learn to recognize when no further rewards are available or that there are no novel, unvisited arms to go in to. pCRITICAL We trigger the doors to close when the mouse tracking point is in an arm 3 cm from the central hub during the working memory task, but 4 cm from the hub during habituation, because mice are more hesitant during habituation and initiating door closing too early can catch and lift the mouse by the door, causing it anxiety and the software to lose track of it. However, these parameters are likely to need adjustment for a different maze apparatus as the speed of door movement might vary. If the mice risk being caught by the moving door they become anxious, tend to move very quickly and perform less well than they are capable of. pCRITICAL It is important to clean the maze floor after one trial to remove any traces of the rewards on the maze floor, faeces and urine using a paintbrush, and dry and wet tissue papers. It is also important to make sure the reward is not on the experimenter’s gloves and pipette body, which will keep the maze clean. Sometimes mice leave urine in the food wells, and it is wise to rinse them with water and dry with a dry tissue paper before the next trial. If it is possible, prepare some extra food wells to replace.

Scoring and statistics

The success rate (%) is calculated by [4/ (number of arms visited in the free run) x 100]. The resulting data of the control and experimental groups are compared using repeated measures two-way ANOVA with a post hoc-analysis (e.g. Bonferroni test) [measuring each subject at 9 different time points]. If comparing groups, at least 10 and preferably ~20 mice in each group should be tested. Where the effect of gene manipulation or making lesions or drug treatment is obvious, the sample sizes can be smaller to conclude. The number of “perfect scores” (zero working memory errors) can also be used as a scoring system. !CAUTION Mice can make quick turns immediately after entering an arm. Do not count and score this as an entry, as it is a quick mistake correction which means working memory is valid⁷⁹ (Movie 9, “Oops turn”). However, if the mouse crosses the mid-point of the arm or stays in the arm for more than 3 sec, count and score the entry.

Troubleshooting

Problem 1: Mice tend to climb on the maze walls.

Solution: This can happen when mice are not habituated to the maze enough. However, showing yourself to the mice (i.e. approaching the maze and letting the mouse know you are there in the room) can discourage this escape behaviour. You may need to push the mouse back into the maze gently, but normally, if your hand is approaching the mouse, it will go back into the maze voluntarily. This must be done quickly if the mouse starts to climb on the maze wall during habituation, otherwise, the mice might develop an interest in the environment outside the maze and escape behaviour becomes habitual.

Problem 2: Mice do not spend much time in the central hub and move from one arm to another very quickly.

Solution: This is because habituation is not enough and mice are scared of the doors. We tried to reduce stress on the mice as much as possible during the task so that they can pass the central hub without pressure or fear from the moving doors or experimenter’s hand. For the working memory task, we let the mouse explore 4 more arms after the mouse collects all the baits in the maze. We managed to let them explore the maze slower than before by adding the 4 extra arm visits. Thus, if the mouse strain is more prone to stress compared to C57BL/6 mice, more habituation is needed.

Problem 3: Mice return to the arm just visited when the door is closing during the forced run and mouse tracking and maze control stop.

Solution: In our protocol, once the tracking point of the mouse is in the central hub 3 cm or more from the arm door, the door closes, preventing the mouse from returning to the arm just visited. When we programmed to close the door just after the mouse entered the central hub, mice sometimes returned quickly to the arm just visited, seemingly they were scared of the doors, the mouse-tracking and door control stopped working because the program is not designed for the return and the trial was stopped. Stopping a trial in the middle of it is not good as mice will have different experiences unnecessarily. During the early stages of habituation, mice often stop at the arm exit point, extending their body to scan around the hub and retreat into the arm they were in, suggesting that mice fear the moving doors. We found by trial and error that 4 cm and 3 cm for habituation and working memory task, respectively, were the optimal distance from the arm entrance to trigger door closing and prevent the mouse from returning to the arm just visited. We set the mouse tracking point to the centre of the body, not those for the tail and the nose, because the software sometimes confuses the nose and tail tracking points.

Problem 4: Body weight, body size and litter size matter.

Solution: We always use mice within the normal body weight range for that age and strain of mouse because otherwise it is difficult to maintain their body weight between 85% and 90%. Also, litter size also could influence spatial memory and anxiety-like behaviour because of early life overfeeding in a strain called NMRI⁸⁰. Small litter size causes over-feeding and increased body weights both in NMRI and C57Bl/6 mouse strains. However, the two strains solved spatial learning tasks differently. C57Bl/6 mice did not show differences between the groups of 4 pups (small litter size) and 10 pups (control litter size) whereas NMRI mice showed significant impairment in the learning phase and the probe test of Morris water maze (MWM) test and slight impairment in Y-maze spontaneous alternation task though it is not significant. Furthermore, small litter size significantly decreased hippocampal brain-derived-neurotrophic-factor (BDNF) protein level and increased hippocampal interleukin-1b (IL-1b) protein level after MWM and Y-maze tests in NMRI strain. However, those changes were not observed in C57BL/6 strain. Additionally, NMRI strain showed a significant increase in anxiety-like behaviours and stress-induced corticosterone if small litter size, while in contrast small litter size resulted in the reduction of anxiety-like behaviours and corticosterone levels in the C57Bl/6 strain. Thus, the normal litter size of your mouse line should be known for behaviour experiments before starting experiments.

Problem 5: Moving mice to another mouse holding room

Solution: If we need to move mice to another animal holding room, we move them at least one week prior to the handling and food restriction to acclimatize them to the new environment. This will minimize stress on the animal, which has widespread effects on their physiology.

Diet restriction and handling: 7 days

Habituation to the end of the experiment: 15 days

After 6 days of habituation and 9 days of maze training, male C57BL/6 mice regularly achieve ≥80% success rate, defined as 4/(4+E) where E is the number of errors. About 70% of trials are without working memory error (4-5 perfect trials/6 trials) at the end of the experiment.

1 Olton, D. S. & Samuelson, R. J. Remembrance of Places Passed - Spatial Memory in Rats. J Exp Psychol-Anim B 2, 97-116, doi:Doi 10.1037/0097-7403.2.2.97 (1976).

2 Kraemer, P. J., Gilbert, M.E. & Innis, N.K. The influence of cue type and configuration upon radial-maze performance in the rat. Animal Learning & Behavior 11, 373–380, doi:https://doi.org/10.3758/BF03199790(1983).

3 Wenk, G. L. Assessment of spatial memory using the radial arm maze and Morris water maze. Curr Protoc Neurosci Chapter 8, Unit 8 5A, doi:10.1002/0471142301.ns0805as26 (2004).

4 Morris, R. G. M. Spatial Localization Does Not Require the Presence of Local Cues. Learn Motiv 12, 239-260, doi:Doi 10.1016/0023-9690(81)90020-5 (1981).

5 Buresova, O., Bures, J., Oitzl, M. S. & Zahalka, A. Radial maze in the water tank: an aversively motivated spatial working memory task. Physiol Behav 34, 1003-1005, doi:10.1016/0031-9384(85)90028-9 (1985).

6 Buresova, O., Bolhuis, J. J. & Bures, J. Differential effects of cholinergic blockade on performance of rats in the water tank navigation task and in a radial water maze. Behav Neurosci 100, 476-482, doi:10.1037//0735-7044.100.4.476 (1986).

7 Olton, D. S., Branch, M. & Best, P. J. Spatial correlates of hippocampal unit activity. Exp Neurol 58, 387-409, doi:10.1016/0014-4886(78)90096-1 (1978).

8 Olton, D. S., Walker, J. A. & Gage, F. H. Hippocampal connections and spatial discrimination. Brain Res 139, 295-308, doi:10.1016/0006-8993(78)90930-7 (1978).

9 Olton, D. S. & Werz, M. A. Hippocampal function and behavior: spatial discrimination and response inhibition. Physiol Behav 20, 597-605, doi:10.1016/0031-9384(78)90252-4 (1978).

10 Olton, D. S. & Papas, B. C. Spatial memory and hippocampal function. Neuropsychologia 17, 669-682, doi:10.1016/0028-3932(79)90042-3 (1979).

11 Honig, W. K. Studies of Working Memory in the Pigeon. 1 edn, 38 (1978).

12 Harper, D. N., Dalrymple-Alford, J. C. & McLean, A. P. Production of a serial position effect in rats using a 12-arm radial maze. J Neurosci Methods 44, 197-207, doi:10.1016/0165-0270(92)90011-2 (1992).

13 Schmitt, W. B., Deacon, R. M., Seeburg, P. H., Rawlins, J. N. & Bannerman, D. M. A within-subjects, within-task demonstration of intact spatial reference memory and impaired spatial working memory in glutamate receptor-A-deficient mice. The Journal of neuroscience : the official journal of the Society for Neuroscience 23, 3953-3959 (2003).

14 Brown, M. F. The effects of maze-arm length on performance in the radial-arm maze. Animal Learning & Behavior 18, 13-22, doi:doi.org/10.3758/BF03205234 (1990).

15 Sasaki, T. et al. Dentate network activity is necessary for spatial working memory by supporting CA3 sharp-wave ripple generation and prospective firing of CA3 neurons. Nat Neurosci 21, 258-269, doi:10.1038/s41593-017-0061-5 (2018).

16 Jew, K. et al. Selective memory and behavioral alterations after ambient ultrafine particulate matter exposure in aged 3xTgAD Alzheimer's disease mice. Part Fibre Toxicol 16, 45, doi:10.1186/s12989-019-0323-3 (2019).

17 Rai, R., Singh, H. K. & Prasad, S. A Special Extract of Bacopa monnieri (CDRI-08) Restores Learning and Memory by Upregulating Expression of the NMDA Receptor Subunit GluN2B in the Brain of Scopolamine-Induced Amnesic Mice. Evid Based Complement Alternat Med 2015, 254303, doi:10.1155/2015/254303 (2015).

18 Harvey, R. E., Thompson, S. M., Sanchez, L. M., Yoder, R. M. & Clark, B. J. Post-training Inactivation of the Anterior Thalamic Nuclei Impairs Spatial Performance on the Radial Arm Maze. Front Neurosci 11, 94, doi:10.3389/fnins.2017.00094 (2017).

19 G. Brolese, P. L., F. Lopes, C.-A. Gonçalves. Chapter 2 - Prenatal Alcohol Exposure and Neuroglial Changes in Neurochemistry and Behavior in Animal Models. 11-22 (Academic Press, 2017).

20 Christina J. Reppucci⁎, L. A. B., Ashley Q. Chambers, Alexa H. Veenema. Wistar rats and C57BL/6 mice differ in their motivation to seek social interaction versus food in the Social versus Food Preference Test. Physiology & Behvior 227, doi:doi.org/10.1016/j.physbeh.2020.113162 (2020).

21 Whishaw, I. Q. & Tomie, J. Of mice and mazes: similarities between mice and rats on dry land but not water mazes. Physiol Behav 60, 1191-1197, doi:10.1016/s0031-9384(96)00176-x (1996).

22 Cressant, A., Besson, M., Suarez, S., Cormier, A. & Granon, S. Spatial learning in Long-Evans Hooded rats and C57BL/6J mice: different strategies for different performance. Behavioural brain research 177, 22-29, doi:10.1016/j.bbr.2006.11.010 (2007).

23 Hok, V., Poucet, B., Duvelle, E., Save, E. & Sargolini, F. Spatial cognition in mice and rats: similarities and differences in brain and behavior. Wiley Interdiscip Rev Cogn Sci 7, 406-421, doi:10.1002/wcs.1411 (2016).

24 Bisby, M. A., Stylianakis, A. A., Baker, K. D. & Richardson, R. Fear extinction learning and retention during adolescence in rats and mice: A systematic review. Neurosci Biobehav Rev 131, 1264-1274, doi:10.1016/j.neubiorev.2021.10.044 (2021).

25 Mizuno, M., Yamada, K., Olariu, A., Nawa, H. & Nabeshima, T. Involvement of brain-derived neurotrophic factor in spatial memory formation and maintenance in a radial arm maze test in rats. The Journal of neuroscience : the official journal of the Society for Neuroscience 20, 7116-7121, doi:10.1523/JNEUROSCI.20-18-07116.2000 (2000).

26 Ellenbroek, B. & Youn, J. Rodent models in neuroscience research: is it a rat race? Dis Model Mech 9, 1079-1087, doi:10.1242/dmm.026120 (2016).

27 Hefner, K. & Holmes, A. Ontogeny of fear-, anxiety- and depression-related behavior across adolescence in C57BL/6J mice. Behavioural brain research 176, 210-215, doi:10.1016/j.bbr.2006.10.001 (2007).

28 Pattwell, S. S. et al. Altered fear learning across development in both mouse and human. Proceedings of the National Academy of Sciences of the United States of America 109, 16318-16323, doi:10.1073/pnas.1206834109 (2012).

29 Pattwell, S. S. et al. Dynamic changes in neural circuitry during adolescence are associated with persistent attenuation of fear memories. Nat Commun 7, 11475, doi:10.1038/ncomms11475 (2016).

30 Riddle, M. C. et al. Caloric restriction enhances fear extinction learning in mice. Neuropsychopharmacology 38, 930-937, doi:10.1038/npp.2012.268 (2013).

31 Cruz, E., Lopez, A. V. & Porter, J. T. Spontaneous recovery of fear reverses extinction-induced excitability of infralimbic neurons. PloS one 9, e103596, doi:10.1371/journal.pone.0103596 (2014).

32 Barbayannis, G. et al. Differential effects of stress on fear learning and activation of the amygdala in pre-adolescent and adult male rats. Neuroscience 360, 210-219, doi:10.1016/j.neuroscience.2017.07.058 (2017).

33 Chen, C. V., Chaby, L. E., Nazeer, S. & Liberzon, I. Effects of Trauma in Adulthood and Adolescence on Fear Extinction and Extinction Retention: Advancing Animal Models of Posttraumatic Stress Disorder. Front Behav Neurosci 12, 247, doi:10.3389/fnbeh.2018.00247 (2018).

34 Schipper, P. et al. Impaired Fear Extinction Recall in Serotonin Transporter Knockout Rats Is Transiently Alleviated during Adolescence. Brain Sci 9, doi:10.3390/brainsci9050118 (2019).

35 Ishiyama, S. & Brecht, M. Neural correlates of ticklishness in the rat somatosensory cortex. Science 354, 757-760, doi:10.1126/science.aah5114 (2016).

36 Borrow, A. P., Bales, N. J., Stover, S. A. & Handa, R. J. Chronic Variable Stress Induces Sex-Specific Alterations in Social Behavior and Neuropeptide Expression in the Mouse. Endocrinology 159, 2803-2814, doi:10.1210/en.2018-00217 (2018).

37 Gouveia, K. & Hurst, J. L. Reducing mouse anxiety during handling: effect of experience with handling tunnels. PloS one 8, e66401, doi:10.1371/journal.pone.0066401 (2013).

38 Gouveia, K. & Hurst, J. L. Optimising reliability of mouse performance in behavioural testing: the major role of non-aversive handling. Sci Rep 7, 44999, doi:10.1038/srep44999 (2017).

39 Gouveia, K. & Hurst, J. L. Improving the practicality of using non-aversive handling methods to reduce background stress and anxiety in laboratory mice. Sci Rep 9, 20305, doi:10.1038/s41598-019-56860-7 (2019).

40 Walf, A. A. & Frye, C. A. The use of the elevated plus maze as an assay of anxiety-related behavior in rodents. Nature protocols 2, 322-328, doi:10.1038/nprot.2007.44 (2007).

41 Võikar, V., Stanford, S.C. The Open Field Test. Vol. Neuromethods 190 9-29 (Humana, New York, NY, 2023).

42 Haga, Y. Effects of food deprivation and food reward on the behavior of rats in the radial-arm maze. Jpn Psychol Res 37, 252-257, doi:DOI 10.4992/psycholres1954.37.252 (1995).

43 David S. Olton, C. C., Mary Ann Werz. Spatial memory and radial arm maze performance of rats. Learn Motiv 8, 289-314, doi:10.1016/0023-9690(77)90054-6 (1977).

44 Vorhees, C. V. & Williams, M. T. Assessing spatial learning and memory in rodents. ILAR J 55, 310-332, doi:10.1093/ilar/ilu013 (2014).

45 Hebert, M., Bulla, J., Vivien, D. & Agin, V. Are Distal and Proximal Visual Cues Equally Important during Spatial Learning in Mice? A Pilot Study of Overshadowing in the Spatial Domain. Front Behav Neurosci 11, 109, doi:10.3389/fnbeh.2017.00109 (2017).

46 Bannerman, D. M. et al. Dissecting spatial knowledge from spatial choice by hippocampal NMDA receptor deletion. Nat Neurosci 15, 1153-1159, doi:10.1038/nn.3166 (2012).

47 Bannerman, D. M. et al. Hippocampal synaptic plasticity, spatial memory and anxiety. Nat Rev Neurosci 15, 181-192, doi:10.1038/nrn3677 (2014).

48 Takeuchi, H. et al. Training of working memory impacts structural connectivity. The Journal of neuroscience : the official journal of the Society for Neuroscience 30, 3297-3303, doi:10.1523/JNEUROSCI.4611-09.2010 (2010).

49 Takahiro Shimizu, S. G. N., Matthew Swire, Yi Jiang, Matthew Grist, Malte Kaller, Cassandra Sampaio Baptista, David M Bannerman, Heidi Johansen-Berg, Katsutoshi Ogasawara, Koujiro Tohyama, Huiliang Li, William D Richardson. Oligodendrocyte dynamics dictate cognitive performance outcomes of working memory training in mice. Nature Communications (2023 (in press)).

50 Wu, D. et al. Persistent Neuronal Activity in Anterior Cingulate Cortex Correlates with Sustained Attention in Rats Regardless of Sensory Modality. Sci Rep 7, 43101, doi:10.1038/srep43101 (2017).

51 Steiner, A. P. & Redish, A. D. Behavioral and neurophysiological correlates of regret in rat decision-making on a neuroeconomic task. Nat Neurosci 17, 995-1002, doi:10.1038/nn.3740 (2014).

52 Sweis, B. M., Thomas, M. J. & Redish, A. D. Mice learn to avoid regret. PLoS Biol 16, e2005853, doi:10.1371/journal.pbio.2005853 (2018).

53 Sultana, R., Ogundele, O. M. & Lee, C. C. Contrasting characteristic behaviours among common laboratory mouse strains. R Soc Open Sci 6, 190574, doi:10.1098/rsos.190574 (2019).

54 Marchette, R. C. N., Bicca, M. A., Santos, E. & de Lima, T. C. M. Distinctive stress sensitivity and anxiety-like behavior in female mice: Strain differences matter. Neurobiol Stress 9, 55-63, doi:10.1016/j.ynstr.2018.08.002 (2018).

55 Brooks, S. P., Pask, T., Jones, L. & Dunnett, S. B. Behavioural profiles of inbred mouse strains used as transgenic backgrounds. I: motor tests. Genes Brain Behav 3, 206-215, doi:10.1111/j.1601-183X.2004.00072.x (2004).

56 Koike, H., Arguello, P. A., Kvajo, M., Karayiorgou, M. & Gogos, J. A. Disc1 is mutated in the 129S6/SvEv strain and modulates working memory in mice. Proceedings of the National Academy of Sciences 103, 3693-3697, doi:doi:10.1073/pnas.0511189103 (2006).

57 Gomez-Sintes, R., Kvajo, M., Gogos, J. A. & Lucas, J. J. Mice with a naturally occurring DISC1 mutation display a broad spectrum of behaviors associated to psychiatric disorders. Front Behav Neurosci 8, 253, doi:10.3389/fnbeh.2014.00253 (2014).

58 Clapcote, S. J. & Roder, J. C. Deletion polymorphism of Disc1 is common to all 129 mouse substrains: implications for gene-targeting studies of brain function. Genetics 173, 2407-2410, doi:10.1534/genetics.106.060749 (2006).

59 Pawluski, J. L., Kokras, N., Charlier, T. D. & Dalla, C. Sex matters in neuroscience and neuropsychopharmacology. Eur J Neurosci 52, 2423-2428, doi:10.1111/ejn.14880 (2020).

60 Gogos, A., Sbisa, A., Witkamp, D. & van den Buuse, M. Sex differences in the effect of maternal immune activation on cognitive and psychosis-like behaviour in Long Evans rats. Eur J Neurosci 52, 2614-2626, doi:10.1111/ejn.14671 (2020).

61 Espinosa-Garcia, C. et al. Sex differences in the performance of cognitive tasks in a murine model of metabolic syndrome. Eur J Neurosci 52, 2724-2736, doi:10.1111/ejn.14751 (2020).

62 Conner, M. R., Adeyemi, O. M., Anderson, B. J. & Kritzer, M. F. Domain-specific contributions of biological sex and sex hormones to what, where and when components of episodic-like memory in adult rats. Eur J Neurosci 52, 2705-2723, doi:10.1111/ejn.14676 (2020).

63 Sensini, F. et al. The impact of handling technique and handling frequency on laboratory mouse welfare is sex-specific. Sci Rep 10, 17281, doi:10.1038/s41598-020-74279-3 (2020).

64 Meaney, M. J. et al. Early environmental regulation of forebrain glucocorticoid receptor gene expression: implications for adrenocortical responses to stress. Dev Neurosci 18, 49-72, doi:10.1159/000111395 (1996).

65 Nunez, J. F., Ferre, P., Escorihuela, R. M., Tobena, A. & Fernandez-Teruel, A. Effects of postnatal handling of rats on emotional, HPA-axis, and prolactin reactivity to novelty and conflict. Physiol Behav 60, 1355-1359, doi:10.1016/s0031-9384(96)00225-9 (1996).

66 Balcombe, J. P., Barnard, N. D. & Sandusky, C. Laboratory routines cause animal stress. Contemp Top Lab Anim Sci 43, 42-51 (2004).

67 Meijer, M. K., Sommer, R., Spruijt, B. M., van Zutphen, L. F. & Baumans, V. Influence of environmental enrichment and handling on the acute stress response in individually housed mice. Lab Anim 41, 161-173, doi:10.1258/002367707780378168 (2007).

68 Ghosal, S. et al. Mouse handling limits the impact of stress on metabolic endpoints. Physiol Behav 150, 31-37, doi:10.1016/j.physbeh.2015.06.021 (2015).

69 Sandi, C. et al. Acute stress-induced impairment of spatial memory is associated with decreased expression of neural cell adhesion molecule in the hippocampus and prefrontal cortex. Biol Psychiatry 57, 856-864, doi:10.1016/j.biopsych.2004.12.034 (2005).

70 Tulving, E. in Organization of Memory (ed E. Tulving & W. Donaldson) (Academic Press, 1972).

71 Panoz-Brown, D. et al. Rats Remember Items in Context Using Episodic Memory. Curr Biol 26, 2821-2826, doi:10.1016/j.cub.2016.08.023 (2016).

72 Fellini, L. & Morellini, F. Mice create what-where-when hippocampus-dependent memories of unique experiences. The Journal of neuroscience : the official journal of the Society for Neuroscience 33, 1038-1043, doi:10.1523/JNEUROSCI.2280-12.2013 (2013).

73 Sorge, R. E. et al. Olfactory exposure to males, including men, causes stress and related analgesia in rodents. Nat Methods 11, 629-632, doi:10.1038/nmeth.2935 (2014).

74 Prior, H. Effects of the acoustic environment on learning in rats. Physiol Behav 87, 162-165, doi:10.1016/j.physbeh.2005.09.012 (2006).

75 Prior, H., Schwegler, H. & Ducker, G. Dissociation of spatial reference memory, spatial working memory, and hippocampal mossy fiber distribution in two rat strains differing in emotionality. Behavioural brain research 87, 183-194, doi:10.1016/s0166-4328(97)02282-1 (1997).

76 Lai, H. Acute exposure to noise affects sodium-dependent high-affinity choline uptake in the central nervous system of the rat. Pharmacol Biochem Behav 28, 147-151, doi:10.1016/0091-3057(87)90205-x (1987).

77 Granon, S., Poucet, B., Thinus-Blanc, C., Changeux, J. P. & Vidal, C. Nicotinic and muscarinic receptors in the rat prefrontal cortex: differential roles in working memory, response selection and effortful processing. Psychopharmacology (Berl) 119, 139-144, doi:10.1007/BF02246154 (1995).

78 Tordoff, M. G., Pilchak, D. M., Williams, J. A., McDaniel, A. H. & Bachmanov, A. A. The maintenance diets of C57BL/6J and 129X1/SvJ mice influence their taste solution preferences: implications for large-scale phenotyping projects. J Nutr 132, 2288-2297, doi:10.1093/jn/132.8.2288 (2002).

79 Yamamoto, J., Suh, J., Takeuchi, D. & Tonegawa, S. Successful execution of working memory linked to synchronized high-frequency gamma oscillations. Cell 157, 845-857, doi:10.1016/j.cell.2014.04.009 (2014).

80 Salari, A. A., Samadi, H., Homberg, J. R. & Kosari-Nasab, M. Small litter size impairs spatial memory and increases anxiety- like behavior in a strain-dependent manner in male mice. Sci Rep 8, 11281, doi:10.1038/s41598-018-29595-0 (2018).

Acknowledgements

We thank our colleagues in the Wolfson Institute for Biomedical Research for helpful discussions and David Bannerman (University of Oxford) for advice on mouse behaviour and for reviewing an early draft of the paper. We also thank Matthew Grist for technical and administrative support. We thank The Wellcome Trust for funding (awards 108726/Z/15/Z and 214286/Z/18/Z to W.D.R.).

Author contributions
T.S. conceived the method for mice, initiated and performed the experiments and analyzed the data and adjusted the protocol. S.G.N. also performed the protocol and confirmed its reproducibility and confirmed that proximal cues did not affect animals’ behaviours. T.S. wrote the first draft of the paper, and T.S. and W.D.R. reviewed and edited it. W.D.R. obtained funding.

Handlingforsubmission.avi
Movie 1: Mouse handling. If mice are well habituated to the experimenter, mice will interact freely with his/her hands.
Habituationstage1forsubmission.avi
Movie 2: Habituation stage 1 (day 1). The video plays at 5 x actual speed. No arms are baited.
Habituationstep2NOTeatingallrewardsforsubmission.avi
Movie 3: Habituation stage 2 (day 3) showing that the mouse is not necessarily eating all the rewards in the food wells at this point and the mouse can go back to the central hub without touching the food wells. All the arms are baited with 70 µl of the rewards. (Actual speed)
Habituationstep2eatingallrewardsforsubmission.avi
Movie 4: Habituation stage 2 (day 4) showing the mouse spending a long time in arms and eating rewards in arms. (Actual speed)
Habituationstep3forsubmission.avi
Movie 5: Habituation stage 3 (day 5) showing the mouse entering all the arms at least once. (5 x actual speed)
Fromreleasingtocatchingforsubmission.avi
Movie 6: Releasing and catching the mouse for the working memory task (day 1) (Playback, actual speed). Red circles indicate the food wells in the four forced-run arms. Green circles indicate free-run goal arms. Blue numbers indicate the extra arm visits. Mice were released and caught on open hands, not by grasping.
RAMbadperformancemouse2day1forsubmission.avi
Movie 7: Bad performance on day 1 of the working memory task (Mouse 2) (5x actual speed). Red circles show forced run arm wells. The mouse in this video makes many working memory errors — he re-entered arms previously visited in the forced runs and/or the free run.
RAMgoodperformerDay8forsubmission.avi
Movie 8: Trial without working memory error on day 8 (Mouse 2) (forced run: 5x actual speed. free run: actual speed). Red circles show forced run arm wells. The mouse did not make any working memory errors in this trial.
Oopsturnandstopforsubmission.avi
Movie 9: Working memory task, day 8 (Mouse 2, actual speed) showing “oops return” and “stop in the last reward arm”. Red circles indicate the forced run arms and the numbers in light blue indicate the free run arm visiting order.

Download PDF

Version 1

posted

You are reading this latest protocol version

An appetitive spatial working memory task for mice, using a semi-automated 8-arm radial maze

Status:

Version 1

Abstract

Figures

Introduction

Reagents

Equipment

Procedure

Troubleshooting

Time Taken

Anticipated Results

References

Acknowledgements

Supplementary Files

Associated Publications

Status:

Version 1

Privacy Policy

Terms of Service

An appetitive spatial working memory task for mice, using a semi-automated 8-arm radial maze

Status:

Version 1

Abstract

Figures

Introduction

Reagents

Equipment

Procedure

Troubleshooting

Time Taken

Anticipated Results

References

Acknowledgements

Supplementary Files

Associated Publications

Status:

Version 1

Privacy Policy

Terms of Service

Manage Cookie Preferences