Behaviors and Anatomy

Win-Shift and Win-Stay Protocol on the Radial Arm Maze

By February 28, 2018 No Comments

The Radial Arm Maze is an experimental protocol that is used to test working memory [1] as well as reference memory or spatial memory [2] in rats and mice. It has been used in studies of aging, dementia, and pharmacology [3][4]. Working memory is a cognitive process in which an individual uses and manipulates information. Reference memory refers to information that is constant over time, which is related to spatial memory. Spatial memory refers to information about how objects are arranged in space. All three of these processes rely on the hippocampus [5-7].

Physiological Basis of Working and Spatial Memory

The hippocampus is part of the medial temporal lobe of the brain and has been associated with many aspects of memory. The hippocampus proper includes three major regions, known as CA1, CA2, and CA3 [8]. The larger hippocampal formation includes the hippocampus proper and the adjacent dentate gyrus, subicular complex, and entorhinal cortex [8]. Connections between the entorhinal cortex and the hippocampus proper allow sensory information to enter the hippocampus. A large variety of pathways within the hippocampus processes the sensory information and generate output from the hippocampus back to the cortex, which is processed through the subicular cortex [9]. Research in human subjects has suggested that different parts of the hippocampal formation are specialized for spatial memory. Functional imaging studies in taxi drivers suggested that the posterior hippocampus plays a role in spatial memory [10][11]. A review of different studies of hippocampal function has also suggested that the right hippocampus plays a role in spatial memory while the left hippocampus is more involved in autobiographical memory [12].

The Radial Arm Maze

The Radial Arm Maze protocol assesses hippocampal function, both by using spatial cues to remember which arms have been visited previously and by using working memory to process and manipulate information. There are several variants on this procedure, but this post will focus on the win-shift and the win-stay procedures.

Materials Needed

  1. Ethanol and towels or wipes to clean the maze after each use.
  2. Experimental animals, typically rats or mice.

    a. Animals are housed under normal conditions, but should be food-deprived to about 80% of normal caloric intake prior to the start of the experiment.
    b. The animals are typically separated into at least two groups, an experimental group and a control group.

  3. Standard food for the experimental animals.

Equipment Needed

  1. Radial Arm Maze apparatus.

    a. The standard version has eight arms radiating off a central chamber.
    b. The length of the arms and size of the central chamber varies depending on whether the experimental subjects are rats or mice.
    c. At the end of each arm proximal to the central chamber is a door controlling access to the arm.
    d. At the distal end of each arm is a well in which food pellets will be placed.

  2. Visual cues to place outside of the Radial Arm Maze to enable the subjects to navigate the environment based on the external cues.
  3. The use of a video camera is recommended to record the movements of the subjects for later analysis.

Experimental Procedure

Familiarization with the Maze

  1. Over the course of two days, the animals are placed into a central chamber of the maze.
  2. Multiple food pellets are placed along each arm to encourage the animal to explore the arm thoroughly and food is placed in the food well at the end of each arm.
  3. The animal is allowed to explore the maze for 10 minutes.
  4. Once an arm has been explored, it is not re-baited to encourage the animal to explore the other arms of the maze.
  5. The animal is removed from the maze after 10 minutes and returned to the home cage.
  6. The apparatus is thoroughly cleaned with ethanol and the protocol is repeated with another animal.
  7. Typically, animals participate in two familiarization sessions per day, separated by four hours. These sessions occur on two sequential days to completely familiarize the animal with the apparatus and the procedure.

Win-Shift Procedure

Training Phase

  1. One day after the Familiarization procedure, the animals are placed back into the central chamber of the maze and encouraged to explore for five minutes [1].
  2. Salient visual cues outside the maze allow the animals to establish a mental map of the maze.
  3. Only four of the arms are baited with a food reward at the end of the arm and the other four arms are closed off by the doors and unavailable for exploration.
  4. The blocked arms are chosen in a pseudo-random fashion, such that no more than two adjacent arms are blocked at a time.
  5. After the five minutes, the animals are removed from the maze for a period of time (the Inter-Trial Interval; ITI), typically about five minutes.
  6. While the rodents are out of the maze, the maze is cleaned with ethanol. The blocks are removed and the arms that were previously unbaited and blocked are now baited and opened/unblocked. The arms that were previously baited no longer contain a reward.

Experimental Phase

  • After the ITI, the animals are returned to the central compartment of the maze and allowed to explore all eight arms for five minutes. The entire experimental phase can be recorded for later scoring or scored in real time.
Win shift-Procedure-2
  • After each animal, the entire apparatus is again washed with ethanol.
  • The experiment is run twice per day per animal, with a break of about five hours between sessions, until the experimental criteria have been met. One standard criterion is that the control group does not make more than one error per session [1].

Common Difficulties

  1. The animals will sometimes fail to explore the maze entirely during the Training phase. This is addressed by providing a longer Training time for all experimental subjects or increasing the food reward for exploring the maze.
  2. Animals are sometimes observed to jump from one arm of the maze to another. To discourage this behavior, the animal should be picked up and returned to the central chamber, if such jumping occurs.

Data Analysis and Scoring

  1. The rodent is considered to have entered the arm if all four limbs leave the central compartment.
  2. A correct response in this experiment would be the animal visiting the arms that were initially not rewarded.
  3. A Working Memory Error occurs when the animal enters an arm of the maze that it previously entered in the same trial.
  4. A Spatial Memory or Reference Memory Error occurs when the animal enters an arm of the maze that was baited during the training phase.
  5. Data from Training and Experimental Phases are analyzed separately, as are data from Working Memory Errors and Reference Memory Errors.

In the standard analysis, the errors made in four sequential trials (spread over two days) are averaged to form a data point. Data from all animals and all data points are then analyzed using an ANOVA.

Win-Stay Procedure

Training Phase

  1. One day after the Familiarization procedure, the animals are placed back into the central chamber of the maze and encouraged to explore for five minutes [1].
  2. Salient visual cues outside the maze allow the animals to establish a mental map of the maze.
  3. Only four of the arms are baited with a food reward at the end of the arm and the other four arms are closed off by the doors and unavailable for exploration.
  4. The blocked arms are chosen in a pseudo-random fashion, such that no more than two adjacent arms are blocked at a time.
  5. After the five minutes, the animals are removed from the maze for a period of time (the Inter-Trial Interval; ITI), typically about five minutes.
  6. While the rodents are out of the maze, the doors are opened and the arms that were previously baited are baited again. The arms that were previously unbaited remain unrewarded [13].

Experimental Phase

  • After the ITI, the animals are returned to the central compartment of the maze and allowed to explore all eight arms for five minutes. The entire experimental phase can be recorded for later scoring or scored in real time.
Win shift win stay Procedure
  • After each animal, the entire apparatus is again washed with ethanol.
  • The experiment is run twice per day per animal, with a break of about five hours between sessions, until the experimental criteria have been met. One standard criterion is that the control group does not make more than one error per session [1].

Common Difficulties

  1. The animals will sometimes fail to explore the maze entirely during the Training phase. This is addressed by providing a longer Training time for all experimental subjects or increasing the food reward for exploring the maze.
  2. Animals are sometimes observed to jump from one arm of the maze to another. To discourage this behavior, the animal should be picked up and returned to the central chamber, if such jumping occurs.

Data Analysis and Scoring

  1. The rodent is considered to have entered the arm if all four limbs leave the central compartment.
  2. Correct responses in this experiment would be the animal visiting the arms that were initially rewarded.
  3. A Working Memory Error occurs when the animal enters an arm of the maze that it previously entered in the same trial.
  4. A Spatial Memory or Reference Memory Error occurs when the animal enters an arm of the maze that was unbaited during the training phase.
  5. Data from Training and Experimental Phases are analyzed separately, as are data from Working Memory Errors and Reference Memory Errors.
  6. In the standard analysis, the errors made in four sequential trials (spread over two days) are averaged to form a data point. Data from all animals and all data points are then analyzed using an ANOVA.

Applications of the Radial Arm Maze

This test has been applied to study cognitive functions under many different conditions. Drugs that affect the hippocampus and memory can be administered to affect performance (Experimental Group). Similarly, the length of the ITI can be altered to make working memory more or less effective [14]. A version of this win-shift protocol has been used in a rat model of depression, to assess how learned helplessness can adversely impact cognitive functions such as memory and attention [15]. A version of this test has even been adapted for human participants in a virtual-reality environment to assess human spatial learning [16].

The Radial Arm Maze is a traditional test of spatial and working memory, two functions that depend on the complex circuitry of the hippocampus and the interaction of this structure with the rest of the cerebral cortex. The win-shift and win-stay protocols enable researchers to manipulate the complexity of the memory task and apply various experimental manipulations that will affect learning, memory, and attention. It provides them with a means to study and understand memory and learning in rats and mice and newer, virtual version of this protocol have been applied to human subjects as well.

References

  1. DeLuca, S.N., L. Sominsky, and S.J. Spencer, Delayed Spatial Win-Shift Test on Radial Arm Maze. Bio-Protocol, 2016. 6(23): p. e2053.
  2. Schmidt, A.T., et al., Dissociating the long-term effects of fetal/neonatal iron deficiency on three types of learning in the rat. Behavioral Neuroscience, 2007. 121(3): p. 475-82.
  3. Ward, M.T., C.R. Stoelzel, and E.J. Markus, Hippocampal dysfunction during aging II: deficits on the radial-arm maze. Neurobiol Aging, 1999. 20(4): p. 373-80.
  4. Yuede, C.M., H. Dong, and J.G. Csernansky, Anti-dementia drugs and hippocampal-dependent memory in rodents. Behavioural pharmacology, 2007. 18(5-6): p. 347-363.
  5. Broadbent, N.J., L.R. Squire, and R.E. Clark, Spatial memory, recognition memory, and the hippocampus. Proceeding of the National Academy of Sciences, USA, 2004. 101(40): p. 14515-20.
  6. Leszczynski, M., How does hippocampus contribute to working memory processing? Frontiers of Human Neuroscience, 2011. 5: p. 168.
  7. Ramos, J.M., Remote spatial memory and the hippocampus: effect of early and extensive training in the radial maze. Learning and Memory, 2009. 16(9): p. 554-63.
  8. Schultz, C. and M. Engelhardt, Anatomy of the hippocampal formation. Frontiers of neurology and neuroscience, 2014. 34: p. 6-17.
  9. Cappaert, N.L.M., N.M. Van Strien, and M.P. Witter, Chapter 20 – Hippocampal Formation, in The Rat Nervous System (Fourth Edition). 2015, Academic Press: San Diego. p. 511-573.
  10. Maguire, E.A., et al., Navigation-related structural change in the hippocampi of taxi drivers. Proceeding of the National Academy of Sciences, USA, 2000. 97(8): p. 4398-403.
  11. Maguire, E.A., K. Woollett, and H.J. Spiers, London taxi drivers and bus drivers: a structural MRI and neuropsychological analysis. Hippocampus, 2006. 16(12): p. 1091-101.
  12. Burgess, N., E.A. Maguire, and J. O’Keefe, The Human Hippocampus and Spatial and Episodic Memory. Neuron, 2002. 35(4): p. 625-641.
  13. UCLA Behavioral Testng Core. Radial Arms Maze. No Date [cited 2017; Available from: https://btc.psych.ucla.edu/behavioral-testing-core-facilitys-available-tasks/mouse-and-rat-services/radial-arms-maze/.
  14. Stafstrom, C.E., CHAPTER 49 – Behavioral and Cognitive Testing Procedures in Animal Models of Epilepsy, in Models of Seizures and Epilepsy, P.A. Schwartzkroin and S.L. Moshé, Editors. 2006, Academic Press: Burlington. p. 613-628.
  15. Richter, S.H., et al., Where Have I Been? Where Should I Go? Spatial Working Memory on a Radial Arm Maze in a Rat Model of Depression. PLOS ONE, 2013. 8(4): p. e62458.
  16. Shelton, A.L., S.A. Marchette, and A.J. Furman, Chapter Six – A Mechanistic Approach to Individual Differences in Spatial Learning, Memory, and Navigation, in Psychology of Learning and Motivation, B.H. Ross, Editor. 2013, Academic Press. p. 223-259.

About John Agnew

John Agnew is a cognitive neuroscientist living in Colorado. He has a Ph.D. in Neuroscience from Georgetown University and a B.A. in Chemistry and Biochemistry from Haverford College. For the past 15 years, he has taught biology, psychology, and neuroscience courses at the community college and university level and is currently a Contributing Faculty Member in the School of Psychology at Walden University and Adjunct Faculty Member in the Biology Department at Front Range Community College. John is also involved in neuroscience research, attempting to understand some of the behavioral aspects of autism spectrum disorders and evaluating the effectiveness of interventions for individuals with autism. As a freelance editor, John has worked with textbook publishers to update and develop new content for their psychology and neuroscience texts, including James Kalat’s Biological Psychology, Michael Gazzaniga’s Cognitive Neuroscience, Bob Garrett’s Brain & Behavior, and for David Eagleman’s Cognitive Neuroscience texts.