We have all been in a place whereby the time we were performing the third test of our protocol’s battery of behavioral tests, we realized that our experimental animals have already been stressed, they are no longer naïve and their performance is not consistent.
There has been a rapid evolution of the rodent behavioral studies leading to the implementation of over 100 behavioral tests used now by scientists all over the world.[1] These tests have been developed to study a wide range of behavioral phenotypes, from learning capabilities and memory capacities to stress coping and aggression. They have also been implemented as a tool to study the behavioral phenotypes of transgenic mouse and rat lines modeling human diseases.
This article firstly states the need for implementing batteries of behavioral tests in neurobehavioral studies. Then, addresses any serious limitations concerning the design of these sequences of behavioral tasks. Finally, it provides information on the appropriate order in which the tests should be performed.
The need for battery of behavioral tests
In an effort to determine and validate behavioral phenotypes and baseline behaviors for various genetic lines of rodents, researchers often choose to perform a battery of behavioral tests. There are many reasons behind this relatively common practice. First of all, the behavioral phenotypes studied and the human diseases often modeled are so complex and multivariate, that a single task or behavioral measure cannot provide all the necessary information on the behavioral repertoire of each animal. Additionally, ethical and welfare reasons or even practical reasons (such as the limited number of available subjects), often leads researchers into exposing the same group of subjects to several different behavioral tests in the course of an experiment.
We should always keep in mind that nearly all the behaviors studied in neuroscience involve brain functions that implicate many different neural systems. An important example is the function of learning and memory, which is supported by at least five different neural systems; hippocampus, amygdala, dorsal striatum, rhinal cortex and cerebellum, each one of them contributing to this crucial brain function.[2] It goes without saying that there is an increasing need for designing sets of paradigms which can provide us with a more comprehensive view of all these multifactorial brain processes.
The use of multitask testing has been gaining ground in behavioral studies, especially in studies modeling human diseases, as mentioned above. By applying a battery of behavioral tests in transgenic animals, researchers are able to examine multiple behavioral components and estimate any relationship between the different behavioral parameters and phenotypes. A variety of batteries of behavioral tests have been used to assess cognitive impairment in transgenic mice modeling Alzheimer’s disease (129S6/Tg2576),[3] motor and non-motor deficits in mouse models of Parkinson’s disease[4] and cognitive rigidity in mouse models of autism.[5] These studies can help us identify any behavioral impairment in transgenic lines modeling human diseases which can, in turn, offer us a valuable insight into the effectiveness of any therapeutic actions taken.
Limitations in the use of the battery of tests
Despite the increasing use of multitasking testing, a series of questions have been raised as to the test order interactions and the performance of the test subjects.
One of the most critical factors to consider when designing and choosing the test battery order is the stress impact that each task has on an animal. Depending on the nature and difficulty of the test, it can be a source of great stress for the subject. Since stress is known to influence the processes of learning, memory, and cognition in general, it is a factor that could introduce a considerable amount of variance in your data and further complicate your analyses. A subject recovering from the stress exposure of the open field test will perform poorly or erratically to a following marbles test. Stressful tests could also inhibit the training trials for other tests or furthermore training for one test could impair the training for all the subsequent.
Another factor which should be taken into consideration is the naivety of your subjects and how this is compromised when each subject is exposed to several different tests. Only a limited amount of studies have investigated the effect of testing on the subjects’ performance to future tests. Macllwain et al[6] showed that mice with battery test experience exhibited different behavior than naïve mice only during specific behavioral tasks. In addition, when testing the effect of the battery order on an animal’s performance, they observed that some but not all the test variables are sensitive to test order. Their conclusion was that only some tests are prone to test experience and those are the open field, rotarod and hot plate. Similar experiments on rats showed that the forced swim test is significantly vulnerable to the test order, in contrast to zero maze which is resistant.
Another study applied an extensive battery of behavioral tests, covering a wide range of the behavioral repertoire in two very popular mice strains, C57BL/6J and 129S2/Sv.[7] When comparing the results obtained from the mice exposed to the entire battery of behavioral tests to naive mice in every paradigm, they observed that the training history and experience significantly affected behavior in a strain-dependent manner. However, performance in each test was altered in a diverse manner. More specifically, non-naïve mice involved in this battery of tests experienced reduced exploratory behavior and emotionality and increased nociceptive sensitivity and coordination ability. In addition, non-naïve mice showed reduced activity in the forced swimming test, as well as reduced conditioned taste aversion.
Contrary to the limited research on the effect of the testing history on performance, there is a significantly larger number of studies focusing on the effect of test-retest on the reliability of anxiety tests. Exposing 5 strains of mouse to repeated testing on three consecutive days in an elevated zero-maze, showed that anxiety increased over the three trials.[8] An interesting finding of this study was that only a repeated-trial zero maze managed to detect anxiety in strains with visual impairment. Since some of the strains use visual or sensory cues to avoid the open arms of the zero-maze; repeated testing in this apparatus is needed in strains with visual impairment.
Recent studies have focused on another issue concerning the batteries of behavioral tests and this is the intra- and inter-specific correlations between the various tasks.[9] Using the multivariate approach, scientists can analyze the vast volume of behavioral data and draw conclusions regarding any effects between the different behavioral tests. It has been shown that correlations exist between the performance in various independent tests, therefore, the behavioral pattern in one behavioral measure can be predictive of the pattern observed in the following ones.
In what order should I perform my behavioral tests?
The golden rule when it comes to the battery test order is to start with the least stressful test and leave the most stressful of all for last. Tests for anxiety-like and exploratory behavior should always be performed first since they can be significantly affected by any previous experience. On the other hand, any task focusing on the cognitive abilities of the animal should be carried out at later stages. This is crucial since the animal’s cognitive performance can be seriously affected if the animal is not properly adapted to the handling procedure and the experimenter. Most of the published behavioral studies abide by this rule of using the least invasive tests before anything else and leaving the more invasive ones for last.[3][6][10][11] Another very important consideration is to allow multiple days of resting time between tests to decrease carryover effects from prior tests.
A very common order of behavioral tests are Y-maze forced alternation, novel object recognition, Morris water maze, radial arm water maze and Y-maze spontaneous alternation.[3] A more extensive example would be the: open field, novel exploration, elevated plus maze, light-dark box, primary SHIRPA, puzzle box, Morris water maze and tail suspension.[11] Finally, an example including some of the most commonly used behavioral tests is the: open field test, elevated plus maze, light-dark exploration, Y-maze, beam walking, coat hanger, rota-rod, hot plate, fear conditioning, water maze, forced swim test and conditioned taste aversion.[7]
An important suggestion is that the tests should be categorized on the basis of the behavioral phenotypes that you want to study. Ideally, if that is possible by the available number of subjects, you should use a different set of animals and separately investigate learning and memory from other functions such as aggression or anxiety. However, if this is not possible, at the end you could repeat some of the tests to see the evolution of the behavior and the impact of the previous experience on the performance for some of the tests.
In the end, the order to which your behavioral tests should be conducted is based on your experimental design and purpose and most importantly on the behavior and characteristic that you want to study.
References
- A. Hånell, N. Marklund, Structured evaluation of rodent behavioral tests used in drug discovery research, Front. Behav. Neurosci. 8 (2014) 1–13. doi:10.3389/fnbeh.2014.00252.
- R.E. Brown, L. Stanford, H.M. Schellinck, Developing standardized behavioral tests for knockout and mutant mice, ILAR J. 41 (2000) 163–174. doi:10.1093/ilar.41.3.163.
- A. Wolf, B. Bauer, E.L. Abner, T. Ashkenazy-frolinger, A.M.S. Hartz, A Comprehensive Behavioral Test Battery to Assess Learning and Memory in 129S6 / Tg2576 Mice, PLoS One. (2016) 1–23. doi:10.1371/journal.pone.0147733.
- T.N. Taylor, J.G. Greene, G.W. Miller, Behavioral phenotyping of mouse models of Parkinson´s disease, Behav. Neurosci. 211 (2011) 1–10. doi:10.1016/j.bbr.2010.03.004.Behavioral.
- Puscian, S. Leski, T. Gorkiewicz, K. Meyza, H.-P. Lipp, E. Knapska, A novel automated behavioral test battery assessing cognitive rigidity in two genetic mouse models of autism, Front. Behav. Neurosci. 8 (2014) 1–11. doi:10.3389/fnbeh.2014.00140.
- K.L. Mcilwain, M.Y. Merriweather, L.A. Yuva-paylor, R. Paylor, The use of behavioral test batteries : Effects of training history, Physiol. Behav. 73 (2001) 705–717.
- V. Võikar, E. Vasar, H. Rauvala, Behavioral alterations induced by repeated testing in C57BL/6J and 129S2/ Sv mice: Implications for phenotyping screens, Genes, Brain Behav. 3 (2004) 27–38. doi:10.1046/j.1601-183X.2003.0044.x.
- M.N. Cook, M. Crounse, L. Flaherty, Anxiety in the elevated zero-maze is augmented in mice after repeated daily exposure, Behav. Genet. 32 (2002) 113–118. doi:10.1023/A:1015249706579.
- D.D. Feyissa, Y.D. Aher, E. Engidawork, H. Höger, G. Lubec, V. Korz, Individual Differences in Male Rats in a Behavioral Test Battery: A Multivariate Statistical Approach, Front. Behav. Neurosci. 11 (2017) 1–8. doi:10.3389/fnbeh.2017.00026.
- M. Peng, C. Zhang, Y. Dong, Y. Zhang, H. Nakazawa, Battery of behavioral tests in mice to study postoperative delirium, Nat. Publ. Gr. (2016) 1–13. doi:10.1038/srep29874.
- H.V. Lad, L. Liu, J.L. Paya-cano, M.J. Parsons, R. Kember, C. Fernandes, et al., Behavioural battery testing: Evaluation and behavioural outcomes in 8 inbred mouse strains, Physiol. Behav. 99 (2010) 301–316. doi:10.1016/j.physbeh.2009.11.007.