A 5-arm maze enables parallel measures of sustained visuo-spatial attention and spatial working memory in mice

A 5-arm maze has been developed to provide parallel tests of sustained visuo-spatial attention and spatial working memory in mice. C57Bl/6 mice were trained to select, either by immediate response (attention) or by delayed-matching response (working memory), one target arm among the five open arms. For attention testing, mice were first trained to acquire the basic task in which one randomly selected baited arm remained lit until a choice was made. Criterion of >80% correct with a response latency <5 s was attained in 52-56 trials. Following this, attention was tested by using trials wherein light signal durations of 2, 1 or 0.5 s were intermixed to vary attentional load. In the working memory test, mice were submitted to a forced visit to a randomly selected baited arm during a presentation phase. Following a variable retention interval (R.I.), mice were replaced into the maze and rewarded for choosing this arm. Criterion of >80% correct was attained in 35-40 trials and mice exhibited high levels of retention for R.I.s up to 4 h. Results validate the 5-arm maze for evaluation of both sustained visuo-spatial attention and spatial working memory in mice. Both the tasks are rapidly acquired and the 20% chance level provides high resolution for evaluating performance. This comparative strategy allows to dissociate attention and memory and to reveal deficits in these processes during ageing or in knockout strains. The high level of retention performance over R.I.s of 4 h enables studies using pharmacological treatments differentially affecting the acquisition, encoding, retention or retrieval phases of working memory. Furthermore, functional brain imaging studies may be used to identify neuronal networks which are differentially activated during these distinct phases.