The Discrete Paired-trial Variable-delay T-maze Task to Assess Working Memory in Mice

Background

The term working memory refers to the type of memory that is active and relevant only for a short period of time, on the scale of seconds, while performing complex tasks such as reasoning, comprehension and learning. The concept of working memory evolved from that of short-term memory and now it stands at the interface between perceptual processes and long-term memory formation. The major components of working memory, as suggested by Baddeley's model (Baddeley, 2010), are: i) a short-term storage buffer for visual-spatial information that provides a virtual environment for physical simulation, calculation, visualization and optical memory recall (often referred as the visuo-spatial scratch pad); ii) a short-term storage buffer for verbal information (referred to as the phonological loop); (iii) a central executive component that is responsible for response selection and for coordinating the outputs of different short-term memory buffers; iv) an episodic buffer, in which complex multimodal events are integrated and stored online. In this model, the maintenance of specific information is governed by the buffer systems, while the regulation and coordination of this information (i.e., updating and maintenance of task goals, management of interference, and manipulation and transformations of stored content) are handled by the central executive processes. Impairments across the domains of phonological, visuo-spatial and central executive working memory are among the most consistently cognitive deficits observed in patients with schizophrenia (Castner et al., 2004; Forbes et al., 2009). The working memory central executive component has been associated in many studies with the function of the dorsolateral PFC. Instead, the storage buffers are thought to depend more on the inferior frontal cortex and parietal cortical areas (Wager and Smith, 2003). The remarkable correspondence between performances of human patients with frontal lobe lesions, PFC-lesioned monkeys and rodents and schizophrenic patients made the PFC-dependent tasks among the most used in behavioral/fMRI studies in schizophrenia and relative translational preclinical research (Callicott et al., 2000; Kellendonk et al., 2006; Papaleo et al., 2008; Barch et al., 2012).

There are numerous working memory tasks that have been employed and validated in rodents in order to reliably measure the maintenance of visuo-spatial information (Dudchenko, 2004; Kellendonk, et al., 2006; Papaleo et al., 2014), for example, the 8-arm radial maze “delayed non-match to sample” or ‘‘win-shift’’ (Seamans et al., 1995; Seamans and Phillips, 1994), the 8-arm maze “random foraging task” (Floresco et al., 1997; Seamans et al., 1995), the odor span tasks (Dudchenko, 2004; Young et al., 2007) and some paradigms of delayed matching and delay non-matching to sample position operant conditioning tasks (Dunnett, 1993). These tasks involve an initial “sample” or “forced run” phase in which the rodent is exposed to a visual target or an arm of the maze. Subsequently, in the “choice” phase that is run after a variable delay, the subject is simultaneously presented with the original sample (the “match”) and another visual target or arm (the “non-match”). These pairs of phases must be presented repeatedly but importantly, with randomly changing cues presented in the sample phase. Thus, the working memory construct is based on the fact that the tested rodent is required to integrate information held online (the sample phase) with the learned rule (non-match or match to sample). This paradigm has been mostly implemented in mice using T-mazes (Aultman and Moghaddam, 2001; Kellendonk et al., 2006). In this context, the discrete paired-trial variable-delay T-maze task seems to be similar to the human delayed response task and also it relies on mPFC functions (Kellendonk et al., 2006). Indeed, it is based on the delayed non-match to position paradigm where the delayed alternations responses are driven by food reinforcement (Ji et al., 2009; Leggio et al., 2019). In particular, during a ''sample'' or "forced run", the experimental subject is forced to explore an arm of the maze. Then, after a variable delay, in the ''choice run'' phase the subject has to choose between the original sample (the ''match'') and the opposite arm (the ''non-match''). Rodents are then presented with a sequence of randomly chosen forced runs, each followed by a choice run. The working memory construct is based on the fact that the experimental animal is required to integrate information held online (the forced run) with the learned rule (non-match or match to sample). In this context, the role of PFC for supporting complex executive functions is well acknowledged (Scheggia et al., 2018 and 2020), as well as the necessity of the hippocampus involvement for different tasks used to study the spatial working memory, such as the Radial Arm Maze (Myroshnychenko et al., 2017) and the Morris Water Maze (Morris et al., 1982; Dudchenko, 2004).