Extensive evidence indicates that stimulus-response (S-R) habits are critically mediated by the dorsolateral striatum of the mammalian brain. Temporary inactivation or permanent lesions of the DLS has been shown to impair acquisition, consolidation, retrieval, and extinction of habit memory in a variety of learning and memory tasks in rodents (for reviews, see Yin and Knowlton, 2006, Packard, 2009, White et al., 2013, Goodman and Packard, 2018, Goodman and Packard, 2019). In human subjects, neural activation of the homologous caudate-putamen has been similarly associated with habitual responding in instrumental learning tasks and S-R navigation in virtual mazes (Iaria et al., 2003, Tricomi et al., 2009, Patterson and Knowlton, 2018). While a role for the DLS in habit memory processes has been well established, a complete understanding of the neurotransmitter systems involved in S-R habit memory requires further investigation. The present study aimed to specifically examine whether N-methyl-D-aspartate (NMDA) receptors in the DLS are involved in the consolidation and retrieval of habit memory in a response learning version of the plus-maze.

Glutamate has been recognized as the chief excitatory neurotransmitter in the mammalian brain and plays a critical role in memory function. Glutamatergic NMDA receptors, in particular, are important for many different kinds of learning and memory, including spatial memory, fear conditioning, and conditioned place preference, as well as learning in other memory tasks (e.g., Morris et al., 1986, Miserendino et al., 1990, Fanselow and Kim, 1994). The role of NMDA receptors in memory function is often attributed to the fact that activation of these receptors is required for synaptic plasticity (Morris, 2013). Importantly, NMDA receptors mediate different forms of synaptic plasticity in the DLS (Picconi et al., 2003, Li et al., 2009).

Extensive evidence has indicated a critical role for glutamatergic mechanisms in the acquisition and consolidation of DLS-dependent habit memory. Systemic administration of NMDA receptor antagonist MK801 impairs acquisition and consolidation in a DLS-dependent cued water maze (Packard and Teather, 1997a, Farina and Commins, 2020; but see also, Mackes & Willner, 2006). Impairment in the cued water maze may also be observed following intra-DLS administration of NMDA receptor antagonist AP5 (Packard & Teather, 1997b). In contrast, glutamate infused directly into the DLS enhances memory consolidation in the DLS-dependent cued water maze (Packard & Teather, 1999) and biases animals toward the use of a DLS habit memory strategy in a task that can be solved adequately with an alternative strategy (Packard, 1999). A similar critical role for glutamatergic NMDA receptors has also been demonstrated for habitual responding in instrumental learning tasks (e.g., Furlong et al., 2018, Ersche et al., 2021).

The present study examined whether DLS NMDA receptor activity is required for memory consolidation and retrieval of habit memory in a DLS-dependent response learning task (for review of response learning, see Goodman, 2021). In response learning, a rat is reinforced to make a consistent body-turn response at the intersection of a plus-maze or T-maze apparatus. Learning in this task has been considered an exemplar of habit memory, given the critical role of the DLS (Packard and McGaugh, 1996, Palencia and Ragozzino, 2005) and the fact that the acquired turning response is insensitive to reinforcer devaluation and changes in the starting position (e.g., Packard and McGaugh, 1996, Kosaki et al., 2018). Previous research has indicated that pre-training intra-DLS administration of NMDA receptor antagonist AP5 impairs acquisition of response learning in the plus-maze (Palencia & Ragozzino, 2005). However, it is unknown whether DLS NMDA receptors are similarly required for memory consolidation or retrieval of habit memory in the response learning task.

The present study consisted of two experiments. In the first experiment, rats received post-training intra-DLS administration of the NMDA receptor antagonist AP5 immediately after each daily training session. Post-training drug administrations were used in order to target the consolidation phase of memory processing, when the labile response learning memory is being “consolidated” into a long-term memory trace (for review see McGaugh, 2000). In experiment 2, different groups of rats received intra-DLS administration of AP5 prior to the last training session in order to investigate whether NMDA receptor activity is required for retrieval of habit memory. Based on previous research (Packard & Teather, 1997a), it was hypothesized that intra-DLS AP5 would impair memory consolidation but not retrieval of habit memory, consistent with the canonical role of NMDA receptors in promoting synaptic plasticity and formation of new memories, as opposed to memory retrieval per se (Kandel, 2001).