The role of working memory and declarative memory in trace conditioning

Abstract

Translational assays of cognition that are similarly implemented in both lower and higher-order species, such as rodents and primates, provide a means to reconcile preclinical modeling of psychiatric neuropathology and clinical research. To this end, Pavlovian conditioning has provided a useful tool for investigating cognitive processes in both lab animal models and humans. This review focuses on trace conditioning, a form of Pavlovian conditioning typified by the insertion of a temporal gap (i.e., trace interval) between presentations of a conditioned stimulus (CS) and an unconditioned stimulus (US). This review aims to discuss pre-clinical and clinical work investigating the mnemonic processes recruited for trace conditioning. Much work suggests that trace conditioning involves unique neurocognitive mechanisms to facilitate formation of trace memories in contrast to standard Pavlovian conditioning. For example, the hippocampus and prefrontal cortex (PFC) appear to play critical roles in trace conditioning. Moreover, cognitive mechanistic accounts in human studies suggest that working memory and declarative memory processes are engaged to facilitate formation of trace memories. The aim of this review is to integrate cognitive and neurobiological accounts of trace conditioning from preclinical and clinical studies to examine involvement of working and declarative memory.

Keywords: Declarative memory; Hippocampus; Medial prefrontal cortex; Pavlovian; Trace conditioning; Working memory.

Fig. 1

Temporal arrangement of stimuli in delay and trace conditioning paradigms (a) delay conditioning occurs when there is a delay between CS onset and US onset, but the CS and US co-terminate and (b) trace conditioning occurs when the CS and US are not contiguous. The CS is followed by a trace interval prior to presentation of the US.

Fig. 2

Hypothetical schematic of brain regions recruited during eyeblink and fear conditioning paradigms. Black arrows indicate projections involved when stimuli are contiguous (delay conditioning). Regions in blue depict core circuitry required for trace and delay conditioning. During eyeblink conditioning auditory sensory information is passed via cochlear nucleus to thalamus (Th), pontine nuclei (Pn), and then Cerebellum (Cer) (Halverson et al., 2008). Critical plasticity occurs within interpositus nucleus of the Cer, mediating CR’s (Steinmetz et al., 1986). During fear conditioning CS and US is sent from thalamus to the amygdala where fear memories are encoded (Kim, Rison, & Fanselow, 1993; LeDoux et al., 1990). The amygdala (Amy) is also critical for mediating CR’s (Phillips & LeDoux, 1992) (additional supporting literature is found in Section 1.1). Red arrows indicate theoretical information flow between regions recruited during trace conditioning. Orange regions indicate brain areas engaged during trace conditioning. During trace conditioning information may pass from the hippocampus via monosynaptic projections to the mPFC. Importantly, the mPFC projects to areas critical for learning-related plasticity, including the Amy (fear conditioning) and Pn, which provides direct input to the Cer (eyeblink conditioning). Thus, while core circuitry differs between fear and eyeblink conditioning, trace conditioning may recruit overlapping neural substrates to support learning. (A) Trace Fear Conditioning: The CS is processed and relayed to the mPFC via thalamic inputs. CS representation may be maintained within the mPFC via sustained neural activation. The dorsal hippocampus receives sensory information from the entorhinal cortex (EC) and processes CS-US temporal information. Direct connections between ventral hippocampus (VH) and amygdala may play an important role in trace fear expression. mPFC and VH signaling converge on amygdala (Amy). (B) Trace Eyeblink Conditioning: CS information is relayed via the Th to the mPFC. The hippocampus may encode and send temporal CS-US information to the mPFC and maintains sustained activation during trace interval. Finally, mPFC may relay CS information to Cer facilitating CS-US plasticity within the Cer (supporting literature is found in Sections 2 and 3).

Fig. 3

Schematic of neural activity patterns observed during trace interval, PFC (Blue) and hippocampus (Orange). Pre-clinical and clinical data suggest differing roles for the PFC and hippocampus in trace conditioning. The PFC shows working memory-like sustained activity across the trace interval. In contrast, the hippocampus shows learning-related changes in activation to the CS and US, suggesting that the hippocampus processes CS-US temporal information (Haritha et al., 2012; Gilmartin & McEchron, 2005a; Knight et al., 2004; Gilmartin & McEchron, 2005b).