Transient inactivation of the thalamic nucleus reuniens and rhomboid nucleus produces deficits of a working-memory dependent tactile-visual conditional discrimination task

Abstract

Working memory depends on communication between the hippocampus and the prefrontal cortex (PFC); however, the neural circuitry that mediates interactions between these brain areas has not been well characterized. Two candidate structures are the thalamic reuniens (RE) and rhomboid (Rh) nuclei, which are reciprocally connected with both the hippocampus and PFC. These known anatomical connections suggest that RE/Rh may be involved in mediating hippocampal-prefrontal communication, and therefore may be critical for working memory processing. To test the hypothesis that RE/Rh are necessary for working memory, we trained separate groups of rats to perform 1 of 2 tasks in a T-maze. The first task was a working memory-dependent conditional discrimination (CDWM) task, and the second task was a nonworking memory-dependent conditional discrimination (CD) task. These tasks took place in the same maze, featured the same number of trials, and utilized the same cue (a tactile-visual maze insert). After rats had learned either task, RE/Rh were transiently inactivated with the GABAA receptor agonist muscimol, and performance was assessed. RE/Rh inactivation caused performance deficits on the CDWM task, but not the CD task. This result suggests that RE/Rh are a necessary component of working memory task performance, which is also thought to depend on the hippocampal-prefrontal circuit. RE/Rh inactivation did not cause a performance deficit on the CD task, suggesting that RE/Rh have dissociable contributions to working memory-dependent and nonworking memory-dependent tasks, independently of the known contributions of these 2 thalamic nuclei to the sensorimotor and attention-related aspects of other memory tasks.

Figure 1

A) Representative photographs detailing cannula placements. The top panel shows a coronal slice stained with cresyl violet, with a red box surrounding the location of the cannula tip The middle panel shows a coronal slice from the same ratvisualized with a confocal microscope with a digital plate from the Paxinos and Watson (2005) rat brain atlas overlaying it. This plate details the boundaries of the thalamic nuclei in relation to the observed spread of the fluorescent muscimol, which remained largely localized to the ventral reunions nucleus (vRE) in this slice. The bottom panel shows the same coronal slice as visualized with a fluorescent microscope. B) Coronal plates showing the placements of the injector cannulae in rats that were trained on the CDWM task (black dots) and rats that were trained on the CD task (stars). All plates are re-printed with permission from The Rat Brain in Stereotaxic Coordinates: 5^th Edition, pages 87-100, by G. Paxinos & C Watson, 2005, Burlington, MA: Elsevier Academic Press. Copyright 2005 by G. Paxinos & C. Watson

Figure 2

A) Schematic of the CDWM (top) and CD (bottom) tasks. During each task, rats were required to run up the central stem, choose a goal arm, consume the chocolate sprinkle reward (black dots on figures), and return to the start box via the return arms where they were confined until the next trial. Rats were required to choose the goal arm that contained reward based on the texture and appearance of a floor insert (e.g., left on mesh, right on wood). For the CD task (bottom), the floor insert covers the stem and goal arms. For the CDWM task (top), the floor inserts only cover the first half of the maze stem. B) Number of sessions needed to reach learning criterion (>80% correct for two consecutive sessions) between the CD (red) and CDWM (blue) tasks. C) Choice accuracy for the group that learned the CD task (red) and the group that learned the CDWM task (blue) in response to microinfusions of PBS (control) and 3 concentrations of muscimol. For all 3 muscimol concentrations, RE/Rh inactivation impaired choice accuracy on the CDWM task, but not the CD task. Error bars represent standard error of the mean (SEM), *p < 0.01