Anterior thalamic circuits crucial for working memory

Abstract

Alterations in the structure and functional connectivity of anterior thalamic nuclei (ATN) have been linked to reduced cognition during aging. However, ATN circuits that contribute to higher cognitive functions remain understudied. We found that the anteroventral (AV) subdivision of ATN is necessary specifically during the maintenance phase of a spatial working memory task. This function engages the AV→parasubiculum (PaS)→entorhinal cortex (EC) circuit. Aged mice showed a deficit in spatial working memory, which was associated with a decrease in the excitability of AV neurons. Activation of AV neurons or the AV→PaS circuit in aged mice was sufficient to rescue their working memory performance. Furthermore, rescued aged mice showed improved behavior-induced neuronal activity in prefrontal cortex (PFC), a critical site for working memory processes. Although the direct activation of PFC neurons in aged mice also rescued their working memory performance, we found that these animals exhibited increased levels of anxiety, which was not the case for AV→PaS circuit manipulations in aged mice. These results suggest that targeting AV thalamus in aging may not only be beneficial for cognitive functions but that this approach may have fewer unintended effects compared to direct PFC manipulations.

Keywords: aging; anterior thalamic nuclei; cognitive functions; prefrontal cortex; working memory.

Fig. 1.

AV thalamus is crucial for the maintenance phase of working memory. (A) Strategy to label AV thalamic neurons by injecting a CreOff-NpHR-eYFP virus in the ATN of C1QL2-Cre mice (Left), immunostaining indicates AV labeling with high specificity (C1QL2 is a marker of AD thalamic neurons, Middle), ex vivo electrophysiology showing light-induced inhibition of AV neuronal activity (Right). (B) Inhibition of AV neurons during each of the three different phases in the T-maze paradigm. (C) Inhibition of AV during the open field behavioral paradigm. (D) Days to criterion before any manipulation. (E) Working memory performance for the various phase-specific AV inhibition tests. (F) Labeling AD neurons by injecting a DIO-NpHR-eYFP virus in the ATN of C1QL2-Cre mice. (G) Days to criterion before any manipulation. (H) Working memory performance for the various phase-specific AD inhibition tests. (I) We injected a mixture of the three Cal-Light viruses in the AV region for activity-dependent labeling. (J) Cal-Light–labeled active neuronal ensembles in AV from home cage, sample phase, delay phase, and choice phase groups (n = 12 mice per group for all behavior panels, n = 4 to 5 mice per group for Cal-Light labeling). Dashed line indicates chance level of performance (i.e., 50% correct choices) in T-maze tests. Data are presented as mean ± SEM; *P < 0.05, ***P < 0.001. Two-tailed unpaired t test (C–E, G, and H) and one-way ANOVA followed by Bonferroni post hoc test (J).

Fig. 2.

Identification of AV circuitry recruited during the maintenance phase of working memory. (A) Terminal labeling of AV neurons in ACC, RSC, PrS, and PaS (Top). S, subiculum. Working memory performance for delay-specific inhibition tests in the different AV terminal groups (Bottom). (B) By injecting an anterograde Syn-Cre virus in ATN combined with a DIO-NpHR-eYFP virus in PaS, we visualized the ATN→PaS→EC circuit. EC is shown in a coronal section. (C) ATN→PaS→EC terminal inhibition during each of the three different phases of the T-maze task (n = 12 mice per group for all behavior panels). Dashed line indicates chance level of performance (i.e., 50% correct choices) in T-maze tests. Data are presented as mean ± SEM; **P < 0.01. Two-tailed unpaired t test (A and C).

Fig. 3.

Activating AV thalamus is sufficient to improve working memory performance. (A) Labeling AV neurons by injecting a CreOff-ChR2-mCh virus in the ATN of C1QL2-Cre mice (Left), ex vivo electrophysiology showing light-induced activation of AV neurons (Right). (B) Activation of the AV→PaS circuit during the sample or delay phases in the T-maze paradigm. (C) Working memory performance in 10-, 60-, and 90-s delay tests with sample phase-specific AV→PaS circuit activation. (D) Working memory performance in 10-, 60-, and 90-s delay tests with delay phase-specific AV→PaS circuit activation (n = 12 mice per group for all behavior panels). Dashed line indicates chance level of performance (i.e., 50% correct choices) in T-maze tests. Data are presented as mean ± SEM; *P < 0.05, **P < 0.01. Two-tailed unpaired t test (C and D).

Fig. 4.

Targeting AV thalamus rescues working memory deficits in aged mice. (A) Working memory performance in young (2 mo) and aged (14 mo) C1QL2-Cre mice. (B) Ex vivo electrophysiology showing decreased excitability of AV neurons in aged mice (15 neurons from n = 3 young mice, 16 neurons from n = 3 aged mice). (C and D) Activation of AV neurons (C) or the AV→PaS circuit (D) during the delay period rescues working memory in aged (14 mo old) mice. (E) Behavior-induced neuronal activity of prefrontal cortex (PFC) neurons was examined using cFos staining (n = 5 mice per group). (F) Labeling PFC neurons in wild-type mice with a ChR2-mCh virus. DAPI staining (blue). (G) Activation of PFC neurons during the delay period rescues working memory in aged mice. (H) Zero maze anxiety test using AV→PaS circuit and PFC neuron activation groups (n = 12 mice per group for all behavior panels). Dashed line indicates chance level of performance (i.e., 50% correct choices) in T-maze tests. Data are presented as mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001. Two-tailed unpaired t test (A, C, D, G, and H), two-way ANOVA with repeated measures followed by Bonferroni post hoc test (B), and one-way ANOVA followed by Bonferroni post hoc test (E).