Midbrain–Hippocampus Structural Connectivity Selectively Predicts Motivated Memory Encoding

Abstract

Motivation is a powerful driver of learning and memory. Functional MRI studies show that interactions among the dopaminergic midbrain substantia nigra/ventral tegmental area (SN/VTA), hippocampus, and nucleus accumbens (NAc) are critical for motivated memory encoding. However, it is not known whether these effects are transient and purely functional, or whether individual differences in the structure of this circuit underlie motivated memory encoding. To quantify individual differences in structure, diffusion-weighted MRI and probabilistic tractography were used to quantify SN/VTA–striatum and SN/VTA–hippocampus pathways associated with motivated memory encoding in humans. Male and female participants completed a motivated source memory paradigm. During encoding, words were randomly assigned to one of three conditions, reward ($1.00), control ($0.00), or punishment (−$1.00). During retrieval, participants were asked to retrieve item and source information of the previously studied words and were rewarded or penalized according to their performance. Source memory for words assigned to both reward and punishment conditions was greater than those for control words, but there were no differences in item memory based on value. Anatomically, probabilistic tractography results revealed a heterogeneous, topological arrangement of the SN/VTA. Tract density measures of SN/VTA–hippocampus pathways were positively correlated with individual differences in reward-and-punishment-modulated memory performance, whereas density of SN/VTA–striatum pathways showed no association. This novel finding suggests that pathways emerging from the human SV/VTA are anatomically separable and functionally heterogeneous. Individual differences in structural connectivity of the dopaminergic hippocampus–VTA loop are selectively associated with motivated memory encoding.SIGNIFICANCE STATEMENTFunctional MRI studies show that interactions among the SN/VTA, hippocampus, and NAc are critical for motivated memory encoding. This has led to competing theories that posit either SN/VTA–NAc reward prediction errors or SN/VTA–hippocampus signals underlie motivated memory encoding. Additionally, it is not known whether these effects are transient and purely functional or whether individual differences in the structure of these circuits underlie motivated memory encoding. Using diffusion-weighted MRI and probabilistic tractography, we show that tract density measures of SN/VTA–hippocampus pathways are positively correlated with motivated memory performance, whereas density of SN/VTA–striatum pathways show no association. This finding suggests that anatomic individual differences of the dopaminergic hippocampus–VTA loop are selectively associated with motivated memory encoding.

Intrinsic functional connections determine how curiosity and prediction errors enhance memory

“…First, VTA-hippocampal as compared to VTA-NAcc connections are more sparse and have a more precise innervation topography (Gasbarri et al, 1994;Zubair et al, 2021), making it more difficult to measure VTA-hippocampal functional connectivity with a seed-based resting-state fMRI analysis. Second, the present finding is in line with recent research suggesting that memory enhancements based on intrinsic curiosity and extrinsic reward share some commonalities but might rely on different functional and structural brain networks (Duan et al, 2020;Elliott et al, 2022;Meliss & Murayama, 2022). For instance and potentially consistent with our findings, Meliss & Murayama (2022) found that changes in VTAhippocampal functional connectivity from pre-to post-learning rest periods were associated with externally rewarded but not curiosity-driven memory enhancements.…”

“…Microstructure indices have provided evidence for structural connectivity of the VTA and NAcc also in humans (see Jbabdi & Behrens, 2013 for review). Furthermore, the strength of both VTA-NAcc and VTAhippocampal structural connectivity (Elliott et al, 2022;Reggente et al, 2018) and variability in cortico-mesolimbic intrinsic functional connectivity (Frank et al, 2019) have been related to individual differences in reward-related memory enhancements. The convergence of structural connectivity, intrinsic functional connectivity, and task-related measures is consistent with the idea that intrinsic functional networks are primarily determined by structural connections, providing a functionally stable basis for the support of task-related cognitive processes (Cole et al, 2014;Gratton et al, 2018).…”

“…The memory-enhancing and neuromodulatory effects of curiosity are highly similar to the effects of extrinsic rewards on memory. Research on reward further suggests that individual differences in reward-related memory enhancements are determined by VTA-NAcc and VTA-hippocampal structural connectivity (Elliott et al, 2022;Reggente et al, 2018) as well as functional connectivity within the dopaminergic corticomesolimbic system irrespective of whether it is measured during a task or at rest (Frank et al, 2019). Based on these findings and following the rationale that intrinsic functional connections measured at rest support cognitive performance due to overlap with task-related coactivation patterns (Cole et al, 2014;Gratton et al, 2018), individual differences in the strength of intrinsic functional connectivity within the mesolimbic system might predict the magnitude of curiosity-driven memory enhancements.…”

In Vivo Structural Connectivity of the Reward System Along the Hippocampal Long-Axis

“…Prior work using human neuroimaging demonstrated VTA-HPC connectivity to be critical for reward-motivated memory encoding (Adcock et al, 2006;Wittmann et al, 2005;Shigemune et al, 2014;Wolosin et al, 2012;Gruber et al, 2016), with studies showing a bias towards engagement of anterior hippocampus (Poppenk et al, 2013;Cowan et al, 2021). Further, previous research has found structural connectivity between the VTA and HPC to be positively correlated with individual differences in reward-motivated memory performance (Elliott et al, 2022b). Our findings integrate these two lines of research by providing a putative mechanism for functional biases towards the anterior hippocampus during motivated memory based on its structural connectivity, which supports future studies using multi-modal imaging approaches to understand function-structure relationships.…”

Anin vivoDissection, and Analysis of Socio-Affective Symptoms related to Cerebellum-Midbrain Reward Circuitry in Humans

“…Things that are rewarding evoke dopaminergic signals that encode information about the discrepancy between reward expectancies and reward outcomes, to facilitate goal-directed behavior (Schultz, 2016). Previous research using probabilistic tractography in humans has implicated VTA-striatum and VTA-hippocampus connectivity in reward-motivated behaviors and psychopathology (Elliott, D’Ardenne, Mukherjee, Schweitzer, & McClure, 2022; Elliott, D’Ardenne, Murty, Brewer, & McClure, 2022; MacNiven, Leong, & Knutson, 2020). We extend these findings by suggesting that medial portions of the cerebellum may help regulate these circuits.…”