Antidepressant mechanism and treatment response define distinct hippocampal-amygdala circuit biomarkers during emotional memory in humans

Abstract

Antidepressant efficacy varies widely, yet the circuit-level mechanisms that distinguish treatment responders from non-responders remain poorly understood in humans. Here, we used high-resolution neuroimaging of hippocampal-amygdala networks during an emotional mnemonic discrimination task that taxes hippocampal pattern separation to examine how antidepressant mechanism of action and perceived treatment response shape emotional memory circuitry (N = 117). Participants included individuals taking single-action antidepressants (selective serotonin reuptake inhibitors), multi-action antidepressants (serotonin-norepinephrine reuptake inhibitors, norepinephrine-dopamine reuptake inhibitors, or polypharmacy), and unmedicated controls matched on current depression severity. Antidepressant mechanism and treatment response were associated with distinct patterns of activity in hippocampal subfields (dentate gyrus (DG)/CA3 and CA1) and amygdala subnuclei, including the basolateral amygdala (BLA) and central amygdala (CEA), during emotional mnemonic discrimination. Among non-responders, the relative balance of hippocampal activity differed by antidepressant mechanism: those taking single-action antidepressants showed greater DG/CA3 than CA1 activity, whereas those taking multi-action antidepressants showed the opposite pattern. This suggests mechanistically specific differences in hippocampal computations associated with ineffective treatment. These effects were localized to the anterior hippocampus, with no significant effects observed in posterior regions. In contrast, responders exhibited stronger DG/CA3-BLA coactivation during negative mnemonic discrimination, a pattern absent in non-responders and unmedicated controls. Antidepressant-associated differences in amygdala subnuclei activity persisted beyond current symptom severity, suggesting medication-related modulation of emotional memory circuits rather than effects driven solely by depression severity. These findings provide evidence in humans that antidepressant use is associated with altered hippocampal-amygdala circuitry in a manner that depends on both pharmacological mechanism and treatment efficacy. Identifying circuit-level signatures of treatment response may inform mechanistically guided approaches to antidepressant selection and monitoring.