D'hers, S.; Ojea Ramos, S.; Robles, A.; Feld, M.

Temporal Order Memory (TOM), the ability to discriminate between events according to when they occurred, is a key component of episodic-like memory. Understanding the molecular mechanisms that support temporal memory requires behavioral approaches capable of capturing the continuous dynamics of natural exploration. Despite extensive evidence implicating the prefrontal cortex (PFC) in temporal memory, the intracellular signaling mechanisms supporting temporal order discrimination remain poorly understood. Here, we combined high-resolution automated behavioral phenotyping with molecular analyses to characterize the behavioral and signaling dynamics underlying TOM in mice. Mice were trained in a spontaneous object-recognition TOM task and tested after short-term (3 h) or long-term (24 h) retention intervals. Exploration was quantified using an artificial intelligence-based behavioral analysis pipeline that enables continuous and unbiased assessment of object exploration. Phosphorylation of extracellular signal-regulated kinase 2 (ERK2) and expression of the ERK phosphatase MKP3/DUSP6 were analyzed in the PFC and hippocampus (HIP) following habituation, a single training session, or two sequential training sessions. Additionally, a Temporal Novel Object Recognition (TeNOR) protocol was used to evaluate the integrity of memory traces. Mice displayed robust TOM performance across sexes and retention intervals. Molecular analyses revealed no significant changes in hippocampal ERK signaling, whereas the cytosolic fraction of the PFC exhibited dynamic, experience-dependent modulation of ERK2 phosphorylation. A single 15-minute training session induced a transient increase in ERK2 activation, while a second session 45 minutes later actively suppressed this peak. This rapid molecular reset was accompanied by increased MKP3 expression, suggesting the targeted recruitment of an active regulatory feedback mechanism. Continuous behavioral tracking further revealed temporal features of memory expression that were not captured by conventional summary measures; it identified an early, rapid decay of discrimination for older object memories in the TeNOR task, suggesting that TOM performance relies on resolving competitive retrieval between co-existing memory traces. Together, these findings identify dynamic ERK2-MKP3 signaling in the PFC as the molecular substrate upon which temporal discrimination can take place, and demonstrate how high-resolution phenotyping in naturalistic behavioral paradigms can reveal mechanistic links between intracellular signaling and the temporal organization of experience.