The hippocampus contributes to retroactive stimulus associations during trace fear conditioning

Abstract

Binding events that occur at different times are essential for memory formation. In trace fear conditioning, animals associate a tone and footshock despite no temporal overlap. The hippocampus is thought to mediate this learning by maintaining a memory of the tone until shock occurrence, however, evidence for sustained hippocampal tone representations is lacking. Here, we demonstrate a retrospective role for the hippocampus in trace fear conditioning. Bulk calcium imaging revealed sustained increases in CA1 activity after footshock that were not observed after tone termination. Optogenetic silencing of CA1 immediately after footshock impaired subsequent memory. Additionally, footshock increased the number of sharp-wave ripples compared to baseline during conditioning. Therefore, post-shock hippocampal activity likely supports learning by reactivating and linking latent tone and shock representations. These findings highlight an underappreciated function of post-trial hippocampal activity in enabling retroactive temporal associations during new learning, as opposed to persistent maintenance of stimulus representations.

Keywords: Behavioral neuroscience; Cellular neuroscience; Neuroscience.

Graphical abstract

Figure 1

Footshock elicits increases in CA1 activity during TFC training (A) Left: GCaMP6f was infused into CA1 (n = 11 mice), and an optical fiber (white dotted line) was implanted above the infusion site. Right: Schematic of the fiber photometry system used to measure bulk fluorescence during TFC training. (B) Bulk calcium response during TFC training trials show a large increase in activity elicited by the footshock. Gray rectangle indicates when tone is presented. Dotted rectangle indicates footshock presentation. (C) GCaMP fluorescence is significantly increased by footshock (paired t-test, t(10) = −6.256, p < 0.001). Light gray lines represent each animal’s mean fluorescence for the 20 s before the shock (pre-shock) and the 20 s after the shock (post-shock). Dark line represents the mean pre-shock and post-shock fluorescence averaged over all subjects. All data are expressed as mean ± SEM; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 2

CA1 inhibition after the footshock impairs TFC memory (A) Representative image of AAV expression and optical fiber placement targeting CA1. (B) Experimental design to silence CA1 after the footshock during TFC. (C) On the first day mice (n = 12 per group) underwent TFC while laser stimulation (561 nm) was delivered to CA1 continuously for 40 s immediately after the footshock on each training trial. The next day mice received a tone test in a novel context B. The following day contextual fear memory was tested in the original training context. (D) ArchT mice froze significantly less during the trace interval and ITI than the eGFP control group during the training session (Group × Phase interaction: F(3,66) = 4.842, p < 0.01; Trace: t(22) = −2.801, p < 0.05; ITI: t(22) = −3.887, p < 0.01). (E) ArchT mice froze significantly less than eGFP mice during the tone test (Main effect of Group: F(1,22) = 11.32, p < 0.01). (F) ArchT mice froze significantly less than eGFP mice during the context test (t(22) = −7.17, p < 0.001).All data are expressed as mean ± SEM; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 3

Delayed CA1 inhibition does not impair TFC memory (A) Representative image of post hoc validation of AAV expression and optical fiber placement (white dotted lines) targeting CA1. (B) Experimental design to silence CA1 during the ITI. On the first day mice underwent TFC while laser stimulation (561 nm) was delivered to CA1 continuously for 40 s starting 140 s after the footshock on each training trial. (C) On the first day mice underwent TFC while laser stimulation (561 nm) was delivered to CA1 continuously for 40 s after a 140 s delay following termination of the footshock. The next day mice received a tone test in a novel context B. The following day contextual fear memory was tested in the original training context. (D) ArchT mice and eGFP performed similarly during training (Main effect of Group: F(1,22) = 0.446, p > 0.05). (E) ArchT mice and eGFP did not differ in their freezing to the tone CS during the tone test (Main effect of Group: F(1,22) = 0.022, p > 0.05). (F) Both groups showed similar freezing responses to the training context During the contextual memory test (t(22) = −0.694, p > 0.05). All data are expressed as mean ± SEM; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 4

Post-shock CA1 inhibition late in learning does not impair TFC memory (A) Experimental design to silence CA1 after the footshock on the second training day. Both groups received TFC training on the first day without any laser stimulation. On the second training day, mice underwent TFC while laser stimulation (561 nm) was delivered to CA1 continuously for 40 s immediately after the footshock on each training trial. The next day mice received a tone test in a novel context B. The following day contextual fear memory was tested in the original training context. (B) ArchT mice and eGFP performed similarly during the first day of training (Main effect of Group: F(1,22) = 0.147, p > 0.05). (C) ArchT mice and eGFP performed similarly during the second training day (Main effect of Group: F(1,22) = 0.690, p > 0.05). (D) During the tone test, ArchT mice and eGFP did not differ in their freezing to the tone (Main effect of Group: F(1,22) = 1.88, p > 0.05). (E) During the context test, both groups showed similar freezing responses to the training context (t(22) = 0.944, p > 0.05). All data are expressed as mean ± SEM; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 5

SWRs increase in frequency after the US (A) Schematic of the electrophysiological setup used to record local field potentials (LFPs) during TFC training. (B) Representative LFP recording. Top: raw data example of LFPs. Bottom: 130-200Hz filtered data. Octothorpe (#) indicates detected SWR. (C) SWR incidence is significantly increased after the shock (t(2) = −19.06, p < 0.01). (D) Freezing levels during the ITI predict SWR incidence (t(31) = 5.87, p < 0.001). All data are expressed as mean ± SEM; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 6

Effects of cholinergic agonism on SWRs, GCaMP activity, and TFC memory (A) Mice (n = 2) were first anesthetized prior to a 5 min baseline LFP recording (baseline) in their home cage prior to receiving an injection of pilocarpine (10 mg/kg, i.p.). After a 10 min delay, mice underwent another 5 min of LFP recording (pilocarpine). (B) Schematic of TFC design. GCaMP was infused into the dorsal hippocampus and an optical fiber was implanted above CA1 prior to mice undergoing TFC. Saline or pilocarpine (10 mg/kg, i.p.) was administered 15 min prior to TFC training (n = 5 per group) wherein bulk calcium activity was measured by fiber photometry. The next day, mice underwent a tone test in a novel context. (C) Trial-averaged GCaMP fluorescence in saline and pilocarpine mice. Inset: GCaMP fluorescence was significantly lower after the footshock in pilocarpine mice compared to saline mice (post – pre shock: t(12) = −2.787, p = 0.016). (D) Mice that received pilocarpine injections prior to training froze significantly during the trace interval than mice that received saline prior to training (Group × Phase interaction: F(2,24) = 3.579, p < 0.05; Trace t(12) = −2.771, p < 0.05).All data are expressed as mean ± SEM; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.