Hippocampal projections to the anterior olfactory nucleus differentially convey spatiotemporal information during episodic odour memory

Abstract

The hippocampus is essential for representing spatiotemporal context and establishing its association with the sensory details of daily life to form episodic memories. The olfactory cortex in particular shares exclusive anatomical connections with the hippocampus as a result of their common evolutionary history. Here we selectively inhibit hippocampal projections to the anterior olfactory nucleus (AON) during behavioural tests of contextually cued odour recall. We find that spatial odour memory and temporal odour memory are independently impaired following inhibition of distinct, topographically organized hippocampal-AON pathways. Our results not only reveal a longstanding unknown function for the AON but offer new mechanistic insights regarding the representation of odours in episodic memory.

Fig. 1

Inhibition of AON-projecting hippocampal neurons impairs context-dependent odour memory recall. a Coronal section depicting site of AAV-ChR2-YFP and -mCherry injections in the hippocampus (top) and the resulting innervation pattern at the AON (bottom). Coordinates mark anteroposterior position from bregma. Scale bars represent 1 mm; ac, anterior commissure; pc, piriform cortex; l, lateral AON; m, medial AON; vp, ventroposterior AON. b Schematic diagram depicting experimental approach. CAV2-Cre was infused into the AON, whereas Cre-responsive AAV-hM4D-mCherry was injected in the hippocampus. c Representative sections depicting AON-projecting HPC neurons expressing hM4D-mCherry. Magnification is indicated on the bottom right of each panel. Scale bars represent 1 mm (black) or 10 μm (white). d The olfactory spatial memory test paradigm. CNO-treated hM4D mice investigated the familiar and novel odour location to a largely equal extent, indicative of impaired spatial odour memory (Independent-samples t-test, t₍₁₆₎ = 3.194, **P < 0.01). e The olfactory temporal order memory test paradigm. CNO-treated hM4D mice were impaired in memory for the temporal occurrence of encountered odours (Independent-samples t-test, t₍₁₆₎ = 2.795, *P < 0.05). f Both groups showed normal performance in a context-independent novel odour recognition test (Independent-samples t-test, t₍₁₆₎ = 0.05644, NS, P = 0.9557). For all behavioural tests, CNO was injected 15 min before the retrieval phase. The odour in the novel spatiotemporal configuration is numbered in red. Positive discrimination ratios indicate preference for the novel odour-context configuration. mCherry control group: n = 8, hM4D group: n = 10. Data are presented as mean ± SEM

Fig. 2

Spatial and temporal information is transmitted along separate HPC-AON circuit elements. a Diagram depicting location of ArchT infusions within the HPC and optic fibre implantations above the AON in both experimental groups. iHPC terminals are predominantly found in the lAON (left) whereas vHPC fibres preferentially innervate the mAON (right). b Inhibition of both iHPC-lAON and vHPC-mAON pathways impaired spatial odour memory (GFP: n = 10; iHPC-lAON: n = 12; vHPC-mAON: n = 12; one-way ANOVA, F_(2,31) = 11.65, ***P < 0.0005). c Inhibition of the vHPC-mAON, but not iHPC-lAON pathway impaired memory for the temporal order of a sequence of odours (GFP: n = 12; iHPC-lAON: n = 12; vHPC-mAON: n = 16; one-way ANOVA, F_(2,37) = 9.235, ***P < 0.005). d HPC-AON pathway is not necessary for context-independent novel odour recognition (all groups: n = 10; one-way ANOVA, F_(2,27) = 1.824, NS, P = 0.1807). Scale bars represent 1 mm. Coordinates indicate anteroposterior position from bregma. Data are presented as mean ± SEM

Fig. 3

HPC-AON pathways contribute distinct spatiotemporal information during episodic odour recall. a Top: illustration of the episodic memory test paradigm. Bottom: heat maps depicting group average position (left) and pie charts representing the percent investigation time (right) during the retrieval phase of the episodic memory test (all groups: n = 12). b Comparison of percent investigation time for all groups during the retrieval phase of the episodic memory test (two-way ANOVA, treatment-by-odour interaction F_(6,132) = 9.532, ****P < 0.0001, main effect of odour F_(3,132) = 5.415, **P < 0.005). c iHPC-lAON pathway inhibition impaired the spatial, but not temporal component of odour memory, while vHPC-mAON pathway inhibition impaired both components to a similar extent (two-way ANOVA treatment-by-memory type interaction F_(2,66) = 9.320, ***P < 0.0005; main effect of treatment F_(2,66) = 6.715, **P < 0.005; main effect of memory type F_(1,66) = 9.620, **P < 0.005). FL/TD, familiar location/temporally distant; FL/TR, familiar location/temporally recent; NL/TD, novel location/temporally distant; NL/TR, novel location/temporally recent. Data are presented as mean ± SEM

Fig. 4

The HPC-AON circuit is necessary for context-driven odour recall. a Top: contextually cued odour recall test paradigm. Bottom: control groups in both experiments showed an increased investigation time between Day 9 and 10. Yet, following hM4D- (left) or ArchT-mediated (right) inhibition of hippocampal projections to the AON, mice fail to exhibit this behaviour (hM4D experiment- hM4D-mCherry: n = 10, mCherry control: n = 8; two-way ANOVA main effect of treatment F_(1,32) = 14.85, ***P < 0.0005; main effect of Day F_(1,32) = 34.5, P < 0.0001; interaction between treatment and Day F_(1,32) = 15.19, ***P < 0.0005; ArchT experiment—GFP: n = 14, iHPC-lAON: n = 12, vHPC-mAON: n = 12; two-way ANOVA main effect of treatment F_(2,70) = 7.538, **P < 0.005; main effect of Day F_(1,70) = 31.45, ****P < 0.0001; interaction between treatment and Day F_(2,70) = 7.052, **P < 0.005). b Mice were trained on a context-odour association and tested in a distinct context B in the absence of an applied odour. All animals showed no increase in investigation of the cotton swab despite all aspects of training being equivalent to mice tested in the training context (hM4D experiment—hM4D-mCherry: n = 10, mCherry control: n = 8; two-way ANOVA main effect of treatment F_(1,32) = 0.5035, NS, P = 0.4831; main effect of Day F_(1,32) = 0.0581, NS, P = 0.8111; interaction between treatment and Day F_(1,32) = 0.01156, NS, P = 0.9150; ArchT experiment—GFP: n = 12, iHPC-lAON: n = 12, vHPC-mAON: n = 12; two-way ANOVA main effect of treatment F_(2,66) = 0.4907, NS, P = 0.6144; main effect of Day F_(1,66) = 0.007502, NS, P = 0.9312; interaction between treatment and Day F_(2,66) = 0.2189, NS, P = 0.8040). Data are presented as mean ± SEM

Fig. 5

Conceptual diagram for the formation of episodic odour memory. A model of episodic odour memory whereby information regarding odour quality and spatiotemporal context merge at the level of the AON producing cellular populations that represent previously encountered odours (what) within the context in which they occurred (when and where)