Real-world memories are formed in a particular context and are often not acquired or recalled in isolation1,2,3,4,5. Time is a key variable in the organization of memories, as events that are experienced close in time are more likely to be meaningfully associated, whereas those that are experienced with a longer interval are not1,2,3,4. How the brain segregates events that are temporally distinct is unclear. Here we show that a delayed (12–24 h) increase in the expression of C-C chemokine receptor type 5 (CCR5)—an immune receptor that is well known as a co-receptor for HIV infection6,7—after the formation of a contextual memory determines the duration of the temporal window for associating or linking that memory with subsequent memories. This delayed expression of CCR5 in mouse dorsal CA1 neurons results in a decrease in neuronal excitability, which in turn negatively regulates neuronal memory allocation, thus reducing the overlap between dorsal CA1 memory ensembles. Lowering this overlap affects the ability of one memory to trigger the recall of the other, and therefore closes the temporal window for memory linking. Our findings also show that an age-related increase in the neuronal expression of CCR5 and its ligand CCL5 leads to impairments in memory linking in aged mice, which could be reversed with a Ccr5 knockout and a drug approved by the US Food and Drug Administration (FDA) that inhibits this receptor, a result with clinical implications. Altogether, the findings reported here provide insights into the molecular and cellular mechanisms that shape the temporal window for memory linking.
This is a preview of subscription content
Subscribe to Nature+
Get immediate online access to the entire Nature family of 50+ journals
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Get time limited or full article access on ReadCube.
All prices are NET prices.
The original videos and datasets generated and/or analysed during the current study are available from the corresponding authors.
The code for concatenation analysis of miniscope videos is available on GitHub.
https://github.com/Almeida-FilhoDG/ConcatMiniscope. All other code is available from the corresponding authors.
Cai, D. J. et al. A shared neural ensemble links distinct contextual memories encoded close in time. Nature 534, 115–118 (2016).
Rashid, A. J. et al. Competition between engrams influences fear memory formation and recall. Science 353, 383–387 (2016).
Abdou, K. et al. Synapse-specific representation of the identity of overlapping memory engrams. Science 360, 1227–1231 (2018).
Yetton, B. D., Cai, D. J., Spoormaker, V. I., Silva, A. J. & Mednick, S. C. Human memories can be linked by temporal proximity. Front. Hum. Neurosci. 13, 315 (2019).
Mack, M. L., Love, B. C. & Preston, A. R. Building concepts one episode at a time: The hippocampus and concept formation. Neurosci. Lett. 680, 31–38 (2018).
Zhou, L. & Saksena, N. K. HIV associated neurocognitive disorders. Infect. Dis. Rep. 5, e8 (2013).
Ellis, R., Langford, D. & Masliah, E. HIV and antiretroviral therapy in the brain: neuronal injury and repair. Nat. Rev. Neurosci. 8, 33–44 (2007).
Jung, W. & Lee, S. H. Memory deficit in patients with schizophrenia and posttraumatic stress disorder: relational vs item-specific memory. Neuropsychiatr. Dis. Treat. 12, 1157–1166 (2016).
Avery, S. N. et al. Impaired relational memory in the early stage of psychosis. Schizophr. Res. 212, 113–120 (2019).
Czajkowski, R. et al. Encoding and storage of spatial information in the retrosplenial cortex. Proc. Natl Acad. Sci. USA 111, 8661–8666 (2014).
Han, J. H. et al. Selective erasure of a fear memory. Science 323, 1492–1496 (2009).
Sano, Y. et al. CREB regulates memory allocation in the insular cortex. Curr. Biol. 24, 2833–2837 (2014).
Zhou, Y. et al. CREB regulates excitability and the allocation of memory to subsets of neurons in the amygdala. Nat. Neurosci. 12, 1438–1443 (2009).
Lanfranco, M. F., Mocchetti, I., Burns, M. P. & Villapol, S. Glial- and neuronal-specific expression of CCL5 mRNA in the rat brain. Front. Neuroanat. 11, 137 (2017).
Zhou, M. et al. CCR5 is a suppressor for cortical plasticity and hippocampal learning and memory. eLife 5, e20985 (2016).
Torres-Muñoz, J. E., Van Waveren, C., Keegan, M. G., Bookman, R. J. & Petito, C. K. Gene expression profiles in microdissected neurons from human hippocampal subregions. Brain Res. Mol. Brain Res. 127, 105–114 (2004).
Shepherd, A. J., Loo, L. & Mohapatra, D. P. Chemokine co-receptor CCR5/CXCR4-dependent modulation of Kv2.1 channel confers acute neuroprotection to HIV-1 glycoprotein gp120 exposure. PLoS ONE 8, e76698 (2013).
Yiu, A. P. et al. Neurons are recruited to a memory trace based on relative neuronal excitability immediately before training. Neuron 83, 722–735 (2014).
Lee, D. et al. Temporally precise labeling and control of neuromodulatory circuits in the mammalian brain. Nat. Methods 14, 495–503 (2017).
Airan, R. D., Thompson, K. R., Fenno, L. E., Bernstein, H. & Deisseroth, K. Temporally precise in vivo control of intracellular signalling. Nature 458, 1025–1029 (2009).
Shideman, C. R., Hu, S., Peterson, P. K. & Thayer, S. A. CCL5 evokes calcium signals in microglia through a kinase-, phosphoinositide-, and nucleotide-dependent mechanism. J. Neurosci. Res. 83, 1471–1484 (2006).
Marozsan, A. J. et al. Mechanisms involved in stimulation of human immunodeficiency virus type 1 replication by aminooxypentane RANTES. J. Virol. 75, 8624–8638 (2001).
Wang, S. W. et al. CCL5 and CCR5 interaction promotes cell motility in human osteosarcoma. PLoS ONE 7, e35101 (2012).
Shen, W. et al. Activation of the chemotactic peptide receptor FPRL1 in monocytes phosphorylates the chemokine receptor CCR5 and attenuates cell responses to selected chemokines. Biochem. Biophys. Res. Commun. 272, 276–283 (2000).
Moyer, J. R. Jr, Thompson, L. T. & Disterhoft, J. F. Trace eyeblink conditioning increases CA1 excitability in a transient and learning-specific manner. J. Neurosci. 16, 5536–5546 (1996).
Oh, M. M., Oliveira, F. A. & Disterhoft, J. F. Learning and aging related changes in intrinsic neuronal excitability. Front. Aging Neurosci. 2, 2 (2010).
Yokose, J. et al. Overlapping memory trace indispensable for linking, but not recalling, individual memories. Science 355, 398–403 (2017).
Rogerson, T. et al. Molecular and cellular mechanisms for trapping and activating emotional memories. PLoS ONE 11, e0161655 (2016).
Mayford, M. & Reijmers, L. Exploring memory representations with activity-based genetics. Cold Spring Harb. Perspect. Biol. 8, a021832 (2015).
Tanaka, K. Z. et al. The hippocampal engram maps experience but not place. Science 361, 392–397 (2018).
Mello, C. V. et al. Fat-storing multilocular cells expressing CCR5 increase in the thymus with advancing age: potential role for CCR5 ligands on the differentiation and migration of preadipocytes. Int. J. Med. Sci. 7, 1–14 (2009).
Yung, R., Mo, R., Grolleau-Julius, A. & Hoeltzel, M. The effect of aging and caloric restriction on murine CD8+ T cell chemokine receptor gene expression. Immun. Ageing 4, 8 (2007).
Wilkin, T. J. & Gulick, R. M. CCR5 antagonism in HIV infection: current concepts and future opportunities. Annu. Rev. Med. 63, 81–93 (2012).
Aharoni, D., Khakh, B. S., Silva, A. J. & Golshani, P. All the light that we can see: a new era in miniaturized microscopy. Nat. Methods 16, 11–13 (2019).
Pnevmatikakis, E. A. & Giovannucci, A. NoRMCorre: an online algorithm for piecewise rigid motion correction of calcium imaging data. J. Neurosci. Methods 291, 83–94 (2017).
Zhou, P. et al. Efficient and accurate extraction of in vivo calcium signals from microendoscopic video data. eLife 7, e28728 (2018).
Almeida-Filho, D. ConcatMiniscope: This is the first release of the ConcatMiniscope Pipeline (v.1.0.0-Beta). Zenodo https://doi.org/10.5281/zenodo.5676164 (2021).
Friedrich, J., Zhou, P. & Paninski, L. Fast online deconvolution of calcium imaging data. PLoS Comput. Biol. 13, e1005423 (2017).
Wei, Z. et al. A comparison of neuronal population dynamics measured with calcium imaging and electrophysiology. PLoS Comput. Biol. 16, e1008198 (2020).
We thank A. Macalino, E. Chen, E. Ramirez, C. Riviere-Cazaux, M. López-Aranda and E. Lu for advice and technical support; M. Sehgal and L. M. De Biase for providing transgenic mice; and A. Luchetti for analysis discussion. This work was supported by grants from the NIMH (R01 MH113071), NIA (R01 AG013622) and NINDS (RO1 NS106969), and from the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation to A.J.S.
The authors declare no competing interests.
Peer review information
Nature thanks Kaoru Inokuchi and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
a, Schematics of hippocampal tissue collection. b-d, qPCR experiment to measure Ccl3 (b), Ccl4 (c), and Ccl11 (d) expression in naïve mice (HC) and in mice at different times after contextual fear conditioning. HC=home cage. HC n = 6, 3 h n = 2, 6 h n = 8, 12 h n = 7, 24 h n = 8 mice. e, Representative images of Ccr5, Slc17a7 (excitatory neuronal marker), and Gad2 (inhibitory neuronal marker) mRNA expression in dCA1 from naïve mice or mice 3–24 h after fear conditioning. Red arrows: cells expressing Ccr5 and Slc17a7. Orange arrows: cells expressing Ccr5 and Gad2. Scale bar, 50 μm. f, Number of Ccr5-expressing excitatory and inhibitory neurons 3–24 h after fear conditioning (HC n = 4, 3 h n = 4, 6 h n = 4, 12 h n = 4, 24 h n = 3 mice; **P < 0.01, ***P < 0.001, two-way repeated measures ANOVA). g, Representative images of Ccl5, Itgam, and Rbfox3 mRNA expression in dCA1 from naïve mice or mice 3–24 h after fear conditioning. Red arrows: cells expressing Ccl5 and Itgam. Orange arrows: cells expressing Ccl5 and Rbfox3. Scale bar, 50 μm. h, Number of Ccl5-expressing microglia and neurons in naïve mice (n = 5 mice; *P < 0.05, paired t-test). i, Number of Ccl5-expressing microglia and neurons in HC mice and 3–24 h after fear conditioning (n = 5 mice per group; ****P < 0.0001, two-way repeated measures ANOVA). All results shown as mean ± s.e.m.
Extended Data Fig. 2 The colocalization of Ccr5 expression and memory ensembles measured with cFos-tTA mice and the optogenetic (ChR2ETTC) pre-activation system.
a, Representative images of Ccr5 and mCherry (neuronal ensemble) mRNA expression in dCA1 from cFos-tTA mice 6h after fear conditioning. Colocalization was labelled with dashed circles. Scale bar, 20 μm. b, Quantification of Ccr5 expression in total cells (DAPI) and neuronal ensemble (mCherry+). (n = 5 mice; *P < 0.05, paired t-test). c, Schematics to use blue light to activate ChR2ETTC-expressing neurons to be involved in neuronal ensemble by pre-activation. INTRSECT system (Cre-off/Flp-on) was used to label non-ChR2ETTC-expressing neurons as the control. d, Representative images of mCherry (pre-activated neurons), c-Fos (neuronal ensembles), and EYFP (non-preactivated neurons) in dCA1 24h after the novel context exposure. Scale bar, 50 μm. e, c-Fos distribution in mCherry+, EYFP+ or non-infected cells. f, Quantification of the colocalization between c-Fos and mCherry or EYFP. Colocalization (of c-Fos and mCherry) = (c-Fos+mCherry+/DAPI)/[(c-Fos+/DAPI)*(mCherry+/DAPI)] (n = 4 mice per group; **P < 0.01, paired Student’s t-test). g, Schematics to detect the colocalization of Ccr5 expression in neuronal ensembles using pre-activation system. h, Representative images of Ccr5 and mCherry (pre-activated neuronal ensemble) mRNA expression in dCA1 6h after fear conditioning. Colocalization was labelled with dashed circles. Scale bar, 20 μm. i, Quantification of Ccr5 expression in total cells (DAPI) and pre-activated neuronal ensemble (mCherry+) (n = 4 mice per group; *P < 0.05, **P < 0.01, two-way repeated measures ANOVA). All results shown as mean ± s.e.m.
a, Schematics of CCR5-iTango2 constructs. b, c, Expression validation of the CCR5-iTango system in HEK-293 cells. DRD2-iTango2 (for Dopamine 2 receptor) was used as a positive control. b, Representative images of tTA immunostaining. Scale bar, 50 μm. c, Quantification of tTA expression (intensity normalized to DAPI). n = 3 slides per group; *P < 0.05, one-way ANOVA. d, HEK-293 cells were transfected with 3 plasmids (see methods) for 24h and then treated with 10 nM CCL5 and blue light to induce EGFP expression. e, Representative images of EGFP expression after different treatments. Light−CCL5− n = 558, Light+CCL5− n = 507, Light−CCL5+ n = 521, Light+CCL5+ n = 500 cells. Scale bar, 50 μm. f, Quantification of EGFP and tdTomato ratio (intensity). Compared to control, light or CCL5 group, only the group with both light and CCL5 showed EGFP expression. Light−CCL5− n = 70, Light+CCL5− n = 97, Light−CCL5+ n = 97, Light+CCL5+ n = 282 cells; ****P < 0.0001, one-way ANOVA. g, Power-dependent EGFP expression. Results were normalized to no light control (30 mW/mm2 n = 320, 90 mW/mm2 n = 307 cells; ****P < 0.0001, student's t-test). h, Duty cycle dependent EGFP expression. The light stimulation was delivered every minute (~0.017 Hz) to induce EGFP expression. Light was kept on for 10–60 s during each stimulation to induce EGFP expression (10 s/min n = 407, 20 s/min n = 377, 30 s/min n = 307, 40 s/min n=383, 50 s/min n = 353, 60 s/min n = 524 cells). i, Dose curve of CCL5 to induced CCR5 activation (measured by EGFP/tdTomato fluorescence ratio) in cultured HEK-293 cells (10−12 M n = 49, 10−11 M n = 39, 10−10 M n = 29, 10−9 M n = 77, 10−8 M n = 86 cells). j, Time course of EGFP expression. The green fluorescence increased monotonically during the different time intervals investigated. Compared to other time intervals (2, 4, 6, 8 and 24 h), the 48h time interval showed the highest EGFP/tdTomato ratio (Light+CCL5+ 0 h n = 58, 2 h n = 194, 4 h n = 282, 6 h n = 310, 8 h n = 316, 24 h n = 396, 48 h n = 345 cells; Light−CCL5− 2 h n = 195, 4 h n = 219, 6 h n = 290, 8 h n = 304, 24 h n = 445, 48 h n = 401 cells; ****P < 0.0001, two-way ANOVA). k, Schematics of CCR5-iTango2 AAVs injected into mouse hippocampus and validated through intra-hippocampal infusion of CCL5 and fiber-optic light stimulation. l, Representative images of CCR5-iTango2-expressing hippocampal dentate gyrus neurons in control condition (no light and CCL5), light only, and light with CCL5. Ligand and light were directly delivered into the hippocampus. Scale bar, 250 μm. m, Left: To test CCR5-iTango2 activation in dCA1 (Fig. 1h), CCL5 was infused into the lateral ventricle (LV) while light was delivered into dCA1 of hippocampus (HPC). Right: Schematics of CCR5-iTango2 AAVs. All results shown as mean ± s.e.m.
a–c, Validation of the leakage in CCR5-iTango2 system in vivo. a, Schematics to test the CCR5-iTango2 system without light activation. b, Representative images of EGFP and CCR5-iTango2-expressing dCA1 neurons after fear conditioning. Scale bar, 50 μm. c, Quantification of EGFP expression (intensity normalized to tdTomato, n = 3 mice per group). d–f, Validation of the maraviroc mediated CCR5 inhibition in vivo. d, Maraviroc was co-infused with CCL5 into mouse dCA1. The CCR5-iTango2 was used to measure CCR5 activation in vivo. e, Representative images of CCR5-iTango2-expressing dCA1 neurons after fear conditioning. Scale bar, 50 μm. f, Quantification of EGFP expression in different treatment (CCL5−Maraviroc− n = 3, CCL5+Maraviroc− n = 4, CCL5+Maraviroc+ n = 3 mice; *P < 0.05, one-way ANOVA). g–i, Analyses of colocalization of c-Fos and CCR5 activation in vivo. g, Schematics to test c-Fos expression in EGFP+ cells after learning with the CCR5-iTango2 system. h, Representative images of colocalization between EGFP and c-Fos in dCA1. Red arrows: c-Fos+EGFP+ cells. Scale bar, 50 μm. i, Percentage of c-Fos+EGFP+ cells in total cells (6 h n = 6, 12 h n = 4, 24 h n = 5 mice; *P < 0.05, two-way repeated measures ANOVA). All results shown as mean ± s.e.m.
a, HEK-293 cells were transfected with Opto-CCR5 and jRGECO1a (Calcium sensor with red florescence) for 24h and then stimulated with blue light to induce a calcium response. b, Representative images at 0 min or 2 min after stimulation, or in the medium with high calcium concentration. Scale bar, 20 μm. c, Quantification of florescence change after light stimulation. In HEK-293 cells, Opto-CCR5-EGFP activation by light significantly increased intracellular Ca2+ concentration reflected by jRGECO1a (Control 2 min n = 95, Control 5 min n = 96, Opto-CCR5 2 min n = 86, Opto-CCR5 5 min n = 89 cells; **P < 0.01, two-way ANOVA). d, e, Opto-CCR5 activation increased pErk1/2 in HEK-293 cells. d, HEK-293 cells were transfected with the Opto-CCR5 construct. After 24h expression, the cells were starved in HEPES buffer for 1h before a 2 min light stimulation to reduce basal pErk1/2 levels. Cells were collected at 0 (no light stimulation), 15, 30 or 60 min after light stimulation and subjected to western blot analysis to investigate the phosphorylation level of Erk1/2 (e). f, Expression of Opto-CCR5 in dCA1 neurons. To express Opto-CCR5 in dCA1 neurons, AAV1-hSyn-Cre was co-injected with Lenti-DIO-Opto-CCR5. NeuN (neuron marker), GFAP (astrocyte marker) and P2Y12 (microglia marker) were co-stained with EGFP in dCA1. Scale bar, 20 μm. All results shown as mean ± s.e.m.
Extended Data Fig. 6 CCR5/CCL5 signalling regulates memory linking in an appetitive place preference task.
a–f, Place preference-based behaviour model to test the linking of contextual memories. a, Schematics of place preference-based linking behaviour. b, Representative trajectory plot (in the 3rd minute) of mice in the pre-exposed context and a novel context with a 5h and 7d interval. c, d, Mice showed a significant preference for pre-exposed context during the 3rd minute in the 5h group compared to the 7d group (5 h n = 13, 7d n = 12 mice; *P < 0.05, one sample paired t-test compared to 50%). e, Representative trajectory plot (in the 3rd minute) of mice in the pre-exposed context and a novel context with Vehicle or CCL5 infusion. f, CCL5 infusion in dCA1 impaired contextual memory linking with a 5h interval (Veh n = 7, CCL5 n = 8 mice; **P < 0.01, one sample paired t-test compared to 50%). g, Ccl5 knockout extended the temporal window of contextual memory linking (WT n = 11, Ccl5−/− n = 16 mice; *P < 0.05, **P < 0.01, two-way repeated measures ANOVA). All results shown as mean ± s.e.m.
a, b, Ccr5 knockdown enhanced memory allocation. a, Schematics of AAV8-shRNA-CCR5-Ef1α-EGFP injection, and representative images of c-Fos and EGFP staining. Two EGFP+c-Fos+ cells were labelled by dotted line circle and two EGFP+c-Fos− cells were labelled by asterisk. Scale bar, 20 μm. b, dCA1 neurons with Ccr5 knockdown had a higher probability of expressing c-Fos after a memory test in context A. Left: The percentage of c-Fos+EGFP+ cells in total (DAPI). Right: percentage of c-Fos+ cells in EGFP− cells (Con) or in EGFP+ cells with Ccr5 knockdown (shCCR5) (n = 8 mice, **P < 0.01, paired t-test). c–h, Expression of c-Fos and Opto-CCR5 or EGFP control in dCA1 (Opto-CCR5 0 mW n = 13, 2 mW n = 3, 4 mW n = 5, 8 mW n = 3 mice; EGFP control n = 4 mice per group). d, Representative images of colocalization between c-Fos and EGFP control after light stimulation and novel context exposure. Scale bar, 50 μm. e, Colocalization between c-Fos+ cells and EGFP+ cells after normalization to chance level. f, Quantification of c-Fos distribution in EGFP+ and EGFP− cells in the Opto-CCR5-EGFP or EGFP control group (*P < 0.05, **P < 0.01, one-way ANOVA). g, Percentage of c-Fos positive cells (normalized to cells with DAPI staining) in dCA1 with light stimulation of different power levels. h, Percentage of EGFP expression cells (normalized to cells with DAPI staining) in dCA1with light stimulation of different power levels. All results shown as mean ± s.e.m.
Extended Data Fig. 8 Analysis of the cumulative distribution of inter-event intervals recorded with miniscopes in wild-type and Ccr5 knockout mice.
a, Schematics used to extract spike information from raw traces of calcium imaging. Plot shows a 3s chunk of data from a single neuron using GCaMP6f calcium imaging and loose-seal cell-attached electrophysiological (Ephys) recordings. b, The average inter-event interval (IEI, from miniscope recordings) is highly correlated with the average inter-spike interval (ISI, by Ephys) (n = 36 cells; R2 = 0.92, P < 0.0001, ρ = 0.96, Pearson's correlation coefficient). c, Cumulative distribution of IEI of the top 10% most active neurons (in Ctx A). The top 10% most active neurons from WT mice showed a significantly different distribution of IEI 5h compared to 2d after the context A exposure. In contrast, this subset of cells showed a similar pattern for both time intervals in Ccr5−/− mice (WT n = 5, Ccr5−/− n = 6 mice; ****P < 0.0001, Kolmogorov–Smirnov test). d, Cumulative distribution of IEIs of the top 10% most active neurons and the remaining 90% neurons (in Ctx A) at 5h or 2d after the context A exposure (WT n = 5, Ccr5−/− n = 6 mice). e, Although neurons may have similar number of spikes during a certain time period of recording, the difference of their coefficient of variation unveils different firing patterns ranging from regular firing (Cell 1) to bursty firing (Cell 2). f, The coefficient of variation of IEI (by calcium imaging) highly correlates with the coefficient of variation of ISI (by Ephys) (n = 36 cells; R2 = 0.38, P = 0.0001, ρ = 0.61, Pearson's correlation coefficient). g, Cumulative distribution of IEI (the first 5s, zoom-in from d) of the top 10% highly active neurons and the remaining 90% neurons (in Ctx A) at 5h or 2d after the context A exposure. The difference between the top 10% most active and the remaining 90% neurons in Ctx A was strongly reduced from 5h to 2d in WT mice but not in Ccr5−/− mice (WT n = 5, Ccr5−/− n = 6 mice). h, Coefficient of variation from the top 10% most active neurons (normalized to the remaining 90%). WT or Ccr5−/− mice were exposed to Ctx B 5h or 2d after Ctx A. WT mice showed a significant decrease in the coefficient of variation of IEI comparing the data for the 2d and 5h intervals, whereas Ccr5−/− mice had similar coefficient of variation of IEI in both intervals (WT n = 5, Ccr5−/− n = 6 mice; *P < 0.05, two-way repeated measures ANOVA). All results shown as mean ± s.e.m.
Extended Data Fig. 9 Analysis of neuronal activity and overlap probability in wild-type and Ccr5 knockout mice.
a, Schematics showing that cells in neuronal ensembles can be sorted into cells with high neuronal activity (red) and low activity (blue), based on their average activity during the exploration of Ctx A and Ctx B which were separated by either a 5h or 2d interval. b, c, Left: Probability of overlap (averaged across mice) between subsets of cells with different levels of activity (y axis) during exploration of Ctx A and Ctx B, in WT and Ccr5−/− mice across time in Ctx B (x axis). Colour bars refer to normalized probabilities (chance = 1). Cumulative values were used for x and y axis (e.g., for x axis, 200s means 0–200 s; for y axis, 30 refers to the neurons within the top 30% of high (b) or low (c) activity). Right: the distribution of SEM across mice for the figures on the left. Asterisks (in the probability of overlap figures) represent the maximum SEM from each plot (WT n = 5, Ccr5−/− n = 6 mice). b, Probability of overlap between high activity cells in Ctx A and high activity cells in Ctx B in WT and Ccr5−/− mice. Note that the top 10% high activity cells in Ctx A are very likely to remain within the top 10% high activity cells in Ctx B 5h later for both WT and Ccr5−/− mice. In contrast, this subset of cells was reactivated around chance levels 2d later in Ctx B in WT mice, but not in Ccr5−/− mice. In the Ccr5−/− mice this subset of cells was still very likely to remain within the top 10% high activity cells in Ctx B. c, Probability of overlap between low activity cells in Ctx A and high activity cells in Ctx B in WT and Ccr5−/− mice. In contrast to high activity cells in Ctx A, the low activity cells in Ctx A were less likely (compared to chance) to be within the high activity cells in Ctx B. d, The probability of overlap between different ensembles (Ctx A and Ctx B) was sorted by neuronal activity in Ctx A and Ctx B, with a 5h or 2d interval between the two contextual exposures. Cells were sorted in percentages from top to bottom mean neuronal activity in the first context (Ctx A, y axis) and from left to right in the second context (Ctx B, x axis). With a 5h interval between Ctx A and B, the likelihood that neurons with high activity in Ctx A remained with high activity in Ctx B was higher than chance for both WT and Ccr5 knockout mice. With a 2d interval, the likelihood that neurons with high activity in Ctx A remained high activity in Ctx B was at chance levels in WT mice. In contrast, Ccr5 knockout mice showed a pattern similar to that observed with the 5 h interval (WT n = 5, Ccr5−/− n = 6 mice). e, Cells were again sorted from high to low activity in Ctx A with a 10% sliding window and 2% steps. The probability of overlap between subsets of cells (10% ensemble size) from Ctx A and the ensemble cells with top 10% high activity in Ctx B was plotted. The probability values were z-scored with respect to a null distribution created by randomly subsampling 10% of cells from Ctx A 10,000 times (i.e., results are represented as standard deviation (SD) from the mean of the null distribution). The 2SD threshold is labelled with a dashed line (WT n = 5, Ccr5−/− n = 6 mice).
a, In young mice, CCR5 signalling increases at a time point more than 5h after learning, and neuronal excitability and memory ensemble overlap remain high at 5h after learning. As a result, memories for context A (neutral context) and context B (shocked context) are linked together, and mice show high freezing during the test in context A. b, In aged mice, CCR5 signalling is higher than young mice at baseline and there is a further increase before 5h after learning, which lead to a reduction of neuronal excitability and memory ensemble overlap at 5h after learning. As a result, memories for context A (neutral context) and context B (shocked context) are not linked, and mice show low freezing during the test in context A.
About this article
Cite this article
Shen, Y., Zhou, M., Cai, D. et al. CCR5 closes the temporal window for memory linking. Nature 606, 146–152 (2022). https://doi.org/10.1038/s41586-022-04783-1